266 



Dr. W. J. Russell. 



evidently arose from a trace of water adhering most persistently to 

 the alcohol. If liquids be used in which the cobalt chloride is not so 

 soluble, dry acetic acid for instance, it will be found that only two 

 spectra are produceable, corresponding to those in the more dilute 

 alcoholic solutions ; and with a liquid in which the cobalt chloride is 

 still less soluble, such as dry ether, there is only one spectrum 

 obtainable. And lastly, with such a liquid as tetrachloride of carbon 

 or benzol, in which the cobalt chloride is absolutely insoluble, the 

 spectrum, which is faintly visible when the liquid and the salt, in a 

 fine state of division, are shaken up together, is that of the pure fused 

 chloride. The action of heat on these solutions is two-fold : first, it 

 increases the amount of colour ; this arises, no doubt, to a considerable 

 extent, if not altogether, from an absolute increase of the amount of 

 anhydrous cobalt chloride present, for if a mere trace of water be 

 present, on adding the blue cobalt salt the first action is for it to com- 

 bine with the water, and to produce a pink and comparatively colour- 

 less solution ; this hydrate, as will be shown further on, is decomposed 

 on rise of temperature, giving the blue chloride. The second action 

 of beat is to transform the saturated solution spectrum or third 

 stage, into the intermediate or second stage, and the intermediate 

 one into the first — the one given by most dilute solutions. Again, 

 in the spectrum of the third stage, the 532 band is visible, but with 

 the alcoholic solution the absorption has to extend nearer to the blue 

 to about 575 instead of only 600, as with the hydrochloric acid solu- 

 tion, before this band becomes visible. This seems to indicate that 

 less of the body producing it is present in the alcoholic than in the 

 hydrochloric acid solution. 



When water is the solvent for the cobalt chloride, the phenomenon 

 is somewhat more complicated, as direct combination occurs and 

 hydrates are formed. The colour of the dilute solution, as is well 

 known, is pink. If this be examined with the spectroscope no 

 sharp bands are visible ; there is, however, very considerable absorp- 

 tion at the blue end of the spectrum, but with certain concentra- 

 tion and depth of solution, it is easy to distinguish a wide absorp- 

 tion band extending from 550 to 485 (fig. 16), and beyond this 

 blue rays are comparatively freely transmitted. This band, how- 

 ever, is never sharply defined, but shades off on both sides ; it un- 

 doubtedly belongs to certain hydrates of the cobalt chloride. To 

 study the formation of these hydrates, solutions containing different 

 amounts of cobalt chloride were prepared and tested as follows : first, 

 a solution containing only O'l grm. of cobalt chloride in 100 cub. 

 centims. of water was made, and the spectrum it gave was compared 

 with that of a solution containing 100 times as much cobalt chloride. 

 These solutions were found to give identical spectra. Again, 

 15 inches of a %\ per cent, solution gave the same spectra as 4^ inches 



