The point, I repeat, to be especially noted about these cases and others which might be cited,[EM] is that the variation produced is a definite variation. Very probably it is generally, or perhaps always, produced in the embryonic or larval period of life. In some cases the variation seems to be transmissible, though definite and satisfactory proofs of this are certainly wanting. Still, we may say that if the changed conditions be maintained, the resulting variation will also be maintained. Under these conditions, at least, the variation is a stable one. It is probable that, apart from preferential mating, the varieties thus produced will tend to breed together rather than to be crossed with the parent form or varieties living under different conditions. In this way varieties may sometimes arise by definite and perhaps considerable leaps under the influence of changed conditions. We must not run the adage, _Natura nil facit per saltum_, too hard, nor interpret _saltum_ in too narrow a sense. It is true, and we may repeat the statement of the fact for the sake of emphasis, that we do not know how or why this or that particular variation should result from this or that change of climate, environment, or food-stuff; nor do we know why certain variations (such as that which produced the ancon breed of sheep) should be stable, while other variations are peculiarly unstable. But in this we are not worse off than we are in the study of inorganic nature. We do not know why calcite should crystallize in any particular one of its numerous varieties of crystalline form; we do not know why some of these are more stable than others. We may be able to point to some of the conditions, but we cannot be said to understand why arragonite should be produced under some circumstances, calcite under others; or why the same constituents should assume the form of augite in some rocks, and hornblende in other rocks. We are hedged in by ignorance; and perhaps one of our chief dangers, becoming with some people a besetting sin, is that of pretending to know more than we are at present in a position to know. Our very analogies by which we endeavour to make clear our meaning may often seem to imply an unwarrantable assumption of knowledge.