208 ANNUAL EEPORT SMITHSONIAN INSTITUTION, 1931 



is not surprising that such partial decomposition occurs at the sun's 

 temperature. It is probable that a few other more familiar mole- 

 cules (notably CO and Nz) exist in the sun's atmosphere, but these 

 molecules, in their normal state, absorb only in the far ultra-violet 

 and can not therefore be detected. 



The hydrides are known to be rather easy to dissociate, and their 

 presence in the sun is additional evidence of the great abundance of 

 hydrogen. All the other compounds which have been detected con- 

 tain at least one constituent which is known to be very abundant in 

 the sun's atmosphere in the free state. 



Last, but not least, comes helium, which shows strong emission 

 lines in the spectrum of the sun's outer atmosphere — the chromo- 

 sphere — but no dark lines on the disk, except occasionally in dis- 

 turbed regions near spots. The emission lines were recognized at 

 solar eclipses 60 years ago, and for 20 years could not be matched 

 in the laboratory, so that the name " helium " was coined to describe 

 the unlaiown solar element. How it was detected in radioactive 

 minerals, found to be a light inert gas — as had been suspected from 

 the height to which it rises in the chromosphere — and finally dis- 

 covered in natural gas in such abundance that it is used in airships, 

 is one of the romances of science which can barely be mentioned here. 

 It is still the outstanding puzzle of solar physics. Its visible lines 

 demand very high excitation — 20 volts or more — so that it is not 

 surprising that they do not show in absorption, but very remarkable 

 that they appear so strongly in emission at the sun's limb. In fact, 

 one line of ionized helium (A4686) with the enormous excitation po- 

 tential of 48 volts appears faintly in the chromosphere; why, no 

 one yet knows. 



All told, 61 of the 90 known chemical elements have so far been 

 identified, positively or with some margin of doubt, in the solar 

 spectrum. Of the remaining 29, 13 have lines so unfavorably placed 

 that they could not well be expected to appear, while for 12 more 

 the spectra have not been accurately enough measured to permit 

 of a decisive test. This leaves four elements — holmium, rhenium, 

 bismuth, and radium — for which the strongest lines are in accessible 

 regions of the spectrum, and do not appear in the sun. These metals 

 must be present only in minute proportions, if at all, in the solar 

 atmosphere. Holmium is one of the rarest of the rare earths. Khen- 

 ium appears to be exceedingly rare on earth, and radium, on account 

 of its short life, must be rare anywhere in the universe, so that we 

 could not expect to find it or any of the other strongly radioactive 

 elements in perceptible amounts. 



So much for the qualitative analysis of the sun or, at least, of 

 its outer layers. What can we do to make our analysis quantitative? 



