THE SOURCE OF STELLAR ENERGY 
295 
and conductivity of the surface rocks. We now know that this escaping 
heat is no more than can be reasonably attributed to the release of energy 
by the radio-active minerals within, so that it is no measure of the rate of 
cooling of the earth. Kelvin, however, supposed that it represented the 
cooling, and calculated backwards to the time when the earth must have 
been molten. This direct method of setting a limit to the age of the earth’s 
crust is no longer tenable. 
But it is not much use extending the age of the earth without extending 
the age of the sun, and here radio-activity helps very little. It is calculated 
that if the sun were composed entirely of uranium and its products (in 
equilibrium proportions) the radio-activity would supply only half the 
sun’s actual rate of radiation; the other half must come from unknown 
sources. 
With regard to the application of these ideas to astronomy there was 
no temptation to formulate any definite theory so long as the only demand 
was for an extended time-scale. Some may have speculated on the exist 
ence of elements of more potent radio-activity in the stars, uranium and 
radium on the earth being the last feeble remnants of an expiring process. 
Others were content to know that an ample store of energy existed in the 
stars whatever their composition, and there was no pressing occasion to 
decide in which of the possible ways it was being released. 
A more intimate contact with the problem seemed imminent in the 
first researches on the radiative equilibrium of the stars in 1916. Then we 
were faced with the question in what manner the source of stellar energy 
is distributed through the interior. If the source were gravitational energy 
of contraction the distribution could be settled. But the contraction 
hypothesis was already becoming obsolete so that allowance had to be 
made for the source being probably subatomic. But we have shown that 
it is not necessary in an approximate solution of the problem to make very 
precise assumptions as to the distribution of the source. Thus the question 
of the law of release of subatomic energy was shelved for a time. 
Reference may here be made to. a very interesting discussion by 
H. N. Russell* in 1919 which indicated some of the conditions which a 
subatomic source must fulfil in order to satisfy astronomical requirements. 
This was perhaps the first sign of serious astronomical interest in the subject 
apart from its bearing on the time-scale. 
In 1920 the researches of Aston, establishing the loss of mass occurring 
when the higher elements are formed from hydrogen, gave a new interest 
to the subject. It provided a much more powerful source of energy than 
any known radio-active change. Annihilation of protons and electrons, 
or the disintegration of unknown elements of intense radio-activity are 
* Pub. Astr. Soc. Pac. 31, p. 205 (1919). See also Eddington, Observatory, 42, 
p. 371 (1919).