TRANSACTIONS OF SECTION A. 943 
in the table of the compound spectrum of nitrogen of the Report already quoted. 
Again, in place of the 115 wave-lengths of haze-band edges, noted as forming the 
whole fluted or compound spectrum of nitrogen in that Report, scarcely less than 
7,000 lines of the same spectrum have been so clearly seen and distinguished in it 
with the bisulphide of carbon spectroscope, as to be all mapped and delineated 
on the lithographic plates of the results obtained with this spectroscope, which will 
ke published in the next, now nearly completed, volume of the Transactions of 
the Royal Society of Edinburgh, 
The measurements of this gas-spectrum’s exceedingly full crowd of lines were 
rendered possible in great measure by end-on vision-tubes of nitrogen of very 
superior purity and brightness, produced for the author by Mr. C. Casella, 
the electric excitation of the tubes being also brought under exceptionally regular 
control by a new battery and five-inch spark induction-coil specially supplied to 
him for this purpose by Mr. Alfred Apps, of London, 
A fourth example of the spectroscope’s great optical efficiency was its complete 
disentanglement of the remarkable web of fine lines forming the green band of the 
electric spectrum of carbonic oxide. When seen through a spectroscope of this 
power a curious crossing presents itself near the band’s brighter edge, arising from 
the concurrence there, apparently, of two really distinct but not easily separable 
series of very similarly arranged linelets, of which the band’s line-cluster must at 
least consist. In Professors Angstrém and Thalén’s drawing of it (14 inch long) 
in their ‘ micrometric measurements,’ ! the details delineated only amount to fourteen 
separately represented light maxima of the gradually decreasing band. But with 
the expansion of this length which the micrometer-screw’s automatic tracing point 
of the bisulphide of carbon prism spectroscope effected on its paper page, from 14 
inch to 26 inches (corresponding in its scale to nearly 4 inches for the interval 
between the two sodium D lines, and to about 220 feet in length for the whole 
visible extent of the spectrum from A to H), no haziness of light was any longer to 
be seen, but instead of ita series of 48 sharp linelets in this space were plotted 
accurately in their true relative positions by the recording pointer on the paper. 
Even with this automatic means of registration, however, the task of projecting 
the prodigious multitude of sharp lines observed in the several different gases sub- 
mitted to examination, demanded the writer’s available diligence towards this end 
too constantly to allow him to give special attention to any harmonic or other 
numerical relations which may exist among the abundant data thus collected, 
although researches of this kind hereafter are not proposed to be omitted. A 
notable example of the probable prevalence of such relations, and a striking corro- 
boration of the clear vision and accuracy reached in the projections, presented itself, 
however, in the mapped places just now described of the 43 carbonic oxide green- 
band linelets. A map of this band’s evidently complete resolution into all its com- 
ponent linelets was sent for inspection to Professor Herschel, and it immediately 
suggested to him a simple law of progression which embraces all the lines mapped 
upon the sheet (of which a figure accompanied this paper) of this band’s visible 
components; if, at least, duplicity of more than half the lines of one of its ranks be 
disregarded, which is so extremely close that it scarcely yet furnishes any material 
occasion for the paired lines’ separate descriptions, The exactness of the surmised 
progression’s agreement with the record is all the more noteworthy when it is stated 
that the latter includes measurements of intervals between some of its lines not ex- 
ceeding the 50th part of the micrometer interval between the two sodium lines, and 
that within about that diminutive quantity also it everywhere agrees with the pre- 
sumed line-places of the theoretical progression, whose range yet extends to many 
intervals between the sodium-lines along the linelet ranks. 
The rank of split-lines of the green carbonic oxide band has successive intervals 
beginning from its first line which forms the leading (or least refrangible) edge of 
the green band, represented by the natural numbers, 1, 2, 3, 4, &c., with a space 
between the first two lines, or an arithmetical difference between the intervals of 
the succeeding lines, of about two British, or three-quarters of a metrical wavye- 
1 Transactions of the Royal Society of Upsala, 1875. 
