WEIGHING THE WORLDS 



forward shift of a star having no visible companion 

 suggests revolution in connection with a dark com- 

 panion; and the classical case will be recalled of 

 Sirius, the observed perturbations of which were 

 thus explained some years in advance of the dis- 

 covery of a companion too dim to be visible until 

 a telescope of unexampled power had been con- 

 structed by Alvan Clark. 



It is obvious from what we have seen as to the 

 means of weighing planets, that observation of the 

 swing of the two stars in a mutual orbit would enable 

 the astronomer to compute the aggregate mass of 

 the two bodies, if the actual size of their orbit can 

 be determined. To make this determination, how- 

 ever, it is necessary to know the distance of the 

 stars from the point of observation, that is to say 

 from the earth; and this as we have seen is only pos- 

 sible in case the stars are so relatively near that their 

 parallax can be determined by observations taken at 

 opposite points of the earth's orbit. Fortunately it 

 happens, however, that several stars of known paral- 

 lax are binary stars. Alpha Centauri, the nearest 

 of all, is a binary; and the fact that Sirius, known 

 to everyone as the brilliant Dog Star, is a binary 

 has just been referred to. In both these cases, as 

 well as in several others, the astronomers have been 

 able to translate the observed orbital shift of the 

 stars into terms of actual size of orbit, and thus to 

 compute the joint mass of the duplex system in ac- 

 cordance with the same formula which gives the 

 mass of a planet and its satellite. 



It appears that Sirius and his companion have 



93 



