186 REPORT—1904. 
(vi) Compounds containing Asymmetric Carbon and Nitrogen Atoms. 
The examination of these compounds was undertaken with a view of 
establishing another analogy between asymmetric carbon and nitrogen 
atoms, and it has been shown! that when an active tertiary amine (methyl 
[amyl aniline) combines with an alkyl iodide (allyl or benzyl iodide), 
unequal quantities of the two possible compounds are produced as anti- 
cipated if the nitrogen behaved like a carbon atom. 
The two compounds formed with benzyl iodide differ somewhat in 
their solubility, but not enough to make a complete separation by 
crystallisation feasible : this can be effected by means of the camphor- 
sulphonates. One of the iodides is dextro- and the other is levo-rotatory ; 
a solution of either in chloroform in the cold or in alcohol on warming 
becomes converted into the other until a state of equilibrium is reached, 
the change from one isomeride to the other being effected by the splitting 
up of the salt into benzyl iodide and amine, which then recombine, as in 
the racemisation of active nitrogen compounds in chloroform solution. 
(vii) The Configuration of Quinguevalent Nitrogen Compounds. 
(a) A fixed Configuration necessary.—The foregoing results demand a 
stable configuration for the molecule of ammonium compounds, in which 
the nitrogen atom is quinquevalent, and is attached to five univalent 
atoms or groups. 
Werner ? still adheres to a modification of the old ‘ molecular compound 
hypothesis’ to account for the ammonium compounds, and regards the 
existence of isomerism and optical activity as an objection to the view 
generally accepted. Dr. J. C. Cain takes the same point of view, 
regards the stability of the ammonium compounds towards alkalies as a 
further objection, and proposes a new hypothesis in which ammonium 
chloride is represented as H;N=CI—H. Two of the numerous important 
objections to this view may be mentioned: first, there is not enough 
evidence to show that compounds with the structure usually assigned to 
ammonium salts would not be stable, and still less for concluding that 
tervalent halogen derivatives would be stable; and, secondly, the 
formule proposed to account for the isomeric hydrindamine salts— 
namely, a,N=X—b and a,.5N=X—a—do not account for the mode of 
formation of these salts, and the formule proposed for active nitrogen 
compounds—namely, bedN=X—a—and acdN=X—b—are not optical 
antimers, and could not by a single process be produced together in equal 
quantities. 
Five points cannot be arranged symmetrically around one point so as 
to be interequivalent ; hence either one or two of the valencies of nitrogen 
must be different from the others : this conclusion finds expression in all 
the configurations proposed, and is supported by the facts. It has been 
found impossible to prepare a quinquevalent nitrogen compound contain- 
ing no electro-negative radical nor one containing more than two such 
radicals.4 
(b) Lhe ‘ Cubie Configuration.’—The first configuration proposed was 
1 Jones, Proc. Camb. Phil. Soc., 1904, xii. 466. 
2 Annalen, 1902, 322, 261. 
3 Memoirs of the Manchester Lit. and Phil. Soc., 1904, 48, No. 14. 
+ Lachmann, American Chem. Jowrn., 1896, 18, 372. 
