Although the currently most popular concept of species and speciation is described as a "synthetic" or "biological" one, it is a population-genetic concept in its essence, since it was in its terms that the synthesis of Darwinian natural selection and Mendelian genetics was realized (Chetverikov, 1926; Fisher, 1930; Haldane, 1932; Dobzhansky, 1937; Huxley, 1942). The basic postulate of this viewpoint is well-known: evolution is a continuous adaptive process of substitution of alleles of polymorphic loci, resulting in gradual divergence of geographically separated populations up to the species level. Reproductive isolation, an important criterion of species, is viewed in this model as only a by-product of such differentiation. However, the results of recent studies in biochemical population genetics and comparative genetics of species permit another hypothesis to be formulated: speciation is not a gradual process of changes expressed in terms of gene frequency dynamics but a qualitatively different phenomenon related to a rapid rearrangement of that part of the genome which is marked by the functionally most important monomorphic loci. The purpose of the present paper is to argue in favor of this viewpoint, resting on the results of our recent studies of natural and experimental populations. Our model of speciation is consonant with Carson's (1975) concepts although it is based on other facts and observations. The following points will be considered: 1. historically formed population systems and their potential evolutionary transformation; 2. genetic monomorphism of species as a real natural phenomenon; and 3. speciation as a saltational process related to rearrangement of the monomorphic part of the genome. Since many publications in the Russian language are apparently unfamiliar to the readers of Evolution I shall deviate slightly from the traditional presentation of material and give a broader account of several earlier works.