under the Influence of the Electric Discharge. 419 



two D-lines. So great an increase of the normal velocity 

 of the molecules is therefore quite improbable *. 



According to my view, we may imagine to ourselves the 

 phenomena of the discharge somewhat as follows: — The elec- 

 tricity produced by the machine, which we may imagine as 

 free aether, is accumulated on the surface of the electrodes 

 partly as free electricity, and there is prevented from passing 

 into the surrounding gas by the mutual action between it and 

 the molecules of the metal ; and a transference can only take 

 place when its density has reached a sufficient magnitude. 

 At the same time the electricity produces a dielectric polariza- 

 tion in the surrounding medium in such a manner that the 

 aether envelopes of the separate gas-molecules become de- 

 formed, and, in consequence of the rotation of the molecules 

 on their axes, maintain a definite position. If a discharge 

 takes place, the sudden change of the dielectric polarization 

 thereby produced -propagates itself from the electrode through 

 the aether envelopes of the gas-molecules, and thereby puts 

 them into vibration. At the same time a transference of free 

 electricity from the electrode may no doubt take place from 

 molecule to molecule. 



As a ray of light produces in the aether envelopes of the 

 molecules of phosphorescent and fluorescent bodies oscillatory 

 movements whose vis viva is considerably greater than corre- 

 sponds to the temperature, so also the case is here ; as there 

 the motions of the aether which produce the luminosity gra- 

 dually transfer themselves to the molecules and produce heat- 

 motions, so also, in an exactly similar manner, with the elec- 

 tric discharge a secondary elevation of the total temperature 

 results. 



If in consequence of this transference two molecules of the 

 gas have a greater motion of oscillation than corresponds to 

 their temperature, in accordance with the normal relation- 

 ships between motion of translation and motion of oscillation 

 (rotation), then upon their collision a portion of the internal 

 motion resolves itself into motion of translation, until at last 

 the normal condition is restored")". 



* These considerations, which I had already (Wied. Ann. ix. p. 1G0) 

 brought forward to prove that there is no propagation of molecules in the 

 direction of the current, have been again adduced by Goldstein in a paper 

 (Phil. Mag. September, 1880) on the phenomena which take place at 

 the negative pole, which only came to my knowledge since the above 

 was in type. 



t In the paper cited above I have compared the luminosity of the gas 

 under the influence of the discharge with the light of a fluorescent sub- 

 stance. Hittorf, in discussing it, has expressed the opinion that one 

 should speak of phosphorescence. The two terms, however, are only dif- 



