126 REPORTS OF INVESTIGATIONS AND PROJECTS. 



question. The computations of double-star orbits are frequently revised, so 

 that there are usually available recently published results based upon nearly 

 all the available observations down to the present. But in the matter of 

 determination of the relative masses for the components of binary systems the 

 conditions are not so favorable. Accordingly, we have determined for our- 

 selves the relative masses of the two components for each binary offering 

 promise of a useful result. The determination of the relative masses from 

 meridian observations is a work of some difficulty in practical application. 

 The observations are liable to peculiar errors, especially in the case of close 

 and unequal pairs. Some of these errors are systematic. Nevertheless, the 

 chief difficulty has always been that of determining the dimensions of a 

 minute orbit from the results of meridian observations, which, except in a 

 few special cases, must be done before the relative masses can be evaluated. 

 The revised, or corrected, observations at our disposal in the present instance 

 have greatly facilitated this work. We have attempted this computation for 

 12 binary systems. For som^e of these the evidence is still entirely insuffi- 

 cient to insure a trustworthy result; but in more than half of the cases the 

 determination seems to be entitled to some confidence. 



Agreeably with the experience of previous computers, we have found, in 

 general, nearer equality between the masses of the two components of a 

 binary star than might have been inferred from their relative brightness. 

 Thus the principal star of Sirius is about 10,000 times as bright as the 

 companion star, while its mass is only about 2.6 times as great as that of the 

 companion. The principal star of v Cassiopeise is about 40 times as bright 

 as the companion, but has a mass only about 1.3 times as great. 



On the other hand, previous computers have found numerous instances in 

 which the fainter star of a pair appeared to have much the larger mass. In 

 no case where the basis of determination in our computation could be 

 regarded as at all adequate has the mass of the companion appeared to be 

 materially greater than that of the brighter star. In fact, of all the stars 

 investigated only two, 85 Pegasi and Bradley 3210, come out with a mass 

 for the brighter star of less than 0.9 (that of the fainter star being the unit), 

 and even in these cases the hypothesis of equal masses is not especially repug- 

 nant to the observations. 



The mass of the sun is more than 1,000 times greater than that of Jupiter, 

 the largest planet ; so that when one star of a pair has not more than three 

 times the mass of the other there may be said to be a tendency toward 

 equality of masses. This is the case with the double stars we have investi- 

 gated ; and with existing ideas of stellar evolution it is rather difficult to 

 devise a satisfactory explanation of this prevailing tendency toward equality 

 of masses. It may be easily admitted and explained for an isolated case, or 

 for a proportion of all ; but why this near equality should be the rule rather 

 than the exception is more difficult to understand. But if we accept the 

 results of computers who have recently found in several instances that the 



