WORK OF GLACIAL WATERS 151 



of the terminal moraine as outwash. This view would restrict the true 

 valley trains to the fillings of valleys beyond the terminal moraine or 

 reach of the ice-sheet. In this latter view the valley-train drift would 

 occur in New York only along the south side of Long Island and in the 

 small area south of the Allegany River. 



LAKES: GLACIAL 



Occurrence. — The term "glacial" is used by the writer to include only 

 lakes which existed by virtue of a glacier ice-barrier. The lakes and 

 lakelets now existing and called "glacial" by some authors should be 

 discriminated mostly as morainal or drift-barrier lakes. 



The conditions necessary for a glacial lake are a valley or depression 

 sloping toward and blocked by the ice-front. These conditions were ful- 

 filled in New York on so large a scale, in area and time, that the State, 

 it is confidently believed, held the largest number of glacial lakes and the 

 most remarkable succession with varied outflow of any .district in the 

 world. The reason for this superiority is found in the peculiar topog- 

 raphy of the western part of the State. In the great Ontario-Erie Basin 

 we have a broad depression, with its lowest passes on the east and west, 

 and with a deeply trenched southern slope, where lie the parallel valleys 

 of the Finger lakes. 



The only glacial lakes of which clear evidnce is preserved are those 

 which lay against the receding front of the latest ice-sheet. But it should 

 be clearly understood that every ice-sheet which transgressed the State 

 blocked the waters both during its advance and its recession. 



We do not know what portions of the Valley-Heads moraine, which 

 now constitutes the divide and forms the south limits of the basin, were 

 left there by pre-Wisconsin ice-sheets, but we may be quite sure that the 

 lakes during the advance of even the last glacier were somewhat differ- 

 ent in dimensions and relations from those of the ice recession which 

 are the subject of our field study. We may also be sure thai the earliest 

 ice invasion found the series of parallel valleys with fairly mature and 

 graded forms and open clear through to their heads, and the larger ones 

 heading in Pennsylvania. Those earliest ice-impounded lakes must have 

 been longer and deeper in the valleys than the lakes of later episodes, 

 when the valleys had become more or less occupied hv glacial and lake 

 deposits. The lacustrine conditions of the episodes antedating the Lau- 

 rentian Ice-retreal are as yet a matter of interesting speculation. One 

 further difference may be noted between the ice-advance and the ice-re- 

 cession lakes. The primitive lakes of the ice-advance were the lowest in 

 altitude and the most northerly in location and with the lowest outlets. 



