r)() II. I,. FAIi;('llII>l) — (il.ACJAL WATKKS IN FINGKR J.AKKS Jtl'XiJOX 



east and eastern faces of the terraces worn by the currents in their north- 

 ward flow to reach the Cedarvale escape. 



LA KE DA NA 



Tlie growth of the Gulf delta, Avhich had been northward in the direc- 

 tion of flow, was terminated by the cessation of the Gulf river. This 

 was caused b}^ the further recession of the ice front, allowing the hy])o- 

 Warren waters to enter the Otisco valley directly from the north. This 

 change had no effect U{)on the Cedarville channel, which remained 

 effective. When in the down-cutting of the Marcellus shales the under- 

 Ijdng limestone was reached, a check was given to rapid erosion, and the 

 great lake that had here its outlet became comparatively stationary for 

 a long time. It is believed that this was the lake which produced the 

 Geneva beach * (see page 44). 



The head of the Cedarvale channel is at the southeastern edge of the 

 village of Marcellus. It is about 50 feet above the Nine INIile creek, and 

 leads directly away from the valley with a wddth between one-fourth 

 and one-third of a mile (plate 6, figure 1). About 50 rods east of the 

 north-and-south road that lies on the channel head the rock sill \vas 

 excavated by a cataract to a depth of 30 or 40 feet. The deepened chan- 

 nel trends east for nearh^ one-half mile, then narrows to perhaps 700 

 feet and trends southeast for a mile, widening to 1,200 feet, then takes 

 an eastward course and spreads out into a branch of the Onondaga val- 

 ley, where the great delta was built. The channel bottom is now con- 

 cave and the erosional form destroyed by the wash from the walls of 

 decomposing shale. The lower part of the later channel, after it had 

 trenched its earlier delta, is shown in plate 6, figure 2. 



The rock strata at the head of the channel dip southward. On the 

 north side of the intake the limestone is exposed, but on the southern 

 and lower part of the rock sill the Marcellus shale is not wholly removed, 

 but cut into three shallow channels with two intervening ridges of shale. 

 Taking as datum the United States Geological Surve}' bench-mark on 

 the coping of the bridge in the village as 653 feet, the elevations at the 

 intake are as follows: 



Limestone sill at road vomers, north side 696 feet. 



Channel, 15 rods wide (not entirelj' open) 684 " 



Ridge of shale 690 " 



Channel, 3 rods wide (onh' a hollow) 684 " 



Ridge of shale 689 " 



Channel, 8 rods wide (free) 685 ' ' 



*See Bull. Geol. Soc. Am., vol. 8, pp. 281-284. 



