THE CARBON MICROPHONE 181 



simply interpreted (just as Gray had found) and, because of the 

 possibility that unknown cohesional or frictional forces were involved, 

 the work was extended by a study of resistance-displacement charac- 

 teristics. Through a comparison of the two sets of data we were 

 led to the conclusions that the stress-strain characteristics are not so 

 simple as those assumed in Pedersen's or Gray's analysis and, there- 

 fore, that a study of the elastic behavior of contacts offered the most 

 promising line of attack on the problem. 



Figure 17 shows the mechanical system developed for this purpose. 

 With it known forces can be applied to a contact element and at the 

 same time its movement can be measured. 



The contact is made between a carbon granule and a polished 

 carbon plate, the granule being attached to the end of a rod R sus- 

 pended by springs 5' from a fixed frame and the plate being attached 

 to the end of a micrometer screw M2 capable of giving to it a transla- 

 tional motion without rotation. 



The force is applied to the granule electrostatically by means of 

 voltage applied between the condenser plates C2, one of which is 

 attached to the rod R and the other to the micrometer screw. This 

 is in principle the attracted disc electrometer of Kelvin and it is 

 capable of applying forces up to 15 dynes without using voltages 

 greater than 200. 



The motion of the granule with respect to the carbon plate is 

 measured electrically through the variation of capacity of the con- 

 denser Ci, of which one plate is attached to the other end of the rod R. 

 Ci forms part of an oscillating circuit of natural frequency Wo (about 

 2000 kc.) which is coupled to a wave-meter circuit adjusted for oscilla- 

 tion at a frequency «i slightly different from n^. Changes in the 

 frequency arising from the changes in capacity C\ alter the energy 

 picked up by the wave-meter circuit and this energy, which is recorded 

 by means of a galvanometer, serves as a measure of the change of 

 capacity or motion of the rod R. With this arrangement it is possible 

 to measure motions as small as 1 X 10"'^ cm. and under the best con- 

 ditions as small as 1 X 10~^ cm. It is necessary to have good damping, 

 which is obtained by means of immersing the drum D in polymerized 

 castor oil. The accessory spring ^2 is used merely for calibrating 

 purposes. 



Figure 18 shows the appearance of the apparatus as set up for 

 measurement. The condenser is contained in the lower housing at 

 the left, the wave-meter in the upper housing. The whole apparatus 

 including the galvanometer is supported on a delicate spring suspension 

 within a second large lead container, the frame of which just appears 

 at the edge of the photograph and which is also supported by springs. 



