The continuous band spectrum of mercury vapor was observed (1) when a drop of mercury was kept rolling in a heated cylindrical quartz tube by means of a horizontal to and fro motion of the tube; (2) when the same tube in an electric oven was excited to luminosity by a high frequency electrodeless discharge. By the electrodeless method, with temperatures ranging from 160°C to over 200°C, and a controlling spark-gap of the order of 1 cm conditions were favorable for the continuous emission, while a greater temperature range was permissible in the first method. In each case the three well-known continuous bands with maximum intensities in the neighborhood of λ4600, λ3300, λ2536, were observed, while in addition, on one or two plates taken by the second method, a fourth band with maximum at about λ2650 was obtained. No evidence was found of the two continuous spectra observed by Balasse and ascribed by him to recombination,—in the one case, of the doubly ionized, in the other, of the singly ionized atom.A discussion concerning the origin of the continuous spectrum considers the relative merits of a recombination theory, in which the atom is the ultimate source, and one according to which radiation is from excited molecules. The authors conclude that the evidence is all in favor of the latter.
JOHN K. ROBERTSON, K. A. MACKINNON, and W. H. ZINN, "THE CONTINUOUS SPECTRUM OF MERCURY," J. Opt. Soc. Am. 17, 417-427 (1928)