M. Dvorak on Acoustic Repulsion. 229 



3. The Acoustic Torsion-balance. — If we hang by a wire a 

 wooden rod provided with a resonator, like the beam of a 

 Coulomb's torsion-balance, in a case having an opening in the 

 side turned toward the resonator, we can compare the inten- 

 sity of notes having an equal number of vibrations by means 

 of the repulsion of the resonator ; but further experiments are 

 necessary to test the practicability of this method. The sound 

 proceeded from an open pipe, having the note A (of 435 vi- 

 brations). To prevent the current of air which passes through 

 the pipe from striking the resonator attached to the balance, 

 we must cut the pipe exactly in the middle of its node, and 

 insert a slack membrane softened with glycerine. To prevent 

 the air issuing from the mouth of the pipe from impinging on 

 the resonator, a broad box is used which surrounds the mouth 

 of the pipe air-tight. This box is open on the side opposite 

 the resonator, so as not to impair the tone. The pipe is 

 sounded by means of a Konig's acoustic bellows with a uni- 

 form blast of air. The distance of the resonator from the 

 mouth of the pipe must be at least 2 or 3 centims., to avoid a 

 change of pitch. 



4. Production of Aerial Currents by Sound. — It may easily 

 be proved by simple theoretic considerations that the mean 

 pressure at the node of a column of air is greater than at its 

 ventre ; and that it steadily diminishes in passing from the 

 node to the ventre, provided that the amplitude of vibration 

 is not infinitely small. 



It would seem that this difference of pressure would be 

 neutralized by the passage of the air from the node to the 

 ventre. There would then be produced a mean pressure in 

 the whole column, which would be greater, however, than that 

 of air at rest ; consequently air would issue from the opening 

 of the vessel in which it forms stationary waves. I have not 

 succeeded, so far, in making the whole process clear ; for in 

 reality no perfect balance of pressure takes place. The mano- 

 meter always shows a slight excess of pressure even at the 

 ventre ; but this excess increases as we pass to the node. All 

 my previous experiments indicate moreover that a current of 

 air passes from the node to the ventre, at least in Kundt's 

 tube, in which the air-waves are very powerful. This prin- 

 cipal current lasts as long as the air vibrates. Besides, the 

 same experiments show a continuous secondary current, close 

 to the walls of the tube and in a direction contrary to that of 

 the principal current ; so that the whole air in the tube is in 

 circulation. The cross section of the principal current is nearly 

 as great as that of the tube, while that of the secondary cur- 

 rent is a very narrow ring. 



