520 



THE POPULAR SCIENCE MONTHLY 



effect of such difference. When the limit of separating power for 

 a telescope of a particular aperture is set at 1" or 2", as the case 

 may be, it is assumed that the stars composing the doubles on 



which the test is made shall be of nearly the same magnitude, or 

 at least that they shall not differ by more than one or two magni- 

 tudes at the most. The stray light surrounding a comparatively 

 bright star tends to conceal a faint companion, although the tele- 



