$2 NEW YORK STATE MUSEUM 



A suggestion can be offered to account for the absence of 

 channels on the salient. It has already been stated that it is 

 morainal territory. East of the salient and northeast of Batavia 

 is a heavy moraine, traversed by the New York Central Railroad, 

 with kettles and kettle lakes. To the north and northeast the 

 land declines gently. It seems altogether probable that when 

 the ice weakened at this locality it was not by recession of a bold 

 and stream-swept front but by ablation and thinning over a wide 

 belt; and that the draining of the second Lake Hall, or the waters 

 under that level, down to the Warren plane was through stagnant 

 and drift-buried ice, and consequently no channels are preserved. 



Warren outflow 



With Lake Warren admitted to the Ontario basin let us again 

 turn our attention eastward, to the locality of its extinction, the 

 critical district in the region of the Split Rock channels and the 

 Gulf and Cedarvale canyons. 



Our problem in this district is the production of the great canyons 

 on the south (west and east of Marcellus) and yet subsequent to 

 the higher (Split Rock district) channels on the north. 



By a glance at plate 4 the reader will understand that the glacial 

 waters could not have been held up to the hight of the highest 

 channels on Howlet Hill and Split Rock if the great passes at Mar- 

 cellus had then existed in the form which they exhibit today. 

 We are forced to the conclusion that the Marcellus gorges are of 

 later production than the Split Rock channels. The simplest 

 explanation is to attribute the two great gorges to the over- 

 flow of the vast Lake Warren, and subsequent to the Vanuxem 

 waters. 



The single but sharp difficulty which we have to meet under 

 this theory is the original hight of the Marcellus passes. When 

 the outflow of the Vanuxem waters cut the Split Rock scourways, 

 from 900 feet downward, the Marcellus passes must have been 

 over 900 feet. But the later Warren waters could not have used 

 them for outflow if they were above the Warren plane, about 

 880 or 890 feet. This difficulty of the altitude of the Marcellus 

 passes is quantitatively slight, but positive. 



The suggestion is now offered of a temporal*}' filling of ice and 

 drift in the Cedarvale-South Onondaga valley which held the 

 pass at a high level during the life of Lake Hall and the early 

 Lake Vanuxem. The ancient valley was sufficiently capacious 



