A large butterfly (_Kallima paralecta_), found in the East Indian forests, has often been described in its position of rest as almost exactly resembling a withered leaf; the resemblance in colour being aided by the markings which imitate the venation of a leaf. These markings are composed of two parts, the upper of which is on the fore-wings, while the lower one is on the hind wings. The butterfly when at rest must therefore keep the wings in such a position that the two parts of each marking exactly correspond, for otherwise the character would be valueless; and as a matter of fact the wings are held in the appropriate position, although the butterfly is of course unconscious of what it is doing. Hence a mechanism must exist in the insect’s brain which compels it to assume this attitude, and it is clear that the mechanism cannot have been developed before the peculiar manner of holding the wings became advantageous to the butterfly, viz. before the similarity to a leaf had made its first appearance. Conversely, this latter resemblance could not develope before the butterfly had gained the habit of holding its wings in the appropriate position. Both characters must therefore have come into existence simultaneously, and must have undergone increase side by side: the marking progressing from an imperfect to a very close similarity, while the position of the wings gradually approached the attitude which was exactly appropriate. The development of certain minute structural elements of the central nervous system, and the appropriate distribution of colouring matter on the wings, must have taken place simultaneously, and only those individuals have been selected to continue the species which possessed the favourable variations in both these directions. It is, however, obvious that sexual reproduction will readily afford such combinations of required characters, for by its means the most diverse features are continually united in the same individual, and this seems to me to be one of its most important results. I do not know what meaning can be attributed to sexual reproduction other than the creation of hereditary individual characters to form the material upon which natural selection may work. Sexual reproduction is so universal in all classes of multicellular organisms, and nature deviates so rarely from it, that it must necessarily be of pre-eminent importance. If it be true that new species are produced by processes of selection, it follows that the development of the whole organic world depends on these processes, and the part that amphigony has to play in nature, by rendering selection possible among multicellular organisms, is not only important, but of the very highest imaginable importance.