282 THEORY OF COLLOIDAL BEHAVIOR 



groups which have a stronger attraction for each other than they 

 have for the solvent. The other groups adhere strongly to the 

 solvent, and hence they cannot come in contact with each other 

 even when the gelatin sets to a jelly. When a 1 per cent solution 

 of gelatin sets to a gel the distribution of the gelatin molecules 

 in the solvent undergoes probably no profound change. What 

 may change is, perhaps, the orientation of the gelatin molecules 

 or ions towards each other, but not their average distance from 

 each other. When, however, too much alcohol is added, i.e., 

 when the solution is on the alcohol side of the critical point, the 

 forces of attraction between gelatin and solvent are weakened 

 to such an extent that the groups which were formerly attracted 

 by the solvent are now more strongly attracted to each other than 

 they are to the molecules of solvent. In the micellae thus formed 

 the protein ions or molecules are in much closer contact than they 

 are in a jelly, and hence they occlude much less water than the 

 micellae formed in pure water or in mixtures of water with little 

 alcohol. The latter micellae increase the viscosity of the solution 

 more than the micellae formed in an excess of alcohol. The gelatin 

 would be precipitated in the latter solutions if it were not for the 

 fact that the coalescence of the protein ions and molecules- is 

 prevented by the forces set up as a consequence of the Donnan 

 equilibrium, namely, forces of osmotic pressure and of P.D., as 

 stated. When, however, a small quantity of salt is added the 

 forces set up by the Donnan equilibrium are diminished and 

 nothing now prevents the forces of attraction between the 

 gelatin molecules from causing the separating out of the gelatin 

 from solution. It should also be recalled that at the isoelectric 

 point not only the P.D. but also the osmotic pressure of protein 

 solutions is a minimum and that salts depress the osmotic pres- 

 sure as well as the P.D. 



The fact that a Donnan equilibrium is established between 

 micellae and surrounding liquid, whereby the opposite ions of 

 electrolytes are distributed in a definite way between the two 

 constituents, makes it clear why in the case of precipitation of 

 colloids some of the precipitating electrolyte must be found in the 

 precipitate, and that there can, as a rule, be no stoichiometric 

 relation between the quantity of salt and the mass of protein in 

 the precipitate. 



