THE LIGHT OF THE STARS 299 



in size, and their average distance apart may perhaps be ten times as 

 great. We will suppose that each galaxy is at the center of an other- 

 wise unoccupied cube 10 andromedes on an edge. The radius of a 

 sphere containing 450,000 such cubes is 760,000 light-years. Now 

 Perrine estimates that there are at least 500,000 nebulae in the heav- 

 ens within reach by the Crossley reflector, and probably nine tenths 

 of these are white nebulae or galaxies. It is therefore safe to say that 

 the light of the stars can travel for one million years before becoming 

 so much reduced by intergalactic absorption as to be beyond the grasp 

 of this powerful instrument. 



The view which I now wish to present is that it is the ether itself 

 which absorbs the radiation from the stars. 



Considered merely as to its volume, the ether is so overwhelmingly 

 immense that all other bodies shrink into nothingness in comparison. 

 The radius of the sun is 



r =7 X (10) 5 kilometers. 



Half the distance to the nearest star is 



r # ==2 X (10) 13 kilometers. 



An ethereal sphere which may be called the sun's own, being bounded 

 by the similar spheres of neighboring stars, may be drawn with the 

 latter radius. The radius of the sun bears to that of its interstellar 

 sphere the ratio 



r : r* = 1 : 30,000,000, 



and the volume of the associated ether exceeds that occupied by the 

 solar substance in the ratio 



(r ) 3 : (r ) 3 = 2.7X (10) 22 : 1. 



Since there are vacant spaces between neighboring galaxies, something 

 must be allowed for these. Let us suppose that the ethereal volume is 

 four hundred times greater than that just given, or that its volume 

 ratio is 



Ether volume : Matter volume = (10) 2i : 1. 



This allows a considerable extension of thinly scattered stars around 

 each galaxy, and places the galaxies at relatively smaller distances from 

 each other than the stars, if distances are expressed in terms of diam- 

 eters, an arrangement which is indicated by the evidence already 

 presented. 



The next step in the argument demands an estimate of the total 

 light from all of the stars. Call this L. Newcomb gave us such a 

 photometric measurement, and found 



L = 600 stars of zero magnitude. 



The brightness of the sun is 



