458 
CAPTAIN W. DE W. ABNEY ON THE SOLAR SPECTRUM. 
that given in the Phil. Trans, for 1880, except that I find the addition of nitric acid is 
not necessary to be present whilst the emulsion is formed, though in the subsequent 
washings it is convenient to use it. This may be avoided, however, by washing first 
with water, and using a dilute solution of iodine to eliminate the veil which is nearly 
always present after the emulsion is boiled. To the emulsion when prepared I add 
about per cent, of good soluble cotton; that known as Anthony’s No. 1 I find the 
most satisfactory. A very sensitive gelatine emulsion plate is coated with the 
collodion emulsion, washed, and then allowed to dry in a warm chamber. The 
washing causes a minute portion of the underlying gelatine to mingle with the 
collodion film surrounding the sensitive salt, and to protect it from premature reduc¬ 
tion by the ferrous oxalate solution. As I have pointed out in other communications, 
gelatine acts as a physical restrainer, in contradistinction to a soluble alkaline bromide, 
which acts as a chemical restrainer. When the former is used there is no partial 
obliteration of the effect of radiation on the salt, as is the case when the latter is 
employed ; and as a consequence a shorter exposure is necessary. 
Apparatus employed. 
At first I used the grating with which I had worked for some tune previously 
in my preliminary work. It had a surface of \\ inches square and about 17,200 wires 
to the inch, and it is excellent in its performance though with spectra of high 
orders, ghosts can be traced. Wishing to use less dispersion to obtain a higher wave¬ 
length, I secured from Mr. Chapman, of New York, a grating ruled on parallel glass 
which Mr. A. Hilger had prepared. For the visible spectrum this glass grating gives 
excellent results, but it is otherwise for the infra-red. When 1 obtained the grating 
I backed it with a mercury amalgam, intending to use it as a reflection grating 
through the glass, but on trying it I found that the spectrum in the first order ceased 
about X 8300. This want of success I laid to the metallic backing, and removed it to 
replace it with one of silver. The glass was carefully cleaned with nitric acid and 
very dilute potash and alcohol, and silvered by Martin’s process. On taking it from 
the bath I was surprised to find that the layer of silver was so delicate that it took 
in the grooves formed by the rulings, and that reflection spectra could be taken from 
both surfaces, i.e., from the silvered surface direct or through the glass. A grating 
such as this would have proved of the very greatest service, since all that is necessary 
to have a perfect reflecting surface is to remove the silver film and to re-silver again. 
Unfortunately the grating thus silvered proved to be useless for my purpose, for there 
was no infra-red spectrum, probably owing to some peculiarity in the shape of the 
grooves formed in ruling. That such phenomena are by no means uncommon would 
now appear to be recognised. Professor Rowland most generously gave me several 
gratings, and amongst them is one ruled on a spherical concave surface, of a radius of 
7 feet 9 inches. This grating is so ruled that there is one really bright spectrum, and 
that one of the first order; all the other spectra are faint. In another grating the 
