As was shown in the first part of the paper, of all possible
populations, the most difficult to sample is a perfectly uniform
distribution, where each age fraction is of the same size. It is
sufficient to date k grains, according to Figure 3, Table 1
or the web-form [9], to reduce f and p to the desired
levels. However, most naturally-occuring populations differ from the
worst-case distribution and it is less likely that statistically
significant fractions of such populations might be missed in a sample.
As a consequence, fewer grains need dating to achieve the same levels
of f and p. A number of numerically-generated random populations are
discussed next, followed by some real detrital age-spectra.