Manchester Memoirs, Vol. Ixiv. (1920), No. 4 



11 



The product should therefore be the exocyclic hydrocarbon 

 (II), whilst Meerwein actually isolates A^dimethylcyc/o- 

 hexene (III). Under the conditions of the transformation 

 the change of position of the ethylene linkage would, how- 

 ever, be anticipated. 



Before leaving the subject of molecular rearrangement, an 

 interesting recent observation of J. Meisenheimer (Ber., 1919, 

 52, (B), 1667) may be noted. Methylallylaniline N-oxide (IV) 

 on distillation with a solution of sodium hydroxide in a 

 current of steam is converted into the allyl ether of phenyl- 

 methylhvdroxylamine (V). This reaction is interesting from 

 the present point of view because, however regarded, it 

 involves the formation of an odd-membered ring. This, as 

 shown below, is rendered possible bv the circumstance that a 

 pentavalent nitrogen atom becomes trivalent in the course of 

 the process. 



CH O 



II II 



CH N Me Ph 



\ch/ 



(IV.) 



CH- 



+ 2 



CH- 



-.0 



NMePh 



Chi 



CH, 



CH 



CH, 



[V.) 



An almost inexhaustible subject for speculation is concerned 

 with the existence of partial valencies and cyclic conjugation 

 of partial valencies in normal molecules as distinct from those 

 which are considered only at the moment of reaction. It has 

 already been pointed out (G. M. Robinson and R. Robinson, 

 Trans. Chem. Soc, 1917, 111, 964; ibid., 1918, 113, 640; 

 Perkin and Robinson, ibid., 1919, 115, 943) that the theory of 

 cyclic conjugation of benzene originated by Thiele, combined 

 with a consideration of the part use of latent valencies of 

 nitrogen, affords an explanation of the aromatic character and 

 degree of basicity of pyrrole, g-lyoxaline, harmine and other 

 heterocyclic types. The extension of such views to even more 

 complex compounds is well illustrated by the case of indigotin, 

 the usual formula of which is of course 



\y 



CO 



NH 



C 



C 



CO 



NH 



\y 



