Spectra and Planck 1 s Law, 439 



energy. In other words, that the energy in the corpuscular 

 radiation will be proportional to the frequency of the primary 

 Rontgen radiation ; the experiments hitherto made on the 

 expulsion of electrons by Rontgen rays are consistent with 

 this law. 



An atom from which an electron has been expelled will 

 be an electron short when it gets free from the atom with 

 which it was temporarily associated, and will be in a condition 

 to give out radiation when an electron falls into it to replace 

 the one that was lost ; the atom is now free, so that the vibra- 

 tion it will emit will have the normal frequency, whatever 

 may have been the frequency of the radiation which supplied 

 the energy to eject the electron. Thus the character of the 

 Rontgen radiation excited by the primary radiation will not 

 depend on the nature of the radiation, but solely on the nature 

 of the atom ; on the other hand, the nature of the corpus- 

 cular radiation will depend solely on the nature of the 

 primary radiation, and not upon the nature of the atom. 

 These conclusions ;;re in accordance with known properties 

 of Rontgen radiation. 



Again, we see that each electron ejected will correspond to 

 one unit of energy of the characteristic radiation of the atom ; 

 this will be given out when an electron falls into ihe atom to 

 replace the one that has been ejected. Thus whatever the 

 energy in the ejected electron, the energy in the corresponding 

 characteristic radiation will be constant. 



Now, the energy in the ejected electron increases with the 

 frequency of the rays ; hence the ratio of the energy in the 

 corpuscular radiation to that in the characteristic radiation 

 will increase with the hardness of the primary Rontgen 

 radiation. We see, too, that when the frequency of the primary 

 Rontgen radiation is the same as that of the characteristic 

 radiation, the energy in the corpuscular radiation will be 

 equal to that in the characteristic. 



We have, in the preceding discussion, neglected the possi- 

 bility of the high-speed electrons ejected by the Rontgen 

 rays exciting by their impact with the atoms the characteristic 

 radiation of the substance. The justification for this is that in 

 some experiments which I made on the excitation of Rontgen 

 rays by cathode rays I found that the energy of the Rontgen 

 radiation was concentrated in a very soft type of radiation, 

 and that only an exceedingly small fraction of the energy of 

 the cathode rays was transformed into a type of radiation 

 comparable in hardness with that which would excite cor- 

 puscular radiation comparable in velocity with that of the 

 cathode rays which produced it. 



