JACQUES LOEB 



401 



influence, therefore, the transport of liquid in alkaline solution in the 

 same sense in electrical endosmose as in free osmosis. This influence 

 is in both cases an additive effect of the oppositely charged ions of the 

 electrolyte. In the alkaline solutions the watery phase of the double 

 layer is positively charged as it is in neutral solutions. If Perrin's 

 rule applied to these experiments, the increasing valency of the anion 

 should have had no effect. 



Table III contains also the transport numbers of solutions of Na 

 acetate, Na aluminate, and K oleate which are all higher than those 

 of NaCl, although the anion is monovalent in each case. Solutions of 



TABLE III. 



Transport of Liquid by Electrical Endosmose to the Cathode in Alkaline Solutions, 

 pH = 10.9 to 11.0. 40 Volts. 



M/512NaCl 



M/512Na2S04 



m/512 Na4Fe(CN)6 



m/512 CaCl2 



m/512 BaClz 



m/512 Na acetate 



m/512 NaA102 



m/512 K oleate (pH = 9.4) 



these salts also attract water more powerfully than solutions of NaCl 

 in the case of free osmosis and the influence of these salts in electri- 

 cal endosmose is parallel to their influence in free osmosis. These 

 salts illustrate the statement that in addition to the valency another 

 constitutional quantity of the ions determines their influence on the 

 transport of liquid in free and electrical endosmose. 



We finally investigated the electrical transport of liquid in acid 

 solutions. The salt solutions were made n/ 1,000 acid by the addition 

 of HNO3; the pH was in all cases exactly 3.0. Table IV gives the 

 results. The anode was in the glass tube. In this case it was neces- 

 sary to use membranes which had received a gelatin treatment. 



If Perrin's rule applied to these cases, the increasing valency of the 

 cation should not have influenced the result in these acid solutions. 



