Chemistry and Physics. 383 



electroscope are independent of the communication of the charges 

 of electricity from the gas to the cylinder, and since the kathode 

 rays fall on the inside of the cylinder, the electroscope would not 

 be affected, even if there were such an effect as is produced when 

 ultra-violet light falls upon the surface of an electro-negative 

 metal when the metal acquires a positive charge. Since any such 

 process cannot affect the total amount of electricity inside the 

 cylinder, it will not affect the gold leaves of the electroscope; in 

 fact, Perrin's experiments prove that the kathode rays carry a 

 charge of negative electricity. 



The other view held as to the constitution of the kathode rays 

 is that they are waves in the ether. It would seem difficult to 

 account for the result of Perrin's experiment on this view, and 

 also I think very difficult to account for the magnetic deflection 

 of the rays. Let us take the case of a uniform magnetic field : 

 the experiments v^hich have been made on the magnetic deflec- 

 tion ot these rays seem to make it clear that in a magnetic field 

 which is sensibly uniform, the path of these rays is curved ; 

 now if these rays were due to ether waves, the curvature of the 

 path would show that the velocity of propagation of these waves 

 varied from point to point of the path. That is, the velocity 

 of propagation of these waves is not only affected by the mag- 

 netic field, it is affected differently at different parts of the field. 

 But in a uniform field what is there to differentiate one part from 

 another, so as to account for the variability of the velocity of 

 wave propagation in such a field ? The curvature of the path in 

 a uniform field could not be accounted for by supposing that the 

 velocity of this wave motion depended on the strength of the 

 magnetic field, or that the magnetic field, by distorting the shape 

 of the boundary of the negative dark space, changed the direc- 

 tion of the wave front, and so produced a deflection of the rays. 

 The chief reason for supposing that the kathode rays are a 

 species of wave motion is afforded by Lenard's discovery, that 

 when the kathode rays in a vacuum tube fall on a thin aluminium 

 window in the tube, rays having similar properties are observed 

 on the side of the window outside the tube ; this is readily 

 explained on the hypothesis that the rays are a species of wave 

 motion to which the window is partially transparent, while it is 

 not very likely that particles of the gas in the tube could force 

 their way through a piece of metal. This discovery of Lenard's 

 does not, however, seem to me incompatible with the view that 

 the kathode rays are due to negatively charged particles moving 

 with high velocities. The space outside Lenard's tube must have 

 been traversed by Rontgen rays; these would put the surrounding 

 gas in a state in which a current would be readily started in the 

 gas if any electromotive force acted upon it. Now, though the 

 metal window in Lenard's experiments was connected with the 

 earth, and would, therefore, screen off from the outside of the 

 tube any effect arising from slow electrostatic changes in the 

 tube, it does not follow that it would be able to screen off the 



