102 W. Gibbs and F, A, Genth 



The fact, that both the acid and neutral oxalo-sulphate of Pur- 

 pureocobalt contain two distinct acids, is also a very instructive 

 one, since it completes the txndlogj between the poljacid bases 

 and the poljbasic acids. A polybasic acid, as, for example, tar- 

 taric acid, may unite with two different bases at once, and we 

 now learn that a polyacid base may in like manner unite with 

 two different acids. Thus we have 



CamOio 1 1% , 5NH. . Co.03 | ^Oa^ , 5NH3 . Co.Oa | ^^O^ ^^ 



The other point to which we refer, is the peculiarity in the 

 constitution of Xanthocobalt, in which one equivalent of oxygen 

 in the secondary radical is not capable of replacement by chlo- 

 rine, so that Ave have for the chlorid of this base the formula 



while for the chlorid of Purpureocobalt, we have 



5NH3.Co2CLCh 



as 



5NH3 , C02O . Cb. 



The appearance of deutoxyd of nitrogen as a conjunct (Paar- 

 ling) is in itself well worthy of attention, and Xanthocobalt 

 forms, we believe, the only known instance in which this occurs. 

 It seems d priori probable, that iridium and rhodium bases cor- 

 responding to Xanthocobalt may be prepared by passing a cur- 

 rent of NOx into amraoniacal solutions of protosalts of those 

 metals, or into solutions of Claus' bases, and we have already 

 instituted experiments with these metals, the results of which 

 we hope hereafter to communicate. 



The theory which we have proposed for the ammonia- cobalt 

 bases has also been brought forward by Claus, and applied to his 

 rhodium and iridium compounds. Claus has extended the view 

 in question to the ammonia compounds of metalHc protoxyds, 

 and we conceive with advantage, in the case of those bases 

 which contain more than one equivalent of ammonia, as for in- 

 stance, the platinum, palladium, and iridium bases, having the 

 formulae 



2NH3 . PtO 2NH3 . PdO SNHa . IrO. 



The discovery of the biacid character of Purpureocobalt and 

 Xanthocobalt, in connection with the views which we have ex- 

 pressed with respect to the molecular structure of these bases, 

 has led us to extend tlie theory of conjugation to the ammonia- 

 platinum compounds. We consider these also as conjugate 



) 



lat their constitution may 

 than upon anv other vie^ 



1 



