154 PROFESSOR A. 8. HERSCHEL ON 
belonging to Cyanogen (unless it may be due partially to marsh-gas), and the proper spectra of 
the oxides of Carbon and Nitrogen if these gases exist at such high temperatures undecomposed 
may in the course of further trials and examinations of the spectra of ignited gases eventually 
come to be discovered.* Multitudes of fixed bright lines in the spectrum of vacuum tubes 
enclosing pure hydrogen, are confidently regarded by Professor Piazzi SMYTH as constituting 
together the low temperature spectrum of hydrogen; and it is assumable that as no attempts 
to produce the spectrum of aqueous vapour in vacuum tubes have yet been attended with 
success, so also the oxides of the metalloids may be too easily decomposable by the electric 
spark to allow the natural spectra of the oxides of Carbon and Nitrogen to be easily exhibited 
in Geissler tubes. 
New end-on tubes of exceedingly hard glass, made and filled this year for Professor Prazz1 
SmytH by M. Sauieron, have afforded new means of measuring the least refrangible section of 
the nitrogen channeled spectrum in Nitrogen, Nitric Oxide, and Cyanogen, with the advantage 
of the fuller conversion tables now constructed for translating into British wave-numbers the 
readings of the several prism-combinations of the aurora-spectroscope. 
The general appearance under high dispersion of each successive escarpment of the 
nitrogen serration is that of a double-notched band,+ whose two teeth or bright edges face 
towards the red; on the downward or fading slope (towards the violet) of each toothlet’s descent 
are two lines besides the leading line at the edge, of diminishing brightness like the haze on 
which they lie. They divide the first slope into three apparently equal parts, while on the 
second slope the two lines are to appearance similarly placed, but the slope extends to once or 
twice their joint range further, before it fades out and leaves a dark space of a little breadth 
before the same double-notched escarpment begins again. ‘There are also two lines preceding 
the new escarpment edge very similarly spaced asunder, and from the edge, to the other linelets 
of the group ; and this space between two adjacent linelets is only a fifth part wider (in wave 
frequency) than the interval between the two chief sodium-lines (Naa,,a,). It embraced on the 
average six divisions of the micrometer screw-head, in the No. 9 prism, whose repeated read- 
ings of the same line seldom varied so much as one division, and in general only a few tenths 
of a division of the screw; but haziness, expansion, division, and supplanting of the lines due 
to coexisting impurities of other spectra in the tubes rendered exact readings of the fainter of 
the above lines mostly difficult and sometimes impossible. The only good measurements 
secured and entered in the accompanying Table were those of the three linelets on the first tooth, 
together with the leading line only of the second tooth. The results for the remaining lines 
lying between the second tooth and the first tooth or edge of the following escarpment, are so 
devoid of regularity, that, perhaps from their faintness and speedy effacement by carbon and 
hydrogen impurities, no fixed system can be recognised among them. The relative bright- 
nesses of the four recurring linelets a, b, b’, c, whose positions are tabulated, are generally about 
5,3,1,3; and it is the triplet of them a, b,c, whose positions, in metrical wave-lengths, are given 
for their recurring groups, in ANGsTROM’s and THALfy’s Table. 
In the present Table’s columns of the linelet a, the metrical wave-length and wave- 
number are followed by the wave-length and wave-number in a British inch, so as to facilitate 
* Professors Liverne and Drwar’s, and Dr Hucerns’ simultaneous recognitions of the remarkable ultra-violet 
speetrum of aqueous vapour in the light of all hydrogen-bearing flames (‘ Proceedings of the Royal Society of London,” 
June 1880), although announced just previously to the presentation of the above reflections, had not yet been received. 
But they afford as yet no certain evidence that the same spectrum, indicative of aqueous vapour, is also producible by 
electrical discharges in gas-vacuum tubes. 
+ (a, ¢, See the accompanying sketch, p. 157). 
