364 PROCEEDINGS OF THE AMERICAN ACADEMY. 



Abegg * carried out under the direction of Arrlieuius. Abegg let 

 alcohol diffuse into a salt solution and found, to his surprise, that the 

 salt, instead of remaining equally divided throughout the liquid, diffused 

 somewhat into the part not yet reached by the alcohol. He concludes 

 that this extraordinary behavior can only be accounted for on the 

 assumption that alcohol increases the osmotic pressure of a dissolved 

 salt. What happens is very simple. When the alcohol has diffused 

 only a little way, one may consider the solution as composed of two 

 parts, one containing a large amount of alcohol, the other very little. 

 The dissolved substance, being in this case less soluble in the first 

 layer than in the second, diffuses into the second only to go back again 

 as the alcohol becomes more evenly divided throughout the liquid. 

 Except that the part containing much alcohol and little water merges 

 insensibly into the part containing much water and little alcohol, and 

 is not in equilibrium with it, the case does not differ from two layers 

 formed by ether and water, where it is well known that the concen- 

 tration of a third substance is not the same in the two layers. The 

 effect of the alcohol is not, as Abegg assumes, to increase the osmotic 

 pressure of the solute, but to diminish its solubility in that portion of 

 the liquid. If, instead of taking salts which were only slightly soluble 

 in alcohol, Abegs; had let water diffuse into water containin"; in solution 

 some substance very soluble in alcohol, slightly soluble in water, he 

 would have observed the opposite effect, and the dissolved substance 

 would have diffused partially into tlie layer rich in alcohol. 



Another line of reasoning which is not quite defensible is that taken 

 by Wildermann,t in his paper, " Ueber cyclische Gleichgewichte." 

 His train of thought is something as follows. Suppose he has a system 

 of three phases, bromine, a solution of bromine in water, and the 

 vapor of bromine and water, it being assumed that the amount of 

 water which dissolves in the bromine can be neglected. He adds to 

 the aqueous solution some substance which does not dissolve in bromine 

 perceptibly, such as potassium bromide or sulphuric acid. The three 

 phases, when in equilibrium, have still the same concentration of liquid 

 bromine and of bromine vapor. Therefore the solubility of the bro- 

 mine in the liquid cannot have changed. It does change experimen- 

 tally ; therefore, in order to reconcile the reasoning with the facts, he 

 concludes that the apparent change, decrease or increase, is due to 

 chemical action, and that the amount of bromine dissolved as such 

 remains unchanged. This may be true in the special examples studied 



* Zeitsclir. f. ph. Chem., XI. 248. 1893. t Ibid., XI. 407. 



