ANALYSIS OF STARLIGHT — ^PAGEL 



309 



into their lowest energy levels. We shall come back to the cool type 

 spectra later. 



Having seen that there are two ways of estimating the surface 

 temperatures of stars, either from the spectral type or from the color 

 index, we now have to consider how we can derive their true bright- 

 nesses from photometric measurements. To do this, we have of course 

 to allow for their distances away from us, since we would not other- 

 wise know whether a certain star was very luminous and very far 

 away or quite faint but very close to us in space. A typical example 

 of the importance of the distance effect is provided by the well-known 

 stars Vega and Rigel, which both have about the same apparent 

 brightness ; but Rigel is about 30 times as far away from us as Vega, 

 and so its intrinsic brightness must be about 900 times as great. 

 If the distance is less than about 100 light-years, we can measure it by 

 a trigonometrical surveying method using the diameter of the Earth's 

 orbit around the Sun as a baseline ; this is referred to by astronomers 

 as measuring the parallax. If they are farther away but in a cluster, 

 we can estimate the distance of the whole cluster, sometimes rather 

 rouglily, by identifying stars in it whose brightnesses we think we 

 know on the grounds that they are the same breed of animal as 

 some nearby star that has been done by the trigonometric method. 



In this way we can take a group of stars and plot the luminosities 

 against the surface temperatures, as estimated from the color dis- 

 tributions or spectral types ; generally the color indices are preferred 

 as being more accurate. Figure 1 shows such a color-lmninosity 



Surface Temperature 



/ 0,000 7,000* 



^,000 



/OO 



10 



Brightness 

 compared 

 ^ith the I 

 Sun 



9/0 



'/lOO 



..• 



X 







i. 



J,^oo' 



04 0-8 



Co/our Index 



1-2 



Figure 1. — Color-luminosity diagram for the Hyades. 

 720-018 — 64 21 



