THE LIOTIS OF THE SOLAE SPECTEUM. 
155 
prominent, did not appear either black or wide. On October 31, again, the atmo- 
spheric lines were not so dark as usual, while the rays beyond C 10 had evidently 
suffered a considerable absorption. On neither of these occasions was there mucli 
colom’ in the sky; but that the phenomena did not depend on either the absence or 
presence of humidity in the atmosphere, is evident from the fact that on the earlier date 
there was a keen ffost, while on the later day the weather was wet, the thermometer 
being 38° F. at the time of observation. That moisture, however, has some influence 
in the production of these bands, is shown by the effect of a fog on the solar radiations ; 
thus on November 20, 1858, at 10 o’clock a.m., at London, the sun loomed red through 
a mist, and a prismatic analysis of its hght showed a and B with extreme distinctness, 
and the characteristic C 6, S, and n. 
When the sun’s rays traverse a mixture of milk and water, though they are dispersed 
and absorbed to a large extent, especially at the more refrangible end of the spectrum, 
these atmospheric lines are in no way exhibited ; a proof, if additional proof be needed, 
that they are not owing to the mere reduction of the light. 
It is a most beautiful and striking sight to observe the gradual appearance of these 
characteristic lines as the sun descends towards the horizon. Professor PiAzzi Smyth 
remarked it ; and from his elevated position on the Peak, he had the peculiar advantage 
of observing the sun when it had sunk beneath the astronomical horizon, even to the 
depth of 1°T. In his drawings^ it is curious to trace the gradual darkening of a and B, 
and the intermediate lines, of what he calls the “ growing ” line beyond C, and of the 
dark band that follows D. In the drawing made by him when the sun was at an altitude 
of — 1°T, the bands between C 6 and h appear to occupy a still larger space than in the 
diagram compiled from the observations in Scotland and England. This, of course, is 
in accordance with what might have been anticipated. 
Measurements. 
The following are the refractive indices of the principal lines of the spectrum, atmo- 
spheric or otherwise, as determined for the flint-glass of which the prism employed by 
Dr. Gladstone is composed : — 
A . . . . 
. . 1-6069 
F 
. 1-6292 
a ... . 
. . 1-6087 
G 
. 1-6404 
B . . . . 
. . 1-6104 
G33 . . . . 
. 1-6464 
C . . . . 
. . 1-6122 
H 
. 1-6501 
C 6 . . . . 
. . 1-61.39 
K 
. 1-6513 
D . . . . 
. . 1-6162 
I 
. 1-6548 
End of § . 
. . 1-6192 
L 
. 1-6567 
End of ^ . . 
. . 1-6219 
2nd of group M . 
. 1-6589 
E . . . . 
. . 1-6234 
5th of group M . 
. 1-6614 
b ... . 
. . 1-6249 
1st of group N . 
. 1-6642 
* Philosophical Transactions, 1858, Plate XXXV. 
