1904.] Sparking Distances between Charged Surfaces. 337 



" The Sparking Distance between Electrically Charged Surfaces. — 

 Preliminary Note." By P. E. Shaw, B.A., D.Sc. Communi- 

 cated by Professor J. H. Poynting, F.K.S. Eeceived 

 March 22,— Eead April 28, 1904. 



Lord Kelvin* first made systematic measurement of the relation 

 between potential difference and sparking distance. He noticed that 

 these factors do not vary in proportion and he surmised that this might 

 be due to " the air near dense bodies being condensed and so becoming 

 a better insulator." 



Later G. A. Liebigf published results bringing voltages used to as 

 low as 800, the discharge distance being 67 micra. Both the above 

 observers used frictional machines to produce E.M.F. and worked in 

 electrostatic units. 



E. F. Earhart| made a new departure, using a continuous current from 

 accumulators, and obtained consecutive results with voltages from 1000 

 to 38 ; the corresponding sparking distances, ranging from 100 /x to \ /x, 

 were measured by Michelson's interferometer. 



One very interesting result of his research was that the potential 

 gradient, dv/dx, suddenly changes when the sparking distance is about 

 2 /x (see fig. 2) ; for greater distances dv/dx is 7 (volts per micron) 

 whereas for less distances it is 200. Thus a distinct "knee" is formed 

 in the curve connecting V and x. I have suggested an explanation of 

 this.§ 



It has been shown that the thickness of the condensed water film on 

 solid surfaces at ordinary pressure and temperature is about 0'8 /x. If 

 two surfaces approach one another to a distance of 2 /x then, deducting 

 0*8 for the film on each surface, we have an air distance left of only 

 0*4 /x, which would be easily bridged by small vibrations of the surfaces, 

 and the bridge would be stable on account of capillarity. Thus any 

 discharge would take place through the water film and not through 

 air. The film on recently cleaned metal would be very pure and would 

 have a dielectric strength greater than air (from the numbers quoted 

 it appears to be thirty times as great), hence the sudden rise in dv/dx. 



Mr. Earhart tried the effect of increasing and decreasing the 

 atmospheric pressure. Increase in pressure causes small increase 

 in dv/dx below the "knee" and large increase above the "knee"; 

 but what is more important to notice is that increase in pressure 

 causes the "knee" point to move up the curve corresponding to 

 greater sparking distance. These last results support the theory 



* Thomson, 1 Boy. Soc. Proc.,' 1860. 



t 'Liebig, 'Phil. Mag.,' vol. 24, 1887. 



X Earhart, ' Phil. Mag.,' vol. 1, 1901. 



§ Parks, ' Phys. Soc. Proc./ Discussion, p. 418, 1903. 



