have keeled breast-bones, which imply their descent from forms adapted to true flight. They are linked to the ostriches, however, and therefore to the still earlier toothed ancestral types, by the South American tinamous, which are intermediate in various anatomical points (too intricate for a lazy man to go into here and now), between the two classes. Put briefly, one may say that these partridge-like Paraguayan birds are ostriches in the bones of their head, but game-birds in those of the breast and body. This line of descent seems to lead us up directly toward the cocks and hens, the pheasants, and the other scrapers. There are more marks of a primitive organization, however, among the penguins, which are almost wingless swimming birds, belonging nearly to the same class as the ducks and geese; and we have reason otherwise to consider the penguins a very early form, since fowls resembling them in many particulars have been unearthed in the upper greensand. Here the wings are reduced to small rudiments, covered with bristly, scale-like feathers, and so rigid that they can be only moved in the mass like fins by a single joint at the base. They are used, in fact, exactly in the same way as the flappers in seals, to assist the bird in diving. The habitual erect attitude of the penguins strongly recalls that of their reptilian ally, compsognathus. From such an incomplete form as this, the gap is not great to the equally erect auks, the guillemots, the grebes, and other web-footed divers, which have short, pointed wings with true quills, but without any extended power of flight. Some species, indeed, can not fly at all, though the puffins and many other kinds can steer their way through the air with comparative ease. Thence to the cormorants, gulls, and ducks the transitions are slight and easy. We are thus led insensibly from almost wingless erect birds, like the penguins, through winged, but mainly swimming forms like the auks and divers, to creatures with such marvelous powers of flight as the frigate-birds, the petrels, and the albatrosses, which pass almost their whole life upon the wing. It must be remembered, however, that in this line of descent the comparatively wingless forms must be regarded as somewhat degenerate representatives of flying ancestors; for the presence of a keeled breast-bone almost conclusively proves hereditary connection with fully-winged progenitors.
By far the greater number of modern birds belong to the still more strictly aërial orders of the perchers, the peckers, and the birds of prey. In almost all these cases, the power of flight is highly developed, and the bird type reaches its highest ideal point of typical excellence. Among the perchers, this perfection of form is best seen in the swallows, whose ceaseless and graceful curved evolutions everybody has seen with his own eyes; while among tropical varieties of the same type the birds-of-paradise, the sun-birds, and the orioles are the most conspicuous. Among the peckers, our own swifts closely simulate the swallow type, while their American relatives, the hum-
