Light consists of waves in the ether of different lengths and making upon the eye different impressions of colour according to the length of the waves. The longest waves are at the red end of the spectrum and the shortest at the blue, or violet, end. But since they all move onward with the same speed, it is clear that the short blue waves must fall in quicker succession on the retina of the eye than the long red waves. Now suppose that the source of light from which the waves come is approaching very swiftly; it is easy to see that all the waves will strike the eye with greater rapidity, and that the whole spectrum will be shifted toward the blue, or short-wave, end. The Fraunhofer lines will share in this shifting of position. Next suppose that the source of the light is retreating from the eye. The same effect will occur in a reversed sense, for now there will be a general shift toward the red end of the spectrum. A sufficiently clear illustration, by analogy, is furnished by the waves of sound. We know that low-pitched sounds are produced by long waves, and high-pitched ones by short waves; then if the source of the sound, such as a locomotive whistle, rapidly approaches the ear the waves are crowded together, or shifted as a whole toward the short end of the gamut, whereupon the sound rises to a shrill scream. If, on the contrary, the source of sound is retreating, the shift is in the other direction, and the sound drops to a lower pitch. This is precisely what happens in the spectrum of a star which is either approaching or receding from the eye. If it is approaching, the Fraunhofer lines are seen shifted out of their normal position toward the blue, and if it is receding they are shifted toward the red. The amount of shifting will depend upon the speed of the star's motion. If that motion is across the line of sight there will be no shifting, because then the source of light is neither approaching nor receding. Now take the case of a binary star whose components are too close to be separated by a telescope. If they happen to be revolving round their common centre in a plane nearly coinciding with the line of sight from the earth, one of them must be approaching the eye at the same time that the other is receding from it, and the consequence is that the spectral lines of the first will be shifted toward the blue, while those of the second are shifted toward the red. The colours of the two intermingled spectra blend into each other too gradually to enable this effect to be detected by their means, but the Fraunhofer lines are sharply defined, and in them the shift is clearly seen; and since there is a simultaneous shifting in opposite directions the lines appear split. But when the two stars are in that part of their orbit where their common motion is across the line of sight the lines close up again, because then there is no shift. This phenomenon is beautifully exhibited by one of the first