Now, besides the silver bromide, we have in the emulsion water,
gelatine, potassium nitrate, and a small excess of potassium bromide;
and in order to find which of these is the cause, we must make different
emulsions, omitting in succession each of these materials. Suppose we
take an emulsion which has just been mixed, and, instead of boiling
it, we precipitate the gelatine and silver bromide with alcohol; on
redissolving the pellicle in the same quantity of water, we have an
emulsion the same as previously, with the exception that the niter and
excess of potassium bromide are absent. If such an emulsion be boiled,
we shall find the remarkable fact that, however long it be boiled, the
silver bromide undergoes no change, neither does the emulsion become
any more sensitive. We therefore conclude, that either the niter or the
small excess of potassium bromide, or both together, produce the change.

Now take portions of a similarly washed emulsion, and add to one portion
some niter, and to another some potassium bromide; on boiling these
we find that the one containing niter does not change, while that
containing the potassium bromide rapidly undergoes the changes
mentioned.

Here, then, by a direct appeal to experiment, we prove that to all
appearance comparatively useless excess of potassium bromide is really
one of the most important constituents of the emulsion.

The following table gives some interesting results respecting this
action of potassium bromide:

__________________________________________________________
Excess of potash bromide. | Time to acquire maximum |
| sensitiveness. |