652 The Philippine Journal of Science 1921 
pairs (probably figure-eights) but not enough to bring these 
pairs together. 
The pairs have strong constraints toward their normal posi- 
tions in each shell, so that independent rotation of the atoms is 
difficult, and probably impossible without dissociation. 
The lateral displacement of the electrons causes large elec- 
tron-free spaces, mentioned below in connection with direct 
addition. The radial displacement, caused by the increased re- 
pulsion of the atomic kernels at the closer approach demanded 
by the double bond, brings the electrons nearer to the shell 
boundaries than in the ordinary single bond. Consequently a 
double bond, or at least one-half of it, is more easily broken than 
a single bond. 
The triple bond. — The triple bond, with three pairs of elec- 
trons, is extremely unstable, due to the extensive displacements of 
the electrons from their normal positions. The formation of 
acetylene and similar compounds at high temperatures is by no 
means a contradiction of this statement. 
Dissociation. — When two atoms are bound by a pair of elec- 
trons whose position of equilibrium is near the boundary of one 
of the shells, only a slight thermal shock is necessary to cause 
the electrons to leave this shell entirely and take up a new posi- 
tion in the other shell. It sometimes happens that the electrons 
are near enough to the boundaries of both shells so that one 
electron goes to each shell. Due to the magnetic stability of a 
pair of electrons, however, and because bonds weak enough to 
dissociate are usually quite unsymmetrical, the result in nearly 
all cases is that both of the electrons go to the shell which attracts 
them most strongly, leaving two vacancies in the weak shell. 
The usual result, therefore, of the dissociation of A — B is 
not A 1 B as assumed in the affinity unit hypothesis, but 
A< -fB . This is the fundamental point on which the writer’s 
theory differs not only from the older theories of reaction mecha- 
nism but also from any of the recent ones that have come to his 
attention. It appears to be a reasonable deduction from Lang- 
muir’s octet theory, but the importance of it in organic chemistry 
seems to have escaped notice. Attempts to apply electronic 
conceptions to organic reactions have resulted in little advance 
over the older theories because of the assumption of dissociation 
+ 
in the manner A 1 B~, which does not explain the observed 
facts in the case of the direct bond. 
Direct addition. — The shell boundaries above defined have 
reference only to the removal of electrons from a normal shell, 
