236 GLACIAL FORMATIONS OF ERIE AND OHIO BASINS. 



The C one wang-o Valley also carries only narrow strips of a filling which 

 reached to a height of about 250 feet above the stream. 



It is probable that much if not all of the Allegheny Vallej' from 

 Warren down to the mouth of the Clarion was filled to a height of 150 to 

 200 feet above the present river level. But now there are only a few 

 remnants to indicate the height to which it was filled. It may be considered 

 doubtful whether the Allegheny was greatly filled; but the fact that a 

 large amount of glacial material was carried by stream action from the 

 middle or glaciated portion of the Allegheny into the lower or unglaciated 

 portion at levels more than 200 feet above the present stream seems to 

 demand a great filling of the glaciated portion. It is hardly probable that 

 an ice sheet would fail to make heavy deposits in a valley which it par- 

 tially covered and with which it had numerous connections, when it was 

 able to contribute enough material to a larger valley (the Lower Allegheny) 

 leading away from the ice margin to produce a filling 80 to 100 feet in 

 depth for a distance of more than 100 miles. 



It seems well within bounds to estimate that the erosion of this old drift 

 on the principal tributaries of the Allegheny which carry it — Conewango, 

 Brokenstraw, Oil, and French creeks — as well as on the Allegheny itself, 

 reached a depth of 150 to 200 feet and a breadth nearly as great as that of 

 their valleys. 



It seems probable also that the attenuated character and patchy 

 distribution of the upland drift has been to a marked degree intensified by 

 erosion. It would be difficult, however, to make any estimate of the erosion 

 there, since the original amount of drift can scarcely be estimated or even 

 closely conjectured. 



The valleys just mentioned have afi"orded discharge for the waters of 

 later glacial stages than the one under consideration, and this should not 

 be overlooked in dealing with the amount of erosion, nor should tlie effect 

 of the glacial floods that occupied the valleys during the melting of the 

 earliest ice field be disregarded. While in general the glacial floods are 

 depositing rather than eroding agencies, not a few instances can be cited 

 where streams which headed in the ice sheet have cut conspicuous trenches 

 across gravel plains that had just been built as outwash aprons along the ice 

 border. But where such marked trenching has occurred the glacial streams 

 were apparently favored by a good gradient, such as is so often afforded by 



