On the Generation of Animals 729b34
"But what occurs in birds and in the oviparous kind of fishes is the best evidence that the seed does not come from all the parts, and that the male does not emit any part such as will remain present within the offspring, but generates an animal merely by the capability in the semen, just as we said of insects in which the female puts a part into the male. For if a hen is gestating wind-eggs and is then mated before the egg has changed from being entirely yellow to being turning white, the eggs become fertile instead of wind-eggs; and if it is mated with a second cock while the egg is still yellow, the chicks turn out to be of the same kind in every respect as the second cock. This is why some who are concerned with the highly bred birds act in this way, changing the first and subsequent matings. It implies that the seed is not mixed in and present within, and that it did not come from every part; for it would have come from both cocks, so that the same parts would have been contained twice. But by its capability the male seed puts the matter and nutriment that is in the female into a particular kind of state. The seed that came in later can do this by heating and concocting, since the egg takes nutriment so long as it is increasing in size. The same occurs in the generation of the oviparous fishes too. When the female has laid the eggs, the male sprinkles the semen over them; those that it reaches become fertile, but those that it does not are infertile, implying that the male's contribution to the animals is not quantitative but qualitative."