Discharge in Rarefied Gases, 179 



it is easy to assume (in accordance with the general belief) 

 that the electricity (following the negative current) starting 

 from a traverses the negative light, then passes into the first 

 layer of the positive light, from that into the second, and so 

 on till it reaches the anode. 



But (2) let the kathode a be a flat surface (a strip of metal 

 for example) whose plane is at right angles to the axis of the 

 cylinder, but whose sides both lie in free gas. With this 

 arrangement a sends out rays towards the remote side of the. 

 anode b exactly in the same way as in the line leading directly 

 to b. The rays radiating from a are just as rectilinear, at 

 X'ight angles to the surface «, and without curvature, as the 

 rays directed immediately towards b, and then expand as the 

 exhaustion proceeds further and further into the space filled 

 with gas in the direction turned away from the anode. 



(3) Fig. 3 represents another case, a is a surface which 

 does not completely occupy the section of the tube, but leaves 

 room for the anode b beside it. 



Then the rays of the negative light do not pass over to the 

 neighbouring anode, but the negative light, as represented in 

 the figure, spreads itself out, without reference to the position 

 of the anode, in rectilinear rays through the whole length of 

 the tube without any visible connexion with the anode. 



How now, in the cases represented by (2) and (3), does the 

 electricity pass from the one pole to the other ? In what path 

 is the electric excitement here propagated ? The rays of the 

 negative light, as Hittorf recognized, are electric currents, 

 not simply a glow which surrounds the path of the actual dis- 

 charge. This is shown by the behaviour of the rays towards 

 the magnet, which is in complete accordance with the laws of 

 Biot, Savart, and Ampere. We are therefore compelled to 

 assume that the rays of light point out to us the path of the 

 electricity, which consequently pursues from the kathode the 

 path to the end of the negative rays. If, then, the current, 

 whether we regard it as consisting in the transport of certain 

 similar electric particles, or only in the propagation of an ex- 

 citement from molecule to molecule, is to reach the anode, it 

 must in (3) return by the path it came. And in (2), though 

 the previous assumption of direct transference may suffice for 

 the rays directed towards b, we must assume a passage of elec- 

 tricity to and fro for the quite similar rays turned away from 

 the anode. 



There is, however, no action of this hypothetical return- 

 current to be observed. The magnet diverts the electric rays 

 only in the manner required by the current flowing from the 

 kathode towards the ends of the rays. The (hypothetical). 



