116 
The Source of Stellar Energy [ch. iy 
Yet the consideration which will carry conviction to most minds that 
stellar radiation arises out of the annihilation of stellar matter is neither the 
fact of its being required by the laws of physics, in so far as we understand 
them, nor the fact of its having seemed to Newton two centuries ago, to be 
“very conformable to the course of Nature,” but the general evidence of 
astronomy that young stars are uniformly more massive than old stars. The 
evidence for this will appear as our book proceeds. 
The Conditions of Liberation of Stellar Energy. 
104. Although all available evidence points to annihilation of matter as 
being the source of stellar energy, we shall not find it necessary to assume 
this to be the source in the following investigation, or indeed in most of the 
investigations or discussions which follow. 
Whatever the source of stellar energy may be, we wish to know in what 
way the liberation of energy takes place, and why it is proceeding so much 
more rapidly in some stars than others. Why should each gramme of the 
sun’s mass liberate on the average 1*9 ergs per second, and why should each 
gramme of Y Puppis average 1100 ergs per second? If, for instance, radiation 
is produced by the annihilation of matter, why is the sun transforming its 
matter into radiation so much more slowly than Y Puppis ? And, if it is a 
general property of matter to transform itself into radiation, why is this 
process so little in evidence on the earth ? 
If the earth generated no energy at all, its surface would assume a steady 
average temperature such that the energy radiated away by its area of 
5T x 10 13 square centimetres was exactly equal to the energy it received from 
the sun. This latter amount of energy is l - 7 x 10 24 ergs per second, and the 
surface-temperature requisite to radiate this away is readily calculated to be 
about 276° abs. or 3° C. If the earth’s surface is not treated as a perfect 
radiator, the form of argument is slightly different, but the final result is 
precisely the same. Suppose, however, that the earth’s mass generated energy 
at an average rate even of a ten-thousandth part of an erg per gramme per 
second. A total internal generation of energy of 0-6 x 10 24 ergs per second 
would now have to be added to the 17 x 10 24 ergs received from the sun, and 
to radiate away the total of 2 3 x 10 24 ergs per second would need a surface 
temperature of 29° C., or higher in so far as the earth’s surface is not a 
perfect reflector. This is far above the average temperature of the earth’s 
surface, so that we may be confident that if the earth’s mass is generating 
energy at all, it does not do so at a rate approaching erg per gramme 
per second. Thus the generation of energy which is proceeding in the sun 
and stars is almost or completely absent from the earth. Whatever causes 
are liberating energy in the sun are out of action on earth. 
This leads us to inquire what determines the rate at which a star 
liberates its energy. It might at first sight seem a quite hopeless task to try