MICHELSON'S RECENT RESEARCHES ON LIGHT.*

By JOSEPH LOVERING, President.

For many generations it was assumed that no sensible time was taken by light in moving over the largest distances. The velocity of sound was found by noting the time which elapsed between seeing the flash and hearing the report of an explosion. It was only in the vast spaces of astronomy that distances existed large enough to unmask the finite velocity of light, and, in extreme cases, to make it seem even to loiter on its way.

The satellites of Jupiter were discovered by Galileo in 1610; and the eclipses of these satellites by the shadow of Jupiter became an interesting subject of observation. It was soon noticed that the interval between successive eclipses of the same satellite was shorter when the earth was approaching Jupiter, and longer when the earth was receding from Jupiter. The change of pitch in the whistle of a locomotive, under similar motions, would suggest to the modern mind an easy explanation. A Danish astronomer, Römer, without the help of this analogy, deciphered the problem in astronomy. The eclipse was telegraphed to the observer by a ray of light, and the news was hastened or delayed in proportion to the distance from which it came. In this way it was discovered that light took about eighteen minutes to run over the diameter of the earth's orbit. This discovery was published by Römer in the Memoirs of the French Academy in 1675. The mathe matical astronomer Delambre, from a discussion of one thousand of these eclipses observed between 1662 and 1802, found for the velocity of light 193,350 miles a second.

Meanwhile Römer's method, after fifty years of waiting, had been substantially confirmed in an unexpected quarter. Dr. Bradley, of the Greenwich Observatory, the greatest astronomical observer of his day, was perplexed by certain periodical fluctuations, of small amount, in the position of the stars. Suddenly the explanation was flashed upon him by something he observed while yachting on the River Thames. He noticed that, whenever the boat turned about, the direction of the

*An address delivered before the American Academy of Arts and Sciences, at the meeting of April 10, 1883, when the Rumford medals were presented to Prof. A. A. MICHELSON. (From the Proceedings of the American Academy; vol. XXIV (n, s. XVI), pp. 380-401.)

H. Mis. 224--29

449

vane altered. He asked the sailors, Why? All they could say was, that it always did. Reflecting upon the matter, Bradley concluded that the motion of the boat was compounded with the velocity of the wind, and that the vane represented the resultant direction. He was not slow in seeing the application of this homely illustration of the parallelogram of motion to his astronomical puzzle. The velocity of light was compounded with the velocity of the earth in its orbit, so that its apparent direction differed by a small angle from its true direction, and the difference was called aberration. In spearing a fish or shooting a bird, the sportsman does not aim at them, but ahead of them. inclination from the true direction is similiar, in angular measure, to what the astronomer calls aberration. Struve's measurement of aberration combined with the velocity of the earth in its orbit gave for the velocity of light 191,513 miles a second. Both of the two methods described for obtaining the velocity of light depend for their accuracy upon the assumed distance of the earth from the sun. The distance adopted was the one found by the transits of Venus in 1761 and 1769, viz. 95,360,000 miles.

This

During the last forty years, the opinion has been gaining ground among astronomers that the distance of the sun, as deduced from the transits of Venus in 1761 and 1769, was too large by 3 per cent. Expeditions have been sent to remote parts of the earth for observing the planet Mars in opposition. The ablest mathematical astronomers, as Laplace, Pontecoulant, Leverrier, Hansen, Lubbock, Airy, and Delaunay, have applied profound mathematical analysis to the numerous perturbations in planetary motions, and proved that the sun's distance must be diminished about 2,000,000 miles in order to reconcile observations with the law of gravitation. Airy reduced the distance of the sun by more than 2,000,000 miles, to satisfy the observations on the transit of Venus in 1874. Glasenapp derived from observed eclipses of Jupiter's satellites a distance for the sun of only 92,500,000 miles. From these and similar data, Delaunay concluded that the velocity of light is about 186,420 miles a second.

These triumphs of astronomical theory recall the witty remark of Fontenelle, that Newton, without getting out of his arm chair, calculated the figure of the earth more accurately than others had done by travelling and measuring to the ends of it. And Laplace, in contemplation of similar mathematical achievements, says: "It is wonderful that an astronomer, without going out of his observatory, should be able to determine exactly the size and figure of the earth, and its distance from the sun and moon, simply by comparing his observations with analysis; the knowledge of which formerly demanded long and laborious voyages into both hemispheres."

The ancients supposed that light came instantaneously from the stars; a consolation for those who believed that the heavens revolved around the earth in twenty-four hours. Galileo and the academicians of Florence obtained even negative results,

While the number of physical sciences has received numerous additions during the last half-century, new affiliations and a more intimate correlation have been manifested. In this mutual helpfulness light has played an important part. The optical method of studying sound, and the many varieties of flame apparatus, have made acoustics as intelligible through the eye as through the ear.

Velocity being expressed by space divided by time, it is evident that in measuring an immense velocity we must have at our command an enormous distance, such as we find only in astronomy, or else possess the means of measuring fractions of time as small as one-millionth of a second. The first successful attempt to measure such a velocity was made by Wheatstone in 1834. Discharges from a Leyden jar were sent through a wire, having two breaks in it one-fourth of a mile apart. The wire was in the form of a loop, so as to bring the breaks into the same vertical line. The sparks seen at these breaks were reflected by a mirror at the distance of 10 feet, and revolving eight hundred times per second. The images of the two sparks were relatively displaced in a horizontal direction. As the displacement did not exceed one-half of an inch, the time taken by electricity to go from one break to the other was less than a millionth of a second. Since the distance was onequarter of a mile, the electricity travelled in that case at the rate of 288,000 miles a second. If this experiment is interpreted to mean that electricity would go over 288,000 miles of similar wire in one second, as it probably often was at that time, the conclusion is fallacious. The velocity of electricity, unlike that of sound or light, diminishes when the length of wire increases.

In 1838, Wheatstone suggested a method for measuring the velocity of light, which he thought was adequate for giving not only the absolute velocity but the difference of velocity in different media.

In that year Arago communicated to the French Academy the details. of an experiment which he thought would give the velocity of light in air or a vacuum. As his own health was broken down (he died in 1853) he appealed to two young French physicists to undertake the experiment. On July 23, 1849, Fizeau, by a method wholly his own, made a successful experiment. A disk cut at its circumference into 720 teeth and intervals, and made by Breguet, was rapidly rotated by a train of wheels and weights. A concentrated beam of light was sent out through one of the intervals between two teeth of the disk, which was mounted in a house in Suresne, near Paris, and was sent back by a mirror placed on Montmartre, at a distance of about 5 miles. The light, on its return, was cut off from the eye or entered it, according as it encountered a tooth or an interval of the disk. If the disk turned 12.6 times in a second the light encountered the tooth adjacent to the interval through which the light went out. With twice as many rotations in the disk the light could enter the eye through the adjacent interval. With three times the original velocity, it was cut off by the next tooth but

༄༣, ༤ ་།་་༥་༣;་་་་

one, and so on. From the number of teeth and the number of rotations in a second the time taken by the light in going and returning was easily calculated. In this way the velocity of light was found to be 195,741 miles per second. In 1856, the Institute of France awarded to Fizeau the Imperial prize of 30,000 francs in recognition of this capital experiment.

In 1862, Foucault succeeded in measuring the velocity of light by a wholly different method, all parts of the apparatus for it being embraced within the limits of his laboratory. The light emanated from a fine reticule, ruled on glass and strongly illuminated by the sun. It then fell upon a plane mirror revolving four hundred times a second, by which it was reflected successively to five other mirrors, the last of which was plane, and returned it back by the same path to the revolv ing mirror and reticule. The total distance traveled was only about 66 feet. As the revolving mirror had moved while the light was making this short journey, the image of the reticule was displaced in reference to the reticule itself; and this displacement was the subject of measurement. Although the time involved was only about one fifteenmillionth of a second, this brief interval was translated by the method of the experiment into a measurable space, and gave 185,177 miles per second for the velocity of light, differing from the best results of astronomical methods by only 1,243 miles. Foucault was prompted to this experiment by Leverrier, director of the observatory. Arago was the first to propose the experiment. To obtain greater accuracy be placed the moving mirror in a vacuum, but without any advantage. He said, "Le mieux est l'ennemi du bien." His modest claim was that he had suggested to Foucault the problem and indicated certain means of resolving it. Babinet thought that the experiment admitted of ten times greater accuracy. With three times only it might correct Struve's value of aberration.

In 1873, Cornu, another French physicist, repeated the experiments of Fizeau with a toothed wheel, the work extending over three years. The observer was stationed at the École Polytechnique. The reflecting mirror and collimating telescope were placed on Mont Valerian, at a distance of about 33,816 feet. Three different wheels were tried, having 104, 116, and 140 teeth respectively, and rotating between seven and eight hundred times a second, the velocity being registered by electricity. Cornu used at times all the eclipses from the first to the seventh order. Calcium and petroleum light were tried, as well as sunlight. A chronograph with three pens recorded automatically seconds, the rota. tions of the toothed wheel, and the time of the eclipse. More than a thousand experiments were made, six hundred of which were reduced. The velocity of light as published by Cornu in 1873, was 185,425.6 miles per second. The probable error was 1 per cent. In 1874, Cornu gave the result of a new set of experiments made by him in conjunction with Fizeau over a distance of more than 14 miles between the Observatory

and Montlhéry. The experiments were repeated more than five nundred times, mostly at night with the lime light. The light was sent through a 12 inch telescope and returned through a 7-inch telescope. The toothed wheel which produced the eclipse was capable of rotating sixteen hundred times a second. From these experiments the velocity of light was placed at 186,618 miles. The probable error did not exceed 187 miles. The time was recorded accurately within a thousandth of a second.

I come now to that which most interests us to-night, viz, the part taken in this country for the measurement of these great velocities. About 1854, Dr. Bache, chief of the U. S. Coast Survey, appropriated $1,000 for the construction of apparatus to be used in repeating Wheatstone's experiment on the velocity of electricity. But those who were expected to take part in the investigation were called to other duties, and the money was never drawn.

In 1867, Professor Newcomb recommended a repetition of Foucault's experiment, in the interest of astronomy, to confirm or correct the received value of the solar parallax. In August, 1879, Mr. Albert A Michelson, then a master in the United States Navy, presented a paper to the meeting of the American Association for the Advancement of Science, on the measurement of the velocity of light. This paper attracted great attention. Mr. Michelson adopted Foucault's method with important modifications. In Foucault's experiment the deflection of the light produced by the revolving mirror was too small for the most accurate measurement. Mr. Michelson placed the revolving mirror 500 feet from the slit (which was ten times the distance in Foucault's experiment) and obtained a deflection twenty times as great, although the mir ror made only one hundred and twenty-eight turns in a second. With apparatus comparatively crude, he obtained for the velocity of light 186,500, with a probable error of 300 miles. This preliminary experi ment, made in the laboratory of the Naval Academy in May, 1878, indicated the directions in which improvements must be made in order to insure greater accuracy. The distance from the slit to the revolving mirror must be increased, the mirror must revolve at least two hundred and fifty times a second, and the lens for economizing the light must have a large surface and a focal length of about 150 feet. With the aid of $2,000 from a private source Mr. Michelson was able to carry out his ideas on a liberal scale.

His new experiments were made in the summer of 1879. The revolv ing mirror, made by Alvan Clark & Sons, was moved by a turbine wheel. Its rapidity of revolution was measured by optical comparison with an electric fork which made about one hundred and twenty-eight vibrations a second, the precise value being accurately measured by reference to one of König's standard forks. The velocity generally given to the mirror was about two hundred and fifty-six turns a second. The distance between the revolving and the fixed mirror was 1,986.26 feet.