188 Dr. E. Goldstein on the Med 



rtc 



even before complete magnetization, does the magnetized light 

 show no clear separation of the curves in the part of the tube 

 under consideration. 



It appears to me that the assumption of the existence of 

 many independent currents in the discharge between the 

 electrodes rinds the strongest support from observation of the 

 special form of the magnetic action on the electric rays. The 

 mode of this action was discovered and explained for the 

 kathode-rays by Hittorf*. My researches have shown that 

 the results obtained by Hittorf, in opposition to the view held 

 since Pliicker's researches, hold also for each separate positive 

 layer, in harmony with the often-repeated statement that each 

 layer is to be regarded as a modified formation of negative 

 light. 



Let us now assume that the discharge actually forms only 

 a single current from anode to kathode. Then the magnet 

 must act upon the discharge, in, for example, the equatorial 

 position, as upon an expansible flexible conductor fixed at its 

 ends (here the electrodes), traversed in the same position by 

 a correspondingly-directed current. The form of the mag- 

 netized column of light would then be an arc extending from 

 the one electrode to the other in the equatorial plane ; but 

 the current could never form itself into a magnetic curve. 

 But if the magnet acted upon a conductor fixed at one end 

 but free to move at the other, the motion of such a conductor 

 would accurately correspond to that of a magnetized kathode- 

 ray or ray of a positive layer ; and a bundle of such linear 

 conductors radiating from one point, and all free at their other 

 ends would, when magnetized, accurately represent the form 

 of a bundle of the kathode-light issuing from a point. 



The magnetic curve into which such a bundle rolls itself 

 together results (according to Hittorf's investigations, which 

 after frequent repetition, I can confirm) in the following 

 manner : — The bundle consists of a cone of diverging rays. 

 The rays situated near to the axis of the cone are clearly dis- 

 tinguished from those lying further away by their greater 

 brightness ; if therefore the axis of the bundle is placed 

 accurately equatorially, the motion of the rays toward the 

 magnet can be recognized in the bright centre bundle. 



The bundle passes, as tlje strength of the magnetic action 

 increases, from a straight thread of light into an excessively 

 narrow flat spiral, whose plane coincides with the equatorial 

 plane. When the magnet is powerful, the diameter of the 

 spiral does not exceed 1 millim., so that it appears nearly as a 

 point of light. If, however, the axis of the cone is inclined 

 * Pogg. Ann. exxxyi. p. 213. 



