18 ART. 10. — N. ICHINOHE : 



and a expressed in the terms of the astronomical unit and the 

 radius of the principal star respectively. Tlius, we obtain 



"="" 7=-(ï+ïT ^ ^ 



Therefore, for the Algols whose orbits are already determined, 

 as in ten Algols which ^vere observed by Graff and their circular 

 orbits were also determined by him (Mitt. Hamburger Sternwarte, 

 No. 11), the density will easily be obtained, as Graff's orbits were 

 used by Ristenpart (A. N., 178, 31). 



But for the stars in which we do not yet know the elements 

 of the orbit, it would be more convenient for us to replace a and 

 '/. by elements of the light-variation themselves. 



As 1 + ;f > « sin ^, 



ß 

 we have o< ^ ^^i (3) 



P' sur -p 



where d represents the duration of the eclipse. As it is easily 

 seen, the value obtained with the formula (3) always exceeds or 

 equals to that given by (2). To determine the limiting value of 

 density according to the formula, it is the first importance for 

 us to determine the constant. H. N. Russell (A. P. J. X, 315) 

 determined this constant, taking the density of water as the 

 standard and assuming the density of the earth 5.53, and also 

 expressing time in hours, as follows : 



44.1 



André, in his '^Traité d'Astronomie Stellaire," (Tome II, 308), 

 deduced the constant from the known density of the Algol, 

 which is previously determined to be 0.23 in the term of the 



