It is not pretended by the advocates of the rival hypothesis, the undulatory theory of light, that they understand exactly the nature of the transference of energy on which they suppose light to depend; but they take from the analogy of sound in air, and of waves in water, the idea of the existence iu all space of a highly elastic fluid (or quasi solid), provisionally named the ether (q.v.), and they suppose light to consist in the prop agation of waves in this fluid. Huygens has the credit of having propounded, and ably develbped and illustrated, this theory.
As we have seen above, no third hypothesis as to the nature of light is admissible. Many strong arguments against the truth of the corpuscular theory had been furnished by experiment, especially in the early pact of the present century; and as they were always met by further and more extraordinary properties which had to be attributed to the luminous corpuscles, the theory bad become complicated in the most fearful man ner; and this of itself was an almost complete disproof. Still, it held its ground, for Newton's old objection to the rival theory, viz., that on the undulatory hypothesis there should be no shadows at all (witness the analogy of sounds heard round a corner), was as yet unanswered. This difficulty was overcome by Young (q.v.), to whose sagacity we are indebted for the idea of interference (q.v.), which completely explained the appar ent discrepancy. But the question between the rival theories was finally by Fizeau and Foucault, who, by processes entirely different, but agreeing in their results, determined the velocity of light in air and in water.
Now, Newton had shown that refraction, such as that of light by water, if predicated of moving particles, requires that they should move faster in water than in air. Huy ghens, again, had shown, that if such refraction be predicated of waves, they must move slower in water than in air. Fizeau and Foucault found, by direct measurement, that light moves slower in water than in air. Hence it is certain that light consists in the transference of energy, not of matter; and the undulatory theory is based upon this fact.
But, as to the manner iu which energy is thus transferred, we are entirely ignorant. The common assumption is, that waves of distortion are propagated in the ether. The nature of this motion described under WAVE. But many other modes have been suggested, one of the most notable of Which is that of Rankine. Here the particles of ether are not supposed to he displaced, but each is merely made to turn about an axis as the wave of light passes it: the particles having polarity (q.v.), by virtue of which they arrange theMselves in similar positions when no light is passing, and by which, also. any rotation of one particle produces a consequent rotation of those in its neighborhood. For the explanation of most of the common phenomena of optics, it is quite indifferent which of these assumptions we make, and, indeed, theory has not yet been carried far enough to enable us to devise experimental methods of testing which is the more likely to be the case in nature. It cannot be too strongly insisted on that all we know at pres ent is, that light certainly depends on the transference of energy from one part of the luminiferous medium to another; what kind of energy is transferred, vibratory or oscil latory motion, or rotation, etc., is a problem which may possibly forever remain unsolved. But vibratory wave-motion being that with which we are most familiar, as in earthquakes, sound, waves in water, etc., we naturally choose this as the most easily intelligible basis of explanation and illustration. And we shall now briefly show how
the laws of linear propagation, reflection, single refraction, interference, diffraction, dis persion, polarization, and double refraction may be accounted for.
'lsume, then, that light consists in a succession of waves, and for our earlier inquiries 2 does not matter whether they be (like those of sound) waves of condensation and rarefact..on, in which the vibrations take place in the direction of the ray, or (like those in water) waves of distortion or displacement without condensation, in which case the luminous vibrations must be assumed to take place in some direction pe7pendicular to the ray. The phenomena of polarization and double refraction show us that the former of these hypotheses is untenable.
Propagation of Light in a Utafortn Isotropic Medium. (An isotropic medium is such that if a cubical portion be taken, it possesses precisely the same properties whatever be the directions of its sides. Glass and water are isotropic, rock-salt and ice are not.)— Suppose AB (fig. 1) to represent at any time the front of a plane wave which is passing. in the direction CD; i.e., suppose all particles of the ether in. the plane AB (perpendicu lar to the plane of the paper) to be similarly and equally displaced. According to Huy gens, we must suppose every particle, P, to be itself the source of a wave, which. from the uniformity of the medium, will spread with the same velocity in all directions. With center P, and radius the space which light passes over in any assigned interval, t, describe a sphere represented in section by a circle in the figure. Do the same for adja cent points etc. Let be the intersection of the circles whose centers are P and that of the circles whose centers are and and so on. Then, as is equidistant from P and and (approximately) from all points of a small circular space between P and on the wave-front AB, all the separate wave-disturbances coining from these points to p1 will be in the same phase (see WavE), and will therefore combine so as to strengthen each other; while in other directions they will be in different phases, and combine to destroy each other. The locus of all such points as etc., will therefore, at the end of the time 1, contain all particles of the ether quality and similarly disturbed, and will thus be the new wave-front. But it is obviously a plane parallel to AB. Also the disturbance at P has passed to and, when the distance is taken as very small, is perpendicular to the wave-front AB. Hence, in such a medium, a plane wave remains plane, and moves with uniform velocity in a direction perpen dicular to its front. [There is a difficulty as to what becomes of the disturbance, which, according to Huygen's assumption, ought to travel back into the dotted portions of the spheres; and it is not easy to account for the absence of this on mechanical principles. But we are content here to take for granted that no waves are propagated backward from the main wave, as a fact clearly proved by experiment.] Since a small portion of the surface of any curved wave may be considered as plane, we now see now any such wave will be propagated in an isotropic medium. Erecting perpendiculars at every point of the surface of the curved wave, and laying off along these lines the space which light passes over in a given interval, the extremities form a new surface, which is the wave-front after the lapse of that interval.