10 HE SEARCHES ON THE 



solution yields, on cooling, small but brilliant crystals of the latter chloride ; in fact 

 even solution in warm water converts a portion of chloride of Eoseocobalt into 

 chloride of Purpureocobalt, as may easily be observed, by the change of color. 

 This transformation is, however, far more striking when a solution of chloride of 

 Eoseocobalt is boiled with a little chlorhydric acid : the solution speedily changes 

 its color from a dull-red. to a beautiful violet-red, and on cooling deposits an 

 abundant crystallization of chloride of Purpureocobalt. There is in this case a 

 direct conversion of Eoseocobalt into Purpureocobalt, an isomeric radical; the 

 reactions of the violet-reel solution being entirely different from those of the same 

 solution previous to boiling with acid, except in one or two particulars, to be 

 pointed out hereafter. The dry chloride of Eoseocobalt is also slowly converted 

 into chloride of Purpureocobalt by keeping, changing its color to violet-red; in 

 this case, however, the change is not complete, even after a long time, unless heat 

 be applied. The formula of the chloride of Purpureocobalt, as we shall hereafter 



show, is 



5NH 3 .Co 2 Cl 3 . 



This differs from that of the chloride of Eoseocobalt only b} 7 containing no water 

 of crystallization. The change which takes place in the conversion of one chloride 

 into the other does not, however, consist in the mere loss of water. As we shall 

 show, the chloride of Eoseocobalt corresponds to a triacid oxide, while that of Pur- 

 pureocobalt yields a biacid oxide. It is to be carefully borne in mind, that the 

 substance which we have called chloride of Eoseocobalt is not the chloride of Eoseo- 

 cobaltiaque of Fremy, Claudet, and other chemists who have studied the subject. To 

 the chloride described by Fr6my under the name of chloride of Eoseocobaltiaque we 

 have given the name of chloride of Purpureocobalt. The necessity of this change 

 of name has arisen from the fact that hitherto two different bases have been con- 

 founded, the chloride of Purpureocobalt having been considered as the chloride 

 corresponding to the sulphate and nitrate of Eoseocobalt. 



The chloride of Eoseocobalt is dichrous, the ordinary being paler than the extra- 

 ordinary image ; both are rose-red, with a faint brownish orange tint. 



Chloride of Eoseocobalt, as already mentioned, has the formula 



5NH 3 .Co 2 Cl 3 +2HO 



as the following analyses show : 



0.7291 grs. gave 0.4235 grs. of sulphate of cobalt = 22.10 per cent, of cobalt. 

 0.6247 grs. gave 0.3619 grs. " " =21.99 " " 



0.9834 grs. gave 0.5742 grs. chloride of silver = 39.57 " chlorine. 



1.8112 grs. gave 2.9095 grs. " " =39.71 " " 



2.2390 grs. gave 1.2829 grs.. water = 6.37 " hydrogen. 



1.4993 grs. gave 0.8784 grs. " = 6.50 " 



1.2235 grs. gave 282 c. c. nitrogen at 22.°5 C. and 766 mm .82 (at 23° C.) = 254.91 c. c. at 0° 

 and 760 mm = 26.16 per cent. 



The formula requires — 





Calculated. 



Found. 



Cobalt . 



. 21.97 



21.99 22.10 



Chlorine . 



. 39.66 



39.57 39.71 



Hydrogen 



6.33 



6 37 6.50 



Nitrogen 



. 26.08 



20.16 — 



