and Attached Water. 207 



§§ 44-46 that the temperature of the constituents has nothing 

 to do with the temperature reached, and in § 49 that the degree 

 of hydration of the salt employed is often without effect. 



Accordingly I unqualifiedly withdraw the expression " ill- 

 defined" as applied to freezing-mixtures. They may be, on the 

 contrary, bodies of precise temperature under widely varying 

 circumstances; 



§ 42. The enormous latent heat of water, the fact that the 

 specific heat of ice is only about half that of water, while the 

 specific heats of all salts are far less than that of ice, and there- 

 fore, a fortiori, less than that of water, together with the good 

 thermal conductivity of water, all argue that, if constantly stirred, 

 all parts of a freezing-mixture will have the same temperature. 

 The fact that the liquid portion of a freezing-mixture of ice and 

 a solid salt is the cryohydrate of that salt, ensures the identity 

 of the resulting temperature under various conditions of propor- 

 tion. The constant tendency to the formation of this cryohydrate 

 by contact between the solids is always seeking to depress the 

 temperature ; while the solidification of the cryohydrate at an 

 indefinitely small fraction of a degree below the temperature of 

 the freezing-mixture, and the consequent liberation of heat, en- 

 sures the temperature against such fall. 



Statements therefore, whether previously made by myself or 

 others, that it is advantageous to weigh the salt and ice in defi- 

 nite proportion, that the ice should be dry, that snow is preferable 

 to ice on account of its state of finer division, that additional cold 

 is produced by previously cooling the ice or salt or both, are to 

 be put aside as untrue — untrue, that is, as far as the temperature 

 or heat-tension is concerned. To obtain the greatest quantity 

 of heat-absorption with a given amount of salt, such a quantity 

 of ice must be taken as will form with the salt a cryohydrate. 

 The proportions can be at once gathered from Table X. § 90. 



§ 43. Further, the fact that all cryohydrates, with, I believe 

 at present, the sole exception of sulphate of zinc and chloride of 

 magnesium, have far more water than that ordinary hydrate 

 which has most water, shows that it cannot matter whether a 

 salt which affects water of crystallization be employed in the 

 anhydrous state or with its crystalline water. 



It may, however, be otherwise with salts containing the ele- 

 ments of water more intimately associated ; and, as we shall see 

 in one case at least, an anhydrous compound may melt ice with 

 which it is in contact and heat the so-formed water far above 

 zero, while the compound so formed, when cool, will when mixed 

 with a fresh quantity of ice absorb heat abundantly. I suppose 

 in such cases double decomposition ensues ; and though nothing 



