DIRECTION OF DRAINAGE IN THE PLEISTOCENE. 461 



gravel being derived from, or belonging to, the large L-shaped kame before 

 mentioned, whose lower end lies about fifty yards west of the quarries. 



The Newcastle and Beaver Valley branch of the Pittsburgh, Fort Wayne 

 and Chicago railway runs along the bluff at an elevation here of 860 feet 

 above tide, and about forty feet below the surface of the base-level plain, 

 which is 180 feet above present river level. 



With the most southerly of the three quarries we have here nothing to do. 

 In the second, about fifty yards north, on the exposed rock surface, there are, 

 or were, numerous potholes, mostly shattered by blasting, but several nearly 

 perfect. The best example was 3 feet wide, 4 feet long, and 24 inches deep. 

 Its steep side is toward the south, the other side being much eroded — this 

 difference in the two sides being commonly observed in modern potholes. 

 All the other potholes were seen to present their steep sides at their southern 

 portion, not one exception being noted in the whole number examined. 



Observations on modern potholes show that they always have their steep 

 side up stream, the lower side being eroded by the water falling upon it from 

 the upper side. 



From these facts we conclude that the current that formed the potholes 

 seen in this quarry must have flowed toward the north along the base-level 

 plain. The question of the date of their formation will be referred to in 

 another place. 



Base-Level Remnants. — It is not alone on the evidence of the potholes, 

 however, that we conclude the forming stream flowed toward the north. We 

 referred above to the difference in absolute level of the base-level plain be- 

 tween the mouth of the Beaver and Lawrence Junction, the base-level plain 

 sloping slightly to the northward. If, however, any allowance be made for 

 the greater postglacial elevation to the northward, a correction must be made 

 to restore the floor-level of the old valley in preglacial times. To the evi- 

 dence shown by the base-level plain on the Beaver, we must add the frag- 

 ments of what appears to be the same plain found both down the Ohio from 

 the mouth of the Beaver, and up that river and its parent streams. The 

 evidence of the Monongahela and Allegheny rivers seems to indicate that 

 the "higher river terraces" of President Chamberlin, or old base-level plains 

 of river erosion, fall with the present streams, but less rapidly.'^ Between 

 Pittsburgh and the mouth of the Beaver several remnants are also found 

 that indicate a general slope with the present stream. Passing down the 

 Ohio as far as Steubenville, the remnants are frequently seen ; few of them 

 have, however, been examined. Opposite Steubenville, Professor Wright 

 gives the elevation of the remnant as 285 feet above low-water mark of the 

 Ohiojf which would be 940 feet above tide. The elevation of the terrace at 



* Bulletin U. S. Geological Survey, no. 58, 1890, pp. 24-32. 

 flbid., p. 81. 



