D O C U M E N T 2 8 9 M A Y 1 9 2 6 4 7 9 ALSX. [70 707]. There are perforations for a loose-leaf binder at the left margin of the document. [1]See experimental setup I with a grid of 0.1 mm in Doc. 284. [2]Setups I with a grid of 0.05 mm and III, respectively. [3]Setup II. [4]In the manuscript of Doc. 278, sent enclosed in Doc. 279. [5]Philipp Lenard, who was ill-disposed toward Einstein. [6]Latin for “as matters stand.” 289. From Dayton C. Miller Cleveland, May 20, 1926. My dear Professor Einstein, I regret exceedingly the long delay in replying to your very kind letter which was received in the midst of a series of observations at Mount Wilson.[1] There is such a mass of observations that I have been completely absorbed in the work of reduc- ing them. The observations were continued through February of this year. Only now have I been able to combine all of the observations so as to form a definite idea of their significance. I hope you will pardon me for not writing before. Your suggestion is quite correct that a difference of 1/10° in the temperature of air in the light-path in the area of the interferometer would produce a displacement of the fringes of the amount observed. It is this possibility that has made necessary such a long continued series of observations: Professor Morley and I felt that this might have been the cause of our observed effect in 1905.[2] Very elaborate precau- tions have been taken to eliminate such an effect of temperature. The effects which are obtained are regularly periodic in each half-turn of the in- terferometer. Usually the apparatus makes one turn in about fifty seconds, so that the effect is periodic in each twenty-five seconds. Such temperature disturbances certainly would vary in time and in azimuth, and thus would be cancelled out in a long series of observations containing several thousand determinations of the peri- odic function. Perhaps, my present argument may be stated by saying that in addi- tion to all temperature and instrumental effects, and effects produced by the diurnal and annual motions of the earth, there seems to be a very real effect which is related to sidereal time and thus must have cosmical relations. In a paper presented to the National Academy of Sciences in April 1925,[3] the conclusion was stated that the ether-drift experiments at Mount Wilson give con- clusive evidence of a positive effect as of an ether-drift, corresponding to a relative motion of the earth and ether of about ten kilometers per second. The observations made at Mount Wilson in 1921 and 1924 consist of 1200 turns of the interferome- ter, while those of April 1925, consist of 800 turns, each turn giving two determi- nations of the ether-drift function. Having found that there is a positive effect, it