IVIR. HOPKINS ON THE THEOET OE THE MOTION OF GLACIEES. 
689 
mena attendant upon it. The formation of crevasses was a necessary consequence of the 
forces acting on the glacier, and the conditions to which it was subjected ; but no reason 
was thus assigned why such crevasses should be obliterated again, as they were frequently 
observed to be, and the continuity of the mass perfectly restored. Moreover, it became 
evident from more accurate and detailed observation, that the continuity of many parts 
of the general mass was preserved in a degree apparently inconsistent with the change 
of form to which a mass so crystalline and brittle as glacial ice did manifestly submit. 
It was to meet this difficulty that Principal Foebes was led to the hypothesis of the 
viscosity ox plasticity of glacial ice. I have already explained my objection to the vague- 
ness with which these terms appear to me to have been used, and to the total want 
of all experimental proof of any property in ice which could be so designated with accu- 
racy, or with a due regard to the propriety of scientific language. Difficulties of this 
kind always remained on the minds of certain glacialists, till the experiments of 
Mr. Faeadat and Dr, Tyndall at once explained to us that regelation, and not viscosity, 
was the real property of ice required for the completion of our general theory of glacial 
motion. This property of regelation belongs essentially to solid bodies, and in treating 
glacial masses as bodies possessing the property of regelation, we must necessarily treat 
them as solid. As such I consider them in the following investigations, the object of 
which is to ascertain, as far as we are able, the internal pressures and tensions to which 
glaciers are subjected, and the phenomena which may result from them, more especially 
those connected with the veined structure of glacial ice, and the formation of cre- 
vasses. 
12. Before I proceed to these investigations, I would here remark that the import- 
ance of a distinct conception of the properties indicated by the terms viscosity or plasti- 
city on the one hand, and solidity on the other, will be at once apparent if we consider 
the difference between the mechanical problems presented to us in the motion of glaciers, 
according as we conceive them to be typified by a viscous or solid mass. In the first 
case we have to determine the continuous motion of a mass the component particles of 
which move with different velocities, but without destroying its continuity. The most 
simple, and the limiting case, would be that in which the tangential action of conti- 
guous particles on each other should vanish. The mass would then become a fluid mass. 
But even in this case we can do little by accurate mathematical investigation, and still 
less in the case in which the mass is viscous. Consequently, the objection against any 
attempt at a mathematical solution of the problem of glacial motion, founded on our igno- 
rance of the motions of viscous masses, is perfectly valid, so long as we treat glacial ice 
as viscous according to our definition of that term. But this same objection has been 
urged against all attempts to apply accurate mathematical processes to the problem, in its 
complete or partial solution, under the supposition of ice having the property of solidity. 
The complete solution of the problem would undoubtedly be far more difficult for a 
solid than for a viscous mass ; for it would involve conditions depending on innumerable 
discontinuities in the mass, resulting from its motion. All that can be attempted is a 
