Generally speaking, it appears that the spectra of these bodies indicate carbon—in some form—as the principal constituent.
As to the particular form of carbon, there is still a considerable doubt, so much that, in describing the spectrum of Morehouse’s comet, Professor Frost says (Astrophysical Journal, xxix., p. 59, 1909):—“We avoid the still unsettled question of the ‘carbon’ bands (of the so-called ‘Swan’ spectrum) which have been so often ascribed to a hydrocarbon, specifically acetylene, and we use for them the simple designation ‘carbon.’”
In addition to this “carbon” there is the cyanogen spectrum present in most cases.
Sodium and iron have been detected in the spectra of some few comets, e.g. Wells (1882, ii.), whilst Holmes (1892) showed only continuous spectrum.
An interesting suggestion is made by Newall, namely, that the spectrum is not indicative of the comet’s composition, but of that of the medium through which the body passes. Thus the persistent identification of the cyanogen bands in cometary spectra is attributed, primarily, to the “heating up” of cyanogen existing, free, in circum-solar space.
Till 1907 most of the cometary spectrograms showed only the “carbon” and cyanogen radiations, but in Daniel’s comet of that year, and in Morehouse’s of 1909, other lines were detected for which origins have not, as yet, been found.
Thus, some form of carbon + unknown + (occasionally) sodium and iron seems to sum up our present knowledge of cometary composition.