Electric Discharge in Gases. 93 



dary furthest from the negative electrode is the brighter, but 



both are equally sharply defined. The line of light d b does 



not reach quite to the end of a/, but cuts this line. Also the 



line af does not extend quite round the tube, but ends where 



the magnet is, bending a little so as to become more nearly 



parallel to the axis of the tube. If the magnet is moved from 



the anode towards the kathode, the two boundaries approach 



each other the more closely the nearer the magnet comes to the 



negative electrode, until at last they coincide. Notwithstanding, 



so long as the magnet-pole is not close to the limit of the dark 



kathode-layer (in one case so long as it was not nearer than 



about 12 millim.), the inclination of the green ring to the axis 



of the tube from the positive electrode remains constant, and 



25 

 in the above case amounted to tan" 1 ^. But if the magnet 



approaches the dark layer, the inclination decreases consider 

 ably, and at the boundary of the dark space attains the value 



25 



tan -1 ^. When the magnet-pole has somewhat overpassed 



the limit of the dark kathode-space, the luminous figure- 

 becomes indistinct, reducing itself to a spot of light, and this 

 recedes further from the pole as the magnet approaches the 

 electrode. 



The phenomenon appears at all pressures within certain 

 limits. The distance of the magnet from the electrode, beyond 

 which the deflection is of such a nature that the half-moon 

 shaped figures are produced, is greater the greater the exhaus- 

 tion, evidently because the dark space extends continually 

 further and further. 



If we examine the deflection of the kathode-rays more closely, 

 we find that it only takes place from the point where the white 

 colour changes into blue. But since the deflection of the ring 

 is so much dependent upon the position of the magnet relative 

 to the electrode, and there is scarcely any ring formed when 

 the magnet is behind the limit of the glow-light, we see that the 

 kathode-rays behave altogether differently within the dark 

 space. And when they have once traversed it they offer, in 

 the first position, a much greater resistance to all lateral dis- 

 placement. 



It is very probable that the glow-rays which are beyond the 

 dark space play some part in the formation of the green ring. 



(2) In the second case, where both magnet-poles a and b 

 deflect the rays towards the centre so that they are there com- 

 pressed, there appears a bright strip of light which, starting 

 from the line joining the poles, extends towards the positive 



