5 
, 2 
% . ee hy ve ‘ 
" * 
f » ee 
| agaite : Seer ee cs 
5 Fan. 30, 1873] 
_ these solid white-light-giving carbon points or that from 
an ordi flame, I show you the spectrum of a light 
- source which is coloured. If, for instance, we burn seme 
4 
tig 
A 
Ad 
L| 
Me ret A 
247 
coloured fire, such as the red fire of our pyrotechnic dis- 
plays. You must not consider that this is sensational, for 
Sir John Herschel, very many years ago, was on the eve 
Fic. 21.—Arrangement for determining the spectra of metals by means of the electric spark. 
of discovering the great point of spectrum analysis which 
I have to bring before you, by merely examining these 
coloured fires. If we examined such a light by means of the 
spectroscope you might expect that we should obtain the red 
| localisations of light or bright lines in different parts of 
the spectrum. Now, the differences in colour are accom- 
panied by differences in the spectra. We have something 
very different from the continuous spectrum we had before, 
and this is, in fact, one of the first practical outcomes of 
f spectrum analysis. It enables you in a moment to deter- 
mine the difference between a solid or liquid body, which 
gives you a continuous spectrum, and a vapour or gas, 
which gives you a spectrum containing bright lines. The 
reason that different vapours and gases are of different 
| 
Fic. 23.—Bunsen’s burner for flame spectra. 
| colours is now clear ; if we examine the light by means of 
| as p2ctroscope, we find that the hght rays wh’ch they emit 
are located in different parts of the spectrum, 
In these instances then, the spectra consist of lines which 
are located in different parts of the spectrum. Let us burn 
some sodium in air, and then examine the spectrum of its 
: vapour, or better still, let us place some sodium, or a 
end of the continuous spectrum ; that on burning green | salt of this metal, such as table salt, in a gas flame 
fire we should see the green portion of the spectrum and/| which is consuming a mixture of air and gas, in a — 
soon. But this is not so ; we findthat the background of | burner known under the name of a Bunsen’s burner, © 
the spectrum is dark or nearly so, and that we have certain | the bluish flamé of which is due to the complete com 
Fic, 22.—The E estric'Arc 
