338 THE WORLD MACHINE 



by a star is dependent simply upon its surface. Surfaces are 

 in turn proportionate to the square of the diameter, volumes 

 to the cube. If Canopus gives out more than ten thousand 

 times the light of our sun, it follows that its diameter is more 

 than a hundred times that of our sun ; that is, perhaps more 

 than a hundred million miles in diameter. This would mean 

 a sun, a body, the distance from whose centre to circumference 

 would be four hundred times the distance from the earth to the 

 moon ; it would be greater than the distance from the earth 

 to the sun. It would be so vast that within its almost un- 

 imaginable shell the earth and all the inferior planets might, 

 in the absence of a deterrent medium, pursue their orbital ways. 

 On present reckonings, its volume is perhaps a million and a half 

 times that of our central orb. 



Reflect that the sphere of the sun would contain a million 

 of our little earths ; then that, on the estimates given, the 

 sphere of Canopus would hold more than a million suns, more 

 than a million million earths. 



All this, it is to be remembered, is simply the apparent 

 lower limit. How much more vast it may be we have no means 

 of knowing. Already the mind reels in its endeavour to com- 

 prehend a body, a single mass, of such stupendous dimensions. 

 Even the greatest of the telescopes probably do not reveal more 

 than a hundred million stars. If we were to suppose that they 

 were of something the average size of our sun, then Canopus 

 would swallow a million of them. And, for aught we now 

 know, Canopus may be a hundred, a thousand, a million times 

 vaster still. It seems probable that some of the stars are 

 distant more than thirty thousand light-years. If the great 

 sun of Argo be as far away, its bulk would necessarily be that 

 of a million million suns to give us the light it does. 



Nor have we any reason to suppose that this colossus is 

 the greatest object in the universe. There are at least two 

 other stars known, which may be as large, or larger, than 

 Canopus. One of these is Rigel, the second brightest star in 

 the constellation of Orion ; the other Deneb, the brightest star 

 of the constellation of the Swan. Like Canopus they yield 

 no appreciable parallax, yet their luminosity is so great that it 

 must be thousands of times that of the sun. 



Consider now that these estimates of actual brilliancy, that 

 is to say, of the true grandeur of the stars, is dependent upon 



