DAYTON C. MILLER
points to within 120 of this apex. One cannot help wondering whether there may be some dynamic significance in these facts.
The argument now being presented can be demonstrated only by means of observations extending over the whole twenty-four hours of the day, in order to determine the exact form of the daily variation in magnitude and azimuth of the effect, and by means of observations made at different times of year, in order to prove that the effect is dependent on sidereal time. The earlier observations of 1887 and 1905 are not sufficiently numerous and are not distributed throughout the day in such a manner as to make it possible to calculate the direction of the drift. These earlier observations were made for the purpose of detecting the earth’s orbital motion and consequently were made at two selected times of day, such that at one time the magnitude of this particular effect would be a maximum and at the other time it would be zero; or, two times of day were chosen in which the azimuth of the orbital component of motion would have very different values. Furthermore, until the year 1925 the experiments have never been carried out at intervals of six months. The reason that a second set of experiments has not been made after this interval before is simply that in no instance has the expected effect been found in a first set.
The observations made at Cleveland by Michelson and Morley in 1887, and later repeated by Morley and Miller, have just been recomputed on the present hypo thesis; while the earlier observations are not sufficient to determine the direction of the drift, they are nevertheless shown to be entirely consistent with the conclusions drawn from the Mount Wilson experiments. Or, to state the converse, the present result wholly confirms the earlier experiments of Michelson and Morley, giving no evidence of the effect of the earth’s orbital motion. In addition to this, the recent experiments, by a thorough study of the residual effects, have shown that there is a systematic cosmical effect as of a true ether drift. This conclusion introduces a new question, “Why is the magnitude of the effect less than would be expected on the classical theories and why is the direction of the effect at Mount Wilson deflected to the westward?” This question certainly is no more difficult than are many others now awaiting solution.
CONFERENCE ON MICHELSON-MORLEY EXPERIMENT 367
The interferometer is being set up again on the campus of Case School of Applied Science in Cleveland, near the place where the original Michelson-Morley experiment was performed in 1887. It is proposed to make a series of observations for four epochs of the year, comparable in every way with the Mount Wilson series. This will give information as to the possible effects of local conditions; it is hoped that it will show more definitely whether there is any effect due to altitude, and whether the orbital motion is appreciable.
IV. DR. ROY J. KENNEDY (CALIFORNIA INSTITUTE OP TECHNOLOGY)
When Professor Miller published the conclusions that he presented to us yesterday, it became necessary, or at least very desirable, that the experiment be repeated independently. It is such a performance of the experiment that I shall discuss this morning.
In this experiment the light-paths were reduced to about 4 m, and the required sensitiveness was obtained by an arrangement capable of detecting a very slight displacement of the interference pattern. The whole optical system was inclosed in a sealed metal case containing helium at atmospheric pressure. Because of its small size, the apparatus could be effectively insulated, and circulation and variations in density of the gas in the light-paths nearly eliminated. Furthermore, since the value of /x — 1 for helium is only about one-tenth that for air at the same pressure, it will be seen that the disturbing effects of changes in density of the gas correspond to those in air at only a tenth of an atmosphere of pressure. Actually it was found that any wavering of the interference pattern was imperceptible, and when temperature equilibrium had been reached there was no steady shift.
The plan of the apparatus is sketched in Figure 8. The optical parts are mounted on a marble slab 122 cm square by 10.5 cm thick, which rests on an annular float in a pan of mercury 77 cm in diameter. This is simply a reduced copy of Michelson’s original mounting. The mirrors Mz, MA9 and Ms are fixed in position; such adjustments of the compensating plate C and mirror M2 as are necessary after the cover is in place can be made from the observer’s position at the telescope. The green light X 5461 is separated by the lens and prism system from the radiation of a small mercury arc lamp