Huggins and Miller suggest, the worlds revolving round this star are also deficient in this element, they would be without water, like our moon.
Upon a very clear night it may be noticed that the stars are not all of the same colour, but that many of them appear to be of a ruddy or yellowish tint. The cause of this is plainly seen when they are submitted to spectral analysis. Thus, Sirius, which is a brilliant white star, shows but three dark lines, while one of the stars in the constellation of Hercules shows several groups of bands in the red, blue, and green portions of its spectrum, fully accounting for its orange tint.
The double star Cygni is a very beautiful example of the distribution of colour between two members of a stellar group. One star shows a strong spectrum with the blue and violet portions almost totally blotted out, while its companion is similarly circumstanced with respect to the yellow and orange portions of its spectrum. The colour of one is consequently orange, while the other is of a delicate blue. If these stars are the principal members of a system, the alternation of blue and orange days must be indeed a singular phenomenon to those who inhabit their planets.
In some of the stars lines have been discovered which do not possess any equivalent amongst those produced by terrestrial matter; they consequently contain elements of which we know nothing; at the same time, however, it has been found that terrestrial elements exist in some of the remote nebulæ, which are so distant that their light takes many thousands of years to reach our earth.
Spectrum analysis has decided the grand question of the physical composition of the nebulæ. Those bodies were supposed, with some reason, to be aggregations of stars, like our Milky Way, which only required telescopes of sufficient power to resolve them. That they
