The process of speciation within insular populations has been discussed by many authors. Hesse, Allee, and Schmidt (1937:517) list the motives for differentiation as, "Special character of insular faunae rests on the conditions common to all islands--isolation, freedom from competition, space restriction, and special insular climates." This combination of characteristics is seldom found elsewhere in nature, and as Murphy (1938:357) points out, an island is the nearest approach to a "man-controlled laboratory." Isolation of small populations is probably the most influential factor in the process of speciation in insular organisms. Lack (1947:134) writes that "in all organisms the isolation of populations is an essential preliminary to the origin of new species." Buxton (1938:265) also stresses this point with regard to the formation of species of insects in Samoa and emphasizes that evolution may occur more quickly in small populations. When mutations appear in such small and isolated populations, they have a greater chance to become fixed than do mutations in less restricted populations in a larger land mass, where such a mutation might be lost by the swamping effects of outbreedings. In addition, Wright (1931 and elsewhere) suggests the possibility of change by accidental elimination and recombination of hereditary characters in micropopulations. This mechanism could well be a factor in Micronesian bird populations, many of which possess no more than a few hundred individuals. Huxley (1938:256) emphasizes that "accidental" mutations may be perpetuated in small, isolated groups. It might be added that such changes might be either advantageous or disadvantageous to the organism concerned. Huxley (1938:263) states also that geographic isolation may promote nonadaptive differentiation, which may be caused by "colonization by a random sample" or by subsequent "preservation of nonadaptive mutations in numerically small isolated groups." Mayr (1942b:237) cites the importance of the "founder" principal for reduced variability in small populations. He points out that if the "founders" of the population carried with them only "a very small proportion of the variability of the parent population," one would expect to see divergence from the ancestral stock.