THE CARBON MICROPHONE 175 



which is just on the limit of resolution of the highest-power micro- 

 scopes. It follows from a consideration of the number of granules in 

 series that the movement between centres of granules w^ould not be 

 greater than 1/lOth of this, viz., 



1 X 10-« cm., ' 



which is in the submicroscopic range. We must, therefore, be able 

 to control and measure movements at least as small as 10"'^ cm.; not 

 an easy thing to do with a "loose contact." 



The contact forces are on the average somewhat less than 10 dynes 

 when the aggregate is in the unagitated state. In the presence of 

 acoustic waves, variable forces of several dynes are superimposed on 

 these fixed forces. The variable forces are smaller than the fixed 

 forces, so that the granules will on the average remain in contact 

 throughout any reversible cycle. We have reason to believe that 10 

 dynes is about the maximum force which is attained at any one 

 contact during a stress cycle. We must therefore be able to control 

 contact forces within the range 1 to 10 dynes. 



Apparatus and technique have now been developed for studying 

 single contacts within the prescribed range of forces and displace- 

 ments, and significant measurements have been made which I will 

 now endeavor to describe to you somewhat in detail. 



Single Contact Studies 



Figure 12 shows the construction of one of the contact tubes used 

 in this study. 



Its essential features are shown diagrammatically in Fig. 13. The 

 contact pieces C\ and Ci are fastened respectively to a movable base 

 M and to the lower end of a helical spring made of fused quartz. The 

 base is supported from a fixed frame by two vertical platinum wires 

 P and two stretched springs as shown. The lower contact piece is 

 moved by heating or cooling the platinum wires through the passage 

 of current. In this way the contacts may be made or broken and 

 any desired contact force applied, the measure of the force being the 

 compression of the helical spring. The temperature of the contact 

 is varied by surrounding the contact region with a metal cylinder 5 

 which may be heated by means of radiation from a coil of platinum 

 wire H, the temperature within the cylinder being measured by means 

 of a thermocouple placed near the contacts. 



In practice the upper contact piece consists of a single granule 

 fastened to the end of a platinum wire and the lower contact piece 

 consists of a number of granules attached to a horizontal metal plate; 



