I9I5.] SURFACES OF TELEPHONIC DIAPHRAGMS. 117 



Temperature Effects. 



It was found that changes of temperature in the air surround- 

 ing a diaphragm had a marked effect, both upon its resonance fre- 

 quency, and upon its amplitudes at any frequency. The curves rep- 

 resenting zv against r^ were apt to differ appreciably in outline from 

 day to day. The degree of tightness of clamping also had a marked 

 effect in these measurements. In general, such disturbances due to 

 temperature and clamping, are likely to introduce tensions in the 

 substance of the diaphragm, and to cause some of the characteris- 

 tics of vibrating membranes to be superposed upon those of a vi- 

 brating plate. It is, therefore, desirable that the clamping should 

 be effected tightly, and that the measurements should then be made 

 before the temperature has changed. Strictly speaking, the Ray- 

 leigh theory shows that there must be a marked difference in both 

 the resonance frequency and in the distribution of amplitudes, if 

 the diaphragm is clamped between circular knife edges, instead of 

 between circular flat rings at the boundary. The experiments have 

 shown that flat-ring clamping is more likely to give consistent re- 

 sults than knife-edge clamping. These clamping difficulties are 

 accentuated in thin glass diaphragms, for the boundary supporting 

 of which, a special technique had to be developed. 



Exploration of Thin Glass Diaphragms. 



From a number of thin glass diaphragms, one Diaphragm No. 4, 

 was selected, on account of its uniformity in thickness. See Table 

 III. It was found very difficult to obtain uniform results with 

 this in the explorer, owing to the above mentioned troubles with 

 clamping. Finally, the glass diaphragm was cemented, with water 

 glass, to a boundary ring of glass, and this was lightly supported be- 

 tween the clamping rings of the explorer. The diaphragm was 

 then excited acoustically by organ-pipes. The natural pitch of the 

 diaphragm was found to be 492 — ', in the fundamental mode. On 

 raising the frequency, the mode of motion was found to change sud- 

 denly, at 968 ^~-', to that of a single nodal diameter, the two halves 

 of the diaphragm then vibrating harmonically in opposite phases. 

 This mode of motion continued until the frequency reached 1,696 '— ', 



