AMMONIA-COBALT BASES. 49 



The salts of Xanthocobalt have a dark sherry-wine or brown-yellow color. 

 They are rather more soluble, both in hot and cold water than the salts of Roseo- 

 cobalt, Purpureocobalt, or Luteocobalt, the solutions when dilute have a yellow 

 color; when concentrated they are dark-brown yellow. Solutions of these salts 

 are decomposed by boiling, though sometimes with difficulty; in some cases, how- 

 ever, a temperature much below the boiling point produces an incipient decompo- 

 sition after a short time, so that great care must be taken in evaporating for the 

 purpose of crystallization. During the decomposition by boiling, ammonia is given 

 off, while a heavy black or dark-brown powder is thrown down. The addition of 

 a mineral acid prevents the decomposition, but produces, even when in very small 

 quantity, a chemical change of a different kind, deutoxide of nitrogen being given 

 off, while a salt of Purpureocobalt is formed, which remains mixed with the unde- 

 composed salt of Xanthocobalt, and which it is difficult to separate. A few drops 

 of acetic acid, however, added to a solution of a salt of Xanthocobalt so as to 

 produce a slightly acid reaction, will usually suffice to prevent decomposition, if 

 the solution be evaporated at a gentle heat. The dry salts of Xanthocobalt heated 

 in a porcelain crucible are easily decomposed, with copious evolution of red vapors, 

 and afterwards of ammonia, while a black powder remains. We have not been 

 able to detect any new ammonia-cobalt bases among the products of the decompo- 

 sition of solutions of these salts, by simple boiling or heating in the dry way. 



Xanthocobalt differs from the other ammonia-cobalt bases by containing deutoxide 

 of nitrogen as a couplet in addition to ammonia. Considering it as a primary 

 radical we attribute to it the formula NOo.SNHg.Coo, and it is evident that it differs 

 from Eoseocobalt and Purpureocobalt only by one equivalent of deutoxide of nitro- 

 gen. The passage of the salts of these two bases into salts of Xanthocobalt by 

 simple absorption of N0 2 is thus easily explained, as is the decomposition of the 

 salts of Xanthocobalt into N0 3 and salts of Purpureocobalt. 



It is, however, proper to remark that, inasmuch as the salts of Xanthocobalt 

 which we have examined in all cases appear to contain water in the proportion of 

 at least one equivalent, it is possible that the view we have taken of their constitu- 

 tion may not be precisely accurate, and that they may contain N0 3 as a couplet in 

 place of N0 2 . This will be evident from a simple inspection of the formulae of the 



When this is heated with sulphuric acid till the excess of acid is expelled, there remains a mixture of 

 sulphates having the formula 



2CoO,S0 3 +3KO,S0 3 . 



We, therefore, dry the salt carefully, burn the filter upon the cover of a deep platinum crucible, and 

 bring the ashes and salt into the crucible itself. A few drops of sulphuric acid are then added, so as 

 to cover the mass, and the whole is then treated precisely as we described on page 4, though in this 

 case a higher heat is to be used. From the weight of the mixed sulphates the quantity of cobalt is 

 easily calculated. This method gives very accurate results, because the cobalt forms only about 1 4 per 

 cent, of the substance weighed. In an experiment made to test the accuracy of Stromeyer's formula 

 0.6611 grs. gave 0.6431 grs. mixed sulphates of cobalt and potash = 13.66 per cent, cobalt. 

 The formula requires 13.63 per cent. 

 1 



