DOUBLE STARS 171 



to them. The result of the investigation was very 

 remarkable. Myers' theoretical curve is given in fig. 15, 

 in which it is directly compared with the observed curve, 

 which is again reproduced in the dotted line. The 

 coincidence is so nearly exact as to leave little doubt 

 that the system imagined actually represents that of the 

 star. The data determined by Myers as giving the best 

 agreement with observations were as follows : 



1. The stars are similar prolate ellipsoids, the shortest 



diameter of each being f of its greatest. 



2. The ellipsoids differ in size, their greatest diameters 



being in the ratio of 8 to 9. 



3. The larger body is ^ as bright as the smaller. 



4. The distance between their centres is f-f the greatest 



diameter of the larger star. 



FIG. 1 6. System of Lyrae (Myers). 



The relative dimensions of the system are represented 

 in fig. 1 6. 



We are tempted, therefore, to regard & Lyrae as repre- 

 senting the earliest stage of separate existence of the 

 components of a double star. There is also independent 

 evidence to suggest that Lyrae as well as variables of 

 the Algol class in general have densities that are very 

 much less than that of the Sun, and this in itself suggests 

 an early stage of development. Further, in the case of 

 Lyrae, the spectrum is so closely allied to that of 



