320 



NATURE 



{Atigust 5, 1880 



CARBON AND CARBON-COMPOUNDS 



THE wayward and inconstant train of coloured light- 

 bands that spectroscopists have noted and dis- 

 tinguished in the spectra of various carbon-compounds in 

 flames and gas-vacuum tubes are as yet far from having 

 all received their full and appropriate interpretations. 

 The extent to which they abound as impurities in almost 

 all spectral vacuum-tubes is a common observation, and 

 in a survey of this kind, aiming at no systematic explora- 

 tion, of a variety of end-on vacuum-tubes in the large and 

 perfect spectroscope erected by Prof. Piazzi Smyth for 

 the examination of auroraj, I have had from time to time, 

 at his kind invitation, excellent opportunities for dis- 

 criminating some of the component groups and clusters 

 of the carbon-denoting series from each other pretty 

 clearly. 



Among the least alterable and changeful in its appear- 

 ance of these coloured ranks is the five-tongued spectrum 

 of wedge-like bands best seen in the end-on prismatic 

 view of a coal-gas blow-pipe flame. Its bands have 

 shaft-lines at the edge and on their fading slopes, with 

 the exception of the last or violet one, just including 

 within its bright edge the solar line of Fraunhofer's 

 spectrum, G. This has a fine-line precursor, nearly coin- 

 cident with Hy, and a faint haze-band preceding it. Close 

 to the place of b-^ in the solar spectrum appears the briglit 

 edge or chief shaft-line of the green band, fitly styled the 

 "green giant," as it is the real .Anak of the coal-gas flame- 

 spectrum. Its less refrangible similitudes in the yellow- 

 green and orange-red arc quite subordinate groups, the 

 latter being only discernible in spectroscopes of large 

 aperture and of very great transparency. The fifth finger 

 of this spectral gauntlet is a blue band, or quintett of five 

 close lines pretty equally spaced and pretty equal in 

 brightness, with little haze between them, lying once or 

 twice its own breadth on the more refrangible side from 

 H/3 (F.). The frontispiece of Watt's " Index of Spectra" 

 contains a figure of this spectrum ; and wave-length 

 positions and symbols and descriptions of its groups 

 are given in the body of the work, under the title 

 " Carbon, Spectrum I." o, y, 8, c, / (/3 and r) careiif) are 

 the five famihar potentates of the blow-pipe flame ; but 

 the two line-bands f, 6, one on each side of f, added in 

 the figure and in the text of Watt's "carbon-spectrum I.," 

 are not visible in the blow-pipe flame-spectrum. Along 

 with a similar ultra-violet cluster just following H K in 

 the solar spectrum, they form a triumvirate, the spectral 

 origin of which Professors Liveing and Dewar have re- 

 cently affirmed to be cyanogen. A reason to question the 

 correctness, however, of Messrs. Liveing' s and Dewar's 

 surmise presented itself to me in my examination of the 

 end-on tubes by the spectacle of the six-lined violet 

 cluster 6 rearing itself, without any accompaniment of 

 its blue associate f, into extraordinary magnificence in a 

 Marsh-gas tube. The grey or ultra-violet member of the 

 trio was indeed weakly discernible at the same time; and 

 in just this relative brightness and condition of extreme 

 isolation from every other spectral feature I have recently 

 observed these two violet and ultra-violet line-clusters in 

 the blue flame part of the arc between particularly pure 

 carbon poles in the Brush's or Anglo-American Company's 

 electric light. 



Another reason for suspecting multiplicity of form in 

 the carbon-spectrum by itself occurred to me in an 

 examination of the spectrum of cyanogen in an end-on 

 tube. A perfect counterpart, it is well known, of the 

 blow-pipe flame spectrum is producible by the induction- 

 spark in vacuum-tubes of olcfiant gas. Accompanying it 

 however is another spectrum which in its fullest purity 

 and intensity is equally well known to be produced by a 

 weak induction-spark in tubes of carbonic oxide and 

 carbonic acid gas. The blue quintett and the violet G- 

 bind are wanting in this spectrum. The edges of the 

 green, citron and orange-red bands are displaced, and 



these bands are devoid of shaft-lines, being composed 

 entirely of haze and fine linelets which smoothly shade 

 them off. The defiant gas and " carbonic oxide " spectra 

 mingle together, usually in divers proportions in the 

 carbon-impurities of gas- vacuum tubes. 



Two cyanogen tubes (one of them of hardest glass) 

 prepared by M. Salleron betrayed alike only the smallest 

 trace of hydrogen by its red line, when they were lighted 

 up by the induction-coil. Aqueous and atmospheric 

 oxygen may therefore be presumed to have been 

 pretty completely expurgated from these tubes, and the 

 gas which charged them to have been an exceptionally 

 pure compound of nitrogen and carbon. Far brighter, 

 notwithstanding this, than in any other vacuum-tube, the 

 smooth-shaded " carbon-oxide " bands made their appear- 

 ance ; and equally splendid with them was the close- 

 ribbed red and yellow fluting forming the less-refrangible 

 part of the spectrum, figured and described by Angstrom 

 and Thalen as that of " nitric oxide." The coincidence 

 with the same spectrum of the bright cyanogen-tube lines 

 in the blue and violet spectral regions was not closely 

 examined ; but as far surpassing in brightness the red-end 

 view of it obtained in any other nitrogen-holding vacuum- 

 tube (nitric oxide itself not excepted), the rasp-like ridges 

 of the so-called nitric oxide spectrum were immediately 

 measured with great care and accuracy. Angstrom's 

 positions and tableau (exactly reproducing that of Plucker 

 and Hittorf) of this region were completely verified ; and 

 the discussion of the vrell-based determinations left no 

 doubt that while a simple order reigns sensibly among the 

 small linelet features of each separate ridge, the ridges 

 have no perceptible connection with each other or with 

 tlie linelet-intervals upon them in the pitch of their wave- 

 iVcquencics, although they follow each other closely in a 

 gradually narrowing succession. In the rest of the 

 nitrogen-spectrum, where the ridge-intervals are much 

 wider, it is again not possible to trace between the ridges 

 any simple wave-period connection. 



Were 1 not from these measures, and from the fore- 

 going considerations disposed to regard shaded spectra! 

 bands as independent systems of vibration, indicating 

 m05t probably particular atomic groupings in a molecule,. 

 I should have beheld with some surprise the complete and 

 thorough metamorphosis shown me by Mr. Lockyer since 

 the above particulars were noted, which the smooth-banded 

 "carbon-oxide" spectrum undergoes by introducing a 

 condensing-jar, or better, a jar and air-break, into_ the 

 circuit of the induction-coil. The smooth shadings disap- 

 pear, the shaft-lines, the " Anak and the sons of Anak " of 

 the olefiant-gas or blowpipe-flame spectrum make their 

 appearance in their place ; even the blue quintett of that 

 spectrum comes forth from its hiding-place ; and, as far 

 as I could examine the spectral appearance of the car- 

 bonic-oxide tube in the now condensed discharge with 

 complete precision, the whole blow-pipe flame, or so-called 

 "hydro-carbon" spectrum, is perfectly reproduced. If we 

 cannot admit, as 1 think that the cyanogen-tube experi- 

 ment forbids us to do, that a chemical transformation has 

 taken place, then we must acknowledge that among the 

 forms which the spectrum of carbon is capable of assum- 

 ing, there may, by subdivision of its molecule into separate 

 vibrating systems, exist not one, but as many difterent 

 " low-temperature " spectra of that Briareus-like, hundred- 

 fiited, or Proteus-like, hundred-visaged element, as the 

 electric discharge is capable of dividing its evidently 

 complex gaseous molecule into separate spectroscopically 

 individual groups. A. S. Herschel 



PHYSICS WITHOUT APPARATUS 



I. 



TT is almost a proverb in science that some of the 



i greatest discoveries have been made by the most 



simple means. It is equally true that almost all the 



