remity of the tube, in consequence of the decomposition of the
water, noted. Without moving the plates, the acid between the copper and
zinc was agitated by the introduction of a feather. The bubbles were
immediately evolved more rapidly, above twice the number being produced in
the same portion of time as before. In this instance it is very evident
that agitation by a feather must have been a very imperfect mode of
restoring the acid in the cells against the plates towards its first equal
condition; and yet imperfect as the means were, they more than doubled the
power of the battery. The _first effect_ of a battery which is known to be
so superior to the degree of action which the battery can sustain, is
almost entirely due to the favourable condition of the acid in contact with
the plates.

1040. A _second_ cause of diminution in the force of the voltaic battery,
consequent upon its own action, is that extraordinary state of the surfaces
of the metals (969.) which was first described, I believe, by Ritter[A], to
which he refers the powers of his secondary piles, and which has been so
well experimented upon by Marianini, and also by A. De la Rive. If the
apparatus, fig. 103. (1096.), be left in action for an hour or two, with
the wire 3 in contact with the plate _p_, so as to allow a free passage for
the current, then, though the contact be broken for ten or twelve minutes,
still, upon its renewal, only a feeble current will pass, not at all equal
in force to what might be expected. Further, if P^{1} and P^{2} be
connected by a metal wire, a powerful momentary current will pass from
P^{2} to P^{1} through the acid, and therefore in the reverse direction to
that produced by the action of the zinc in the arrangement; and after this
has happened, the general current can pass through the whole of the system
as at first, but by its passage again restores the plates P^{2} and P^{1}
into the former opposing condition. This, generally, is the fact described
by Ritter, Marianini, and De la Rive. It has great opposing influence on
the action of a pile, especially if the latter consist of but a small
number of alternations, and has to pass its current through many
interpositions. It varies with the solution in which the interposed plates
are immersed, with the intensity of the current, the strength of the pile,
the time of action, and especially with accidental discharges of the plates
by inadvertent contacts or reversions of the p