THE MASSES (^E VLSUAE BINARY STARS. 



Bv Robert TiiuRr.uRx Avton Innks, F.R.A.S., F.R.S.E. 



For a general description of donl)le stars reference nia}' be 

 made to any text-book on astronomy. It must suffice if -ve .^ay 

 here that some 250.000 stars have been examined, and of these' 

 15,000 have been found to be double; but many of these 

 double stars are not very near to each other, and it would lie 

 hopeless to expect to find any sign of gravitational motioni 

 between them. A small percentage of the total numlv^n- are- 

 optical pairs, meaning that apparent jjroximity is only due to the 

 point of view. When the stars of a double star revolve around" 

 each other, they form a binary system. When about half the 

 revolution of one star around the other has been described, the 

 actual orbit can be calculated. Although measurements of 

 doi;ble stars have now accumulated for nearly a century — and 

 in great number during the last half-century — very few orbits- 

 have been found, simply because in nearly every case the motion 

 is so small. Besides this, the errors of measurement are con- 

 siderable, so that until half a revolution of a j:)air has been 

 accomplished it is impossible to com])ute a reliable orbit. Never- 

 theless, astronomers are certain that, with very few exceptions, 

 all close pairs of stars are connected systems — in very many cases 

 this is easily proved, because both stars of the ])air are nic>ving 

 through space together, or technically, are moving with common 

 proper motion. 



In dealing with double stars, certain simple fornrcike are 

 wanted, and for purposes of easy reference these are relegated to- 

 an appendix to this paper. 



Table I contains all those double stars for which fairly 

 reliable orbits have been computed. The magnitudes and spectral 

 classes have been taken from the Harvard Annals. Col. 8 of 

 this table ( H} pothetical Radial) gives the distance in radials 

 for parsecs ) to which the Sim would have to be removed to shine 

 with the same magnitude as the double star. With this hy])o- 

 thetical radial and P, the period, col. 9 has been computed ; it 

 gives the mass or the gravitative power — the expression " gravi- 

 tative power "' appears to be preferable, as mass suggests a body 

 of matter, and it is an assumption to consider mass as equivalent 

 to gravitative power. Col. 10 gives the measured parallax for 

 those stars for which it is available ; col. 1 1 is the reciprocal of 

 this parallax or the radial. Col. 12 gives the magnitude of the 

 Sun if removed to the distance in col. 11. Col. 13 repeats the 

 calculation of col. 9, using, instead of the radial derived from 

 magnitude, that derived from parallax measures. So far as this, 

 table alone goes, it is evident that many of the parallax measures 

 are very weak, and that the distance, upon the simple assumption 

 that all stars have the same brightness as the Sun. is about as 

 trustworthy. 



