tially in the union or conjugation of two diverse cells (sperm and germ cell) to produce one cell (ovum). Now, in the lowest forms of sexual reproduction, among unicelled organisms, the conjugating cells are not perceptibly different; so that the element of diversity in the conjugating cells may be eliminated from the essential conditions of this mode. In parthenogenetic reproduction of female offspring, as in the case of moths and phyllopods, we have the other element, i, e., the necessity of two cells, eliminated; so that there remains nothing which is absolutely essential.
(c.) Facts which hear on the First Step, viz., Derivation of NonSexual Modes from Ordinary Processes of Growth.—The transition between the lowest form of non-sexual reproduction, viz., fission, and ordinary growth, is so obvious that it is hardly necessary to insist on it. A single cell divides itself into two; each half grows, and again divides itself into two, and so on. Now, if the cells cohere, we call it growth; if they separate, we call it reproduction. Again: a mass of cells grows by continued cell-multiplication, as above. Finally, the increasing mass or community becomes too large to be managed well from one center; it therefore divides itself into two masses or communities, each of which continues to grow as before. It is plain that a slight difference only in the degree of cohesion determines whether the same process be called growth or reproduction.
Thus we have shown the easy gradation, and therefore the probable derivation, of the highest mode of sexual reproduction—the unisexual—from the ordinary processes of growth, through the different grades of asexual and bisexual reproduction. The derivation of different modes of sexual reproduction from each other will not, I think, be questioned. Still clearer is the fact that non-sexual reproduction is but a modification of the ordinary process of growth. The only place where there is any gap is between the asexual and the sexual modes. Throughout growth and non-sexual modes of reproduction we have everywhere only cell-multiplication—everywhere we have division of one to form two: in sexual reproduction, on the other hand, we have the contrary process, viz., the union of two to form one. Yet this gap is certainly partly filled by the larval reproduction of aphids, by those cases of parthenogenesis in which unfertilized ovules produce females, and those cases of true sexual generation in which the conjugating cells are similar.
5. Outline of Probable History of the Process.—The gradual evolution of the higher forms of sexual reproduction probably took several different roads. There is little doubt that in some cases sexual reproduction in its simplest form was reached at a very early period. It is probable, for example, that in very early times unicelled organisms multiplying usually by fission (asexual) from time to time conjugated (sexual). The simple form of sexual reproduction thus reached was afterward perfected. But it is also probable, nay, judging from the
