116 LABRADOR 
attaches to a regional glacier will surely and amply repay 
the explorer who heads his steamer for Frobisher Bay. 
The Grinnell Glacier lies only a long half-day’s journey 
by steamer from Cape Chidley; in a sense it is at the very 
door of civilization, yet it is far less known than the ice of 
northern Greenland or the distant glaciers of Alaska. 
Whether or not the north land bears any remnant of the 
ice which once overwhelmed Labrador, the recency of the 
glacial retreat from the peninsula is most strikingly evident. 
This is especially true on the northeast coast, where the gla- 
cialist, no less than the worker in bed-rock, is blessed with 
that negative virtue of the earth’s surface, the absence of a 
forest-cover. He who runs may read the glacial records 
from one end of the coastal belt to the other. 
To gain a vital idea of ice-work even on the Greenland 
scale or the Antarctic scale, one needs not the training of a 
professional glacialist. A first approach to the understand- 
ing of glaciers may be profitably made in the recognition 
of their analogy with rivers. Upstream, a river scours its 
channel, batters, grooves, and wears away the solid rock, 
so deepening its bed and in time excavating a valley of a 
size appropriate to the stream. In its lower course on 
flood-plain or delta, the river lays down the rock-fragments 
worn out of the rocky channel. Throughout the length 
of the river, increasingly, this débris, in the form of gravel, 
sand, or mud, is moving deltawards. A water-stream has 
thus three main functions — to scour, to carry the scoured 
rubbish down the valley, and then to deposit that same 
rubbish in lake or sea or other basin, where the stream’s 
velocity is finally checked. In like manner the gliding ice- 
stream, whether flanked by valley-walls or blanketing 
