250 THE GLACIAL LAKE AGASSIZ. 



The present writer differs with Mr. Upham also in his conception of tlie 

 waviu which tlie enghxcial debris becomes at k^ngtli exposed and deposited. 

 Instead of rising toward the surface of the glacier, it is believed, on the 

 basis of observations in Greenland, to pursue a course nearly parallel with 

 the base, on the whole, and to come out at the extremitv of the glacier. 

 To some slight extent it may become superglacial by ablation, but <nily to 

 a limited degree. Consonant with this conception, it is believed that eskers 

 and similar formations of gravel are essentinlK' sul)glacial or marginal. 



These fundamental considerations add to the difficulty which the pres- 

 ent writer experiences in accepting the idea of the contemporaneity of the 

 Herman beach with tlie formation of so many important moraines and of 

 such massive sheets of till, together with so great a retreat of the ice margin. 

 This difficulty is enhanced by a comparison of the uppermost beach with 

 the lower ones, which are correlated with much less important glacial action, 

 and yet do not seem to be correspondingly inferior in magnitude. 



The difficulty is somewhat relieved by the jirobability that while the 

 lake was glacier-bound on the north the surface was covered throughout a 

 large part of the year with ice, and during the rest by bergs and berg- 

 lets, and that, as a consequence, wind action was reduced to a minimum. 

 This may possibly be the line of solution of the problem. Mr. Upham, 

 however, finds but limited evidence of berg deposits. These need not be 

 gi'eat if the engdacial debris was of small amount; but if it was large the 

 protection of the lake surface from wave action by bergs should find expres- 

 sion in very considerable berg deposits. 



A further difficulty is presented by the slightness of the cutting down 

 of the outlet during the period of the Herman beach. If, however, the 

 alternative hypothesis is adojited, the trench of 40 or 50 feet existing in 

 the drift plain previous to the formation of the Herman beach should per- 

 haps be regarded as the work of the outlet during the earlier j:)art of the 

 glacial retreat, instead of Ijeing referred to preglacial or interglacial, or at 

 least preretreatal, times. The work done by the outlet and by the ice would 

 be thus brought more nearly into harmony. 



In respect to the lowering of the outlet of Lake Agassiz by stages, 

 instead of steady progress, which has been previously discussed, it appears 



