Page:Popular Science Monthly Volume 64.djvu/452

448 more uniform progress. A single species may have as great a variety of characters as a dozen related groups which have been segregated. Two species may be quite distinct and yet differ much less than the connected extremes of another. That a species differs in different parts of its range does not necessarily mean that a subdivision will take place; it means merely that characters are originating more rapidly than they spread over the whole species. The integrity of a species is not destroyed by 'inconstancy' of characters, but because geographical or other barriers make a gap in the series.

The failure of the extremes of a widely distributed species to breed when brought together does not prove the attainment of specific distinctness, nor the approach of it, since internal diversity does not weaken the species, but is an evolutionary advantage, and both extremes may continue to cross freely with the connecting forms, which constitute the bulk of the species. Neither does the power to form fertile hybrids prove that two species occupying distinct ranges are one. Faith in such criteria is simply a remnant of the pre-evolutionary theory of the separate creation of species. The only way to ascertain that two groups of organisms are separate species is to find the gap between them. Whether they will breed together or not, and whether the hybrids are fertile and vigorous, or weak, sterile and aberrant, may indicate the period and degree of divergence of the types crossed, but affords absolutely no evidence as to whether the series to which they belong in nature are continuous or interrupted. Specific distinctness is a question much more geographical than evolutionary. Evolution continues whether the species is divided or not; the divergence of the parts is rendered possible by the cessation of the interbreeding which would otherwise maintain the coherence and relative uniformity of the undivided group.

The systematist 'separates' species because they are 'different,' but the evolutionary significance of species does not appear from formal descriptions of these biological islands; it lies in the fact that isolated groups of organic individuals universally acquire diagnostic differences. Isolation has furnished millions of these tests of the universality of biological motion, but it does not cause the motion. Evolution is independent of isolation, and without it has often brought about great diversity of form and structure, as witness the dissimilar sexes, castes, dimorphic and alternating generations of many species of plants and animals. Without evolutionary progress there would have been no species as we now know them, but the causes of the segregation of species are not causes of evolution; segregation merely permits this universal tendency to become more manifest. If it should be found that evolutionary divergences sometimes assist natural selection or