ACOUSTICS AND GRAVITATION. 43 



this reason it is much more prominent for the wider quill-tube than for the 

 aluminum probe. Vibration is thus communicated through the pin-hole and 

 favored by width of probe-tube. 



The work with these resonators has not been completed, but it is already 

 obvious that the smaller resonators must be regarded as adjutages or flares 

 of the necks connecting them with the telephone. By making the connector 

 very short, for instance, the resonator, which quite failed to respond before, 

 gave me a strong %c" maximum, with no fringe displacement at /". 



37. Correlative experiments 1 with the torsion balance. The following 

 pretty experiment is very instructive in its bearing on the Mayer-Dvorak 

 effect, as well as on the experiments of the present paper. In figure 66, b is 

 the light wooden beam (30 cm. long, counterpoised at a) of a horizontal 

 torsion balance, the torsion wire (of brass 0.02 cm. in diameter and 18 cm. 

 long on either side normal to the diagram) being seen at w. A light disk of 

 cardboard d is suspended in equilibrium from the end of the balance. Below 

 this is the telephone T, to which the brass pipe p (13 cm. in length and 2.6 cm. 

 in diameter) has been cemented, to the form of a closed c" organ-pipe, of which 

 the telephone plate is the bottom. The open top of p is surrounded by a 

 fixed annular disk cc of metal parallel and close to the movable disk d. 



When the telephone is strongly energized and emits a rising note (motor- 

 break and rheostat) no effect is produced until its frequency is in resonance 

 with the pipe p, whereupon the disk d is at once attracted. Since the pipe 

 p is closed above by this process, the telephone frequency must be slightly 

 reduced to keep the disks in cohesion. On breaking the current, d is at once 

 released . 



This is, of course, nothing further than a modified example of the familiar 

 pneumatic paradox. When the pipe howls, the distance from which d 

 may be attracted and held is perhaps 2 cm., beyond which the couche of di- 

 minished static pressure is ineffective. The thickness vanishes with the 

 intensity of sound. 



If now cc is removed and the disk d is replaced by the closed paper cylinder 

 e of a diameter (2.1 cm.) sufficiently small to enter the mouth of p easily, the 

 results of the experiment are the same. Here, however, the cylinder e may be 

 made to enter the pipe as much as i cm. or more by successively decreasing 

 the pitch conformably with the gradually stopped mouth of p. Supposing 

 the total displacement to be 2 cm., the force indicated by the torsion balance 

 would be 0.7 dyne and the mean pressure decrement for the area 3.5 cm. 2 , 

 therefore 0.2 dyne /cm. 2 . But as both the disk and cylinder come down with 

 a jerk, the maximum forces are probably larger. 



If there were a pin-hole in the bottom of e, the density of air contained would 

 tend to increase and the cylinder would fall for this reason also, but the present 

 experiment is relatively too crude to show this. For the content of the cylinder 



Reference to the allied experiments of Herbert G. Dorsey (Phys. Review, xv, p. 512) is 

 in place here. 



