STUDIES ON THE VARIABLES OF THE ALGOL-TYPE. 



49 



of light necessary to cause the sensation to our eyes of an un- 

 mistakable change of light, when that amount is reduced from 

 the normal brightness. Therefore, the limit becomes less and 

 less as the method of investigation is developed. At present, 

 the change of light, which is less than half magnitude in its 

 range, is very difficult to distinguish from those caused from 

 accidental origins independent of the real change of light in the 

 star. If the companion of an Algol is absolutely dark, and its 

 dimension equals or exceeds that of the principal star, then the 

 system will show tbe minimum in the phase of absolute darkness, 

 when the eclipse is central. In fact, the range of variation 

 observed in Algols extends from a half magnitude to four full 

 magnitudes in our stars, excepting RR Draconis whose minimum 

 is not yet determined as it sinks more than 1ö'".4. Arranging 

 the range of variation in the increased sense, the actual counting 

 of the number gives us the distribution of the range of variation 

 as follows : — 



TABLE XVL 



Thus, there is a tendency for the number of Algols to 

 decrease with the increase of the range of variation. This may 

 be due to the facts that the central eclipse occurs very seldom, 

 and that the probability of the chances for observing the eclipse 

 phenomena by observers on the earth, increases as the distance 

 of the two components, projected on the celestial sphere at the 



