172 THE EVOLUTION OF DOUBLE STARS 



a nebula, that it appears probable, from that fact alone, 

 that the star is not far removed from the nebulous con- 

 dition. These conclusions clearly support the view of 

 development that we have adopted, but it is unfortunately 

 impossible to enter here into the details of the attractive 

 branches of astronomical research upon which they are 

 based. 



The very considerable eccentricities of the orbits of 

 telescopic double stars appear at first sight to be a 

 difficulty in connexion with the theory, but, although the 

 problem of eccentricity has not been completely worked 

 out, it seems possible that it may actually tend to con- 

 firm it. Darwin has shown that a gradually increasing 

 eccentricity of orbit is itself a necessary result of tidal 

 friction, and that it will be most marked as the com- 

 ponents approximate to each other in mass. Whether, 

 however, this is sufficient to account for the high eccen- 

 tricities actually observed is a matter upon which it is at 

 present impossible to express a definite judgement. It 

 is, however, a suggestive fact, that, upon the whole, high 

 eccentricities are characteristic of the telescopic doubles 

 in which separation is great. In /3 Lyrae, and in stars 

 of the Algol class, the details of light variation demand 

 orbits that differ but slightly from the circular form. 



Whether or not a double star is actually formed by the 

 division of a rotating nebula in the manner suggested by 

 our study, it is certain that tidal friction must play a part, 

 and possibly an important one, in the separation of the 

 components, and in the introduction of eccentricity into 

 their orbits ; and there appears little doubt that the 

 ellipsoids of Maclaurin and Jacobi, as well as the pear- 

 shaped figure of Poincare\ play important parts in the 

 general scheme of stellar evolution. 



