hence follows one of the most just and valuable of all his principles, that a genus should furnish a character, not a character form a genus; or, in other words, that a certain coincidence of structure, habit, and perhaps qualities, among a number of plants, should strike the judgment of a botanist, before he fixes on one or more technical characters, by which to stamp and define such plants as one natural genus. Thus the Hemerocallis cærulea, Andr. Repos. t. 6, and alba, t. 194, though hitherto referred by all botanists to that genus, are so very different from the other species in habit, that a discriminative character might with confidence be expected in some part or other of their fructification, and such a character is accordingly found in the winged seeds. Yet in the natural genera of Arenaria and Spergula, winged or bordered seeds are so far from indicating a distinct genus, that it is doubtful whether they are sufficient to constitute even a specific character. See Engl. Bot. t. 958, 1535 and 1536. So Blandfordia, Exot. Bot. t. 4, is well distinguished from Aletris, with which some botanists have confounded it, by its
