Investigations dealing with the State of Aggregation. 401 



in the following manner: Before coagulation takes place, the colloidal 

 particles must lose a certain part of their charge, so that the difference of 

 potential between colloid and dispersion medium does not exceed a certain 

 critical maximum (Powis*). This partial discharge can be brought about, 

 when the precipitating capacity of various salts is considered, by the 

 adsorption of equivalent quantities of ions with a charge opposite to that of 

 the colloids ; if one particular active ion is more readily adsorbed by a colloid 

 than another of equal valency, a salt containing the former ion will produce 

 the degree of discharge necessary for flocculation in a lower concentration of 

 solution than will the salt which contains the ion that is less readily 

 adsorbed. By means of the adsorption hypothesis, Freundlicb has endea- 

 voured to explain the numerical relationship between the flocculating 

 capacities of salts containing ions of different valencies.f The greater 

 precipitating capacity of salts containing organic ions as compared with 

 those containing inorganic ions has also been explained by the assump- 

 tion that the former are more readily adsorbed than the latter. Furthermore, 

 typical adsorption isotherms have been obtained, when the distribution of 

 salts between colloid and dispersion medium has been determined.^ 



A totally different conception has been put forward by Duclaux,§ who 

 regards the colloid particle as a micella containing a "granule" of the 

 colloid proper associated with " active " ions derived from a substance used in 

 the preparation of a colloidal solution. Thus, in the case of the ferric 

 hydroxide sol, prepared by the dialysis of a solution of ferric hydroxide in 

 ferric chloride, the granules consist of ferric hydroxide, and they are 

 .associated with active chlorine ions, and flocculation on addition of a salt is 

 regarded as taking place by the substitution of the active ions by equivalent 

 quantities of others by a purely chemical reaction. It is not necessary to 

 discuss in detail Duclaux's hypothesis, as certain of his assumptions are no 

 longer tenable, as, for example, those dealing with the osmotic pressure of 

 colloidal solutions, which are not in accord with subsequent researches of 

 Perrin on the equipartition of energy in solutions. 



More recently still, Pauli and Matula|| have put forward a somewhat 

 different hypothesis, according to which flocculation is regarded as due mainly 

 to a chemical action. Their conceptions are founded on a study of the ferric 



* ' Zeitsch. Physikal. Chera.,' vol. 89, p. 195 (1914). 



t Freundlich, ' Kapillarchemie,' Leipzig, 1909, pp. 354-355. 



J See e.g. Maffia, ' Kolloid-chem. Beihefte,' vol. 3, p. 85 (1911), and Gann, ibid., vol. 8, 

 p. 63 (1916). 



§ ' Jouin. Chim. Physique,' vol. 5, p. 29 (1907), and vol. 7, p. 405 (1909). 

 II ' Kolloid Zeitsch.,' vol. 21, 49 (1917). 



