1886.] electric discharge in a uniform electric field. 409 



field varies as the square of the electro-motive intensity, the ratio 

 of the greatest possible electro-static energy to the amount of 

 energy required to raise the temperature of the gas 1 per cent, will 

 diminish very rapidly with the density, so that the capacity of 

 radiation to dissipate the electrical energy will increase as the 

 pressure diminishes. 



The discharge would also take place more readily at low 

 pressures than at high ones, for at low pressures the free path 

 is greater, and therefore the electric forces have a better chance 

 of producing that abnormally great velocity which is necessary for 

 radiation, and in addition, the collisions are not so frequent; so 

 that when it has got it, it does not run so great a chance of losing 

 it by collision with other molecules. 



The discharge by radiation will be much more continuous than 

 that by the decomposition of the molecules, and probably a much 

 larger number of molecules will participate in it. 



The decay in the energy in the medium will destroy the 

 equilibrium of the stresses, which according to Faraday and 

 Maxwell, exist in the medium, but as the decay is much slower 

 than if the energy were dissipated by the decomposition of the 

 molecules, the motion of the dielectric which the disappearance of 

 the stress produces will be much less violent. 



Whether the discharge of the electric energy takes place by 

 radiation or by the decomposition of the molecules will evidently 

 depend very much upon the nature of the gas ; if the gas is one 

 whose particles are easily dissociated, then we should expect the 

 discharge to take place chiefty by the decomposition of the mole- 

 cules ; if, however, the molecules are dissociated with great difficulty 

 then we shall expect the discharge to take place chiefly by radiation. 

 From what we have said, it will be seen that the discharge is more 

 likely to take place by decomposition at the negative than at the 

 positive electrode; so that we may have a case where the discharge 

 is by one means at one electrode and by the other at the other elec- 

 trode. This I think may be the reason why stratifications are pro- 

 duced more easily in the electrically weak gases, such as the vapours 

 of turpentine and alcohol, than in those which are electrically 

 stronger, for according to our view of striae, their formation depends 

 upon the decomposition of the molecule, and at the positive elec- 

 trode the discharge does not take place so easily by this means as 

 it does at the negative ; so there would be a greater tendency for 

 the discharge to take place in the other way, whatever that way 

 may be, and the formation of the striae would be much facilitated 

 if the molecules of the gas were easily decomposed, as gases of 

 complicated chemical constitution usually are. 



