24 



However, if we are to give up the theory that the phenom- 

 enon is due to ionization, how sliall we explain the change of color 

 when other chlorides are added to cupric and cobaltous chlorides 

 and when other ])romides are added to cui)ric bromide? 



In order to find what the actual facts in the case are, solutions 

 were prepared of a large number of acids and salts, comprising 

 bromides, chlorides, nitrates, and sulphates, all of the same 

 strength, namely, thrice molecular-normal. These were made up 

 in a number of ways and care was taken only that they should 

 not be in error by more than 1 per cent. To equal volumes of 

 these solutions the same small quantity of cupric Ijromide was 

 added. The resulting solutions ranged from blue to green, except 

 in the cases of the four bromides used, in which the solutions were 

 all of an olive brown. Similar experiments were tried with copper 

 and cobalt chlorides, using all the solutions except the bromides. 

 The cupric chloride was green in the chloride solutions, blue in all 

 the others. The cobaltous chloride was a bluish red in the 

 chlorides, a pure red in the others. 



If the differences of color are due to differences in hydration, we 

 should, at first thought, expect foreign salts to have no effect on 

 the color, except as through their own hydration they might 

 deprive the colored salt of the water coml)ined with it. ^^"e 

 should therefore ex})ect that the strongest dehydrating agents 

 would have the most effect on the color, rather than substances 

 with a common ion. However, as a matter of fact, such 

 hydroscopic substances as sulphuric acid and calcium nitrate 

 seem to have little if any greater effect than other salts. 



Let us examine the matter more closely, and suppose that 

 during the dilution one or several reactions of the following type 



take place: CuBi\ + nI{fi = Ci\BY,(Ji,0)„ = Cn(lCo)„ + 2BY. 



As a matter of fact, the color change ma}^ be due chiefly not to 

 a change from the anhydrous salt to a hydrated one, but from one 

 with less water to one with more, or the ions may exist in any 

 degree of hydration, but whatever the actual reactions are, we may 

 assume the al)Ove reaction merely for the sake of concreteness. 

 Let us further assume that the brown color is due to tlie CuBr.^ 



and the blue to the CuBr,(H,0)„ and the Cu(H,0)„. 



Anything which would make the reactions run from right to left 

 would cause the brown color to increase at the expense of the blue. 



