CHEMISTRY: STIEGLITZ AND COLLABORATORS 
209 
apparently, even attempted; for the 'benzidine' and related rearrange- 
ments of hydrazines are of a different type. This seemed the more 
strange since derivatives of the third member^^ of the group, hydrogen 
peroxide, have been found to undergo the same typical rearrangement^^ 
(see part I, equation 5) as the hydroxylamine derivatives. 
When we undertook to bridge this gap in the chemistry of the hydra- 
zine derivatives, no difficulty was anticipated in securing such rear- 
rangements. Not only the fundamental analogy between the groups of 
compounds indicated this, but, from the point of view of the univalent 
nitrogen theory of the rearrangements, it seemed as if there should be 
no difficulty in removing ammonia from triphenylmethylhydrazine and 
thus starting a rearrangement: 
(C6H5)3C.NH.NH2 - NHs (C6H5)3C.N (C6H5)2C iNCeHs. (10) 
We were very much surprised to find that all our attempts to cause a 
rearrangement of this hydrazine and of a great many others by means 
of acids or zinc chloride (reagents to combine with ammonia) were com- 
plete failures. Although the experiments are being continued in spite 
of the original failures, these were sufficiently numerous and impressive to 
lead us to reconsider in greater detail the theoretical relations involved. 
Now, for rearranging chloroamines (e.g., (C6H5)3C.HN" +C1), hydroxy- 
lamines (e.g., (C6H5)3C.HN- +0H) or peroxides (e.g., (CH3)2C^ I ), 
the distribution of the charges on the atoms in the unstable parts of 
the molecules can be determined with a great degree of rehability.^^ 
For a hydrazide, this is not the case, no criteria having as yet been es- 
tablished by means of which we can distinguish, for instance, between 
the electromeric structures: 
Obviously, however, the question of the molecular rearrangement of 
triphenylmethyl hydrazine could very well, and probably would, depend 
altogether on its electronic structure. For instance, a decomposition 
(CeH5)3C.HN.NH2 (C6H5)3C.NH2 + (HN) would not lead to a rear- 
rangement and the electronic structure of the hydrazide might insure 
this decomposition in place of the one expressed in equation 10. 
These considerations made it appear desirable to attempt the rear- 
rangement of a symmetrical molecule, di-triphenylmethylhydrazine, 
about whose electronic structure there could be no doubt: (C6H5)3 
C.HN+.~NH.C(C6H5)3. A loss of triphenylmethylamine from either 
side would leave a univalent nitrogen derivative (C6H5)3C.N and rear- 
(C6H5)3C.HN-.+NH2 and (C6H5)3C.HN+.-NH2 
