SECTION III., 1907. [29] Trans. R. S. C. 
VLEquihbrium Between Ice and Water. 
By Elo). BARNps;  D:S0: 
Associate Professor of Physics, McGill University, Montreal. 
Read May 15, 1907. 
In previous papers which the writer has had the honour to read 
before the Royal Society, it was shown that the temperature of equili- 
brium between ice and water is not absolutely constant, but is dependent 
on the relative amounts of ice and water, and the rate at which freezing 
or melting is progressing. This is illustrated on a large scale in our 
rivers and lakes in winter, when small deviations from the freezing 
point are observed to accompany the formation or disintegration of ice. 
The production of frazil-ice, with its characteristic adhesive proper- 
ties, is observed only when the water in which it is being produced is 
in a slightly undercooled state. Thus our ice crystal becomes a centre 
of crystallization, imparting its latent heat to the surrounding water. 
A vessel of pure water may be supercooled when maintained free 
from agitation, but it is well known that the introduction of a small 
erystal of ice will cause the formation of ice throughout the mass, until 
the temperature arrives at the freezing point. Here the latent heat is 
liberated and serves to warm the water to the equilibrium temperature. 
In this case freezing takes place along the surface of the ice, and the 
ice is formed at the freezing point. 
There is another condition which is of some interest to investigate, 
where the ice itself is supercooled and brought in contact with water. 
In this case freezing takes place along the surface of the ice, and the 
heat which is liberated is conducted away, since ice is nearly four times 
as good a conductor for heat as water is. The water therefore becomes 
supercooled and remains so unless there is a general crystallization 
throughout the mass. It is probable that freezing proceeds along the 
surface of ice already present in cohtact with water rather than that 
new crystals form in the mass, so that a condition of supercooling might 
exist for a considerable time where the ice mass was large enough. This 
point was observed experimentally in a cake of clear river ice which had 
become supercooled by being left out during a part of the winter. The 
bleck, which measured about 14 inches square, was brought into the 
laboratory and a hole about half an inch in diameter was drilled into the 
centre. The whole block was covered with flannel and paper, and a 
sensitive mercury thermometer, reading to thousandths of a degree 
inserted in the hole. Some water produced by melting at the surface 
