Although calling attention to the force of the evidence that evolution is brought about by mutation (incl. chromosome re-arrangement) followed by natural selection among the mutants, the author doubts that this is the full explanation. The problem of species origin is unsolvable, badly put, since "species," in the Linnean or taxonomic sense, is an undefined category, varying in its inclusive-ness from one group to another. The only case of a definable taxonomic difference between 2 forms is that of difference between a homozygous strain and one of its mutations. Genetical research has given information concerning only 1 category, the Rassenkreis: and here, the author''s work has shown that in Lymantria dispar, and contrary to Osborn''s opinion, the characters which differentiate subspp. within a Rassenkreis are not directly produced by the environment, and are not the beginnings of more pronounced species-characters. In L. dispar "a series of adaptations to the conditions of a series of typically different environments" (e.g., length of diapause and a period requisite for the reproductive cycle, as adapted to length of winter and summer), are caused by differences in genotype, thus supporting the conception that adaptation is brought about by (1) the origin of preadapted forms, perhaps outside the region they are fitted for, and (2) the migration of these into the environments in which their preadapted characters have positive survival value. The formation of geographical subspecies is not the beginning of speciation, for, while "the different adaptational characters" (e.g., length of diapause, intensity of pigmentation) "are of a quantitative nature, and show a plus-minus character, .... different species .... may solve one and the same adaptational problem by entirely different methods."[long dash]Speciation requires (1) permanent changes in genetic make-up (i.e., mutations) ; (2) ability of these changes to pass the test of selection; and (3) since a considerable number of the stages between egg and adult are usually affected by an evolutionary modification, every one of these changes must be compatible with life and orderly development. Assuming that the genes act (in controlling development) by controlling "reactions of definite velocities, properly in tune with each other"; and keeping in mind that experimental embryology reveals the existence of 2 types of differentiation, independent and dependent, the author suggests that mutations affect "certain developmental processes" (i.e., reaction velocities) "in a direction which lies within the ordinary range of changes which might occur within the developmental system under purely environmental influences." Orthogenetic changes are, in this view, "a necessary consequence of the way in which the genes control orderly development[long dash]a way which makes only a few directions available to mutational changes, directions which if once started and not act;d upon by counter-selection will be continued." [long dash]Occasionally mutations affect embryonic structures in a non-lethal way, thus profoundly affecting the mature organism. "Immense evolutionary effects could be brought about by changing the differential growth rates of the whole body or organ at an early point in development." Examples suggestive of this sort of change are discussed.