THE STRUCTURE OF THE UNIVERSE 389 



a long way. Some millions of dark bodies might lie within this 

 vast abyss. One of these might be huge enough to be a centre 

 of revolution for our system ; we have no present evidence 

 of its existence. 



So much for fancy. Scientific method has not been more 

 successful. 



The first to attempt a conception of the structure of the 

 universe by induction from observed facts was William Herschel. 

 It was, indeed, the great problem of his life. To gain some sort 

 of an idea, he set about systematically counting the number of 

 stars in different telescopic fields to determine whether there is 

 any wide variation in their apparent number. This process 

 he called " star gauging." 



The actual count of the whole heavens was, of course, a 

 physical impossibility with such telescopes as he employed. 

 These were, for^this purpose, not greatly inferior to the best 

 which we now have. He therefore counted selected portions 

 of the field, and made his estimates accordingly. His industry 

 was amazing. Before he was through he had actually made 

 computations for 3400 telescopic divisions of the sky. What 

 he found was that the stars seemed very much more numerous 

 in the direction, or as we may say, the plane, of the Milky Way 

 than in the heavens at right angles to this. He therefore con- 

 jectured that there might be some definite distribution in a 

 word that the universe, at least such of it as the telescope 

 may disclose, does possess a definite shape. 



If it may be supposed that the stars lie at something like 

 a general average distance one from the other, the result of 

 Herschel's investigations was to confirm in a measure the ideas 

 of Wright and Kant. He thought that, looking in the direction 

 of the Milky Way, the stellar universe might extend perhaps 

 five times as far as in the direction perpendicular to this. 

 He did not conceive, however, that this arrangement was per- 

 fectly uniform, but that in one direction there is an immense 

 bifurcation. The figure which he drew was bizarre enough. 

 It resembled a two-legged body with no head. 



Herschel did not himself hold by his two-legged universe. 

 Further consideration revealed to him that the assumption of 

 equal spacing, or, as we may say, equal star density, might be 

 wholly erroneous. His discovery of binary systems and the 



