Great Ice-Dams. — Taylor, - 31 
level would have resulted and this may have been the cause 
of the dual form of the Forest or Crittenden beach.* 
]^Ir. Warren Upham has contended strongly for the view 
that lake Warren covered the whole area of the four upper 
lakes. The course of the ice-front at the time of this lake as 
he represents it, however, passses from a point east of Buffalo, 
N. Y., presumably on the Lockport moraine, directly north- 
ward, east of Georgian bay and lake Nipisssing and far to the 
east of lake Superior. t But this hypothetical course is nov.' 
known to be widely at variance with that indicated by the mo- 
raines. At the time of the Albion dam the ice must still have 
covered the whole of Georgian bay, and nearly half of lake 
Huron, the north side of the straits as they were then and prob- 
ably more than half of lake Superior. 
Without offering any proof that they really are so extended, 
Dr. J. W. Spencer has claimed an extension of all these glacial 
lake beaches, and others supposed to have been found at 
higher levels, over the whole lake region and on to an indefi- 
nite distance over the lower St. Lawrence valley. He also 
denies that the abandoned outlet channels associated with the 
beaches are old river courses. On his view all the beaches are 
of marine origin. As applied to the features of the region here 
considered, however, such a hypothesis falls far short of ac- 
counting for the facts in a satisfactory manner. For among 
the facts gathered from this region there are so many things 
which are clearly characteristic products of glacial agencies 
and as clearly impossible products of marine agencies that a 
mere statement of the facts ought to be a sufficient and con- 
clusive answer. Some of the most important features to be ac- 
counted for demand for their explanation the supposition of a 
movable, water-tight barrier or dam retreating across the lake 
basin in a northerly direction. The relation of the beaches to 
the outlet channels plainly demands such a receding barrier. 
*If the straits were opened during the time of lake Warren this 
event probably occurred when the ice withdrew from the Hagenville 
moraine. Agreeably to this idea, Leverett found the upper Crittenden 
beach ending near Alden and Fairchild found only the lower member 
extending on to Lima. This suggests that lake Warren should be 
considered as having had two stages, falling from the upper to the 
lower when the ice-dam withdrew from the Hagenville-Alden moraine 
and opened the straits of Mackinac. 
fAm. Jour. Sci., vol. XLIX, Jan. 1895. Boundary No. 4 on Plate I. 
