THE COEFFICIENT OF OPACITY 
245 
/ 
of small mass are older stars, and the hydrogen has been gradually used 
to form heavier elements. 
Hydrogen is the only element which can make these changes; admixture 
of helium would give very little increase of k t /k a . 
Some writers have thought that hydrogen is unable to remain in the 
interior of a star and necessarily rises to the surface. This would be fatal 
to the foregoing suggestion. It seems, however, that there is no such 
separation of the hydrogen (§ 195). 
I was formerly attracted to the view that stars, especially in the giant 
stage, contain a large proportion of hydrogen—the idea being that the 
stars are the main, if not the only, seat of the manufacture of the higher 
elements from protons and electrons, the star’s heat being incidentally 
provided by the process. But the low molecular weight involved is out 
of keeping with the general trend of astronomical evidence. It upsets 
altogether the relation which we have found between the masses of the 
stars and the critical values of 1 — /3. And it leaves room for haphazard 
fluctuations depending on how much hydrogen is left which seems contrary 
to the general uniformity of the mass-luminosity diagram. I would much 
prefer to find some other explanation of the discordance between k t and k a . 
The Theory of Nuclear Capture. 
170. Before Kramers’ theory of electron capture was put forward 
I had proposed a theory of nuclear capture. The interest of this theory is 
that it gives full agreement with astronomical observation. That almost 
automatically brings it into conflict with laboratory experiment, since we 
have seen that the discordance really lies between the two classes of 
observation. A brief account of this theory may be given here, although 
I do not think it can be accepted. 
We return to the apparent target for iron at the centre of Capella 
( 151 ' 94 ) a = 1-20.10- 10 cm., 
and follow up the first idea that this is an actual sphere at the centre of 
the atom. The electron tracks which if undisturbed would have just grazed 
the apparent target will curve towards the nucleus and envelope a much 
smaller true target. 
Since these tracks approach close to the nucleus it is necessary to take 
account of change of mass with velocity. Let an electron of initial mass 
m and velocity V be aimed at the edge of the apparent target so that its 
angular momentum is mVo. Let the pericentron distance be o' and the 
mass and velocity there be m ', V '. Then a' will be the radius of the true 
target.