IN CONNEXION WITH THE SPECTRUM OF THE SUN. 
569 
was sent to the School of Military Engineering at Chatham, where, by the kind 
permission of the authorities, it was photographed down to the scale determined upon. 
IV. IMPORTANCE OF THE PHOTOGRAPHIC METHOD. 
The importance of applying photography to the violet and ultra-violet portions of 
the solar spectrum seems to have been fully appreciated by Angstrom.'" 
The classical “Spectre Normal” is avowedly incomplete about this region, both with 
regard to the number of lines and their wave-lengths. Even in portions of the spec¬ 
trum distinctly visible—such, for example, as from G to F—a good photograph shows 
a much greater number of lines than the corresponding region of the “ Spectre Normal.” 
Similarly with regard to metallic spectra : the most complete spectral maps hitherto 
published, those of Thalen,+ contain in a given region a much smaller number of lines 
than are mapped for the same region by means of photography, even when the 
spectra are purified to the greatest possible extent by the elimination of all known 
impurity lines. Numerical comparisons illustrating this superiority of the photo¬ 
graphic method over eye observation have already been given in the preliminary note 
before referred to. 
O 
Since the publication of the “Spectre Normal,” and of Angstrom and Thalen s 
map of the violet portion of the solar spectrum, a map of the ultra-violet region has 
been constructed by MascartJ by means of photography ; but this map possesses the 
disadvantage of an arbitrary scale, and no metallic lines are introduced. A diffraction 
spectrum obtained by means of photography was published by Dr. Draper in 1872. 
This spectrum extends from beyond G to O, and although the best diffraction spec¬ 
trum hitherto published, the wave-length scale is too indistinct and blurred for use in 
very accurate measurements. The allocation of the lines with those of metallic 
spectra is, moreover, not shown, although the author states that he has photographed 
some of the metallic spectra, both by means of a grating and of a train of quartz 
prisms. The excellent photograph of the solar spectrum taken by Mr. Pujtherfurd, 
as might have been expected, contains a much greater number of lines than the 
“ Spectre Normal” for the same region, but being a refraction spectrum is of course 
not available for the purpose of directly determining wave-lengths. The most perfect 
map of the violet and ultra-violet solar spectrum at present in existence is that 
recently published by Cornu, who has determined absolutely the wave-lengths of 
thirty-six of the principal lines in the portion of the spectrum included between O 
(wave-length 3440) and beyond h (wave-length 4120), the remaining lines, about 650 
in number, being introduced by interpolation. This map was constructed by means 
* “ Quant a la lumiere violette et extra-violette, j’espere pouvoir, a l’aide de la pbotog’rapbie, en publier 
bientot un aper^u exact et detaille, surtout comme les tentatives, deja faites a cet egard par M. Thalen, 
confirment ces esperances.”—(‘ Recbercbes sur le Spectre Solaire,’ Upsal., 1869.) 
t ‘ Nova Acta Regise Societatis Scientiarum.’ Upsaliensis, 1868. 
+ “On the Rays of the Ultra-violet Solar Spectrum.’’ ‘Compt. Rendus,’ Nov., 1863. 
4 e 
MDCCCLXXXT. 
