These results are mainly independent of the assumed chemical constitution of the stars, owing to the fact that the average molecular weight for all elements except the lightest turns out to be nearly 2 when account is taken of ionization; but for stars composed mainly of hydrogen the mass-luminosity relation would be greatly altered. It may be inferred, therefore, that the stars do not contain an excessive proportion of hydrogen.
It is interesting that we should be able to learn anything at all about the chemical composition of the inaccessible interior; but the fact has also a definite bearing on theories of the source of a star's heat. One theory suggests that the store of energy needed to maintain the star for thousands of millions of years is provided by the evolution of higher elements from hydrogen. (Hydrogen, which consists merely of protons and electrons is perhaps best re garded as unformed matter.) By cutting down the admissible pro portion of hydrogen to, say, io%, we reduce the supply from this source to one-tenth, and it becomes very doubtful whether it will suffice. Secondly, by admitting that 90% of the star must consist of "formed" matter, we must suppose that the main pro cess of formation of matter occurs in a pre-stellar stage. Evolu tion of the elements is, therefore, in the main a low temperature process and, perhaps still more surprisingly, a low density process.
it is the same in all parts of the interior, and is independent of the star's density; the molecular weight being given, it depends only on the star's mass. It is remarkable that for stellar masses there is no great disparity between radiation-pressure and gas pressure, whereas for masses of a different order would begin .00 . or .99 . It seems clear that the condition of rough equality of the two pressures has in some way determined the size of the aggregations of the primitive matter of the uni verse. The simplest view is to suppose that, when radiation pres sure amounts to say one-tenth of the whole pressure, it renders the aggregation of further material more difficult or makes the mass more liable to break up. Hence larger masses are more and more rare. Gravitation by drawing together the primitive nebulous material will, in general, build up masses to a size at which radia tion pressure first begins to offer serious opposition to further increase. It is now widely believed that the stars radiate away a large part of their mass during their life-time; hence it is fair est to consider only the youngest stars—those in the most diffuse state ; 90% of these have masses (deduced by the mass-luminosity relation) between two and one-half times and five and one-half times that of the sun, corresponding to values of I-13 between .17 and .35. If the rise of radiation to importance plays a part in limiting the masses of the stars, this is just the range which we should expect to be most critical.