under unequal heating.[2] This explanation has recently been called in
question by Mr. Preece,[3] who has expressed the opinion that
although vibrations may be produced in the disks by the action of the
intermittent beam, such vibrations are not the cause of the sonorous
effects observed. According to him the aerial disturbances that produce
the sound arise spontaneously in the air itself by sudden expansion due
to heat communicated from the diaphragm--every increase of heat giving
rise to a fresh pulse of air. Mr. Preece was led to discard the
theoretical explanation of Lord Raleigh on account of the failure of
experiments undertaken to test the theory.

[Footnote 1: Amer. Asso. for Advancement of Science, August 27, 1880.]

[Footnote 2: _Nature_, vol. xxiii., p. 274.]

[Footnote 3: Roy. Soc., Mar. 10, 1881.]

[Illustration: Fig. 1. A B, Carbon Supports. C, Diaphragm.]

He was thus forced, by the supposed insufficiency of the explanation, to
seek in some other direction the cause of the phenomenon observed, and
as a consequence he adopted the ingenious hypothesis alluded to above.
But the experiments which had proved unsuccessful in the hands of Mr.
Preece were perfectly successful when repeated in America under better
conditions of experiment, and the supposed necessity for another
hypothesis at once vanished. I have shown in a recent paper read before
the National Academy of Science,[1] that audible sounds result from the
expansion and contraction of the material exposed to the beam, and that
a real to-and-fro vibration of the diaphragm occurs capable of producing
sonorous effects. It has occurred to me that Mr. Preece's failure to