STUDIES ON THE VARIABLES OF THE ALGOL-TYPE. ö7 



Thus, if we speak only about the Algols whose spectrum 

 became already known, it would be quite safe to conclude that 

 the Algol, in general, displays the spectrum of tlie type A. 

 Besides this predominant type of spectrum, we see some stars in 

 the earlier tyjie, but on the contrary, the more advanced type 

 of the spectrum is very rarely met with. This peculiarity may 

 be either connected w^ith the physical properties of the Algol- 

 system or caused accidentally on account of our imperfection to 

 detect the light variation for the advanced stars. Prof. Campbell 

 lately studied spectroscopic binaries statistically and touched on 

 this question and he said as follows (Publ. A. S. P., XXII, 55) : 

 " The two members of an Algol-system are in general so near 

 each other and so large in diameter that eclipses occur with 

 great ease : they are observable by those of us w^io are not 

 situated exactly in the plane of orbit ; the eclipses last long 

 time, so that even the unsystematic observations of the past have 

 readily detected variable brightness. As the two members of a 

 system grow further and further apart, corresponding to the 

 longer periods, the number of eclipsing pairs decreases very 



rapidly, Algol stars of the type G-M appear not to have 



been observed for the reasons that the binary stars of these 

 types have their components relatively far apart; and the eclipses 

 would be of relatively short duration both because they are 

 smaller in diameter, being more condensed. The observer would 

 have to be situated very closely in the plane of the orbit in 

 order to witness an eclipse and his chances for observing an 

 eclipse would be small, as compared with eclipses of close 

 system." 



Prof. Schlesinger and Baker expressed the same opinion in 

 their paper " A comparative Study of Spectroscopic Binaries " 



