ery deep well--the power which its elevation indicates
must be only a small fraction of that enclosed in the air reservoir.

It will be one great point in favour of compressed air, as a form of
stored energy for the special purpose of pumping, that by making a
continuous small flow of air take the place of the water at the lowest
level in the upward pipe, it is possible to cause it to do the pumping
without the intervention of any motor.

One means of effecting this may be simply indicated. The air under
pressure is admitted from a very small air pipe and the bubbles, as
they rise, fill the hollow of an inverted iron cup rising and falling
on a bearing like a hinge. Above and beneath the chamber containing
this cup are valves opening upwards and similar to those of an
ordinary force or suction pump. The cup must be weighted with
adjustable weights so that it will not rise until quite full of air.
When that point is reached the stroke is completed, the air having
driven upwards a quantity of water of equal bulk with itself, and, as
the cup falls again by its own weight, the vacuum caused by the air
escaping upwards through the pipe is filled by an inrush of water
through the lower valve. The function of the upper valve, at that
time, is to keep the water in the pipe from falling when the pressure
on the column is removed. The expansive power of the air enables it to
do more lifting at the upper than at the lower level, so that a larger
diameter of pipe can be used at the former place.

Cheap motors working on the same principle--that is to say through the
upward escape of compressed air, gas or vapour filling a cup and
operating it by its buoyancy, or turning a wheel in a similar
manner--will doubtless be a feature in the machine work of the future;
and for motors of this description it is obvious that compressed air
will be very useful as the form of power-storage. Excepting under
very special conditions, steam is not available for such a purpose,
seeing that it condenses long before it has risen any material
distance in a column of cold water.

"The present accumulator," remarked Prof. Sylvanus P. Thompson in the
year 1881, referring to the Faure storage batteries then in use,
"probably bears as much resemblance to the future accumulator as a
glass bell-jar used in chemical experiments for holding gas does to
the gasometer of a city gasworks, or James Watt's first model
steam-engine does to the engines of an Atlant