THE NEW THEORY OF SOLUTIONS. 183 



in B. Moreover, the diminution in solubility should follow 

 the same law as the diminution in vapour-pressure. Thus, 

 to take an example, using- water as solvent, if L be solu- 

 bility of ether, and L' the solubility of ether containing- a 

 dissolved substance insoluble in water, then in all the 

 equations we have discussed dealing with the vapour- 

 pressures of solutions, we may substitute (L-L')/L' for 

 (p-p')'iP', and hence from (3) vol. i., p. 411, we may obtain 



(L-L')/L' = «/N (5). _ 



Consequently, the relative diminution in the solubility 

 of the ether should be the same as the ratio of the number 

 of molecules of dissolved substance to the number of mole- 

 cules of ether present in the solution. In 1890 Nernst 

 gave a thermodynamical proof of the above equation ; 

 experiments to test its validity, although easy to perform, 

 are not very numerous, however, on account of the difficulty 

 of obtaining suitable solvents. The solvent B must be 

 practically insoluble in the liquid A. In his experiments 

 Nernst used water as solvent and found the solubility of 

 valeric acid, ether, and ethyl acetate, containing different 

 dissolved substances. So far as these experiments go 

 they accord with equation (5). 



We have here, therefore, another indirect mode of 

 proving that the gas-equation is applicable to dilute solu- 

 tions and another mode of estimating molecular weights. 

 Quite recently (1894) Kiister has attempted to make such 

 measurements on solubility the basis of a laboratory method 

 of determining molecular weights. 



Osmotic pressure and the partition-coefficient.- -The 

 analogy between dissolution and evaporation leads to an- 

 other and indeed the most direct indication that osmotic 

 pressure is independent of the nature of the solvent. 

 Berthelot and Jungrleisch found in 1872 that the partition- 

 coefficient of a substance between two mutually insoluble 

 solvents was independent of the amount of dissolved sub- 

 stance. Thus at 1 5 , if iodine be shaken with water and 

 carbon bisulphide, the ratio of the concentration of the 

 iodine in the bisulphide, to that in the water, is 410, no 

 matter how much iodine be employed. This result leads 



