146 Niagara. (April, 
Post-GLAcIAL PHENOMENA. 
When the ice had retired so far back as to leave the 
channel to the St. Lawrence valley again open, the waters 
of Lake Ontario began to re-flow in that direction. From 
the whirlpool, northward, they did not run in their old 
channel, but took a more easterly course. This may have 
been because the lowest outlet through the moraines left 
by the ice was in that direction, as Dana has suggested; 
but I think it more likely that it was because the ice retired 
from the eastward first. If we look at a map of the 
southern side of Lake Ontario, we shall find that most of 
the rivers have been diverted in the lower part of their 
courses to the eastward, indicating that the cause was not 
one of accidental configuration of the ground, but some such 
general one as the early retirement of the icy barrier from 
the eastern part of the lake. 
Wherever the river first commenced to flow, there it was 
likely to cut down through the rocks, for it would soon 
make for itself a channel through the loose drift lying on the 
surface; and between the banks of that channel it would be 
confined, and there only operate on the hard rocks below, 
just as, in copper engraving, the acid only a¢ts on the plate 
in the lines cut through the soft wax covering it. To make 
clear the argument in the question we have to discuss, 
namely, how much of the gorge in which the river now 
runs has been excavated out of the solid rocks since the 
glacial period, I must, in the first place, direct attention to 
the sketch plan (Fig. 3) of the old and new gorges at the 
whirlpool, four miles below the falls. The sketch is founded 
principally on a small plan in Lyell’s ‘‘ Travels in North 
America,” and partly from my own observations and sketches 
on the spot. I regret that I cannot give an accurate plan, 
and I could not learn that any complete survey has ever 
been made. 
Standing at the summer-house, on the American side, at 
the point where the river takes a sudden bend to the east- 
ward, I looked across the whirlpool to the old gorge oppo- 
site, and the question at once presented itself to my mind— 
from this point there are two channels downwards, one 
excavated before, the other after, the glacial period; to 
which does the one upwards to the falls belong? This 
question does not appear to have occurred to the authors of 
the theory, that the whole of the gorge through which the 
river now runs, from the falls to Queenstown, is post- 
glacial. But why might not the old pre-glacial river have 
