All who have learned a little of chemistry doubtless remember the
experiment with vortex rings produced by phosphorus trihydride mixed
with a little phosphide of hydrogen. As this curious phenomenon
evidently does not depend upon the peculiar properties of this gas, I
have been trying for some time to reproduce it by means of tobacco
smoke, and even with chemical precipitates, which are, in a way,
liquid smoke. After a few tentatives made at different times, my
experiment succeeded perfectly. The following is, in brief, the mode
of operating:

Take up a little hydrochloric acid in a pipette and put a few drops of
it into a very dilute solution of nitrate of mercury, and you will
obtain rings of mercurial chloride that will, in their descent, take
on the same whirling motion that characterizes the aureolas of
phosphureted hydrogen.

The drops of acid should be allowed to fall slowly, and from a feeble
height, to the surface of the liquid contained in the vessel. It is
unnecessary to say that the result may be obtained through the use of
other solutions, provided that a precipitate is produced that is not
very thick, for in the latter case the rings do not form. If need be,
we may have recourse to milk, and carefully pour a few drops of it
into a glass of water.

[Illustration: FIG. 1.--PRODUCTION OF SMOKE RINGS.]

As regards smoke rings, it is easy to produce these by puffing cigar
smoke through a tube (Fig. 1). But, in order to insure success, a few
precautions are necessary. The least current of air must be avoided,
and this requires the closing of the windows and doors. Moreover, in