244 Sir J. J. Thomson on 



are produced and where they have not had an opportunity of 

 acquiring a high velocity from the electric field, they will 

 only be moving slowly when struck by the Canalstrahlen. 



Experiments on the ionization produced by the collision 

 of corpuscles against molecules at rest, or rather moving 

 very slowly in comparison with the corpuscles, have shown 

 that for the corpuscle to ionize the molecule the velocity of 

 the corpuscle must be greater than a certain value : thus, 

 according to the results obtained by Townsend and H. A. 

 Wilson, if a corpuscle is to ionize a molecule of air by col- 

 liding against it, the velocity of the corpuscle must exceed 

 that which it would acquire by falling through a potential 

 difference of two volts. It would probably not require so 

 large a velocity to ionize the constituents of the beam of 

 Canalstrahlen, as some of these are probably much more 

 loosely connected systems than a molecule of air; we should 

 expect, too, that the velocity would vary from one constituent 

 to another. Let us suppose that to ionize a particular con- 

 stituent requires a potential difference of n volts. If now 

 the corpuscle is at rest and the molecule moving, the relative 

 velocity when there is ionization must be the same as in the 

 previous case ; but since the molecule has a much greater 

 mass than the corpuscle, the molecule to acquire the same 

 velocit}" must fall through a much greater potential difference; 

 if it requires n volts to give this velocity to the corpuscle, it 

 will require 10 x ?iX 1*7 x 10 n volts to give this velocity to a 

 system whose electric atomic weight is iv ; we have taken 

 the mass of the hydrogen atom as 1*7 XlO 3 times that of the 

 corpuscle. Thus the greater the electric atomic weight of 

 any constituent of secondary radiation the greater is the 

 potential difference through which it must have fallen in the 

 discharge-tube in order that it should be dissociated in the 

 deflexion-tube. Thus the secondary radiation of least electric 

 atomic weight, that of the atom of hydrogen, would require 

 to fall through a smaller potential difference than any other. 

 This explains why at comparatively high pressures, when 

 the potential difference in the discharge-tube is small, the 

 hydrogen atom is the only type of secondary radiation to be 

 seen ; the others are, as it were, latent in the beam of Canal- 

 strahlen, but the potential difference in the discharge-tube 

 was not sufficient to make them move quickly enough to be 

 ionized when they came into collision with a corpuscle. 



For the uncharged Canalstrahlen to be dissociated, and 

 thus give rise to secondary rays, they must travel with a 

 velocity greater than a certain critical velocity, and to 

 acquire this velocity they must fall through a potential 



