Spectra and Planck's Law. 441 



Let us suppose that the atom is being ionized by radiation 

 whose frequency is n, where n is also the frequency of the 

 characteristic radiation of the atom. The ejected electron 

 begins by absorbing an amount of energy equal to 2hn, of 

 this, when it is liberated, it retains as kinetic energy hn — 

 this energy appears as the energy of corpuscular radiation ; 

 the other half of the energy appears subsequently as 

 characteristic radiation when an electron drops into the 

 atom to take the place of the one displaced. In this case 

 the energy in the corpuscular radiation is equal to that in 

 the characteristic. 



Next suppose that the frequency n Y of the incident 

 radiation is not the same as ?i, that of the characteristic 

 radiation. The absorption of energy will take place when 

 the atom has formed a temporary alliance with another atom, 

 of such a kind that the forces between the atoms have altered 

 the frequency of the electron from n to n x . The absorption 

 of energy by the electron is now 21111^ of this Jm 1 appears as 

 the energy of corpuscular radiation, the atom recovering an 

 electron after it has again become free will give out hn 

 units of characteristic radiation ; the difference 



2/mj — h(n + ni) = h(n l — n) 



will be given up to the atoms, making them separate with 

 more energy than they had before they came together, a 

 fraction /5 of this energy might be converted into slow 

 corpuscular or soft Rontgen radiation. 



Thus the ejection of a high-speed electron will be accom- 

 panied by the following energy effects : 



Energy in corpuscular radiation = hii. 

 Energy in characteristic radiation = hn. 

 Energy in soft Rontgen radiation =yS/t(n 1 — ?i). 



The corpuscular radiation will bear to the characteristic 

 radiation the proportion of n^ to n and will thus increase 

 indefinitely as n Y increases. If we include the soft Rontgen 

 radiation with the corpuscular, and in many experiments 

 they would not be distinguished, the ratio of the corpuscular 

 energy to the characteristic energy would be 

 ni + Pfa-n) 

 n 



The view that when an electron absorbs radiant energy 

 it is a constituent of a system in which it vibrates in 

 unison with the radiation, is not without difficulties. It 

 requires, for example, concentration of the energy of the 

 light-wave in certain places instead of uniform distribution 



