OF THE LAEVAL POLTPTERUS. 327 



The centra themselves are at this stage very thin and hourglass-shaped, expandino- 

 widely anteriorly and posteriorly ; they are not formed previously in cartilage 

 (PL XXXIV. fig. 8, G.) except for the small neural and lateral cartilages which they 

 enclose ; they directly surround the chordal sheath. There is, however, a single layer 

 of cells separating them from the latter in the position where the membrana elastica 

 externa should be, which somewhat resemble cartilage-cells ; they do not, however, 

 stain in the very characteristic way in which all other cartilage in this specimen has 

 stained (PI. XXXIV. fig. 6, m.c). The cartilages of the neural arches play but a very 

 small part in the formation of the bony neural arches, which are also formed directly 

 in connective tissue. 



In specimens 13 cm. long the vacuolated portion of the notochord has become 

 constricted vertebrally, while the sheath of the notochord, retaining a considerable 

 thickness in the vertebral region, intervertebrally becomes greatly thickened 

 (PI. XXXIV. fig. 6, Ch.sh.). It is this thickening of the notochordal sheath which 

 is the chief cause of the great expansion of the bony vertebrae in the intervertebral 

 regions. At this stage all trace of cartilage has disappeared from the composition 

 of the vertebrae (PI. XXXIV. fig. 5). 



The formation of the vertebral centra in Polypterus is very much on the same lines 

 as that of Lepidosteus and the Teleostei. It differs from Lepidosteus in the absence of 

 the intervertebral rings of cartilage which give rise to the procoelous articulations in 

 that form, and from Amia in the absence of intercalated cartilages. In the possession 

 of three pairs of vertebrally placed cartilages resting upon the notochordal sheath 

 before the commencement of bone formation, Polypterus difi'ers from all living 

 Vertebrates. 



That Polypterus is provided with a double set of ribs has long been known ; and by 

 some authors the upper series has been homologized with the ribs of Elasmobranchii, 

 Amphibia, and Amniota, and the lower series with the ribs of the Dipnoi, Ganoidei, and 

 Teleostei. Goppert especially ^ has developed this view regarding the reduction of 

 the lower series in the former groups as due to the suppression of the ventro-lateral 

 muscles. The starting-point which Goppert took was the condition in Calamoichthys 

 at the stage where the lower ribs are attached to the under sides of the upper ribs 

 shown in PL XXXIV. fig. 4 of Polypterus. The discovery that all the lower ribs in 

 Polypterus have their origin in a ventral series of cartilages resting directly on the 

 notochordal sheath seems to me to very much strengthen his views. The absence of 

 intercalated cartilages, such as are found in the majority of Vertebrates exclusive of the 

 Teleostei or of the intervertebral cartilage-rings of Lepidosteus, may possibly be 

 connected with the early and great development of membrane-bone which is 

 so characteristic of the development of all the skeletal structures of Polypterus. 



' Morph. Jahrb. Bd. xxiii. 



