1911.] 



The Properties of Colloidal Systems. 



247 



the fact of the diffusibility of the sodium chloride, viz., 18 hours at least. If 

 the outer solution be replaced by water while the sign is reversed, this latter 

 is changed back again to the original direction, but is naturally greater in 

 numerical value owing to the previous diffusion of sodium chloride into the 

 inner solution, which gradually escapes again into the water. 



The fact that the E.M.F. given by Congo-red solutions of unequal 

 concentration separated by a membrane can be abolished or reversed by 

 Na" ions does not, it seems to me, warrant the conclusion that the dye 

 solutions necessarily dissociate with the production of Na" ions. If I rightly 

 understand the rationale of the process, the production of the potential! 

 difference under discussion is a matter of electric charge of positive sign, so 

 that the precise chemical nature of the carrier of the charge is a matter of 

 indifference. That this point of view is correct is shown by the fact that 

 potassium chloride, as well as sodium chloride, is effective in reducing the 

 E.M.F. A Congo-red solution in dilution of about 21 litres gave, with 

 water outside the membrane, an E.M.F. of 005 volt ; when the water was 

 replaced by ?n/20 potassium chloride, this value was reduced to less than 

 O'OOl volt. This being the case, Ci and c 2 in the equation must apparently 

 be regarded as expressing the concentrations of positive charges. The 

 abolition of E.M.F. is, of course, only a temporary one, since the K ' ions 

 diffuse through the membrane and thus produce again a concentration of 

 positive ions greater on the inner side of the membrane. A potential 

 difference due to the greater migration rate of the potassium ion than that 

 of the sodium ion would not, I think, be perceptible under the conditions 

 of the experiment. It is of interest to note that this experiment shows how 

 sodium ions are enabled to escape by interchange with similarly charged ions 

 in the outer fluid. 



The possibility that the membrane itself may carry an electric charge must 

 not be neglected, although the concordance of the most reliable of the 

 experimental results given above with the calculated values shows that, if so 

 charged, this charge plays no appreciable part in the production of the 

 E.M.F. Moreover, it would only be very small in the presence of the large 

 electrolyte concentration on both sides. As far as osmotic pressure is 

 concerned, a priori considerations, for which I must refer to Mr. Hardy's 

 forthcoming work on " Colloids," show that a charge on a membrane could 

 only affect the time taken to attain equilibrium and would have no effect on 

 the permanent result, unless the permeability of the membrane were affected. 

 The total constraint exerted by the membrane, upon the non-dissociated mole- 

 cules and anions directly, upon the cations indirectly, would be unaffected. 



Since the cation is held within the membrane by electric forces alone, it is 



VOL. LXXXIV. — B. U 



