By descent the banana bush is a developed tropical lily, not at all remotely allied to the common iris, only that its flowers and fruit are clustered together on a hanging spike, instead of growing solitary and separate as in the true irises. The blossoms, which, though pretty, are comparatively inconspicuous for the size of the plant, show the extraordinary persistence of the lily type; for almost all the vast number of species, more or less directly descended from the primitive lily, continue to the very end of the chapter to have six petals, six stamens, and three rows of seeds in their fruits or capsules. But practical man, with his eye always steadily fixed on the one important quality of edibility--the sum and substance to most people of all botanical research--has confined his attention almost entirely to the fruit of the banana. In all essentials (other than the systematically unimportant one just alluded to) the banana fruit in its original state exactly resembles the capsule of the iris--that pretty pod that divides in three when ripe, and shows the delicate orange-coated seeds lying in triple rows within--only, in the banana, the fruit does not open; in the sweet language of technical botany, it is an indehiscent capsule; and the seeds, instead of standing separate and distinct, as in the iris, are embedded in a soft and pulpy substance which forms the edible and practical part of the entire arrangement. This is the proper appearance of the original and natural banana, before it has been taken in hand and cultivated by tropical man. When cut across the middle, it ought to show three rows of seeds, interspersed with pulp, and faintly preserving some dim memory of the dividing wall which once separated them. In practice, however, the banana differs widely from this theoretical ideal, as practice often _will_ differ from theory; for it has been so long cultivated and selected by man--being probably one of the very oldest, if not actually quite the oldest, of domesticated plants--that it has all but lost the original habit of producing seeds. This is a common effect of cultivation on fruits, and it is of course deliberately aimed at by horticulturists, as the seeds are generally a nuisance, regarded from the point of view of the eater, and their absence improves the fruit, as long as one can manage to get along somehow without them. In the pretty little Tangierine oranges (so ingeniously corrupted by fruiterers into mandarins) the seeds have almost been cultivated out; in the best pine-apples, and in the small grapes known in the dried state as currants, they have quite disappeared; while in some varieties of pears they survive only in the form of shrivelled, barren, and useless pips. But the banana, more than any other plant we know of, has managed for many centuries to do without seeds altogether. The cultivated sort, especially in America, is quite seedless, and the plants are propagated entirely by suckers.