xii TOWARDS INFINITY 323 



tional. It happens that stars of this type have their 

 light partially obliterated by the interposition of dark 

 bodies between us and their luminous surfaces at regular 

 intervals ; but if these dead worlds passed a little above 

 or below the line of sight from the earth to the bright 

 star there would be no periodic eclipse, and, therefore, 

 no visible indication which would lead us to suspect the 

 existence of dark globes near the stars we see. Dark 

 stars may revolve around bright stars in orbits inclined 

 at any angle to the level in which the earth travels around 

 the sun, but only in those cases in which they are near 

 this level when they pass between us and their luminous 

 companions can a variation of light caused by them be 

 noticed. 



It may seem, therefore, that the score or so stars 

 which are periodically dimmed by eclipse are the only 

 objects which are known to possess dark partners ; but 

 this is not the case. If two globes a bright one and a 

 dark one are really a united couple, each influences 

 the movements of the other, no matter what the direction 

 may be in which they revolve around their common 

 balancing point. The spectroscope shows that Algol is 

 swung back and forth by the invisible mass near it. By 

 what may almost be termed a fortunate accident, this 

 dark body passes in front of the bright star every three 

 days, and causes a sudden loss of light, but even if it 

 did not come directly between us and the luminous 

 surface the spectroscope would still detect a periodic 

 increase and decrease of pitch of the light rays. The 

 spectroscopic evidence of the movements of stars towards 

 the earth or away from it is, in fact, independent of 

 variations of light. If upon analysing determinations of 

 movements of approach or recession, it is found that a 

 star has its velocity increased and decreased in a definite 

 G.D. x 2 



