492 Mr. W. C. M. Lewis : Experimental Investigation 



It is possible, therefore, that some irreversible effect of 

 the nature o£ gelatinization is the real cause of the discre- 

 pancy between observed and calculated values in the adsorption 

 of sodium glycocholate, Congo red, and methyl-orange. 



The Adsorption of Inorganic Salts from Aqueous 

 Solution, 



It has already been pointed out that for certain purely 

 inorganic salts, discrepancies are found between observed and 

 calculated values of the adsorption, but that the existence of 

 these discrepancies might be considered as not quite proved 

 owing to possibility of great experimental error. On the 

 whole, however, it is believed that there really does exist a 

 small discrepancy which requires explanation. Attention 

 might be drawn to the fact that in the cases measured the 

 metal, i. e. the cation, shows an excess, while in the case of 

 potassium chloride, and possibly also in barium chloride, the 

 chlorine (the anion) is much nearer to the calculated value. 

 This at once suggests some separation of ions, that is partial 

 selective adsorption due to some electrical effect. The value, 

 however, for the adsorption of the chlorine in copper 

 chloride is scarcely in accord with the preceding statement. 

 Granting that there is a tendency for selective ionic ad- 

 sorption, it is evident that the phenomenon cannot be com- 

 pletely covered by Gibbs's capillary expression, in the 

 deduction of which no allowance is made for any such 

 ionic separation. Of course for non-electrolytes, where no 

 ionic separation is possible, Gibbs's expression ought to 

 hold quite readily? and experiment confirms this, although 

 the determinations are unfortunately of a more qualitative 

 character than one could wish. In electrolytes, the sepa- 

 ration of cation from anion may possibly be regarded as 

 local electrolysis at the surface of the oil — that in fact we 

 are dealing with a cell of small dimensions the poles of 

 whicn are the two homogeneous masses, and the hetero- 

 geneous layer separating the masses functions as the " elec- 

 trolytic medium." If this is so, we must treat the question 

 as one of electrolytic conduction, but first of all it must be 

 shown that there exists an electrical potential -difference 

 between the electrodes, i. e. between the oil and the water. 

 Owing to the non-conducting nature of the oil one cannot 

 employ the usual methods, but by emulsifying the oil (i. e. 

 converting it into small particles suspended in water) and 

 placing the emulsion in an electrostatic field, it can be- 



