234 Sir J. J. Thomson : Farther 



calculated on this assumption is far too small to be consistent 

 with the equation 



A dX 



ax 



where p is the density and X the electric force in the 

 neighbourhood of the cathode. 



The reflexion of the positive rays involves the presence 

 near the cathode of a number of positive ions with com- 

 paratively small velocities. As close to the cathode the 

 cathode rays are also moving slowly, they would be able to 

 combine with slowly moving positive particles, and thus 

 give rise to the luminosity which is to be seen close to the 

 cathode. 



On the Difference in the kind of Ionization produced by 

 Cathode Particles and by Positive Rays. — We have seen 

 that on the photographs there are sometimes marked differ- 

 ences between the lines due to the atoms and those due 

 to the molecules. This suggests that the atoms and the 

 molecules get their charges in different ways. It seems, 

 on a priori grounds, that w r e should ascribe the formation 

 of the atoms mainly to ionization produced by positively 

 charged atoms or molecules. For at the end of their paths 

 in the cathode dark space these have in our experiments 

 energy represented by the fall of the atomic charge of 

 electricity through a potential difference of more than 

 20,000 volts. This is thousands of times the energy de- 

 veloped per molecule in any chemical reaction. It is 

 therefore probable that when a charged particle of atomic 

 mass collides with another atom, it will give to it sufficient 

 energy to tear it from any compound, and therefore produce 

 dissociation. We should expect therefore that the fast 

 positive rays would disintegrate into atoms any molecules 

 against which they struck. On the other hand, a fast 

 cathode particle would be able to penetrate into the atom, 

 and would collide rather with individual corpuscles than with 

 the atom as a whole. We should expect then that in general 

 the result of such a collision would be to detach a corpuscle 

 from the molecule without necessarily splitting it up into 

 atoms. The question arises, if the molecule were split into 

 atoms by the impact of the positive rays, would these atoms 

 be positively charged, or would when the atoms split off some 

 be charged positively, others negativel}< ? or, as a third alter- 

 native, would they be uncharged when liberated and acquire 



