ON COLLOID CHEMISTRY AND ITS INDUSTRIAL APPLICATIONS. 35 
Or solid may adsorb the solvent itself and be peptized, while the 
other possibilities, all of which tend to produce peptization, are 
adsorption of a non-electrolyte, an indissociated salt, or a second 
colloid.** Since electrical endosmose has to do with electrically 
charged surfaces, we are concerned chiefly with preferential adsorp- 
tion of ions. 
It is necessary to postulate that every solid has a specific adsorbing 
power fora given ion, depending upon the specific surface*® of the 
solid, upon the temperature, upon the concentration of the particular 
ion in the solution and upon the other ions present, or adsorbed 
previously by the solid. When the ion content of a liquid is 
vanishingly small, we shall have but little ion adsorption and little 
electrical endosmose. On the other hand ‘pure’ water shows 
marked endosmose through many diaphragms. This is a case of 
preferential ion adsorption where the irons are produced both from 
the ionization of water itself and the solution of the solid, which is 
avery important matter in some cases (notably glass).*”7 Since the 
majority of solids are charged negatively against water, hydroxyl 
ions are probably adsorbed in preference to hydrogen ions. In 
discussing the potential of a solid against water originally pure, we 
must accordingly take two factors into account.*® 
(1) The specific adsorption capacity of the solid for hydrogen 
and hydroxy] ions produced by the dissociation of water. 
(2) The dissolution of the solid, which, though extremely slight 
in many cases, may produce ions that are strongly adsorbed. 
Hydrogen ions are often adsorbed preferentially from solutions 
containing them, especially from acids, though we have seen that the 
rule is by no means a general one. Experiments by the author have 
shown that aluminum, which is positive in dilute hydrochloric acid, 
is weakly negative im citrate acid where the equivalent selective 
adsorption of the citrate ion must be greater than that of the hydrogen 
ion. Moreover, we know that metal sulphides are peptized by 
hydrogen sulphide, an acid, yet the adsorbed ion is sulphur and not 
hydrogen ; for the particles in suspension are electro-negative.”® 
In general the same statement applies to the adsorption of 
_hydroxy! ions. Solids seem to have a somewhat greater adsorption 
affinity for hydroxyl ions than for hydrogen ions, though there are 
notable exceptions to this generalization. The general theory cover- 
ing the electrical endosmose of all liquids and solutions may be 
formulated as follows :— 
(1) Electrical endosmose depends upon the preferential or selec- 
tive adsorption of ions and is inflnenced onty by those ions which 
are adsorbed by the diaphragm. 
45 Cf, Bancroft, Jour. Phys. Chem., 20, 85 (1916). 
46 Wo. Ostwald, Grundriss der Kolloidchemie, 29 (1912).| 
47 Cf. Briggs, Bennett, and Pierson, Jowr. Phys. Chem., 22, 256 (1918); Zrans. 
Am. Electrochem. Soc., 31, 257 (1917). , 
48 In this connection, compare the statement of Schwerin, Brit. Pat. 10793 (1909) 
wherein he emphasises the importance of taking into account the dissolving of the 
diaphragm. ‘ : \ 
49 Winssinger, Bull. Soc. chem., Paris (3), 49, 452 (1888); Linder and Picton, Jour. 
Chem. Soc., 61, 116 (1892). 
In this connection compare the beautiful experiments of Lottermoser, Jowr. 
prakt. Chem. (2), 72, 53 (1905). 
20895 B2 
