103. AFTER having acquired a knowledge of the true significance of the index of refraction as the relation of velocity of propagation in the first to that in the second medium, it is easy to express the facts of the dispersion of colour in the language of the undulatory theory. To say that waves of different colour undergo an unequal amount of refraction is equivalent to stating that in a colour-dispersing medium the various homogeneous kinds of light are propagated with different velocities.

The proposition, that all kinds of light are propagated with equal rapidity, which we were formerly compelled to admit in regard to the free æther of the universe, is thus no longer admissible for the æther contained in the interior and occupying the interstices of the particles of natural substances.

The action which the particles of a body exert upon the undulations of æther propagating themselves in it, may be conceived to be dependent on the nature of these particles. In very many fluids and solids, especially in the colourless and transparent ones, as water, glass, &c., the rays produced by more rapid undulations are more strongly deflected, that is to say, experience

a greater amount of retardation than the rays produced by a smaller number of undulations. Prisms composed of such substances exhibit a spectrum with the ordinary succession of colours, from the least refrangible red to the most refrangible violet rays. The specific nature of the substance is however rendered evident even here by the different arrangement of the lines of Fraunhofer. (See fig. 106.)

The dispersion of colour in atmospheric air and in gaseous bodies generally is (according to Ketteler) so insignificant that we may admit in them, as in free æther, equal velocity for all kinds of light, being smaller than that of universal space in the proportion of 1: 1.000294. This number expresses the index of refraction of a ray of light in its passage from empty space, that is to say, space filled with free æther alone, into air at a temperature of 0° C. and under 760 millimeters pressure.

The influence of the nature of the material particles on the velocity of propagation is remarkably

exhibited in coloured substances, especially in those in whose absorption spectra one or more very dark lines appear. If we introduce, for example, an alcoholic

solution of the anilin colour Fuchsin' into a hollow prism (fig. 53) and look through it at a brightly illuminated slit, we obtain a spectrum in which blue and violet are less deflected than yellow and red. What is elsewhere the end of the spectrum here appears at the commencement, towards the middle it fades, and in the centre the green, being absorbed, is absent (fig. 140). From this behaviour the conclusion may be drawn that in Fuchsin the blue and violet rays are propagated with greater velocity than the red and yellow.

This phenomenon, which was discovered by Christiansen, and was shown by Kundt to be presented by a great number of absorbing substances, has been called anomalous dispersion of light.

104. The phenomena of anomalous dispersion renders us strongly disposed to the opinion that neither the refrangibility nor the length of undulation, but the number of vibrations, is to be regarded as the characteristic of a homogeneous ray of light. The number of undulations by which the impression of colour perceived by our eyes is conditioned does not undergo any alteration in the passage of light from one medium into another. In fact we observe no change of tint (Tonhöhe) when, for example, the yellow light of Sodium passes from air into water.

The length of the waves, however, does undergo a change. The wave-length is, it is to be remembered, always obtained by dividing the velocity of propagation by the number of vibrations. As the latter remains unchanged, whilst the velocity of propagation in water is only three-fourths of the velocity in air, the wavelength in water can only amount to three-fourths of

the wave-length in air. The wave-length of a ray of light in any given substance is consequently obtained by dividing the wave-length in air by the index of refraction of the substance itself.

105. We possess no means of changing the number of vibrations, that is to say, the colour of a homogeneous ray of light. But that such an alteration may and does occur under certain circumstances may now be demonstrated.

The sensation of a definite colour is conditioned by the number of waves of æther that penetrate into the eye in a second, just as the pitch of a musical note depends on the number of waves of sound which enter the ear in the same space of time. As long ago as 1841 Doppler called attention to the fact that the pitch of a musical note or the colour of an impression of light must be raised or lowered when the resounding or luminous body approximates or recedes from the observer. In the former case the organ of sense is struck in the course of a second by a greater, in the latter case by a less, number of waves than if the source of light or sound be stationary. As regards sound the truth of the principle of Doppler can easily be demonstrated by experiment; it is only necessary to allude to what may perhaps have been noticed by many. During the passage of a train through a station it may be observed that the whistle of a locomotive becomes higher in pitch as it approximates to, and lower in pitch as it recedes from the observer than when it is at rest. It is impossible, no doubt, to make a similar experiment in the case of light, because the greatest velocity we can attain is vanishingly small in comparison with its enormous speed. Nevertheless the possibility of its

occurrence in the case of the waves of light cannot be doubted.

Let it be conceived that in free space a sphere of glowing Sodium vapour is moving with sufficient velocity towards our earth, its light would appear more green than that of a terrestrial Sodium flame, whilst if it were receding it would assume a reddish tint. And if this light fell upon a prism instead of our eye it would reach the prism in the former case with a greater and in the latter case with a smaller number of undulations than that of a Sodium flame at rest, and in correspondence with this would experience a stronger or weaker deflection. Hence it follows that if a spectroscope be directed towards the moving source of light the bright Sodium line would appear to have changed its position and to be advanced towards the more, or towards the less refrangible end of the spectrum, according as the source of light was approximated to, or made to recede from the observer.

Just as the bright Sodium line in this example undergoes a change of position, so also, when the fixed star moves in the direction of the visual line with sufficient velocity, do the dark lines in the spectrum of a fixed star become altered and no longer coincide with the bright lines of the elementary substances to the absorbing action of which they owe their origin. From the direction and amount of this dislocation both the direction and the velocity of the movement of the star can be deduced.

106. Huggins, on comparing the F line of the spectrum of Sirius with the blue-green line of the spectrum of a Geissler's tube filled with hydrogen, found the former as compared with the latter moved