1908] on the Carriers of Positive Electricity. 173 



not be due to A acting- at times as a cathode in virtue of the negative 

 charge given up to it by the cathode rays starting from B. That 

 this is not the explanation is proved by the following experiments. 

 When the cathode rays were diverted by a magnet so that they no 

 longer fell upon A, the brightness of the phosphorescent patch due 

 to the positive rays going down the tube in A was not appreciably 

 diminished, although the tendency for A to become cathode must 

 have been almost entirely removed. The conclusions drawn from 

 this experiment were confirmed by the results obtained when the ob- 

 stacle in the side tube was used as a cathode instead of B. When 

 the obstacle was not pushed far enough across the tube for its normal 

 to pass down the tube in A, no positive rays passed down the tube, 

 but as soon as the obstacle had advanced into such a position that its 

 normal went down the aperture in A, the phosphorescence on the 

 screen due to the positive rays appeared. The contrast between the 

 brightness of the phosphorescence when the normal to the obstacle 

 went down the hole in A and when it did not was very sharp, though 

 there was very little variation in the number of cathode rays striking 

 against the anode as a whole. These experiments show that the 

 positive rays under discussion are not due to reversals of the induc- 

 tion-coil nor to the negative electrification of A by the bombardment 

 of cathode rays, but that they originate at the cathode and travel away 

 from it down the tube. 



By means of the rotating cathode B we can determine whether 

 the positive rays coming from the cathode are emitted normally to its 

 surface, or whether, like some of the rays observed by Goldstein, they 

 come off in all directions. When the normal to the cathode went 

 down the tube in A, a plentiful supply of positive rays went down the 

 tube. When the cathode was rotated, the phosphorescence due to 

 the positive rays did not immediately disappear, although it became 

 very much fainter ; it could, however, be detected until the normal 

 to the cathode made an angle of about 15° with the axis of the tube. 

 The positive rays under discussion appear to follow much the same 

 path as the cathode rays, for it was found that the angle of rotation 

 required to prevent these getting dow^n the tube was much the same 

 as that required to extinguish the phosphorescence due to the positive 

 rays. 



Properties of these Positive Rays. 



These rays get exceedingly faint at very low pressures, and cease 

 to be observable at pressures when the Canalstrahlen are still quite 

 bright. It is probably due to this that I have never been able to 

 observe the resolution of the phosphorescence, under the action of 

 electric and magnetic forces, into separate patches as in the case of 

 Canalstrahlen when the pressure is low.* The spot of phosphorescence 



* Phil. Mag. xiii. p. 561, 1907. 



