l x x i v I N T R O D U C T I O N T O V O L U M E 1 5 random thermal motion (Atkinson 1926). Philipp Lenard, director of the Heidel- berg institute where Rupp carried out the experiments, later remarked that Rupp made his observations of the interference fringes with the naked eye, but that he himself could not see anything because of their low intensity (Dongen 2007a, p. 112). In Doc. 223 (Einstein 1926p), Einstein proposed an experiment in which the light emitted by the canal rays would be directed at a wired grid and analyzed by a Michelson interferometer using Rupp’s techniques. Einstein expected that, if the wave theory was right, this should result in an “intermittent wave train.” If the in- terference pattern was not influenced by the presence of the grid, then, Einstein ar- gued, it would follow that the interference patterns of light had nothing to do with the atom emitting it. This, he believed, would speak against the wave theory of light. Given a grid spacing of 1/10 mm, Einstein determined that the wave trains produced by hydrogen canal rays should be 6 cm long that is, within reach given the precision obtained by Rupp for this type of experiment. Einstein’s “Proposal” elicited criticism. Georg Joos, skeptical of Rupp’s results, intended to show that the experiment would lead to the same result regardless of whether light is emitted in waves or in quanta (Doc. 266). Paul Ehrenfest shared this opinion (Doc. 248). Gustav Mie offered a special relativistic consideration that predicted needle radiation, that is, quantum emission (Doc. 268). Einstein approached Rupp on 20 March 1926, before these critical remarks had reached him (Doc. 231), and asked him to perform his proposed experiment. Rupp accepted and said that he had been planning a similar experiment (Doc. 233). Einstein was pleased. He sketched several setups and proposed to publish the even- tual results jointly (Doc. 240). The present volume contains as full text thirty letters exchanged between Einstein and Rupp, at the time Privatdozent in physics at the University of Heidel- berg (see Illustration 26), whose daring papers eventually turned out to be based on non-reproducible experimental results (Dongen 2007a). Rupp repeatedly wrote that he had experimentally proved what Einstein wanted to see, and Einstein re- sponded with detailed criticism, pointing out how the experimental setup ought to be modified before the results could be trusted. Rupp carried out the experiments suggested by Einstein between late April and early November 1926. Throughout, they kept in written contact. Rupp used two types of experimental setup: one contained a wire grid, while a second, also pro- posed by Einstein, contained a lens in place of the grid. In both cases a Michelson- type interferometer was attached to the detector. In the second type, the Doppler shift resulting from the motion of light sources along the beam was compensated by