Now let us glance at Haeckel’s genealogical tables. We find eight of them, artistically drawn, at the end of the second volume. The “genealogical tree” is given in the form of a branching tree, or as a huge forest-like growth of stems some of which only meet in the ultimate roots. There is no trace in Haeckel’s designs of the sort of Eiffel-Tower arrangement that the Linnean system involved. At the bottom we find the protists, the most primitive forms of life. From this point two parallel stems diverge, that of the animals and that of the plants; they never touch each other after this point, and so cannot be expected to be reproduced in the embryonic forms. Then the animal stem is split up almost at the root into at least five independent branches, each of which pursues its separate line of development. One culminates in the insects, above the worms and the crustacea. A totally independent stem issues in the vertebrates, and this in turn breaks into many different branches. Beyond the lizards, for instance, we find the development of the mammals and birds, which run on as separate and parallel lines. It was mere nonsense to expect a mammal in its embryonic development to assume the form of a bird, or a crab, or a beetle, or a mussel, or a medusa, even if the biogenetic law were established ten times over. The second mistake made by Oken was to declare that, whatever it cost, the law must be observed everywhere. He examined the butterfly. It passed through two curious embryonic stages: first the caterpillar, then the pupa. The caterpillar corresponded to the worm; that might be plausibly contended. But the pupa also must stand for something. Between the worm and the insect in classification was the crustacean. It had a hard shell: so had the pupa. Consequently, the pupa is a reproduction of the crustacea-stage. Such were the bold chess-moves of the older theorist. Haeckel first established that there was such a thing as the biogenetic law. There is a fundamental norm, which is made clear to us in embryology and can at the same time (remember the instance of the lizard-like teeth in the bird-embryo) give us most wonderful suggestions as to the line of ancestral development. But it has certain limitations, as we will now show.