186 Prof. Kirchhoff on the Chemical Analysis 
contained in the atmosphere, and from the presence of these 
lines should infer that of these various metals. The more intense 
luminosity of the sun’s solid body, however, does not permit the 
spectrum of its atmosphere to appear; it reverses it, according 
to the proposition I have announced; so that instead of the 
bright lines which the spectrum of the atmosphere by itself 
would show, dark lines are produced. Thus we do not see the 
spectrum of the solar atmosphere, but we see a negative image 
of it. This, however, serves equally well to determine with cer- 
tainty the presence of those metals which occur in the sun’s 
atmosphere. For this purpose we only require to possess an 
accurate knowledge of the solar spectrum, and of the spectra of 
the various metals. 
“T have been fortunate enough to obtain possession of an appa- 
ratus from the optical and astronomical manufactory of Steinheil 
in Munich, which enables me to examine these spectra with a 
degree of accuracy and purity which has certainly never before 
been reached. The main part of the instrument consists of four 
large flint-glass prisms, and two telescopes of the most consum- 
mate workmanship. Jy their aid the solar spectrum is seen to 
contain thousands of lines; but they differ so remarkably in 
breadth and tone, and the variety of their grouping is so great, 
that no difficulty is experienced in recognizing and remembering 
the various details. I intend to make a map of the sun’s spec- 
trum as I see it in my instrument, and I have already accom- 
plished this for the brightest portion of the spectrum—that por- 
tion, namely, included between Fraunhofer’s limes Fand D. By 
painting the lines of various degrees of shade and of breadth, I 
have succeeded in producing a drawing which represents the 
solar spectrum so closely, that, on comparison, one glance suffices 
to show the corresponding lines. 
“The apparatus shows the spectrum of an artificial source of 
light, provided it possess sufficient intensity, with as great a 
degree of accuracy as the solar spectrum. A common colourless 
gas-flame in which a metallic salt volatilizes, is m general not 
sufficiently luminous; but the electric spark gives with great 
splendour the spectrum of the metal of which the electrodes are 
composed. <A large Ruhmkorff’s mduction apparatus produces 
such a rapid succession of sparks, that the spectra of the metals 
may be thus examined with as great facility as the solar spectrum. 
“‘ By means of a very simple arrangement, the spectra of two 
sources of light may be compared. The rays from one source of 
light can be led through one half of the vertical slit, whilst those 
from another source are led through the other half. If this is 
done, the two spectra are seen directly under one another, sepa- 
rated only by an almost invisible dark line. By this arrange- 
