The Evolution of the Sciences 



variation of light, with long or short periods. 

 Spectroscopic investigation shows that also in 

 this case the variable star describes a closed 

 orbit; but occupation by a single companion 

 does not appear sufficient to explain the con- 

 tinuous variation of brightness if the stars 

 are considered to be spherical and their orbits 

 circular. An ingenious hypothesis, which 

 deserves great consideration from the name 

 of its author, Sir Norman Lockyer, considers 

 each of the bodies revolving round a common 

 centre to consist of clusters of meteorites, 

 agglomerated with increasing density from the 

 periphery to the centre, and suggests that the 

 continuous variation of luminosity is caused 

 by the interpenetration of two such clusters, 

 which, by giving rise at the same time to 

 numerous collisions between the meteorites, 

 would rapidly increase the temperature of the 

 system, and thus explain the sudden blazes 

 which sometimes affect the brightness of vari- 

 able stars. 



This hypothesis is superfluous; astronomers 

 prefer, as a rule, to continue to consider all 



variable stars as binary systems to which 



250 



