Origin of Itslccrs. — Crosby. 3 
g^lacier, the type of piedmont j^laciers, is simply a lake of ice 
existing at a level where permanent ice could not form, due 
to the confluence on the lowlands of the powerful alpine gla- 
ciers of the St. Elias range, and deriving its movement, in 
part at least, from the thrust of these tributary ice-streams. On 
Greenland, which appears to have mountainous borders with 
an inner lowland, we find a true ice-cap, with an area estimat- 
ed by Peary at 600,000 square miles, and a maximum thickness 
of, probably, several thousand feet and jiossibly a mile or more; 
and it is well known that in tlie recent past this ice-cap, which 
has evidently passed its culmination or maxinuim stage, has 
covered the whole of Greenland and the islands which fringe 
its coast, extending, possibly, far into the adjoining seas. But 
observation, naturally, has been chiefly conhned to the mar- 
gins of the ice, and to the overflow portions of the great mcr 
dc glace descending as lol)es and valley glaciers to and toward 
the coast, and, as in the case of the Malaspina glacier, to lev- 
els at which permanent ice can not form under existing climat- 
ic conditions. 
In both Alaska and Greenland, the drainage of the ice is 
chiefly subglacial ; and at many points powerful streams of 
water, carrying heavy burdens of detritus, are seen to issue 
from beneath the margins or extremities of the ice lobes, while 
the superficial streams, due to ablation of the tipper surface of 
the ice, rarely if ever reach its margin, being swallowed by 
crevasses to form iiwiiliiis and become tributary to, and the 
main sources of, subglacial rivers. In the case of the Malaspina 
glacier, the principal rivers discharging from its front or sea- 
ward margin undoubtedly liave their sources high up in the 
valleys of the St. Elias range; and, as described by Russell,* 
are seen in several instances to pass betieath the upper margin 
of this great piedmont glacier, in well-formed, wide-mouthed 
tunni'ls, the subglacial course of such a stream as the I'ountain 
or Vahtse being merely an incident of its history. lUit the main 
point, of course, is that we have here more indubitably than 
anywhere else ice-tiumels of consi(leral)le length — 5 to 25 
miles at least — occu]iied by large and rapid streams, the outlets 
of which are being olxstrucled and raised by the deposition of 
coarse detritus swept out of the tunnels by the torrents or dis- 
* Jour. Geol., 1, 240. 
/ 
