THE SOURCE OF STELLAR ENERGY 
297 
guilty of vain speculation. If our critic possessed in his laboratory similar 
fountains of energy whose output he could measure and whose physical 
state he could calculate, he would not be backward in speculating on the 
processes occurring. 
Unfortunately the facts as yet do not fall into satisfactory order, and 
we are still groping for a clue. I have no particular theory to advocate in 
the following sections and the general result of the arguments is entirely 
inconclusive. But the discussion will bring out the intricacy and difficulty 
of the subject. 
Astronomical Difficulties. 
207. Consider the following comparison of Capella and the sun— 
1. Capella liberates 58 ergs per gram per second compared with 
1-9 liberated by the sun. 
2. The density of the sun is 620 times the density of Capella. 
3. The temperature of the sun at corresponding points is 4-3 times the 
temperature of Capella. 
The first two are immediate results of observation. The third is a conclusion 
from the present theory which it is difficult to distrust. Now it is generally 
believed that the liberation of subatomic energy, if it depends on tempera 
ture and density at all, will increase with temperature and density. Why 
then is there this decreased output in the sun in spite of the apparently 
more favourable conditions ? 
Presumably the answer is that the sun is a much older star, and that 
Capella is drawing on a more prolific source of energy which has become 
exhausted in the sun. We may note incidentally that (judging from the 
output) not more than ^ of this original source remains in the sun, so 
that if the sole supply of energy is the conversion of hydrogen to higher 
elements the hydrogen must now be pretty well used up in the sun. But the 
main point is that the interpretation of the astronomical results is likely 
to be complicated by a third factor in addition to temperature and density, 
viz. exhaustion of the supply. 
Turn now to the two components of Capella. At some epoch a single 
star divided and the two components started life with material in the same 
stage of exhaustion. By the mass-luminosity relation the more massive 
component has radiated more energy per gram and has accordingly suffered 
greater exhaustion. Yet we find now that the more massive component, 
with lower temperature, lower density, and more exhausted, is liberating 
more energy per gram than the fainter. This is a very awkward paradox. 
Several points in this last argument require amplification. It has been 
urged that the massive component will take for its share more of the central 
region of the original star where the heaviest and most intensely active