On the Composition of the Sun's Atmosphere

The energy of binding of an electron in different quantum states by neutral and singly ionized atoms is discussed with the aid of tables of the data at present available. The structure of the spectra is next considered, and tables of the ionization potentials and the most persistent lines are given. The presence and abseice of the lines of different elements in the solar spectrum are then simply explained. The excitation potential, E, for the strongest lines in the observable part of the spectrum is the main factor. Almost all the elements for which this is small show in the sun. There are very few solar lines for which E exceeds 5 volts; the oniy strong ones are those of hydrogen. The abundance of the various elements in the solar atmosphere is calculated with the aid of the calibration of Rowland `s scale developed last year and of Unsold's studies of certain important lines. The numbers of atoms in the more important energy states for each element are thus determined and found to decrease with increasing excitation, but a little more slowly than demanded by thermodynamic considerations. The level of ionizatioii in the solar atmosphere is such that atoms of ionization poten- ~ hal 8.3 volts are 50 per cent ionized. Tables are given of the relative abundance of fifty-six elements and six compounds. These show that six of the metallic elements, Na, Mg, Si, K, Ca, and Fe, contribute 95 per cent of the whole mass. The whole number of metallic atoms above a square centi- meter of the surface is 8 X ic20. Eighty per cent of these are ionized. Their mean atomic weight is 32 and their total mass 42 mg/cm2. The well-known difference between ele- ments of even and odd atomic number is conspicuous-the former averaging ten times as abundant as the latter. The heavy metals, from Ba onward, are but little less abundant than those which follow Sr, and the hypothesis that the heaviest atoms~ sink below the photospifere is not confirmed. Tue metals from Na to Zn, inclusive, are far more common than the rest. The compounds are present in but small amounts, cyanogen being rarer than scandium. Most of those elements which do not appear in the solar spectrum should not show observable lines unless their abundance is much greater than is at all probable. There is a chance of finding faint lines of some additional rare earths and heavy metals, and perhaps of boron and phosphorus. The abundance of the non-metals, and especially of hydrogen, is difficult to estimate from the few lines which are available. Oxygen appears to be about as abundant by weight as all the metals together. The abundance of hydrogen may be found with the aid of Menzel's observations of the flash spectrum. It is finally estimated that the solar atmosphere contains 6o parts of hydrogen (by volume), 2 of hdium, 2 of oxygen, i of metallic vapors, and o.8 of free electrons, practically all of which come from ionization of the metals. This great abundance of hydrogen helps to explain a number of previously puzzling astrophysical facts. The temperature of the reversing layer is finally estimated at 5600° and the pressure at its base as 0.005 atm. A letter from Professor Eddington suggesting that the departure from the thermo- dynamic equilibrium noticed by Adams and the writer is due to a deficiency of the num- ber of atoms in the higher excited states is quoted and discussed