[The validity of the Willis Law for taxonomy]

Different explanations of Willis' law are discussed. The common view on the observed relationship between species and genus frequencies suggests that Willis' rule is artifact of two independent processes. However, actual data on similarity of the diversity patterns in different taxa contradicts with this of view and show that there are intrinsic laws regulating biological diversity. The analysis of statistical parameters of diagnostic keys reveals that distribution of differential features according to the frequency of their usage appeas to be ranked. Considering taxon size as extensional characteristic and the set of differential features as intensional characteristic one can conclude that nature of disctribution is identical in both cases. Existed data indicates that biological diversity is self-organizated and self-ordered. It allow to reconsider the conception of Natural system which could be considered as fractal with the properties of the whole composed of parts (taxa) similar by their properties to each other and to the whole. The lowest parts, units of Natural system are taxa of generic rank, their characteristics is subjected to the Willis' rule.