1864. | Stewart on Radiant Light and Heat. 597 
many rays, it will give out a good many. We have further to observe, 
that the rays it gives out must be precisely of the same kind as those 
which it arrests: it stops the red rays. Now, if it give out dark heat, 
but not red, then, looking towards the glass, we shall get no light from 
it, since it stops the red rays from behind, and gives out none of its 
own. But this is evidently impossible in a red-hot chamber, for 
universal experience teaches us, that any substance left in such a 
chamber will ultimately appear red hot. If, therefore, the glass be 
of such a nature as to stop any particular ray of light or heat, it will, 
when at a high temperature, give out on its own account that very 
ray. This may be exemplified by several very simple experiments. 
Take a piece of ruby glass, coloured with gold. This glass, as its 
name sufficiently denotes, allows all the red rays to pass, but stops 
the green. Heat it in the fire to a good heat, and examine it in 
the dark, when it will be found to give out greenish rays. On the 
other hand, green glass (which stops red rays) will, when heated, 
give out a dark-red light. These curious facts may be noticed at 
any place where coloured glasses are spun. Another good illustra- 
tion of this law is obtained by puttimg a number of differently- 
coloured glasses into the fire, and it will be found that they all 
appear to lose their colour when they become of the same temperature 
as the coals around them. Not that the glasses have changed their 
nature in the least, for the red glass still stops the green, and the 
green glass the red rays, from the coals behind ; but each glass gives 
out, on its own account, precisely those rays which it stops, so that 
the light (transmitted and radiated together)-which comes from the 
glass is just the same as would have come from the coals alone. 
Kirchhoff has beautifully extended this law to those individual rays 
which compose the spectra of luminous bodies. We have already 
stated that the spectrum of sodium consists of two bright lines p, and 
in general that the spectrum of a heated vapour consists of bright 
lines on a dark ground. Now, if instead of using these vapours as our 
source of heat, we use them in a comparatively cold state as a screen, 
our source being an incandescent solid body which gives out all rays, 
then we shall find that each vapour stops precisely those rays which 
it gives out when heated. Thus, a salt flame, or incandescent sodium, 
gives out the double line p; vapour of sodium (when cold) also 
stops the same line, and a similar rule holds for all vapours. When- 
ever, therefore, we have a source of light containing all rays, but 
which is surrounded by an atmosphere of metallic vapours at a low 
temperature, each of these vapours will stop its own appropriate rays, 
and the spectrum of such a system will exhibit a bright ground inter- 
sected by dark lines, This is precisely the character of the solar 
spectrum, and we have seen that one of its lines, namely p, is that 
which is produced by the absorptive power of sodium vapour. We 
are therefore entitled to conclude that this vapour exists in the 
atmosphere of the sun. Again, the vapour of iron gives out certain 
bright lines which appear as dark in the sun’s spectrum, we conclude 
therefore that iron vapour exists there in a comparatively cold state ; 
and in like manner we find that a number of the solar lines are due to 
