indication connected with the more or less perfect action of the
battery, I may mention that no gas ought to rise from the zinc plates. The
more gas which is generated upon these surfaces, the greater is the local
action and the less the transferable force. The investing crust is also
inconvenient, by preventing the displacement and renewal of the charge upon
the surface of the zinc. Such zinc as, dissolving in the cleanest manner in
a dilute acid, dissolves also the slowest, is the best; zinc which contains
much copper should especially be avoided. I have generally found rolled
Liege or Mosselman's zinc the purest; and to the circumstance of having
used such zinc in its construction attribute in part the advantage of the
new battery (1134.).

1145. _Foulness of the zinc plates._--After use, the plates of a battery
should be cleaned from the metallic powder upon their surfaces, especially
if they are employed to obtain the laws of action of the battery itself.
This precaution was always attended to with the porcelain trough batteries
in the experiments described (1125, &c.). If a few foul plates are mingled
with many clean ones, they make the action in the different cells
irregular, and the transferable power is accordingly diminished, whilst the
local and wasted power is increased. No old charge containing copper should
be used to excite a battery.

1146. _New and old plates._--I have found voltaic batteries far more
powerful when the plates were new than when they have been used two or
three times; so that a new and an used battery cannot be compared together,
or even a battery with itself on the first and after times of use. My
trough of twenty pairs of four-inch plates, charged with acid consisting of
200 water, 4-1/2 oil of vitriol, and 4 nitric acid, lost, upon the first
time of being used, 2.82 equivalents per plate. When used after the fourth
time with the same charge, the loss was from 3.26 to 4.47 equivalents per
plate; the average being 3.7 equivalents. The first time the forty pair of
plates (1124.) were used, the loss at each plate was only 1.65 equivalent;
but afterwards it became 2.16, 2.17, 2.52. The first time twenty pair of
four-inch plates in porcelain troughs were used, they lost, per plate, only
3.7 equivalents; but after that, the loss was 5.25, 5.36, 5.9 equivalents.
Yet in all these cases the zincs had been well-cleaned from adhering
copper, &c., before each trial of power.

1147. With the rolle