CHEMISTRY: STIEGUTZ AND COLLABORATORS 
203 
have the tautomeric structures, RC(OMe) (:NHal) or RCO.N(Me)Hal. 
This complication does not exist in the case of triphenylmethyl halogen 
amines (C6H5)3C.NHHal, whose metal derivatives could only have the 
definite structure (C6H5)3C.N(Me)Hal. A successful molecular rear- 
rangement of these halogen amines by bases would give convincing evi- 
dence as to the nature of the salt formed in initiating the rearrangement 
and would give strong presumptive evidence in favor of the view that 
in the rearrangement of acyl halogen amines, the salts RCO.NMeHal 
are the ones directly concerned in the rearrangement. 
Further: chloroimidoketones (C6H5)2C :N~+C1 have been found to 
resist rearrangement^^ in spite of the presence of the unstable positive 
chlorine atom (see part I). This result was not unexpected from the 
point of view of the theory of the one of us concerning the mechanism of 
these molecular rearrangements:* the theory postulates that the easiest 
path to a rearrangement is through the intermediate formation of unival- 
ent nitrogen derivatives RCO.N<, R3C.N< (see equation 3), the free 
valences of the univalent nitrogen being considered sufficiently powerful 
to tear the radicle R from the carbon and thus effect the rearrangement 
to a more stable molecule (see equation 3). The significance of the 
intermediate formation of a univalent nitrogen derivative in relation to 
the electron theory is discussed later on in the text. Triphenylmethyl 
halogen amines, R3C.NH(Hal) and R3C.N(Hal)2, in contrast with the 
chloroimidoketones R2C:NC1, present the opportunity for an unham- 
pered formation of the univalent nitrogen derivative R3C.N< by the 
loss of H(Hal) or (Hal)2 under suitable experimental conditions. They 
were therefore expected to be easily subject to rearrangement. 
This investigation was undertaken with the object of deciding the 
questions indicated and thus throwing light on the whole problem of 
these molecular rearrangements. The experimental results confirmed in 
every respect the theoretical anticipations. Of the facts established 
and their bearing, the following are the most important : 
1. A number of triphenylmethyl halogen amines — (C6H5)C.NHBr, 
(C1C6H4)2(C6H5)C.NHC1, (BrC6H4) (CIC6H4) (C6H5)C.NHC1, and 
(C6H5)3C.NCl2 and (CIC6H4) (C6H5)2C.NCl2 — were found to undergo un- 
der suitable conditions a molecular rearrangement into phenylimido 
benzophenones R2C:NR (see equation 3). The rearrangement was es- 
tablished in the case of the first compound listed by the identification 
of the phenylimido benzophenone (C6H5)2C :NC6H5 (yellow crystals, 
m.p. 110°; the melting-point of a mixture of the product with synthetic 
substance (m.p. 111°-112°) was found at 111°) and by the identification 
of the benzophenone (m.p. 47°) and aniKne (bleaching-powder test; the 
