tup: elasmobranch ftsiies 



71 



The anterior section of the cohunn differs in the various Elasniohranchs in 

 its relation to the eraniinn. In the notidanids, as we have seen, the two are 

 more or less continuous. This coiitiunity is further emphasized in all general- 

 ized forms by the fact that this fused region is perforated by a number of 

 oceipitospinal nerves which arise l)etween the vagus and the first spinal nerve. 

 In a type like Acanthias in which there are a dorsal and two lateral ridges 

 contiiuiing directly from the column to the cranium, the relation of the two 



A. BJiinohatis producius. (Chester Stock, orig.) 



iW^ 



ho. 



B. Keierodontus francisci. (Duncan Dunning, del.) 



Fig. 77. Transitional vertebrae. 

 hd., dorsal basal (basidorsal) plate; hv., ventral basal plate; f.d., foramen for dorsal root 

 nerve; f.v., foramen for ventral i-oot nerve; iv., ventral intercalary; n.s. (and) sid., neural 

 spine ; r., ril). 



regions is not so clear; while in still more specialized forms tlie column is more 

 or less clearly separated from the cranium, attachment of the two being made 

 by processes of the cranium and the column (rays) . 



This segment of the column (fig. 76a, Heterodontus), unlike that of Hep- 

 ianchus, is nsnally clearly divided into centra (c. ) . In practically all forms the 

 relation of the dorsal and ventral arches to the centra is equally clear. In a 

 type like Pseudotriacis, how^ever, the segmentation is irregular, both in the 

 centra and in the neural arches of the column. This irregularity is especially 

 marked in the most anterior vertebra which has fused into a solid ring. In the 

 rays (Rhinohatis, fig. 76b) the larger part of this anterior region has second- 

 arily fused into a solid vertebral plate, the segmentation of which is made out 

 only through a study of the foramina and certain lateral processes. 



