242 



SCIENCE 



[N. S. Vol. XXXVII. No. 946 



In the uplands south of the recessional moraines 

 foreign fragments are much more rare, and in 

 some parts of the uplands a careful search is 

 required to find even a small pebble of crystalline 

 rock, while bowlders are practically absent (page 

 16). 



To whatever extent the ice in the margin 

 of the snow field was produced by the cen- 

 trifugal, anticyclonic winds from the in- 

 terior of the ice cap, as suggested by Hobbs 

 from study of the existing continental gla- 

 ciers," it also favored lack of drift in the 

 periphery of the ice body. "With the wan- 

 ing and thinning of the ice cap the drift- 

 loaded lower ice was finally uncovered so 

 as to constitute the marginal belt, and was 

 then subjected to thiiist or push from the 

 thicker body on the north. At this stage 

 the heavy moraine deposits were made in 

 the valleys, producing the present drain- 

 age divide, and the lobations of the ice 

 front built the crescentic lateral-terminal 

 ridges north of the divide. At a later stage 

 of the waning, when the required factors 

 were properly combined and balanced, the 

 drumlins were constructed on the low- 

 lands, northward. 



Many facts are cited by Tarr showing the 

 impotency of the latest ice sheet, and he 

 finally admitted that the 



Wisconsin ice sheet failed to notably modify 

 the topography in the greater part of this area 

 (page 16). 



This would seem to terminate the debate 

 about glacial erosion in the Finger Lake 

 district. But it does not, as the responsi- 

 bility for the anomalous topography is 

 shifted back to the Prewisconsin glaciers. 



This carries with it the necessity of believing 

 in 1,500 feet of vertical erosion in the Seneca 

 Valley by the continued ice work of at least two 

 periods of glacial occupation, separated by an 

 interval of gorge cutting several times as long 

 as the postglacial interval (page 16). 



The statement is warranted, therefore, that 



•"Characteristics of Existing Glaciers," 1911. 



these valleys have been profoundly modified by 

 glacial erosion, both by deepening and broadening 

 (page 30). 



But here, as in the process of glacial stream 

 erosion, the bulk of the work was done by an 

 earlier ice advance (page 31). 



It is admitted that ice sheets may have 

 some individuality and that successive 

 sheets on the same territory may have some- 

 what different behavior and produce dif- 

 ferent effects, due to differences in the cli- 

 matic, topographic and drift factors. But 

 it does not seem reasonable that one ice 

 sheet could deepen Seneca Valley 1,500 

 feet, while its successor did practically no 

 eroding at all. If the Prewisconsin ice 

 sheet had such remarkable excavating 

 power it should have produced conspicuous 

 ei'osional effects elsewhere than in the val- 

 leys, and specially in the southern part of 

 the state, and should have piled heavy 

 "old drift" deposits beyond the reach of 

 the Wisconsin ice. 



The drift burden of the Laurentian ice 

 sheet is represented not merely by the mass 

 of the moraines and the volume of detritus 

 carried away by the glacial drainage, but 

 also by the enormous bulk of drift built 

 into the drumlins. Even if the drumlins 

 were partly constiiicted by the earlier ice 

 sheet they can not, because of their loca- 

 tion, represent any product of deep ero- 

 sion of the sections of Seneca and Cayuga 

 valleys in question. There are no heavy 

 moraine deposits south of the "Valley- Heads 

 moraine, for the terminal moraine is not 

 massive, and the ancient drift in Pennsyl- 

 vania and New Jersey is not excessive in 

 volume. The only other disposal of the 

 great volume of debris that should have 

 been produced by deepening of the valleys 

 1,500 feet must have been by outwash of 

 the glacial drainage. But when the valley- 

 train and outwash deposits attributed to 

 the latest ice are considered there is no 



