I44 Niagara. (April, 
Lawrence and changed it into animmense lake, so, during its 
retreat, it must have done the same. Probably it did so toa 
greater extent, not only because, in its retirement, it had 
left moraines and deposits of till, blocking up the deep chan- 
nels draining into the Ohio and the Mississippi, but because, 
during the greatest accumulation of ice, the land north- 
wards, and especially the area of the St. Lawrence, had 
been depressed, and an immense sheet of fresh water, 
dammed back by the lower part of the valley of the St. 
Lawrence, being still filled with ice, stretched south-west- 
ward and northward. On the northern shores of this great 
lake glaciers still came down from the Lawrentian hills, 
and gave birth to icebergs that floated southward, dropping 
boulders of granite, gneiss, and other crystalline rocks, on 
the bed of the lake, or stranding onits shores, and there de- 
positing their freight. During this time were also formed 
many stratified beds of sand and gravel that lie above the 
till, and to it belong most of the deposits of the “‘ terrace 
epoch ” of Dana which were formed, not after, but during 
the glacial period. 
Before leaving this branch of my subject, I must again 
advert, as I have done in previous papers, to the great im- 
portance of a proper appreciation of the effect of the stop- 
page of the drainage of the northern parts of the continents 
during the glacial period. It was not only a period of 
erosion and transportation of rocks, but of great fresh water 
deposits ; and I fully believe that the fresh-water and inland 
sea beds that Professor Ramsay proves to have been de- 
posited in old red sandstone and Permian times-were due to 
former glacial periods, that of the Permian epoch being 
greater than the last one, and resulting in such a lowering 
of the level of the ocean, that there was great destruction of 
marine life by the increased salinity of the sea. 
There are many proofs that the ice was thickest and 
highest during the glacial period in the bed of the Atlantic. 
That which advanced up the valley of the St. Lawrence 
came from the direction of Greenland, and the whole of the 
eastern coast of America, down as far south as New York, 
must have been blocked up by it. This is proved, not only 
by the many fresh-water beds and terraces due to the dam- 
ming back of the rivers, but by the direction taken by the 
continental ice. Thus, over the higher summits of New 
England, the scratches point to the south-east and not to 
the east, as they would have done if the ice had been free 
to move directly towards the ocean. I think that this 
shows that the bed of the ocean was then occupied by ice, 
