Certain structural features of flowers have been so developed that fertilization of the ovary by the pollen of the same flower is impossible. The commonest case is in those flowers where the stamens are shorter than the pistils, as they always are in the common snowdrop, hyacinth, the sassafras tree, and in hundreds of others. There can be no consummation of the reproductive process in such flowers without some outside aid. More futile still without this aid are those flowers where the stamens are well above the pistils, but the time of maturing in both differs by a few days or even hours. Nothing could be more helpless than the pistil under these circumstances, for if its instinct for maternity were ever so strong, it would be doomed to barren sterility by the premature development of the males. Sometimes, too, the female is prematurely ripe for impregnation, and the stamens lag behind a day or two. Her time passes and with it her only chance of fertilization--by her own haste she has rendered impotent the now useless pollen which appears doomed to fall aimlessly upon the unreceptive stigma. [Illustration: FIG. 69--THE VIOLET Note the showy, often partly infertile upper flowers and the much smaller cleistogamous ones at the base, which never open and yet produce a good crop of seeds.] But, perhaps, the most hopeless of all is the well-known partridge berry, whose red berries are common in the woods during August and September. This seems as though it fought off any chance of securing a mate by a flower structure and behavior that would certainly so result if some way out of the difficulty were not at hand. The partridge berry bears two kinds of flowers that outwardly look much alike, but whose sexual organs differ in this way: in some flowers the stamens are all shorter than the pistil, and in others the pistil is much outtopped by the stamens. The extraordinary feature of it is not so much this structural difference, however, but the fact that pollen from the short-stamened flower is useful only to its neighboring short-styled relative, while the pollen from this long-stamened but short-styled neighbor is nearly useless where it is found and really useful only on the long-styled plant. By this device, but again only with outside aid, this plant does not prevent maternity, but increases its chances of being fruitful, for, as we have already seen, cross-fertilization appears to be the rule rather than the exception, and the partridge berry not only needs it, but can exist only when its offspring are the result of such crosses.