324 THE BINARY STARS. 



knows that their revolution is performed in a hundred and four 

 days ; he is therefore able to proceed with his calculation in the 

 same way as I have described in the case of Sirius, and he finds 

 that the combined masses of the two stars forming the spectro- 

 scopic double which we call Mizar is no less than forty times as 

 great as the mass of the sun. 



I will now describe one other very wonderful result which has 

 followed the application of this spectroscopic method of research 

 to one other star, which, although very different from any I have 

 before mentioned, must yet be placed in our fourth class among 

 the spectroscopic binaries. This star is the famous and well- 

 known variable Algol in the constellation Perseus. Although 

 beheved to be a binary before the spectroscope was ever applied 

 to it, yet it seemed impossible that the supposition could ever be 

 proved, until the spectroscope furnished us with most convincing 

 evidence of the fact. The remarkable variability in the light of 

 Algol had attracted attention to it for centuries, as its changes of 

 lustre are quite perceptible to the unaided eye ; and several 

 theories accounting for the vagaries of the star have been pro- 

 pounded. But the one which has commanded most respect, and 

 which was generally believed by astronomers to be the correct solu- 

 tion of the mystery attaching to Algol, was one which assumed that 

 Algol was one of the binary stars having a large dark companion 

 revolving round him, so that it sometimes passed in front of him 

 as seen from the earth, thereby cutting off a portion of his light, 

 or in other words partially eclipsing him ; and it is this theory 

 which has received such wonderful confirmation from the spectro- 

 scope. The variability of Algol has been found to be almost 

 perfectly regular, and the star passes through the complete cycle 

 of its changes in a period which has been ascertained with great 

 accuracy to be 2 days 20 hours 48 minutes and 51 seconds. For 

 a period of 2 days 13^ hours, Algol shines with a steady light as 

 a star of the second magnitude ; then his brightness begins gradu- 

 ally to decline, and for 3-| hours he slowly becomes fainter, till he 

 has lost about three-fifths of his light. At this, the lowest point, 

 his brightness appears to remain stationary for about twenty 

 minutes, and then it begins to increase in the same way that it 

 before diminished, until in 3^ hours from the time of minimum 



