COMPOSITION OF THE SUN — RUSSELL, 213 



Among- the rare elements there is a general parallelism between 

 solar and terrestrial abnndance. Some elements, of wliich the most 

 noteworthy is scandium, appear to be much more abundant in the sun 

 than here ; but small quantities of these, widelj^ disseminated among 

 the rocks, may well have escaped the search of the ordinary chemist, 

 even though skilled. The actual amounts even of the rare elements, 

 in the sun's atmosphere, are very great. For example, platinum is 

 represented only by three faint lines in the solar spectrum; but to 

 produce these there must be something like 500 million tons of the 

 precious metal above the photosphere. This amounts, however, to 

 less than 8 ounces per square mile of surface, or one-eightieth of an 

 ounce per acre. An equal amount, rained down in thin dust on the 

 earth's surface, would not be worth the labor of sweeping up. 



The amounts of the nonmetallic elements are much harder to 

 determine in the sun, for the onl}?- available lines come from atoms 

 in highly excited states. For every atom in such a state there are a 

 very great number in the normal state — sometimes millions — and 

 these huge factors are hard to determine accurately. As has already 

 been said, carbon, sulphur, and nitrogen must be as abundant as the 

 commoner metals, oxygen more so, and h3^drogen far more abundant 

 than all the rest. 



In the hotter stars, where lines of ionized oxygen, nitrogen, etc., 

 and of neutral helium, are conspicuous, a better comparison can be 

 made. Miss Payne, of Harvard, working a few years ago by meth- 

 ods developed by Milne, derived values for the relative abundance 

 of a number of elements and reached the very important conclusion 

 that the stars are remarkably similar in composition. The great 

 differences in their spectra arise, not from differences in the abun- 

 dance of the elements, but in their ionization and excitation at dif- 

 ferent temperatures. Her results, and those of the writer's later 

 work, are in remarkably good agreement. 



Combining them in a final summary, we may conclude that at 

 least 90 per cent of all the atoms in the sun's atmosphere — and per- 

 l\aps 95 per cent or more — are atoms of hydrogen. Of the remainder, 

 helium and oxygen contribute some two-thirds, and all the metals, 

 together with carbon, sulphur, etc., the rest. 



These are the proportions of the various constituents by volume 

 in the gas. By weight, the metals, whose atoms are much heavier, 

 make up a quarter of the whole (or less, with the higher estimates 

 of the abundance of hydrogen). 



Whether the sun's interior is of the same composition is a harder 

 question. The great ascending cyclonic whirls which produce sun 

 spots must come from a considerable depth, but from only a very 

 small fraction of the way to the center. Whether the deeper layers 

 are sufficiently stirred by currents to undo the very small tendency 



