324 Prof. J. J. Thomson on the Discharge of Electricity 



describes the light produced in a tube round which the wire 

 connecting the coatings of a Leyden jar is twisted ; the lumi- 

 nosity in Hittorf s experiments seems to have filled the tube, 

 and not, as in the experiments described in this paper, been 

 confined to a ring. It seems possible that the difference in 

 the appearance in the tubes may have been due to the exist- 

 ence of an electrostatic action in Hittorfs experiments, the 

 primary coil getting raised to a high potential before the 

 discharge of the jar, and inducing a distribution of electricity 

 over the inside of the glass of the tube ; on the passage of the 

 spark the potential of the primary coil will fall, and the 

 electricity on the glass redistribute itself ; to effect this re- 

 distribution it may pass through the rarefied gas in the 

 discharge-tube and produce luminosity. 



In my experiments I took two precautions against this 

 effect : in the first place I connected the primary coil to earth, 

 so that its potential before discharge took place was unaltered, 

 and as an additional precaution I separated the discharge- 

 tube from the primary by a cage made of blotting-paper 

 moistened with dilute acid ; the wet blotting-paper is a suffi- 

 ciently good conductor to screen off any purely electrostatic 

 effects, but not a good enough one to interfere to an appre- 

 ciable extent with the electromotive forces arising from 

 rapidly alternating currents. In this way we can screen off 

 any electrostatic effects due to causes which operate before 

 the electrical oscillations in the jars begin ; when once these 

 have commenced, there ought not, I think, to be any separa- 

 tion of the electromotive forces into two parts, one being- 

 called electrostatic, the other electrodynamic. As this is a 

 point on which it is desirable to avoid any misunderstanding, 

 I hope to be excused if I treat it at some length. 



In the mathematical treatment of the phenomena of the 

 " Electromagnetic Field," it is customary and not incon- 

 venient to regard the electromotive force as derived from two 

 sources, or rather as consisting of two parts, one part being 

 calculated by the ordinary rules of electrostatics from the 

 distribution of electricity in the field, the other part being the 

 differential coefficient of the vector potential with respect to the 

 time. From a mathematical point of view, there is a good 

 deal to be said for this division ; the two forces have very 

 distinct and sharply contrasted analytical properties. Thus the 

 electrostatic force possesses the property that its line-integral 

 taken round any closed curve vanishes, while the surface- 

 integral of its normal component taken over a closed surface 

 does not in general vanish. The " vector potential force/' on 

 the other hand, does not in general vanish when integrated 



