' 
24 SURVEY OF THE PROBLEM 
In the laboratory we can produce X rays of the quality occurring in the 
stars, but we cannot produce them with the same intensity. Thus we can 
only perform the experiment of feeding the atoms with quanta at a very 
slow rate. Long before the atom has the chance of a second mouthful it 
has had time to set its trap in order and is ready again. But in the 
stars the X rays fly by so fast that the atoms are gorged and cannot take 
advantage of their abundant chances. The moment the trap is ready it is 
sprung again. Only a small proportion of the absorption traps are in 
working order at any one moment and the absorption coefficient is pro 
portionately reduced. This saturation effect is responsible for the stellar 
coefficient falling below the values derived by extrapolation of Table 3. 
In the main, terrestrial experiments are adapted to measure the atom’s 
activity in catching X rays, and astronomical experiments to measure its 
power of recovery afterwards. The consumption of food by the hungry 
hunter depends on his skill in catching it; the consumption by the 
prosperous citizen depends on the strength of his digestion. 
For stellar investigation we therefore find it best to fix attention not 
on the process of disruption but on the process of repair (electron capture). 
In other words, we turn from absorption coefficients to emission coefficients. 
In the equilibrium condition of a star absorption and emission must 
balance, so that it is indifferent which of them we study. It is only a 
question of following the line of greatest practical simplicity. 
Details will be given later of attempts to calculate the emission or 
absorption at given temperature and density from a purely physical theory 
of the process of electron capture. As happened for molecular weight, the 
chemical composition of the material appears not to have a large influence 
on the result. If we could solve this problem fully we could extend the 
problem set to the cloud-bound physicist in § 14 and ask him to predict 
the heat and light emitted by his globes of given mass and density with 
no assistance from astronomical observation. The theory which seems most 
in keeping with our general physical knowledge leads to predictions pretty 
near the truth, but, I think, not so near as they ought to be. This part of 
our aim is still in the interesting state when we cannot help feeling that 
we are not far off the right track and the true solution is waiting just 
round the corner. 
20 . Jumping as well as we can the difficulties raised by the discordance 
above alluded to, we combine our helping and hindering factors and 
determine the theoretical rate of outflow of radiation from a star. The 
result is the mass-luminosity relation explained and compared with 
observation in Chapter vu. We shall not here anticipate the results which 
follow when at length the theory reaches direct contact with observation. 
The purpose of this preliminary survey is that in entering on the mathe- 
matici 
of the 
factori 
Ot 
varial 
ment 
White 
of a s 
driver 
sub ate 
dispos 
observ 
object 
Th 
essent 
Chapt 
place 
elemei 
probal 
has p: 
quant 
tions. 
ideas 
given 
when, 
close 1 
inner < 
that t 
degree 
possib 
in a b 
Ch 
the in 
where 
mecha 
afterw 
energy 
far aw 
what i 
settlec 
of hoT 
not th 
cross-]