262 ANNUAL REPORT SMITHSONIAN INSTITUTION, 19 31 



stars 10 times as massive as the sun, and nearly all have at least one- 

 tenth its mass. 



Now the true brightness of a star clearly depends on how big 

 it is, but it also depends just as much on how bright a square inch 

 of its surface is. And the brightness of each square inch is almost 

 entirely a question of how hot it is. As a piece of iron is heated in 

 a blacksmith's forge it gives out no visible light at first, but in time 

 it begins to glow with a feeble reddish light. As the temperature 

 increases two things happen. It gives out more light from each 

 square inch, and the quality of the light changes from reddish to 

 yellowish, and finally at very high temperatures is intensely bright 

 and looks nearly white. These changes always happen together. It 

 is impossible to heat a poker until it is very bright and to have it still 

 look red. And it is equally impossible for a feebly shining poker to 

 look white. Now the astronomer, imprisoned as he is upon the earth, 

 far from his celestial experiment, is quick to take advantage of such 

 a reliable connection between the surface brightness, the surface 

 temperature, and the color of an object. 



In order to find the temperature of a star it is only necessary to 

 measure its color. This is a straightforward process if a prism is 

 used to break up the colors. Then by means of a photographic plate, 

 or, still better, a radiometer, thermocouple, or photoelectric cell, the 

 relative strengths of the red and the blue are measured. The propor- 

 tion of blue light to red light increases as the temperature rises, and 

 if a star is found to have a certain proportion of blue to red light its 

 temperature can easily be inferred. It is worth while to note that 

 we have here a means of finding the temperature of a star without 

 knowing anything at all about its distance. For the color of an 

 object ordinarily depends in no way on whether it is close or far 

 away. 



As a matter of fact there are two exceptions to this rule. The set- 

 ting sun looks red when the air is filled with smoke from a forest fire 

 because the short waves of blue light are scattered by the tiny 

 particles of smoke, while the longer red waves pass through to our 

 eyes with little interference. The same thing may happen to star- 

 light. It has been shown recently that some stars, particularly those 

 at great distances, look redder than others, which in every other 

 respect appear to be the same kind of star. This is taken as evidence 

 for the existence of clouds of very small particles of material in ihe 

 otherwise empty spaces between us and some of the distant stars. If 

 such clouds of material are at all uniformly distributed through 

 space, stars will look redder the farther away they are. 



The only other known case in which the color of objects is in- 

 fluenced by their distance is that of the extremely distant nebulae, 



