Although mention is made in the previous paragraph of a spectrum there called the "Hydro-carbon" spectrum, its real origin is still unsettled. There are four carbon spectra: (1) a simple line spectrum undoubtedly due to carbon itself; (2) a band spectrum given by the base of a candle flame and often called the "Swan" spectrum after the name of its discoverer; (3) a spectrum usually known as the carbonic oxide spectrum, which also is banded; and (4) a banded spectrum due to cyanogen.

Much discussion has raged around the second, the so-called hydro-carbon spectrum, and its real origin is still so uncertain that Frost, in a recent discussion of Morehouse's

Comet, refuses to go further than to call it a "carbon" spectrum. Smithells has apparently shown it to belong to carbonic oxide, but the question is too involved for discussion here.^[1]

The question however arises, with which of the carbon spectra do the cometary spectra coincide? With regard to the brighter comets the testimony is clear; the spectrum yielded has been that given by the blue base of a candle-flame, or by a Bunsen burner—the spectrum which Lockyer terms that of hot carbon, but which Hasselberg and others consider as characteristic of a hydro-carbon, probably of acetylene. The