36 FORESHADOWING OF THE ELECTRON [CH. iv. 



readily occur through direct encounters. And the 

 time that thus elapsed before the whole of the 

 conductivity disappeared from dust-free air suggested 

 that the moving particles must be very small, so that 

 intimate collisions were comparatively infrequent. 



The mobility or diffusiveness of the molecules 

 of a gas depends on their mean free path, and 

 that depends on their atomic size ; the smaller 

 they are, the more readily can they escape 

 collision. Hence it is that collisions are so rare in 

 astronomy : the bodies are small compared with the 

 spaces between them. The behaviour of charged par- 

 ticles seemed to indicate that they must in some 

 cases be something smaller than atoms : it seemed 

 hardly likely that material atoms could behave in 

 the way they did. It was recollected that it had 

 occurred to some philosophers, among them Dr. 

 Johnstone Stoney, that electric charges really existed 

 on an atom in concentrated form, not diffused all 

 over its surface but concentrated at one or more 

 points, perhaps acting as satellites to the main 

 bulk of the atom ; so on that view it was j ust 

 possible that these flying particles might be not 

 charged atoms at all, but charges without the atoms, 

 the concentrated charges detached knocked off as 

 it were in the violence of the discharge, and after- 

 wards going about free. Such particles would 

 naturally travel at an immense pace, because they 

 would still be exposed to the full electric force that 

 they had experienced before, and yet would have 

 shaken off the encumbrance of the material atom 

 with which they had been associated. It is true that 

 no such disembodied charges, or electric ghosts, had 

 ever been observed. All the experiments that had 

 been made in electrostatics had been made on charged 



