MEMBRANE POTENTIALS 149 



exclusively determined by the Donnan equilibrium, since 

 otherwise the quantitative agreement between the observed 

 P.D. and the values calculated from Donnan's equation would 

 be impossible. 



It may be stated, finally, that since P.D. = 29 log (l + |) 



millivolts, it follows that the addition of a non-electrolyte to a 

 gelatin chloride solution cannot influence the P.D. since the 



addition of a non-electrolyte cannot affect the value of - 



The expression for P.D. also explains why the P.D. is zero 

 at the isoelectric point of a protein, since z becomes zero at that 

 point. Moreover, it is obvious that the addition of a neutral 

 salt to a solution of isoelectric gelatin cannot further depress 

 the P.D. or cause a reversal in the sign of the charge. 



Donnan's theory of membrane equilibrium therefore explains 

 mathematically and quantitatively the P.D. observed at equi- 

 librium between a protein solution and a watery solution free 

 from protein, separated by a membrane. It does not happen 

 very often that every postulate of a theory is fulfilled by the 

 observation as it is in this case. 



The bearing of this fact upon the theories of colloidal behavior 

 is as follows: The P.D. is influenced by the hydrogen ion con- 

 centration, by the valency of the ion, by the addition of neutral 

 salt in the same way as are the other properties of protein, such 

 as osmotic pressure, viscosity, and swelling. Since the Donnan 

 theory explains this influence of pH, valency, and salt effect on 

 P.D. with mathematical accuracy, it seems at least highly 

 probable that it may also explain the other colloidal properties 

 of proteins in the same way, with this difference only, that the 

 experimental error is much greater in the measurements of the 

 other properties than in the measurements of P.D. 



