Ice at 32 deg. may, as already stated, be crushed with extreme facility, and glacier-ice with still more readiness than lake-ice: it may also be scraped with a knife with even greater facility than some kinds of chalk. In comparison with ice at 100 deg. below the freezing point, it might be popularly called _soft_. But its softness is not that of paste, or wax, or treacle, or lava, or honey, or tar. It is the softness of calcareous spar in comparison with that of rock-crystal; and although the latter is incomparably harder than the former, I think it will be conceded that the term viscous would be equally inapplicable to both. My object here is clearly to define terms, and not permit physical error to lurk beneath them. How far this ice, with a softness thus defined, when subjected to the gradual pressures exerted in a glacier, is bruised and broken, and how far the motion of its parts may approach to that of a truly viscous body under pressure, I do not know. The critical point here is that the ice changes its form, and preserves its continuity, during its motion, in virtue of _external_ force. It remains continuous whilst it moves, because its particles are kept in juxtaposition by pressure, and when this external prop is removed, and the ice, subjected to tension, has to depend solely upon the mobility of its own particles to preserve its continuity, the analogy with a viscous body instantly breaks down.[A] FOOTNOTES: [A] "Imagine," writes Professor Forbes, "a long narrow trough or canal, stopped at both ends and filled to a considerable depth with treacle, honey, tar, or any such viscid fluid. Imagine one end of the trough to give way, the bottom still remaining horizontal: if the friction of the fluid against the bottom be greater than the friction against its own particles, the upper strata will roll over the lower ones, and protrude in a convex slope, which will be propagated backwards towards the other or closed end of the trough. Had the matter been quite fluid the whole would have run out, and spread itself on a level: as it is, it assumes precisely the conditions which we suppose to exist in a glacier." This is perfectly definite, and my equally definite opinion is that no glacier ever exhibited the mechanical effects implied by this experiment. REGELATION. (23.) [Sidenote: FARADAY'S FIRST EXPERIMENT.]