﻿Liquid Films letween Iridio-platiniim .Electrodes. 327 



7. The contact sound is invariably clearer and sharper for 

 litj vi ids ihan for air. and discharge is more definite. Air 

 discharge often occurs partial!}', i. e. there is a rush of material 

 from th^ calho le an I the contacts approach one another but 

 do not touch. Discharge does not continue; in fact, thi 

 partial discharge seems to retain the tension and for a time 

 stop the complete and final breakdown. 



1 heory- 



Very short spark -lengths in gases have now been studied 

 by Earhart, Shaw, Kinsley, and Hobbs. The accumulated 

 results of these and the present experiments on short spark- 

 lengths in liquids should lea 1 inductively to a theory of the 

 con itions of matter at and between the discharge surfaces. 

 Recently T. Schwedoff (Ann. d. Physik, pp. 918-934, April 

 1906) has propounde I a ballistic theory of the spark-discharge. 

 The fun lamental principle he uses is that the discharge 

 between the surfaces is due to the shattering of the molecules 

 of the gas into free ions ; this occurs when the electric 

 strength between the electrodes jointly with the kinetic 

 energy of the ions already existing free, reaches sufficient 

 maonitu le. He uses the ballistic law of resistance, in which 

 the fractional force opposing the passage of an ion depends 

 on its -peed and on the density of the gas. An expression is 

 finally uerived 



3007/n . 10" 4 . 1 





c 



1+^(1--") 



where v = potential difference on the surfaces ; 

 1 = spark-length ; 

 g = .lensity of the gas ; 

 /t , c, X are constants. 

 The values of these constants are derived from consideration 

 of the experimental results of Earhart ami Shaw. Putting 

 in the constants and the density of the gas, the P.D. required 

 fur any given spark-length is derived. 



Schwedoff thus obtains a curve which fairly closely follows 

 that of Earhart, having a knee in the curve at voltage 

 about 350. 



At first sight the similarity between Earhart' s experimental 

 curve and Schwedoff's theoretical curve appears to be a 

 striking confirmation of the theory; but careful comparison 

 shows that the two curves are far from coincident near the 



