42 NEW YORK STATE MUSEUM 



limestone outcrops and to the Portage sandstone summits is approxi- 

 mately 1600 feet in a distance of some 12 miles along a north-south 

 line. The difference in elevation between the Lake Plains and the 

 Appalachian plateau uplands is well shown in the photograph of the 

 relief model (plate i). 



It must not be understood, however, that this escarpment pre- 

 sented in preglacial times a solid unbroken cliff. Drainage collecting 

 on the upper levels and flowing down over the edge of the steep 

 north-fronting outcrops in streams of considerable volume, cut 

 notches in the edge of the cliff, which in time were deepened and 

 widened into considerable valleys ; and such valleys then were also 

 extended back, or southward, by headwater erosion for significant 

 distances into the upland country. Accordingly the escarpment in 

 preglacial times probably presented a number of broad salients pro- 

 jecting northward separated one from the other by narrower valleys, 

 widest at the north ; and pinching out into the upland in the form of 

 the letter V laid flat with its open end toward the north and drawn 

 very tall and narrow. In a general way such has continued to be the 

 topography of the escarpment front in glacial and postglacial time 

 except that the valley feature has been much accentuated in depth, 

 width and length and most extraordinarily modified in extension 

 and form. The occasion for these changes in the valleys is to be 

 found in the action of the ice. 



In the movement of the glacial ice it would be manifestly improb- 

 able that the thin edge of its front, on encountering so high and steep 

 a barrier as that of the escarpment described above, could be shoved 

 up over such a slope and then across the broader uplands of the Appa- 

 lachian plateau itself. It is evident, accordingly, that the escarpment 

 slope acted as a dam, holding back the glacial flood until so much 

 ice had been brought forward from the north as to make the ice thick- 

 ness along the barrier as great as the height of the topographic 

 obstruction. Only then could the glacial flow southward be recom- 

 menced in the form of a submerging ice sheet. 



Meanwhile the water-cut and weather-widened valleys of the main 

 streams draining to the north and notching the escarpment slope, them- 

 selves of considerable dimensions both as to depth and width, must 

 have afforded channels into which the ice could thrust projections of 

 notable length before the general front of the advance had over- 

 topped the plateau level. As the ice became thicker and thicker such 

 projections were elongated more and more and when the ice finally 

 overflowed the upper levels the deepest and presumably most freely 

 moving parts of the mass necessarily continued to follow the path- 



