19, G 
Perkins: Addition to the Double Bond, I 
659 
H 
Compound B 
0 
H 3 C - C - 0 - C 2 H 5 
0 
H 
Further discussion of this reaction, including the subsequent 
stage of regeneration of the double bond, will be postponed to 
a later paper. It is sufficient for the present to point out the 
influences which cause this type of reactivity to vary in differ- 
ent compounds. The reaction does not depend on large spaces 
in the undissociated shell, but only on dissociation, so that factor 
(2) mentioned above is primarily important. Factor (3) is 
relatively unimportant, and factor (1) has no influence. 
CONCLUSION 
The discussion of the two typical reactions has been very 
brief, the purpose of this paper being to present a hypothesis, 
not to justify it. The hypothesis is admittedly in a rough form, 
and doubtless will require certain additions and modifications 
of the less-important details. Nevertheless it forms a definite 
and consistent basis for many of the various “rules” and “prin- 
ciples” that have been applied to organic reactions, in so far as 
they are consistent with the facts, and is particularly valuable 
in dealing with the “exceptions” to these rules, among which 
may be considered the benzene nucleus. The writer expects to 
take up some detailed applications in a series of papers, but 
before continuing this series the subject of one-electron and 
three-electron bonds, not provided for in Langmuir’s postulates, 
will be considered. 
SUMMARY 
1. A theory of the reaction mechanism of the direct union 
is proposed, based principally on Langmuir’s octet theory of 
valence. 
2. The distinct features of the theory are certain working defi- 
nitions of the electronic forces and the suppositions (a) that 
dissociation of the direct bond usually takes place in such a 
