284 THEORY OF COLLOIDAL BEHAVIOR 



sufficiently raised, a change occurs nobody knows of what 

 nature in the molecule, the solutions become opalescent and 

 the albumin may set to a solid gel. In this case "oily" groups 

 must be activated or formed which were not active or did not 

 exist before and their natural attraction must cause the gel 

 formation. 



In the case of casein solutions in acid all the properties of the 

 solution, stability, viscosity, osmotic pressure, and P.D., possess 

 colloidal character. The forces of attraction between the mole- 

 cules or ions of casein on the acid side of the isoelectric point for 

 each other are very strong and for the water molecules compara- 

 tively feeble. In the case of casein chloride or casein phosphate 

 and to a lesser extent casein nitrate, they are just strong enough to 

 make a solution possible ; but the stability of the solution depends 

 largely on the forces set up by the Donnan equilibrium between 

 the nascent micellae or the existing micellae and the surrounding 

 solution. In the case of casein trichloracetate or casein sulphate 

 the forces of mutual attraction between the casein molecules for 

 each other are so much greater than those for water that these 

 two salts are practically insoluble. On account of this great 

 attraction for each other the granules of casein cannot swell in 

 sulphuric acid or trichloracetic acid. Solutions of Na caseinate 

 behave like crystalloidal solutions in regard to stability but like 

 colloidal solutions in regard to viscosity, P.D., and osmotic 

 pressure. 



This shows the complications which may be due to the large 

 size and complex constitution of large molecules, especially of 

 proteins. The problems of gel formation or of precipitation are 

 not colloidal problems, they are a part of the more general prob- 

 lem of solubility. These problems enter only in a secondary 

 way into the problem of colloidal behavior, since the phenomena 

 of aggregation are only a means of preventing the diffusion of an 

 ion, thereby creating the conditions for the establishment of a 

 Donnan equilibrium. 



We now understand why it is not correct to define colloidal 

 solutions as solutions in which the ultimate unit is a micella, 

 i.e., an aggregate of molecules or ions. The colloidal behavior of 

 solutions of salts of crystalline egg albumin in regard to osmotic 

 pressure and P.D. is only due to the non-diffusibility of the pro- 



