196 W. Gibbs on the Hexatomic compounds of Cobalt. 
The potassium salt is also formed, as I shall show, in various 
other cases; the similarity of some of its reactions to those of a 
solution of Co,(NO,),.Nag in sodic nitrite for a long time waietod 
me; but its relations to salts of silver, mercury and thallium 
enable us to recognize its presence with absolute certainty. 
The salt does not enter into combination with iodine. 
XANTHOCOBALT. 
9. Genth and I have shown in our seis that the salts of 
xanthocobalt may be formed either directly by the action of 
nitrous acid vapors upon salts of cobalt, or by the action of the 
same acid upon salts of purpureocobalt and roseocobalt, in each 
case in the presence of free ammonia. I propose now to give 
the results of a more detailed study of the subject. 
With respect to the constitution of this class of salts, I may 
remark, in the first place, that Genth and I left it undecided 
whether the salts in question contain NO or NO,, pointing out 
the fact that the analyses do not decide in hock of either view, 
and adopting the former provisionally. Braun first prove 
snelsiliveky that ig salts of xanthocobalt contain NO,, an 
this view has since been generally adopted. I have hie 
shown (§ 1) that en cobaltic chloride, CoCl,, is mixed wi 
ammonia and ammonic nitrite and nitrate, the solution beers 
oxygen from the air, while the nitrate of the octamin series, 
Co,(N Hs),(N' O,),(NO,)a, i is formed. I have not observed in this 
reaction the formation of a salt of xanthocobalt. If present at 
all, such salts must be in very small relative quantity. Gent 
and I have shown, on the other hand, that when the red gases 
resulting from the action of nitric acid upon starch, sawdust 
or arsenous oxide are passed into solutions of cobaltic salts in 
presence of an excess of ammonia, salts of xanthocobalt are 
formed in a very short time, and in large quantity. 
we consider > red gas to consist of hyponitric oxide, 
N,O,, we may hav 
16S, HONEA, O,= Cols) sol NOs)aNUolacs 
salts of this base among the products of the reaction. In one 
case, however, in which I employed cobaltic sulphate and added 
so large a quantity of ammonic sulphate that the solution gave 
no precipitate with ammonia, I obtained a very large relative. 
quantity of Erdmann’s salt Co,(NH,),NO.). In other cases in 
which cobaltic chloride was present I detected crystals of the 
chloronitrate Co(NHs)io(N Os)(NO, )zCl The solutions after 
the action of the red gases also contain small quantities of the 
ammonia-cobalt-nitrite of Gorge CoN (NE) UNONEN Hyg)s, 
as well as ammonic nitrite and n 
¥ 
