THE ASTRONOMY OF THE INVISIBLE 375 



example, the period of Mercury is eighty-eight days, and its 

 orbital motion is nearly thirty miles a second. It is the fastest 

 member of our system. If the twin stars of Polaris are travel- 

 ling at an average speed of only a mile or two per second and 

 encompass their mutual revolutions in four days, they cannot 

 be separated by more than a small fraction of the distance 

 of Mercury from the sun. 



These revelations bear with them many interesting implica- 

 tions. The detection of a binary system either by means of 

 the spectroscope or by measures of actual displacement, is a 

 matter of the most delicate observation. The shifting of the 

 lines in the spectral image is exceedingly minute ; the inferences 

 one may draw are complicated by many physical problems, 

 each of which may introduce a possible error. It is obvious that 

 each system must be treated by itself one does not lead to 

 another. All this takes time. 



On the otner hand, the actual displacement of the double 

 stars from their observed positions is measured for the most 

 part by hundreds of seconds of arc, and we have already seen 

 what accuracy these micrometric measures involve. Sometimes 

 the series of observations must extend over half a century or 

 more before any reasonable inference may be drawn. Here 

 again the chances of error are great, and progress is necessarily 

 slow. 



If, in spite of all this, a little more than a dozen years has 

 sufficed to reveal more than sixty spectroscopic binaries, the 

 obvious inference is that their number is enormous. Let us 

 supplement this with the fact that out of the comparatively 

 small number of stars whose positions have been determined 

 with very great accuracy, at least two hundred and fifty are 

 known to be double systems, with fifty or sixty of their orbits 

 measured from their displacements. Add to this that per- 

 haps ten thousand stars are known, situated so nearly together 

 that for the most part they can only be resolved into two by 

 means of telescopes of the highest powers of magnification 

 double stars in the Herschelian sense. One might almost con- 

 clude from this that the double-star system was the general 

 pattern of the universal arrangement, and that our solar system, 

 so far from being the type, was relatively unique. One 

 might go further, and conclude that a companion of our sun 



