JACQUES LOEB 719 



which is an amphoteric electrolyte {e.g. gelatin or silk) and those 

 which consist of non-amphoteric material. Both types are not 

 necessarily influenced in the same way by acids and alkalies. Am- 

 photeric electrolytes form salts with acids as well as with alkalies 

 and this salt formation reverses the influence of the diaphragm 

 on the sign of electric endosmose. HCl transforms gelatin into 

 gelatin chloride, and NaOH transforms it into sodium gelatinate. 

 In the case of gelatin chloride the gelatin ion is positive and in 

 the case of sodium gelatinate it is negative. We assume that it 

 is primarily the sign of the gelatin ion which determines the 

 sense of the electrification of water and that the OH and H ions 

 play only a secondary part. This would explain why Perrin could 

 find no other monovalent ion except H or OH which was able to re- 

 verse the electrification of the gelatin diaphragm, since only acids and 

 alkalies can transform a gelatin cation into a gelatin anion, or vice 

 versa, while neutral salts cannot produce such an effect.^ 



We are, therefore, of the opinion that the experiments on electrical 

 endosmose through diaphragms of colloidal substances like gelatin, 

 pig's bladder, silk, or any other amphoteric electrolyte, do not warrant 

 the assumption of a specific influence or predominant position of 

 the H and OH ions in the electrification of a membrane bounded on 

 both sides by water, except in so far as acids and alkalies are the only 

 substances which can reverse the sense of ionization of an ampho- 

 teric electrolyte. 



Girard, in accepting Perrin's idea of a direct electrification of the 

 wall by H and OH ions, and of the specific action of these two ions, 

 tried to ascribe the phenomena of abnormal osmosis he observed with 

 membranes of pig's bladder to the acid or alkaline reaction of the 

 solution. Thus neutral solutions, like those of NaCl and Na2S04, 

 should be without any other except a purely osmotic effect and he 

 publishes data apparently supporting this conclusion. 



The writer's results with collodion membranes do not agree with 

 the conclusions of Girard, and we shall see that the solutions of neutral 

 salts act as powerfully on the rate of diffusion of water through col- 

 lodion membranes as the solutions of alkalies, acids, or acid salts. We 



6 Loeb, J., /. Gen. Physiol., 1918-19, i, 39, 237. 



