of the Compounds of Ammonia. 123 



extended so as to embrace a large class of the ammoniacal 

 compounds newly discovered, — but can it embrace all? When 

 calomel is treated by water of ammonia there is formed the 

 body, Hg 2 CI + Hg 2 N H 2 . If the ammonium theory be ap- 

 plied to this body, are we to make a new compound radical, 

 and say it is CI + N H 2 Hg 4 , and to say that the correspond- 

 ing body S 3 . Hg 2 O + Hg 2 Ad is S 3 . O N H 2 Hg 4 ? 

 Here, then, is no parallel whatever ; these bodies lie altogether 

 out of the possibility of replacement connecting them with or- 

 dinary ammonium ; and it would be far too violent a supposi- 

 tion to assume the existence of a body, N H 6 , in order to sup- 

 port the disputed existence of a more likely body, N H 4 . I 

 consider this example as being fatal to Graham's view. Sub- 

 limate CI Hg gives white precipitate CI Hg . Hg Ad, calomel 

 CI Hg 2 gives black precipitate CI Hg 2 . Hg 2 Ad ; relations so 

 simple, so natural, that it should require very strong reasons 

 indeed to prove that they are not those consonant to truth. 



If the theory of metallic ammoniums were adopted, it would 

 be only just to give to it its proper form. A sulphate of am- 

 monia perfectly isomorphous with sulphate of potash contains 

 S0 3 .NH 3 + 2HO. The replacing element of the potassium 

 is therefore (N H 5 O) ; there is no doubt of this ; it is one of 

 Mitscherlich's best-established determinations ; (N H 4 ) and 

 (N H 5 O) are equally isomorphous with potassium. Hence as 

 sublimate treated in the cold with ammonia gives CI N H 4 and 

 CI . N H 2 Hg 2 , so in boiling water we get CI 4- (N H 2 Hg 4 2 ) 

 equally equivalent, and a sort of complex ammonium. The 

 ammonia turbeth is thus : S 3 + O (N H 2 Hg 4 2 ) ; there is 

 N0 5 + (NH 2 Hg 4 2 ) and I(NH 2 Hg 4 2 ), and so on. 

 Now the correspondence of the common oxychloride of mer- 

 cury to these bodies has been proved by Ullgren, and it 

 should therefore be looked upon as a chloride of a compound 

 radical CI + (Hg 4 O s ). This is truly the principle involved 

 in Graham's idea of compound ammoniums; for no matter 

 where we begin, we find the chain by which the common salts 

 of ammonia and the common basic salts are connected so per- 

 fect, that whatever principle we apply to one includes the others. 

 Already, two years ago (May, 1838, Annalen der Pharmacie) , 

 I started the question, Were basic salts salts of compound radi- 

 cals ? and Liebig, in adapting a theory to the salts of platina dis- 

 covered by Gros, approached nearly to the embodying of the 

 same idea ; but further examination showed me that it is one 

 which is at present quite unfit for science, the even partial 

 adoption of which would throw into confusion the most posi- 

 tive and simplest systematic arrangements that chemistry pos- 

 sesses, and hence do much harm and no good. Having so 



