i64 



NATURE 



[June £4, 1900 



America,^ with the giant telescope of the Yerkes Observatory's 

 enormous power, that the green and citron hydrocarbon's chief 

 band-lines can be observed dark on the photosphere at the sun's 

 edge, and close by, bright in the chromosphere to a height from 

 the sun's edge which they estimated not to exceed i" of arc, or 

 about 500 miles, to completely ratify the foregoing views that 

 those low gas-flame's fluted bands are produced by carbon vapour 

 at an exceedingly high temperature ; and fully to justify the first 

 observers in England and America of the presence of carbon in 

 the sun, in the opinion which they independently expressed, that 

 the temperature of the glowing region of the sun's atmosphere 

 where this carbon vapour is produced and made to incandesce, 

 must certainly approach nearly to, and at the same time not 

 much exceed, that of the electric arc. 



Carbon substance furnishes yet another known form of gaseous 

 spectrum, which consists only of a few sharp bright lines, quite 

 free from bands of shaded light, or flutings ; and this linear form 

 of its spectrum maybe pretty certainly ascribed to carbon vapour 

 in its simplest molecular, perhaps even monatomic, state of 

 aggregation, since it is only obtained by heating carbon in a 

 condensed electric spark to the highest possible artificial tem- 

 peratures. No indications, however, of carbon's occurrence at 

 this exceedingly high temperature in any celestial spectra, appear 

 as yet to have been met with. Although no gaseous spectra 

 produced at such high temperatures as those of the condensed 

 electric spark were spectroscopically measured by the late Prof. 

 Piazzi Smyth, yet a depiction of this carbon spectrum as it was 

 first seen in the Leyden-jar spark between carbon poles by 

 Angstrom in 1863, and as it was represented by Angstrom and 

 Thalen in their "Spectres des Metalloides" in 1875, is given 

 with the Hne-spectra of common air, hydrogen, nitrogen and 

 oxygen and of vapour of mercury, by diff"erent authors, in the 

 Plate of full-length spectra of his " High Dispersion Spectrum " 

 Paper of 1884, by the late Prof. Smyth, to compare with his 

 own measurements of low temperature spectra of the same 

 elementary gases or their compounds. The map of the carbon 

 line-spectrum given by Angstrom and Thalen shows a spectrum- 

 field extremely bare of lines, but terminated, at its two ends, by 

 two very bright ones, a red, closely double line almost coinci- 

 dent with Fraunhofer's C, or Ho, and a violet one close-follow- 

 ing G and the violet hydrogen line H7, and like the hydrogen- 

 lines appearing to be easily widened into a diff"use, broad line by 

 taking the spark in gases at increasing pressures. ^ 



A faint single line near E, and two groups of three and four 

 moderately bright, pretty close-packed lines near the beginnings 

 of the two brightest (green and citron) flutings of the Bunsen- 

 flame, or " hot-carbon " band-series, are all the remaining 

 visible portions of this spectrum figured, as they had excellently 

 observed and studied it, by Profs. Angstrom and Thalen. But 

 the latter two isolated line-groups appear to fit on remarkably 

 well to the view already apparently borne out and substantiated 

 by what precedes, that with rising temperatures and increasing 

 disgregation of carbon-vapour molecules, the interval between 

 the beginnings of the green and citron flutings becomes wider in 

 passing from the "cool" to the "hot" carbon-band series. 

 For while those band-beginnings are respectively at A. = 51970 

 and 56075 (distance = 41-05/^/4) in the cool, or oxy-carbon set, 

 and at A. = 5 16 -40 and 563-34 (distance ^ 46-94yUM) in the hot, 

 or hydrocarbon set of bands, the front-lines of the " Excelsior" 

 Carbon-spectrum's (as the late Prof, Piazzi Smyth poetically 

 termed it), or still hotter and more broken-up carbon-vapour 

 molecules' two small solitary line-groups, are at A =: 515-05 and 

 569-41, in Angstrom and Thalen's Table of these carbon lines ; 

 both shifted again slightly in position in the same left and right 

 directions as before, and with the interval between them again 

 increased a little, now, from its last measures, 41-05 and 46*94, 

 to 54 •36mm- 



But a most industrious explorer, and a describer and recorder 

 unsurpassed in the skill of his depictions of the surprising beau- 



1 Bulletins of the Yerkes Observatory, No. 12, 1899. 



^ This widening of the carbon violet line to a " broad band " at A=4272 

 (Angstrom, 42660) is very distinctly recorded in Dr. W. M. Watts' 

 ' Index of Spectra," 1872, "Carbon-Spectrum, No, IV."; where the groups 

 and single lines, a, /3 {plus the two next lines), -y, and t, compose together 

 Angstrom and Thalen's line-spectrum of pure carbon. With four or five 

 exceptions, all the many lines contained in the several other line-groups 

 besides these, in the same Carbon-Spectrum Table, can, however^ be readily 

 identified with lines of the oxygen line-spectrum mapped by Angstrom and 

 Thalen on the same Spectrum-Plate (" Spectres des Metalloides," Upsala 

 Nova Acta, vol. ix. 1875) with their line-spectrum of carbon, and also with 

 lines in Dr. Schuster's map {Philosophical Transactions of the Royal 

 Society, 1879) of the line-spectrum of oxygen. 



NO. 1598, VOL. 62] 



ties of all this rich domain of matter's spectroscopic radiations, 

 we must again here grieve to note, has passed away. Besides 

 his already-mentioned extremely perfect measurements of 

 "gaseous spectra," the late Prof. Smyth's published spectrum- 

 maps and spectroscopic writings comprised long descriptions too 

 of not less than five full series of measurements with high disper- 

 sion, in .southern skies, and with great magnifying powers, of the 

 dark lines of the solar spectrum.' The.se graphic solar-spectrum 

 maps and those of the "gas-spectra," and separate papers treat- 

 ing also of the oxygen-gas spectrum singly, and of the dark line 

 group " (5 " in the solar spectrum by itself,^ together form a 

 lasting store of precious materials for spectroscopic study too 

 variously instructive and often suggestive of interesting theo- 

 retical deductions from their well-recorded details, to be here 

 dealt with shortly and concisely. It is with a sense of doing only 

 very partial justice to the exceedingly high merit and scientific 

 value of those other important spectrum records and researches, 

 that as much space as could be accorded to these short notes has 

 been devoted here to pourtraying only the increasing cosmical 

 significance and the widely-spreading applications in spectro- 

 scopic astronomy, of his valuable investigations of the ordinary 

 forms of carbon spectra. In his effectual unravelling of the 

 mazy linelet systems of those familiar spectra's bands, a plain 

 and simple law of sequence in the linelets' spectrum-places was 

 disclosed, which some years later also proved the proper clue to 

 elicit order from the complex-looking linelet structures of the 

 dark absorption-bands, " A " and " B " (both due to oxygen in 

 our terrestrial atmosphere), at the red end of the solar spectrum. 

 Although those shaded groups' constructions were only perfectly 

 made known at last in 1893, by Mr. G. Higgs, of Liverpool,^ 

 from the beautiful figures of them given in his then published 

 " Photographic Studies of the Solar Spectrum," yet the drawings 

 of those bands in Prof. Smyth's Madeira and Winchester Solar 

 Spectrum Plates in 1881 and 1884, only second to Mr. Higgs' 

 photographs in their clear discriminations and accurate positions 

 of the bands' details, would have certainly afforded ripe enough 

 materials to establish at least the major portions of their simple 

 featured laws of linelet sequence by themselves, if they had been 

 searchingly examined, and carefully enough discussed and studied 

 for the purpose. 



Further examples of the same simple law of linelet intervals in 

 such " shaded " bands (where each distinguishable suite ox tier of 

 linelets exhibits simply a fixed and uniform excess or growth of 

 interval — of each suite's own amount or proper measure — in 

 every pair of adjoinmg lines, over that of its immediately pre- 

 ceding line-pair, as the rank of lines advances from the brighter 

 to the dimmer region of the shading) occur, moreover, not only 

 in the brightest, green, but also in the citron and the blue band- 

 figures, very plainly, of Prof. Smyth's full-plate "high dis- 

 persion " maps of those three most notable light-ridges in the 

 "carbon oxide" (or "cool carbon") spectrum. Another 

 interesting indication of line-systems also can be traced in his 

 full-length mapped array of the four then known low tempera- 

 ture lines of oxygen, three of which he discovered to be finely 

 triple, and to which he contributed three more just similarly 

 triple lines. Two Balmer's series of three lines each can be 

 pretty certainly distinguished in this strikingly peculiar group of 

 six mapped triplets, converging approximately to a nearly 

 common progression-head, or series-limit, at about A = 430 MM- 

 Possibly these two line-sequences which his much extended 

 range and finely inultiplied line-features of the ordinary tube- 

 spectrum of oxygen appear to show, may have been ^Iready 

 recognised and fittingly comprised by Messrs. Kaiser and Runge 

 among the many such line-series which they have found indi- 

 cations of in the spectrum-field of oxygen. But these and many 

 more such philosophical results may be looked for to be richly 

 gleaned and brought to light by coming years' discussions of the 

 minute and copious information which with Mr. T. Heath's 

 skilful assistance in their draughtsmanship and computations, is 

 lucidly unfolded in Prof. Smyth's noble works of well resolved 

 and accurately measured ranks of lines both in the solar and in 

 gaseous spectra. In those several sound and stalwart opera 



1 At Lisbon, in 1877-78, with glass prisms (the whole visible solar spec- 

 trum), Edinburgh Philosophical Transactions, vol. xxix. 1880 ; in 

 Madeira, in 1881, with a Rutherford's diffraction grating (21 special 

 " subjects," or small regions of the solar spectrum); "Madeira Spectro- 

 scopic," Edinburgh, 1882 ; and at Winchester, in 1884, wiib a Rowland's 

 diffraction grating (the whole visible spectrum, mapped thrice), Edinburgh 

 Philosophical Transactions, vol. xxxii. 1886. 



^ Edinburgh Philosophical Transactions, vols. xxx. Part i, and xxxii. 

 Part I. 



3 Proceedings of the Royal Society, vol. liv. p. 200, October 1893. 



