598 Dr. G. Hevesy on the Diffusion and 



This value must as usual be increased by 13 per cent, to 

 eliminate the retarding influence of the alcohol. 



Sometimes it is difficult to distinguish from diffusion data 

 alone whether colloidal particles or ions with multiple charges 

 are present ; in both cases diffusion takes place slowly, in the 

 first case because an aggregate of molecules which acts as a 

 single molecule diffuses but with greater friction, and in the 

 latter case, on account of the increased hydration of the ions 

 which has the result of increasing the friction. 



Colloidal particles are often considered as large ions *. If 

 the same considerations that have been used above to deter- 

 mine the number of charges on an ion are applied to find 

 the number of charges on a colloidal particle, then the great 

 number of charges is proved by the smallness of the diffu- 

 sion velocity and by the proportionate great mobility of 

 colloids in an electric field. If the colloidal particles had 

 only one charge they would probably move in an electric 

 field with a small velocity. This velocity can be calculated 

 from the diffusion coefficient according to Nernst's equation. 

 Other considerations likewise lead to the conclusion that the 

 colloidal particles have numerous charges t- 



In the case of the alkaline solutions of the members of the 

 lead type, it is evident from the diffusion data without 

 further discussion that a colloid is present, since Pb(0H) 2 

 can only form ions with at most two charges. It is other- 

 wise with polonium, the chemical properties of which are 

 little known, and which can be present in solution in different 

 electrolytic forms. Nevertheless a definite decision was 

 possible when the sign of the charge on the polonium 

 particles was determined. It was then found that in alkaline 

 solution the great part of the polonium is charged negatively, 

 in acid solution it is positively, and in neutral and very 

 weakly acid solution it is predominantly negatively charged. 

 If the negatively charged particles formed an electrolyte 

 they must produce a complex anion, and such a solution can 

 scarcely be as stable in presence of strong alkali as in weakly 

 acid solution, since it has six charges — the minimum number 

 which is compatible with any of the above given diffusion 

 constants. 



The diffusion constants calculated from the different layers 

 show considerable variation. This is caused partly by the 

 precipitation of some particles on the glass walls during 



* J. Billiter, Zeit.plujs. Chem. iv. p. 365 (1906) ; Freundlich, Kapillar- 

 chemie, p. 340. 



t Smoluchowski, Bull, cle VAcad. de Science de Cracovie, p. 182 

 (1903) ; Zsigmondy, Kolloidchemie (Leipzig, 1912), p. 51. 



