We have the reverse of this experiment in a Mexican gilled salamander, the fish-like axolotl (Siredon pisciformis). It was formerly regarded as a permanent gilled amphibian persisting throughout life at the fish-stage. But some of the hundreds of these animals that are kept in the Botanical Garden at Paris got on to the land for some reason or other, lost their gills, and changed into a form closely resembling the salamander (Amblystoma). Other species of the genus became sexually mature for the first time in this condition. This has been regarded as an astounding phenomenon, although every common frog and salamander repeats the metamorphosis in the spring. The whole change from the aquatic and gill-breathing animal to the terrestrial lung-breathing form may be followed step by step in this case. But what we see here in the development of the individual has happened to the whole class in the course of its stem-history. The metamorphosis goes farther in a third order of Amphibia, the Batrachia or Anura, than in the salamander. To this belong the various kinds of toads, ringed snakes, water-frogs, tree-frogs, etc. These lose, not only the gills, but also (sooner or later) the tail, during metamorphosis. The ontogenetic loss of the gills and the tail in the frog and toad can only be explained on the assumption that they are descended from long-tailed Amphibia of the salamander type. This is also clear from the comparative anatomy of the two groups. This remarkable metamorphosis is, however, also interesting because it throws a certain light on the phylogeny of the tail-less apes and man. Their ancestors also had long tails and gills like the gilled Amphibia, as the tail and the gill-arches of the human embryo clearly show. For comparative anatomical and ontogenetic reasons, we must not seek these amphibian ancestors of ours--as one would be inclined to do, perhaps--among the tail-less Batrachia, but among the tailed lower Amphibia. The vertebrate form that comes next to the Amphibia in the series of our ancestors is a lizard-like animal, the earlier existence of which can be confidently deduced from the facts of comparative anatomy and ontogeny. The living Hatteria of New Zealand (Figure 2.264) and the extinct Rhyncocephala of the Permian period (Figure 2.265) are closely related to this important stem-form; we may call them the Protamniotes, or Primitive Amniotes. All the Vertebrates above the Amphibia--or the three classes of reptiles, birds, and mammals--differ so much in their whole organisation from all the lower Vertebrates we have yet considered, and have so great a resemblance to each other, that we put them all together in a single group with the title of Amniotes. In these three classes alone we find the remarkable embryonic membrane, already mentioned, which we called the amnion; a cenogenetic adaptation that we may regard as a result of the sinking of the growing embryo into the yelk-sac.