The first observations of americium spectra by Tomkins and Fred (49.3) proved the relative simplicity of Am I and settled the analogy with Eu I in the lanthanide sequence, the lowest configurations being built on an isolated 5f7 8S07/2 core level. The initial steps of the Am I analysis were described by Fred and Tomkins (57.1): Over 3000 lines of Am I and Am II were measured on the 9.15 m Paschen-Runge spectrograph at Argonne National Laboratory and 112 of them were classified as transitions between 12 odd and 39 even levels. Several kinds of light sources were used and the line shapes of americium lines under various conditions of excitation of electrodeless discharge tubes were compared in (57.1). Due to the lack of theoretical analysis of the excited configurations, only ten levels were interpreted.
The analysis of Am I was resumed two decades later by using Fourier transform spectra at high resolution recorded at the Kitt Peak National Observatory. The present status of the analysis is from B.V. Pulliam (unp.25), whose unpublished work was briefly described in (83.6). At the present time 122 odd and 74 even levels are known and the determination of A and B hfs constants supports the line assignments. These constants which are tabulated here are accurate to 0.0005 cm-1 for the A's and to 0.001 cm-1 for the B's, unless the decimal point is missing (accuracy 0.01 cm-1). About two dozens of tentative levels, partly from (57.1), have been discarded since they need further confirmation. Although the parametric interpretation is still limited to ab initio predictions for 5f77s2 (80.4) the lowest levels of both parities could be labelled by analogy with neighbouring spectra. Similarities are obvious between Am I 5f66d7s2 and Pu II 5f66d, and between Am I and Cm II for 5f76d7s (unp.9). Isotope shift data support the identification of a few 5f67s27p and 5f76d2 levels, but these isotope shifts are still tentative for many upper levels. The 5f77s7p levels are better grouped in jj than in LS coupling scheme.
High resolution investigations of hyperfine structures were performed by using a hollow cathode light source and a Fabry-Perot interferometer (61.2). The atomic beam magnetic resonance method was applied to the 241Am I ground state by Marrus et al. (60.7) . Improved measurements for 241Am and 242Am are reported by Armstrong and Marrus (66.5). Values for the ground state of 241Am complete those from optical measurements (unp.25).