Rays of Positive Electricity. 243 



called secondary rajs, for these particles would have their 

 maximum velocity when they reached the cathode, and after 

 passing through the cathode would continue to produce rays 

 of the same type while passing through the electric and 

 magnetic fields. 



Though some secondary rays are probably produced in 

 this way, yet I am of opinion that the rays which produce 

 the curves on the photographic plate arise in a different 

 way, for the ions produced by the impact of one molecule 

 against another would probably start off in different direc- 

 tions ; this would make the curves on the photographic plate 

 very fuzzy, while, as a matter of fact, they are very often 

 beautifully sharp. Again, if the ions forming these rays 

 came from the dissociation of the molecules of the gas in the 

 deflexion chamber, we should expect that the brightness of 

 the curves due to the mercury ions, for example, would be 

 much more dependent on the presence of mercury vapour 

 in the deflexion-chamber than in the discharge-tube. Expe- 

 riments on the behaviour of the secondary radiation due to 

 mercury show that this is not the case ; on the contrary, if 

 we abstract the mercury vapour from the discharge-tube by 

 charcoal cooled with liquid air, we produce a much greater 

 diminution in the brightness of the secondary lines due to 

 mercury than we do if we abstract by the same means the 

 mercury vapour from the deflexion-chaniber. 



The evidence is, I think, in favour of the view that the 

 secondary rays are due to the dissociation of systems in the 

 undeflected Canalstrahlen, rather than to the dissociation of 

 the gas in the deflexion-chamber through which the Canal- 

 strahlen pass. The dissociation of the systems in the 

 Canalstrahlen is produced by the collision of these systems 

 with corpuscles and not with ordinary molecules ; for if the 

 collision were with ordinary molecules, the direction of 

 motion of the systems and their velocities would change 

 when the collision took place, and the result would be that 

 the lines on the plate would be very fuzzy and ill defined, 

 a collision with a body having as small a mass as that of a 

 corpuscle would leave the direction of motion and the velocity 

 of a body as massive as a molecule practically unaltered. 

 The corpuscles when struck by the Canalstrahlen may be 

 regarded as practically at rest in comparison with the systems 

 which strike against them ; for though in the space between 

 the parallel plates used to produce the electrostatic deflexion 

 of the rays there is an electric field strong enough to make 

 the corpuscles move with very great velocities, yet as the 

 path of the Canalstrahlen is the region where the corpuscles 



112 



