23 



salt, the undissociated molecules and the ions hein^j; supposed 

 to have different colors. As far as one can judge by the eye, 

 all the intermediate colors which occur during the dilution of 

 these solutions in every case can be considered mixtures in different 

 proportions of two colors. In the case of cobaltous chloride the 

 two are blue and red, of cupric bromide yellowish Ih'owu and l)lue. 

 The theory is that the undissociated cupric l)romide, for exami)le, 

 is brown, the cupric ion lilue. As the solution is diluted the 

 degree of dissociation increases until finally all trace of brown 

 disappears and the color is the pure l)lue of cupric ion. This 

 theory is further borne out l)y the fact that tlic aildition of a 

 bromide, which would be expected to jiush ])ack the di^sociation 

 of cupric l)romide, does in fact turn a blue solution to green or 

 even brown. 



1 was first led to distrust the adequacy of this ex])]anation by a 

 research of Prof. H. M. (ioodwin, read before the Boston Physico- 

 Chemical Club, and as far as I know not yet published, in which 

 he showed, liy a quantitative study of the alisorption spectra of 

 copper salts in solution, that the greatest change in color which 

 occurs on dilution does not conic where the change in dissociation 

 is the greatest, but at a much liigher concentration, and that in 

 some cases little change in the al)sorjjtion spectrum accompanies 

 the further dilution of a solution which is still to a large extent 

 undissociated. 



These experiments force us to seek another exjjlanation of tlie 

 color change, and the simplest is that of Wiedemann previously 

 mentioned, according to which the change of color is due to cliange 

 in the degree of hydration. There is much evidence in favor 

 of this view. Anhydrous cobalt compounds are in general blui — 

 for exam])le, the borate and the silicate. The solid chloricK' of 

 cobalt changes from red to blue on losing its water of crystallization. 

 Unquestionably hydration is almost universally accompanied by an 

 evolution of heat, wherefore by the principle of I^e Chatelier the 

 degree of hydration in solution should decrease with increasing 

 temperature, and in fact a red solution of cobaltous chloride turns 

 l)lue on heating. According to the otiiei- theory this change of 

 color would have to be attributeil to ;ui enormous diminution 

 in electrolytic dissociation. The experiments of Isaachsen' render 

 this improbable. 



^Xrll. Pli>,s. riuw. 8. Ho (1891). 



