nators vibrates in response, and the apparatus begins to rotate. Rotation will take place even if there is only one resonator, properly balanced; but the phenomenon is more marked and certain if there are four.
A second style of apparatus—the "rotating resonator"—is represented in plan and elevation in Fig. 2. It consists of a short cylindrical box of stiff paper, having four projections, each of which bears at its side a short open tube of paper. It is hung on a silk fiber, and is supplied with a small needle, projecting below to steady the motion during its rotation.
The operation of these instruments depends on the principle which has been pointed out by Lord Rayleigh and Professor Mayer as well as by Professor Dvorák, that "when sounds of great intensity are produced, the calculations, which are usually carried only to the first order of approximation, cease to be adequate, because now the amplitude of motion of the particles in the sound-wave is not infinitely small as compared with the lengths of the sound-waves themselves. Mathematical analysis shows that under these circumstances the wave of the pressures in the condensed part, and in the rarefied part of the sound-wave, is no longer equal to the undisturbed atmospheric pressure, but is always greater. Consequently, at all nodal points in the vibrations of the air in tubes or resonant boxes, the pressure of the air is greater than elsewhere; and therefore any resonator closed at one side and open at the other is urged along bodily by the slight internal excess of pressure on the closed end." The apparatuses therefore rotate by reaction.
To produce vibrations of sufficient intensity. Professor Dvorák uses heavy tuning-forks mounted on resonant cases, and excited electrically. For this purpose he places between the prongs of the fork an electro-magnet, in which the core is composed of two plates of iron, separated by a sheet of paper, and cut of such a breadth as to lie between the prongs without touching them. The core is overwound with insulated copper wire, as shown at E, Fig. 3, and the electro-magnet is mounted by a bent piece of wood, a b c, upon the sounding-box, K, of the fork. The wires are connected in a circuit with the battery, and with the electro-magnet of a self-exciting tuning-fork of the same note. The sounding-boxes of the forks must not touch the table, but the arm a b c is clipped at about the point b in a firm support; and particular care must be taken to have the wood of the resonant boxes tuned into exact accord with the tone of the fork and of the air within the cavity of the box.
The third apparatus is the "sound-radiometer," and was described by Professor Dvorák before the Imperial Viennese Academy in 1881. It is more simple than the two instruments previously described, but its cause of action is less easily explained. It is shown in Fig. 4. It consists of a light cross of wood pivoted by a glass cap upon a vertical
