THE THIRST OF SALTED WATER 653 



the water. A simple solution of the problem would be given, 

 if it were supposed that water is a mixture of molecules of 

 different orders of complexity, in equilibrium — a mixture of 

 simple hydrone (OH 2 ) molecules with " polyhydrones " of various 

 orders of complexity (see Science Progress, No. 11). The in- 

 trusion even of an entirely neutral substance (provided it were 

 soluble) into such an assemblage would disturb the equilibrium 

 in the direction represented by the generalised equation : 



(H,0)x -> xOH, 



and the extent of the disturbance — the extent to which the 

 polyhydrones would be dissociated — would depend on the 

 number of individual neutral molecules introduced. To restore 

 equilibrium in the direction 



(0H,)x <- xOH 2 



it would be necessary to cool such a solution more or less 

 below the temperature at which ice would separate from pure 

 water. The boiling point of the solution would be higher than 

 that of water because of the presence of a larger proportion of 

 attractive hydrone molecules and of molecules of the simpler 

 polyhydrones. " Thirst " of hydrone would be conditioned 

 in such a solution by the presence of an increased proportion 

 of attractive hydrone molecules seeking to form water. The 

 properties of aqueous solutions of non-electrolytes were thus 

 referred to the fact that molecules of " Fair Hydrone " are 

 eminently attractive of their own kind. 



It thus became obvious that the method adopted by Van't 

 Hoff and others of using volume normal solutions must be a 

 fallacious one, except in the case of very dilute solutions, owing 

 to the great variation in the relative molecular proportions of 

 solvent and solute in such solutions, a substance of small 

 molecular weight (such as alcohol) displacing but little water, 

 one of high molecular weight (such as cane sugar) a large 

 amount. It was clear that the solvent was the test substance — 

 that all thought of analogy with the gaseous state must be put 

 aside and the effect ascertained of this and that substance on 

 this and that property of some definite mass of the liquid. 

 Raoult's work, for example, had been done from such a rational 

 point of view ; its real significance had been obscured, however, 

 by the use of a false analogy. 



