474 ELECTRICAL CHARGES AND ANOMALOUS OSMOSIS 



possibility later and show that there is no adequate support for this 

 second assumption, though it cannot be absolutely excluded. 



Third, the difference between the curves for CeCls and CaCla is 

 smaller in Fig. 2 than in Fig. 1. It is possible that Ce increases the 

 value of € beyond that accounted for by the ionization of gelatin 

 chloride. 



Aside from these discrepancies we can say that Helmholtz's formula 

 explains the curves for anomalous osmosis given in Fig. 1 when the 

 values for c used are those to be expected on the basis of the Donnan 

 equilibrium. It may, therefore, be stated that the Donnan theory 

 is able to explain the phenomena of anomalous osmosis more com- 

 pletely than any other theory thus far offered. 



II. The Transport Curves on the Alkaline Side oj the Isoelectric Point. 



In these experiments the salt solutions were rendered alkaline by 

 adding enough KOH to bring the salt solution to a pH of 11 .0. The 

 outside solution was a pure KOH solution also of pH 11.0 but free from 

 salt. Fig. 3 gives the curves for the transport of liquid in the solution 

 during the first 20 minutes. The curves show a rise — until the con- 

 centration of the salt is about m/64 — followed by a drop, and then a 

 second rise follows at a concentration of about m/8. The general 

 character of the curves in Fig. 3 is about the same as that in Fig. 1 

 but the relative efficiency of the cations and anions is reversed. In 

 solutions whose pH is on the alkaline side of the isolectric point of 

 gelatin, the "attraction" of the solution for water increases with the 

 valency of the anion but inversely with the valency of the cation; 

 while on the acid side the relative efficiency of the two oppositely 

 charged ions is the reverse. Thus in Fig, 1 the curve for Na2S04 

 is flat while that for CaCl2 rises; in Fig. 3 the curve for Na2S04 rises 

 while that for CaCl2 is flat. The reason for this reversal is the fact 

 that the sign of charge between the liquid inside the pores of the 

 gelatin film and the gelatin wall of the pore is reversed on the opposite 

 sides of the isoelectric point. At pH 3.0 the gelatin is positively 

 charged and the liquid inside the pores is negatively charged; while at 

 pH 11.0 the gelatin is negatively charged and the liquid inside the 

 pores is positively charged. The sign of charge of the solution in tlie 

 p.D. across the membrane, i.e. of E, remains, however, the same in 

 alkali and acid solutions (Table V). 



