206 ANNUAL REPORT SMITHSONIAN INSTITUTION, 1931 



Sodium vapor, for example, when cool, absorbs only the familiar 

 l)air of lines in the yellow, and others in the ultra-violet, which are 

 produced by transitions from the normal state. But at a white heat, 

 a minute fraction (perhaps 1 in 100,000) of the atoms are in an 

 exited state and absorb other pairs of lines in the red, orange, and 

 green. In the electric furnace those lines are very feebly absorbed, 

 but in the far hotter atmosphere of the sun they are stronger in pro- 

 portion to the great principal pair in the j^ellow. 



Similar phenomena occur in the more complicated spectra, such 

 as those of titanium and iron, where there are dozens of different 

 exited states. The behavior of the lines at different laboratory tem- 

 peratures has been of great importance in working out the intricate 

 problem of their analysis. 



Once more the sun-spot spectrum confirms the theor}^ perfectly. 

 Lines absorbed by unexcited or slightly excited atoms are strength- 

 ened most in the spots and those of excited neutral atoms less so. For 

 elements of easy ionization the increase in the number of neutral 

 atoms in the spots is so great that it swamps the loss due to dimin- 

 ished excitation; but for elements harder to ionize, such as silicon 

 and zinc, the reverse is the case, and the lines of high excitation 

 potential are weakened in spots. 



The same principles apply to enhanced lines. As might be ex- 

 pected, enhanced lines of high excitation are greatly weakened in 

 spots, and often obliterated. 



By a detailed study of the behavior of lines of all these sorts it 

 is possible to find both the temperature and pressure in the spots. 

 Miss Moore (of Mount Wilson and Princeton) has completed a very 

 successful study in which she finds that the temperature of a typical 

 spot is 1,000° lower than that of the disk, the total pressure 70 per 

 cent higher (since the gas is more transparent and we can see deeper), 

 but the pressure due to the free electrons 40 per cent less, owing to 

 the diminished ionization. The agreement of theory and observation 

 is satisfactory to the finest details. 



Even at the solar temperature, 5,740°, only about 1 atom in 

 20,000 should be excited to the extent measured by 5 volts (that is, 

 by the energy imparted to an electron by 5 volts' potential drop). 

 It is easy to understand, then, why only 46 solar lines out of nearly 

 6,000 for which the energy relations are known are absorbed by 

 atoms more highly excited than this. 



Almost all the metals (except gold and mercury) have arc lines 

 of low excitation in the accessible part of the spectrum, so that 

 they are, so to speak, on an even footing in our solar lists. Some 

 of them have similar enhanced lines accessible, (as do calcium, 

 scandium, titanium, and the rare earths) ; but for many others, such 



