I N T R O D U C T I O N T O V O L U M E 1 5 l x i ter of the Earth completely drags the ether along with it, so that no ether drift could be detected at the surface of the Earth, still it must be true that in space far from the Earth the ether must be unaffected by the Earth’s motion. Accordingly, it was nat- ural to suppose that a gradient must exist, so that, at sufficiently high altitudes above the Earth, some ether drift ought to be measurable. At the time, Miller had already begun a new series of experiments in a lightly constructed building at Mount Wilson, using substantially the same apparatus. It was there that he claimed to have detected a positive ether drift, in contrast to the earlier null result in the Cleveland basement. He always claimed, however, that there had been a similar, but smaller, positive effect in the slightly elevated location in Cleveland. The results of Miller’s 1921 experiments were published the next year (Miller 1922) and reached Einstein through a letter from Max Born of 6 June 1922 [Vol. 13, Doc. 320]), who commented: “The Michelson experiment belongs to things that are ‘practically’ a priori I believe not a single word of the rumor.” Einstein took a similar view. Over the next three years Miller published only brief reports. By 1925, finally convinced that he was consistently seeing an ether drift of some kind on Mount Wilson, he published at greater length and in multiple places. It was the editor of the journal Science, Edwin E. Slosson, who sent Einstein the proofs of two papers by Miller in June 1925 (Miller 1925a and 1925b), asking for Einstein’s opinion (Doc. 12). Einstein replied cautiously: experiment being the su- preme judge, he was awaiting more complete details (Doc. 13). The same reserve can be seen in a letter to Robert A. Millikan: “Miller’s experiments rest on sources of error. Otherwise the entire theory of relativity collapses like a house of cards” (Doc. 20). Most physicists expressed serious doubts, first among them Millikan and his staff at Caltech, who were in a position to see the apparatus on Mount Wilson. Al- ready in July 1925 they launched a “counterattack,” both on the heights of Mount Wilson and on the lower level of Caltech’s laboratories, because, as Paul Epstein wrote to Einstein, “our circle accepts Miller’s somewhat daring statements with great reserve we hope we will be in the position to check his measurements by other observers with other kinds of instrument” (Doc. 31). Upon Einstein’s inquiry as to whether Millikan and Epstein were intending to “gather new evidence on the problem of ‘ether drift’” (Doc. 58), Epstein disclosed that they intended to repeat Michelson’s classical experiment with a modified apparatus and that the idea be- longed to the young student Roy Kennedy (Doc. 72). A year later, Epstein reported that the “repetition of the Michelson experiment by Kennedy had led to a complete- ly negative result, even though the sensitivity of the apparatus was four to five times higher than that of Miller’s” (Doc. 372 and Abs. 662). Epstein’s impression had grown more and more certain that “the whole story is a humbug,” and that Miller did not understand his results and had drawn conclusions not supported by