cause of all these things. Since, moreover, sacrifices had to be made to
the gods in order to obtain benefits, the flow of milk, wine, or water,
as well as the hissing of the serpent, when the sacrificial flame was
lighted, appeared to demonstrate clearly the existence of the gods."

In another analogous apparatus of Heron's, it is steam that performs the
role that we have just seen played by dilated air. But the ancients do
not appear to have perceived the essential difference, as regards motive
power, that exists between these two agents; indeed, their preferences
were wholly for air, although the effects produced were not very great.
We might cite several small machines of this sort, but we shall confine
ourselves to one example that has some relation to our subject. This
also is borrowed from Heron's _Pneumatics_. (Fig. 3.)

"Fire being lighted on an altar, figures will appear to execute a round
dance. The altars should be transparent, and of glass or horn. From the
fire-place there starts a tube which runs to the base of the altar,
where it revolves on a pivot, while its upper part revolves in a tube
fixed to the fire-place. To the tube there should be adjusted other
tubes (horizontal) in communication with it, which cross each other
at right angles, and which are bent in opposite directions at their
extremities. There is likewise fixed to it a disk upon which are
attached figures which form a round. When the fire of the altar is
lighted, the air, becoming heated, will pass into the tube; but being
driven from the latter, it will pass through the small bent tubes and
... cause the tube as well as the figures to revolve."

Father Kircher, who had at his disposal either many documents that we
are not acquainted with, or else a very lively imagination, alleges
(_Oedip. AEg._, t. ii., p. 338) that King Menes took much delight in