mediums. I have, at the end of Chapter iii., given the reasons which prevent my assenting to the assertion of a special analysis of light by absorption. In the same manner, with regard to other effects produced by media upon light, it is sufficient for the defence of the theory that it should be consistent with the possibility of the laws of phenomena which are observed, not that it should explain those laws; for they belong, apparently, to another province of philosophy.

Some of the optical properties of bodies which have recently attracted notice appear to be of this kind. It was noticed by Sir John Herschel,1 that a certain liquid, sulphate of quinine, which is under common circumstances colorless, exhibits in certain aspects and under certain incidences of light, a beautiful celestial blue color. It appeared that this color proceeded from the surface on which the light first fell; and color thus produced Sir J. Herschel called epipolic colors, and spoke of the light as epipolized. Sir David Brewster had previously noted effects of color in transparent bodies which he ascribed to internal dispersion: and he conceived that the colors observed by Sir J. Herschel were of the same class. Professor Stokes of Cambridge applied himself to the examination of these phenomena, and was led to the conviction that they arise from a power which certain bodies possess, of changing the color, and with it, the refrangibility of the rays of light which fall upon them: and he traced this property in various substances, into various remarkable consequences. As this change of refrangibility always makes the rays less refrangible, it was proposed to call it a degradation of the light; or again, dependent emission, because the light is emitted in the manner of self-luminous bodies, but only in dependence upon the active rays, and so long as the body is under their influence. In this respect it differs from phosphorescence, in which light is emitted without such dependence. The phenomenon occurs in a conspicuous and beautiful manner in certain kinds of fluor spar: and the term fluorescence, suggested by Professor Stokes, has the advantage of inserting no hypothesis, and will probably be found the most generally acceptable.*

It may be remarked that Professor Stokes rejects altogether the doctrine that light of definite refrangibility may still be compound, and may be analysed by absorption. He says, "I have not overlooked the remarkable effect of absorbing media in causing apparent changes

1 Phil. Trans. 1845.

3 Phil. Trans. 1852 and 1854.

2 Edinb. Trans. 1833.

* See Phil. Trans. 1852.

of color in a pure spectrum; but this I believe to be a subjective phenomenon depending upon contrast."

CHAPTER XIII.

UNDULATORY THEORY.

Direction of the Transverse Vibrations in Polarization.

IN the conclusion of Chapter xiii. I have stated that there is a point

in the undulatory theory which was regarded as left undecided by Young and Fresnel, and on which the two different opinions have been maintained by different mathematicians; namely, whether the vibrations of polarized light are perpendicular to the plane of polarization or in that plane. Professor Stokes of Cambridge has attempted to solve this question in a manner which is, theoretically, exceedingly ingenious, though it is difficult to make the requisite experiments in a decisive manner. The method may be briefly described.

If polarized light be diffracted (see Chap. xi. sect. 2), each ray will be bent from its position, but will still be polarized. The original ray and the diffracted ray, thus forming a broken line, may be supposed to be connected at the angle by a universal joint (called a Hooke's Joint), such that when the original ray turns about its axis, the diffracted ray also turns about its axis; as in the case of the long handle of a telescope and the screw which is turned by it. Now if the motion of the original ray round its axis be uniform, the motion of the diffracted ray round its axis is not uniform and hence if, in a series of cases, the planes of polarization of the original ray differ by equal angles, in the diffracted ray the planes of polarization will differ by unequal angles. Then if vibrations be perpendicular to the plane of polarization, the planes of polarization in the diffracted rays will be crowded together in the neighborhood of the plane in which the diffraction takes place, and will be more rarely distributed in the neighborhood of the plane perpendicular to this, in which is the diffracting thread or groove.

:

On making the experiment, Prof. Stokes conceived that he found, in his experiments, such a crowding of the planes of diffracted polarization towards the plane of diffraction; and thus he held that the

hypothesis that the transverse vibrations which constitute polarization are perpendicularly transverse to the plane of polarization was confirmed.

1

2

But Mr. Holtzmann, who, assenting to the reasoning, has made the experiment in a somewhat different manner, has obtained an opposite result; so that the point may be regarded as still doubtful.

Final Disproof of the Emission Theory.

As I have stated in the History, we cannot properly say that there ever was an Emission Theory of Light which was the rival of the Undulatory Theory: for while the undulatory theory provided explanations of new classes of phenomena as fast as they arose, and exhibited a consilience of theories in these explanations, the hypothesis of emitted particles required new machinery for every new set of facts, and soon ceased to be capable even of expressing the facts. The simple cases of the ordinary reflexion and refraction of light were explained by Newton on the supposition that the transmission of light is the motion of particles: and though his explanation includes a somewhat harsh assumption (that a refracting surface exercises an attractive force through a fixed finite space), the authority of his great name gave it a sort of permanent notoriety, and made it to be regarded as a standard point of comparison between a supposed "Emission Theory" and the undulation theory. And the way in which the theories were to be tested in this case was obvious: in the Newtonian theory, the velocity of light is increased by the refracting medium; in the undulatory theory, it is diminished. On the former hypothesis the velocity of light in air and in water is as 3 to 4; in the latter, as 4 to 3.

But the immense velocity of light made it appear impossible to measure it, within the limits of any finite space which we can occupy with refracting matter. The velocity of light is known from astronomical phenomena;-from the eclipses of Jupiter's satellites, by which it appears that light occupies 8 minutes in coming from the sun to the earth; and from the aberration of light, by which its velocity is shown to be 10,000 times the velocity of the earth in its orbit. Is it, then, possible to make apparent so small a difference as that between its passing through a few yards of air and of water?

Mr. Wheatstone, in 1831, invented a machine by which this could

Camb. Trans., vol. ix. part i. 1849.

2 Phil. Mag., Feb. 1857.

be done. His object was to determine the velocity of the electric shock. His apparatus consisted in a small mirror, turning with great velocity about an axis which is in its own plane, like a coin spinning on its edge. The velocity of spinning may be made so great, that an object reflected shall change its place perceptibly after an almost inconceivably small fraction of a second. The application of this contrivance to measure the velocity of light, was, at the suggestion of Arago, who had seen the times of the rival theories of light, undertaken by younger men at Paris, his eyesight not allowing him to prosecute such a task himself. It was necessary that the mirrors should turn more than 1000 times in a second, in order that the two images, produced, one by light coming through air, and the other by light coming through an equal length of water, should have places perceptibly different. The mechanical difficulties of the experiment consisted in keeping up this great velocity by the machinery without destroying the machinery, and in transmitting the light without too much enfeebling it. These difficulties were overcome in 1850, by M. Fizeau and M. Léon Foucault separately and the result was, that the velocity of light was found to be less in water than in air. And thus the Newtonian explanation of refraction, the last remnant of the Emission Theory, was proved to be false.

BOOK X.

THERMOTICS.-ATMOLOGY.

CHAPTER III.

THE RELATION OF VAPOUR AND AIR.

THE

Sect. 4.-Force of Steam.

HE experiments on the elastic force of steam made by the French Academy are fitted in an especial manner to decide the question between rival formulæ, in consequence of the great amount of force to which they extend; namely, 60 feet of mercury, or 24 atmospheres: for formula which give results almost indistinguishable in the lower part of the scale diverge widely at those elevated points. Mr. Waterston' has reduced both these and other experiments to a rule in the following manner :-He takes the zero of gaseous tension, determined by other experimenters (Rudberg, Magnus, and Regnault,) to be 461° below the zero of Fahrenheit, or 274° below the zero of the centigrade scale: and temperatures reckoned from this zero he calls "G temperatures." The square root of the G temperatures is the element to which the elastic force is referred (for certain theoretical reasons), and it is found that the density of steam is as the sixth power of this element. The agreement of this rule with the special results is strikingly close. A like rule was found by him to apply generally to many other gases in contact with their liquids.

But M. Regnault has recently investigated the subject in the most complete and ample manner, and has obtained results somewhat different.2 He is led to the conclusion that no formula proceeding by

Phil. Trans. 1852.

2 Mém. de l'Institut, vol. xxi. (1847). M. Regnault's Memoir occupies 767 pages.