88 ELECTRIPICATION OF WATER AND OSMOTIC PRESSURE 



which is to be applied to the solution to make the rate of diffusion 

 of water in both directions equal must also be doubled. 



This influence of the concentration holds strictly only as long as 

 the solute influences the rate of diffusion of water simply by the 

 number of its molecules (e.g. in preventing a number of water mole- 

 cules from impinging on the solution side of the membrane this 

 number being equal to the number of molecules of solute impinging 

 during the same time). Van't Hoff's law, however, must become 

 inadequate if the molecules of the solute can modify the rate of the 

 diffusion of water by other forces than mere gas pressure; e.g., by 

 electrical forces varying with the nature of the molecule. This is 

 probably true for any solute but in a much smaller degree when the 

 solute is a non-electrolyte than when it is an electrolyte; and in the 

 case of electrolytes it holds in a smaller degree when the electrostatic 

 field around the individual oppositely charged ions is nearly the same 

 (e.g. in the case of NaCl) than when it is very different as in the case 

 of Na4Fe(CN)6 or LaCls. This influence of the electrical field sur- 

 rounding the ions upon the rate of diffusion is due to the electrification 

 of the water molecules. 



We use the term electrification of water merely as a short expression 

 of the fact that electrostatic forces cause water to migrate in a 

 definite sense through a membrane. Whether this electrification of 

 water particles is due to a cluster formation of water molecules 

 around an ion as a nucleus, or to some other cause, may for the 

 present remain outside the discussion. 



It is the purpose of this paper to show that the electrification of 

 water molecules by ions in solution must in certain cases result in a 

 deviation of the osmotic pressure of a solution from that to be expected 

 on the basis of van't Hoff's theory; and that the sense and relative 

 quantity of deviation can be predicted. 



The writer has recently investigated the influence of various ions 

 on the rate of diffusion of water through a collodion membrane 

 separating pure water from a watery solution. Before being used 

 for the experiment the collodion membranes were filled over night 

 with a 1 per cent solution of gelatin and kept in water. The next 

 day the gelatin solution was removed and the collodion flasks were 

 rinsed a considerable number of times with warm water to remove 



