STELLAR LABORATORIES DUNHAM 263 



but here the apparent reddening seems to be due to properties of 

 space which reveal themselves when great intervals are involved. 

 Neither of these effects needs to be considered in studying stars rela- 

 tively close to us within our own system. And so we may quite 

 confidently infer the temperatures of these stars from their colors. 



We come now to the question, "How big are the stars?" Since we 

 can not see the disk of any star with any telescope, we may proceed 

 by an indirect method. 



Supjjose that I had a cannon ball and also a small bullet of 

 something harder even than tungsten to melt. Suppose I were to 

 place them side by side a quarter of a mile away where you could 

 not see the disk on either, and heat them both equally until they 

 were faintly red hot. The cannon ball would be quite easily visible, 

 but the bullet would be so faint that you could hardly see it. If I 

 assured you that they were equally distant and you could see that 

 the color of both was the same you would at once know that the 

 bullet must be smaller than the cannon ball, and if you had an instru- 

 ment to measure how much fainter the bullet looked 5'ou could tell 

 just how much smaller it must really be than the cannon ball. 



But now suppose I should somehow heat the bullet to a much 

 higher temperature, so that it looked brilliantly white hot and should 

 adjust the temperature until the bullet gave out as much light as the 

 much larger, dull red cannon ball. Again suppose that I assured 

 you the two were equally far away. If you were color-blind you 

 would merely see two equally bright points of light and might quite 

 naturally suppose them to be of equal size. But if your eyes were 

 normal eyes you would almost instinctively allow for the fact that 

 the white-hot bullet must be far brighter for every unit of its surface 

 than the cooler, redder cannon ball. You would know that if the 

 two seemed equally bright the white-hot bullet must be the smaller. 



In just the same way the astronomer who has measured the dis- 

 tance and the color of a star can calculate its diameter in miles. 

 Within recent years these calculated diameters have been beautifully 

 confirmed by direct measurement of the diameters of several of the 

 larger stars with the interferometer, an instrument which makes it 

 possible to measure the apparent diameter of a star too small and 

 too distant to show a disk in a telescope. 



With this confirmation of the method, the diameters of many stars 

 have been determined. They show a remarkable range in size, from 

 stars like the faint companion of Sirius, which is not much larger 

 than the earth, all the way up to Antares, which is so tremendous a 

 star that if the sun were at its center the orbit of the earth and even 

 that of Mars would still be well below its surface. 



