26 PROCEEDINGS OF THE AMERICAN ACADEMY 



pureocobalt by simple heating. I can recall no single instance 

 in which a true hydrate behaves in a similar manner. Nitrate of 

 roseocobalt, when heated in solution with a little free nitric acid, 

 undergoes a similar conversion into nitrate of purpureocobalt. 



3. Sulphate, chromate, oxalate, &c., of roseocobalt in solution yield, by 



double decomposition in the cold with baric chloride and nitrate, 

 salts of roseocobalt only. Is a mere state of hydration trans- 

 mitted from salt to salt ? 



4. According to the determinations of F, Rose,* one part by weight 



of chloride of roseocobalt requires at 10° C. 4*8 parts of water for 

 solution. At the same temperature, one part of chloride of pur- 

 pureocobalt requires 287 parts of water for solution. Rose re- 

 marks that the identity of chloride of roseocobalt and chloride of 

 purpureocobalt can only be maintained by admitting that the 

 same salt can exist both in the hydrous and anhydrous condition 

 m a solution at the same temperature, an assumption directly 

 opposed to the numerous observations of Riidorft' and Wiillner 

 on the tension of the vapor of aqueous solutions. 



5. Solutions of the two chlorides form with the same reagents in 



many cases different salts. Thus, chloride of roseocobalt foruis 



Co,(NH3)ip,+2PtCl,-|-50H2 with platinic chloride, 



Co2(NH3)ioCl6+2AuCl34-20H2 with auro-chloride of sodium, 



Co2(NH3)jy(C^O^)3-|-60H^ with ammonic oxalate. 



With the same reagents respectively, chloride of purpureocobalt 

 gives, 



Co,(NH3),„Cl,+2PtCl„ 



Co,(NH3),,Cl,+2AuCl3, 



Co,(NH3),„Cl,(C,Oj3. 



C. The greater number of the salts of the roseocobalt series contain 

 water of crystallization. The greater number of the salts of pur- 

 pureocobalt are anhydrous. If it be replied that this statement 

 involves a petitio principii, I reply tliat the presence or absence 

 of water of crystallization is, in most cases at least, coexistent 

 with other jjroperties tending to establish a clear distinction be- 

 tween tiie two classes of salts. 



* Untersuchungen iiber ammoniakalische Kobalt-Verbindungen. Heidel- 

 berg, 1871, p. 47. 



