44 



PROCEEDLNGS OF THE AMERICAN ACADEMY 



THEORETICAL VIEWS. 



Since the appearance of the first part of this paper many chemists 

 have given expression to theoretical views of the constitution of tlie 

 ammonia-cobalt salts. Of these I think it will be necessary to notice 

 onlj' those in which the atomicity of cobalt is taken into account, the 

 older theories having passed away with the chemistry of which they 

 formed a part. So far as I can determine, Frankland * first endeav- 

 ored to reduce the formulas of the cobaltamines to atomistic expres- 

 sions. In the first edition of his lecture notes he gives for the chlorides 

 of jjurpureocobalt and luteocobalt respectively the formulas, — 



Co., 



NH.,(NHJC1 

 NH,;(NHJC1 

 NHXl 



^^h'ci 



NH.;(NH,)C1 

 NH^(NH;)C1 



Co., 



NH„(NH,)C1 



nh:.(nhjci 



NH."(NHJC1 

 NH"2(NHJC1 



NH.,(NHJC1 



nh;(niijci 



It is easy to see that the other series of salts may be formulated in 

 a similar manner, Franklaud's view was an important step in ad- 

 vance. It may fairly be objected to it, however, that it involves the 

 replacement of hydrogen in ammonium by ammonium and by chlorine, 

 a view which was not new, and which is certainly defensible, but 

 which has never been generally received by chemists. If we replace 

 in ammonium, NII^, one atom of hydrogen by one atom of clilorine, 

 and another atom of hydrogen by an atom of ammonium, it is dithcult 

 to see how the new ammonium, NH2(NH^)C1, can possess a suffi- 

 ciently well-marked chlorous power to unite with the highly zincous 

 cobalt so as to form an extremely stal)le compound. 



In a paper on the theory of atomicities, f I have given another view 

 of the constitution of the cobaltamines and of the analogous platin- 

 amines. If nitrogen be regarded as pentatomic, ammonia will be 

 diatomic, and any number of atoms of ammonia may be regarded as 

 constituting a single diatomic whole. Taking the atomicity of cobalt 

 (Co = 59) as 6, two atoms of the metal may be supposed to unite to 



vi vi 



form a comijlex with eight units of affinity, since we have =Co = Co= ; 

 of these eight units two will be saturated by the diatomic ammonia. 



* Lecture Notes, 1st edition, p. 196. t Am. Journal, Vol. xliv. Xov. 18G1 



