338 



Dr. P. E. Shaw. The Sparking Distance [Mar. 22, 



put forward, for the film would increase and decrease in thickness 

 directly with pressure. 



W. R. Carr* made similar investigations, but as he worked with 

 reduced pressures and his discharge distance was never less than 1 mm., 

 whereas mine is never greater than 1 /x, our work does not overlap. 



Suppose two surfaces having a constant potential difference (V) 

 approach one another, the tension between the charged surfaces at last 

 breaks down the medium when the distance apart is x. The space 

 between the contact surfaces may be occupied by one or more media, 

 each medium having dielectric strength, either constant or varying 

 from point to point. Two cases concern us — 



1. One medium, uniform in dielectric strength; we can express the 

 critical potential in terms of potential gradient. The curve between 

 V and x is a straight line to the origin. 



2. Two or more media uniform in dielectric strength. The curve 

 is a succession of straight lines. 



Referring to fig. 2 it will be seen that Earhart's curve is composed of 

 two straight portions. Hence on the above hypothesis the water film 

 has a constant dielectric strength and so has the air outside, the 

 strength of the former being about thirty times that of the latter, 



Fig. 1. 



* ' Phil. Trans. ' yol. 201, 1903. 



