465 
of Edinburgh, Session 1870 - 71 . 
The last of those portions is so widely separated from the rest as to 
induce a suspicion that it may be more refracted than any rays in 
the solar spectrum. This separated portion has a pale undecided 
hue. I should hardly have called it violet were it not situated at 
the violet end of the spectrum. To my eye it had a somewhat 
whitish or greyish appearance.” 
This was followed by another paper of mine “ On Prismatic 
Spectra” (Phil. Mag. 1836, vol. ix. p. 3), in which the spectra of 
gold, copper, zinc, boracic acid, and barytes are described. 
Wheatstone, nearly at the same time, published some interesting- 
analogous researches. I regret not to have his paper at hand at 
present, in order to give a full account of it. 
Brewster then took up the subject, and described the spectra 
produced by the combustion of a great variety of substances, in a 
paper printed in the Manchester meeting (1842) of the British 
Association (see Proceedings of the Sections, p. 15). But in the 
same page there is another short paper by Brewster, of surpassing- 
interest, since he there announces the fact that the bright rays 
which are characteristic of artificial flames are for the most part 
those which are deficient in solar light, a fact previously confined 
to the line D, and discovered, as we have said, by Fraunhofer. 
These observations of Brewster deserve to be quoted textually. 
His paper is entitled “ On Luminous Lines in certain Flames cor¬ 
responding to the defective Lines in the Sun’s Light.” 
After noticing Fraunhofer’s beautiful discovery as to the phe¬ 
nomena of the line D in the prismatic spectra, Sir David said— 
“ He had received from Fraunhofer a splendid prism, and upon 
examining by it the spectrum of deflagrating nitre, he was surprised 
to find the red ray discovered by Mr Talbot, accompanied by several 
other rays, and that this extreme red ray occupied the exact place 
of the line A in Fraunhofer’s spectrum, and equally surprised to 
see a luminous line corresponding to the line B of Fraunhofer. 
In fact, all the black lines of Fraunhofer were depicted in the 
spectrum in brilliant red light. The lines A and B in the spectrum 
of deflagrating nitre appeared to be both double lines, and upon 
examining a solar spectrum under favourable circumstances, he 
found bands corresponding to these double lines. He had looked 
with great anxiety to see if there was anything analogous in other 
