COMPOSITION OF OUR UNIVERSE — BROWN 199 



of many constituents. The complexity of the treatment is snch that 

 even in the most favorable cases the relative numbers of atoms of tAvo 

 elements cannot be determined with a precision which is better than 

 a factor of two. Nevertheless, although the precision is not so great 

 as might be desired, the results possess considerable significance. 



Unsold recently determined the relative abundance of elements in 

 tlie sun's atmosphere as 560 atoms of hydrogen for each atom of 

 ox5^gen, and 0.37 atom of carbon, 0.037 atom of silicon, 0.76 atom of 

 nitrogen, 0.0035 atom of sodium, and 0.062 atom of magnesium for 

 each atom of oxygen, and so on down to 0.000021 atom of vanadium 

 for each atom of oxygen. The significance of these abundances will 

 be apparent later on. 



Whether the composition of a star's atmosphere is representative 

 of the composition of the interior can only be answered directly. 

 Numerous arguments have been presented to favor the conclusion 

 that they are the same — and that they are different. In general, the 

 arguments which favor fairly complete mixing of the elements within 

 the sun appear to be somewhat stronger than the others, particularly 

 with respect to elements heavier than oxygen. This is so in spite 

 of the fact that small traces of lithium and boron, which have been 

 detected in the sun's atmosphere, could not possibly exist for an ap- 

 preciable length of time at the temperatures of the sun's interior. 

 The sun probably sweeps up small traces of these observed elements. 

 Secondly, the amounts of lithiiun and boron observed are so minute, 

 and the region in wdiich they could be consumed is so relatively small, 

 that the amounts observed need not be incompatible with a relaxation 

 time for mixing adequate to result in fair homogeneity. Additional 

 evidence favoring good mixing will be presented when we compare 

 and find similarities in the composition of the sun with that of other 

 material in our solar system. 



A method independent of observed spectral intensities exists for 

 the determination of the abundances of hydrogen and helium relative 

 to other elements in stars. The method depends first upon the gen- 

 eral theory of stellar structure, a fundamental result of which is that 

 for a given mass the radius and luminosity of a star will depend 

 strongly upon the mean molecular weight of the matter of which the 

 star is composed. At the temperatures which exist in stellar interiors 

 atoms are completely ionized, so the mean molecular weight of a given 

 element will be its ordinary molecular weight divided by the total 

 number of particles produced by the ionization (electrons plus nu- 

 cleus). The molecular weights of most completely ionized elements 

 lie very close to 2 because of the fact that in general the mass num- 

 bers of nuclear species are nearly double their atomic numbers. Thus, 

 the mean molecular weight of completely ionized iron will be 



922758—51 14 



