On the Absorption Spectra of Cobalt Salts. 267 



of the 10 per cent, solution, and 16 inches of 1 per cent, solution was 

 compared to 1^- inch of 10 per cent, solution. So that in all these 

 dilute solutions, as far as the spectroscope can tell us, there are the 

 same compounds present. With stronger solutions this is not the case ; 

 for if the solution contains, say, from 20 to 30 grms. in 100 cub. 

 centims. of water, and a thin layer of it be compared with a thick 

 layer of a more dilute solution, the spectra will be seen to be essen- 

 tially different. This new spectrum, which proves to be a very in- 

 teresting one, is best seen when a saturated or nearly saturated 

 solution of cobalt chloride is used (100 cub. centims. at 16° dissolve 

 32 grms. of cobalt chloride), then there are two bands — one at 625 

 and the other at 610 — and an absorption stopping sharply at 652. 

 Evidently we have again the spectrum of the dissolved cobalt chloride, 

 a spectrum which has already been shown to be perfectly indepen- 

 dent of water, consequently it appears that the anhydrous cobalt 

 chloride is capable of existing in an aqueous solution. 



If a solution saturated at ordinary temperatures with the chloride of 

 cobalt be diluted with half its volume of water, the bands at 625 and 

 610 disappear ; but if to this solution, or even to one more dilute, any 

 substance be added capable of combining with water, such as sul- 

 phuric acid, calcium chloride, or even ammonium or sodium chloride, 

 the bands at 625 and 610 again appear. 



Heat acts on these dilute solutions in the same kind of way. If, for 

 instance, this saturated solution, diluted with half its bulk of water, 

 be heated up gradually at 49°, the 625 band first becomes visible, and 

 then the 610 band. As the temperature increases the absorption in 

 the green diminishes, and at the boiling point you have a spectrum 

 very similar to that obtained by fusing together chloride of cobalt 

 and chloride of potassium. 



The reverse action also occurs. A solution giving strongly the 

 bands at ordinary temperature ceases to show them when cooled. In 

 both cases as the solution returns to its original temperature so does 

 the original spectroscopic appearance return. In many cases the 

 spectroscopic variations by temperature are very delicate and very 

 striking in appearance. Any change tending to break up the hydrates 

 seems always to diminish the amount of absorption taking place 

 towards the blue end of the spectrum. The well known change of 

 colour of strong cobalt solution on heating bears this out. Pass 

 hydrochloric acid gas, for instance, into a dilute cobalt solution, and 

 the gradual formation, and the chloride spectrum and disappearance of 

 the absorption in the green and blue, is well shown. 



In proportion as one compound, the hydrate, ceases to exist, so the 

 other, the anhydrous compound, is apparently produced. The ordinary 

 crystals of chloride of cobalt contain six molecules of water, and they 

 give a spectrum shown at fig. 9. There are certain cases in which the 



