A HOLLOW globe, or other vessel, may be constructed, into which if any liquid be poured, it will be forced aloft spontaneously and with much violence, so as to empty the vessel, though such an upward
motion is contrary to nature. The construction is as follows.
Let there be a globe, containing about 6 cotylae (3 pints),
the sides of which are of metal plate, strong enough to sustain the
pressure that will be exerted upon them by the air. Let A B (fig.
9.) be the globe, resting on any base C.
Through an aperture in the top of the globe insert a tube, D E, soldered
into the globe at the aperture, and projecting a
little above it; and reaching to the other extremity, except an
interval sufficient for the passage of water. At its upper extremity
let the tube D E branch into two tubes, D G and D F, to which two
other pipes, G H K L, F M N X, are fastened transversely,
communicating with D G, D F. Again, into these transverse pipes, and
communicating
with them, let another pipe, P 0, be fitted, from which a small pipe, R
S, projects perpendicularly, communicating with it, and terminating in a
small orifice at S. If, then, we take hold of R S and turn round the tube
P 0, the connection between the corresponding holes will be shut off, so
that the liquid which is to be forced up will have no outlet. Now,
through another aperture in the globe, let another tube, T U Q, be
inserted, closed at the lower extremity Q, and having a hole in the side
near the bottom at W. In this hole must be fixed a valve, such as the
Romans call assarium, the construction of which we will explain
presently. Into the tube T U Q insert another tube, Y Z, fitting
tightly. If the tube V Z be drawn out, and water poured into T U Q, it
will enter the vessel through the aperture W, (the valve opening into the
interior of the vessel), and the air will escape through the pipe 0 P,
which communicates, as we have explained, with the apertures of the
pipes G H K L aml F M N X. When the globe is half full of liquid, turn
the small tube N S so as to break the connection between the
corresponding apertures: then depress the tube Y Z and drive out the air
and liquid collected in T U Q, which will, on exertion of some force,
(as the vessel is full of air and liquid), pass through the valve into
the hollow of the globe; and this passage is made possible by the
compression of the air into the void spaces dispersed among its
particles. Draw up the tube Y Z, in order again to fill T U Q with air,
and then, depressing it again, we shall force this air into the globe.
By repeating this frequently we shall have a large quantity of air
compressed into the globe; for it is clear that the air forced in does
not escape again when the rod is drawn up, as the valve, pressed on by
the air within, remains closed. if, then, we restore the pipe R S to
its up-right position, and re-open the communication between the
corresponding apertures at L and X, the liquid will now be forced out,
as the condensed air expands to its original bulk and presses on the
liquid beneath; and if the quantity of condensed air be large, it will
drive out all the liquid, and even the superfluous air will be forced
out at the same time.