This contribution serves to advertise the empirical solar-spectrum determinations of the oscillator strengths of 860 Fe I lines by Gurtovenko and Kostik (1981), by showing that these Kiev data contain just the lines needed in cool-star abundance analyses, and by explaining why they are so good. For details, see Rutten and Kostik (1982) and Rutten and Zwaan (1983).

Figures 1 and 2 show why the Kiev gf-values are so useful. While the precise Oxford measurements (see Simmons and Blackwell, 1982, for references) cover the low-excitation lines (Fig. la) which are mostly too violet and too strong for abundance studies, the Kiev workers have added the best lines in the solar visual, thus precisely the weaker lines of higher excitation one needs (Fig. 1b). Figure 2 demonstrates this for the solar Fe I curve of growth: the Doppler part, which is the only really useful part, is made up of Kiev lines (dots).