CALCULATION OF INTENSITY RATIOS OF OBSERVED INFRARED [Fe II] LINES

Two recent observational studies of the [Fe II] lambda 12567/lambda 16435 line ratio by Smith & Hartigan and Rodriguez-Ardila et al. have suggested that the available theoretical A-values could be incorrect to 10%-40%. We have carried out an extensive configuration interaction calculation of [Fe II] lines to investigate this claim, as well as the variability in observed line ratios for lambda 8617/lambda 9052 and lambda 8892/lambda 9227 of Dennefeld. For these transitions, we are generally in good agreement with the results of Nussbaumer & Storey, less so with those of Quinet et al. In comparison, the ratios derived from observations appear either to be less secure, or other factors influence those results.