OF SOME OF THE CHEMICAL ELEMENTS. 
145 
constant, with reference to the position and relative characteristics of its lines, with all 
the metals which have been employed. The air-spectrum as a whole , however, varies 
considerably in intensity and distinctness with electrodes of different metals. As the lines 
are due to the stratum of air separating the points of the electrodes, it is to be expected 
that these lines will appear strongest and most distinct when those metals are employed 
which, being less volatile, will therefore in a less degree displace the air between the 
electrodes with their own special vapour. This consideration appears to be confirmed by 
observation. The air-spectrum is especially intense and distinct when the spark is taken 
between points of platinum, gold, iridium, and rhodium ; whilst, of all the metals which 
I have employed, mercury and sodium, perhaps, are those with which the intensity of 
the air-spectrum is most diminished. With these comparatively very volatile metals, 
the air between and about the electrodes must be, to a very considerable extent, replaced 
by the metals themselves in a state of vapour. It accords with this suggested explana- 
tion of the differences in brightness of the air lines with different metals, that, if the 
electrodes be mercury or sodium and a platinum wire, the air-spectrum is observed to 
be weaker when the current is so directed that the greater heating effect of the dis- 
charge shall be at the mercury or sodium electrode, and to become perceptibly stronger 
when the current is reversed. It is known that, within certain limits, the air-spectrum 
is rendered more intense by the separation of the electrodes. 
The following experiments have been made to refer the lines of this compound spec- 
trum to the components of common air to which they severally belong : — 
a. Hydrogen . — The strong line of the air-spectrum at 589*5 is coincident with Fraun- 
hofer’s C, and with the red line of hydrogen. 
When the spark is taken in air that has passed over sulphuric acid, this line becomes 
very faint. A larger surface of acid being employed, the line faded out so completely 
that no trace of it could be perceived. Steam was then mixed with air, when this line 
became much brighter and the other lines of hydrogen appeared. 
The presence and comparative brightness of this line form a delicate test for aqueous 
vapour. 
b. Carbonic add .— Air that had passed through a solution of caustic potash was 
examined, but its spectrum was not observed to differ from that of ordinary air. When 
carbonic acid is added to air, several prominent lines make then* appearance. These 
are due to carbon, since they coincide with lines in the spectrum of graphite. One of 
the strongest and most characteristic of these lines, and a test for carbonic acid, is a red 
line a little less refrangible than the hydrogen line. Its number is 580*5. 
[Though a good indication of the oxygen and nitrogen compounds of carbon, the 
absence of this line must not always be accepted as a proof that no carbon is present. 
I have recently found that, when carbon is subjected to the induction spark in the pre- 
sence of hydrogen, this line in the red is not seen. Further details of these experiments 
will be given when the spectrum of carbon is described. — February 7, 1864.] 
c. Nitrogen . — In the spectrum of the electric spark when taken in a current of pure 
