easy to find. But this anatomical chain corresponds exactly with the paleontological facts; selachians and ganoids are already found in the Silurian formations, dipneusta in the Devonian, amphibia in the Carboniferous, reptiles in the Permian, mammalia in the Trias.
These are historical facts of the first rank. They attest in the most gratifying manner the successive steps of the development of vertebrates, as they have been made out by the comparative researches of Cuvier and Meckel, of Johannes Müller and Gegenbaur, of Owen, Huxley, and Flower. The historical succession of the principal steps in the vertebrate stock is thereby definitely established, and this success is much more important for an understanding of the human family tree than if we had succeeded in placing, in a hundred fossil skeletons of lemurs and apes, the entire series of our Tertiary primate ancestors in coherent succession before our eyes.
Much more difficult and dark is the oldest history of our stock, the derivation of a vertebrate stem from an invertebrate ancestry. As none of these possessed any hard and petrifiable parts of the skeleton (resembling in this respect the lowest vertebrates, the cyclostomata and acrania) the evidence of paleontology entirely fails us here; we must rely alone upon the other two records of our family history, upon comparative anatomy and ontogeny. To be sure, their value is here so great in many respects that for every expert and discriminating zoologist they throw the clearest light upon many great features of our older phylogeny. Of the greatest value are these far-reaching inferences which modern comparative ontogeny has drawn during the last thirty years by the aid of the fundamental biogenetic law. Already the older embryology has made clear the elements of vertebrate development by the thorough work of Baer and of Bischoff, of Remak and Kölliker. Then, in 1866, came the important discoveries of Kowalevsky, which confirmed the suspicion of Goodsir and pointed to the close relationship of vertebrates and tunicates; the comparative anatomy of Amphioxus and of the ascidians has since that time been the constant starting point for all further investigations concerning our invertebrate predecessors.
Five years investigation of the structure and development of the chalk-sponges (1867–1872) had led me at that time to a reform of the theory of the germinal layers and to advance the gastræa theory. It first appeared in 1872 in my monograph on the chalk-sponges or Calcispongidæ. These views obtained the most earnest support and the most fruitful development by the excellent comparative researches of many other embryologists, especially those of E. Ray-Lankester and Francis Balfour, as well as those of the brothers Oscar and Richard Hertwig. I had already then concluded from these comparative researches that the first step of development in all Metazoa, or tissue-building animals, is essentially the same, and that we may from this obtain definite insight into the common origin and the older ancestral series of the same. The
