On the Absorptio?i Spectra of Cobalt Salts. 259 



in hydrochloric acid gas, when in place of hydrochloric acid, an atmo- 

 sphere of carbon dioxide was used ; in all these cases the absorption 

 spectrum remained precisely the same : these experiments then show 

 that the above is the visible spectrum belonging to the true cobalt 

 chloride when in a fused state. The chloride fuses at so high a tem- 

 perature that the glass becomes red hot, and it has consequently not 

 been found possible to observe the absorption spectrum of the chloride 

 when in a liquid state. 



One point further: this spectrum is not in any way due to the 

 action of the fused chloride on the glass tube, for water entirely dis- 

 solves the salt, leaving the glass colourless. 



Whether in fig. 1 both bands are due to the same substance, and 

 whether the small absorption at the red end of the spectrum is acci- 

 dental or not, cannot be discussed at present. 



If bromide of cobalt be treated in a like manner to the chloride, it . 

 yields a very interesting spectrum (shown in fig. 2), for on comparing 

 it with the chloride spectrum a strong similarity between them is at 

 once seen ; but the bands of the bromide corresponding to those of the 

 chloride have approached the red end of the spectrum, the most re- 

 frangible band more than the other one. Owing to the opacity and 

 the ease with which the iodide of cobalt decomposes, its absorption 

 spectrum cannot be obtained in the same way. 



The strong power which the cobalt chloride has of combining with 

 water, and the change of colour and of spectrum which it undergoes, 

 is well known. The alteration thus brought about in the absorption 

 spectrum will be discussed further on ; but it should here be stated 

 that simply on drying the moist chloride without fusing it, a spectrum 

 is obtainable similar to that from the fused salt. The dehydration of 

 the ordinary crystallised chloride can be effected even at a compara- 

 tively low temperature ; if the heat be continued for a considerable 

 length of time, a temperature of 150° if continued for 4J hours 

 Tenders the salt dry. It is not easy to see the absorption spectrum of 

 the powder thus formed ; the plan found to answer best is to have the 

 salt in a fine state of division, suspend it in melted paraffin, and then 

 when cold cut slices of the paraffin of proper thickness. 



Assuming, then, that the above is the true spectrum of the solid 

 cobalt chloride, the next point was to study spectroscopically the 

 changes it underwent when acted on by different substances, or even 

 when merely its physical state was altered. 



The action of water on the chloride is complicated, and it is better 

 to consider first the simpler case of the chloride fused with potassic 

 chloride. When these two substances are mixed, especially if mixed 

 in atomic proportions, the whole mass fuses at a considerably lower 

 temperature than either salt alone ; it forms a greenish-blue mass, 

 and even when fused for some length of time in air it undergoes no 



