208 
CHEMISTRY: STIEGLITZ AND COLLABORATORS 
form (equation 4) . The aniline obtained gave the characteristic bleach- 
ing-powder test and formed a chloroplatinate which was analyzed (33.35, 
32.32% platinum found, theory 32.78%); the benzophenone was identi- 
fied by its melting-point 47.5°, which was unchanged when the product 
was mixed with some known, pure benzophenone. 
(3) The third question raised in this investigation concerned itself 
with the following relations: in the rearrangement of a substitution 
product of triphenylmethyl azide, such as (C1C6H4)(C6H5)2C.N(N2), a 
certain part of the substance would form an aniline derivative (CIC6H4) 
(C6H5)C:NC6H5, another part would yield a derivative of chloroaniline 
(C6H5)2C :NC6H4C1 (see above). The ratio may be easily ascertained 
by hydrolysis of the product and titration of the anilines with bromine. 
Now, the same situation exists in regard to the rearrangement of the 
corresponding hydroxylamine (CIC6H4) (C6H5)2C.NHOH and the chloro- 
amine (CIC6H4) (C6H5)2C.NHC1. If in each instance the path to the 
rearrangement lies through an intermediate univalent nitrogen deriva- 
tive, we should have the identical product (CIC6H4) (C6H5)2C.N in each 
of these cases, the azide losing nitrogen, the hydroxylamine water and 
the chloroamine halogen acid. The environment of the univalent- 
nitrogen derivative would be different in the three cases on account of 
differences in the physical state, the temperature, etc. But since the 
rearrangement is an intramolecular one, it might be little affected by 
environment, once the compound (CIC6H4) (C6H5)2C.N is formed. A 
comparison of the ratios of anHne to chloroaniline formed in the differ- 
ent series could then shed some light on the question of the probability 
of the formation of the hypothetical intermediate univalent nitrogen 
derivative. On account of the difficulty of securing experimental mate- 
rial only a beginning has been made on this phase of the problem of 
these rearrangements. We have already found, however, that in the 
rearrangement of (ClCftH4) (C6H5)2C.N(N2) the ratio of aniline to chloro- 
aniline formed is 70/30 in molar terms. This compares with the ratio^^ 
71/29 found for the rearrangement of (CIC6H4) (C6H5)2C.NHOH. The 
result speaks for itself. 
That hydrazine and hydroxylamine are fundamentally similar com- 
poimds^i is a well-established fact, shown both by their reactions and 
also especially by their forming parallel derivatives, of similar proper- 
ties, with a great many organic substances. One strange difference in 
the chemistry of these derivatives, as thus far developed, seems to have 
been overlooked, and that is that molecular rearrangements of hydra- 
zines, corresponding to the rearrangements of hydroxamic acids, ketox- 
imes, triphenylmethylhydroxylamines, etc., have not been observed, or. 
