1 877. [ The Great Ice Age and Origin of the “ Till ” 
221 
relation of thickness to superficial extension. If a thin 
sheet of ice can be bent to a given arc, a thick sheet may 
be bent in the same degree, but the thicker ice demands a 
greater radius and proportionate extension of circumference. 
But we have direCt evidence that ice of great thickness — • 
aCtual glaciers — may bend to a considerable curvature before 
breaking. This is seen very strikingly when the uncrevassed 
ice-sheet of a slightly inclined neve suddenly reaches a pre- 
cipice and is thrust over it. If Mr. Geikie were right, the 
projecting cornice thus formed should stand straight out, 
and then, when the transverse strain due to the weight of 
this rigid overhang exceeded the resistance of tenacity, it 
should break off short, exposing a face at right angles to the 
general surface of the supported body of ice. Had Mr. 
Geikie ever seen and carefully observed such an overhang or 
cornice of ice, I suspeCt that the above quoted passage 
would not have been written. 
Some very fine examples of such ice-cornices are visible 
from the ridge separating the Handspikjen Fjelde from the 
head of the Jostedal, where a fine view of the great neve or 
sneefornd is obtained. This side of the neve terminates in 
precipitous rock-walls ; at the foot of one of these is a 
dreary lake, the Styggevand. The overflow of the neve here 
forms great bending sheets that reach a short way down, 
and then break off and drop as small icebergs into the lake.* 
The ordinary course of glaciers afford abundant illustra- 
tions of the plasticity of such masses of ice. It spreads 
out where the valley widens, contracts where the valley 
narrows, and follows all the convexities or concavities of the 
axial line of its bed. If the bending thus enforced exceeds 
a certain degree of abruptness crevasses are formed, but a 
considerable bending occurs before the rupture is effected, 
and crevasses of considerable magnitude are commonly 
formed without severing one part of a glacier from another. 
They are usually V-shaped, in vertical section, and in many 
the rupture does not reach the bottom of the glacier. Very 
rarely indeed does a crevass cross the whole breadth of a 
glacier in such a manner as to completely separate, 
even temporarily, the lower from the upper part of the 
glacier. 
If a glacier can thus bend downwards without “ sundering 
its connection with the frozen mass behind,” surely it may 
bend upwards in a corresponding degree, either with or 
* See “ Through Norway with a Knapsack,” chapters 11 and 12, for further 
descriptions of these. 
