358 THE WORLD MACHINE 



reach, the extent of the starry universe is an impenetrable 

 enigma. 



This may well be ; but restless minds have not hesitated 

 to attempt a guess. One is based upon the fact that we cannot 

 as yet be certain that light is in any way absorbed in passing 

 through empty space. It is an old remark that if it is not, 

 and the number of stars is infinite, the heavens would be ablaze 

 with light night and day, with a marvellous white light at that ; 

 it would give to the disk of our sun the appearance of a sickly 

 yellow. 



Nothing of the sort exists, and one inference we might draw 

 would be that the number of the stars is finite. The explana- 

 tion may be quite otherwise. Space, as we shall see, holds 

 more things than was imagined. The light of the stars may be 

 cut off or dimmed in several ways. If it should turn out that 

 it is not, in a vague way the number of the suns would be 

 measurable ; and there has been more than one attempt in 

 this way to indicate the limits within which their number must 

 lie. One of these is based upon the amount of star-light which 

 reaches us. 



Even in the old Alexandrian days, Hipparchus had adopted 

 a method of classifying the stars according to their apparent 

 brightness. Thus the Dog-star, Arcturus, Vega, and their like, 

 were said to be of the first " magnitude " ; those a little fainter, 

 like the Pole-star, the " pointers " of the Dipper, of the second ; 

 and so on. All that are visible to the naked eye were classed 

 in six such divisions ; the faintest of them were broadly grouped 

 in the sixth magnitude. With the invention of the telescope 

 this method of classification has been carried out to the seven- 

 teenth and eighteenth magnitude. It was noted that, roughly 

 speaking, the number of stars catalogued in each order of mag- 

 nitude was about three times that of the next. Calculating 

 as nearly as he might the total intensity of star-light, the 

 French astronomer 1'Hermite endeavoured to estimate the 

 number of suns which would be required to shed the light they 

 do. He fixed the outside limit of the possible number at 66,000 

 millions. If star-light reaches us integrally and undimmed, this 

 limit is certainly far outside of the reality. 



A very ingenious method of attacking the problem in another 

 way was worked out by the distinguished American astronomer, 



