Wave-cut Terraces in Keuka Valley 
37 
THE ALTERATION OF SHORK LINES BY LATER ICE-INVASIONS. 
Partial or complete efFacement of the constructional and destruc- 
tional products of wave and current work in these pre-Wisconsin 
ice-dammed lakes would be expected. The sweep of an ice-inva- 
sion, followed by the destructional work of the slowly falling bodies 
of water marking the period of ice-recession; would necessarily 
modify, remove or cover such features as terraces in unconsoli- 
dated materials, as bars, spits, cusps, etc.; whereas the cliffs and 
terraces in rock would be much less altered. 
The potency of ice as a factor in erosion does not make an iden- 
tical appeal to all observers; this is when the sculpturing of bed- 
rock is under consideration. So it is possible that all will consent 
to the general, though not complete, removal by erosion of the 
constructional products of lake waves and currents. As a matter 
of field study, however, it may as well be granted that these con- 
structional forms have been entirely obliterated; the differentia- 
tion of a bar, or delta belonging to some pre-Wisconsin lake, from 
the water-laid portions of glacial drift would require an environ- 
ment unusually free of other deposits. But we must grant that 
cliffs and terraces formed in rock would be less affected by glacial 
erosion. 
The extent to which these cliffs might be modified by erosion 
would depend upon their topographic relations. Ice abrasion 
is more effective on the slopes opposed to ice motion; it is more 
effective also along the lower contours of the walls of the valleys 
trending with the direction of the moving ice. Hence in a series 
of terraces along a valley wall, the lowest one would be the most 
modified by glacier ice. 
The beach structures of these former lakes have suffered fur- 
ther from wave work of more recent water bodies, especially of 
the high-level lakes. The degree of effacement through this 
agency depends upon the coincidence of the surface-planes of the 
two bodies of water, or upon their approximation to coincidence; 
if these planes intersected at a very slight angle, the vertical range 
of beach agents would at least partially overlap for a considerable 
horizontal distance; if the planes were actually coincident, then 
