th in so many of the substances referred to, as, for instance, the
metals, water, air, &c., would lead to many very curious theoretical
generalizations, which I must not indulge in here. One point, however, I
shall venture to refer to. Conduction appears to be essentially an action
of contiguous particles, and the considerations just stated, together with
others formerly expressed (1326, 1336, &c.), lead to the conclusion, that
all bodies conduct, and by the same process, air as well as metals; the
only difference being in the necessary degree of force or tension between
the particles which must exist before the act of conduction or transfer
from one particle to another can take place.

1612. The question then arises, what is this limiting condition which
separates, as it were, conduction and insulation from each other? Does it
consist in a difference between the two contiguous particles, or the
contiguous poles of these particles, in the nature and amount of positive
and negative force, no communication or discharge occurring unless that
difference rises up to a certain degree, variable for different bodies, but
always the same for the same body? Or is it true that, however small the
difference between two such particles, if _time_ be allowed, equalization
of force will take place, even with the particles of such bodies as air,
sulphur or lac? In the first case, insulating power in any particular body
would be proportionate to the degree of the assumed necessary difference of
force; in the second, to the _time_ required to equalize equal degrees of
difference in different bodies. With regard to airs, one is almost led to
expect a permanent difference of force; but in all other bodies, time seems
to be quite sufficient to ensure, ultimately, complete conduction. The
difference in the modes by which insulation may be sustained, or conduction
effected, is not a mere fanciful point, but one of great importance, as
being essentially connected with the molecular theory of induction, and the
manner in which the particles of bodies assume and retain their polarized
state.

* * * * *

P xi. _Relation of a vacuum to electrical phenomena._

1613. It would seem strange, if a theory which refers all the phenomena of
insulation and conduction, i.e. all electrical phenomena, to the action of
contiguous particles, were to omit to notice the assumed possible case of a
_vacuum_. Admitting that a vacuum c