64 NEW YORK STATE MUSEUM 



weathering with some modification by the ice rubbing. Through 

 considerable stretches the Warren beach lies on the top of the 

 ledges and is marked by the deposits of angular chert gravel, which 

 the farmers call "chawed rock." Such is the case from Indian 

 Falls eastward for 4 miles. In some stretches the shore is a lime- 

 stone cliff, some striking examples of which are seen south of South 

 Alabama (east and west of "Pond" survey station, northeast and 

 east of Batavia, and west of Morgan ville) . 



In 1896 the Warren shore was traced in a general way, chiefly to 

 prove its continuity and altitudes, from Crittenden around into the 

 Genesee valley. The topographic maps were then not available 

 and the plotting was done with reference to the roads. The bars 

 are so weak and discontinuous that mapping is difficult and unsatis- 

 factory even with the help of the topographic sheets. 



As the province of this treatise is only the portion of the Erie basin 

 lying in New York we will leave the Warren shore at its extreme 

 north point, the limestone ledge at Pond station, i mile south of 

 South Alabama, where its altitude is 887 feet. 



Discussion of Whittlesey and Warren levels 



The duality or multiplicity of the bars on the Warren shore, 

 giving occasion for different names and the conception of two 

 distinct lake levels, has been stated on page 45. Having now 

 before the reader the detailed description of the shore features we are 

 better prepared to discuss the causes and significance of their 

 complexity and of the lake history. 



The difference in complexity between the Whittlesey and 

 Warren beaches is very marked. In all the Erie basin the Whit- 

 tlesey beach is commonly a single ridge, which doubtless represents 

 the wave -work at the surface and margin of the lake, while on the 

 same slope the Warren shore, about 45 feet lower, has several 

 ridges ranging through 10 to 30 feet vertically. However, the 

 Warren shore is commonly more simple from Crittenden eastward, 

 and the interesting problem discussed in this chapter is the cause of 

 Warren complexity west of Crittenden as compared with the relative 

 simplicity eastward. 



The following table shows the vertical spacing of the bars, both 

 Whittlesey and Warren, at many points along the shore. The 

 figures are not precise, some being by aneroid, but all are carefully 

 checked. Doubtful figures have been omitted from the table, and 

 the errors are not greater than is the local variation in the individual 

 bars. The tendency to read the aneroid in multiples of 5 feet is 



