ICE-SHEET EROSION IN NEW YORK 63 



Certainly rock cliffs, with or without angles, are not probable features 

 by glacial abrasion, though they might be produced along the edges of 

 stream glaciers aided by weathering. However, there were no stream 

 glaciers in the Finger Lakes valleys. These cliffs are either the work of 

 preglacial streams or of postglacial waves ; they are not ice work. 



Islands and capes of rock in Cayuga basin. — On the east side of 

 Cayuga valley, at Union Springs, the shore has several sharp rock pro- 

 jections and one island of limestone, Frontenac island, the latter a half 

 mile from shore. These features are shown in figure 6. The boat pilots 

 say that other submerged rocks lie offshore. Such features are emphat- 

 ically inconsistent with the idea of glacial enlargement of the valley at 

 this point, as an}' severe effect of the glacial plane would straighten the 

 valley walls. They are normal products of the work of atmosphere and 

 water. 



This point might properly be included under the next main heading 

 as direct proof of non-erosion by ice. 



Direction of the ice flow due to the topography. — The radial or 

 digital attitude of the Finger lakes has been regarded by erosionists as a 

 possible effect of the spreading flow of the ice sheet. If the valleys were 

 radiating on downward slope the conception would not be entirely un- 

 reasonable, but they are spreading on upward slope. Undoubtedly the 

 ice flow in the earlier and later stages may have been guided by the val- 

 leys. During the waning of the Ontarian ice body the spreading flow of 

 the ice was up the radiating valleys; but this direction of flow was an 

 effect, and not a cause, of the valley topography. The ice favored the 

 lowest ground, and the deeper ice and stronger flow was naturally along 

 the depression of which Seneca and Cayuga are the broad axis. 



During the most forceful stage of glaciation the direction of flow was 

 obliquely across the valleys. 



Transverse valleys. — These have already been mentioned (pages 55- 

 56). The transverse valleys are positive proof that the ice did not 

 carve the more elevated topography. South of the divide, where the ice 

 moved on downward slope, the transverse valleys and the irregular 

 topography are proofs that the ice did not obliterate the relief features, 

 even where it had greater fluency. 



In the belt of the Finger lakes the transverse valleys exist, although 

 somewhat obscured by drift filling. Along the divide and southward 

 the preglacial topography is conspicuous. It would seem unreasonable 

 to claim for a continental ice sheet the power of cutting trenches below 

 some plane while above that plane the relief features are scarcely affected ; 

 yet the claim of glacial origin of valley discordance in central New York 

 requires essentially that assumption. 



