DOUBLE STARS 167 



repeated indefinitely with almost mathematical regularity, 

 though there is distinct evidence that the period has 

 increased since the first observations were made. 



A little consideration will show that it is possible to 

 explain the variation of ft Lyrae by an eclipse theory, so 

 far as its main features are concerned. The brightness 

 at the chief minimum is almost exactly ^, and that at 

 the secondary minimum almost exactly f of the maxi- 

 mum, and we will, for the purpose of illustration, assume 

 that these simple values are exact. Now imagine two 

 stars A and B to revolve in circular orbits round their 

 common centre of mass, or more simply, and this comes 

 to the same thing as far as the light changes are con- 

 cerned, imagine B to revolve round A, A remaining 

 fixed. Suppose further that the plane of revolution is 

 presented edgeways to the eye, and that the stars are of 

 equal size, but that A is twice as bright as B. Let us 

 call the light of A, 2, and that of B, i. At a particular 

 instant in its revolution B will pass between A and the 

 observer, and will completely eclipse it. The light 

 received is now i, that of B alone. This corresponds 

 to the chief minimum. In a quarter of the period of 

 revolution B will have moved a quarter of the way round 

 A, the observer will receive light from both stars, and its 

 amount will be 3, that of A and B together. This is the 

 maximum. In the next quarter period B will move 

 behind A, and will be entirely concealed. The light, 

 now that of A alone, is 2. This is the secondary mini- 

 mum. By the end of the next quarter, A and B will 

 again be clear, the light rising to 3, and at the end of the 

 fourth quarter the original condition will be restored. 



It is possible to explain two equal maxima alternat- 

 ing with two unequal minima by the hypothesis of 

 central eclipses in every conceivable case, though it is 



