Passage of Electricity through Gases. 169 



perfectly dark and therefore cold, in what manner the gas 

 conveys the electricity before becoming heated to the tem- 

 perature at which it begins to conduct like a metal, and how 

 it thus becomes heated. Besides, the increase of temperature 

 in tubes of different diameters must be precisely such at each 

 pressure, that, in addition to the heat formerly produced in 

 them, so much more heat should be produced in the gases 

 which now conduct metallically that the combined bodies of 

 heat in the tubes would be equal to each other — a supposition 

 equally complicated and improbable. 



Even the phenomenon of the aureole and the "trait de 

 feu " is no proof of the metallic conductivity of the gas, as 

 the former is equally produced by single discharges in lumi- 

 nous gases, and the latter by metallic discharges which at 

 high pressure are added to the former. 



Rather do the above-mentioned experiments allow us to 

 conclude almost with certainty that the conveyance in gases is 

 not carried on by any such double process, but actually 

 continues essentially in the same manner from the beginning 

 to the end of the discharge. 



The process of the discharge in my opinion maybe supposed 

 to be somewhat as follows. First of all, we cannot suppose that 

 the conveyance of electricity in gases can take place through 

 the ether alone ; for then a discharge would pass in all vacua, 

 however perfect, which is not the case. The particles of the 

 gas themselves must therefore take part in the conveyance of 

 electricity. If, then, the charge of the electrodes at one place 

 exceeds a certain limit, and if the scattered particles of gas 

 lying nearest to the electrode, between it and the other parts 

 of the gas, be also sufficiently strongly charged with electri- 

 city of the same name, the whole of the electrified mass of 

 gas there will be thrown up, as when a liquid is boiled. But 

 even when the electrodes are considerably heated the number 

 of discharges with an equal supply of electricity does not 

 alter, provided the pressure remains constant. Hence the 

 overcoming the attraction of the gas stratum condensed upon 

 the electrodes, which is certainly altered, plays only a subor- 

 dinate part in the phenomenon. 



The body of gas electrified at the electrode moves away in 

 the direction in which the tension most rapidly diminishes, 

 with a velocity proportional to the force which immediately 

 before was pressing the gas particles upon the electrodes. 



Here the original impulse occurs only upon the electrodes ; 

 and in the case of a pair placed opposite to one another in a 

 free gas space, it will depend only upon the tension existing 

 upon particular points of them. If the electrodes are en- 



