MICHELSON-MORLEY EXPERIMENT. A celebrated experiment carried out by A. A. Michelson and E. W. Morley, to attempt to measure a velocity of the earth through the ether by the effect which such a velocity might be anticipated to have on the velocity of light. Had such a velocity been established, axes fixed in the ether could have been taken as standards of fixed position, to which all velocities could be referred, and in this sense we should have been able to talk of the absolute velocity of a material body, as distinct from the relative velocity of two material bodies. The failure of Michelson and Morley to detect any influence of the earth's motion on the velocity of light really formed the starting point of Einstein's theory of relativity (see RELATIVITY), and it is on account of its importance for this theory that the experiment is so fundamental, and has been so often repeated.
In the experiment light is sent along two paths at right angles to one another. If the luminiferous ether were a fixed medium, obeying the laws of such a wave-carrying medium as an elastic solid (or, to be more correct, if less suggestive, of a medium obey ing the laws represented by Maxwell's original equations) then the time of the passage of the light should be greater for a to-and-fro path in the direction of the earth's motion through the ether than for a to-and-fro path, of equal length, at right angles to the direction of motion. More precisely, when light goes from a source to a mirror at distance 1 and back, and both source and mirror are moving with the same velocity v, the time taken is c—v when the motion is in the direction of the rays, and is smallness of all terrestrial velocities compared to the velocity of light, is very small and requires the greatest refinement of experi mental skill for its detection. (See RELATIVITY.) If, therefore, the apparatus can be set so that one path lies in the direction of motion,. and can then be turned through a right angle, the two beams of light will interchange their roles, the one which in the first position travelled the faster travelling the slower in the second position. If, further, a system of inter ference fringes is formed by the two beams, as is done in the Michelson interferometer (see INTERFEROMETER), then it follows at once that this fringe system will shift when the apparatus is slowly rotated, so that it must pass through the two positions just specified twice per revolution. Michelson and Morley looked
for, but failed to find, any shift of fringes of the order of mag nitude to be expected on the hypothesis of a fixed ether.
Preliminary experiments were carried out by Michelson alone in 1881; the experiment with Morley, of a superior order of accuracy, was performed in 1887. In the experiment of 1887 the interferometer was mounted on a heavy block of stone which was carried by a disc of wood floating in a tank of mercury : this arrangement enabled the apparatus to be rotated smoothly with out causing any strains to be set up. To increase the light path, and so the effect to be sought, each of the interfering beams was reflected several times backwards and forwards : four mirrors were used to replace each of the single mirrors in the simple form of interferometer, and their supplementary mirrors. The total length of the path of either beam was about r,roo cm. With such a path length the difference of times to be expected on the fixed ether hypothesis, though only about r in roo,000,000, should give rise to a displacement of .4 of a fringe. But the actual displacement measured was at most a fortieth of this value. In carrying out the experiment the orbital motion of the earth alone was considered, the sun being considered at rest. (The rotational motion of the earth produces a velocity which is, of course, negligibly small compared to the orbital motion.) Read ings were made for sixteen different directions, the stone being kept slowly turning and the observations taken by the experi menter walking round with it. The experiment was carried out at midday, when the velocity relative to the sun is in the plane of the apparatus, and at 6 P.M., when the velocity is normal to the plane of the apparatus. To see if there was a velocity of the solar system as a whole through the ether, which might, by a remote chance, have cancelled out the orbital motion at the particular month of the experiment, it was proposed to repeat the experiment every three months. Recent experiments have put this possible explanation of the negative effect out of court.