d no idea whether the action was a rapid or a
slow one; a point, however, of the utmost importance, since, if it were a
slow one, my theory might prove but a philosophic dream.

Such were, as nearly as I can now remember, the reflections which led me
to the invention of this theory, and which first impelled me to explore a
path so deeply hidden among nature's secrets. And the numerous researches
which were afterwards made--whatever success may be thought to have
attended them--cannot, I think, admit of a comparison with the value of the
first and original idea.

In January 1834, I returned to England from my continental tour, and soon
afterwards I determined to put my theories and speculations to the test of
experiment, and see whether they had any real foundation.

Accordingly I began by procuring a solution of nitrate of silver, and with
a brush spread some of it upon a sheet of paper, which was afterwards
dried. When this paper was exposed to the sunshine, I was disappointed to
find that the effect was very slowly produced in comparison with what I
had anticipated.

I then tried the chloride of silver, freshly precipitated and spread upon
paper while moist. This was found no better than the other, turning
slowly to a darkish violet colour when exposed to the sun.

Instead of taking the chloride already formed, and spreading it upon
paper, I then proceeded in the following way. The paper was first washed
with a strong solution of salt, and when this was dry, it was washed again
with nitrate of silver. Of course, chloride of silver was thus formed in
the paper, but the result of this experiment was almost the same as
before, the chloride not being apparently rendered more sensitive by being
formed in this way.

Similar experiments were repeated at various times, in hopes of a better
result, frequently changing the proportions employed, and sometimes using
the nitrate of silver before the salt, &c. &c.

In the course of these experiments, which were often rapidly performed, it
sometimes happened that the brush did not pass over the whole of the
paper, and of course this produced irregularity in the results. On some
occasions certain portions of the paper were observed to blacken in the
sunshine much more rapidly than the rest. These more sensitive portions
were generally situated near the edges or confines of the part that had
been washed over with the brush.

After much consideration as to the cause of t