182 



Professor Sir J. J. Thomson 



[April 10, 



right angles to their direction of projection : thus if they are positively 

 charged they will be bent to the right, if they are negatively charged 

 to the left, while if they are uncharged they will be undeflected. The 

 path of the rays when the pressure of the gas is not too low, can 

 readily be traced by the luminosity they produce 

 in the gas as they pass through it. On trying 

 the experiment, it was found that when the 

 gauze was connected with the anode the path 

 of 'the few rays which got through the gauze 

 was a straight line, coinciding in direction with 

 their path before passing through the cathode. 

 An easy way of seeiug this is to connect by 

 means of a key the gauze in quick succession 

 with the anode and the cathode, when it is easily 

 seen that though the Canalstrahlen are much 

 more numerous in the latter case than in the 

 former, the paths of those which do get through 

 are identical, so that even when the gauze is 

 connected with the anode some of the rays get 

 through the space between c d and cf without 

 suifering any deflexion, showing that they must 

 be uncharged as they pass through this region. 

 It is thus evident that a considerable u umber 

 of the positively charged Canalstrahlen lose their 

 positive charge by attracting when in the neighbourhood of tlie 

 cathode a negatively electrified cor])uscle ; the mass of the corpuscle 

 is so small in comparison with that of the particles foruiing the 

 Canalstrahlen, that the addition of the corpuscle will not materially 

 reduce the velocity of the Canalstrahlen. These rapidly moving 

 uncharged particles will soon get ionized by collision, and by losiiig a 

 negatively electrified corpuscle again become positively charged. 



In my first paper on the positive rays,* I sliowed that at not too 

 low pressures the appearance presented by the phosphorescence on 

 the screen indicated that many of the particles in the Canalstrahlen 

 were positively charged for only a portion of their path ; the experi- 

 ments just described are very direct evidence of this eifect. 



Again, at not too low pressures the Canalstrahlen are accompanied 

 by negatively electrified particles having masses and velocities equal 

 to those of the positive particles ; the negatively electrified particles 

 in my experiments always being less numerous than the positive ones, 

 in most cases very much so, in others the difl'erence was not very 

 great. We should expect the negatively electrified particles to be less 

 numerous than the positive ones, since they would more readily lose 

 their charges by collision. 



I think that the positive rays which travel away from the cathode 



Fig. 5. 



*_Phil. Mag. xiii. p. 561. 



