108 VARIABLE STARS. 



in point, being now distinctly faiuter than its neighbor, Beta, although 

 presumably when Bayer assigned the letters, about three hundred 

 years ago, the reverse was the case. A second <dass consists of those 

 wiiich oscillate in brightness, but without any api)arent regularity, 

 like Eta Argus, and Alpha Orionis. The third class is made up of the 

 so-called "temporary stars," stars which suddenly shine out and after 

 a few months disappear or become very faint. There have been about 

 a dozen instances of this iiind, the last being the "new star in Auriga,^'' 

 which appeared in 1892. The fourth class, and by far the most numer- 

 ous, is made up of those which vary periodically in brightness in a 

 manner which admits of mathematical prediction. Among these 

 periodic stars, at least three subclasses are to be distinguished: {a) 

 Those which at regular intervals brighten up for a comparatively short 

 time and then fall back to their normal condition until the next maxi- 

 mum. These are often referred to as the " Omicron Geti type," 

 because that star usually called Mira, or the "wonderful," is the most 

 conspicuous and longest known of all that belong to it. The majority 

 of the periodic stars belong to this sub-class, and geuerally repeat 

 their changes about once a year, though their periods range all the 

 way from one hundred and ten days to six hundred and ten. {b) The 

 second sub-class includes the stars which, like Beta Lyra', are in a 

 continual state of activity; they often present more than one maxi- 

 mum of brightness within a single complete period, which is usually 

 somewhere between five days and sixty, (c) Finally, we have the ten 

 stars of the so-called ^^ Algol type," which behave in a manner pre- 

 cisely the opposite of that of the stars of th.e first sub-class, i. e., they 

 undergo periodical diminutions or eclipses of their light, remaining 

 for the remainder of the time steadily bright. Their periods are 

 always short, ranging from seven hours to ten days. The type of the 

 class is Beta Perscl, better known as Algol. 



The variations of the stars of this class can be very simply explained 

 as merely due to actual eclipses, and the truth of this explanation has 

 been practically established within the last three years by the spectro- 

 scopic work of Vogel, at Potsdam, Tiie period of Algol is a trifle under 

 sixty-nine hours, and the photographs show that seventeen hours before 

 the obscuration the lines of its spectrum are displaced toward the red 

 by an amount which indicates that it is then receding from us at the 

 rate of about 27 miles a second, while seventeen hours after the obscu- 

 ration it is found to be approaching at the same rate; in short, the 

 observations pro\e that the star is moving in an orbit nearly circular, 

 a little more than 2,0(M>,000 miles in diameter, and lying in a plane 

 nearly i^assing through the observer. Now if the periodical obscura- 

 tion is really caused by the interposition of an eclii)siug body, the 

 bright star must necessarily circulate around the common center of 

 gravity between itself and its dusky attendant in just this sort of way, 

 and the detection of such motion falls little short of a demonstration of 



