OF ARTS AND SCIENCES. 11 



I propose therefore to speak of them more in detail hereafter, and to 

 contiue myself at present to their relations to the salts of the octamin 

 series. 



"When a solution of Erdmann's salt, Co2(NH3)^(]Sr02)8K2, is added 

 to one of the octamin nitrate, a beautiful crystalline precipitate is 

 formed, which after washing with cold water may be redissolved in hot 

 water and then separates in fine orange-yellow granular crystals. The 

 equation representing the reaction is here 



Co,(NH3),(NO.0,(NO3),+ Co.,(NH3),(NO.,),sK., = 2KNO3 + 



The new salt gives with reagents the reactions of the salts of the 

 octamin series. The relation between the two complex atoms which 

 form a molecule of the new salt is worthy of notice, the number of 

 atoms of ammonia and nitroxyl in the one corresponding to the num- 

 ber of atoms of nitroxyl, NOj, and of ammonia in the other ; one 

 complex atom being, to use Graham's convenient expression, chlorous 

 and the other zincous. We have furthei'more the relation expressed 

 by the equation : — 



Now I shall show, farther on, that there exist several other salts, the 

 empirical constitutions of which may be represented also by multiples 

 of the formula 



C02(NH3)3(N02)e, 



so that we have here, for the first time, I believe, in inorganic chemistry, a 

 series of strictly metameric bodies. In the salt of the octamin series, 



0-2600 gr. gave 0-1622 gr. CoSO^ = 23-74 per cent cobalt. 



The formula requires 23-79 per cent. 



The salts which I have described are not the only ones which con- 

 tain 8 atoms of ammonia with 2 atoms of cobalt. In our memoir 

 Genth and I made mention of a leek-green crystalline body which we 

 obtained in more than one reaction in quantities too small for analysis, 

 and which we termed, provisionally, Praseocobalt. Braun subsequently 

 denied the existence of any such substance ; but, in an excellent paper 

 on the ammonia-cobalt compounds, F. Rose has not merely described 

 and analyzed the body in question, but has given a method of pre- 

 paring it in quantity. Rose gives for the formula of this salt CogClg 

 N^H^2 (old style). I should write this 



Co2(NH3),Cle. 

 and give it the atomistic formula 



