66 BELL SYSTEM TECHNICAL JOURNAL 



light scattered from an incident beam, of which the wave-length is 

 varied while the intensity is kept constant, increases very rapidly in- 

 deed with decreasing wave-length — in some cases, as the inverse fourth 

 power of the wave-length. (This is the reason why the sky is blue; 

 the molecules of oxygen and nitrogen are not especially tuned to 

 blue light, but the waves from the sun are more powerfully scattered the 

 nearer they lie to the violet end of the visible spectrum.) The scat- 

 tered light is more or less polarized, even when the primary light is not 

 polarized at all. In some cases the light deflected through 90° is 

 perfectly plane-polarized; by the undulatory theory this signifies a 

 very simple sort of vibrator in the atom, a vibrator which is attracted 

 by an equal restoring-force whichever way it is displaced from its 

 centre of vibration. In other cases the polarization is different, and 

 other inferences about the atom-model may be drawn from it. As for 

 the "coherence," it is a very important property — important for the 

 theorist. If the scattering atoms contain vibrators which the infalling 

 waves maintain in forced oscillations, and which themselves send out 

 the scattered light, then these scattered or "secondary" wave-trains 

 should interfere with one another. If in particular the vibrators, or 

 let us say the atoms, form a regular lattice in space — a cubic lattice, 

 for example — there should be destructive interference; the secondary 

 wavetrains should completely destroy one another in all directions 

 save that of the ongoing primary beam,^ and there should be no per- 

 ceptible scattering at all. The perceptible scattering is then a measure, 

 to speak rather vaguely, of the irregularity in the arrangement of the 

 atoms. This theory seems to be confirmed, for the light scattered with- 

 out change of wave-length. Whether it is true also for the shifted 

 light will probably soon be known. 



Scattering of light with unchanged frequency is easy to explain by 

 either wave-theory or corpuscle-theory. To those who think of light 

 as waves and of atoms as systems of vibrators, it is a consequence of 

 forced vibrations. The fluctuating or alternating electric field which — - 

 coupled with an alternating magnetic field — constitutes the beam of 

 light, seizes upon an electrified portion of the atom and swings it to and 

 fro in synchronous vibration. From this swinging electric charge, the 

 scattered waves originate. It is evident that the forced vibrations 

 and the scattering should be especially intense, when the frequency of 

 the light coincides with a natural frequency of vibration of the atom, 

 for then there is resonance. Now it is a fact that scattering is espe- 

 cially intense, when the infalling light agrees in frequency with any of 



^ More precisely, in all directions save those of the Laue diffraction-beams, which 

 however for crystals and visible light do not occur. 



