442 Sir J. J. Thomson on the Origin of 



over the wave front. It has, however, the great advantage 

 that it accounts for the electrical effects produced by light 

 without introducing changes in our conception of light 

 which give rise to difficulties in connexion with ordinary 

 optical effects such as interference or scattering. These 

 would not be affected by the resonance- hypothesis. 

 In favour of this hypothesis may be urged : — 



1. That it is an exceedingly minute fraction of the number 



of atoms or molecules of a substance which plays any 

 part in the electrical effect. Even what is regarded 

 as very intense ionization by Rontgen rays would be 

 produced if on an average each atom got ionized once 

 in some hundred thousand years. So that if each 

 atom were to form a part of a suitable system once 

 in this period it would be all that is required. 



2. The number of lines in the spectrum of a substance 



which are theoretically possible is very large ; the 

 electrons have a large number of possible frequencies, 

 each of which would be disturbed by the electric and 

 magnetic forces exerted by a neighbouring atom, so 

 that the frequencies continuously covered by the 

 system would extend over a wide range. 



3. The difficulty of supposing that the influence of two 



nei<rhbourini>- atoms of a light element could be so 

 great that the frequency of an electron in one of the 

 atoms could rise to that of the K radiation of a heavy 

 element has been very greatly lessened by some 

 recent researches. Barkla (Phil. Trans, vol. 217. 

 p. 315) has shown that the lighter elements give out 

 a type of radiation much harder than the types pre- 

 viously detected, and that this new type is comparable 

 in hardness with the K type given out by the heavier 

 elements. Again, Shearer (Phil. Mag. xxx. p. 644), 

 who has investigated the ionization of hydrogen by 

 hard Rontgen rays, has shown that the amount of it 

 is, at most, surprisingly small, even when the density 

 of hydrogen is taken into account, and that it is 

 doubtful whether even this small amount is a genuine 

 ionization of hydrogen ; the experiments seem to 

 suggest that it was due to a trace of some heavier gas. 

 This is an exceedingly important result, because it 

 shows that the impact of Rontgen rays against an 

 electron does not necessarily communicate very much 

 energy to it, even though there may be a large amount 

 of energy in the rays ; it differentiates the action of 

 Rontgen rays from that of either cathode or a-rays. 



