402 
NATURE 
[Marcu 24, 1923 

microscopic ripples encounter a wooden ship. We 
should expect that they would produce no effect, 
especially as they may have passed many other ships 
without having affected them, but, for some reason, 
as these tiny ripples reach the ship, a plank of the 
same weight as the log is hurled out of the ship to a 
height of exactly one hundred feet, and the whole 
energy which was originally supplied by the log falling 
into the water is concentrated upon the ejection of 
the plank. It will be seen at once how inadequate 
the wave theory is to account for this phenomenon. 
Similar difficulties occur in connexion with photo- 
electricity or the liberation of electrons under the 
influence of light. 
The method by which a photographic emulsion 
adds up light during a long exposure has always been 
a great problem when it is considered from the point 
of view of the classical wave theory. If we accept 
the idea that the grains of silver halide in an emulsion 
are exposed to a continuous flood of light from a distant 
star, for example, then each grain must be imagined 
to be integrating light until it has received enough 
to make it developable. Since the exposure required 
in astronomical photography is frequently very long, 
we must consider that the grains continue to integrate 
the light for many hours, and it is difficult to imagine 
any mechanism which would enable them to do this. 
The difficulty is enhanced by the fact that even a 
very brief exposure continues to produce an effect 
after an interruption of a long period, so that if all 
the grains have been affected by the first exposure, 
they must be capable of storing energy quite in- 
sufficient to make them developable and to hold this 
energy for a long period, and then resume its accumula- 
tion at the level where the interruption occurs. In 
the same way, when we study the exposure of the 
individual grains, even if we could imagine some 
mechanism by which the grains could store up the 
energy falling upon them until they became develop- 
able, we should expect that all the grains of the same 
size would become developable at the same time, 
unless, indeed, we assume the process of exposure to 
be autocatalytic in nature. When grains are examined 
under the microscope, however, some of them are 
found to have been affected before others. If we 
imagine that they all have become exposed to a uniform 
flood of light, we must consider that these grains differ 
in sensitiveness among themselves, and that the 
possibility of change on exposure, so that they become 
developable, is due to the presence of a sensitiser. 
This may be either concentrated unequally in the 
different grains or may form centres of sensitiveness 
similar to those supposed to exist by Svedberg and 
other workers in the field, who think that the centres 
found at the beginning of development are the origin 
of sensitiveness, and are present from the time of 
making the emulsion. 
If we had no prior knowledge of the wave theory of 
light, however, it is clear that the simplest explanation 
of the sensitiveness of different grains would be that, 
instead of a continuous flow of light in the form of 
waves on to sensitive films, the light was falling upon 
it as a rain of projectiles, and that these projectiles 
made developable any grains that they hit, the grains 
that were missed not being developable, but being 
NO. 2786, VOL. 111] 


hit later if they continued to be exposed to the radia- 
tion. Naturally, the bigger the grains the more 
likely are they to be hit, so that a calculation can be 
made of the relation between the size and the per- 
centage number of grains which will become develop- 
able after a given exposure. . 
Silberstein suggests that the projectiles, rather than 
being called “ corpuscles,” which gives the idea that 
they are round, should be called “light darts,” and 
should be imagined to consist of a long train of waves 
of very small diameter travelling with the velocity 
of light. ; 
It is obvious that this theory of light darts would 
meet the difficulties which are offered by the phenomena 
of X-rays and photo-electricity to the idea of a con- 
tinuous wave front, while not excluding the possibility 
of the formation of interference and diffraction effects. 
At first sight it would seem to offer a solution of the 
problem of the integration of exposure by the silver 
halide grains of the emulsion, since we might assume 
that, instead of a grain integrating energy falling 
upon it until it had received enough to make it de- 
velopable, it was not affected at all until struck by a 
quantum of light, and then became developable com- 
pletely. If this was so, however, we should expect 
that the amount of energy necessary to make a grain 
developable would be, on the average, one quantum, and 
at most a few quanta, more than one being necessary 
because of the chance that a fresh grain would not be 
struck by every “light dart ” falling upon the emul- 
sion, some falling between the grains and others 
striking grains which were already developable. 
In some work which has just been started in our 
laboratory we are getting results from which I think 
we may conclude asa preliminary statement that, for 
high-speed emulsions, several hundred quanta of violet 
light are necessary per grain in order to make the grain 
developable. If this is confirmed, the light dart 
hypothesis would seem to be scarcely sufficient by 
itself to explain the integration of energy by the emul- 
sion, and we are thrown back on to the idea of differ- 
ential sensitiveness among the grains, or of spots of 
limited area on the grains, so that of the hundreds 
of quanta striking a grain only one may be considered 
to be operative, the rest falling upon the insensitive 
portions of the grain. Suppose that the fraction of 
a grain which is sensitive is e, and this consists of 
an average of K spots of area, then 
Ko=aa. ; 
Now, if a grain has no spots, it will be quite insensitive 
and will not be developable, no matter how long it is 
exposed, so that the value of K and can be deter- 
mined experimentally by counting the grains left 
over after a very prolonged exposure. “a 
In any case, a question of great importance in con- 
nexion with the latent image is the amount of energy 
required to make the silver halide developable. If the 
new determinations show that several hundred quanta 
of violet light per grain are necessary, then a revision of 
ideas relating to the latent image itself will follow, as. 
compared with those ideas derived from the belief that 
the energy available is only one quantum per grain, in 
which case it is clear that the latent image must depend 
upon a change occurring in a single atom of silver or 

| 
