“…The temporal dynamics of several lines is shown in Figure 2. While the temporal decay of some lines can be fitted, at least roughly, by an exponential function (Curtis and Theodosiou, 1989) b Laser-induced fluorescence from sputtered metal vapors (Hannaford and Lowe, 1983) c Numerical Coulomb-like approximation (Lindgård et al, 1980) d Relativistic Hartree-Fock calculation with account for core polarisation effects (Migda lek and Baylis, 1978) e Relativistic Hartree-Fock calculation with model potential accounting for exchange and core polarisation (Migda lek, 1978) f Relativistic Hartree-Fock calculation with model potential and core polarisation (Migda lek and Baylis, 1979) g Calculation using the level-crossing measurement of lifetimes and configuration coupling coefficient deduced by fitting other experimental measurements of lifetimes and relative oscillator strengths (Siefart et al, 1974) h Atomic absorption measurements on flames (Lvov, 1970) i Atomic-beam absorption (Bell and Tubbs, 1970) j Critical survey of experimentally-determined oscillator strengths (Corliss, 1970) k Curves of growth in absorption measurement (Moise, 1966) l Rozhdestvenskii's hook method (Slavenas, 1966) m Atomic-beam absorption (Bell et al, 1958) n Arc emission measurements (Allen and Asaad, 1957) a Time profile demonstrates significant deviation from the exponential decay b The decay curve has essentially non-exponential form with a plateau or secondary maxima; τ value is absent or roughly approximate several lines display essentially non-exponential behavior including some "plateaux" or even secondary maxima at 35-50 µs after the laser shot. Their decay time, T , values are therefore estimated in Table 3 in a rough approximation; it is seen from this table that for essentially non-exponential decays the uncertainity ∆T is of the same order of magnitude as T itself.…”