Marcu 24, 1923] 
NATURE 
401 

before development by deep red light, then develop- 
ing for a short time and removing the undeveloped 
halide. On the plate there are then left small dots, 
and comparisons with the first plate showed these 
to correspond with the silver halide grains originally 
present. 
Svedberg has shown that the number of centres 
produced in this way by initial development increases 
with the exposure in accordance with the usual photo- 
graphic laws, and it might be assumed that the dis- 
covery of these centres produced during development 
is a proof of discreteness in the action of light upon 
the grain, and that they must result from a structure 
in the silver bromide grains existing either before 
exposure or produced during exposure, and corre- 
sponding, for example, to spots of sensitiveness. 
While the evidence for this seems very great, it must 
be remembered that we know nothing about these 
centres until development takes place, and that even 
if the whole grain were equally affected by the action of 
light and changed to the same extent, we should still 
expect development to take place first at some local 

Fic. 4. 
spot corresponding to slight surface differences in the 
grain. A sheet of metal immersed in acid, for example, 
will not be attacked uniformly all over the surface. 
Because of the impurities, action will start at in- 
dividual points. 
In a very important paper, Toy has given measure- 
ments showing that the number of nuclei produced 
on initial development are proportional to the number 
of grains which become developable on complete 
development, and that the larger grains not only 
have more nuclei on account of their size, but that 
these nuclei are also more sensitive to light than those 
in the smaller grains, the sensitivity of a grain being 
the sensitivity of its most sensitive nucleus. Svedberg 
considers that the number of developable centres 
per unit area of grain surface is a measure of the light 
sensitivity of the silver halide of the emulsion. From 
Toy’s work it would seem to be doubtful whether we 
can speak of the light sensitivity of the halide itself 
in terms of the nucleus theory, since this will vary 
with the size of the grain. 
Recently, a number of phenomena have been observed 
which are very difficult to explain by the use of the 
classical wave theory of light, and it seems not unlikely 
that it may be necessary to turn to a theory having 
some analogy to the corpuscular theory. As a first 
NO. 2786, VOL. 111] 
step towards this, Max Planck suggested his now 
well-known quantum hypothesis, according to which 
an atom radiating energy liberates it in discrete quanta, 
the amount of energy corresponding to each quantum 
being a constant multiplied by the frequency of the 
light. Bohr adopted Rutherford’s theory of the 
structure of the atom, considering the atom to consist 
of a nucleus containing an electron carrying a positive 
charge of electricity, and to be surrounded by one or 
more electrons carrying a negative charge, the electrons 
revolving about the positive nucleus itself. He 
imagined that the electrons revolve without radiating, 
but that when an electron suffers some violent shock 
it gives up energy, and this energy is radiated and has 
the value of Planck’s quantum. Thus, if an electron, 
by the sudden impact of another electron, for example, 
is thrown out of an atom and is attracted back to its 
place by the nucleus, then, as it falls back, it will 
send out a pulse of energy, and it will be seen at once 
that, if light is produced by such a behaviour of elec- 
trons, it is inherently probable that it will be radiated 
in pulses rather than continuously. Since, according 
to Bohr, the frequency of the vibration emitted is 
exactly proportional to the energy which the electron 
releases, Planck’s quantum condition is fulfilled, and 
we have the famous equation, 
Ve=hv, 
where V is the voltage acting on the electron charge e, 
v is the frequency, and h is Planck’s constant. 
In an X-ray tube the discharge of electricity is 
in the form of a stream of corpuscles travelling with a 
very high velocity, which depends upon the voltage 
of the electric current applied to the tube. When 
these corpuscles strike the target their energy is 
radiated in the form of X-rays, and we know that 
these X-rays partake very closely of the nature of 
light, except that the length of the waves is about 
one-thousandth of those of light, or, what is the same 
thing, their frequency is a thousand times as great. 
It is to this that they owe their great penetrating power. 
On the classical wave theory of light, then, we 
should imagine that an X-ray tube having its target 
bombarded by the stream of corpuscles produced by 
the current would emit waves of X-rays spreading 
into space, just as waves of light are imagined to 
spread from a source ; but now comes a great difficulty. 
When these X-ray waves travelling out pass through 
a gas and are absorbed, they cause the molecules of 
the gas to emit electrons, and these electrons are 
emitted with almost exactly the same velocity as the 
electrons in the tube which produced the X-rays 
themselves. The extraordinary nature of this phengme- 
non is well illustrated by Sir William Bragg in a recent 
article. He takes as an analogy the dropping of a 
log of wood into the sea from a height of one hundred 
feet. A wave radiates away from where it falls ; the 
wave spreads ; its energy is more and more widely 
distributed ; the ripples die away ; at a short distance, 
a few hundred yards, perhaps, the effect will apparently 
disappear. If the water were perfectly free from 
viscosity, and there were no other causes to fritter 
away the energy of the waves, they would travel 
indefinitely, always diminishing in their height. Now, 
at some point, say a thousand miles away, these now 
