190 ELECTRIFICATION OF WATER 



salts. The writer has never observed negative osmosis when solu- 

 tions of neutral salts were separated from pure solvent by collodion 

 membranes. 



The writer has recently investigated the influence of the concen- 

 tration in acids and alkalies on the rate of diffusion of water through 

 collodion membranes (previously treated with gelatin), with the re- 

 sult that the phenomenon of negative osmosis described for tartaric 

 and oxalic acids is very common in acids and alkalies, that it occurs in 

 exactly the same range of concentration where the drop in the curves of the 

 neutral salt solutions occurs, namely within a range between M./256 to 

 about m/4, and that the phenomenon is in reality nothing else but this 

 drop. The difference between the nature of the drop in the case of 

 solutions of neutral salts on the one hand and of solutions of acids and 

 alkahes on the other is that in the case of alkalies and acids the drop 

 is not only relative but absolute. Titration experiments show that 

 acid diffuses from the acid solution into distilled water and that the 

 concentration of acid in the solution is considerably less after 20 

 minutes than at the beginning. Since at the same time the total 

 volume of solution diminishes — this being the nature of negative 

 osmosis — we must conclude that the diminution of the volume of 

 the solution is due to the fact that the combined volume of acid 

 and water diffusing out from the solution is slightly larger than the 

 volume of water diffusing simultaneously into the solution. We 

 shall return to this problem in a subsequent paper. Since acids as 

 well as bases diffuse into the pure solvent the phenomenon of nega- 

 tive osmosis can only be observed during a short period at the begin- 

 ning of the experiment. 



In all the experiments described in the preceding pages it was nec- 

 essary to put the solution into the collodion flask and to dip the latter 

 into distilled water in order to observe the influence of the solution on 

 the initial rate of diffusion of water. More water diffuses into the 

 solution than diffuses simultaneously in the opposite direction and as 

 a consequence the level of the water in the glass tube rises. When 

 we put solutions of certain acids or alkalies into the collodion flasks 

 and dip these flasks into distilled water the level of the Hquid in the 

 manometer falls instead of rising. If, however, we put the solution 

 of the same acids or alkalies outside, filhng the collodion flask with 



