END-ON VIEWS OF GAS SPECTRA UNDER HIGH DISPERSION. 153 
good end-on views of it, I had abandoned the project of repeating the Swedish experimenters’ 
measurements, and I obtained instead some rapid measures of a superb series of red and yellow 
flutings in a new end-on cyanogen tube, and had completed its astonishing survey before it 
occurred to me to compare the series with the imperfect but very similar colonnades visible in 
the other nitrogen-containing tubes. A rather weaker display seen in the tube marked 
“ Nitrogen” was measured, and it was immediately seen to be identical with the series in the 
cyanogen tube. The identity of the same series in all the compound-of-Nitrogen tubes which 
I have since tested also leaves no doubt of the absolute constancy of its appearance, as far as 
it is visible without confusion and obliteration by other substances in these tubes. Thus the 
wished-for end of its remeasurement was already attained in the set of readings noted of the 
magnificent array of linelet groups seen ruling the red and yellow portions of the cyanogen 
spectrum.* 
In this marvellously beautiful array of triplets (the tube has now lost its original perfec- 
tion), one following another at a little interval, a small intrusive line in one of the intervals 
did not conform to the measures. A suspicion of its origin being thereupon excited, the hydro- 
gen comparison tube was lighted up, and it immediately proved to be Ha. A mere trace of 
hydrogen so weak and feeble as this was, is, from. the ordinary prevalence of aqueous vapour, 
rather an exceptional occurrence in a vacuum tube. The conclusion, however, which may be 
drawn from its scarcity in this instance is a point of special moment to the theory of these 
channeled spectra, the explanation of which was given to me on the occasion of this occurrence 
by Professor Prazzi Smyru. As the proposal to contribute this Appendix has prompted and 
invited me to the fullest freedom of communication, I gladly avail myself of the liberty with 
which I am thus entrusted to reproduce it here, in order to show upon what small and 
apparently insignificant appearances, sound and just views of the nature and origin of gaseous 
spectra may sometimes come to be correctly founded. 
The absence or deficiency of hydrogen is demonstrative of the sensible absence or remark- 
able deficiency of aqueous vapour, and consequently (admitting the purity of the included gas from 
air) of disposable oxygen in the cyanogen tube. Yet not only is the serried nitrogen colonnade 
most resplendent in its electric spectrum, but so also are those vacuum tube carbon bands which 
are best known as constituting the electric spectrum of carbonic oxide and carbonic acid vacuum 
tubes. Of all the lateral and end-on tubes examined, these latter bands, like those of Nitrogen, 
were noticed by Professor P1azzi Smytu ¢o be most brilliant in this Cyanogen one, in which yet 
there can hardly be assumed to exist more than a mere trace of Oxygen set free, to combine 
with the carbon and nitrogen, by the electric spark! A new Cyanogen tube supplied to me quite 
recently by M. SaLLeron yields an electric spectrum of the very same description. The opinion 
held by ANGSTROM and THALEN, therefore in their Memoir appears to be scarcely tenable, that the 
fluted and banded spectra just mentioned are those of the oxides of Nitrogen and Carbon, but 
it seems more probable that these are in fact the true low temperature spectra of those metal- 
loids. The coal-gas or blow-pipe flame spectrum, on the other hand, is probably attributable to 
olefiant-gas.f A similar band spectrum is recognised by Professors Liveinc and Dewar as 
* SALET, as well as Picker and Hirrorrr, struck by the identity of this spectrum in all the nitrogen-bearing 
tubes, was led to the opinion from its constancy that the real source of the fluted spectrum is nitrogen itself. 
+ A different view of this spectrum is, however, taken by Mr Lockyer, in whose opinion it is one form of the 
spectrum of elemental carbon. The smooth-shaded tube-carbon bands, in fact, resolve themselves into the line-bearing 
gas-flame ones on simply strengthening the induction discharge with a condenser, and especially on introducing at the 
same time an air-break also in its course. The experiment was tried after the present paper was read, with the above 
described carbonic acid and cyanogen vacuum-tubes, on July 23, 1880, on its prescription by Mr Lockyer to the writer 
of the paper, and to Professor Prazzi Smyru, and it succeeded in the presence of its suggester, literally as he expected ! 
WOT XexeOXG) PAIR Te Z 
