Pleistocene of the Winnipeg Basin. — Tyrrell. 27 
to the bottom was afterwards covered with a sheet of fine alluvium, 
just as has occurred in the deposition of the alluvial clay on the 
east shore of lake Winnipeg. 
While the face of the glacier was not far removed from the 
Manitoba escarpment, and when in places the highest beaches 
were in process of formation a sand delta was deposited in front 
of the mouth of Short creek in the depression now drained by 
Valley river. While the sand plain was being formed here the fine 
material brought down by the river would, under ordinar}' con- 
ditions, have been carried out into the lake and have fonned a 
fringe of alluvial clay which would now be found covering the 
lower country. This does not appear to have been the case how- 
ever, for but a few miles from the front of the sand plain of the 
delta, gravel ridges rest directly, or almost directly, on the till, 
and it is probable that the finer alluvium was spread out for a con- 
siderable distance north and south from the mouth of the old delta- 
forming river. 
As the glacier, in its intermittent recession, drew back from the 
escarpment, the lake followed it, dropping more or less suddenly 
from beach to beach, and probably being drained northwestward 
as well as southward. A northerly current in the lake or a con- 
traction in its width is clearly evidenced on the face of the Duck 
mountain b}^ the graduall}' diminishing size of the pebbles of 
which the beaches are composed as they are followed towards the 
Swan river valley from the Riding mountain ; the general small 
size of the pebbles is also remarkable as larger pebbles and 
boulders are found in the underhing till. The absence or scarcity- 
of boulders on these beaches has already been recorded by Mr. 
Upham and the writer, and has been adduced as evidence that 
lake Agassiz did not freeze over, Ijut even then it is difficult to 
conceive why masses of ice laden with boulders should not be 
thrown on the shore. Dr. Dawson has suggested that the water 
was at no time sufficientl}' deep to float icebergs, and it is pos- 
sible that this may be one reason for the difference between these 
old ridges and the present beaches of lake Winnipeg. The 
beaches are almost unbroken, and are rarely more than ten or 
fifteen feet in hight. These facts lead to the supposition that 
they had not been acted on by waves such as would be generated 
on a lake with an area of more than 100,000 square miles, for on 
the shores of an}' of the three adjoining lakes of coniparativelv small 
dimensions, higher beaches composed of larger pebbles are of com- 
