That this is the true explanation is demonstrated by the fact that in amphibians this very change actually takes place before our eyes in the _individual history_. We have already seen that the individual frog, in its tadpole state, is a gill-breather. It has therefore its gill-arches (Fig. 38), three on each side, like a fish, and for the same reason, viz., the aëration of the blood. But when its gills dry up and lung-respiration is established, its now useless gill-arches still remain as aortic arches, to attest their previous condition (Fig. 39). Now, the lizard undoubtedly came from an air-breathing, tailed amphibian, and therefore inherited this form of arterial distribution. In both lizard and amphibian the ultimate cause is an origin from fishes, in which such arches are obviously necessary. The diagrams, Figs. 38 and 39, are illustrations somewhat idealized, showing the manner in which the change actually takes place in air-breathing amphibians. Fig. 38 represents the tadpole stage, and Fig. 39 the mature condition. In the former the gills are mostly external, G G′, etc., but also internal, _g g′_, as in the fish. Observe in this condition the small connecting vessels, _c c′_. When the external gills dry up, these are enlarged, and the whole of the blood passes through them, as shown in Fig. 39. It is seen, also, in Fig. 38, that a small branch, _p_, goes from the lower gill-arches to the yet rudimentary lung, _l_. When the gill-fringes have disappeared, the whole of the blood of the lower arch goes through the now enlarged pulmonary branch to the lungs, L, now in full activity, and the remainder of this arch disappears, as shown by the dotted lines in Fig. 39. The change which actually took place in the family history of the lizard probably differed from the above only in being more simple, the gills being only internal like the fish. The external gills complicate the process a little in the case of the frog, but the principle is precisely the same. As already explained (pages 82-85), the large gap between fishes and reptiles, as regards mode of respiration, is completely filled both in the taxonomic series--i. e., in ganoids, dipnoi, and the mature condition of the different orders of amphibians--and in the ontogeny of the higher amphibians. Now, we add that the same is true of the arterial distribution. We have just traced the change in the ontogeny of the frog, but the steps of the same change are traceable in passing from the typical fish (teleosts), through dipnoi and amphibians to reptiles. Thus, again, the phylogeny, the taxonomy, and the ontogeny, are in complete accord.