yo 



which is the larger of the two stars, is redder than 

 the bright central object. If w r e may accept, as 

 generally valid, the result that for all systems of 

 known mass-ratio the brighter object is invariably 

 the more massive, it follows that the fainter and 

 redder companion of RR Draconis is of a much 

 lower density than the brighter object. Further, 

 if the redder color really represents more advanced 

 spectral type, commonly associated with a later 

 stage of development, we should have the unex- 

 pected case of a star of low density associated with 

 a less-developed star of higher density. As this 

 is contrary to the usual view that a star's density 

 increases with its age, further investigations of 

 such systems may prove of importance in the 

 study of stellar evolution. 



LIGHT-SCATTERING IN SPACE. 



The collection of nearly 4,000 photographs of 

 stellar spectra thus far obtained on Mount Wilson 

 is available for many classes of work. It affords 

 Kapteyn material for the further investigation of 

 star-streaming, which is now being studied with 

 reference to the association in streams of stars in 

 different stages of physical development; but it 

 also permits many other problems to be attacked. 

 One of these is the question of light-scattering in 

 space. We have already seen that the fainter 

 stars are redder than the brighter ones, but the 

 meaning of this result is not yet certain. 



One of the methods used to determine the rela- 

 tive colors of near and distant stars was to photo- 



