D O C . 2 7 8 O N L I G H T E M I T T E D B Y C A N A L R A Y S 4 6 7 Completed before 9 May 1926 (see Doc. 279). A manuscript, [1 054] and [1 055], is also available. [1]Einstein had stated this assumption in Doc. 262. [2]See Wien 1923. [3]Einstein 1926p (Doc. 223). [4]The equation is obtained from a Taylor expansion of the general equation for the relativistic Doppler effect see also Doc. 240, note 11. [5]In the manuscript, “die so gebildete Lichtquelle.” replaces “das so gebildete zweite Bild.” [6]The sentence is missing in the manuscript. [7]The angle β is first introduced in Doc. 240. [8]Given the arrangement sketched, one will only obtain an interference pattern if the mirror is to be turned in the opposite direction from that indicated by Einstein in this figure. For with the canal ray moving upward, the light emitted at the top end of the figure will be slightly red-shifted, whereas light emitted at the bottom end will be slightly blue-shifted. Thus, in order to compensate for this dif- ference in frequency, the red-shifted light ought to have a slightly shorter path until it reaches the mir- ror than the blue-shifted light hence, the mirror should be turned counterclockwise. When summing up the results of the Munich group in re-doing the experiment in the 1930s, which gave results com- pletely at odds with those reported by Rupp, Gerlach and Rüchardt 1935 pointed out the mistake in Einstein’s diagram and concluded that Rupp could not have observed an interference pattern if he had indeed followed Einstein’s instructions. See the Introduction, sec. VII, for further discussion. [9]Einstein has shifted the question compared with Einstein 1926p (Doc. 223). There, he had argued that the wire grid experiment would put pressure on the wave theory of light by showing that the frequency of the emitted radiation does not coincide with the frequency of the electron assumed to orbit the emitting atom core. Indeed, he assumed exactly this relationship on p. 335 of the present paper. His conclusion at this point of the paper, however, concerns the question of whether the emitted radiation is emitted instantaneously or continuously. The connection is likely via the following line of thought: if light is emitted according to the classical theory, then the emission of one full light wave corresponds to several oscillations of the emitting atom if the emission takes place via Bohr’s quan- tum theory (see Doc. 223, note 1), then light might well be emitted instantaneously, given that the change in (discrete) energy levels of the atom bringing about the emission of light is supposed to take place instantaneously. The mention of the views of Bohr and Heisenberg might refer to a conversation Einstein had with Werner Heisenberg after the latter’s talk in Berlin on 28 April 1926 (see Heisenberg 1969). [10]Footnote 1 is missing in the manuscript. [11]Einstein had used this terminology for contrasting the geometric and the energetic properties of light emission before, especially in Vol. 13, Doc. 29 and Vol. 14, Doc. 484. By “energetic” proper- ties—Einstein means properties and processes in which the energy, momentum, and direction of a light beam play the primary role—in particular, the processes of absorption, emission, and scattering of light, most easily explained using a quantum account of light. The “geometric” properties of light relate to processes like refraction, diffraction, dispersion, and interference, and get their name from geometric optics, which is a limiting case of the classical wave theory of light. See Doc. 49, note 5, for more on the Bohr-Kramers-Slater theory of radiation. [12]In the original manuscript, the caption giving the diameter is given as , like in fig. 1 above. [13]See also Einstein 1926p (Doc. 223), note 1. [14]For Einstein’s “original expectations” regarding the parallel displacement of the light source, see Doc. 262. [15]Rupp 1926b. Einstein presented it on 21 October 1926 (see Einstein 1926w [Doc. 387]). S1 d 2 -- -