1. Similarity seems at first sight easy to determine—as though we had only to open our eyes; yet by this elementary procedure we hardly pass beyond the level of generic images, and there is risk of falling into many errors. It is necessary to penetrate into resemblances deeper than the superficial; and here is the first degree of complexity. Buffon observed that “the horse and the donkey, which are distinct species, resemble each other more than the water spaniel and the harrier, which are of the same species.” The facts which our contemporaries denote by the name of _polymorphism_, entirely baffle the criterion of similarity. Not to speak of the obvious difference between the larva and the perfect insect, the caterpillar and the butterfly, or between the males, females, and neuters of bees, ants, and termites; there are cases in which the disparity between the two sexes is so great that the male and the female, taken respectively as two different creatures, have been classified in distinct _genera_, and even _orders_: e. g., the lampyris or glow-worm, _Lernea_, and many others. The character of the resemblance is thus too often vague, sometimes deceptive, nearly always inadequate: it follows that we must resort to the other, to filiation. 2. This, the physiological criterion, again appears to leave no opening for equivocation, since it can be materially stated. Generally speaking, one is imbued with the notion that children resemble their parents, that the immediate product is the reproduction of the type of the progenitors. But the alternating generations (metagenesis, geneagenesis) discovered in the course of the present century, show that this conception is too simple, and often fallacious. This mode of reproduction is by no means rare; we meet with it among a great number of the lower plants, infusoria, worms, and even insects. “The dominating fact in the reproduction of all these creatures, is that a _sexual_ being, of definite form, gives birth to _a-sexual_ beings which do not resemble it, but which in their turn produce by a sort of budding, or by fission of their bodies, the sexual creatures similar to those from which they issued.” Vogt, accordingly, in his definition of species, is forced to include the case of alternate generation by saying: “Species is the reunion of all the individuals that originate from the same parents, and are in themselves, _or in their descendants_, similar to their primordial ancestors.” In brief, the general notion of species depends upon two ideas, complex notwithstanding their apparent simplicity, fluctuating in spite of their apparent precision.