56 Opacity of the Yellow Soda-Flame to Light of its own Colour. 
which they prove that the luminous lines which are produced in 
the spectrum of a spirit- or gas-flame, when salts of the alkalies 
or alkaline earths are caused to volatilize therein, become reversed 
(2. e. that the bright lines become changed to dark ones) when 
a source of light of sufficient intensity, and giving a continuous 
spectrum, is placed behind the coloured flame. 
During some researches on the spectra of artificial flames which 
I have been carrying on simultaneously with MM. Kirchhoff and 
Bunsen, one or two forms of experiment suggested themselves 
which, while they perfectly corroborate most of the facts men- 
tioned by them, seem to merit attention, from the facility with 
which they may be performed, and the striking manner in which 
the phenomena can be exhibited to several persons at once with- 
out the necessity of employing any optical apparatus. The 
atmosphere of the room—or rather that part of it in which the 
illustration is performed—is to be first impregnated with soda- 
smoke, by igniting a piece of sodium, the size of a pea, on wet 
blotting-paper. Any flame, whether of gas, spirit, a candle, &e., 
which may now be burning anywhere in that part of the room, 
will exhibit in a marked manner the well-known yellow soda- 
flame; and if the full amount of gas in an ordinary wire-gauze 
air-burner is turned on and ignited, it will give a uniformly 
brilliant yellow flame, upwards of a foot high and 3 inches in 
diameter. 
If a smaller flame be now moved in front of this large one, it 
will exhibit a curious phenomenon. Those parts of it which are 
ordinarily seen to be luminous will suffer no change, other than 
that slight dimimution in intensity which might be anticipated 
from their projection in front of a broad but not very brilliant 
source of light; but beyond these there will appear a sharply- 
cut and intensely black narrow border, closely surrounding the 
visible flame, and presenting the curious appearance of the latter 
being set in an opake frame. A closer scrutiny will show that 
the position of this black rim is not, as I at first supposed, in 
that outer cone in which the yellow soda-flame is most distinctly 
seen, but that it lies in the dark space immediately outside the 
luminous part of the flame, affording proof of the existence of 
another invisible cone of vapour. The flame from a tallow candle 
shows this appearance better than that of wax or sperm, pro- 
bably on account of its inferior luminosity. A small spirit- or 
gas-flame will also answer very well; but I think a tallow candle 
shows the phenomenon in a more striking manner. 
The fact of the cone of yellcw soda-flame being transparent, 
while the outer, non-luminous space is perfectly opake to the 
same kind of light placed behind it, appears worthy of attention. 
It seems to show that the yellow flame caused by the presence 
