Supplement to ‘ Nature,” May 12, 1923 xi 

about, meeting other atoms ; after a time a burst atom 
meets a loose electron under suitable conditions and 
induces it to stay and heal the breach. The atom is 
now repaired and ready for another mouthful as soon 
as it gets the chance. 
From this cause a big difference arises between 
absorption of X-rays in the laboratory and in the stars. 
In the laboratory the atoms are fed very slowly ; the 
X-ray bundles which they feed on can be produced by 
us only in small quantities. Long before the atom has 
the chance of a second bite it is repaired and ready 
for it. But in the stars the intensity of the X-rays is 
enormous ; the atoms are gorged and cannot take 
advantage of their abundant chances. The consump- 
tion of food by the hungry hunter is limited by his 
skill in trapping it ; the consumption by the prosperous 
profiteer is limited by the strength of his digestion. 
Laboratory experiments test the atom’s skill in catch- 
ing food ; stellar experiments test how quickly it re- 
covers from a meal and is ready for another. That is 
why the absorption follows a different law in the two 
cases. 
CapTuRE OF ELECTRONS. 
To predict:the stellar absorption-coefficient we must 
accordingly fix attention on the rate of repair of the 
burst atoms. The atom is wandering about advertis- 
_ ing a vacancy for an electron, and numbers of ejected 
electrons are rushing about on holiday. Many elec- 
trons will come up, look at the situation, and go off 
again. How is the atom to trap the electron into 
taking up the situation ? I will give you the solution 
of this problem which I am inclined to think fairly 
probable, though I have not found many who agree 
with me. We may compare the electron to a stray 
planet entering the solar system from outside, bearing 
in mind, however, that the planets (satellite electrons) 
must be supposed to repel the invader, and the sun 
(positive nucleus) attracts it. Dynamics teaches us 
that, provided no actual material collision occurs, the 
intruder will scarcely ever be captured, but after 
stirring up things a little will retreat again towards 
infinity. There are exceptions, as when the sun and 
Jupiter conspire to capture a comet, but these would 
be very rare in the conditions corresponding to an 
atom. In some cases the intruder would turn the 
tables by carrying off a regular planet, thus compen- 
sating for the occasions when it was itself captured. 
Probably, as regards repair of the atom, as much harm 
as good would be done on the average. 
More delicate persuasion being of no avail, there 
seems nothing left but for the atom to secure its 
electron by brute obstruction. For this reason I take 
the view that usually the capture of an electron occurs 
through its running against the positive nucleus of the 
atom. This nucleus has a highly complicated struc- 
ture, the iron nucleus, for example, consisting of 86 
distinct charges arranged in some kind of equilibrium. 
If by accident an electron runs full tilt into this packed 
mass, it will agitate it and lose energy in so doing ; 
it will rebound, no doubt, but with smaller velocity 
insufficient to carry it out of the sphere of attraction 
of the atom.’ By a process of exclusion this seems 
the only method consistent with dynamical laws by 
which the atom can secure the electron needed for its 
repair. Therefore I have concluded that the actual 
electron trap is none other than the positive nucleus— 
a region at the centre of the atom known to be about 
10-12 cm. in radius. It must be remembered that 
the nucleus attracts the electrons and will sweep 
into the trap many which were not initially aimed 
at it. 
This theory has been adversely critised mainly on 
the ground that it is entirely accordant with the laws 
of dynamics. At first sight that might not seem a 
grave objection ; but we have got so used to the atom 
behaving in a way which violates the classical laws, 
that any theory which does not violate them is liable 
to be viewed with suspicion. While admitting that 
there are uncertain possibilities in the mysterious 
region in the interior of an atom, we must note that 
the present problem belongs to a class of investigations 
in which the usual dynamical laws are applied by 
physicists, often with much success. It concerns the 
motion of a free electron—not yet forming part of 
any permanent quantised system—a problem which 
occurs in the theory of conduction of electricity in 
metals, in thermionic phenomena, and in the scattering 
of a- and f-particles. In these problems physicists 
are accustomed to assume (rightly or wrongly) that 
the classical laws of dynamics are observed, and we 
have only followed their (good or bad) example. In 
particular in Rutherford’s experiments on scattering, 
the classical laws of force are found to hold good almost 
to the boundary of the nucleus itself. There seems 
to be a fair presumptive evidence that our stellar 
problem should be attacked in the same way; 
although we admit that unknown circumstances may 
intervene.* 
The strong point in our favour is that this theory 
? The kinetic energy at the moment of collision with the nucleus is 
enormously greater than the kinetic energy before entering its sphere of 
attraction; so that a very small proportionate change of kinetic energy by 
collision would wipe out the original energy of the electron. The imperfect 
elasticity of the collision is a dynamical consequence of the complex structure 
of the nucleus. A collision of two simple charges may be perfectly elastic, 
except that that would apparently prevent a hydrogen nucleus from ever 
recovering its electron. 
* While the fast-moving particles undoubtedly penetrate the atom in 
the way we have assumed, it is held by some that slow-moving electrons 
(as in the stars) are turned back at the surface. The idea seems to have 
originated at a time when the positive charge of the atom was thought to 
be a large sphere coextensive with it; and it seems out of keeping with 
modern views. It is ignored in current theories of conduction of electricity, 
Even if it were conceivable that a neutral atom could so ward off an electron, 
the strongly positive atoms in the stars could scarcely exclude it. 
