· aperture in a screen so as to cross one of these intervals and fall upon a reflector placed at a known and considerable distance from the wheel. The reflector was so arranged as to throw back the beam through the notch in the wheel exactly opposite to that through which it first passed. Through this notch could be perceived the reflected ray, which had traversed a distance double that of the reflector from the wheel. When, now, the wheel was revolved with increasing velocity, the reflection at first seen continuously, gradually became feebler and presently entirely disappeared. This occurred when the velocity of the wheel was such that, the light transmitted through the notch on one side was intercepted by the tooth adjacent to the opposite aperture on its return; that is, when the velocity of rotation carried a tooth over its own breadth whilst the ray was going and returning. This velocity is readily measured; indeed, it may be registered by the mechanism used to drive the wheel. If the wheel, as was actually the case, makes twelve and six-tenths revolutions in a second and has fourteen hundred and forty divisions (teeth and notches), the time of the passage of a tooth across its own breadth is found by taking the reciprocal of the product of these numbers. In this fraction of a second the ray has traversed twice the distance between the mirror and the wheel,— which amounted in M. Fizeau's experiments to eighteen thousand, eighteen hundred and eighty yards. But this total distance measured or divided by the time, will give the distance gone over in a second, or, in other words, the velocity of the ray. The mean results of the experiments established a velocity of one hundred and ninety six thousand miles.

The far more refined and delicate method of M. Foucault has shown, however, that this result is too great. This method is essentially that first used by Arago, in an experiment determining the relative velocities of light in air and water. A horizontal ray of light is admitted into a darkened chamber, and falls upon a mirror arranged to revolve on a vertical axis lying in its own plane. As the mirror turns, the reflected ray will move, of course, in a horizontal plane passing through the point of incidence and the aperture of admission, and by an easy geometrical consideration its angular velocity is known to be double that

of the mirror. In this horizontal plane, a second mirror is placed perpendicular to a line itself drawn perpendicular from the centre of the last to the axis of the first;-placed, in other words, so as to return a ray to the first mirror upon the same path in which it is first reflected from it. If, now, the revolving mirror be supposed at rest and be so arranged as to reflect the ray (received through the aperture) upon the second mirror, the ray will manifestly be returned by the latter upon the same path, and will be again reflected by the first directly towards the aperture. But if, whilst the ray has been passing between the mirrors, the first has revolved through a small angle, the ray in passing back towards the aperture will deviate from its original path by an angle double that described by the mirror. This angle is readily measured; and the fraction of a second required by the light to traverse the distance between the mirrors, to and fro, multiplied by the angle described by the mirror in any small fraction of a second taken as a unit, will give a product equal to one-half this measured angle. But the number of rotations in a second being registered, the unit angle may be readily deduced from the unit of time. The product and one factor of it being thus known, we derive the other factor, or the time of the ray; this, with the space passed over-the double distance between the mirrors-affords one factor of another known product; so that finally dividing the space by the time we obtain the velocity.

We may here mention, in passing, that by the interposition of a column of water between the mirrors, through which the ray is passed, we have the means of ascertaining the velocity of the propagation of light through water, and that this is found to be less than its velocity in air.

The result of M. Foucault's experiments was a velocity of 185,172 miles; so that taking all the methods into consideration, neglecting only M. Fizeau's as subject to important errors from mechanical imperfections, we may conclude that the velocity of light in interplanetary and cosmical space is about one hundred and eighty-six thousand miles in a second!

This enormous velocity takes hold upon the infinite, and is beyond any adequate comprehension. The greatest speed we

can generate in a body moving through the air or over the surface of the earth, is a mere bagatelle to a velocity which will belt the globe in the eighth part of a second. Yet it is some consolation to know that we can always halt one immensity, however overbearing, with the qui va là of another. Space, in fact, is infinite, and we can, by bringing it face to face with this vast velocity, not only reduce the scale of numbers upon which the latter plumes itself, but even turn the tables upon it and obtain tremendous results in the opposite direction, so as to leave. the impression that, after all, light is really slow! To achieve this desirable result, we count off space in units of this velocity. We thus find that light will reach us from the moon in about one second and a quarter; from the sun, in eight minutes and thirteen seconds; from the fixed star, Alpha Centauri, in about three years; from 61 Cygni, in nine years; from Alpha Lyræ, in twelve years; whilst from the remotest nebulæ, as surmised by Sir Wm. Herschel, it will require not less than two million years! These facts develop a singular field of contemplation. When we view, in an unclouded night, the starry dome, we are really looking upon a historical chart reaching back into the far-distant ages of the past. The rays that reach our eyes started upon their journey, some an hour, some a day, some years, some centuries, ago from the stellar bodies in our field of vision. This light may be partially that reflected by these bodies from rays which left the earth just twice those periods past. So that the whole history of our earth may now be illustrated onthe vault of heaven.

The two leading theories of light are known as the corpuscular, emission, or Newtonian, and the wave or undulatory theory. The first seems to be attributed to Newton on insufficient grounds. He, indeed, advanced it as a means of comprehending certain phenomena of light, but he explicitly says: Tis true that from my theory I argue the corporeity of light; but I do it without any absolute positiveness as the word "perhaps" intimates; and make it at most but a very plausible consequence of the doctrine, and not a fundamental supposition, nor so much as any part of it.' Newton, indeed, whilst urging objections against 1 Phil. Trans. Vol. x. p. 5086: quoted by Prof. Baden Powell.

1

the undulatory theory, still held a particular hypothesis of undulations, consentaneous with corpuscular emission, as possible. In fact, however, he adopted positively no hypothesis. Were I', says he, to assume an hypothesis, it should be this, if propounded more generally, so as not to determine what light is, further than that it is something or other capable of exciting vibrations in the ether; for thus it will become so general and.comprehensive of other hypotheses as to leave little, room for new ones to be invented.' 2

The corpuscular theory assumes that a luminous body projects. or emits, in all directions, extremely minute particles, which falling upon the retina of the eye produce the sensation of light, just as minute particles of any perfume excite sensation in the organs of smell. A leading objection to this view arises from the extreme velocity under which these particles must move. 'If each luminous molecule', says l'Abbè Moigno, 'should weigh a grain, its momentum, endowed as it is with so excessive a velocity, would equal that of a ball of seventy-five kilogrammes (165 pounds avoirdupois) traversing more than one thousand feet a second. The weight of the luminous molecule must be, in reality, some millions of times less than we have supposed it; but as, on the other hand, we can make efficient at the same moment several millions of these molecules collected in the focus of a lens, the mechanical effect produced by the sum of their momenta ought to be rendered sensible, which result it has been impossible to obtain under the most favorable circumstances.' (Vol. i: p. 71.) This negative evidence is, however, not conclusive. It is a part of this theory, in its explanation of the passage of light from air into water or glass,- that the luminous molecules which escape reflection from the surface of the second medium, advancing still more closely to the particles of water or glass, reach the sphere of their attraction, and then enter the substance. This entrance is made with the original velocity increased by the powerful attraction of the particles of water or glass. The undulatory theory, on the contrary, in its explanation of the same phenomenon, holds that light traverses the new medium, if denser than that from which it is received,

2 Pbil. Trans. x. 5089.

A

under a diminished velocity. Here is then, as between these theories, the crucial test. First applied by Arago, the result has already been mentioned as obtained by Foucault's apparatus. A diminished velocity in the denser medium is established by experiment; and the corpuscular theory, without essential modification, 'must be abandoned.

The wave theory of light assumes the existence of a fluid of great tenuity pervading cosmical space, which is called the ether. The luminous body is supposed to put the ether into vibration, as a sonorous body excites vibrations in the air. The luminiferous waves travel forward, as waves raised upon a surface of water, with the velocity we have already established. There is no transmission of the ethereal particles; each, after suffering the vibratory motions necessary to carry it through all the phases of the wave, subsides to its former position. Light is, therefore, upon this theory, a property, a mere vibration of an assumed highly elastic fluid having a fixed relation to our organs of sight, just as sound is a vibration of sensible matter communicated to the auditory organs.

The vibration of the air caused by a sonorous body proceeds. by alternate compression and expansion along the lines of communication; in other words, it is longitudinal, like that propagated by expansion and contraction along the length of an elastic cord. Waves excited upon the surface of water are, on the other hand, transversal, the vibration of each particle of water being perpendicular to the rectilinear advance of any face of the "wave. Now, certain phenomena of light plainly show that the vibrations of the ethereal molecules must be regarded as of the latter class, with, however, this remarkable difference, that there must be simultaneous vibration of particles of the luminiferous ether in all directions at right angles to the lines of progression; that is, alternate expansion and contraction transverse to the line of the ray. Each particle may be considered as vibrating under two forces acting in lines perpendicular to each other. If, then, the force along one line is separated in its action by an interval from that along the other, curvilinear motion will result; circular, if the interval be just one-fourth of a vibration; elliptical, if it be any other amouut. These results, theoretically