This was fitted perfectly tight in my coating-box.

I placed some pure undiluted bromine water and the agents necessary for producing chlorine gas (in small quantity) in the retort.

The result was that my first experiment produced an impression completely solarized in all its parts by an exposure of four seconds of time, which would have required an exposure of twenty seconds to produce a perfectly developed impression by the usual process.
Another trial immediately produced one of the finest toned impressions I ever saw, perfectly developed in one second of time.
My next two or three experiments proved total failures.

I was unable to produce even a sign of an impression.

By accident my retort was broken, and not being in a locality convenient to obtain another, my experiments were necessarily suspended.
My attention was not called to this subject again for several years, when I noticed an account of some similar experiments by F.A.P.
Barnard and Dr.W.H.Harrington, the latter of whom is now of the firm of Dobyns & Harrington, of New Orleans.
From reading this article, I found my own difficulties explained.