52 



Dr. W. J. Russell. 



[Nov. 18, 



On fusing cobalt chloride with potassic chloride, a greenish-blue 

 mass is formed, which gives a spectrum entirely different from that of 

 the chloride when alone. Judging from the fusing-point of the mix- 

 ture being lower than that of the components, and the cobalt salt not 

 decomposing in this mixture on fusion in contact with air, it seemed 

 natural to suppose that a new compound had been formed, and that it 

 gave rise to the new spectrum. Further experiment showed, how- 

 ever, that this is not the case for other solid chlorides, such as of 

 sodium and zinc give with cobalt chloride the same spectrum ; and 

 liquids in which the cobalt chloride easily dissolves, such as ordinary or 

 amylic alcohol, the saline ethers, glycerine, and hydrochloric acid, also 

 give this same spectrum ; in fact, this spectrum is produced whenever 

 cobalt chloride dissolves freely in any menstruum without definitely 

 combining with it. A careful set of experiments were made in the 

 case of the solid chlorides, to exclude the presence of water, for it 

 was possible that the spectrum in all the above cases might be due to 

 a trace of water, which, by its combining with the cobalt chloride, 

 formed in every instance the spectrum-giving body. 



The bromide and iodide of cobalt, when fused with potassium 

 bromide and potassium iodide respectively, give results corresponding 

 to those of the chloride, but the bands in the spectrum of the 

 bromide, and still more so those of the iodide, are nearer to the red 

 than the corresponding bands of the chloride. The action of heat 

 and of water on the bodies producing these spectra is discussed, and it 

 is pointed out how the definite compound with zinc was indicated by 

 the spectrum. The action of liquids which easily dissolve the cobalt 

 chloride is next described, and as all give the same spectrum, and 

 this spectrum is identical with that obtained with the fused chlorides, 

 the conclusion drawn is, as before stated, that this spectrum must 

 be that of the cobalt chloride, only, owing to solution, it is in a mole- 

 cular state, different from that obtained on fusing this salt alone. 

 This spectrum, when obtained in the hydrochloric acid solution, is 

 remarkable for its persistence under varying circumstances, and for its 

 being a reaction of great delicacy. 



Hydrochloric acid as a solvent for the cobalt chloride differs in one 

 respect from all the other solvents which have as yet been examined, 

 namely, that, whether much or little cobalt be dissolved in it, the spec- 

 trum is the same ; whereas with dry alcohol, for instance, a saturated, 

 or nearly saturated, solution gives the spectrum above mentioned, but 

 a dilute solution, one containing about 20 grms. of the chloride in 100 

 cub. centims. of alcohol, gives a spectrum somewhat different : a new 

 band appears and others which were present fade out : if this dilution 

 be carried on still further, so that only about 0*008 grm. of the 

 chloride be present in 100 cub. centims. of alcohol, an entirely different 

 spectrum is obtained, but on carrying the dilution beyond this no 



