2 3 2 D O C U M E N T 2 2 3 E X P E R I M E N T O N R A D I A T I O N 223. “Proposal for an Experiment on the Nature of the Elementary Process of Radiation Emission” [Einstein 1926p] Dated 16 March 1926 Published 2 April 1926 In: Die Naturwissenschaften 14 (1926): 300–301. According to the classical wave theory of light, the interference capacity of monochromatic light emitted by an atom arises because charged masses execute harmonic oscillations that have the same frequency as the emitted light. Pursuant to the Maxwell equations, the emission process is such that each wave peak in the spatial process corresponds to one part of an oscillation of the atom’s charged masses by which the relevant wave peak was generated. The periodicity of the atomic oscillation hence appears to be the cause of the ordering relation that seems to exist between the various parts of the emitted train of waves (interference phe- nomena). On the other hand, however, according to quantum theory, the frequency of the radiation is connected to the amount of energy emitted in the elementary process. According to Bohr’s original theory of spectra,[1] it is impossible to assume the ex- istence of a frequency of electron motion equal to the frequency of light. The Compton effect particularly speaks against the existence of a periodic process by a diffracted electron, however, which produces diffracted radiation in the sense of classical wave theory. Thus we are led to assume that, generally, the sinuslike char- acter of the kind of field of waves corresponding to an elementary process (as is manifested in interference phenomena) is not determined by the emitted atom or electron, but rather by an intrinsic lawfulness of the space-time continuum. In the following, I indicate an experiment whose negative result would be irrec- oncilable with the classical wave theory. I have long had this experiment in mind. But an investigation performed by E. Rupp (interference analyses of canal rays, Ann. d. Phys. IV, 79, 1, 1926), to which Mr. Grotrian kindly drew my attention,[2] gave me the initial proof that it really was practicably possible to perform this ex- periment successfully. The underlying idea is the following: [p. 300]