In many of the plants already noted cross-fertilization is accomplished by virtue of the fact that the stamens mature before the stigma. But in the common strawberry the reverse is true. As the insect stands on the white petals, it must, in order to reach the honey at the base of the flower, put its head down so that if pollen were available there could only result self-fertilization. But while the stigma is ready, the pollen at this stage never is, and the insect which comes usually laden with pollen from other flowers cannot avoid impregnating the stigma with this foreign pollen. Later the stamens mature and, as insect visits continue so long as honey is to be found, they become dusted with pollen which is used for the fertilization of other plants. Scores of other plants also produce ripened stigmas before the pollen matures, and they must rely on insects to cross-fertilize them. The cross-fertilization of the strawberry is such a comparatively simple process--seems in fact almost inevitable--that we are lost in wonder at the almost mathematical complexity of the act in the common purple loosestrife, which has been introduced into American gardens from Europe, and sometimes runs wild. In this plant there is a long terminal spike of showy, purplish-pink flowers, the color of which is sufficient to attract many insects from even a fairly swift flight. The petals are streaked with “pathfinders” toward the center of the flower. This consists of a tubular calyx; at the bottom of this is the honey, which secures the insect’s further interest once the color has attracted it. But it finds a condition of the reproductive organs almost without parallel. In some plants the style is hidden down in the calyx tube, while one set of stamens just peep out of the end of the tube, and a second set are still further and quite obviously protruded. In a neighboring plant the style will be found outside the tube, one set of stamens hidden in it, and the other set outtopping everything, except the petals. In still a third type of the loosestrife, the style exceeds everything but the petals, one set of stamens just emerge from the tube and the lowest set are hidden. It would appear as if the loosestrife could scarcely escape self-fertilization, except possibly in that form where the stigma outtops all the stamens, and this would result always if the pollen were indiscriminately useful from all three lengths of stamens. But it never is, only that from the short-stamened plants will fertilize the short-styled one, the mid-stamened ones the mid-styled counterpart, and the long-stamened ones the long-styled flower. The pollen grains of the different-lengthed stamens are even of different size and color. If this were not so, it is a simple mathematical problem that, with three different sets of style lengths and six sets of stamens, two in each flower, eighteen different crosses might be possible. As it is, only six crosses are ever possible and these only by