Abstract
The first description of vertebral development in a plethodontid salamander is presented. Eurycea bislineata has larvae that hatch at a rather early stage of development. Somites and the notochord appear early. Somitic differentiation is slight, and no distinct sclerotome can be found. As a result, there is no clear primary segmentation of the skeletogenous tissue. No evidence of a sclerocoele can be found. The amount of sclerotomal cells surrounding the notochord is very low, relative to other tetrapods. Yet discrete perichordal rings of cells do form, in nearly midsegmental positions, and these give rise to the intervertebral cartilages. Osteogenesis of the centra is initiated prior to hatching and is coincidental with ossification of the neural arch. There is no sign of a neurocentral suture. The centrum forms as a thin shell of bone directly from sclerotomal cells. The notochord is a prominent feature of the vertebral column throughout life, retaining its integrity until late in life when some disintegration occurs locally. The notochord is filled with cartilage midvertebrally in late larval stages, and some additional cartilage forms later in life. The intervertebral cartilage enlarges greatly in late larval life. An opisthocoelous joint forms in this cartilage, apparently as a result of differential changes in the cells of the perichordal ring rather than by an invasion of cells from an external source. The intervertebral cartilage is a dominant structural and highly important functional feature of the adult vertebra. In metamorphosed individuals it may become extensively mineralized, and it consists of many different structural kinds of cartilage.The cranio‐vertebral joint seems to form in a single segment, contrary to the condition reported by some early investigators. It is complex, and consists of articulations between the odontoid process of the atlas and the occipital arch, as well as between the occipital condyles and atlantal cotyles. The notochord plays a dominant role in the early development of the odontoid, but then changes radically and is absent in the adult process.The anterior trunk region seems to be much more conservative than posterior parts of the column. The patterns of nerve routes and nature of development of the ribs and rib bearers differs greatly from conditions elsewhere in the column. The rib patterns are similar to presumed ancestral conditions. Rib development on the sacral and caudosacral vertebrae is in some ways more similar to that of the anterior vertebrae than of the central trunk vertebrae.Quantitative aspects of variation in the vertebrae of adult salamanders are presented. There is more regional variation and less site variation than would be expected from literature reports.Evolutionary aspects of the origin of the cranio‐vertebral joint, transverse process and ribs, patterns of segmentation, and centrum development are considered in the light of the new information on Eurycea. There is no evidence that more than one vertebra is involved developmentally or evolutionarily in the cranio‐vertebral joint. The most generalized condition of rib bearers in living salamanders is one in which the dorsal and ventral bearers are in cartilaginous continuity during development. There are many variations on this theme in living species. It is inappropriate to speak of a resegmentation of the sclerotome in Eurycca, even though the adult vertebra is a transsegmental structure, because there is no primary segmentation of the scanty sclerotome. The important feature found in vertebral development in all tetrapods is the perichordal tube and its subsequent differentiation. Questinos concerning precise homologies of the salamander vertebral centrum with those of other vertebrates cannot be answered by data from development sequences with currently used criteria of homology. On the other hand, it appears that all centra, regardless of subdivision, are homologous in all tetrapods.

This publication has 29 references indexed in Scilit: