24 NEW YORK STATE ' MUSEUM 



feet, correlates with the sand plains at or near 1300 feet. The latter 

 are well developed contiguous to the higher plains, and represent 

 the lower level of the Herkimer lake. 



With the opening of the Cedarville channel the Herkimer waters 

 fell about 100 feet. The amount of downcutting is not evident, but 

 the present channel is 140 feet lower than the Summit lake channel. 



The Cedarville channel was located theoretically in advance, sim- 

 ply from the map.^ Its excellent character even surpassed expecta- 

 tion. It is a handsome and typical river channel of mature form. 

 It is at least one-fourth of a mile wide; steep banks, uniform width 

 and grade, with the long stretch of level summit occupied by a 

 swamp of cedar which has given name to the village. The valley 

 below the col, through East Winfield, West Winfield and Unadilla 

 Forks, has the characters produced by a strong river with low 

 gradient. 



It is possible that the Ontarian ice lobe backed away along the 

 southern divide sufficiently to allow the Herkimer waters to find 

 escape by the Bouckville pass, at 1150 feet, but it is not probable. 

 A high ridge between the Unadilla and Sangerfield rivers extends 

 north nearly to the parallel of Clinton, ending in Crow hiH, six miles 

 southwest of Utica. The ice lobe in the Utica district probably 

 pressed against this salient to the close of Herkimer time.^ 



The Cedarville channel is believed to have been the final outlet 

 of the Herkimer lake, and to have persisted all the time required for 

 the ice lobe in the Hudson valley to recede sufficiently to open a 

 lower pass in Albany county, initiating the next stage. 



The extent of the Herkimer sand plains is suggested by plates 

 2 and 7, where they are shown ranging from 1440 down to 1280 feet. 

 They will be found in many districts, either where heavy streams 

 poured into the lake or where there was outwash from the glaciers. 

 The differential uplift must be taken into account, which may not be 

 uniform far to the north. They should be found in large develop- 

 ment through the Adirondacks, but will not occur in strength on 



iThe certain proof of the correctness of the glacial lake philosophy is 

 found in the fact that nearlj^ all the more important glacial stream chan- 

 nels in the State, described by the writer, have been located in advance 

 of any visit, by the studj^ of the topography in relation to the position 

 of the ice front. 



2 The writer has described certain west-leading channels at Crow hill 

 (see title 17, page 22), but they are regarded as having carried only the 

 small volume of water from the Sauquoit valley on the east over into the 

 Oriskany valley and Bouckville escape. 



