Manchester Memoirs, Vol. xlviii. (1904), No. 14. 3 



are still the isomeric inactive phenyl benzyl allyl methyl 

 ammonium iodides of Wedekind {^Ber., 1899, 32, 518), 

 one of which has been resolved into its optically active 

 constituents by Pope and Peachey (/. C. S., 1899, 75) 

 1 1 27), and the several sets of isomers described by 

 Kipping (/. C. S., 1900, 77» 861. 1901, 79, 430. 1903, 83, 

 873), which are well defined. The latter has obtained 

 two isomers of hydrindamine bromocamphorsulphonate, 

 chlorocamphorsulphonate and cis-7r-camphanate as well 

 as others from substituted hydrindamines, none of which 

 contain an asymmetric nitrogen atom. The existence of 

 such isomeric forms recalls the experiments of V. Meyer 

 and Lecco (1876) who showed the identity of the 

 quaternary ammonium iodides obtained from trimethyl- 

 amine and ethyl iodide and ethyl dimethylamine and 

 methyl iodide, and concluded that this compound was 

 properly represented by the ammonium theory and not 

 to be regarded as a substituted ammonia hydrochloride. 

 This conclusion must now be doubted in view of the 

 many cases of two isomers here quoted. 



Werner proposes a new ammonium theory which, 

 involving as it does such terms as " Principal Valency " 

 (" Hauptvalenz "), " Supplementary Valency " (" Neben- 

 valenz "), and " Co-ordination Constant," is apparently 

 not very different from the older idea of molecular 

 attraction or residual affinity. According to Werner, the 

 formula of ammonium chloride would be 



H3N HCl, 



a supplementary valency (indicated by the dotted line) 

 existing between the nitrogen atom and the hydrogen of 

 the hydrochloric acid ; an assumption for which there is 

 at present no experimental evidence. 



The constitution which I suggest for the ammonium 

 compounds would, I believe, account in a very simple 



