of the Compounds of Ammonia. 125 



hypothetical about them ; everything follows from experi- 

 ment. 



Now the existence of amidogene is one of the best- esta- 

 blished and most universally received hypotheses in chemistry ; 

 moreover, it is adopted also by Graham and by Rose. But 

 how many more hypotheses must they adopt for the theory of 

 ammonium ! I shall only count up a very few, or better, state 

 that the number of hypothetical bodies necessary for the com- 

 plete ammonium theory would be equal to the number of pos- 

 sible metallic amidides, therefore equal to the number of me- 

 tallic chlorides and oxides at present known ; and this without 

 at all touching on the theory of basic salts, which, as I have 

 shown, is a necessary consequence of the theory of complex 

 ammoniums. 



If general and simple laws can be obtained by the intro- 

 duction of an hypothesis, and according as experimental re- 

 search proceeds, the new facts gained are found to regulate 

 themselves according to it, a real and important service is con- 

 ferred upon science by him to whom we are indebted for it. 

 But when a theory must change its shape and make a new 

 assumption for each new fact discovered, as soon as the di- 

 rect tendency of the theory is at an angle with that of research, 

 and it must tack from side to side to keep the course of dis- 

 covery in its line, its day has passed ; and notwithstanding 

 the services rendered to chemical theory by the hypothesis of 

 ammonium, it is now, as I conceive, incapable of retaining its 

 old position. Its great utility was in fixing attention on the 

 relations of the ammoniacal and potash salts; but for explain- 

 ing the immensely extended classes of compounds which am- 

 monia is now known to form, it is insufficient. 



In a paper lately published by Mitscherlich, he describes a 

 compound of chloride and azoturet of mercury 2 Hg CI -f- N Hg 3 , 

 and he makes an observation which, as connected with the 

 present subject, I shall here notice. He says that the equivalent 

 of white precipitate is not CI Hg -f- Hg Ad, but three times that, 

 because it requires 3 (ClHg+ HgAd) to give CI Hg-f 2NH 3 

 and 2 CI Hg + N Hg 3 : he says that also amide of potassium 

 KNH 2 should be taken 3 K Ad, because it gives 2 N H 3 and 

 K 3 N. Now this appears to me to be a very irrational me- 

 thod, for then the body CI Hg N H 3 , the simple formula of 

 which Mitcherlitch admits, should be 8 (ClHg NH 3 ), because 

 it gives 2 CI Hg + 3 CI Hg 2 + N + 4 N H 3 + 3 CI NH 4 . The 

 hydrate of phosphorous acid should be 4 P 3 + 3 H O, be- 

 cause it gives P H 3 and 3 P 5 ; and a crowd of other exam- 

 ples might be brought forward. The equivalent formula of a 

 body cannot be fixed thus from a single action or property. 



