solution and porous vessel.

[Illustration: Fig. 3.--COURTOT'S ARRANGEMENT OF THE BICHROMATE PILE.]

The whole is covered with a sheet of ebonite connected with the zinc and
the two carbon plates in such a way that when the pile is not in operation
the whole can be lifted from the liquid. Under such circumstances the
deposit of oxide is notably diminished, and the duration of the discharge
is consequently greatly increased.

Fig. 3 shows the details of a windlass that permits of lifting, according
to circumstances, all the elements of the same trough or only a part of
them. To effect this, the drum around which the chain winds that carries
the carbons is mounted upon a sleeve fixed upon the axle. This latter is
actuated by a winch; and a ratchet wheel, R, joined to a click which is
actuated by a spiral spring, prevents the ebonite plates from falling back
when it is desired to place the bolt under the button, B, of the spring.

When it is desired to put an element out of the circuit, it is only
necessary to act with the finger upon the extremity of the lever, D. Under
the action of the latter, the piece, _s_, which carries a groove for the
passage of the screws that fix it to the upper cross-piece, takes on a
longitudinal motion and consequently gears with the drum through the
toothed sleeve, E. When an experiment is finished the zinc may thus be
lifted from the liquid, and the deposit of oxide be prevented from forming
upon the carbon. As may be seen, the arrangements which we have just
described exhibit nothing that is particularly original. The windlasses
used for removing the elements from a pile when the circuit is open have
been employed for a long time; the bichromate pile is itself old, and, as
we said in the beginning, it has been modified in its details a number of