Drainage Features of the Upper Ohio Basin. 277 



shelves, or available cols which would afford opportunities for 

 the re-descending river to locate itself on other lines than its 

 old track with all its meanders. When these possibilities are 

 added to the preceding, it becomes exceedingly strange that, 

 below the mouth of the Clarion, no abandoned channel is 

 found which retains any old filling comparable in depth to the 

 present trench. We find numerous old channels containing 

 gravels ranging from 50 to a little over 100 feet that represent 

 such old courses on the higher plane. This demonstrates the 

 truthfulness of the principle here urged and shows its applica- 

 tion to this particular field. 



If any other proof of the tendency to take new courses in 

 such situations were necessary, we might find it in the Rock 

 River valley in Wisconsin, which was filled up south of the 

 kettle moraine to a depth of from 350 to 400 feet with glacial 

 wash. The stream twice within ten miles of the head of the 

 filling cuts across rock spurs on the side of its old channel, and 

 it makes a third cut within the next ten miles. On the Chip- 

 pewa river, a similar number of rock cuts within a similar dis- 

 tance show the ready applicability of the principle. In these 

 cases the old valley filling is retained in its integrity, and the 

 fact of such filling is demonstrated. 



So also in the middle Allegheny basin, along the line of 

 drainage which we interpret as leading northwesterly to the 

 Lake Erie basin, there are remnants of deeply filled abandoned 

 channels ; the present streams occupying, in part, newer rock 

 cuts. At the mouth of the French creek, the old trenches 

 reach within about 65 feet of the present rock floor of the 

 Allegheny adjacent. The ox-bows and recesses retain here 

 and there a full section of the old filling. These seem to us to 

 emphasize the significance of the absence of similar sections 

 reaching the bottom of the present valley below the mouth of 

 the Clarion. 



The strength of this negative evidence, in the lower Alle- 

 gheny and the upper Ohio valley, is such that we are unpre- 

 pared to accept the existence of thin sheets or patches of 

 gravel on the slopes of the valley as establishing even a pre- 

 sumption that there was a trenching to anything like the pres- 

 ent depth previous to the formation of the high terraces. 



The improbability of the river re-establishing so completely 

 its old course would hot be applicable to a hypothesis which 

 supposed the formation of the upper gravels on a high rock 

 floor and the trenching of the valley afterwards to any depth 

 below and a re-filling to a point somewhat less high than the 

 summit of the earlier terrace plane for then it would be kept 

 in its old course. Such an hypothesis would comport with 

 two interglacial epochs, and if these remnants on the slopes 



