26 
SIR J. F. W. HERSCHEL ON THE CHEMICAL ACTION OF THE 
tain conclusion then that can be drawn from these trials is, the nonexistence of any 
large inactive spaces in the spectrum. 
7 0 . With the apparatus in its ordinary state, the following are the dimensions of 
the several principal elements of the luminous spectrum, which it will be necessary to 
bear in mind as data in the experiments to be detailed. They are given in parts of a 
scale each equal to -4g-th of an inch, and are reckoned from a point, selected as a fiducial 
centre, at that particular definite yellow which the cobalt glass above noticed insu- 
lates in great abundance and with considerable sharpness of termination*. It was 
not mere caprice which determined me to depart in this particular instance from my 
hitherto invariable standard, the extreme red ray. The solar image in that ray is 
far better defined than in the yellow in question. But from many phenomena ob- 
served in the action of coloured glasses I was led to suspect, and the suspicion was 
borne out by the results, that this ray may be considered as marking a sort of che- 
mical centre, a point of equilibrium, or rather of change of action, in the spectrum. 
In fact, in a large proportion of the preparations submitted to the action of the spec- 
trum, that action, as far as it is marked on the paper by traces of darkness, termi- 
nates precisely, or very nearly indeed, at this point, allowing for the sun’s diameter. 
. „ Parts each 
Dimensions, &c. inch . 
Focal length of the achromatic lens used 774 
Breadth of the fixed spectrum, or diameter of the sun’s image . . . 7'20 
Distanc efrom the fiducial point or centre of the sun’s image formed by 
the yellow ray, 1st, to the centre of the image formed by the ex- > — 13-30 
treme red + J 
2 nd, to that of the image formed by the mean red — 6’66 
3rd, to that of the image formed by the last violet transmitted by the -j 
same glass, a tolerably well-defined ray; the centre of the termi- >+40-62 
nating semicircle of the spectrum judged of as well as practicable . J 
Total length of the luminous spectrum from centre to centre of the 53-99 
terminating solar images J 
71. It would be mere waste of time to recount the almost numberless experiments 
on particular preparations of paper made with this apparatus, especially as they lead 
to no general laws. At the same time, a number of interesting facts have been en- 
countered in particular cases ; and we are led to form a very high idea of the vastness 
of this new field of inquiry, and the extensive bearings and connexions of any theory 
which shall give a rational account of these complex phenomena, from the way in 
which these facts, so various and so remarkable each in itself, stand insulated from 
one another. I shall now therefore proceed to state the phenomena observed onpre- 
* See Transactions of the Royal Society of Edinburgh, 1822. 
f The sign — indicates that the distance is measured towards the least refracted end of the spectrum, the 
sign + towards the most refracted. 
