On the A bsorption Spectra of Cobalt Salts. 



271 



instance, the sulphate, nitrate, and acetate all act in this manner. An 

 interesting reaction is visible with the cobalt carbonate. It is stated 

 in works on chemistry that on adding sodic or potassic carbonate to a 

 cobaltous solution it is not the normal carbonate, but a mixture of the 

 carbonate and hydrated cobaltous oxide, that is formed.* If now this 

 precipitation be made in front of the spectroscope, the changes which 

 go on are perfectly visible. On the addition of the alkaline carbonate 

 a precipitate forms which gives no bands, but after a very short time, 

 perhaps half a minute or less, the bands of the oxide become distinctly 

 visible ; apparently then it is a pure carbonate which is thrown down, 

 but that it rapidly, by the action of the water, undergoes decomposi- 

 tion, and oxide is formed. 



Under special conditions, however, some of the oxygen salts may 

 be made to give band spectra, for instance, on fusing chloride, oxide, 

 or phosphate of cobalt with microcosmic salt and allowing it to cool, a 

 pink-coloured glass is obtained, which gives a spectrum consisting of 

 bands with much general absorption. Since all three of these cobalt 

 compounds give under these circumstances the same spectrum, it 

 apparently leaves little doubt that it is the spectrum of a phosphate of 

 cobalt. If this pink bead be heated, it undergoes an interesting change, 

 the absorption in the blue entirely disappears, and you get a spectrum 

 identical with that of the oxide ; for instance, that of an ordinary 

 borax or glass bead (fig. 10). As this phosphate glass cools it again 

 becomes pink, and the banded spectrum returns ; evidently the phos- 

 phate is dissociated, recombination taking place as the temperature 

 falls. When sodic phosphate is added to a cobaltous solution, a 

 precipitate is formed which is soluble in excess of the cobalt salt ; on 

 continuing the addition of the phosphate, first a permanent pink 

 precipitate is obtained, and afterwards when the phosphate is in excess 

 the precipitate is of a blue colour. If the cobalt salt be added to the 

 disodic phosphate, the blue precipitate forms at once. Now, on 

 examining this blue precipitate spectroscopically, it is at once evident 

 that oxide of cobalt must be present, for there is exactly the same 

 spectrum as is given by the precipitate formed on adding caustic 

 alkali to a solution of cobalt. The pink precipitate gives a faint band 

 about 590, but the bands seen in the fused phosphate were not visible 

 in the precipitate. 



The foregoing account of these different spectra is but very imper- 

 fect ; many most interesting questions with regard to the nature and 

 meaning of the spectra have to be left for the present unsolved for 

 want of sufficient data. Further experiments with other compounds 

 are required, and above all it would be of the highest interest to 

 investigate fully the parts of the spectrum not visible to the eye ; 



* Miller's " Chemistry," p. 536. 



