On the Absorption Spectra of Cobalt Salts. 263 



*is, however, this point of interest about it, that it gives the spectrum 

 of the simple fused chloride ; this may arise either from some dry 

 chloride having escaped the action of the calcium chloride, or what 

 seems more probable, that it arises from the action of the high tem- 

 perature on the cobalt. 



There are a few liquids which appear to dissolve the anhydrous 

 cobalt chloride without combining with it, and it was of much 

 interest to see what spectra they would give. The liquids ex- 

 perimented with were hydrochloric acid, methylic, ethylic, and 

 amylic alcohols, the saline ethers, amylene, glycerine, and glacial 

 acetic acid. 



First, with regard to the action of the ordinary strong hydro- 

 chloric acid. As is well known, cobalt chloride dissolves readily in 

 this acid, yielding an intensely blue solution. The spectrum which 

 this liquid gives is shown at fig. 6, and it will at once be seen to be 

 similar to the spectrum obtained from the fused chloride (fig. 3), 

 the two bands in the orange now being a little nearer the blue. The 

 similarity of these two spectra is remarkably complete, for on 

 increasing gradually the thickness of the solution the bands in the 

 red are gradually hidden, and when the absorption has crept up to 

 600, then a band at 540 to 525 becomes visible. The spectrum is 

 remarkable for its permanence, if the solution be either weak or 

 strong, hot or cold, the spectrum remains unchanged. This agrees 

 with what one would naturally expect, that, under these circum- 

 stances, the cobalt remains as chloride. Moreover, from the similarity 

 of this spectrum with that obtained with the fused chlorides, the 

 inevitable conclusion is that the cobalt is in the same state in both 

 cases. 



This spectrum of cobalt chloride is very characteristic in its 

 appearance and very easily produced ; without special care and with a 

 common spectroscope "0005 grm. of cobalt chloride can easily be 

 detected, and even the presence of other coloured chlorides does not 

 interfere with it ; for instance, even the presence of iron, nickel, 

 chromium, manganese, does not hide the characteristic appearance 

 produced even by the above small amount of cobalt chloride. 



Supposing the 532 band be produced by a specific compound, it 

 seemed likely that at least a rough estimate of the amount of that 

 body, present under different circumstances, might be judged, by 

 bringing always the absorption in the red up to the same point and 

 comparing then the intensity of this 532 band ; practically, however, 

 this did not lead to results of any value. 



It is, however, important to note that the spectrum of the f used 

 chloride is never seen without the 532 band ; and that in the other 

 two cases already described, this band is only visible when a con- 

 siderable amount of the cobalt salt is present. It seemed, therefore, 



vol. XXXII. T 



