COMPOSITION OF THE SUN KUSSELL, 207 



as iron and nickel, the " ultimate " lines of low excitation are beyond 

 the observable limit in the ultra-violet, and only lines of considerable 

 excitation are accessible. For magnesium and aluminum the only 

 accessible enhanced lines are of very high excitation and are natu- 

 rally faint, while for the accessible lines of the alkali metals the 

 excitation is so high as to banish all hope of finding them. 



Passing to the nonmetals — which are so hard to ionize that only 

 arc lines could be expected in the sun — we find that the ultimate 

 lines are in all cases hopelessly out of reach and the accessible lines 

 have excitation potentials ranging from 6 volts upward. Our spec- 

 troscopic tests for these elements are therefore at best perhaps a 

 hundred-thousandth part as sensitive as they would be if we could 

 observe the far ultra-violet, and the wonder is not that we fail to 

 find so]ne of them but that we do find any. Carbon, sulphur, and 

 nitrogen, though they show only faint lines, must be as abundant as 

 the commoner metals. Oxygen, whose lines are stronger, must be 

 still more so. But hydrogen is really amazing. Despite the very 

 high excitation potential of 10.15 volts, its lines are among the 

 strongest in the solar spectrum. It may be that a large allowance 

 should be made for the peculiar character of its lines, which are 

 exceptionally susceptible to widening by the electric fields of neigh- 

 boring ions, and hence to apparent strengthening. But, even so, there 

 appears to be no escape from the conclusion that hydrogen is far 

 more abundant in the sun's atmosphere than any other element — 

 much more so, indeed, than all the others together. 



For the elements which do not appear in the solar spectrum, the 

 excitation potentials of the best available lines are in most cases 

 hopelessly high. There may be some chance of finding phosphorus 

 when its spectrum has been investigated in the infra-red, but this is 

 all. 



An unexpected aid in our search has, however, recently appeared. 

 More and more band spectra are being identified in the sun; that 

 is, more species of molecules detected in the atmosphere. Most of 

 these are compounds of hydrogen, oxygen, or carbon, but boron 

 oxide (BO) was detected a few years ago in sun spots by Nicholson 

 and Perrakis, and silicon fluoride (SiF) more recently by Richard- 

 son, both in the spots and on the disk — adding two more elements to 

 the list of those positively identified in the sun. The other com- 

 pounds so far identified (some in spots only) are the hydrides CH, 

 NH, OH, MgH, AlH, SiH, CaH, the oxides AlO, TiO, ZrO, and 

 the carbon compounds CN and Co. It is possible but not certain that 

 molecules of hydrogen (H^) may exist in sun spots. 



The chemist will gaze with open eyes at this list. All but the last 

 are mere fragments of molecules, already partly decomposed and too 

 active chemically to be isolated in ordinary temperatures. But it 



