Page:The story of the comets.djvu/231

XII. wave-lengths of the edges of the three principal bands of this spectrum are given by Kayser and Runge as 5635 for the band in the yellow, 5165 for that in the green, and 4737 for that in the blue respectively. But the complete spectrum contains 5 bands, one in the red at 6188 and one in the violet at 4311. These have not often been detected in the spectra of comets, but in the case of one or two of the brightest (the first instance being that of Coggia's Comet in 1874 when Secchi recognised their presence) they have been made out. It appears, therefore, that it is only a question of brightness, and that the whole series may be expected to be shown by any comet of sufficient brilliancy.

In the case of some of the earlier and fainter comets the bright bands were recorded by some observers as coincident with that spectrum of carbon called by Lockyer that of cool carbon, by others that of carbonic oxide, in distinction to the hydro-carbon spectrum referred to above. The wave-lengths of the bands in this spectrum are as follows:—yellow, 5609; green, 5198; blue, 4834. It is not yet absolutely certain whether these observations are to be accepted. It appears probable in some cases, if not in all, that the faintness of the spectrum prevented any accurate measurement of the cometary bands, and that being compared with the carbonic oxide spectrum with only low dispersion, and a close approximation being noted, it was assumed that the correspondence was exact. Thus Christie observing the spectrum of Hartwig's Comet on Oct. 7, 1880, measured the green band as at wave- length 5201, or as being the band of the carbonic oxide spectrum: four days later he measured the same band as at 5169, or as being the band of the hydro-carbon spectrum; and on Oct. 12 the observation was repeated. A comparison of the individual measures renders it all but certain that there was no change in the place of the cometary band, and that the difference in the positions recorded was due entirely to the difficulty of securing an accurate measurement of an object so faint and diffused.