1 8 THE SALAMANDER 



To sum up then, there would seem to be some truth in each of the 

 two earHer and apparently conflicting views, since the first vertebra 

 of the adult has a double origin while a part of the intercalated arch 

 also fuses with the skull. 



3. The Trunk Vertebrae. 



The trunk vertebrae are all essentially similar. The actual one 

 figured (Figs. 13—17) is the last pre-sacral and may be taken as 

 typical of the whole group. The centrum (cen.) — or corpus — is well 

 formed and rather skittle-shaped. It is oval in cross-section and very 

 strongly opisthocoelous, the posterior concavity tending to become 

 rather deeper towards the ventral side. The ends of the centrum 

 remain cartilaginous, but the body is ossified and hollow, the cavity 

 being filled with a fatty marrow. The neural canal (can.nl.) — or 

 foramen vertebrale — is somewhat pentagonal, the zygapophyses — or 

 processi articulares — are well formed and wing-like, especially those 

 at the posterior end, which are very broad and form a roof over the 

 spinal ganglia. The obliquity of the sliding surfaces is very slight. 

 The neural spine — ox processus spinosus — forms a well-defined ridge 

 anteriorly but fades away posteriorly. The mesial portion of the 

 neural arch between the posterior zygapophyses is thickened, and 

 bears on its posterior surface a pair of small but well-marked depres- 

 sions, one on each side of the middle line. From these concavities 

 the MM. inter spinales arise (see p. 93). The anterior edge of the 

 neural arch is V-shaped, so that a triangular space exists on the dorsal 

 side between one vertebra and the next which is closed by membrane. 

 The transverse processes (pr.tr.) which are stout and very deep dorso- 

 ventrally protrude in a postero-lateral direction and bear ribs (rb.). 



4. The Ribs. 



The question of the articulation of the Urodele rib with the trans- 

 verse process, and of the homology of the ribs with those of Amniota, 

 has long been a somewhat vexed problem, but it has recently been 

 cleared up by Gray (1931), who worked on the development of the 

 ribs in Triton. An examination of the adult condition only would 

 lead one to suppose that both heads of the rib articulate with the bifid 

 transverse process, and indeed this was the view generally held until 

 the appearance of Gray's paper. It was supposed that the attachment 

 of the rib to the chorda, i.e. the capitulum of the rib, was represented 

 by a transitory vestige only, but Gray showed that the reverse is 

 actually the case, and that the chondrification of the rib rudiment is 

 a centripetal process, the capitular process growing inwards towards 



