12 THE ROYAL SOCIETY OF CANADA 
and they carried this burden of rock debris and “‘shavings’’ along 
until they reached some low place where the ice was sufficiently 
buoyed up by water to allow the load of debris to sink and accumulate 
and thus to form till, or until other conditions might supervene which 
would cause till to be deposited in hollows beneath the glacier, or as 
sediment in low places in front of it where possibly its readvance 
might obliterate most of the evidence of sedimentation, or finally 
until it was completely carried away by streams flowing from the 
glacier. If the debris was dropped beneath the glacier it might 
possibly be stratified if the ice was actually floating on the water. 
In cases where the ice may have advanced over sediments de- 
posited in water it undoubtedly tended to spread these deposits over 
the inequalities of the surface if such existed, and so while we would 
find an undulating surface of till over an undulating rock surface, 
we might expect to find a fairly level surface of till over a level floor 
of rock. Thus the even surface of till underlying the Archudsonian 
Swamp is itself probably underlain by a fairly even rock floor. . 
The till under this great swamp is a highly calcareous clay composed 
of subangular grains of quartz and limestone mixed with a large quan- 
tity of decomposed argillaceous material. It contains many small frag- 
ments of marineshells, as well as many smallimperfectly rounded pebbles. 
It also contains a number of boulders, some of which are distinctly 
polished and scored with glacial grooves and scratches. The boulders 
are chiefly of granite and Paleozoic limestone, but some are of a very 
characteristic quartzite weathering with light coloured depressions, 
while others are of red conglomerate, white and red sandstone, dia- 
base, banded jaspilite, etc. It is unstratified, and often breaks into 
small angular fragments along short vertical joints. In some places it 
is distinctly divided into upper and lower portions by a horizontal plane 
or by stratified sands which may contain layers of mossand lignite giving 
evidence of very shallow water conditions, or the immediate proximity 
of a shore during the temporary retirement of the glacier which formed 
the till. Where now exposed this bed of sand with moss and lignite 
hundreds of feet below the highest old post-glacial shore line indicates 
that the land stood higher during the period of deglaciation than after- 
wards. It probably raised on the retreat of the ice, and then sank 
again with its last advance down at least to the level of the highest 
beaches, and possibly considerably below that level. 
On the Nelson river till is well exposed both in cliffs beside the 
stream, and on the hard tidal flats. 
Beginning at the shore of Hudson Bay and ascending the river 
which gradually narrows from a width of twelve miles at its mouth 
to a mile and a quarter at the head of tide water below Seal Island, 
