The structure of the band spectrum of helium

Abstract

The band spectrum of helium was first described by Goldstein and by the present writer in 1913. The accounts given of the spectrum, and of the conditions favouring its production, were closely similar, but in neither case was any analysis of its structure attempted. In view of the somewhat complex character of the spectrum in certain regions, it appeared advisable for this purpose to secure a series of high dispersion plates, while the apparently abnormal nature of many of the bands provided a further incentive to their study on the largest scale possible. Accordingly, a number of exposures were made in the first, second, and third orders of the 10-foot concave grating of the Royal College of Science, London. Of these plates some seven or eight were suitable for measurement, only one of which, however, was in the third order. The outbreak of the War prevented the completion of the series of exposures contemplated, but sufficient material was accumulated to permit of a reasonably complete investigation of the spectrum between the limits 6480 — 3350 Å. A preliminary examination of these plates was made by Prof. Fowler in 1914 (the author being on military service), and resulted in the detection of two series of double-headed bands presenting certain unusual features, of which the most striking was their distribution according to a line-series law. Nicholson, using Fowler’s measures, calculated new formulæ for these series, and concluded that they were analogous to principal series in line spectra. No other work relating to the structure of the spectrum seems to have been published. Apart from the doublet bands studied by Fowler, the most prominent features of the spectrum are a red band near λ 6400 and a green band near λ 5730, both of which appear, under moderately high dispersion, to have a quite abnormal structure. As these two bands seemed to promise more immediately interesting results than the remainder, they were the first to be studied, a band in the blue (near λ 4550), which appeared to be related to λ 6400, being subsequently included in the investigation. The results obtained in connection with these three bands form the subject of the present communication. The structure of the remainder of the spectrum, including Fowler’s doublet bands, will be discussed, it is hoped, in a future paper.