11 PROCEEDINGS OF THE AMERICAN ACADEMY 



This salt is isomorphous with the corresponding potassium saU, the 

 ciystals belonging to the rhombic system. Erdmann does not explain 

 the reaction which takes place in the formation of this or the corre- 

 sponding potassium salt, and regards the compounds in question as 

 double salts. When slightly acid solutions were employed, Erdmann 

 obtained, in addition to the above-mentioned salt, an ammonic salt 

 corresponding to Fischer's salt, Co2(N'02)i2(NHJg-|-20H2, as we 

 should now write it. The existence of this salt was first remarked by 

 Genth and myself.* Sadtler studied the action of ammonic nitrite on 

 acid sortitions of cobaltic chloride, and obtained two salts having 

 respectively the formulas : — 



Co2(N02),o(NH,),+ 20H2, 



Co,(N02),2(NH,)3+20H2, 

 but did not observe the formation of Erdmann's ammonium salt. In 

 repeating these experiments, I always obtained Erdmann's ammonium 

 salt, Co2(NH3)^(N02)8(NH^)2, in largest quantity. The crystals are 

 uncommonly beautiful and well defined. Of these crystals, 



0-3390 gr. gave 0-1783 gr. SO^Co = 20-02 per cent. 

 The formula requires 20-00 per cent. In one experiment, in which a 

 little free acetic acid was present, I obtained lai-ge, dark sherry-wine 

 colored prismatic crystals, which after solution and recrystallization 

 gave only very thin lozenge-shaped tabular crystals, the form and 

 appearance of which are highly characteristic. These crystals gave 

 no reactions with salts of luteocobalt, purpureocobalt, and roseocobalt, 

 and none with potassic chromate and dichromate, ammonic oxalate or 

 arsentic nitrate. The absence of the first-mentioned reactions shows 

 that they do not contain Co2(NH3)4(N02)8 or Co2(N02)i2, while the 

 fact that they give no reactions with alkaline chromates and oxalates 

 shows that they do not contain any known cobaltamin. Of these 

 crystals, 



0-1554 gr. gave 0-0974 gr. 80^00= 23-86 per cent cobalt. 



0-3081 gr. gave 0-0635 gr. NH3= 20-61 per cent ammonia. 



The formula Co2(NH3)g(N02)e requires : — 



Cobalt, 23-79 23-86 



Ammonia, 20*56 20-61 



These analyses are sufficient to identify the salt in question with one 

 which Erdmann has described in the paper referred to as formed by 

 the action of ammonia and potassic nitrite upon cobaltic chloride, 



* This Journal, 2d Series, vol. xxiv. p. 86. 



