ty amongst dielectrics will be found to be in
intimate relation, the whole difference probably depending upon the mode
and degree in which their particles polarize under the influence of
inductive actions (1338. 1603. 1610.).

* * * * *

P x. _Convection, or carrying discharge._

1562. The last kind of discharge which I have to consider is that effected
by the motion of charged particles from place to place. It is apparently
very different in its nature to any of the former modes of discharge
(1319.), but, as the result is the same, may be of great importance in
illustrating, not merely the nature of discharge itself, but also of what
we call the electric current. It often, as before observed, in cases of
brush and glow (1440. 1535.), joins its effect to that of disruptive
discharge, to complete the act of neutralization amongst the electric
forces.

1563. The particles which being charged, then travel, may be either of
insulating or conducting matter, large or small. The consideration in the
first place of a large particle of conducting matter may perhaps help our
conceptions.

1564. A copper boiler 3 feet in diameter was insulated and electrified, but
so feebly, that dissipation by brushes or disruptive discharge did not
occur at its edges or projecting parts in a sensible degree. A brass ball,
2 inches in diameter, suspended by a clean white silk thread, was brought
towards it, and it was found that, if the ball was held for a second or two
near any part of the charged surface of the boiler, at such distance (two
inches more or less) as not to receive any direct charge from it, it became
itself charged, although insulated the whole time; and its electricity was
the _reverse_ of that of the boiler.

1565. This effect was the strongest opposite the edges and projecting parts
of the boiler, and weaker opposite the sides, or those extended portions of
the surface which, according to Coulomb's results, have the weakest charge.
It was very strong opposite a rod projecting a little way from the boiler.
It occurred when the copper was charged negatively as well as positively:
it was produced also with small balls down to 0.2 of an inch and less in
diameter, and also with smaller charged conductors than the copper. It is,
indeed, hardly possible in some cases to carry an insulated ball within an
inch or two of a charged plane or convex surface without its receiving a
charge of the contrary