) 
200 CHEMISTRY: J. STIEGLITZ 
and, since the reagent used remains in the same condition of oxidation 
throughout the reaction, these electrons must be taken from some neigh- 
boring atom or atoms in the original molecule. This, obviously, must 
disturb the valences holding the atoms in the molecule, and a molecular 
rearrangement is altogether inevitable. 
The evidence as to the charges on the individual atoms and radicles 
is clearest for triphenylmethylchloroamines and its analogs,^ and we 
must formulate the rearrangement in the following way, the charges 
that are changed being indicated by heavy type : 
(QH5+)2lCt-NI+HZE!lH (CaH5+)2lCi-N ->(C6H5+)2lCj-Nl 
(C6H5+) C1+ (CeH5+) ^^'^'^ 
^(C6H5+)2lCilN-(+C6H5) (3). 
It is assumed that C1+ takes the two electrons first from the nitrogen 
originally holding it. (The same result would be obtained if the elec- 
trons were taken directly from the carbon by the positive chlorine. 
Evidence on this part of the problem is presented in the following article.) 
The nitrogen is thus rendered incapable of holding the H+ and Cl~ and 
these are lost, forming sodium chloride and water with the reagent so- 
dium hydroxide, as indicated in equation 2. The unsaturated (unival- 
ent^^) nitrogen, in turn, takes two electrons from the carbon atom hold- 
ing it, and the new positive charge produced thereby on the Carbon 
atom must release one of the positive phenyl groups and this must move, 
under the force of the electrical field, to the negative nitrogen. 
We recognize that the disturbing, labile element in the original struc- 
ture is just the unstable, positive chlorine, so admirably represented by 
Thomson's first theory, and that its undeniably great tendency to go 
over into its common stable negative form is the original source of the 
molecular disturbance and rearrangement. 
The product of the rearrangement, phenylimido benzophenone, in 
harmony with the electronic structure resulting for it according to equa- 
tion 3, is uniformly hydrolyzed by acids to benzophenone and aniline: 
(C6H5)2CtN-(+C6H5) + HOH + HQ -> (C6H5)2Ct ZO -f- 
(CeH5+-NEH3+)+-Cl (4). 
It is evident that the appHcation of the original theory of Thomson to 
the valence relations in triphenylmethylchloroamine gives us a complete 
and precise picture of the cause, mode, and product of its molecular 
rearrangement. 
If we attempt to express the rearrangement in terms of the theory 
