DOUBLE STARS 137 



orbit of the second body. In the study of double stars it 

 is, except in a very few instances, the only one that it is 

 possible to obtain, and for many purposes it is as valuable 

 as a record of the separate orbits. As illustrations of 

 relative orbits the reader may refer to fig. 10 (p. 160), in 

 which the orbit traced is the relative orbit of the star C 

 in fig. 12. Again in fig. 3 (p. 136) the stars are those of 

 fig. 4, but the ellipse drawn is the relative orbit of Z. In 

 comparing the corresponding figures it should be noticed 

 that the same scale has been retained in the actual and 

 relative orbits, and that the distances between the stars, 

 as well as the directions in which either would appear if 

 seen from the other, are the same whether the actual 

 orbits of the pair, or the orbit described by one of them 

 relatively to the other, are taken to represent the motion. 



If then the components of a double star are in actual 

 and not merely apparent proximity, each should, in 

 obedience to the laws of motion and gravitation, de- 

 scribe an elliptical orbit (the relative orbit) round the 

 other, or, more definitely, both should describe elliptical 

 orbits round their centre of mass as a common focus. In 

 our illustrations we have already anticipated that this 

 mutual revolution is actually observed. It was first 

 detected by Herschel in 1802 in the components of 

 Castor, and at the present time several hundreds of 

 mutually revolving pairs are known. It will be sufficient 

 to illustrate these by the detailed description of one case, 

 that of Sirius. 



In 1862 a small star was detected by Alvan Clark in 

 the immediate neighbourhood of Sirius. Observations 

 have since shown that it is in motion, following an ellip- 

 tical path round Sirius, and from repeated measurements 

 of its distance and direction from Sirius it has been 

 possible to trace its relative orbit with considerable 



