54 DUGAN— RESULTS FROM THE [April 24, 



unknowns. The probable errors are of satisfactory smallness, con- 

 sidering the amount of asymmetry. 



It seems necessary in this case to face the fact of asymmetry 

 unflinchingly. The secondary minimum gives no evidence of an 

 eccentric orbit and the consequent possibility of a magnified peri- 

 astron effect. The most conspicuous asymmetry is a greater bright- 

 ness from middle of primary eclipse through to secondary than on 

 the other side. A sine term of the first order with an amplitude 

 of three hundredths of a magnitude takes care of the greater part 

 of this. The existence of this sine term would probably indicate 

 that the advancing side of the bright star is brighter than the follow- 

 ing. If such were the case, then the loss of light during the early 

 stages of eclipse — when the brightest part of the disk is being 

 covered — and the gain of light right after totality would both be 

 more rapid than would be the case if the disk were uniformly bright, 

 as was assumed for the least square solution. The divergence from 

 the theoretical curve of the well observed branches of primary 

 minimum is in this anticipated direction and it is of about the right 

 amount. A similar asymmetry has already been remarked in the 

 curve of RT Persei and is probably also, at least in part, due to this 

 cause. This suggested explanation is of course not new, but the 

 evidence is apparently strong that the advancing side of the brighter 

 component of some Algol variables is brighter than the following 

 side. After this sine term is removed from the curve of RV 

 Ophiuchi, there seem to be other changes in brightness of an amount 

 small but seemingly guaranteed by the probable error. 



In the system RT Persei, one star is one and one third times as 

 large and five times as bright as the other ; in s Draconis, one star is 

 thirteen times as bright as the other but just about equal to it in 

 size. The eclipse is very deep and nearly total. In RV Ophiuchi 

 one star is twelve times as bright as the other but smaller. The 

 brighter star is entirely hidden behind the fainter for about an hour. 

 The two stars are farther apart than in the other two systems but 

 they are apparently much more elongated. 



During the seven years since completing the light-curve of RT 

 Persei I have observed through an occasional primary minimum 

 both visually and photographically. Recently I have observed two 



