Remembering the fact that among the higher classes of organisms, fertilization is always effected by combining the sperm-cell of one individual with the germ-cell of another; and joining with it the above fact that among hermaphrodite organisms, the germ-cells developed in any individual are usually not fertilized by sperm-cells developed in the same individual; we see reason for thinking that the essential thing in fertilization, is the union of specially-fitted portions of _different_ organisms. If fertilization depended on the peculiar properties of sperm-cell and germ-cell, as such; then, in hermaphrodite organisms, it would be a matter of indifference whether the united sperm-cells and germ-cells were those of the same individual or those of different individuals. But the circumstance that there exist in such organisms elaborate appliances for mutual fertilization, shows that unlikeness of derivation in the united reproductive centres, is the desideratum. Now this is just what the foregoing hypothesis implies. If, as was concluded, fertilization has for its object the disturbance of that approaching equilibrium existing among the physiological units separated from an adult organism; and if, as we saw reason to think, this object is effected by mixture with the slightly-different physiological units of another organism; then, we at the same time see that this object will not be effected by mixture with physiological units belonging to the same organism. Thus, the hypothesis leads us to expect such provisions as we find.
§ 94. But here a difficulty presents itself. These propositions seem to involve the conclusion that self-fertilization is impossible. It apparently follows from them, that a group of physiological units from one part of an organism ought to have no power of altering the state of approaching balance in a group from another part of it. Yet self-fertilization does occur. Though the ovules of one plant are generally fertilized by pollen from another plant of the same kind, yet they may be, some of them, fertilized by pollen of the same plant; and, indeed, there are plants in which self-fertilization is the rule: even provision being in some cases made to prevent fertilization by pollen from other individuals. And though, among hermaphrodite animals, self-fertilization is usually negatived by structural or functional arrangements, yet in certain _Entozoa_ there appear to be special provisions by which the sperm-cells and the germ-cells of the same individual may be united, when not previously united with those of another individual. Nay, it has even been shown that in certain Ascidians the contents of oviduct and spermiduct of the same individual produce, when united, fertile ova whence evolve perfect individuals. Certainly, at first sight, these facts do not consist with the above supposition. Nevertheless there is something like a solution.