Drainage Features of the Upper Ohio Basin. 275 



Hydetown on Oil Creek, near Coopertown on Sugar creek, 

 near Utica on French creek, where, however, the connection 

 is not very well displayed, near Raymilton on Sandy creek, 

 above Homewood on Beaver creek, near Gallilee on Little 

 Beaver creek, and near New Lisbon in another valley of the 

 Little Beaver. Besides these twelve localities, at which the 

 relationship of this low system of terraces to the moraines may 

 be seen demonstratively, there are subordinate trains of gravel 

 that make still more certain, if possible, the relationship. 



All of these headings of the terraces are much lower than 

 the tipper terraces at points nearest to them. In the upper 

 part of the basin, the higher terraces, even in the valleys, head 

 125 to 200 feet above the heads of the terraces of the lower 

 system. If we compare with these the overwash trains on 

 Tidioute and on Pit Hole creeks, which we interpret as being 

 genetically the same, the difference in altitude is 350-400 feet 

 (see section B, fig. 1). If we compare the head of the high- 

 level gravels of Clarendon with the head of the low-level ter- 

 race at Russellburg, which is nearest to it and on the same 

 ice-course, being only thirteen miles distant, the difference in 

 the altitude is 210 feet (see section A, tig. 1). From this it 

 appears to us perfectly certain that the high-level terraces had 

 no connection with the morainic overvmsh, hut that, after the 

 high-level gravel planes were formed, they were trenched to 

 levels below the horizon of the lower terraces {presumably to 

 the present rock bottom) ; otherwise, the lower terraces could 

 not have been formed. The ttvo systems are, therefore, sepa- 

 rated in time by the iyiterval required for this erosion. 



Thus far we think all careful students of the region must go 

 with us, whatever they may think respecting the time required 

 for the erosion, or regarding the kind of material removed. 



Intermediate Phenomena. — The pronounced characters and 

 the wide separation in time of the higher and lower systems 

 of terraces being thus definitely determinate, we may turn to 

 intermediate phenomena that may be less clear in their gene- 

 sis and import. Between the base of the undisputed high- 

 level gravels and the summit of the low-level systems, gravel 

 is found at numerous points on the sides of the Allegheny 

 trench. This gravel is commonly found on sloping points in 

 the inner bends of the river and in other localities where, in 

 cutting down its valley, the river would be likely to leave 

 remnants of gravels, if they were there before, or would per- 

 mit their lodgment in the process of sinking its bed, if not 

 there before. Herein lies the radical difficulty of their inter- 

 pretation. A winding stream, which is cutting down its bed 

 at a moderate rate, tends to extend its meanders as well as 

 deepen its floor, and so it cuts outwards as well as downwards 



