DOUBLE STARS 165 



visible. At a quarter period later, recession and approach 

 cease, as the components travel directly across the line of 

 sight, and, their spectra being superposed, the result 

 appears single. In the greater number of these cases 

 the period of a complete oscillation of the spectral lines, 

 and therefore that of mutual revolution of the component 

 stars, is measured by a few days. In the case of Spica it 

 is four, and in that of Mizar, twenty days. 



From the variations of their light, and through the 

 movements of the dark lines in their spectra, we have 

 thus become acquainted with double stars, the compo- 

 nents of which are so close as to evade all possibility of 

 telescopic separation. In many of these cases the actual 

 distance separating the stars must be very small : in the 

 case of Algol and its companion for instance, it is, as we 

 have seen, less than five times the radius of the bright 

 star. We shall conclude our study by considering the 

 case of a star that has received a great deal of attention, 

 and which presents features that have not even yet 

 received an entirely satisfactory explanation. 



In 1 784 Goodricke observed that the light of the bright 

 star Lyrae varied between fairly wide limits. The 

 remarkable character of its variation has attracted a great 

 deal of attention since, and in no other instance has the 

 law of light variation been determined with so high a 

 degree of accuracy. Between 1840 and 1860 Argelander 

 made more than 1,500 comparisons of the star with 

 others in its neighbourhood, and from them expressed 

 the law of light variation by means of a curve. This 

 curve is reproduced in fig. 13 1 . 



Referring to the figure, it will be seen that the varia- 

 tion is essentially different from that of Algol. Instead 



1 The variation may be easily followed by the naked eye, by com- 

 paring the star with others in its neighbourhood from time to time. 



