CONTEMPORARY ADVANCES IN PHYSICS 67 



certain spectrum lines which the atom may emit spontaneously. To 

 those who think of light as a hail of corpuscles — "quanta" — scattering 

 is rebounding of the quanta from atoms which they strike "elas- 

 tically"; that is to say, as one elastic sphere striking another. This 

 is easy to picture ; but then we are left without any obvious explana- 

 tion of the fact which was just mentioned — the fact that this sort of 

 rebounding takes place especially often, when the quanta agree in 

 frequency with those which the atom can naturally emit. Finally, for 

 the single case in which the incident frequency agrees with that of a 

 spectrum line and the scattering is very abundant, one can employ a 

 compromise-theory; the atom is struck as by a bullet which sets it to 

 vibrating freely with one of its own natural frequencies, as a bell 

 which is struck by its clapper. 



Scattering of light with change of frequency is certainly more com- 

 plicated. The advocates of waves and oscillators must conceive that 

 in the atom there goes on a process similar to what, in the art of elec- 

 trical communication, is known as modulation. The frequency of the 

 infalling light is modulated with some frequency characteristic of the 

 atom. If the compromise-theory is valid, there are several cases in 

 which one easily sees how this happens. Thus if a straight spring is 

 alternately contracting and expanding with a frequency «o, and at the 

 same time is revolving around an axis perpendicular to its length with 

 a constant angular velocity lirni, its ends will seem to a stationary 

 observer to be moving with a motion compounded of two frequencies — 

 (wo + fix) and (wo — Wi) ; and if waves are sent out, they will have 

 these frequencies jointly.^ If the "spring" is an electrical doublet 

 lying perpendicular to a magnetic field, it revolves automatically as it 

 vibrates, and sends forth electromagnetic waves which are discrim- 

 inated by the spectroscope into two lines of these two frequencies. 

 Such is the explanation by wave-theory of the "normal Zeeman effect" ; 

 and while the actual effect of magnetic fields upon the light emitted 

 by the atoms which they influence is not often exactly thus, it is 

 sufficiently nearly so to prove that this interpretation is a step on the 

 right path. If however one wishes to maintain the uncompromising 

 wave-theory, and suppose that the vibrators in the atoms are kept 

 going in forced vibration by the continually-acting waves of light, then 

 modulation does not necessarily occur — not at least with the con- 



^ This idea was introduced by the elder Lord Rayleigh in the course of some 

 speculations on the emission of light by rotating atoms, and was later turned to ac- 

 count in explaining the fine-structure of the bands which constitute the spectra of 

 molecules. 



