cancels the full-period effect and all odd harmonics, giving the shorter curve which is the desired half-period effect (together with any higher even harmonics which may be present). Inspection shows clearly that these curves are not of zero value, nor are the observed points scattered at random; there is a positive, systematic effect. These full-period curves have been analyzed by the mechanical harmonic analyzer, which determines the true value of the half-period effect; this, being converted into its corresponding value for the velocity of relative motion of the earth and ether, gives a velocity of 8.8 kilometers per second for the noon observations, and 8.0 kilometers per second for the evening observations. In Fig. 4, the smooth
Fig. 4. Velocity of ether drift observed by Michelson and Morley in 1887, and by Morley and Miller in 1902, 1904 and 1905, compared with the velocity obtained by Miller in 1925.
curve shows the value of the ether-drift throughout the day for the latitude of Cleveland, as determined by the specifications of the drift which are derived later in this report from the observations made at Mount Wilson. The two circles on this chart show the magnitude of drift actually obtained by Michelson and Morley for the noon and evening observations, indicating a result wholly consistent with the later work here reported.
The fact that the result obtained by Michelson and Morley was not negligibly small was very fully set forth by Professor Hicks of University College, Sheffield, in 1902, in his important theoretical examination of the original experiment.2 Hicks also called attention to the presence of a full-period, first-order effect, which has never been sufficiently investigated; this first-order effect will be considered later.
The Lorentz-Fitzgerald Hypothesis
The Michelson-Morley experiment, which indicated that the theory of the ether was either incomplete or incorrect, attracted world-wide attention because of its fundamental character and because the result was wholly unexpected. Professor FitzGerald of Dublin, in 1891, offered an explanation for the small effect on the hypothesis that the forces binding the molecules of a solid might be modified by the motion of the solid through the ether in such a way that the dimension of the stone base of the interferometer would be shortened in the direction of motion and that this contraction might be such as to neutralize the optical effect sought in the Michelson-Morley experiment. FitzGerald did not publish this theory in a scientific journal but he expounded it in his lectures. This hypothesis was given publicity by Sir Oliver Lodge in his address on Aberration Problems and New Ether Experiments, presented to the Royal Society on March 31, 1892, which address was published in the Philosophical Transactions for the year 1893.5 Lodge has given further details of this historical fact in his recently issued autobiography.6 In 1895 Professor Lorentz of Leyden developed the theory in a systematic manner, on the supposition that the particles of all solids are held together by electrical forces; and that a motion of the body as a whole would superpose upon the electrostatic forces between the atoms a magnetic effect due to the motion. There would result a contraction of the body in the direction of motion which is proportional to the square of the ratio of the velocities of translation and of light and which would have a magnitude such as to annul the effect of ether-drift in the Michelson-Morley interferometer.7 If the contraction depends upon the physical properties of the solid, it was suggested by others that while the expected effect might be annulled in one apparatus, it might in an apparatus of different material give place to an effect other than zero, perhaps with a contrary sign.
5 G. F. FitzGerald, see O. J. Lodge, Aberration Problems, Phil. Trans. Roy. Soc. 184, 749 (1894).
6 Sir Oliver Lodge, Past Years, 204 (1932).
7 H. A. Lorentz, Versuch einer Theorie der electrischen und optischen Erscheinungen in bewegten Körpern (Leyden, 1895); Theory of the Electron, 195 (1909).
The Morley-Miller Experiments, Cleveland, 1902-1906
The interferometer of wood, 1902
At the International Congress of Physics held in Paris in connection with the International Exposition of 1900, Lord Kelvin gave an address in which he expounded certain theories of the ether, and he explained the significance of the results of the Michelson-Morley experiments as related to these theories.8 Professor Morley and the writer were present and in a later conversation with Lord Kelvin he strongly urged the repetition of the ether-drift experiment with a more powerful apparatus. Morley and Miller then constructed an interferometer designed especially to test the Lorentz-FitzGerald hypothesis. The base of this instrument was in the form of a cross, made of planks of white pine wood about 430 centimeters long, providing a light-path more than three times as long as that used by Michelson and Morley in 1887. The general dimensions, optical parts and methods of observing with this apparatus were the same as for the steel interferometer described in detail in following sections of this paper. The instrument was mounted in the northwest corner room of the basement of the Main Building of Case School of Applied Science and three series of observations were made in August, 1902, and in June, 1903, consisting of 505 turns of the interferometer. A small positive effect was observed, indicated by the square in Fig. 4, which, while slightly larger than that of the previous experiment, was still so small as to indicate that if the reduction of the observed velocity is to be attributed to the hypothetical contraction, the pine is affected by about the same amount as is the sandstone. The changes in the wooden support due to variations in humidity and temperature made it difficult to obtain accurate observations and it was decided to abandon the pine apparatus and to construct one having a base of metal for supporting the heavy parts, while the length of the optical path could be determined by various substances, wood or metal, as desired.
While planning a new apparatus, experiments were made to show that differences of magnetic
8 Lord Kelvin, Rapports présentés an Congrès International de Physique 2, 1 (1900).
attraction on the iron parts of the instrument could not influence the observations. Massive bars of iron were suspended at the opposite ends of one of the long arms of the cross, so that one bar should be parallel to the earths magnetic field while the other was transverse to this field, these relations being reversed on reversing the azimuth of the apparatus. Observations with this load gave the same results as before. In a further experiment, an analytical balance was placed on one arm with which to weigh a bar of iron having a mass of about 1200 grams. It was so oriented that at one azimuth of the apparatus the bar was parallel to the lines of the earths magnetic field, while at another it was transverse to the field. A difference of half a milligram could have been detected but no such difference existed. By observing the effect produced by a known weight on one arm of the interferometer, it was shown that the earths magnetism could not be a disturbing factor.
Description of the new steel interferometer
An appropriation from the Rumford Fund of the American Academy of Arts and Sciences made possible the construction, in 1904, of an entirely new apparatus of steel. The design for the base of the interferometer, made by Professor F. H. Neff of the Department of Civil Engineering of Case School of Applied Science, provided that all optical parts and accessories should be carried by two girders of structural steel, Figs. 5, 10 and 14, each about 430 centimeters long, which intersect in the form of a cross. The purpose of this design was to secure structural symmetry and the utmost rigidity.
The steel cross rests on a circular float of wood, Fig. 5, 150 centimeters in diameter; on the under
Fig. 5. Cross section of the mercury float for the interferometer.
