182 Mr. A. Schuster on the Disruptive Discharge 



are proportional to the currents induced in the movable 

 wire, and hence represent the intensity of field traversed by 

 the wire in the successive positions which it occupies around 

 its axis. The time-intervals between the ordinates corre- 

 sponding to the dots forming the curve are hundredths of 

 a second. 



XIX. The Disruptive Discharge of Electricity through Gases. 

 By Arthur Schuster, F.R.S.* 



npHE phenomena connected with the disruptive discharge 

 JL of electricity through gases present some special diffi- 

 culties which have hitherto withstood all efforts to form a 

 consistent theory. Some time ago f I presented to the Royal 

 Society an outline of a general theory of electric discharge, 

 which has been so favourably received by nearly every one 

 who has been working at the subject \ since, that I am en- 

 couraged to trace out its consequences in various directions. 



Prof. Chrystal § concludes an interesting paper on the dis- 

 ruptive discharge as follows : — 



" With the same end in view, I think it would be desirable 

 to reduce the experiment of Bailie and others with spheres, in 

 cases where the surface-density can be calculated, and to 

 examine the values of the dielectric strength in these cases. 

 Until this or something equivalent is done, it is clear that we 

 have really no experimental ground for asserting the exist- 

 ence of such a constant as the dielectric strength of a medium, 

 and therefore cannot take the very first step towards a physical 

 theory of the disruptive discharge, which appears to me to be 

 the next great advance to be made in the science of electricity." 



The present paper is divided inio two parts. In the first 

 place I have reduced the numerical results obtained by Bailie 

 and Paschen for the discharge between spheres. The result 

 quite justifies Prof. Chrystal's caution, for it appears that the 

 so-called dielectric strength of a medium depends on so many 

 circumstances that there seems to be no sufficient reason to 

 consider it at all as a special property of the medium, except, 

 perhaps, in the case of a perfectly uniform field. In the 



* Communicated by the Author. 



t Proc. Roy. Soc. xxxvii. p. 317 (1884). 



X Dessau, Wied. Ann. xxix. p. 362 (1886) ; Warburg, Wied. Ann. 

 xxxi. p. 545 (1887) ; Elster and Geitel, Wied. Ann. xxxvii p 315 

 (1889). r ' 



§ Proceedings Royal Society Edinburgh, 1881-82, p. 487. 



