Drainage Features of the Upper Ohio Basin. 281 



Supposing such trenching to have taken place, the first 

 invasion of the ice by its wash filled these lower trenches and 

 covered the gradation-plane to the heights now indicated by 

 the gravel sections previously given. After this, the sequence 

 of events is the same as in the first hypothesis. 



This hypothesis reduces the rock cutting, as compared with 

 the first hypothesis, by the amount of this preglacial trench — 

 say by the amount of 100 feet in depth and a part of the 

 breadth of the valley. By so much as it reduces the inter- 

 glacial rock cutting, it magnifies the work of the first glacial 

 epoch because of the greater amount of filling. 



We have had under consideration, in the field investigations, 

 as one of our working hypotheses, a fourth interpretation. 

 This was suggested partly by the phenomena themselves, and 

 partly by the division of the old drift in southwestern Ohio 

 and the region west into two parts, as before indicated. 



Hypothesis IV. — This hypothesis presumes (1) a deposition 

 of the high-level gravels in essentially the same way as in the 

 first hypothesis, or possibly the third. (2) After this, there 

 was a retreat of the ice and a period of erosion carrying the 

 trench down to the lower rock shelves, which we have before 

 described. (3) During the chief loess-depositing epoch, this 

 was again filled up to heights as yet undetermined, perhaps to 

 the base of the high level gravels, or possibly a little above, 

 but not to the full height of these gravels, otherwise the 

 river would not presumably have resumed a second time ex- 

 actly its earlier course. (4) After the close of the loess- 

 depositing epoch, a second stage of erosion carried away almost 

 completely the valley deposits and cut the innermost trench 

 down to essentially the present level. (5) Then followed the 

 ice invasion which produced the adjacent moraines, and the 

 late gravels, and after that, (6) the postglacial erosion, as in 

 the preceding hypotheses. 



We do not feel at liberty to make those geologists who have 

 expressed opinions concerning the history of the region, 

 responsible for any of these hypotheses, for they may not be 

 formulated precisely in accord with their views.* We think 

 however, that with minor modifications they embrace essen- 

 tially all the types of hypotheses which the facts, at present 

 known, warrant. We feel the more at liberty to leave the 

 hypotheses to stand by themselves, without special connection 

 with personal acceptance, since, in the limited space left us, 

 we think it will be more profitable to discuss their common 

 import than their differences and relative merits, for they all 

 tell a common story respecting the main features of the history 



* The first is the view advanced by the senior writer, in which he follows Mr. 

 Chance in the main. Bui. 58, U. S. Geol. Survey. 



