STAR. 503 STAR. nitucle appear, not with small disks as all the planets do. but as liiininous points without any visible diameter, and always the smaller the better the telescope. We are, therefore, totally ignorant of the real size of the fixed stars; nor could it be determined thouj:;h we were sure of their distances, for the apparent diameter is an essential element in the oalonlatinn. We cannot, then, say whether the greater brilliancy of one star, when coni]iared with another, arises from its greater nearness, its greater size, or the greater intensity of its light. But it is certain tliat the fixed stars are self-luminous. The num- ber of the stars is beyond determination. Those visible to the unaided eye in the Northern Hemi- spliere amount only to about 3000. Stars of the first magnitude visible north of 35° south decli- nation are reckoned bv Heis at 14, of the second at 4S, of the third at" 152. of the fourth at 313, of the fifth at 854, and of the sixth at 2010. But in the following classes the numbers increase rapidly, so that it is impossible to say how many there are. Recent photographic observa- tions seem to indicate that if the exposure of the sensitive plate in the telescope was continued long enough the entire plate would be light- struck. If this be so we must conclude that the entire heavens would be found covered with stars if we were able to see the faintest ones in ex- istence. That the fixed stars are not really immovable, as their name would imply, is seen in the phe- nomenon of doiihle or multiple stars, which are systems of two or more stars that revolve about one another, or rather about their common cen- tre of gravity. As they can be seen separate only by means of a telescope, and in most eases require a very powerful one. their discovery was possible only after the telescope was invented. Galileo himself discovered their existence and proposed to make use of them in determining the yearl.y jjarallax of the fixed stars. After a long lapse of time, Bradley, JIaskelyne. and Mayer again directed attention to the phenomena of double stars, but nothing important was made out respecting them until the elder Herschel made them the subject of a protracted series of obser- vations, which led to the most remarkable con- clusions as to their nature. At jiresent over 10.000 stars are known to be double: and in many cases more than one companion accom- panies the principal star. The apparent angular distance between two stars must be less than about thirty seconds in order that they may be counted a double. The theory of probabilities renders it almost certain that this vast number of double stars could not exist unless there were some real physical connection between the component nmn- bers of a double. Still, there is always a possi- bility of duplicity being merely apparent, the two stars lying nearly in the same direction in space, but at widely different distances from the earth. Such apparent doubles are called optical doubles, and those in which a real physical con- nection has been demonstrated by observed grav- itational rotation of the component members are called binai-y stars. Of these only about 250 are known. In some cases one of the components of a double star is much larger than the other, as in the star Rigel. in Orion, and in the polar star ; but very often the connected stars are nearly equal in luminous power. The two mem- bers of double stars are mostly of one color when the two components arc nearly equal ; but a difference of color is often observed when the com])onents are widely different in size. In many of these cases the one color is the complement of the other. It was Sir W. Herschel who first advanced the view, which has been confirmed since, that double stars are connected systems of two or more stellar bodies revolving in regular orbits around one another, or rather round their common centre of gravity. Their motions are found to follow the same laws as prevail in the solar system, and the orliits are elliptical. These distant bodies are therefore subject to the New- tonian law of gravitation. The period of revolu- tion has, in many cases, been computed ; the shortest is estimated at 5% years; others are set down at hundreds. In cases where the paral- lax is known the size of the orbits can be deter- mined, and thus the astronomer is able to assert in regard to the double star a Centauri that the orbit described bj' the two components about each other is 24 times as large as that of our earth around the sun. Even the masses of these stars have been calculated as being together 2, that of our sun being 1. It is a consequence of these revolutions that some stars are now seen double that formerly seemed single, and vice versa. If the plane of revolution has its edge presented to the earth the stars will seem to move in a straight line, and at times to cover one another. In addition to these double stars that can be seen to revolve and change their relative positions there are others of which the components are so close together that even our most powerful telescopes fail to separate them. These doubles are known to exist from peculi- arities in their spectra. The}' show a doubling of the spectral lines that must be caused by a duplicity in the source of light, and where this doubling is shown to occur periodical!}' we must conclude that the two sources of light are re- volving in some way or other. The proper motion of stars is of another kind. It consists in a displacement in various direc- tions of the individual stars, so that the configu- ration of constellations is slowly changing. The annual proper motions yet observed vary from nothing to 8.7". The proper motion of the binary star 61 Cygni amounts to 5.2", so that in 360 years it would pass over a space equal to the moon's diameter. It must thtis take thousands of years to alter sensibly the aspect of the heav- ens; although, taking into account the enormous distances, the actual velocities must be great. But the observed proper motions of the stars do not give us very accurate information as to their real motions and velocities. In the first place, it is only the angular change of the star's position that we observe; and we cannot tell the corre- sponding linear shift unless we know the distance of the star from the earth. And even where this distance (parallax) is known we can obtain only the transverse motion, as projected on the sky. There may be also a component of motion directly toward us or away from us. This radial component remained entirely unknown until recently, when it became possible to meas- ure it with the spectroscope. It was first observed by Sir W. Herschel that there is a perceptible tendency in the proper motions, as observed, to make the stars generally diverge or open up in one quarter of the heavens, and draw together in the opposite quarter; and
