W. Gibbs on the Hexatomic compounds of Cobalt. 289 
after I added PtCl,Na, toa solution of sulphate of xantho- 
cobalt, hoping to obtain a salt with the formula Co,(NH,),, 
(NO,),(SO,)Cl, (PtCl,), analogous to a platinum salt of roseo- 
cobalt, which I shall hereafter describe, and which has the for- 
mula Co,(NH,), ,(SO,),PtCl,. The beautiful crystalline pre- 
cipitate formed proved to be only the salt Co, (NHL), (NO,), 
Cl, PtCl,+OH,, described in the first part of this memoir. 
03882 gr. gave 0°1612 gr. Co+Pt=4152. The formula re- 
quires 41°39 per cent. 
Nitrite of xanthocobalt.—W hen argentic nitrite is boiled with 
a solution of chloride of purpureocobalt, the liquid soon loses 
its fine violet color, and assumes the wine-yellow tint of the 
salts of xanthocobalt. The filtrate from the argentic chloride 
gave, on careful evaporation, two distinct salts—a salt in beau- 
tiful scaly crystals, and one in octahedral crystals. The two 
salts were separated by crystallization. Of the scaly salt 
0°2854 gr. gave 0°2286 gr. SO,Co+SO0,Ag,—79°97 per cent. 
The formula of the ammonia-cobalt-nitrite, Co,(NH,),(NO,), 
Ag, requires 80°75 per cent, and the salt was easily identified, 
by its appearance and properties, with the silver salt of Erd- 
ann’s series. he octahedral salt was rather difficult to 
obtain perfectly pure by this method, I had recourse to the 
decomposition of sulphate of roseocobalt by baric nitrite. A 
solution of the last named salt is to be added to one of the sul- 
phate as long as a precipitate is formed. The sherry-wine- 
colored filtrate is then to be cautiously evaporated, when fine 
dark wine-colored octahedral crystals form. Of these crystals 
0°4750 gr. gave 0°2303 gr. SO,Co=18°46 per cent cobalt. 
071220 gr. gave 0°0594 gr. SO, Co=18°54 per cent cobalt. 
0°3129 gr. gave 0°0403 gr. water, when heated to 140° C, =12'87 
per cent. 
9°4289 gr. gave 071141 gr. ammonia—26°60 per cent. 
The formula Co,(NH,), ,(NO,),+40H, requires 
culated. Found. 
‘obal 2 18°55 18°46 18°54 
Ammonia, 10 26°72 6°60 
ater, 4 «1182 12°87 
The percentage of water in the analysis is too high, and would 
seem to show that a slight decomposition of the salt had taken 
place. I attempted to determine the percentage of NO, in this 
salt by titrition with potassic hypermanganate, but though the 
analyses were made with the greatest care, I obtained as a mean 
of three determinations, agreeing well together, only 11-24 tied 
cent, which would correspond to less than two atoms. Tn other: 
cases also I found that the method could not be employed. 
Am. Jour. Sc1.—Tarp Saums, Vox. VIII, No. 46 —Oct., 1874. 
