322 BELL SYSTEM TECHNICAL JOURNAL 



monness of the Compton effect depends not merely on the nature of 

 tlie atoms and on the angle at which the scattering is observed, but 

 also upon the frequency of the radiation. High-frcciiicncy quanta 

 are liable to rebound in the manner prescribed 1)\' Compion's assump- 

 tions, but low-frequency quanta are not. Light of the visible spec- 

 trum suffers no change in wavelength when it is scattered. 



Must we now concede that radiant energy travels about through 

 space in the form of atom-like units, of corpuscles, of quanta every 

 one of w'hich, for a radiation of a specific frequency v, possesses always 

 the same energy hu and always the same momentum hv, c? How 

 indeed can we longer avoid admitting it? The phenomena which 

 1 have cited do certainly seem to close the case be>ond any possi- 

 bility of reopening it. Vet they might be interpreted in another 

 way — a way which will probably seem e.xtremely elaborate and artificial 

 to the reader, a way which will seem like a mere e.xcuse to avoid 

 a simple and satisfying explanation; and yet this would not be 

 sufficient to condemn it utterh'. We might lay the whole blame and 

 burden for all these "quantum" phenomena upon the atom. We 

 might say that there is some mysterious mechanism inside every 

 atom, which constrains it never to emit radiation of a frequency v 

 unless it has a quantity of energy hv all packed up and ready to deliver, 

 and never to absorb radiation of a frequency v unless it has a special 

 storeroom ready to receive just exactly the quantity of energy bf. 

 This indeed is not a bad formulation of Bohr's theory of the atom. 

 It would be necessary to go much further, and to say that not only 

 e\ery atom, but likewise every assemblage of atoms forming a liquid 

 or a solid body, contains such a mechanism of its own; for the phe- 

 nomena which I have called the "photoelectric effect" and the "inverse 

 photoelectric effect" are qualities not of individual atoms, but of 

 l)ieces of solid metal." And it would be necessar\- to go much funlRi 

 yet, and make mechanisms to account for the transfer of nionuntuin 

 from radiation to electrons. 



\ (.1 i\cii ill is would not be sufficient; for the most surprising and 

 inexplicable fad of all is still to be presented. Here is the crux of 

 the great dilemma. Imagine radiation of the frequency v emerging 

 from an atom, for a length of time determined by the condition that 



" It was formerly contended that this explanation, while applicable to the be- 

 havior of free atoms which respond only to certain discrete frequencies, would not 

 avail for a solid substance like sodium which delivers up electrons with energy hv, 

 whatever the frei|uency v may be. This contention, however, is probably not 

 forcible, as it can be supposed that the solid has a very great number of natural 

 frequencies very close together. This in fact was the inference from Epstein's 

 theory of the photoelectric elTecl. 



