1 64 THE EVOLUTION OF 



its greatest speed, and again at a quarter of the period 

 after its minimum, when it should be at Q, and approach- 

 ing with greatest speed. As was anticipated, it was found 

 that in the first photograph the spectral lines were dis- 

 placed towards the red, while in the second they were 

 displaced towards the violet. Further, the amount of 

 displacements indicated an orbital speed for the star of 

 twenty-six miles per second. It follows from a simple 

 calculation that, to travel with this speed and to occupy 

 68-8 hours in describing its orbit, the radius of the orbit 

 the distance OA in fig. 12 must be approximately 

 a million miles. 



In many other stars phenomena are presented very 

 similar to those of Algol, and almost all of these have 

 been discovered during the last quarter of a century. In 

 some cases, as in that of Spica, a regular oscillation of 

 the spectral lines indicates an orbital motion of the star, 

 and consequently necessitates the presence of a com- 

 panion, but the motion is unaccompanied by any change 

 of light, so that the orbits must be so inclined to the line 

 of vision that eclipses are avoided. In several instances 

 a variation of light takes place, as with Algol, but the 

 stars that exhibit it are too faint to furnish spectra cap- 

 able of sufficiently exact measurement to establish the 

 fact of motion by their means. In a third class of stars, 

 illustrated by Mizar, the dark lines of the spectra regu- 

 larly open out into doubles and close again with a con- 

 certina-like motion. There can be no hesitation in 

 interpreting this effect as due to a double star, of which 

 both components are luminous. When the line joining 

 them is at right angles to the line of sight, one of the 

 components is approaching and the other is receding from 

 the Earth, so that the lines of their separate spectra, 

 becoming displaced in opposite directions, are separately 



