follow them. (I should explain
that for a reason which will appear later, I made a loud note by
whistling into a key at the time that this photograph was taken.)

[Illustration: Fig. 39.]

[Illustration: Fig. 40.]

Lord Rayleigh has shown that in a stream of water one twenty-fifth of an
inch in diameter, necks impressed upon the stream, even though
imperceptible, develop a thousandfold in depth every fortieth of a
second, and thus it is not difficult to understand that in such a stream
the water is already broken through before it has fallen many inches. He
has also shown that free water drops vibrate at a rate which may be
found as follows. A drop two inches in diameter makes one complete
vibration in one second. If the diameter is reduced to one quarter of
its amount, the time of vibration will be reduced to one-eighth, or if
the diameter is reduced to one-hundredth, the time will be reduced to
one-thousandth, and so on. The same relation between the diameter and
the time of breaking up applies also to cylinders. We can at once see
how fast a bead of water the size of one of those in the spider's web
would vibrate if pulled out of shape, and let go suddenly. If we take
the diameter as being one eight-hundredth of an inch, and it is really
even finer, then the bead would have a diameter of one sixteen-hundredth
of a two-inch bead, which makes one vibration in one second. It will
therefore vibrate sixty-four thousand times as fast, or sixty-four
thousand times a second. Water-drops the size of the little beads, with
a diameter of rather less than one three-thousandth of an inch, would
vibrate half a million times a second, under the sole influence of the
feebly elastic skin of water! We thus see how powerful is the influence
of the feebly elastic water-skin on drops of water that are sufficiently
small.

I shall now cause a small fountain to play, and shall allow the water as
it falls to patter upon a sheet of paper. You can see both the fountain
itself and its shadow upon the screen. You will notice that the water
comes out of the nozzle as a smooth cylinder, that it presently begins
to glitter, and that the separate drops scatter over a great space (Fig.
41). Now why should the drops scatter? All the water comes out of the
jet at the same rate and starts in the same direction, and yet after a
short way the separate drops by no means follow the same paths. Now
instead of explaining this, and then showing experiment