ME. HOPKINS ON THE THEOET OF THE MiOTION OF GLACIEES. 
711 
The maximum tension in the marginal portions of the glacier will be increased by F ; 
and it is in those portions, near the lower as well as near the upper surface of the mass, 
that crevasses will be most likely to be formed. The forces tending to form them may 
probably be much the same in both cases ; but the tendency of the incipient fissures to 
open into wide crevasses must almost necessarily be much counteracted near the lower 
surface by the action of the bottom of the valley on that surface. The sides of a fissure 
will not there be able to separate from each other with the same facility as at the free 
surface of the mass. In fact any incipient fissure formed within the glacier and not 
extending to its external surface, will have much less of this facility of subsequently 
opening, than those which are formed nearer to the outer surface and directly commu- 
nicate with it. I therefore conceive it to be very improbable that crevasses exist in the 
deeper portions of a glacier, equal in number and magnitude to those which are seen in 
its more superficial portions. 
35. Effects of the Crushing of the Mass. — It may frequently happen that the maximum 
pressure at any proposed point of a glacial mass shall be much greater than the 
maximum tension pi, or the maximum tangential force Tj (art. 18) at that point; and 
the dislocation may be produced by the crushing power of ^ 2 - The element may thus 
be broken into a greater or smaller number of fragments, and the constraint of the mass 
removed. Its onward motion will then be continued, and its continuity restored by 
regelation, till a repetition of the process becomes necessary to overcome a new state of 
constraint. It is in this manner that the property of regelation enables us so beautifully 
to account for the molecular mobility of a glacial mass, and the consequent freedom of 
its central to move more rapidly than its marginal portions, without any reference what- 
ever to the property of viscosity or plasticity. 
36. Dislocation 'produced hy the Tangential Force. — The third way in which the mass 
may be dislocated is, that the tangential cohesion may yield to the tangential force on 
any proposed element, as explained in article 28. For the greater simplicity of explana- 
tion, I shall suppose the sides of the valley parallel, and A=0, and B=0. Also we 
take C=0 and E=0 as in our first approximate solutions. Let^g'rs(fig. 6) be the 
5 E 2 
