MEDIAN FINS 



105 



type (p. 480). The name gephyrocercal has been applied to these 

 pseudo-diphycercal tails. 



Turning now to the relation between the axial and the appen- 

 dicular skeleton of the median fins, which has already been alluded 

 to above(pp. 69 and 101), wefind that in the Pleuracanthids amongst 

 the Chondrichthyes, in the Dipnoi, and in many Teleostomes (Coela- 

 canths, many Teleosts) the radials of the fins correspond to, and 

 may articulate with, the neural and haemal spines. As a rule, the 

 more complete is the fin, the more thorough is the continuity 



etff 



c. 



The development of the caudal liii of Lrp'ulosteus. (After A. Agassiz.) A, young larva with 

 .interior sucker, s; yolk-sac, y.s; continuous dorsal and ventral tin folds, d.f and v.ff; and 

 straight notochord. The latter stages, B, C, and L>, show the upbending of the notochord, 

 the dwindling of the axial lobe, o, which disappears in the adult (cp. Fig. 65), and the 

 great development of the hypochordal tin, h.f. n.f, anal, d.f, dorsal, ec.f, epichordal, and pt.f, 

 pectoral tin. 



between the two sets of skeletal rods. Now it is important to 

 notice that in the Elasmobranchs, where the median fin breaks up 

 into isolated dorsal and anal fins, the neural spines above the longi- 

 tudinal ligament and the haemal spines tend to disappear com- 

 pletely (except perhaps in the epichordal lobe of the caudal fin), and 

 the radials are free to concentrate and coalesce far from the vertebral 

 column (Fig. 46, B). Traces of the spines remain only in such forms 

 as the liajiformes, where the fin skeleton is closely connected by 

 them to the arches (Fig. 52). On the contrary, in the Teleostomi 

 the series of neural and haemal spines usually remains complete 

 and unaffected by the concentration of the fin-radials. These may 



