We have so far assumed that intercrossing of parents and interblending of their characters in the offspring always go together. This, we must now notice, is not always the fact. If a blue-eyed Saxon marry a dark-eyed Italian, the children will have blue eyes _or_ dark eyes, not eyes of an intermediate tint. The characters do not interblend. The _ancon_, or otter-sheep, a breed with a long body and short, bandy legs, appeared in Massachusetts as a chance sport in a single lamb. The offspring of this ram were either ancons or ordinary sheep. The ancon characters did not blend. Hence for a time a definite breed was maintained. We may call this mode of isolation _isolation by exclusive inheritance_. A further mode of isolation or segregation, for which Mr. Romanes[AR] claims a foremost, indeed, the foremost, place, is _physiological isolation_ as due to differential fertility. One among the many variations to which organisms are subject is a variation in fertility, which may reach the climax of absolute sterility. But it is clear that a sterile variation carries with it its own death-warrant, since the sterile individual leaves no descendants to inherit its peculiarity. Relative infertility, too, unless it chances to be correlated with some unusual excellence, would be no advantage, would be transmitted to few descendants, and would tend to be extinguished. The same is not true, however, of differential fertility. "It is by no means rare," said Darwin,[AS] "to find certain males and females which will not breed together, though both are known to be perfectly fertile with other males and females." Mr. Romanes assumes, as a starting-point, the converse of this, namely, that certain males and females will breed together, though they are infertile with all other members of the species.