HYDROGRAPHY OF THE EARLY PLEISTOCENE 35 



in another epoch would be widened very considerably. Meanwhile the 

 drainage from the edge of the ice farther south naturally found its out- 

 let through the Big Sioux, which at that time we believe to have flowed 

 southward into the Vermilion, and the Split Rock joined it in the north- 

 ern part of Lincoln county. 



The reasons for this supposition are found : (1) In the course of the 

 ice-sheet in the next epoch, with the evident changes of drainage attend- 

 ] ng it ; (2) in the fact that the level of the Big Sioux in its bend south of 

 Sioux Falls is as high as the general surface 8 or 10 miles farther south ; 

 (3) in that topographic maps show a distinct sag across the present divide 

 west of Worthing, South Dakota, leading to an ancient lake bed in Turner 

 county connected with the Vermilion river. 



During the early Pleistocene the streams were, no doubt, very active- 

 Much of the James River valley was excavated at that time, though we 

 may not suppose that it was as low as at present. The White river and 

 Cheyenne show broad and high terraces about 200 feet above their present 

 levels in the vicinity of the Missouri river. These may readily be 

 believed to correspond to the level of these streams during the later 

 Pliocene. 



Another, terrace well marked and covered heavily with western gravel 

 is found on White river 325 feet above the present level, and a 

 similar terrace is found on the Cheyenne, near its mouth, 370 feet 

 above the stream. These are believed to mark the level of the streams 

 before the advent of the Wisconsin ice-sheet. If the country has 

 been raised toward the west since the Pliocene we are unable to say 

 what the gradient of those streams was during that epoch. If the 

 altitude has not changed, the gradient from the late Pliocene terrace on 

 White river to the present level of the James at Rockport would be 

 about 4.7 feet a mile, allowing the usual crooks in the stream along the 

 course indicated on the map, and the gradient for the early Pleistocene 

 over the same course would be about 3.5 feet per mile, and from the 

 mouth of the Cheyenne 2.6 feet, and 2, to the same point for correspond- 

 ing times. If this is thought too great for the large streams which we 

 have shown flowing over those channels, we may find recourse in pre- 

 suming that the region west of this was not so much elevated at that 

 time. 



The streams radiating from the Black hills, like those about the 

 Rocky mountains generally, have several prominent terraces, marking 

 successive stages in the erosion of their valleys. 



Rapid creek may be taken as a type. It exhibits at least three well 

 marked terraces. The highest one noted in the hills is a broad one 450 

 feet above the present stream. This may correspond to the early Plio- 



VI— Bur.r,. Geot.. Soc. Am., Vor.. 13, 1901 



