In his further study of the development of _Blennius viviparus_, Rathke observed some important facts about the development of the vertebral column and skull. He found that the vertebral centra were first formed as rings in the chorda-sheath, which give off neural and haemal processes. The vertebra later ossifies from four centres. The chorda (notochord) is prolonged some little way into the head, and the base of the cranium is formed by the expanded sheath, which reaches forward in front of the end of the notochord. This cranial basis shows a division into three segments, in which Rathke was inclined to see an indication of three cranial vertebrae. (It turned out that this division into three segments did not really exist, and Rathke later acknowledged that he had made an error of observation.) The side walls of the skull grow out from this base and form a fibrous capsule for the brain. The cranial section of the chorda itself shows no sign of segmentation; but later on the cranial portion of the chorda-sheath ossifies, like the vertebrae, from several centres. The vomer, which, in the classical form of the vertebral theory of the skull, was the centrum of the fourth, or foremost, cranial vertebra, does not, according to Rathke, develop in continuity with the cranial basis and the chorda sheath, but develops separately in the facial region. Von Baer, like Rathke at this time, was also to some extent a believer in the vertebral theory of the skull. In his second volume (1834, pub. 1837) he holds that the development of the skull, as the sum of the anterior vertebral arches, is in general the same as that of the other neural arches, and is modified only by the great bulk of the brain (_Entwickelungsgeschichte_, ii., p. 99). He had, however, some doubts as to the entire correctness of the vertebral theory, doubts suggested by a study of the developing skull. "In the course of the formation of the head in the higher animals, something additional is introduced which does not originally belong to the cranial vertebrae. At first we see the vertebration in the hinder region of the skull very clearly. Afterwards it becomes suddenly indistinct, as if some new formation overlaid it" (i., p. 194). Even more clearly is his doubt expressed in his paper on _Cyprinus_. "Upon the formation of the vertebral column only this need be said, that at this stage the notochord is very clearly seen, and the upper and lower arches and spinous processes are visible right to the end of the tail, but the separation into vertebrae ceases abruptly where the back passes into the head. I do not hesitate to assert _that bony fish, too, have at this stage an unsegmented cartilaginous cranium_ (as cartilaginous fish have all their life), the prominences and hollows of which constitute its only resemblance with the vertebral type" (1835, p. 19).