1911.] 



The Properties of Colloidal Systems. 



241 



where (s) stands for the non-diffusible anion. It is probable that there is an 

 uncertainty about the concentration of the dye-acid in the positive flask 

 owing to precipitation. 



But this equation, even if otherwise acceptable, goes too far, since the 

 osmotic pressure of the dissociated salt would be less than that of the 

 non-dissociated. If it were 



2(§)Na 2 ^ Nad)' + ®Na 3 ', 



the experimental facts would be satisfied as far as the osmotic pressure is 

 concerned. There is difficulty, however, with the electrical conductivity. 

 "We have seen that the molar conductivity at infinite dilution is practically 

 tbe same as that of sodium sulphate, if we allow for the slower migration 

 rate of the large organic anion. Now one dissociated molecule of this salt 

 carries four times as many charges as one of Congo red, according to the 

 last equation, if sodium sulphate dissociates thus : — 

 Na 2 S0 4 ^2Na- + S0 4 ". 



Moreover, if we suppose that at infinite dilution Congo red is completely 

 and normally dissociated, an osmotic pressure considerably exceeding the 

 " molecular " as dilution increases would be given. This is so only to a very 

 small degree, if at all, as we have seen. On the other hand, if the dissocia- 

 tion were of the type suggested, with complex ions, it seems impossible to 

 explain the facts illustrated by the following case : at 28 litres dilution the 

 molar conductivity of a Congo-red solution is half that at infinite dilution, 

 whereas, on the supposition of complex ions, it would have only one-quarter 

 the conductivity which it has at infinite dilution, even if it were completely 

 dissociated at the higher concentration, a very unlikely assumption. 



If, however, it were allowable to assume that the complex ions could retain 

 the combined charges of their components, thus : — 



2(S)Na 2 ^ Na"(§)" + (S)"Na 3 - ' ', 

 the matter could be explained. 



A similar difficulty was met with by Brailsford Bobertson* in his investi- 

 gation on the conductivity of the potassium salt of caseinogen. Under 

 certain conditions the addition of potassium chloride to such salts causes no 

 diminution of ionisation. The conclusion drawn is that potassium caseino- 

 genate does not form K' ions. But, at the same time, it is pointed out that, 

 in order to explain other experimental facts, it must be supposed that the 

 complex ions are able to transport an increased number of atomic charges. 



* ' Journ. Phys. Chem.,' 1910, vol. 14, pp. 556 and 557. 



