The heliostat is a mirror, mounted on an axis, which moves at the same rate as the sun appears to travel, so that wherever the sun is, the reflector, once adjusted, automatically throws the beam into the instrument, so that the light of the moving sun can be observed without moving the spectroscope. [Illustration: FIG. 189.—Comparison Prism fixed in the Slit. ] An average solar spectrum is thus obtained, and, by means of a prism over one-half of the slit, it was quite possible for Kirchhoff and Bunsen to throw in a spectrum from any other source for comparison, and so they compared the spectra of the metals and other elements with the solar spectrum, and tested every line they could find in the spectra. They first found that the two lines of sodium corresponded with the two lines called D in the spectrum, then that the 460 lines of iron corresponded in the main with dark lines in the solar spectrum; and so they went on. [Illustration: FIG. 190.—Foucault’s Heliostat. ] There is, however, a method of varying the attack on this body altogether, by means of the spectroscope and telescope. We saw that Kirchhoff and Bunsen contented themselves with an average spectrum of the sun—that is to say, they dealt with the general spectrum which they got from the general surface of the sun, or reflected from a cloud or any other portion of the sky to which they might direct the reflector; but by means of some such an arrangement as is shown in Fig. 192, we can arrange our spectroscope so that we shall be able to form _an image_ of the sun by the object-glass of a telescope, on the slit, and allow it to be immersed in any portion of the sun’s image we may choose. We then have a delicate means of testing what are the spectroscopic conditions of the spots and of those brighter portions of the sun which are called faculæ, and the like. And it is known that, by an arrangement of this kind, it is even possible to pick up, without an eclipse, those strange things which are called the red prominences, or the red flames, which have been seen from time to time during eclipses. If we wish to observe any of the other celestial bodies, we must employ a telescope and form an image on the slit, or else use the heavenly body itself as a slit. In the former case spectroscopes must be attached to telescopes, and hence again they must be light and small, unless a siderostat is employed. In the latter case the prism is placed outside the object-glass, and the true telescope becomes the observing telescope.