ova which remain dormant through the winter, is more favourable to the preservation of the race than would be a further continuance of agamogenesis. On the other hand, among the higher animals living on food which, though dispersed, is more or less aggregated into large masses, this alternation of gamic and agamic reproduction ceases to be useful. The development of the germ-product into a single organism of considerable bulk, is in many cases a condition without which these large masses of nutriment could not be appropriated; and here the formation of many individuals instead of one would be fatal. But we still see the beneficial results of the general law--the postponement of gamogenesis until the rate of growth begins to decline. For so long as the rate of growth continues rapid, there is proof that the organism gets food with facility--that expenditure does not seriously check accumulation; and that the size reached is as yet not disadvantageous: or rather, indeed, that it is advantageous. But when the rate of growth is much decreased by the increase of expenditure--when the excess of assimilative power is diminishing so fast as to indicate its approaching disappearance--it becomes needful, for the maintenance of the species, that this excess shall be turned to the production of new individuals; since, did growth continue until there was a complete balancing of assimilation and expenditure, the production of new individuals would be either impossible or fatal to the parent. And it is clear that "natural selection" will continually tend to determine the period at which gamogenesis commences, in such a way as most favours the maintenance of the race. Here, too, may fitly be pointed out the fact that, by "natural selection," there will in every case be produced the most advantageous proportion of males and females. If the conditions of life render numerical inequality of the sexes beneficial to the species, in respect either of the number of the offspring or the character of the offspring; then, those varieties of the species which approach more than other varieties towards this beneficial degree of inequality, will be apt to supplant other varieties. And conversely, where equality in the number of males and females is beneficial, the equilibrium will be maintained by the dying out of such varieties as produce offspring among which the sexes are not balanced. NOTE.--Such alterations of statement in this chapter as have been made necessary by the advance of biological knowledge since 1864 have not, I think, tended to invalidate its main theses, but have tended to verify them. Some explanations to be here added may remove remaining difficulties.