27 



predicted fi-oiu its urcat tVeezing-})()int depivpsion, wliidi is at least 

 90°, it gave a dee}) olive-l)rown color. 



All these experiments were repeated, using cobaltous chloride in 

 place of ciipric. The color differences here are less marked. The 

 solutions of the chlorides of the first grou]) all had practically the 

 san.e red color with a tinge of blue. But in the second group the 

 solutions Avere a more decided i)ur])le and arranged themselves in 

 the same order as before — that is, BeCl, gave the solution with most 

 blue in it. AlCl, gave a deep purjjle, and a four-times-normal 

 solution of AICI3 gave a i)ure blue. 



All these facts speak for the truth of the above explanation of the 

 color changes, and give strong evidence of the hydration of the 

 three substances CoCl,, CUCI2, and CuBr^. It may be remarked 

 that these substances, according to the work of Biltz,^ give un- 

 usually large depressions of the freezing point. 



I believe that a quantitative study of these color changes by 

 means of a spectro-radiometer might throw a great deal of light on 

 the Avhole question of hydration, although any attempt to determine 

 the formulas and the amounts of hydrates in solution is rendered 

 difficult by the probability of the simultaneous existence as a rule 

 of several different hydrates in each solution. 



'Z,'if. PInis. Clwm. 40, 198 (1902). 

 O 



