172 



Professor Sir J. J. Thomson 



[April 10, 



a metallic obstacle is placed at the end of a rod, this end is fastened 

 to a closed glass vessel containing a piece of iron. By moving this 

 vessel along D by means of a magnet, the obstacle can either be in- 

 serted in the line of fire of the rays coming from B and passing down 

 the hole in A, or withdrawn into the tube ; the obstacle is in metallic 



Fig. 1. 



connexion with a wire leading out of the tube, so that it can be used 

 as a cathode if required. The discharge through the tube was pro- 

 duced in some cases by a large induction-coil, in others by a Wims- 

 hurst machine. 



If the stopper carrying the electrode B were turned so that the 

 normal to the electrode coincided with the axis of the tube A, or 

 made only a small angle with it, and if B were made cathode and a 

 discharge sent through the tube, then in addition to the cathode rays 

 other rays passed through the tube in A and excited phosphorescence 

 on the screen. The direction of the deflexions of the phosphorescence 

 under electric and magnetic forces showed that these rays were charged 

 with positive electricity. 



If A were made cathode, the ordinary Canalstrahlen produced 

 bright phosphorescence on the screen. The first point investigated 

 was to make sure that the positive rays observed when A was anode, 

 were not due to reversals of the induction-coil making A at times a 

 cathode and sending ordinary Canalstrahlen down the tube. Very 

 simple observations, liowever, showed that this could not be the ex- 

 planation. In the first place, the positive rays still passed down the 

 tube when A was disconnected from the coil and the auxiliary elec- 

 trode C used as an anode ; secondly, A being connected with the coil, 

 the rays down the tube disappeared when B was twisted round, so 

 that the normal to its plane made a considerable angle with the axis 

 of the tube ; and thirdly, the rays down the tube were stopped when 

 the obstacle in the side tube was pushed forward so as to be in the 

 line between the cathode and the aperture in the anode. 



The next point investigated was to see whether the effect might 



