134 
The Source of Stellar Energy [ch. iv 
of the sun and stars. The only possible reservation to this statement arises 
in connection with the radioactive elements. We have, however, seen that 
even if the sun were built of pure uranium, its radiating power would be only 
about one-half of that observed, and would only last for a minute fraction of 
what is believed to have been the sun’s life. A sun of pure radium would 
radiate more than enough for the moment, but its life would be limited to a 
few thousand years. No possible combination of terrestrial elements can give 
the combination of high radiation and of staying power which is observed in 
the sun and stars. 
According to practically all theories of the origin of the solar system, the 
earth and the other planets must have been formed out of the body of the 
sun at some epoch which we may date at approximately 1500 million years 
ago. The question then arises why the atoms of the sun, which are generating 
energy at an average rate of some 1'9 ergs a second, should be of different 
type from those of the earth, which are certainly generating less than O’OOOl 
ergs per second. 
If they are of different type now, they must have been of different type 
when the earth was born out of the sun. The sun’s present rate loss of mass 
by radiation indicates that the present average life of solar atoms is of the 
order of 15 x 10 12 years. Thus the chemical composition of the sun, or of a 
fair sample of solar atoms, can hardly have changed appreciably in the 15 x 10 8 
years of the earth’s existence. The earth’s atoms cannot have formed a fair 
sample of the sun’s atoms when the earth was born, otherwise they would 
still do so, and the phenomenon of energy-generation shews that they do not. 
Thus we must suppose that the atoms in the outer layers of the sun, 
from which the earth was formed, do not constitute a fair sample of the sun 
as a whole. This supposition can hardly be considered unreasonable, any 
more than the supposition that the earth’s atmosphere is not a fair sample 
of the earth as a whole. Whatever mixture of elements may constitute the 
sun, the heavier elements are likely to sink to its far interior*, so that the 
atoms in its outer layers are selected for their light atomic weight. If for 
simplicity we regard the sun as containing only two kinds of atoms—light 
elements which rise to its surface, and heavy elements which sink into its far 
interior—then the light elements, being similar to those found on earth, have 
practically no capacity for generating energy, so that the sun’s generation 
of energy must originate in the heavy elements which reside in its interior 
regions. But if, as we have supposed (§ 120), ionisation of stellar atoms par 
tially or wholly inhibits their generation of energy, then the problem of the 
distribution of this energy generation becomes very intricate. 
123. More generally we may suppose that the matter of the sun and stars 
consists in its earliest state of a mixture of elements of different atomic 
* See § 87 (p. 93) above.