308 E. C. ANDREWS. 



have been impounded and have lost its stream character- 

 istics. Inasmuch however, as it preserved its stream 

 character over the summit of the range we must seek for 

 an explanation of its movement in the mechanics of the 

 general stream. The case is not that of a brooklet which 

 has filled a great basin and proceeds to trickle over at the 

 opposite end, but rather that of an immense stream which 

 meets an obstacle lying across the channel bed and surmounts 

 it easily. In this case it is evident that the basal layers 

 do not lie inert while the upper ones are sli eared over them 1 

 but that they are forced to flow over the summit with only 

 a lesser degree of strength than are the upper layers. Of 

 course the basal stream layers will not ascend the upstream 

 aspect of the obstacle so quickly as they might descend 

 the slope, nevertheless a great percentage of the stream 

 energy is translated into corrasion during this surmounting 

 of an obstacle (see note on Obstacles in Part I of this 

 Series). 



Again the general corrasive effect on the upstream aspect 

 of an obstacle may not be pronounced if that portion of the 

 obstacle present a broad face of even slope to the ascend- 

 ing stream, but if definite channels exist in the upstream 

 slope the local increase of velocity, with consequent local 

 concentration of energy, will occur along these channels. 



In the case under consideration, the ice sheet met an 

 opposing range which was trenched with stream channels 

 discharging northwards. As the ice sheet surmounted the 

 range its basal layers were forced to ascend the old stream- 

 developed channels. Towards their heads the channels 

 were somewhat constricted. The deepest portions of the 

 moving ice mass were vertically above the channels also, 



1 Such an inert stage may arise, however, during a waning glacial 

 stage, when the ice moves along a grade which has been adjusted to the 

 strength only of the flood stream. 



