OF SUNLIGHT THEOUGH THE EARTH’S ATMOSPHERE. 
3 
therefore,, from such observations k could be at once determined. It was similarly 
shown that when a photographic silver salt was employed to register the photo¬ 
graphic spectrum it was equivalent to measuring the integrated spectrum with a 
coefficient of absorption jx', and that this was equivalent to observing the alteration 
in intensity of a ray X 4540. 
XXV. — Application to the foregoing Results. 
This being the case, if by any photographic means we can measure the total 
intensity of white light affecting a photographic compound after passing through 
various air thicknesses, we ought to be able to find the value of /Tj in the formula 
and if we a,lso observe optically the value of light transmitted through various air 
thicknesses we shall get the value of g in the formula As already stated we 
know that g = 105/c. Similarly, g = Xp‘^.K, being the equivalent ray observed. 
Therefore 
g 105 
or 
and 
K 
jf 
Ay^’ 
that is, w'e can find the single ray which is the equivalent of the vhole of the spec¬ 
trum which is impressed on any photographic compound, and having thus found A^“^ 
we can at once deduce the value of k — or the coefficient of scattering by the fine 
particles. 
XXVI .—Sensitive Compound employed. 
The question which presented itself was as to the best form of sensitive salt 
to use for convenience and accuracy. A process with a silver salt recjuiring 
development was almost impracticable. The exposure necessary to give to sunlight 
would have been so small that accuracy in timing it would have become a very 
difficult problem. The chloride of silver paper v/hich darkens by suidight was experi¬ 
mented with, but it had a very serious drawback, Roscoe and Bunsen have shown 
that chloride of silver paper may be prepared, which, when exposed to the same 
intensity of light for the same time, will always give the same blackness, but it is 
impossible to keep this paper for more than a few hours, and it would often be 
inconvenient to prepare it. Further it can never be pi'edicted when a day will be 
suitable for making observations, and measurements of its blackness would be difficult 
except at a fixed observing station. There were also two other desiderata which 
had to be taken into account; one was that the record of the action of light should 
be as permanent as possible, and the other that it should be easy to measure the 
action produced by the light. Chloride of silver paper fulfilled neither of these 
desiderata, since fixing the chloride altered all measurement and the darkening was 
liable to fade even when the paper was fixed, and it was not easy to obtain accurate 
measures of the darkening owing to its ruddy colour. 
In the comparatively new process of platinotype we have, however, a process which 
B 2 
