In the early days, before the other claimants to the dignity of ancestral form to the Vertebrates--_Balanoglossus_, Nemertines and the rest--had put in an appearance, there were two main views on the subject, one upheld by Haeckel, Kowalevsky and others, to the effect that the proximate ancestor of Vertebrates was a form somewhat resembling the ascidian tadpole, the other supported principally by Dohrn and Semper that Vertebrates and Arthropods traced their descent to a common segmented annelid or pro-annelid ancestor. The former view is historically prior, and arose directly out of the brilliant embryological investigations of A. Kowalevsky, who proved himself to be a worthy successor of the great comparative embryologist Rathke. His work was indeed a true continuation of Rathke's. It was not directly inspired by evolution, though it supplied much useful confirmation of the theory--you may read Kowalevsky's earlier memoirs and not realise that they were written several years after the publication of the _Origin of Species_. His first paper of evolutionary importance was a note in Russian on the development of Amphioxus, published in 1865. This subject was followed up in two papers which appeared in 1867[387] and 1877.[388] In his papers on Amphioxus Kowalevsky made out the main features in the development of this primitive form, and showed that the chief organs were formed in essentially the same way as in Vertebrates; he described the formation of the archenteron by invagination, the appearance of the medullary folds, which coalesced to form the neural canal, the formation of the notochord and of the gill-slits. At first he made the mistake of supposing that the body-cavity arose from the segmentation-cavity, but in his later paper he rightly surmised that it was formed from the cavities of the "primitive vertebrae," or mesodermal segments. The origin of the notochord from the endoderm was also not made out by Kowalevsky in his paper of 1867. Although many important details remained to be discovered by later investigators,[389] Kowalevsky's work at once made the development of Amphioxus the key to vertebrate embryology, the typical ontogeny with which all others could be compared.