On some historical and theoretical foundations of the concept of chordates.

The concept of chordates arose from the alliance between embryology and evolution in the second half of the nineteenth century, as a result of a theoretical elaboration on Kowalevsky's discoveries about some fundamental similarities between the ontogeny of the lancelet, a putative primitive fish, and that of ascidians, then classified as molluscs. Carrying out his embryological studies in the light of Darwin's theory and von Baer's account of the germ layers, Kowalevsky was influenced by the German tradition of idealistic morphology that was concerned with transformations driven by laws of form, rather than with a gradual evolution occurring by means of variation, selection and adaptation. In agreement with this tradition, Kowalevsky interpreted the vertebrate-like structures of the ascidian larva according to von Kölliker's model of heterogeneous generation. Then, he asserted the homology of the germ layers and their derivatives in different types of animals and suggested a common descent of annelids and vertebrates, in agreement with Saint-Hilaire's hypothesis of the unity of composition of body plans, but in contrast with Haeckel's idea of the Chordonia (chordates). In The Descent of Man Darwin quoted Kowalevsky's discoveries, but accepted Haeckel's interpretation of the ascidian embryology within the frame of a monophyletic tree of life that was produced by the fundamental biogenetic law. Joining embryology to evolution in the light of idealistic morphology, the biogenetic law turned out to be instrumental in bringing forth different evolutionary hypotheses: it was used by Haeckel and Darwin to link vertebrates to invertebrates by means of the concept of chordates, and by Kowalevsky to corroborate the annelid theory of the origin of vertebrates. Yet, there was still another interpretation of Kowalevsky's discoveries. As an adherent to empiricism and to Cuvier's theory of types, von Baer asserted that these discoveries did not prove convincingly a dorsal position of the nervous system in the ascidian tadpole larva; hence, they could not support a homology between different animal types suggesting a kinship between ascidians and vertebrates.