62 Jl. L. KAIIICIIIIJ) — (ILACIAI. WATKKS I \ KIN(;|.;i; LAKKS KK(iI<)\ 



visible from tlie raili-oad upon the east side of the Jiutternut valley, but 

 no study has been made of the Avater-levels. The walls are steep and 

 tiie side streams weak, and the shore phenomena is not ))ronounced. 

 There are no villages in the valley, and the lake may be named the But- 

 ternut lake. 



The Apulia outlet probably remained effective while the ice barrier 

 receded 10 miles, or to within 3 miles of Jamesville. Then a somewhat 

 lower pass westward was uncovered, which has been cut to 1,200+ feet 

 elevation, according to the Tull}^ sheet. This channel is about four miles 

 south-southwest of Jamesville and the divide is upon a north-and-soutli 

 road near the house of Mr Edward Russell. It is a rock channel, in 

 shale, perhaps 15 rods wide at the col, leading southwest, then north- 

 west, and debouching into the Onondaga valley at the Indian village. 

 It carried the waters of the second-phase Butternut lake over westward 

 into whatever waters then occupied the Onondaga valle}^ either the 

 Cardiff lake or most probably lake Warren (see figure 1, page 54). 



The local Butternut waters were possibly drained to a third level by 

 an eastward-leading channel two miles east of Jamesville and on the 

 south side of the Manlius road. Being by the " Green " station of the 

 New York State survey, it may be called Greens channel. Here is a 

 trench 20 rods wide in the black shale which carried the overflow of 

 Butternut waters east-northeast to the Limestone valle}'' at Manlius. 

 Tlie exact elevation of this waterway is not determined, but it is in the 

 neighborhood of 890 feet. The initial height of the overflow was proba- 

 bly 50 feet above the present channel bottom. The reasons for not con- 

 fidentl}^ regarding this .channel as primarily a member of the hypo- 

 Warren or hyper-Iroquois series are : (1) its probably too great initial 

 elevation, and (2) its position. The relative position and trend of the 

 several large h3'per-Iroquois channels (see figure 1) and the orienta- 

 tion of the drumlin masses indicate that the ice-front had a trend some- 

 what northeast and southwest, and the Greens channel might have l^een 

 open for eastward escape of the high local Butternut waters, while the 

 Reservoir channel was still ice-covered and the Warren waters barred 

 from the Butternut valley. However, it is j)ossible that this cut was 

 deepened by the eastward escape of Warren waters, and if so it was the 

 first spilhvay. The honor of being the channel of first eastward flow 

 certainly lies between this Greens channel and the Reservoir channel. 

 This can not be positively determined Avithout better knowledge of the 

 relative land deformation and the precise elevation of the Warren plane 

 in this locality and a close examination and comparison of the two 

 channels. 



^y way of review of the glacial lake history u]) to the stage now under 



