VARIABLE STARS ROBINSON 129 



analyses of the light received from them by means of the spectro- 

 scope. Still more complicated are those systems which the telescope 

 shows as double or multiple, where one or more components of the 

 visual system are in reality spectroscopic binaries. Thus, the astron- 

 omer is able to discover that what appears as a single star to the 

 unaided eye is in some cases a complex system of stars revolving 

 about their centers of gravity, with periods from a fraction of a day 

 to hundreds or even thousands of years. Such a system is Castor, 

 the fainter member of Gemini, the twins. In this instance the tel- 

 escope shows two bright stars exceedingly close together where the 

 eye only sees what appears to be a single star. These two stars 

 revolve about each other with a period of about 350 years. The third 

 and fainter member of this same system is found a little farther 

 away revolving about the common center of gravity probably with 

 a period of more than a thousand years. Each of these three stars 

 is a spectroscopic binary, so that in reality we have six stars be- 

 longing to a system where the unaided eye would see a single bright 

 star. It is estimated that of all the stars in the sky, one of every 

 three or four is a double or a multiple system. This is only one bit 

 of evidence suggesting gregarious tendencies among the stars. 



Astronomers believe that the distinction between visual telescopic 

 double stars and the class of eclipsing and spectroscopic binaries is 

 more real than apparent. Members of the former class have periods 

 measured in years while those of the latter class, with a few ex- 

 ceptions, have their periods restricted below 100 days. But why the 

 difference ? This is one of the questions of modern astronomy. Un- 

 less the universe is very much older than we believe it to be, near 

 approaches and captures are not frequent enough to account for the 

 large number of visual and telescopic doubles even if, under such 

 circumstances, captures are possible. Serious doubts indeed arise if 

 we assume that capture of one star by another may result at all from 

 near approach. Probably the best suggestion is that visual doubles 

 result from condensations about two separate nuclei in an early nebu- 

 lar stage of evolution. The revolution of one star about another, 

 with reference to spectroscopic and eclipsing binaries as well as to 

 visual doubles, is not more strange than the much more general fact 

 now established by modern astronomy, that all the stars of the Milky 

 Way system are revolving about their common center of gravity. 

 To carry this idea further, it must also be concluded that visual and 

 telescopic binaries were formed at a time when the Milky Way 

 system was yet young and when all the stars were closer together 

 than at present. At least we now see no evidence of the formation of 

 such systenis, 



