549 



or ten) molecules. In most of the substances bound by carbon snch 



a liypotliesis lias no sense. Besides it does not become clear wlij 



the carbon works the belter as it is more finely divided; this must 



then be accounted for as a consequence of the easier diffusion. 



In 1907 Freundlich showed ') that the binding isotherm can be 



represented by the formula: 



1 



X " 



— = « . c 

 • m 



for not too great values of c (?h is the quantity of carbon, x the 

 substance bound by it, e the concentration of this substance in the 

 solution which is in equilibrium with the carbon, a and n are con- 

 stants). He showed that we had to do here with real equilibria 

 which are established within a very short time. The degree in which 

 a solid substance binds, varies greatlj' with ihe absorbed substance, 

 but is Utile dependent on the nature of Ihe solid phase. Freunduch 

 demonstrated that these facts become perhaps most easily compre- 

 hensible when il is assumed that the binding rests on surface 

 adsorption, on a becoming denser of the surface of the solid phase. 

 But in 1909 he himself does not exclude the possibility that the 

 phenomenon rests on the formation of a dissociable chemical bond 

 or a solid solution; he only calls these explanations "wesentlich 

 nnvorleilhafter" '). 



In course of time, however, in default of new arguments for the 

 other conceptions, this view has gained so many adherers that it 

 often makes the impression as if it were an established fact that 

 the sorption by carbon rests on a real surface-adsorption. 



In 1910 I succeeded') in showing that a deviation from the laws 

 of Henry and Nernst in solid solutions can have another cause than 

 the dissociation of the bound substance into molecules, viz. when 

 the mixing in solid solution is chiefly caused by the attraction 

 between the molecules of solvent and dissolved substance; whereas 

 in the ordinary diluted solutions the mixing is brought about parti- 

 cularly by the diffusion impulse (because mixing is a more probable 

 state, one that takes place with increase of entropy — also when 

 the attraction may be neglected). In this case the decrease of free 

 energy is about equal to the heat eflfect that takes place in the 



') Zeitschr. f. physik. Chemie 57, p. 385 (1907). 



') Kapillarcliemie, Iste Aufl. p 289, Akadein. Verlagsgesellschaft Leipzig 1909. 

 ') These Proc. Vol. Xlll, p. 958: Address at the Meeting of the Bunsen-Gesell- 

 schaft. Kiel, 1911; Gesetze der Quellung, Kolloidchem. Beihefte Bd 9. 



