JACQUES LOEB 



265 



I 



g 



to 



t 



s 



IB 



4096 2048 1024 512 256 128 64 32 16 "8 T 2 ^^^ ^^'■ 



Concentration 



Fig. 7. Negative osmosis when solutions of acids (tartaric, citric, and oxalic) 

 are separated from water by membranes treated with gelatin. The weak acetic 

 acid behaves practically like a non-electrolyte. 



The table shows that the amount of add which diffused out from 

 the collodion bag into the distilled water was (within the limits of 

 accuracy of these experiments) the same for membranes treated and 

 not treated with gelatin. Oxalic acid diffused a little more slowly 

 than hydrochloric acid, but the rate of diffusion for each acid is approx- 

 imately in direct proportion to its concentration. The fact that the 

 same acids caused positive osmosis when the membranes were not 

 treated with gelatin and negative osmosis when they were treated 

 cannot be ascribed to differences in the permeability of the two types 

 of collodion membranes. 



Since solutions of AI2CI6 and of LaCls caused a vigorous diffusion 

 of water from solvent into solution when the collodion membrane 

 was treated with gelatin, but not when the membrane had not received 

 a gelatin treatment, the idea suggested itself that the collodion mem- 

 branes not treated with gelatin were so permeable for the molecules 

 of the salt that an increase in volume of the solution by a diffusion 

 of water into it was impossible. Titration experiments with Al2Br6 

 proved, however, that there was only a slight difference in the rate 



