Abstract
Summary: There are several kinds of criteria for determining whether an average family of insects is comparable to an average family of mammals. Different combinations of criteria are appropriate for different purposes and give different results. A new metataxonomic criterion is explored for all phyla of animals and plants and for various subordinate taxa. Differences in it can be caused by sampling error, different degrees of splitting, and different patterns of evolution; these causes are more or less confounded but indicate the domain of the criterion's applicability. Body weight and other factors affect the number of species in genera and families. It is not yet known to what extent boundaries between taxa correspond to real subcontinuities in properties of the organisms. We should know what characters are, or have been, important to an organism, even for angiosperms, where little research has been directed at the problem as formulated here.