14A» The American Geologist. Feb. 1891 
gigantic islands north of Hudson bay would prove the former presence 
of a dam holding in the waters of that bay, wore the whole country ele- 
vated" (p. 450). Although I have done no fir-Id work worth mentioning 
on the shore-lines of these ancient lakes, the question of the origin of 
the lakes has interested me greatly, and I have for some time regarded 
them as most probably the combined result of pre-glacial erosion, glacial 
excavation and postglacial warping, with perhaps some assistance from 
local glacial depression, as suggested by Chamberlin. The share that 
warping may have had in this result depends in part on the former 
altitude as well as on the altitude of the continent, and professor Spen- 
cer claims to have shown that the continental altitude at a certain time 
may be determined by means of the old shore-lines, which he interprets 
as having been made on the margin of a bay, and therefore at sea-level. 
Hence the interest attaching to his explanation of the outlet by the 
Mohawk valley. The comparison that he makes between Iroquois and 
Hudson bays is in one respect perfectly fair, but there are four considera- 
tions that appear to me to invalidate the comparison completely. The 
first is the close correspondence of the Iroquois shore-lines with the 
level of the Mohawk passage; this correspondence would be of accidental 
origin if Iroquois were a bay; but it would be an essential feature if 
the passage were an outlet. Second, if Iroquois were a bay, the shore- 
lines should be found at the same level, except for their subsequent 
warping, on both sides of the passage; and as far as observation goes, 
no Iroquois shore-lines are reported on the eastern or outlet side of 
the Mohawk valley. Third, the Mohawk valley shows signs of former 
occupation by a river of much larger volume than the present Mohawk. 
Fourth, it is not only in Iroquois that the correspondence in the levels 
of old shore-lines and outlets is found; the same correspondence is 
known to have occurred in the case of the expanded shores of Erie, 
when it rose to the level of the Maumee-Wabash passage (Spencer's 
Algonquin); and still better in the case of lake Agassiz, when its waters 
rose so as to cross the pass between Big Stone and Traverse lakes, by 
which its basin was separated from the Minnesota-Mississippi valley. 
Large post-glacial lakes with outlets over low passes appear therefore 
to have characterized a recent epoch in our history. Arms of the sea 
were very likely expanded at about the same time, but they did not 
reach Iroquois or Agassiz. 
The question then remains, by what barrier were these lakes held up 
at their other end; and the answer generally given is, by the retreating 
ice-sheet. It is eminently proper that we should be slow to accept so 
curious an explanation; it is certainly true that existing glacial lakes 
are small and evanescent; and it is very questionable what thickness of 
ice would be needed to hold in the waters of Iroquois and its fellows; 
but at present, we must regard these bodies of water as lakes, and no 
explanation for the lakes has yet been suggested that can displace the 
glacial hypothesis. W. M. Davis. 
Harvard College, Cambridge, Mass., Jan. 3, 1891. 
