156 
the climax of the iceage of more southern träets, he may have had Baffin Land 
in mind beside Greenland. That the glaciation of southern BafEin Land, as well 
as of the adjacent parts of Labrador, has lasted up to a comparatively recent time, 
is shown by Barton, Glac. Action, and especially by Tarr, Evid. of Glac, who 
adduces among other facts also the existence of lakes with more than one outlet 
to prove that the ice has only a short time ago withdrawn from these regions. 
But I can bardly find that this ean be used as a proof of a late beginning of 
glaciation, and it also would be difficult to find out why southern Baffin Land 
should not have had its iceage about the same time as northern Labrador. An 
elevation of 3,000 feet would, indeed, close Davis Strait, but between southernniost 
Greenland and the american regions in question still a wide bay of deep water 
would cut in from the Atlantic, and its evaporation must have been large enough 
to produce a considerable precipitation in the southern part, at least, of Baffin Land. 
That it lasted longer here than to the south, I have been inclined to believe after 
study of the flora, already before I found that Tarr had expressed the same opinion. 
The glaciation farther north, of small amount as it is, certainly has begun in 
a låter period, when the elevation of the arctic area had come to an end, and a 
general subsidence had set in in its stead. This reversion of the inovement may 
have been directly due to the enormous pressure exerted by the icesheets on the 
enrths crust, as some geologists think, or to other causes, alone or in connection 
with it, it is shown, however, to have begun before the maximum of the iceage 
and to have contiuued long enough to lower the coastlines considerably even below 
their present level. This subsidence allowed the formation of glaciers farther north, 
on bolh sides of Lancaster Sound, on the northeastern part of North Devon and 
along the eastern side of Ellesmereland. The glaciation of these regions never 
reached to form large icesheets or continuous inlandices, only isolated glaciers in 
the valleys and thin ieecaps over the summits of some mountains. In Ellesmereland 
at least this glaciation may be said to stånd about its maximum even now, there 
being no traces of any considerable expansion of the present glaciers in former 
times. As I have previously (p. 30) alluded to, the glaciers of present time are 
entirely restricted to träets where a strong current through a more or less narrow 
channei keeps the water open most part of the year, or perhaps all through the 
winter, and thus forms a local area of evaporation, feeding precipitation on adjacent 
high land. The North Devon and Ellesmereland glaciers generally lie on Archeean 
rocks, and this circumstance probably accounts for the often expressed opinion that 
causahty should exist between the underlying rock and the glaciation. For the 
region here in question this view seems first to have been introduced by Suther- 
land, Glac. Phsen., p. 302, who points out the entire absence of glaciers on the 
Silurian rocks west of Lancaster Sound. Now for that region we must seek the 
explanation in the small amount of precipitation, and if wo turn instead to the 
western part of Jones Sound we shall find all existing glaciers actually on rocks 
of the Silurian system, while the local glaciation at the Heureka Sound is to be 
