sun's place among the stars — ADAMS 145 



temperature of a star is very quiclily recognized is from the analysis 

 of its light into a spectrum. The distribution of the light of dif- 

 ferent colors throughout the spectrum, the presence or absence of 

 certain lines, and many other features determine the temperature of 

 the star rather definitely (pi. 4). So stars of the same spectral type 

 have closely the same temperature. Now, it is a fact of observation 

 that the masses of stars of the same spectral type do not differ very 

 greatly. A factor of 10 would cover the vast majority of cases. We 

 know this from a great variety of evidence, mainly from double 

 stars for which the masses can be calculated accurately. On the 

 other hand, we know that the luminosity or candlepower of stars 

 shows enormous variations, stars of the same spectral type some- 

 times differing as much as hundreds of millions of times in the 

 amount of light they give out. 



Since the surface brightness, or the amount of light each unit of 

 area gives out, is nearly the same for stars of the same temperature 

 or spectral type the only explanation for the immense difference in 

 luminosity is a great difference in size. In other words, the very 

 luminous stars must be very large as compared with the fainter 

 stars. Since the masses, however, do not differ very greatly, the 

 brighter stars must be very much less dense than the fainter stars. 

 We conclude, therefore, that many of the brighter stars in the sky 

 must be enormous masses of gas of very low density, a conclusion 

 fully borne out by measurements of diameter with Michelson's inter- 

 ferometer, which show the existence of great red stars as much as 

 200,000,000 miles in diameter, or more than 220 times the diameter 

 of our sun. (Fig. 3.) 



Such stars are recognized through a study of their spectrum. 

 When a star gives out light, the atoms of the gases in its atmosphere 

 are in an excited state and absorb light in the particular wave lengths 

 that correspond to the spectrum lines of each element involved. So 

 we get a pattern of lines of all the elements in the star's atmosphere. 

 When the temperature of the star is low, we obtain the lines of 

 what is called the neutral atom, the atom in its normal state, but 

 if the temperature is high the atom is modified by having one or 

 more of its electrons pulled off and becomes what we call ionized. 

 Ionized atoms give rise to a different class of lines from neutral atoms, 

 but in stars of ordinary temperature both sets of lines are usually 

 present, some of the atoms being neutral and some ionized. In 

 stars of low temperature the neutral lines are the stronger, in stars 

 of high temperature the lines due to the ionized atom. 



There is, however, another factor that favors the detachment 

 of electrons from atoms, and this is the density of the star's atmos- 

 phere. If the density is low, there are fewer collisions and fewer 



