417. Consider what follows from these observations. Snow consists of small particles of ice. Now if by pressure we squeeze out the air entangled in thawing snow, and bring the little ice-granules into close contact, they may be expected to freeze together; and if the expulsion of the air be complete, the squeezed snow may be expected to assume the appearance of compact ice.
418. We arrive at this conclusion by reasoning; let us now test it by experiment, employing a suitable hydraulic press, and a mould to hold the snow. In exact accordance with our expectation, we convert by pressure the snow into ice.[H]
[H] A similar experiment was made by the Messrs. Schlagintweit prior to the discovery which explains it, and which therefore remained unsolved.
419. Place a compact mass of ice in a proper mould, and subject it to pressure. It breaks in pieces: squeeze the pieces forcibly together; they re-unite by regelation, and a compact piece of ice, totally different in shape from the first one, is taken from the press. To produce this effect the ice must be in a thawing condition. When its temperature is much below the melting point it is crushed by pressure, not into a pellucid mass of another shape, but into a white powder.
420. By means of suitable moulds you may in this way change the shape of ice to any extent, turning out spheres, and cups, and rings, and twisted ropes of the substance; the change of form in these cases being effected through rude fracture and regelation.
421. By applying the pressure carefully, rude fracture may be avoided, and the ice compelled slowly to change its form as if it were a plastic body.
422. Now our first experiment illustrates the consolidation of the snows of the higher Alpine regions. The deeper layers of the névé have to bear the weight of all above them, and are thereby converted into more or less perfect ice. And our last experiment illustrates the changes of form observed upon the glacier, where, by the slow and constant application of pressure, the ice gradually moulds itself to the valley, which it fills.
423. In glaciers, however, we have also ample illustrations of rude fracture and regelation. The opening and closing of crevasses illustrate this. The glacier is broken on the cascades and mended at their bases. When two branch glaciers lay their sides together, the regelation is so firm that they begin immediately to flow in the trunk glacier as a single stream. The medial moraine gives no indication by its slowness of motion that it is derived from the sluggish ice of the sides of the branch glaciers.
424. The gist of the Regelation Theory is that the ice of glaciers changes its form and preserves its continuity under pressure which keeps its particles together. But when subjected to tension, sooner than stretch it breaks, and behaves no longer as a viscous body.