52 ^ AMPHOTERIC COLLOIDS. I 



then freed again from the salt by perfusion with water. ^ In this case 

 the same resuhs were obtained as those stated in this paper; namely: 

 When pH of the gelatin solutions is >4.7, the gelatin can exist in an 

 ionized condition only as an anion capable of combining with cations 

 only to form metal gelatinates, while on the more acid side of the iso- 

 electric point ipH <4.7) gelatin can exist in an ionized form only as a 

 cation capable oj combining only with the anion of the neutral salt. 

 This law holds probably for all amphoteric electrolytes. The question 

 then arises: How does an amphoteric electrolyte react with neutral 

 salts at its isoelectric point? Before we answer this question we will 

 point out an important fact shown by our curves, though this has no 

 direct bearing on the subject under discussion. 



III. 



A comparison of the curves for the conductivities with those for the 

 other properties of gelatin in Figs. 1 to 4 shows that the curves for 

 conductivity are practically identical in all four figures as are also the 

 values for pH; while the curves for osmotic pressure, etc., differ 

 according to the valency of the ion combining with the gelatin. 

 Since the conductivity curves are the direct expression of the degree 

 of electrolytic dissociation, we can now say that the difference in the 

 effect of univalent and bivalent ions demonstrated in the writer's 

 previous publications is not merely a difference of the degree of ioni- 

 zation of the gelatin but is due to another effect upon the gelatin mole- 

 cule. The important work by Langmuir and by Harkins on surface 

 tension has suggested if not proved the idea of polarized molecules 

 and of different degrees of solubility for different parts of the same 

 organic molecule. It seems possible from their work that the solu- 

 bility is confined to certain groups of the gelatin molecule or the 

 protein molecule in general. It is quite possible that this will ulti- 

 mately explain why the entrance of a univalent anion or cation into a 

 gelatin molecule increases its solubility in water (or its attracting 

 power for water molecules) and with this its osmotic pressure, alcohol 

 number, viscosity, etc., while the entrance of a bivalent ion (com- 



' In order not to increase the bulk of this paper we may be pardoned for omitting 

 the details of these experiments. 



