430. It is well to converge by means of a concave mirror a good light upon the ice, and to view it through a magnifying lens. You already see the result. Hazy surfaces are formed in the very body of the ice, which gradually expand as the pressure is slowly augmented. Here and there you notice something resembling crystallisation; fern-shaped figures run with considerable rapidity through the ice, and when you look carefully at their points and edges you find them in visible motion. These hazy surfaces are spaces of liquefaction, and the motion you see is that of the ice falling to water under the pressure. That water is colder than the ice was before the pressure was applied, and if the pressure be relieved, not only does the liquefaction cease, but the water re-freezes. The cold produced by its liquefaction under pressure is sufficient to re-congeal it when the pressure is removed. 431. If instead of diffusing the pressure over surfaces of considerable extent, we concentrate it on a small surface, the liquefaction will of course be more rapid, and this is what Mr. Bottomley has recently done in an experiment of singular beauty and interest. Let us support on blocks of wood the two ends of a bar of ice 10 inches long, 4 inches deep, and 3 wide, and let us loop over its middle a copper wire one-twentieth, or even one-tenth, of an inch in thickness. Connecting the two ends of the wire together, and suspending from it a weight of 12 or 14 pounds, the whole pressure of this weight is concentrated on the ice which supports the wire. What is the consequence? The ice underneath the wire liquefies; the water of liquefaction escapes round the wire, but the moment it is relieved from the pressure it freezes, and round about the wire, even before it has entered the ice, you have a frozen casing. The wire continues to sink in the ice; the water incessantly escapes, freezing as it does so behind the wire. In half an hour the weight falls; the wire has gone clean through the ice. You can plainly see where it has passed, but the two severed pieces of ice are so firmly frozen together that they will break elsewhere as soon as along the surface of regelation. 432. Another beautiful experiment bearing upon this point has recently been made by M. Boussingault. He filled a hollow steel cylinder with water and chilled it. In passing to ice, water, as you know, expands (§ 45); in fact, room for expansion is a necessary condition of solidification. But in the present case the strong steel resisted the expansion, the water in consequence remaining liquid at a temperature of more than 30° Fahr. below the ordinary freezing point. A bullet within the cylinder rattled about at this temperature, showing that the water was still liquid. On opening the tap the liquid, relieved of the pressure, was instantly converted into ice.