The gills are the organs by which the young tadpole breathes, and each little gill-thread is seen, when viewed with a low power of the microscope, to contain tiny loops of blood-vessels. Every time the tadpole uses his jaws, wags his tail, or, in short, does anything at [Pg 345] all, he uses up some oxygen and produces some carbon dioxide. The water of the aquarium contains dissolved oxygen—green water-weeds are kept in aquaria for the express purpose of liberating oxygen (p. 51)—and this oxygen makes its way through the excessively thin membrane which divides the blood of the gills from the water, to be swept away by the current of the blood to the various parts of the body. The carbon [Pg 346] dioxide produced is brought to the gills by the blood stream, and passes through the membrane into the water, where it is utilised by the water-weed as food (p. 51). Animal and plant are thus mutually serviceable. A series of slits soon opens through the sides of the neck, and along their margins are formed folds which are usually called the internal gills. The external gills dwindle and shrivel up as the internal gills are being formed, and at the same time a flap of skin grows backwards from each side of the head and covers over the slits so that they cannot be seen. Presently the two flaps fuse at their edges—except at one point on the left side, where a spout is left—and so enclose a chamber. The water which enters the tadpole’s mouth pours through the gill-slits, into the chamber, and out through the spout. As it swills over the folds of the internal gills there is an exchange of carbon dioxide for oxygen in the manner just described. The tadpole in the meantime is growing strong and active, and the tail has grown out to form a powerful organ, the sinuous motion of which propels the animal with relatively great speed through the water. From the point of view of the biologist, perhaps the most interesting feature of this stage of tadpolehood is the almost entire correspondence of the structure with that of a fish, although the adult frog is not in any sense a fish. This curious state of things is explained by supposing that frogs have descended from fish-like ancestors, and that every frog, in the course of its development, is under the necessity of repeating, in a more or less modified manner, the chief stages of its ancestral history. As Marshall happily expressed it,[27] a frog during its development climbs up its own genealogical tree.