3 0 6 D O C U M E N T 2 9 0 M A Y 1 9 2 6 into account. It is equally incomprehensible why the omission of the lenses L1, L2 destroyed the interference.[2] The setup with grating and one lens is theoretically totally non-transparent. The destruction of the interference seems logical in this case.[3] It would be good for you to look into my considerations more closely.[4] A joint publication absolutely requires a clear confrontation of theory and experiment. If the experiments do not match the theory in part, the systematic aspects of the non- agreement must be shown. How sorry I am that I can’t be there, and only because of the fickle-mindedness of Mr. Principal![5] But there is nothing we can do. Maybe one day you will have an opportunity to come to Berlin, where we can calmly talk things through. In any case, the grating experiment rebus sic stantibus[6] cannot be regarded as finished despite the one successful attempt. Kind regards, A. Einstein 289. From Dayton C. Miller Cleveland, 20 May 1926 [See documentary edition for English text.] 290. From Emil Rupp [Heidelberg,] 20 May 1926 Dear Professor, A thorough consideration of your theoretical paper[1] prompted me to perform the grating experiment again this week. I am in receipt of your valuable letter of May 18th[2] just as I am busy setting these experiments up. I can now report findings to you that agree with the theory, as well as seemingly systematic discrepancies. In setup no. I,[3] as described, I have now inserted a 2 mm wide diaphragm in place of the wide diaphragm B. Furthermore, I have substituted a trough made of mirror plates with cuproammonia for the green-glass filter. 1) If this diaphragm stands directly at the end of the canal (as in Ann. p. 7),[4] then I obtain the following decrease in intensity: