OLD MIDDLE ALLEGHENY DRAINAGE SYSTEM. 133 



The force of this evidence has also been noted in showing either that a 

 disproportionately small gradation plain with high cliff borders lay in this 

 narrow gorge (having a breadth of only one-third to one-half that of the 

 Clarion and Lower Allegheny gradation plain) or that there has been a 

 reversal of drainage by which a small stream that formerly flowed north- 

 westward through this gorge to join the French Creek-Conneaut outlet was 

 reversed and its valley recut to fit the new an.d larger stream. The' balance 

 of probabilities in favor of the latter has also been set forth. 



There seems to be evidence that the portion of the present Allegheny 

 between the supposed col near the mouth of the Clarion and the French 

 Creek outlet has been derived from parts of two valleys, each discharging 

 northwestward, but separated by a divide below the mouth of East Sandy 

 Creek. This view is supported not only by a notable constriction there 

 (to a width of scarcely 1,000 feet) but also by an abandoned valley leading 

 northward from the bend of West Sandy Creek at Polk (Waterloo) to 

 French Creek Valley, just above the mouth of Sugar Creek, which would 

 afford a northward outlet for the western stream. The relationships of the 

 present streams to this abandoned valley and to the supposed col may be 

 seen by a comparison of figs. 5 and 6. This comparison will also serve to 

 show how natural is the restored system compared with the disturbed and 

 unnatural present system. 



Turning next to the line of evidence found in the rock shelves and 

 terraces, a general inspection of the French Creek Valley shows that there 

 has been broader and deeper excavation than on the Middle Allegheny. 

 But inasmuch as the French Creek Valley lies within the glacial boundary, 

 and its lower course nearly coincides with the direction of the ice flow, it 

 seemed necessary to determine whether its greater size may not be due in 

 the main to glacial excavation. An examination of the valley with this 

 question in mind led to the discovery of old channels and ox-bow curves 

 of preglacial streams whose preservation is so complete as to furnish decisive 

 evidence that glacial excavation has been of little consequence in causing 

 the size of the southern end of the valley. 



Along the valley which led i^ast Waterloo from the highland tract near 

 the mouth of the Clarion to the present French Creek Valley there are 

 remnants of an old valley floor near the supposed divide at an altitude of 

 375-400 feet above the river, or 1,275 to 1,300 above tide, while at Waterloo, 



