Close Binary Systems. 115 



RR Puppis, in the Southern Hemisphere, and both stars were dis- 

 covered at the Royal Observatory, Cape Town — is due to the 

 rotation of a large dark body round a small bright companion. 



We have said that at first sight this seems a strange and con- 

 tradictory state of things ; yet I venture to offer a rational and simple 

 explanation. 



Let us consider a mass of tenuous matter rotating on its axis. 

 It is not unreasonable to suppose that the nucleus of such a system 

 might be nearer one side than the other : and that in eccentric layers 

 the whole mass condensed round this nucleus. That is, we would 

 have a body such as is imperfectly represented in Fig. 2. 



It is evident that such an unstable condition of mechanical 

 adjustment would favour bipartition : whereas a sphere-shaj^^ed 

 body, perfectly symmetrical and uniform in structure, would present 

 no line along which cleavage might take j)lace. 



Now, constriction setting in along the line AH (Fig. 2), and 

 this leading, after the lapse of countless ages, to bipartition, there 

 would result two bodies, one dark, large, light : the other bright^ 

 small, dense. 



We would have the very condition of things which exists in 

 stars of the type of S Velorum. In process of time the larger body 

 might still further divide and sub-divide — the irregular density of the 

 mass would favour such disruption. 



Is it possible that in a star like S Velorum we have the first 

 beginnings of another solar system? 



(B.) The question of the density of close Binary stars is intimately 

 connected with the preceding inquiry, and, therefore, with all pro- 

 blems dealing with the evolution of stellar systems. Indeed, the 

 density of an Algol Binary is a question of fundamental importance in 

 Astrophysics. 



It is fortunate that this inquiry admits of rigorous treatment. 

 The equations of orbital movements and dimensions can be related 

 directly to the density of occulting binary stars. The investigation 

 premises great refinement of observations, as all cosmical problems 

 do. but granted this accuracy of data, the deductions are incontro- 

 vertible. 



It is found that the densit\ of all close binary systems is, with- 

 out exception, many times lighter than that of the sun. For 

 example, it is found that the mean density of eight southern Algol 

 stars comes out as one-eighth that of the sun. 



This is a remarkable result : and none the less remarkable from 

 the fact that it is the state of things we would expect to find in a 

 binar) system that just have been evolved from some irregular ga,seous 

 mass. The result proves that at their genesis close l)inarv stars are 

 composed of matter "light as air." 



(C.) We have stated that the light curve of RR Centauri, and 

 of four other Algol stars, indicate that the light changes of this class 

 of binary stars are caused by the revolution of two bodies in 

 contact. 



