The experiment is of such fundamental importance that it has been much repeated, with a view both to increasing the accuracy and to investigating suggested explanations of the negative effect, other than that offered by the theory of relativity. In 1904, that is, a year before Einstein first put forward his theory, Morley and Miller carried out new measurements. It had been suggested that the contraction in the direction of motion demanded by the Fitz gerald-Lorentz hypothesis (see RELATIVITY) might be a property of particular materials only, and that it would be desirable to try the effect of replacing the stone on which the interferometer parts were mounted by other substances. Morley and Miller used both a wooden and a steel framework to support the apparatus, which was floated on mercury, as in 1887. The path length was increased to 3,224 cm., which is 55 million wave lengths of yellow (sodium) light, so that the fringe displacement should, on the hypothesis of a stationary ether, be r•I fringes in place of the .4 fringes of 1887. In the cellar at Cleveland where the apparatus was mounted the effect was less than one hundredth of this amount, which is alternatively expressed by saying that these results show that the velocity of the earth with reference to a hypothetical stationary ether is at most 3 kilometers (say 2 miles) a second. (The shift is, of course, proportional to the square of the velocity, and the earth's orbital velocity is about 3o kilometres per second.) Morley and Miller took into account the motion of the solar system towards a certain fixed point in the heavens, and compounded this with the velocity of the earth in the solar system when computing the magnitude and direction of the velocity of the apparatus.
In 19o5 and 1906 Morley and Miller removed their apparatus from the cellar and installed it in a light shed at a height of 300 feet above Lake Erie, to see if possibly the nature of the imme diate surroundings, whether massive or no, affected the "ether wind" which the experiment was designed to detect. A possible effect of about of the calculated value was detected, but it was not clear whether this was a spurious effect of temperature. It would be a result of the first importance, constituting a grave blow to the theory of relativity, if an actual effect, which varied with the height above sea level or with the surroundings, could be definitely established, and therefore in 192r and again in 1925 D. C. Miller carried out experiments of the Michelson and Morley type at Mount Wilson, about 1,800 metres (6,000 feet) above sea level. To avoid possible magnetic effects the steel frame used in the earlier experiments was later replaced by a concrete frame. Miller, as a result of several thousand readings, arrived at the conclusion that there was a small positive effect, about 9 kilo meters per second, as against the full orbital velocity of 3o kilo meters per second, and also considered that the experiments in the cellar at Cleveland showed a still smaller positive result. He attempted to explain his results by a drift of the solar system in a direction nearly normal to the plane of the ecliptic. These experiments of Miller naturally attracted much attention, but it appears almost certain that the positive effect must be due to unrecognised sources of error, possibly to temperature effects. It has, for instance, been pointed out, notably by Thirring, that the measurements of Miller are not consistent among themselves, from the point of view of his explanation.
To check Miller's results R. J. Kennedy repeated the experi ment, using an ingenious modification of the interferometer, pro duced by making a small step, only one-twentieth of a wave length in height, on one of the totally reflecting mirrors. This produces two fringe systems bordering on one another, which enables a very slight displacement of the interference patterns, of the order of 4 thousandths of a fringe, to be detected. Ken nedy's apparatus was comparatively small, the light path being only 4 metres, and sealed up in a case containing helium, the re fractivity of which is only one-eighth that of air, at atmospheric pressure. The experiments were carried out at times of day when Miller's effect should be greatest, but no positive effect was ever found, although a shift one-quarter as great as that meas ured by Miller would have been detected. K. K. Illingworth has repeated Kennedy's experiments, likewise finding no posi tive result (1927). Both Kennedy's and Illingworth's experi ments were carried out in a constant temperature basement at Pasadena, California, which gives defenders of the reality of Miller's effect a possible line of criticism, namely that while no positive result can be found in a sheltered spot, where massive surroundings check the "ether wind," this does not prove that a small positive effect does not exist at a height, in a free situa tion. This last possibility is, however, removed by the experiments of Piccard and Stahel, carried out at intervals from 1926 to 1928. These workers placed the apparatus in a free balloon, the chief measurements being made at a height of 2,500 metres. The apparatus was a small interferometer of Michelson-Morley pat tern, with multiple reflections. The fringes were registered photo graphically, on a moving film, and the balloon was kept in steady rotation by small air screws. No trace of an effect was ever found.
In January, 1929, Michelson published a preliminary account of a repetition of the Michelson-Morley experiment carried out by himself in collaboration with Drs. Pease and Pearson, with new refinements : the results, in contradiction to Miller's, were negative.
We may theref ore say that the broad result of a series of experiments of Michelson-Morley type is that no motion of the earth through the ether can be detected by an influence on the velocity of light. For theoretical implications see RELATIVITY.
BIBLIOGRAPHY.—Brief accounts of the Michelson-Morley experiment are given in the standard text-books of light, cited under LIGHT, and the experiment is discussed in all books on relativity (q.v.). The most important original papers are:—A. A. Michelson and E. W. Morley, Philosophical Magazine, 449 (1887) ; E. W. Morley and D. C. Miller, Philosophical Magazine, 9, 68o (19o5) ; D. C. Miller, Proceedings National Academy of Sciences, ii, 3o6 (1925) ; H. Thir ring (criticism of D. C. Miller's results), Zeitschrift fur Physik, 35, 723 (1926) ; J. R. Kennedy, Proceedings National Academy of Sci ences, 12, 621 (1926) ; K. K. Illingworth, Physical Review, 3o, 692 (1927) ; A. Piccard and E. Stahel, Naturwissenschaften, 14, 935 (1926) ; 15, 140 (1927) ; 16, 25 (1928) ; A. A. Michelson, Nature, 123,