34 CARNEGIE INSTITUTION 



computers and a somewhat larger staff of mere routine computers. 

 Under these arrangements it should be possible in a good climate to 

 turn out at least 20,000 accurate star positions each year, and to 

 make all the computations necessary to put them in catalogue form. 

 Since not more than 200,000 observations would be required, the 

 entire work should be completed within ten years from beginning — 

 possibly in less time. 



(3) Measuremeitt of Stellar Parallax. 



The determination of the distance of individual stars is one of the 

 severest tasks in the science of obser\'-ation. The results already 

 obtained during a period of more than sixty years, but chiefly during 

 the last fifteen , are not free from troublesome discordances that seem 

 to encourage pessimistic views. Yet a review of the work which has 

 been accomplished, together with that of the larger amount which 

 is now in progress, affords great encouragement that we shall be 

 able to determine average distances of classes of stars with a very 

 satisfactory degree of precision. This is the important thing we 

 need to know in our first studies of the structure of the sidereal 

 system. The two great questions to be solved are : 



(a) What is the relation of brightness to distance ? Are the stars 

 of the sixth magnitude as far removed on the average from those of 

 the second as the relation of brightness alone would lead us to 

 think? According to this relation, the sixth magnitude stars 

 should be on the average rather more than six times as far removed 

 from us as are the stars of the second magnitude. Is this really the 

 fact ? It is a question for measurement to decide. 



(J)) What relation to distance has apparent motion on the face of 

 the sky? Are the stars which appear to move athwart the sky 

 nine tenths of a second per year three times nearer us than the 

 stars which move three tenths of a second per year ? It seems very 

 probable that this is approximately the case, and that apparent 

 motion is a better criterion of distance than apparent brightness. 

 Whether this conjecture is correct or not can be settled only by 

 actual measurements of distances. 



The decision upon these points is a fundamental necessity in the 

 stellar problem, because it opens the way for a much more powerful 

 and economical solution of questions in relation to distribution in 

 distance through the discussion of meridian observations. That 

 way is, to some extent, open now, but we need that certainty upon 



