but quite unlike the lily or tulip, because the stamens are very numerous in each blossom. In this plant there is no need of speculating as to the origin of the petals. They arise in large numbers from the failure of the stamens to develop as such. The filaments broaden out upon opposite sides and a petal results. In nearly half such petals the remnant of an anther can be seen at the tip of the petal, which is somewhat notched, often deeply, and in the center is the abortive anther. Near the center of the flower the transition is more evident, for here the filament-wings are not much broadened, and the anthers more prominent. Still nearer the great center pistil the ordinary stamens may be found, with their anthers bearing pollen. Occasionally the poppy illustrates a modification of the stamen in the opposite direction to that given above—namely, the inner ones become small simple pistils, which are either closely applied to the surface of the large central compound pistil, or adherent to it and blending with the stigmas.
The rose family and the crowfoots both furnish a long list of plants which uniformly produce double flowers under cultivation, and for this reason these two orders are rich in ornamental garden species. Both the roses and the buttercups abound in stamens; and, from what we have seen in the poppy, it should be expected that doubling would be easy in such plants. The examples of doubled flowers in these two families are so familiar that no further comment need be made. Among the hardy cultivated roses, for example, it is rarely that a blossom can be found not exhibiting all gradations between perfect stamens and unmistakable petals. It may, however, be stated that in a member of the rose family grown for its fruit—namely, the apple—petal-stamens were frequently met with. In the Tallman sweeting variety, upon one tree, the doubling was found as frequently as one flower in ten. Usually one stamen was transformed, but rarely so much so as to be distinctly petaline.
The abnormities which we have been considering, both generically and specifically, are rarely met with in wild plants in a state of nature. They are, therefore, transformations in flowers concomitant with culture. It is a well-established fact that culture induces changes in those parts for which the plant is cultivated and it might be added that they are cultivated because of this response. Varieties of any cereal differ mostly in the grain; beets, carrots, and turnips in the roots; apples, plums, and peaches in the fruit; and so on. In accordance with the general rule, plants grown for their flowers should vary most in the blossoms. A plant when under cultivation has been removed from the conditions which obtain in the wild state and is relieved from that fierce struggle for life which is everywhere in progress among