$4 LHE JOURNAL OF GHOLOGY: 
concerned in the production of the drift, it does not follow that 
they may not have played a subordinate part. 
Neither winds, nor streams, nor lakes and seas, acting alone, 
can account for the drift. Neither can their combined activities, 
attributing to each its maximum, go far toward explaining the 
remarkable series of phenomena bound up in the drift, or closely 
associated with it. Even if shore-ice, pan-ice, and icebergs be 
added, the whole combination falls far short of adequacy, though 
pan-ice and icebergs might account for some phenomena which 
water and wind cannot. It is evident therefore that some other 
agency must have been concerned in the production of the drift, 
though wind and water and shore-ice and icebergs may have 
played subordinate roles. 
GLACIERS. 
A possible agent concerned in the production of the drift is 
glacier ice. If glacier ice be responsible in whole or in part for 
the drift which has so great a development in North America 
and Europe, the study of the results effected by glacier ice to-day 
should give us confirmatory testimony. It is well known that 
many existing glaciers increase and diminish sensibly within 
periods of a few years. Even seasonal fluctuations in size are 
readily observed. These fluctuations of accessible glaciers facili- 
tate the study of their work, and the careful study of the results 
which existing glaciers effect, affords a means of testing the 
possibility of the glacier origin of the drift. 
After a glacier has retreated for a distance up the valley 
which it occupies, an area which was formerly covered by the 
ice is left bare. The area from which the ice has withdrawn is 
readily accessible, and the effects of former ice occupancy and 
activity may be seen. In such situations, the surface is found 
to be more or less generally covered with a mixture of bowlders, 
gravel, sand and clay. If many valleys occupied by glaciers be 
examined, these various constituents of the surface material may 
be found to co-exist in all proportions. They may be in approxi- 
mately equal quantities in one valley, or in one part of one 
