1886.] electric discharge in a uniform electric field. 403 



and as the discharge took place in a large bell-jar between pointed 

 electrodes, it is hardly possible, I think, to imagine that anything 

 like so large a number of molecules were dissociated by the spark. 

 In addition to this, in some experiments on the subject which 

 Prof. Threlfall and myself have been making during the past year, 

 we got effects which were too large to be explained even by sup- 

 posing that all the molecules were dissociated, and in these ex- 

 periments only a small fraction of them became luminous. These 

 results seem to point to a kind of explosion taking place at the 

 place where the spark passes, which projects the surrounding 

 gas away from the place of explosion. We should expect from 

 Faraday and Maxwell's theory of stress in a medium that some- 

 thing of this kind should take place. For according to this theory 

 there is a tension along the lines of force and a pressure at right 

 angles to them ; this distribution of stress producing when the field 

 is steady equilibrium at a place where there is no electrification. 

 When however the electric field disappears at one place and not 

 at another the stresses will no longer be in equilibrium, and since 

 the original state was that of tensions along the lines of force, 

 the effect produced by the disappearance of this stress from some 

 part of the field will be much the same as if there was an 

 explosion at the place of discharge, at least as far as the motion 

 of the gas parallel to the line of force is concerned. This effect 

 too will be increased by the decomposition of the molecules which 

 takes place when the spark passes, because this decomposition 

 produces a sudden increase of pressure. For these reasons we 

 conclude that there is a violent projection of the molecules paral- 

 lel to the lines of force from the neighbourhood of the place 

 where the electric field is discharged. 



Let us now consider what takes place when a spark passes 

 through a gas at a moderately high pressure. The electric strength 

 of the gas will break down in the neighbourhood of the negative 

 electrode through some of the molecules there being decomposed. 

 Some of these dissociated atoms will get projected away from 

 the negative electrode, parallel to the lines of force ; after going 

 for a short distance, the length of which depends on the density 

 of the gas, they will recombine, giving out heat in so doing ; this 

 warms the neighbouring molecules, they are then more easily 

 dissociated, that is they are electrically weaker; the spark passes 

 through them, and the same process is repeated until the spark 

 reaches the positive electrode. 



The explosion which takes place when the electric field gets 

 discharged is perhaps one of the reasons why polarization is not 

 produced by the passage of an electric spark through a gas, as 

 it is when an electric current passes through an electrolyte, though 

 according to our view, the process in both cases is very much the 



28—2 



