The first observations of curium spectra emitted by an arc resulted in a list of about 200 lines between 251.6 and 500 nm (51.1; 56.1) and isotope shifts 244Cm - 242Cm of 148 lines between 305 and 525 nm (56.1). Atomic beam resonance method was used for measuring the g value of four low levels and this indicated the ground term to be 5f76d 7s2 9D (59.5). All energy levels of this term were first obtained by Worden et al. at Livermore and Berkeley and reported in (62.4). The energy of the lowest even level was just mentionned in a conference abstract (67.6).
Extensive observations of the curium emission spectra were made possible with electrodeless lampe as sources on the Paschen-Runge spectrograph at the Argonne National Laboratory in the range from 240 to 910 nm. Supplementary observations were made with a 3.4 m Ebert spectrograph and the number of lines (Cm I and Cm II) available for the first extended term analysis was 13260 (76.1). The lines of the atom and of the ions were discriminated by varying the excitation conditions of the electrodeless tube and over 6800 lines were attributed to Cm I. Their analysis, supported by Zeeman effect measurements, produced 335 odd levels and 348 even levels, the attribution to configurations and LS terms being well extablished below 16000 cm-1 (76.7). The infrared emission spectrum was observed at Laboratoire Aimé Cotton by means of Fourier transform spectrometry (FTS) (76.12).
After 1976, investigations have been focused on:
a) extended isotope shift measurements for about 6000 lines at Livermore (80.3),
b) visible FTS recordings at the Kitt Peak National Observatory.
Consequently, the earlier results were checked, 311 of the 335 odd levels were confirmed and 16 even levels of (76.7) were rejected. Then, 174 new odd and 90 new even levels were found and the number of configurations identified increased from 12 to 18, 5f77s29p being a possible other one (88.4). The 244 Cm I levels reported here are from this recent publication. They are characterized by 246 Cm - 244 Cm isotope shifts and g factors with very few exceptions.
The interpretation of the levels is supported empirically by the shifts and by the observed intensities of the lines. Only two configurations have been calculated theoretically by means of the Slater-Condon method. First calculations of 5f87s2 by Crosswhite are mentionned in (76.7). In a survey of 5fN7s2 configurations, 6 additional levels were attributed to 5f8 7s2 (80.4) . These labels have been changed into 5f86d7s for two of them. The sub-configuration 5f7(8S,6P,6D,6I)7s27p was calculated in (76.7) with a good agreement between experiment and theory for both energies and g-factors of the first six levels.
Parallel researches have been led in USSR at the I.V. Kurchatov Institute and resulted in a wavelength list of 6774 lines (79.5) and isotope shifts for 97 lines (79.4).