MEDIAN FINS. 



21 



relation between the endoskeleton of the dorsal fin and the 

 neural arches is shown in the primitive genus Coccosteus (p. 67, 

 fig. 52); and it is likely that this relation would also be observed 

 in Cladoselache (p. 30) if the axial skeleton of that fish were 

 known. On the other hand in Pleuracanthus, the fin-supports 

 bear no numerical relation to the filiform dermal rays (actino- 

 trichia) stiffening the fin-membrane ; except that occasionally 

 the latter are grouped in bundles round the extremities of the 

 supports. 



FIG. 18. 



Pleuracanthus (Xenacanthw) decheni ; skeleton of anal fins, about nat. size. 

 L. Permian; Bohemia, a, b, c, hsemal arches of vertebral axis; the suc- 

 cessive segments numbered. Eight border preaxial, left border postaxial. 

 (After A. Fritsch.) 



The severance of two small anal fins from the dorso-caudal 

 membrane in Pleuracanthus adds further interesting informa- 

 tion in reference to the process by which this membrane becomes 

 subdivided. The endoskeleton of these fins is carefully drawn 

 in fig. 18. The supports of the anterior anal are connected only 

 with two haemal arches (b, c), and the preaxial one (!') is com- 

 paratively simple, differing from an ordinary support merely 

 in the duplication and transverse articulation of the distal 

 element. The postaxial support (1") is remarkably complex and 

 must evidently be regarded as a cluster of primitive supports, 



