genus than for the other. Among fossil organisms, also, the more generalized the types the wider was the distribution, the separation of local genera and species following with less favorable circumstances or greater competition. Segregation multiplies species by separating groups of organic individuals, just as the ocean might form many islands from a partially submerged continent. Species are biological islands, but we do not go farther in biology than in geography by the discovery that islands must be isolated. Isolation permits evolutionary progress to be made on separate lines until the differences become of diagnostic utility to the systematist, but that isolation is responsible for the changes which bring about the divergence of characters is a deduction no more logical than that the differences between islands are due to the waters which separate them.
Too narrow zeal in the descriptive task has led many systematists to act on the assumption that the same amount of difference should everywhere receive the same systematic recognition, a method sometimes defended on the ground that all variations of form or structure indicate incipient species budding out from the parent stock, and sure to become separate groups like other now segregated types, a supposition quite unsupported by evidence. Far more rational and more secure would be the progress of systematic biology if recognition as species were limited to groups of individuals separate in nature, regard being given to the completeness of segregation rather than to the amount of difference.
It is to be admitted, of course, that when specimens from a new locality offer tangible differences from any previously known, the working systematist must describe and name them as representing new species. To crowd them into an old species by 'emending the description' or by calling them a 'variety' is to guess at an integration in advance of knowledge; while to refuse to unite 'species' which have been shown to belong to a continuous series in nature is to prefer technical fiction to biological reality. A coherent group of interbreeding individuals is the unit of evolutionary biology to which the term species finds its most proper application. The tendency of some systematists to refer also to intergrading, unsegregated subdivisions of such groups as 'species' shows how easily conventional taxonomic methods may obscure evolutionary distinctions.
Criteria of Specific Distinctness.
Species differ, of course, in the variability of their characters, but, other things being equal, the uniformity of the individuals of a species might be expressed by a ratio between the range and the facilities for interbreeding. A widespread species of sedentary animals or plants will become locally diversified; more frequent intercommunication per-
