360 THE WORLD MACHINE 



tance. The actual ranges of the stars were, of course, not 

 known to him ; but he seems to have made a splendid guess. 

 According to present-day photometric methods, a star of the 

 first magnitude sheds about a hundred times the light of a sixth 

 magnitude, and so on, so that a star of the sixteenth magnitude 

 sends only one- millionth as much light. So far as we now know, 

 the brilliancy of a star varies evenly and inversely as the square 

 of the distance. If this were strictly true, on the average, 

 therefore, a sixteenth magnitude star would be about a thousand 

 times as far away as one of the first. 



Herschel's forty-foot reflector, with an aperture of four feet, 

 showed stars up to the seventeenth or eighteenth magnitude 

 of our present-day classifications. 



There are twenty or twenty-one stars in the heavens accounted 

 of the first " magnitude." The parallax of three-fourths of 

 these has been fairly determined. With one exception, they 

 are all under fifty light-years. Arcturus comes out at between 

 two and three times this distance. The average of these known 

 stars would fix the mean distance of first magnitude luminaries 

 at about twenty-two light-years, or very evenly five times the 

 distance of the nearest one we know of. This is more than one 

 million times the space between the earth and the sun that 

 is, ninety-three million million miles. 



But there are at least three of the first magnitude stars 

 whose parallax evades our present resources. Present methods 

 would certainly disclose the shift of a star thirty or forty times 

 the nearest, so that each of these three Canopus, Rigel, and 

 Deneb must be considerably more than this. Sir David Gill 

 considers that present limits of error will not permit us to suppose 

 that Canopus is nearer than two hundred and ninety-six light- 

 years. The fact that Rigel has practically no apparent motion 

 seems to fix it at a still greater distance, and Deneb may not 

 be much less. If we set the average of these three at three 

 hundred light-years, this would raise the average for all of the first 

 magnitude stars to above sixty light-years. We may conclude 

 that the mean distance is somewhere between these upper and 

 lower limits. It cannot be less, and it is probably not much more. 



Merely to fix our ideas of this distance, it may be noted 

 that our shining sun would be of the first magnitude only within 

 five or six light-years, and it would quite cease to be visible 

 to the naked eye at much beyond thirty light-years. 



