“…93, 95,102 In order to resolve this discrepancy, one phenomenologically introduces an ad hoc distribution of A that goes such as A µ , with µ a small fraction, typically around 0.3. 15,95,96,[124][125][126][127][128][129][130] This yields hole widths whose temperature dependence is dominated by a T 1+µ power law, in accord with experiment. Although the behavior of this distribution function at A f 0, where it goes to zero, is surely unrealistic, the main effect of this modification is to make the distribution rise slowly, rather than being a constant throughout most of the range of values of A.…”