from D to G, with a very slight condensation evidently due to the familiar green band at wave-length 515, and a very faint trace of the yellow band. Keeler examining the spectrum again on January 29, 1893, after an anomalous brightening of the comet on January 16, saw a very faint suspicion of the green band of which he could never be certain.[1]
The spectrum of a strongly moonlit sky was exactly similar to, but fainter than, that of the comet. It would appear then that in this case they were dealing with an object almost wholly illuminated by reflected light. It had been suggested that this peculiar comet might have been formed by the collision of two asteroids, but, as Keeler pointed out, the observed spectrum negatives this hypothesis; such a cataclysmic birth would involve the production of a bright line, or banded, spectrum.
The spectrum of Rordame's Comet (1893, ii.) was very fully investigated, photographically, by Campbell,[2] who measured 28 bright lines, 14 of which were found to correspond with lines and bands given by Kayser and Runge in the spectra of carbon and cyanogen.
Visual observations by Keeler[3] showed that the spectrum was a beautiful example of the hydro-carbon type, and he called special attention to the sharpness of the bands at their less refrangible edges, a phenomenon not usually seen in cometary spectra.
The spectrum of the Comet 1902 (ii.) (Perrine-Grigg's) was of special interest, because it was in photographing this that Count De La Baume-Pluvinel first employed the short- focus prismatic camera, now recognised as a sine quâ non in the photographic recording of the spectra of faint comets.
Experiments with larger instruments having proved fruitless, Baume-Pluvinel built up a camera[4] wherein the focal length of the objective was but four times the aperture, and to which he applied a prism of 20° refracting angle. Of
