414 MR WILLIAM SWAN ON THE PRISMATIC SPECTRA OF THE 
to tinge a flame with bright yellow light; and as the equivalent weights of sodium 
and chlorine are 23 and 35:5, it follows, that a quantity of sodium not exceeding 
250000 Of a troy grain renders its presence in a flame sensible. If it were possible 
to obtain a flame free of yellow light, independently of that caused by the salt 
introduced in the experiment, it is obvious that a greatly more minute portion of 
sodium could be shown to alter appreciably the colour of the fame. It therefore 
follows, that much caution is necessary in referring the phenomena of the spectrum 
of a flame to the chemical constitution of the body undergoing combustion. For 
the brightest line in the spectrum of the flame of a candle,—the yellow line R* of 
FRAUNHOFER,—can be produced in great brilliancy, by placing an excessively 
small portion of salt in a flame, in whose spectrum that line is faint or alto- 
gether absent. The question then arises, whether this line in the candle flame 
is due to the combustion of the carbon and hydrogen of which tallow is chiefly 
composed, or is caused by the minute traces of chloride of sodium contained in 
most animal matter. When indeed we consider the almost universal diffusion of 
the salts of sodium, and the remarkable energy with which they produce yellow 
light, it seems highly probable that the yellow line R, which appears in the 
spectra of almost all flames, is in every case due to the presence of minute 
quantities of sodium. 
The view, which would attribute a great portion of the light of the envelopes 
of flames to the adventitious presence of minute traces of foreign matter, may 
possibly serve to explain certain anomalous diversities of colour which are 
observed in the envelopes of flames arising from the combustion of the same ele- 
ments. Thus tallow, coal gas, anhydrous alcohol, and weak spirit of wine, all 
contain the same combustible substances, carbon and hydrogen: yet the envelope 
of the flame of a candle is bright yellow, that of a coal gas flame is purple, and 
those of strong alcohol and weak spirit differ greatly in luminosity. 
It is important also to remark, that while the luminosity communicated to 
the exterior envelope of a flame by such substances as the salts of sodium or of 
copper, may be so great as to disguise that of the inner bright cone of the flame, 
or in some cases to render it altogether invisible; yet I have ascertained that the 
light of the blue portion of the flame, or of the inner cone, remains absolutely 
unchanged in colour and intensity. The proof of this curious property of flame 
will be given in the sequel. 
Prismatic Analysis of Flame. 
Reserving, meantime, a more complete description of the apparatus I have 
employed, it may be sufficient to premise, that, in what I shall have to say regard- 
ing the spectra of flames, the object observed is supposed to be a narrow illumi- 
nated slit, viewed through a glass prism mounted before a telescope, which has 
been adjusted to focus on the slit. 
* Scuumacuer’s Astronomische Abhundlungen, 1828, p. 18. t See p. 419. 
