154 
behaviour is perhaps to be sought in the peculiar eflPect of 
pressure on the melting of ice, I have here an experiment 
in progress in which a weight of 56 lbs. is suspended by a 
wire passing over a block of ice, and the wire is gradually 
making its way through the ice, but without destroying its 
continuity; the ice melts at the points of pressure, and 
freezes again as soon as the pressure is relieved. 
Professor James Thomson first pointed out, from theo- 
retical considerations, that the melting point of ice must be 
lowered by pressure; and calculated from the expansion which 
takes place, when water passes into ice, that an increase in 
pressure of the atmosphere would lower the melting point of 
ice by 0*0075° C, and this result was experimentally verified 
by Sir W". Thomson, who subjected a mixture of water and 
ice at 0° C. to pressure, and found that the mixture became 
colder ; some of the ice melting, and the latent heat thus ab- 
stracted from the remainder, lowered the temperature of the 
whole. Under a pressure of 16"8 atmospheres, the melting 
point was lowered 0"129° C. 
^^ext, Mousson succeeded in maintaining water in a liquid 
state at — 5° C.,and by a pressure of 13,000 atmospheres caused 
ice to melt at — 18° C. The apparatus employed consists 
of a strong iron vessel with a cylindrical cavity, in 
which ice can be compressed by means of a powerful 
screw. The apparatus being filled with water, and a 
loose piece of copper having been dropped in, was inverted, 
and in this position surrounded by a freezing mixture 
so as to freeze the water. When the water was frozen, 
the apparatus was placed upright, and the piece of copper was 
of course now frozen into the ice at the top. Pressure was 
then applied, the whole apparatus being kept at a temperature 
several degrees below freezing ; on relieving the pressure and 
opening the apparatus it was found still full of ice, but the piece 
of copper was now at the bottom, showing that under the 
