246 THE GLACIAL LAKE AGASSIZ. 



were formed, but were obliterated or buried by the advancing waters. It 

 may be further supposed that this advance continued until after all the 

 adjacent moraines iu Minnesota and North Dakota were formed, and that 

 it reached its maximum some little time later; and this may perhaps find 

 some siipport in the crustal movements of the Atlantic border regions. In 

 this wav it is easy to understand how the uppermost or Herman beach 

 might have essentially the same strength in all parts of its length of about 

 250 miles from soutli to north, and might ride over the several moraines 

 with seeming impartiality and without notable variation in character. This 

 hvpothesis also relieves the interpretation of the necessity of supposing that 

 the retreat of the ice and the formation of the moraines was especially 

 rapid. It has the advantage of not l)eing burdened with any hypothesis 

 at all regarding the rapidity or slowness of the formation of the moraines, 

 nor with any of the hypotheses necessary to account for the extraordinary 

 rapidity of morainic formation and glacial retreat which are postulated iu 

 the foregoing pages. 



Mr. Upham urg-es against this hypothesis the formation of the rather 

 large sand deltas around the border of the lake, which he thinks were 

 deposited contemporaneously with the existence of the ice on the adjacent 

 land, for without the presence of the ice in some cases, he urges, the sti'eam 

 which produced the deltas could not have existed. It appears to the 

 present writer, however, that these deltas would have been formed in very 

 nearly the same way under either hypothesis. The chief difference between 

 the two hypotheses, so far as deltas are concerned, would lie in a possible 

 slight variation in the height of the delta surfaces. Under the hypothesis 

 just suggested, the greater part of the material of the deltas must have 

 been deposited when the lake stood somewhat lower than its maximum. 

 The delta sunmiits would not therefore originally have reached to the full 

 height of the uppermost or Herman beach. However, as the shore con- 

 tinued to advance, the streams would have continued to build up these 

 original deltas, and their later deposits, being contemporaneous with the 

 Hei-man beach, would have brought the deltas up to an accordant eleva- 

 tion, or at least would have tended to do so. An exception to this would 

 perhaps be found in the case of deltas whose rivers became extinct before 



