Davis.] 
496 
[May 18, 
17. HEIGHT OF ESKERS. 
It is noteworthy that the height reached by the esker ridge is 
frequently less, but seldom greater than that of the delta surface. 
No reason appears for this relation if the ridge was formed 
in a superglacial channel, while it is a not unnatural result of 
conditions attending a subglacial origin. If the esker gravels 
accumulated m a channel on the ice surface, no reason can be 
assigned for their taking a definite relation of height after they 
had subsequently settled down to the ground. If the gravels 
accumulated in a subglacial channel, it is most natural for them 
to have built a ridge up to the level of the plateau, because their 
waters would rise to that height if the roof of the channel were 
high enough, particularly near the outlet ; it is not to be expected 
that the ridge should rise to a higher level except at a distance 
back from ice front, where the subglacial tunnel might rise and 
fall over rolling ground, and thus form a ridge of gravel above 
the plateau level. This may be clearer if it is remembered that 
a subglacial stream, flowing below the level of the standing water 
into which it discharges, is acting under a 14 negative gravity,” 
and is therefore attempting to cut away the roof as well as the 
floor of its tunnel. 
18. BOULDERS IN SAND PLATEAUS. 
Brief mention may be made of the large boulders that are 
found at various levels in the sand plateau. These sometimes 
still bear the glacial striations that we may suppose once were on 
nearly all the now water- worn stones of the esker and delta head. 
The boulders could not have been washed to their present posi- 
tion. They must have been floated by ice rafts or bergs ; not 
over the delta surface, which it must be remembered rose a little 
above water level, but from some free margin of the ice front, 
across the open part of the pond or arm of the sea in which the 
delta grew. The ice standing here or there, or dropping its load 
on the way across the water, the boulders would lie in an acci- 
dental relation to the sands. 
19. CHANGES IN ESKERS SINCE DEPOSITION 
It is important to emphasize the conclusion that when the delta 
had reached some indefinite size, the feeding stream was diverted 
