262 THEORY OF SOLUTION AND ORYSTALLISATION. 



And even at 165^ the amount of salt retained in solution is not 

 appreciably diminished. An experiment of my own shows 42.6 

 parts of Na2 SO4 to 100 parts of water, in a solution saturated at 

 165*^0. There can be no doubt the result is complicated by the 

 increasing solubility of the anhydrous salt at higher temperatures, 

 so that if the solubility of sulphate of sodium could be traced 

 through a much greater range of temperature, it would be found 

 that at a certain temperature, at present unknown, the decrease of 

 apparent solubility, due to the decomposition of the hydrates, 

 would be more than compensated by the increasing solubility of 

 the anhydrous salt. 



One main point upon which it appears to me to be desirable to 

 insist is this, that a solution contains a tnixfure of several hydrates, 

 the constitution ofivhich depends partly upon the temperature of the 

 Liquid, and partly upon the proportion of water present. Such 

 hypotheses as those of Lowel, for example, to the effect that at one 

 temperature the solution contains one hydrate alone, and at another 

 temperature another entirely distinct, or indeed that any one 

 hydrate can exist in solution by itself, appear to me to be unsupported 

 by reason or the facts of the case. 



If we desire to understand why it is that dissociation of dissolved 

 hydrates occurs at elevated temperatures, and how it is that even 

 anhydrous salts may remain in solution, although usually in 

 relatively small quantity, it is only necessary to appeal to the 

 ordinary kinetic theory of liquids. The discussion of these details 

 would, however, extend the paper to undue length. 



There is one other topic upon which, in connection with this 

 subject, I find myself at variance with all those writers whose 

 opinions I have been able to consult in print. The constitution of 

 super-saturated solutions is an interesting question which has led 

 to much discussion. Viewed by the light of the theory which has 

 been briefly sketched in the preceding pages, it appears to me 

 highly probable, that a super-saturated solution is a liquid in a state 

 of unstable molecular equilibrium resulting from dissociation. 



