THE LOCAL ORIGIN OF GLACIAL DRIFT 43 r 



If the ice passing over 2 encounters a rough topography, so that 

 drift was introduced into the ice far above its base, a large 

 amount of matter from 2 would go forward in the more rapidly 

 moving upper part, and be found in the ice, and finally in the 

 drift, over 4, or beyond. 



There is at least one other factor involved. Much of the ice 

 which passes over 4 (Fig. 2) never passed over 1, but accumu- 

 lated on 2, 3, and 4. This does not preclude the passage of 

 some ice from I to 4, but simply means that much of the 

 ice at 4 has never had a chance to work on 1. The idea 

 involved may be gathered from Fig. 3. Let the lines a, b, and. 



Fig. 3. 



c, respectively, represent the successive profiles of a growing ice- 

 sheet. Much of the upper part of the ice, the profile, of which 

 is c, accumulated over areas 2 and 3 and did not come from i v 

 as already stated. This upper ice which never worked on 1, is 

 moving more rapidly than the ice at lower levels which came in 

 part from I. Furthermore the conditions of snowfall and move- 

 ment are such as to develop a high marginal and a low central 

 gradient. The excessive marginal snowfall, for such was prob- 

 ably the condition, will in some sense check movement from the 

 center, and make the marginal portion of the field the effective 

 dynamic center of movement. The origin of much of the ice 

 far from the center of the ice-sheet, where it had no chance to 

 work on rock formations near the center, tended still further to 

 make the material carried and deposited by the ice at any point 

 of local origin. The effect of this factor is to still further dif- 

 ferentiate the fractions of the preceding series. 



When the spread of the ice, the constant tendency to lodg- 

 ment of drift in transit, the differential movement, and the 

 marginal origin of much of the ice are all considered, it is 

 probably true that the fractions given on pp. 428, 429 give much, 



