'.''. Prof. J. ,J. Thomson. On the Propagation of the [Jan. !.">. 



ury is in the tube is something of the following kind. The 

 positive electricity rushes from the anode down the tube until it 

 reaches the first mercury plug; it attracts the negative electricity to 

 the nearer end of this plug, and repels the positive to the other end ; 

 this positive electricity begins to leave the mercnry immediately a?id 

 travels down to the next mercury ping. 



The positive electricity which travels up to the first mercury plug 

 and the negative electrification it produces on the mercury form an 

 i ical double layer which takes some time to disappear, longer 

 probably than the time taken by the electricity to travel from one 

 end of the tube to the other. The time the luminosity takes to travel 

 from B to G will thus not be much affected by the mercury plugs; 

 but, as the discharge leaves behind it a series of electrical double 

 layers on the sides of the mercnry columns nearest the positive 

 electrode, the difficulty of forcing electricity through the tube will 

 be temporarily increased. 



It is, I think, worthy of remark that the effects produced by dis- 

 placement currents render it impossible to predict the velocity of the 

 discharge of electricity through a rarefied gas. For, if we consider 

 the processes which accompany this discharge, we have, first, the 

 production of the electric field ; this causes an increase in the electric 

 displacement, and in consequence produces magnetic effects ; and the 

 displacement current behaves as if it had inertia, travelling through 

 the medium with the velocity of light. When the intensity of the 

 field is increased sufficiently to cause discharge, the electricity passes 

 through the gas, and the electric field disappears. The convective 

 current formed by the passage of the free electricity is balanced by 

 the displacement current in the opposite direction, due to the dis- 

 appearance of the electric displacement. The discharge, therefore, 

 does not produce a magnetic field, and has, therefore, no inertia. 

 The velocity of propagation of this discharge will, therefore, be 

 governed by different laws from those which control currents pro- 

 ducing a magnetic field, and need not, therefore, have anything 

 to do with the velocity of propagation of light through the 

 medium. 



By adjusting the circumstances under which the preliminary 

 charging takes place, we can separate the magnetic force due to the 

 charging by as long an interval as we please from the discharge. We 

 can also, by charging sufficiently slowly, make the magnetic force at 

 any instant as small as we please ; thus it is conceivable that we 

 might have a copious discharge of electricity through a gas prac- 

 tically unaccompanied by magnetic force. 



The very remarkable action of a magnet on the electric dis- 

 charge is not inconsistent with this view, as on it the discharge 

 consists of two equal and opposite currents, of which only one is 



