248 Chamberlin and Lever ett — Studies of the 



'Lewis,* W right, f Lesley,^: Spencer,§ Randall, || Foshay,*[ and 

 Hice.** 



Probable old Drainage Basins. 



We follow Carll in the general view that the present drain- 

 age system of the upper Ohio basin has been formed by the 

 union of several pre-glacial systems that formerly flowed into 

 what is now the Lake Erie basin. These were blocked up by 

 the ice of the early part of the glacial period, which invaded 

 their lower courses and forced them to flow over low divides 

 and unite to form a common southwestward flowing system 

 nearly parallel to the border of the ice. In this there is a 

 strong analogy to the upper and middle Missouri system 

 believed t'o have been formed out of numerous northward, 

 northeastward and eastward flowing streams forced to unite 

 along the edge of the ancient glacier, as determined by the 

 studies of Todd, Salisbury and the senior writer. 



If this view be correct, and a part of the purpose of this 

 paper is to add data in support of it, each of the united sec- 

 tions must be considered separately in determining their pre- 

 glacial and glacial history. Each section may indirectly help 

 to interpret the others but it will be rather by the suggestive- 

 ness of analogy than by the direct force of specific evidence. 

 We regard the evidence of the union of separate pre-glacial 

 basins as clear and decisive in at least three cases, as highly 

 probable in another, and as rather weak in another, at the 

 present stage of investigation. These may be treated briefly 

 in succession beginning with the uppermost. 



The Upper Allegheny Basin. — Three specific lines of evi- 

 dence support each other in showing that the uppermost sec- 

 tion of the present Allegheny formerly discharged into the 

 Lake Erie basin. (1) There is a notable constriction of the 

 valley near Kinzna, Pennsylvania, from its usual breadth of a 



*Penn. Second Geo]. Survey. Rept. Z, 1884, pp. 141-202, 279. 



fThis Journal, July, 1883; Proc. A. A. A. S., 1883, pp. 202-207; The Gla- 

 cial Boundary, Proc. Western Reserve Hist. Soc, 1884, 86 pages; Bull. U. S. 

 Geol. Survey, No. 58, 1890, pp. 39-110: This Journal, Nov., 1892. Also discus- 

 sions of the drift and drainage features of this region embraced in the "Ice Age 

 in North America" (1889), and in abbreviated form in "Man and the Glacial 

 Period" (1892). 



\ Penn. Second Geol. Survey, Rept. Q, 1878, pp. xxiv-xliii; Rept. Ill, 1880, 

 pp. xin-xvii: also various footnotes in (III) and (IIII); Rept. Z, 1884, pp. v-xltx. 



§ Penn. Second Geol. Survey, Rept. QQQQ, 1881, pp. 387, 405. 



|| Penn. Second Geol. Survey, Rept. IIII, 18S3, pp. 1-30, 309. We are also 

 indebted to Mr. Randall for notes on a detailed study of the distribution of the 

 attenuated drift in the vicinity of Warren, Pennsylvania, iu Which the distribu- 

 tion of the bowlders together with their altitude has been determined with care. 



If This Journal, Nov. 1890. Bull. Geol. Soc. Amer., vol. ii, 1891, pp. 457-464. 



**Bull. Geol. Soc. Amer., vol. ii, 1891, pp. 457-464: Science, vol. xxii. Sept. 

 2 9, 1893, p. 170. "We are also indebted to Mr. Hice for favoring us with the 

 esults of careful observation on the distribution of the drift and the depth of 

 hannels in the vicinity of Beaver, Pennsylvania. 



