734 ELECTRIFICATION AND DIFFUSION OF WATER 



pressure as a m/64 solution of cane sugar. If we surround the bag 

 containing the m/128 solution of KCl with a solution of m/64 cane 

 sugar approximately equal numbers of molecules of water should im- 

 pinge on both sides of the collodion membrane and no change in the 

 level of the two liquids should occur, if only the osmotic pressure p 

 were active. As a matter of fact the water will diffuse from the 

 m/64 sugar solution into the m/128 solution of KCl owing to the 

 electrical force e due to the electrolyte. By using higher concen- 

 trations of cane sugar we arrive at finding one concentration which 

 balances the combined osmotic pressure and electrical force of a m/128 

 solution of NaCl. The concentration of cane sugar which balances 

 the combined osmotic and electric attraction of m/128 NaCl for 

 water was found to be approximately m/8. Assuming that the 

 osmotic attraction p of m/128 NaCl for water is equal to the osmotic 

 pressure of a m/64 solution of cane sugar, the force e of the m/128 

 KCl solution must be equal to the osmotic pressure m/8 — m/64 = 



7 X 22.4 

 7 m/64, or ■ =2.4 atmospheres. In this way we have de- 

 termined the value of e for a number of electrolytes by selecting as the 

 balancing concentration that concentration of a cane sugar solution 

 in which the level of liquid does not rise when a beaker filled with the 

 cane sugar solution is dipped into a solution of the electrolyte. This 

 method has the advantage that the concentration of the balancing 

 solution of cane sugar can be roughly estimated in about 5 minutes, 

 although in fact most of our experiments were continued for half an 

 hour or an hour. We used the same set of membranes for a large 

 number of experiments and two sets were used in all; occasionally a 

 membrane began to leak and we had to replace it by another membrane. 

 Measurements of the height of the column of solution inside the tube 

 were taken at first every 5 and later every 10 minutes. The sugar 

 solutions were in the bag and 350 cc. of the m/128 NaCl solution were 

 outside. The initial level of the Hquid in the manometer tube was 

 about 25 mm. above the level of the sugar solution. Table V gives 

 the results of our measurements. The balancing solution is always 

 that concentration of cane sugar in which the level of liquid in the 

 manometer tube falls slightly in the first 10 minutes, while in the 

 next higher concentration of cane sugar a marked rise occurs. It 



