OF ARTS AND SCIENCES. 9: 



3.Co,(NH3)io(N03)e+3Co,(NO,),2Na,=lCo,(NH3)„XNO,),^3 



I shall allude to its composition again, when speaking of the metameric 

 compounds of the cobaltamines. 



The salt described is not the only one which is formed by the action 

 of a solution of Co^(NOJi2Nay on nitrate of purpureocobalt. In sev- 

 eral experiments I obtained an orange-red granular salt readily soluble 

 in cold water. The solution of this salt gave with nitrate of luLeo- 

 cubalt a beautiful crystalline precipitate of 



|Co,(NH3),,^Co,(NO,),,^ 



With thallous nitrate it gave the characteristic scarlet salt Co^(N02)i2 

 Tl„. On the other hand, it gave all the reactions of the salts of 

 xanthocobalt with great distinctness, including the highly characteris- 

 tic crystalline ferrocyanide. This salt could in no way be distinguished 

 from one obtained by the action of Co2(N02)i2Nag upon sulphate of 

 roseocobalt which, as I shall show, has the formula • — 



^Co2(NH3),,nCo2(NO,),2^ or Co,(NH3),„(NO,)„-f C.o^CNO^)^. 



In acting upon solutions of purpureocobalt with solutions of cobalto- 

 nitrite of sodium in excess of sodic nitrite, it often happens that 

 neither of the salts above described is obtained in appi-eciable quantity, 

 but only uiti-ate of xanthocobalt and cobaltic salts formed by a total 

 reduction of the cobaltamines. It is therefore important to avoid an 

 excess of sodic nitrite as much as possible. 



Ghloro-jiuosilicate. — When a hot solution of chloride of purpureo- 

 cobalt containing a few drops of free chlorhydric acid is j^oured into a 

 hot solution of fluosilicate of zinc, no precijiitate is produced at first, 

 but after a time a beautiful violet-red crystalline salt separates. This 

 salt is very slightly soluble in cold water, but dissolves with a violet 

 tint in a large quantity of boiling water ; it is readily soluble in a hot 

 solution of sodic carbonate. It is decomposed very quietly by heat, 

 giving off white condensible vapors, and leaving a dull violet residue. 

 Of this salt : — 



0-6379 gr. gave 0-2861 gr. SO^Co = 17-10'^ cobalt. 

 0-8221 gr. gave 0-2479 gr. silver = 9-91% chlorine. 



The formula Co2(NH3)io(SiF,.),,Cl,-l-30H3 requires 16-93% cobalt, 



and 10-18% chlorine. 



The determination of the chlorine was made by dissolving the salt in 

 sodic carbonate, adding an excess of argentic nitrate, and afterward a 



