160 SEE— DYNAMICAL THEORY [Ai>rii 19, 



double star orbit with a mean distance of 30 must be considered small 

 compared to many orbits which exist in the heavens. For there are 

 physically connected stars which show very little motion in a cen- 

 tur}-, and others which remain (juite fixed, as may be clearly estab- 

 lished by comparing modern measures with those of Herschel and 

 Struve. 



The conclusion from this calculation is that the observed mean 

 distance of wide double stars has not been developed by the transfer 

 of moment of momentum of axial rotation to moment of momentum 

 of orbital motion. By such transfer of moment of momentum the 

 orbit may indeed be expanded, but not to many times its original 

 size. On this tidal frictional theory the larger orbits of double stars 

 could not be explained satisfactorily. The difficulty encountered 

 some twenty years was therefore first overcome in developing the 

 second volume of my " Researches," along the lines of thought result- 

 ing from the extension of Babinet's criterion in 1908. 



Looking at the problem in the light of recent progress it is evident 

 that the large and highly eccentric orbits of double stars do undoubt- 

 edly point to capture ; that is, the formation of separate nuclei at a 

 great distance, and the revolution of the two stars in narrowing 

 orbits about the center of gravity of the system. If this process of 

 revolution in the original nebula should continue long enough, the 

 size and eccentricity of the orbit would be much reduced ; and we 

 should thus obtain systems of the type commonly observed to be in 

 comparatively rapid revolution. There is thus established a real 

 connection between the revolving visual double stars and the much 

 larger number of physical systems which have remained nearly if not 

 quite fixed since the epoch of Herschel and Struve. 



This inference is also sustained by recent progress in double star 

 astronomy, which shows that the longer the period the higher the 

 eccentricity, and the same tendency holds for the rapid spectroscopic 

 binaries, as I pointed out in 1907 {Monthly Notices, Roy. Astron. 

 Soc, Nov., 1907). This unbroken continuity among all the classes 

 of double stars shows that the cause is everywhere the same. If 

 therefore the wdder visual double stars have formed from separate 

 nuclei, in the condensing nebul?e, the explanation becomes valid also 



