One of the most interesting of these is Sirius, from the irregularity of whose movements Peters and Auwers had con cluded that it was accompanied by an attendant star or at least a large planet; and in 1862 such a companion was dis covered by the younger Clark in the very direction in which theory had predicted it to be at that time. Its great minute ness renders detection hopeless except in very favorable circumstances. The mo tions of Procyon indicate a similar doubleness, but its companion has not yet been seen. Triple and multiple stars forming one system are also not un common. For instance, u Herculis, re cognized as a double star by Herschel, is really a triple star, the small Herschel lian companion having been resolved into two in 1856 by Alvan Clark. An inter esting optical property of these binary systems is that the color of the one com ponent is frequently the complement of that of the other. A very remarkable double star is 61 Cygni, famous as the first star whose distance from the sun was calculated directly from its parallax, i. e., from the angular distance between its positions as viewed from opposite ex tremities of the earth's orbit. The most probable mean of the various results gives about .5 of a second for this paral lax.
This, however, is not the nearest star; it is nearly twice as far away as a Cen tauri, which according to Henderson has a parallax of 0".91, corresponding to a distance of 226,000 times the radius of the earth's orbit, or more than twenty millions of millions of miles (20 X 10"). In other words, light, which travels at the rate of 236,000 miles per second, takes nearly three years to pass from the nearest fixed star to us. The paral lax of at least half the stars of the first magnitude is probably less than one-tenth of a second, so that their average dis tance is greater than two million radii of the earth's orbit. Though, to the naked eye, stars have probably occupied the same relative positions from the earliest of historic times, the telescope reveals that they are not strictly fixed. Many have appreciable proper motions, among others 61 Cygni, as was before noticed.
Its annual motion is 5" .2, which is ex
ceeded, however, by that of Groombridge, 1830, a star of the 7th magnitude, whose proper motion is 7". As a general rule, the brightest stars have the great est proper motion, and stars below the 3d magnitude move only a few seconds in 100 years.
This motion in the heavens cannot be taken even as an approximate estimate of the velocity with which any star is moving, since only the portion of the real velocity which is resolved perpendicular to the line of sight is measurable. Spec trum analysis has, however, given us a means of measuring the velocity along the line of sight. The physical meaning of the dark absorption lines which cross the spectra of the sun and stars has been pointed out under SPECTRUM ANALYSIS. They indicate the absence of certain rays of definite refrangibility and wave length from the solar or stellar light which reaches us. Since each star spectrum has its own peculiar lines, these must be due to absorption by the star's own outer envelope or atmosphere. Now, exactly as ocean waves will appear shorter or longer than they really are according as the observer is sailing against or with them, so will the wave length of each in dividual ray be shortened or lengthened according as we approach or recede from the source of light, and each absorption line in the spectrum, corresponding as it does to a ray of definite wave length, will be displaced toward the violet or red respectively. By comparison with the spectrum of the substance to which a determined portion of these lines cor responds, this displacement may be measured, and from its relation to the particular wave length the ratio of the relative velocity of approach or recession of the star to the velocity of light can be easily calculated. Stars which show periodic variations of magnitude are called variable stars. Among these are Mira and Algal. Up to 1920 about 750 of these stars were known. Their average period of variation is about 300 days and some are 1,000 times brighter at maximum than at minimum. It is be lieved that these are the giants among the stars. At their maximum some are 150 times brighter than the sun.