Iterating Bonferroni bounds (Q1070639)

An elementary method is presented for deriving a lower bound from an upper one (and vice versa) in Bonferroni type inequalities. The method, of course, does not transform a best lower inequality into a best upper inequality, but, in some cases, improvements do result. A specific numerical example is included in which improvements on the original Bonferroni bounds are demonstrated. However, there are several inequalities superior to the original Bonferroni bounds to which the new bounds are not always favourable; see chapter 1 in the reviewer's book ''The asymptotic theory of extreme order statistics.'' (1978; Zbl 0381.62039) for a large number of Bonferroni type inequalities. (The second edition of this book, to be published by R. Krieger Publ., Inc., has an extensive list of additional references.)