The principle of this inquiry is as follows: If a source of light is approaching us very rapidly, then the waves of light coming from it necessarily appear a little shorter than they really are, or, in other words, that light appears to be slightly more blue--the blue waves being shorter than the red--than it really is. A similar thing with regard to the waves of sound is often noticed in connection with a railway train. If an express train, the whistle of which is blowing the whole time, dashes past us at full speed, there is a perceptible drop in the note of the whistle after it has gone by. The sound waves as it was coming were a little shortened, and the whistle therefore appeared to have a sharper note than it had in reality. And in the same way, when it had gone by, the sound waves were a little lengthened, making the note of the whistle appear a very little flatter. Such a change of colour in a star could never have been detected without the spectroscope; but since when light passes through a prism the shorter waves are refracted more strongly, that is to say, are more turned out of their course than the longer, the spectroscope affords us the means of detecting and measuring this change. Let us suppose that the lines of hydrogen are recognized in a given star. If we compare the spectrum of this star with the spectrum of a tube containing hydrogen and through which the electric spark is passing, we shall be able to see whether any particular hydrogen line occupies the same place as shown by the two spectra. If the line from the star is a little to the red of the line from the tube, the star must be receding from us; if to the blue, approaching us. The amount of displacement may be measured by a delicate micrometer, and the rate of motion concluded from it. [Illustration: THE HALF-PRISM SPECTROSCOPE ON THE SOUTH-EAST EQUATORIAL.]