1382 THE BELL SYSTEM TECHNICAL JOURNAL, NOVEMBER 1954 



tion function / should be unique if plotted against NvfF^ where the a 

 and h are constants to be determined experimentally. 



Experiment 



Cotton lint fibres and clean contact surfaces were used exclusively in 

 this study. The fibres were essentially in the form of ribbons a few 

 microns thick and of variable width and length. By using a dust sepa- 

 rator,* lint fibres A\ith a well controlled size distribution were collected 

 on a glass plate. The setting used gave a rather uniform monolaj^er of 

 fibres 80 per cent of which had a A\idth between 10 and 20 microns. 



The contacts tested were flat and made of palladium, f They were 

 cleaned with methyl alcohol and distilled water, then dried. The col- 

 lected lint fibres were transferred to the surface of one contact by a 

 special adapter which allows the pressing of the contact on the glass 



LU 0.60 



a. 



< 



LU 



do.50 



z 

 o 



0.1 o9- 



10 10"^ 



NUMBER OF OPERATIONS, N 



Fig. 6 — Effect of force. 



* Based on controlling sedimentation by adjusting air speed in a two-stage 

 separator. 



t Contact surface roughness was controlled by frequently polishing the contact 

 surfaces by a fixed process. 



