c,._A CONSIDERATION OF CLOSE BINARY SYSTEMS 

 IN RELATION TO LIGHT VARIATION. 



By Alex. \V. Robekis, D.Sc, F.R.A.S.. P\R.S.E. 



If two stars revolve round one another in such a manner that at 

 e\ery revolution each component comes between its companion star 

 and the earth, it is evident that the total light which comes from 

 the system will undergo periodic and regular variation. 



This phenomenon of stellar variation, of a well-defined type,, 

 is exceedingly interesting, not only because it affords a striking" 

 manifestation of celestial movement, but also because it yields data 

 which may be of service in investigating some of the most important 

 problems of modern Astronomy, for it will be plain that the 

 character of the periodic eclipse due to orbital movement will be 

 determined to a considerable extent by the relative size and bright- 

 ness of the two stars forming the system ; or. to put the matter the 

 other way, light variation of a certain well-defined character will be 

 no untrustworthy indication of the dimension and brightness of the 

 stars producing the variation. From the form and dimensions of 

 any stellar orbit to the weight and densitv of the stars circling in 

 this given orbit is but a step. 



Thus the obscuration of the light of any star by a revolving 

 companion is an occurrence of far more significance than at first 

 sight appears. It is this intimate relation of stellar variation to 

 celestial mechanics that has during the past ten years raised the 

 study of variable stars from being a pastime to a science. 



It is no doubt unnecessary to point out here that all light 

 variation is not due to eclipse ; it is only variation of a certain 

 definite type that is caused by the revolution of one star round 

 another. Further, all f)inary stars do not exhibit light eclipse. It 

 is onl\- when they move in such an orbit tTiat at every revolution one 

 or other of the two stars comes into the line of sight that eclipse 

 takes place. When, however, any star revolves round another in 

 an orbit whose plane is coincident, or nearly so. with the plane 

 of sight, eclipse will occur ; and this eclipse will, as we have said, be 

 of a certain definite type, of a character so marked that there will be 

 no chance of mistaking the light changes for variation of a different 

 type, and due to totally different causes. 



A moment's thought will indicate what the prominent features 

 of this type of variation — called Algol variation from the remarkable 

 close binary Algol — are : — 



(i) First, Ave will have great regularity f)f light changes. One 

 cycle of variation will be exactly like another cycle. 



(2) The full cycle of light changes will be completed in a. 



few days, sometimes in a few hours. 



(3) The ascending and descending periods of variation will 



be, approximately, equal in duration. 



