184 PHENOMENA ACCOMPANYING 



are attracted or repelled by magnets as electric currents circulating in metallic 

 wires would be. In a word, tins action is subject to all the laws of electro- 

 dynamics, Avitli this difference, that all the parts of the mobile conductor being 

 independent of one another, instead of being united with one another, as they 

 arc in a rigid wire, they completely obey the forces which act upon them, and 

 take the positions of equilibrium which result therefrom. Hence it is that the 

 luminous iilament takes the form of a magnetic curvcj a necessary condition, in 

 order that the equilibrium should take place, since the action of the magnet on 

 the element of the current is then nothing, the direction of the action being per- 

 pendicular to that element when it is a tangent to the magnetic curve. 



I have verified in sundry cases the hnv just recited, and have even succeeded 

 in showing that, conformably to the law of Ampere, tAvo electric streams having 

 the s.ame direction in a rarefied gas attract each other as two voltaic currents 

 transmitted across movable metallic Avires would do. I have not realized the 

 repulsion of two electric streams passing in contrary directions, by reason of the 

 practical difllculty Avhich I have hitherto encountered in constructing an appa- 

 ratus for the purpose. 1 do not, however, renounce the hope of being able to 

 do so. I shall return to this subject in an article in which I propose to consider 

 the mutiial action of electric currents on one another. I restrict myself, for the 

 present, to an investigation of the effects of magnetic action on those currents. 



My researches on this subject comprise two series of experiments : first, those 

 in whifli the electro-magnet from which the electric action emanates is placed 

 externally to the rarefied gas through which the electric stream is propagated; 

 secondly, those in which the magnetized iron is situated in the gas itself. 



One of the most simple cases is that in which one of the tubes of which I 

 have spoken in preceding experiments is placed either axially or equatorially in 

 relation to the poles of a strong electro-magnet. The following is what is ob- 

 served when care has been taken to rarefy well the gas which transmits the 

 electric current. The portion of this current submitted to magnetic action is 

 condensed towards the walls of the tube in the part nearest, or that most remote 

 from, the magnetic poles, according to the direction of the current and that of 

 the magnetization; the striai become much more compressed and more brilliant. 

 If the portion of the tube placed in the neighborhood of the electro-magnet is 

 that Avherc the negati\'e electrode happens to be, the obscure space is immedi- 

 ately seen to become luminous, and to present close and brilliant stria; as would 

 be the case with the constantly luminous portion of the current which seems to 

 advance. At the same; time, the bluish photosphere which surrounds the neg- 

 ative ball contracts to at least half its size, becoming more brilliant, and the sort 

 of bluish sheath which surrounded the metallic rod, at the extremity of which 

 is the negative electrode, com])letely disappears. All that bluish atmosphere is 

 concentrated on the ball. It seems that all the gaseous filaments, which may 

 Ix; considered as so many conductors of the discharge, instead of radiating from 

 i'll points of the negative ball and rod, and being disseminated through the 

 (entire gaseous mass as far as the positive electrode, radiate only, when tin; 

 magnetic action is exerted on them, from the negative ball, becoming condensed 

 towards the Avails of the tube, on one side or the other, as far as that portion of 

 their course at which, the action being no longer sensible, they resume their 

 normal position, 'i'his condensation explains why the part of the current which 

 was obscure because the gas was there too much dilated, becomes luminous, and 

 why thfit part Avhich Avas already luminous becomes more slender and brilliant, 

 with stratifications more closely compressed. The action of the magnet pro- 

 duces the same effect Avhich Avould be produced by a local augmentation of 

 density in the rarefied gaseous matter. Further, it is not necessary that the 

 action of the magnet should take place exactly on the obscure part in order to 

 its becoming luminous ; it equally becomes so, even Avhcn the magnetism acts 



