194 



tals of small size. It is generally obtained, however, as a crystalline 

 powder of a fine pink colour, as it is usually rapidly formed, and 

 deposited too quickly to form regular crystals. 



The chemists who have analysed the red or pink salt are not agreed 

 as to its composition, for while Claudet found it to contain no oxy- 

 gen and no water, Fremy admits 1 eq. of water, and Genth con- 

 siders it as a compound of sesquioxide of cobalt, ammonia, and chlo- 

 rine. I have made a number of analyses of this salt, prepared in 

 different ways, and when it has been slowly ignited in a current of 

 hydrogen, to determine the cobalt which is left in the metallic state, 

 I have not in any case obtained a trace of water. Consequently the 

 salt cannot have the formula given to it by either Genth or Fremy. 

 As that of Genth is absolutely erroneous, I shall give here the em- 

 pirical formulse of Fremy and Claudet, with my own results. 



It is very difficult to form any distinct idea of the rational for- 

 mula, whichever empirical one we adopt. The most interesting point 

 is this, that from this compound analogous ones with oxygen acids 

 may be formed, and that from these, by the action of alkalies, a base 

 may be separated, although it has not yet been isolated in a state of 

 purity, which appears to consist of ammomsb plus some oxide of cobalt. 

 If such bases exist, they will probably, like other oxidised bases, 

 yield, with hydrochloric acid, water and chlorides, and thus our red 

 salt would be the chloride of the radical, which, with oxygen, forms 

 the base in the oxidised salts. But we must not dwell on possibili- 

 ties ; and my object is to shew, first, that the red compound does not, 

 as Fremy states, contain oxygen (at least that which I have examined), 

 and that before we can speak with confidence as to its true formula, 

 we must have more certainty as to the empirical one. 



With regard to the yellow salt, this, according to Fr^my, contains 



