EROSION SOUTH OF ST. LAWRENCE-MISSISSIPn WATERSHED 281 



In the Missouri Eiver^ we have direct evidence, in the boulders which 

 were carried to Tuscumbia, 60 miles up its southern tributary (the Osage 

 Kiver), that there were annual floods, in the latter part of each summer 

 during the closing stage of the lowan epoch, rising to a height of 200 

 feet. Floods to this extent are also made to seem credible from the amount 

 of ice which the sun would be capable of melting over the glaciated area 

 tributary to the Missouri River. Floods of similar dimensions must also 

 have poured through troughs of the Ohio and Alleghany rivers. So enor- 

 mous were these floods that it is difficult to set limits to the work accom- 

 plished by them. Wherever they or their tributaries were eroding chan- 

 nels they would effect results which can not be measured at all by the 

 work accomplished by present comparatively insignificant streams. 



On the other hand, in the Alleghany Valley there are certain features 

 of deposition, pointed out by Prof. E. H. Williams, which indicate an 

 entirely different interpretation from that ordinarily given of the high- 

 level gravels which border the Alleghany Eiver throughout its entire 

 middle and lower course. Mr. Williams has pointed out that these de- 

 posits uniformly occur where a tributary glacial stream came in with 

 sufficient force to throw coarse gravel to a high elevation on the other 

 side, as at Kenerdell, where tributary streams from the melting ice-front 

 near by came into the Alleghany River at right angles, through Scrub 

 Grass Creek, with power sufficient to push gravel up 300 feet upon the 

 opposite side. Lower down the stream, beyond the direct influence of 

 glacial tributaries, the high-level gravel terraces occur below bends in 

 the trough of the river where the direct action of the swollen current 

 would throw gravel on and over the rock shelves which furnished the 

 most direct outlet for the rushing torrent at that elevation. 



In detailing the phenomena at Kenerdell,^ I had shown the impossi- 

 bility of considering this gravel deposit as a remnant which had existed 

 during the entire period demanded for the rock erosion of the gorge, but 

 had supposed that it did involve the filling of the gorge with glacial 

 gravel and the subsequent erosion of the trough. But Mr. Williams' 

 explanation is more credible. The plunging current from the glacial 

 border, coming down through the trough of Scrub Grass Creek, kept the 

 trough of the Alleghany scoured out and threw the material upon the 

 opposite bank. Below this point the high-level gravels which I had noted, 

 at Gates Ferry, Emlenton, Bradys Bend, Orrsville, Kittanning, Taren- 

 tum, Springfield post-office, and Verona, besides numerous intervening 

 places, are accounted for by the shifting angles at which the glacial tor- 



2 See American Geologist, vol. 33, April, 1904, pp. 205-222. 



3 See American Journal of Science, vol. 47, March, 1804, p. 175. 



