240 DISCHARGE OF ELECTRICITY. 



the magnetic force is on a small discbarge preceding the main one; 

 the action of the magnetic force on the chain of polarized molecules 

 which are formed before the discharge passes might produce an effect 

 equivalent to that which we have supposed was produced on an actual 

 discharge. 



The chain of polarized molecules would be affected in the following 

 way: The magnetic held due to the electro-Diagnet consists of tubes of 

 electrostatic induction moving about these tubes, as well as the direc- 

 tion in which they are moving, are at right angles to the lines of mag- 

 netic force. The short tubes of electrostatic induction wliich join the 

 atoms in the molecules of the gas will, under the influence of the electric 

 forces, set themselves parallel to the direction of the electro-motive in- 

 tensity at each point. 



Thus, when the magnetic force is at right angles to the line of dis- 

 charge, tubes of electrostatic induction parallel to those in the molecules 

 will be moving about in the field; and since parallel tubes exert attrac- 

 tion and rei)ulsiou on each other, the molecular tubes will be knocked 

 about and their efforts to form closed chains made much more difficult 

 by the action of the magnet. On the other hand, when the lines of 

 magnetic force are parallel to the discharge, the moving tubes are at 

 right angles to those in the molecules, and widnot disturb them in the 

 attempt to form chains along the line of magnetic force; they will in 

 fact assist them in doing so by preventing all attempts in directions 

 across the lines of force. 



Prof. G. F. Fitzgerald has suggested to me in conversation that tliis 

 action of a magnet on thedischarge might be the causeof the " streamers" 

 which are observed in the aurora; the rare air being electrically weaker 

 along the lines of magnetic force than at right angles to them will cause 

 the discharge in the direction of those lines to be the brightest. 



Discharge through different gases. — I have examined the discharge 

 through air, carbonic acid, hydrogen, oxygen, coal gas, and acetylene. 

 As I have already mentioned, at the highest pressures at which the 

 discharge passes through air, the discharge is reddish, and gets brighter 

 and whiter at lower pressures. If the discharge is examined through 

 a spectroscope, the lines in the spectrum coincide with those obtained 

 by sparking through air in the ordinary way with a jar in the circuit. 

 The relative brightness of the lines in the spectrum of the discharge 

 without electrodes varies very much with the pressure of the gas and 

 the length of spark in the jar circuit. With a long spark in this cir- 

 cuit, and the pressure such as to give a bright white discharge, the 

 spectrum is very much like that of the ordinary jar discharge in air. 

 When however the pressure is so low that the discharge passes with 

 difficulty, a few lines in the spectrum shine out very brightly, whilst 

 others become faint, so faint indeed sometimes that if the air spectrum 

 were not thrown into the field of view of the spectroscope at the same 

 time, they might pass unnoticed. Three lines which are very persist- 



