STELLAR ASTRONOMY 



399 



These are quantities which we may hope to measure with some 

 precision. 



It is true that, for the elucidation of most questions, we require 

 accurate proper motions for multitudes of stars not observed by 

 Bradley or other early astronomers, but the difficulty is not a formid- 

 able one. As de Sitter and myself tried to demonstrate elsewhere, 1 

 photography enables us to obtain, in a dozen years, proper motions 

 of as many stars, down to the faintest we can photograph, as will be 

 required for our purposes. The precision need no way be inferior to 

 that of the bulk of the Bradley stars. 



It is also true that a still greater base-line would be acceptable; 

 but we may provisionally be content. The base as it is will enable us 



P * 



(Apex) a 



R 



*~X {Anlapex) 



FIG. 1 



to reach many conclusions, and even while we struggle on with our 

 problem the base is lengthening out, the precision of the observa- 

 tions is increasing. We may look forward to enormously improved 

 data by the time we have exhausted the treasures virtually contained 

 in the data already now available or obtainable in a short time. 



Of course the difficulty of using the parallactic motion as a meas- 

 ure of distance is, that, for individual stars, we do not know what 

 part of the observed motion is parallactic, that is, due to the solar 

 motion, what part is peculiar to the star. The two become separable 

 only for numerous groups of stars, and then only by the introduction 

 of some hypothesis. 



The most plausible thing to do seems to be to adopt as such the 

 hypothesis already in general use in the derivation of the precession 

 and the sun's motion in space: viz., " The peculiar motions of the stars 



1 See Publications of the Astronomical Laboratory at Groningen, no. 14, preface. 



