regularly recurring form of heterogenesis; and it is interesting to see how survival of the fittest has established a mixed form. In the spring, as well as in the autumn, there is in some cases a formation of resting or winter eggs; and evidently these provide against the killing off of the whole population by summer drought. Meanwhile, by ordinary males and females there is a production of summer eggs adapted to meet the incident of drying up by drought and subsequent re-supply of water. And all along successive generations of parthenogenetic females effect a rapid multiplication as long as conditions permit. Since life and growth are impeded or arrested not by lack of food only, but by other unfavourable conditions, we may understand how change in one or more of these may set up one or other form of genesis, and how the mixture of them may cause a mixed mode of multiplication which, originally initiated by external causes, becomes by inheritance and selection a trait of the species.[33] And then in proof that external causes initiate these peculiarities, we have the fact that in certain _Daphnideæ_ "which live in places where existence and parthenogenesis are possible throughout the year, the sexual period has disappeared:" there are no males. Passing now to animals which multiply by homogenesis--animals in which the whole product of a fertilized germ aggregates round a single centre or axis instead of round many centres or axes--we see, as before, that so long as the conditions allow rapid increase in the mass of this germ-product, the formation of new individuals by gamogenesis does not take place. Only when growth is declining in relative rate, do perfect sperm-cells and germ-cells begin to appear; and the fullest activity of the reproductive function arises as growth ceases: speaking generally, at least; for though this relation is tolerably definite in the highest orders of animals which multiply by gamogenesis, it is less definite in the lower orders. This admission does not militate against the hypothesis, as it seems to do; for the indefiniteness of the relation occurs where the limit of growth is comparatively indefinite. We saw (§ 46) that among active, hot-blooded creatures, such as mammals and birds, the inevitable balancing of assimilation by expenditure establishes, for each species, an almost uniform adult size; and among creatures of these kinds (birds especially, in which this restrictive effect of expenditure is most conspicuous), the connexion between cessation of growth and commencement of reproduction is distinct. But we also saw (§ 46) that where, as in the Crocodile and the Pike, the conditions and habits of life are such that expenditure does not overtake assimilation as size increases, there is no precise limit of growth; and in creatures thus circumstanced we may naturally look for a comparatively indeterminate relation between declining growth and