PERSISTENT EASTWARD DRAINAGE 31 



and Bad river ; and, secondly, by traces of shallow channels crossing 

 the divide. 



The divide is strewn with pebbles and cobblestones generally at an 

 altitude of 550 to 600 feet above the present Cheyenne * in northern 

 Ziebach county, and 350 to 400 feet above it fifty miles farther south, 

 the difference being mainly due to the descent of that stream. " Cedar 

 mountain," near White river, bears on its summit, 300 feet above that 

 stream, near its junction with Porcupine creek, a heavy deposit of gravel 

 and boulders of various crystalline rocks from the Black hills. Similar 

 deposits of smaller size occur 150 miles due east of that point, capping 

 terraces 500 to 600 feet above the Missouri, and near it a few miles south 

 of White river in Lyman county. Pebbles of porphyry resembling vari- 

 eties in the Black hills occur on the general upland level near Bonesteel, 

 Gregory county. Similar crystalline erratics occur along the Niobrara 

 as far east as Long Pine, and doubtless farther. 



Some higher points in the Pine Ridge agency and along the divide 

 between the White and Niobrara rivers are probably not strewn with 

 these erratics. 



Referring to the Pliocene map, we see all of these streams making 

 their way eastward to the James River valley. f This feature is attested 

 by the evidences of the change described in the next stage, and we only 

 need to state here that the Bijou hills are clearly a continuation of the 

 range of buttes, remnants of the old Miocene plain, which form an old 

 divide between the White and Niobrara rivers. Moreover, this highland 

 continues to a point south of Plankinton. So also the Tertiary beds in 

 the Ree hills and Wessingtons, already alluded to, are evidently a con- 

 tinuation of the divide between the Bad and White rivers. So also 

 Fox ridge west of the Missouri finds its continuation in the high ridge 

 east of that stream from Gettysburg to Faulkton. 



It should not be thought that the lines of these streams as represented 

 on the Pliocene map should be found as buried channels, filled with 

 drift, as is frequently the case with pre-Glacial streams in other parts of 

 our country, for the baselevel for them all (except perhaps the Niobrara) 

 was at a considerably higher level than in the present corresponding 

 streams. 



BEHEADING OF STREAMS FARTHER SOUTH BY THE CHEYENNE 



At what time the South fork of the Cheyenne attained its present re- 

 lations we are unable to say to a certainty. Possibly it may not have 



♦Bulletin 2, South Dakota Geological Survey, p. 519, and Bulletin 1, South Dakota Geological Sur- 

 vey, p. 123. 



fThis conclusion was reached by the writer and published in 1884. Proc. Am. Assoc. Adv. Sci., 

 vol. xxxiii, p. 391. 



