And how beautiful is the minute mechanical adaptation by which this end is accomplished! This species of Andromeda is a shrub of about four feet in height, its blossoms being borne in close panicled clusters at the summit of the branches. The individual flower is hardly more than an eighth of an inch in diameter. From one of three blossoms I made the accompanying series of three sectional drawings (Fig. 6). The first shows the remarkable interior arrangement of the ten stamens surrounding the pistil. The second presents a sectional view of these stamens, showing their peculiar S-shaped filaments and ring of anthers--one of the latter being shown separate at the right, with its two pores and exposed pollen. The freshly opened blossom discloses the entire ring of anthers in perfect equilibrium, each with its two orifices closed by close contact with the style, thus retaining the pollen. It will readily be seen that an insect's tongue, as indicated by the needle, in probing between them in search for nectar, must needs dislocate one or more of the anthers, and thus release their dusty contents, while the position of the stigma below is such as to escape all contact. [Illustration] In most flowers, with the exception of the orchids, the stamens and pollen are plainly visible; but who ever sees the anthers of the blue-flag? Surely none but the analytical botanist and the companion insect to whom it is so artfully adjusted and so demonstrative. This insect is likely to be either a bumblebee or a species of large fly. In apt illustration of Sprengel's theory of the "path-finder" or honey-guide, the insect does not alight at the centre of the flower, but upon one of the three large drooping sepals, whose veins, converging to the narrow trough above, indicate the path to the nectar. Closely overarching this portion is a long and narrow curved roof--one of three divisions to the style, each surmounting its veined sepals. Beneath this our visiting bee disappears, and a glance at my sectional drawing shows what happens. Concealed within, against the ridge-pole, as it were, the anther awaits his coming, and in his passage to and from the nectar below spreads its pollen over his head and back. Having backed out of this segment of the blossom (A, Fig. 7), he proceeds to the next; but the shelf-like stigma awaits him at the door, and scrapes off or rubs off a few grains of the pollen from his back (B). Thus he continues until the third segment is reached, from which he carries away a fresh load of pollen to another flower. It will be seen that only the outer side of this appendage is stigmatic, and that it is thus naturally impossible for the blue-flag to self-fertilize--only one instance of thousands in which the anther and stigma, though placed in the closest proximity, and apparently even in contact--seemingly with the _design_ of self-fertilization--are actually more perfectly separated