Manchester Memoirs, Vol. xlviii. (1904), No. 14. 9 



also explains its reactions much better than does the 

 usual one (c.p. Armstrong). 



The fact that a comparatively small number of 

 isomers indicated by the above formulae exists, is to be 

 expected, for the isomers would in many cases differ 

 greatly in stability. There would be a " stable " and a 

 " labile " form, and often the labile would pass into the 

 stable form so quickly as to prevent its isolation. 



The ammonium theory has been freely used in the 

 past in order to arrive at a formula for analogously 

 constituted bodies, thus we have the phosphonium, 

 arsonium, stibonium, sulphonium, oxonium, and dia- 

 zonium compounds. We should, in view of the new 

 theory, write the general formulae 



H3P = C1H, H3As = ClH, H3Sb = ClH, H,S = C1H, Hp = ClH 



corresponding to 



HgN^ClH. 



The diazonium salts which are now considered to be 

 derived from, e.g., aniline hydrochloride by the replacement 

 of three hydrogen atoms by one nitrogen, thus 



CeHfiN . CI -> CeHfiN . CI 



III 111 



H3 N 



a mode of representation which does not account for the 

 explosive nature of these compounds would be better 

 written, in my opinion, as follows : 



C6H6.N = C1-^C6H6N = CI 



II I \/ 



H, H N 



resembling the formulae of the azoimides 



CeHsN - N or perhaps CeHgN = N 



N N 



This mode of representation of the diazonium salts 



