References and Notes for Indium ( In )
B69 K. S. Bhatia, Ph. D. Thesis, Univ. British Columbia (1969) and K. S. Bhatia, J. Phys. B 11, 2421 (1978). Although this thesis has In III and In IV as the main subjects, new wavelength measurements for In II lines are included. Bhatia's wavenumber for the 5s5p 3P°2-5s6s 3S1 line, 48095.0 cm-1, has a Ritz-principle disagreement of about 1.2 cm-1 with determinations of the 3P° intervals given by his measurements in the 1607-1977 Å region. We have not used the observed wavenumber of this line in determining the level value and, instead, give a Ritz-principle wavelength of 2078.608 Å for the line. Bhatia's experimental procedures are described in his published paper on In III: K. S. Bhatia, J. Phys. B 11, 2421 (1978).
C00 L. J. Curtis, Phys. Scr. 62, 31 (2000).
C52 H. E. Clearman, J. Opt. Soc. Am. 42, 373 (1952).
DMZ53 G. V. Deverall, K. W. Meissner, and G. J. Zissis, Phys. Rev. 91, 297 (1953).
FW96 J. R. Fuhr and W. L. Wiese, NIST Atomic Transition Probability Tables, CRC Handbook of Chemistry & Physics, 77th Edition, D. R. Lide, Ed., CRC Press, Inc., Boca Raton, FL (1996).
G54 W. R. S. Garton, Proc. Phys. Soc. (London) A 67, 864 (1954); W. R. S. Garton, W. H. Parkinson, and E. M. Reeves, Can. J. Phys. 44, 1745 (1966). Garton's wavelength measurements for the four 5s25p 2P°-5s5p2 2P autoionization-broadened transitions have Ritz-principle consistency of a few cm-1. This multiplet was first identified in [C52].
JL67 I. Johansson and U. Litzén, Ark. Fys. 34, 573 (1967).
LS31 R. J. Lang and R. A. Sawyer, Z. Phys. 71, 453 (1931). The wavelength 2079.26 Å tabulated for the important 5s5p 3P°2-5s6s 3S1 line in this paper is the vacuum wavelength, contrary to the column heading.
M00 D. C. Morton, Astrophys. J. Suppl. Ser. 130, 403 (2000).
MCS75 W. F. Meggers, C. H. Corliss, and B. F. Scribner, Natl. Bur. Stand. (U.S.), Monogr. 145 (1975).
ND81 J. H. M. Neijzen and A. Donszelmann, Physica 106C, 271 (1981).
P38 F. Paschen, Ann. Physik (5) 32, 148 (1938). We have adjusted the values of the 5s5p2 4P levels to the more accurate 5s25p 2P° ground term fine-structure splitting given by the measurements in [DMZ53].
PC38 F. Paschen and J. S. Campbell, Ann. Phys. (5) 31, 29 (1938). Wavelengths of resolved hyperfine components are tabulated for many of the In II lines in this paper, together with the corresponding hfs center-of-gravity wavenumber for each line. The wavelengths for such lines quoted here were derived by converting these center-of-gravity wavenumbers back to air wavelengths using, in effect, the same (now superseded) air-dispersion formula as Paschen and Campbell.
SM02 J. E. Sansonetti and W. C. Martin (this work). Paschen and Campbell observed the spectrum from 2078 to 9246 Å, but adopted a value of 43349 cm-1 for the 5s2 1S0-5s5p 3P°1 separation, as determined earlier by Lang and Sawyer [LS31]. The value for the 5s2 1S0-5s5p 3P°1 wavenumber obtained from Bhatia's more recent measurement [B69] of the corresponding line is 43351.00 cm-1, in good agreement with the value 43350.85 cm-1 obtained for the wavenumber of this line by Paschen and Campbell. We have thus adopted a value of 43350.9 cm-1 for the 5s5p 3P°1 level and used this value together with measurements of 5s5p 3P°-5s6s 3S, 5s5d 3D, 5p2 3P, and 5p2 1D wavelengths (1607-1977 Å) from Bhatia to redetermine the 5s5p 3P°0, 3P°2, and 5s6s 3S1 levels. The Ritz-principle consistency of six determinations of 5s5p 3P° level separations from Bhatia's measurements indicates a wavenumber uncertainty of the order of 0.3 cm-1, corresponding to a wavelength error of order 0.010 Å at 1800 Å. Our new position for the 5s6s 3S1 level is 4.37 cm-1 above the value in Atomic Energy Levels [M58], which was taken from [PC38]. Since Paschen and Campbell evaluated the 5s5p 1P°1 level and all levels above the 5s6s 3S1 level with respect to the 5s6s 3S1 level, all levels given to two decimal places by Moore [M58] should be increased by 4.37 cm-1. The ionization energy is also increased by 4 cm-1. The corrected levels can be used to obtain Ritz vacuum UV wavelengths for In II having much greater accuracy than the measured values of Lang and Sawyer and, for wavelengths below about 1600 Å, greater accuracy than those given by Bhatia. Thus, for example, the new Ritz wavelength of 734.773 Å for the 5s2 1S0-5s8p 1P°1 line should be accurate to about 0.002 Å, whereas the values from Bhatia [B69] and Lang and Sawyer [LS31] are longer than this Ritz value by 0.005 and 0.17 Å, respectively.