OF ARTS AND SCIENCES. lO 



presented orange-red crusts of indistinct octuliedraJ crystals. These 

 were rubbed to powder, washed with a little cold water, and dissolved 

 in hot water with a little free sulphuric acid. Fine yellow crystals 

 separated, which could not be distinguished in appearance from sid- 

 phate of luteocobalt, but appeared to be less soluble in water. Of 

 these crystals, — 



0-5018 gr. gave 0-2341 gr. SO.Co = 17-76^ cobalt. 



0-7GU gr. gave 0-7992 gr. SO, Ba = 43-25% So,. 



1-1G18 gr. gave 213 c.c. nitrogen at 13'^C and 7G()-47""" =21-G2%. 



The formula Co,(NH.,),„(SO,),,4-50II,, which is also that of the 

 ordinary and of the soluble modification of sulphate of roseocobalt, 

 requires : — 



The formula Co,(NII,)„(S0,),-|-50IT, — sulphate of luteocobalt — 

 requires IG.85% cobalt, 41-14%SO,, and 24-00% nitrogen. The 

 analyses show clearly that the salt was not sulphate of luteoco- 

 balt. The solution of the new sulphate gave beautiful yellow or 

 orange-yellow crystalline precipitates with various I'eagents closely 

 resembling in color the corresponding salts of luteocobalt. Thus with 

 potassic bromide it gave a buff yellow, and with potassic iodide a fine 

 orange-yellow crystalline precipitate. With the platino-chloride and 

 auro-chloride of sodium it gave beautiful yellow salts, the analyses of 

 which will be given in connection with those of the ordinary salts 

 of roseocobalt. By double decomposition with baric nitrate the yellow 

 sulphate gave a fine yellow solution, which on standing yielded crystals 

 of the nitrate of this series readily soluble in hot water. In these 

 crystals, — 



2517 gr. gave 0-1092 gr. SO,Co = 16-50% cobalt. 



This percentage of cobalt would indicate the formula Co2(NH,),q 

 (NOg)j-|-30H2, which requires 16-52%. Nitrate of luteocobalt con- 

 tains 17-00% cobalt. 



The sulphate gave also with baric chloride a fine yellow solution 

 and baric sulphate. The solution of the chloride gave beautiful yellow 

 salts, with the chlorides of gold and of platinum. Unlike the ordinary 

 and soluble modifications of chloride of roseocobalt, it did not give 

 chloride of purpureocobalt by boiling with chlorhydric acid. The yel- 



