282 THE JOURNAL OF GEOLOGY. 
from above results in a progressive thickening of the glacier 
from above downward. Though the increase in the thickness 
and velocity may have begun in the upper regions, the end may 
still be melting back; when, however, this increase in thickness 
reaches the lower end the advance will begin, for there will be a 
greater quantity of ice and it will have a greater velocity ; it will 
therefore have to travel further before being melted away. After 
a while, the snow-fall becoming less, the glacier will gradually 
decrease in thickness, and consequently in velocity, until the rate 
of melting exceeds the rate of flow, when the retreat will begin. 
Comparatively small changes in snow-fall result in marked 
movements of the glacier’s end; for not only is the thickness of 
the ice at the end changed, but the velocity is also, and these 
= 
Geers 
two factors so interact that an increase of one produces an 
increase of the other, and the reverse. This can be made clearer 
by a diagram. 
In Fie. 1, letline 1 represent the surtace of the glacier when 
the névé fields are lowest; suppose now the snow-fall increases ; 
the névé fields get thicker, but the end is not affected, and it 
continues to melt back; the lines 2, 3, etc., represent the surface 
at successively later times. After a while the thickening and 
consequent increase in velocity of the lower end causes it to 
advance. 
Similarly the continued advance after the snow-fall has 
begun to diminish, and the later retreat, may be illustrated by 
Euiges 2: 
The question of glacial erosion is a very important one; the 
