Shore Lincx of Aiu-ienf Gladdl L<il'<')<. — Todd. 2iit> 
side moraines upon the more recent drift. Lake Agassiz,and tlie 
lake of the Bhie Earth region in Minnesota, and lake Dakota and 
James lake in Dakota, readily come to mind in this connection. But 
can similar lakes l)e recognized in the much eroded and fragment- 
ary deposits external to the great terminal moraines? Some, as 
one with whom I was talking a few years ago, when discussing 
Prof. Wrighfs hypothesis of lake Ohio, said, '-Glacial lakes are 
a delusion and a snare, " yet the same person has mapped such a 
lake in central Wisconsin. Others would refer most of theextra- 
morainic drift to this cause. 
One dirticulty, and one which some consider insuperable, is the 
absence of distinct barriers and shore-lines and old water levels. 
The beaches of lake Agassiz have been readil^^ traced, but where 
are there an}' such traceable about lake Ohio or lake Missouri, or 
anywhere upon what has been called the older drift? The even, 
flat topography impresses one with lacustrine character in trav- 
ersing the Blue Earth region, or that between Scotland and 
Micchell, S. D. ; but we can readily see that if a lake has been of 
transitory duration it would fail of producing a plain. 
Before dwelling on a few recent observations, which it is the 
main purpose of this paper to present, let us consider briefly a few 
reasons for the common obscurity of the shore-lines of old glacial 
lakes. 
1. The surface of such lakes would usually l)e very inconstant. 
The ice would have been a very uncertain barrier. The chance of 
depositing a beach or cutting a clifT would, therefore, have 1>een 
small. 
2. The accumulation of shore deposits would not only be 
slight, but l)eing made largely by floating ice would be quite un- 
equally distributed, especially in wide and shallow lakes. Preva- 
lent winds would drive the drift-laden ice to certain shores much 
more than to others. If the lakes contained islands, the more 
remote shores might receive no erratics. 
'is. The ditferencein the ease of erosion between glacfo-natanl 
drift clays and the formations bordering them, may produce 
marked changes in topograph}- aiul drainage. The regions of 
Dakota and Nebraska illustrate this well. Loose sands and easily 
erode:! clays border the western edge of the comjjact and often 
l)oulder clad drift clays. While the latter are little afl'ecte.l by 
rains and streamlets, the former are rapidly removed. ."Nloreovcr. 
