MEDIAN FINS 



reaches, or articulates with, the neural spines (p. 105), and in the 

 Dipnoi there is no distinction between the two (p. 234). Moreover, 

 the ventral lobe of the caudal fin is almost always supported by 

 direct prolongations of the haemal arches; and the dorsal lobe 

 in Selachians is usually borne^ by cartilaginous median 'spines/ 

 articulated to the vertebral column, yet more numerous than the 

 vertebrae (Fig. 46, B). All these facts seem to uphold the second 

 hypothesis, according to which the 'neural spines' and the fin- 

 radials represent the proximal and distal regions of an originally 



Fio. 46. 



Diagrams showing the modifications of caudal fins, and the relations of the cndoskeletal 

 radials of median tins to the axial skeleton. A, diphycercal type, with equal dorsal and ventral 

 lobes (Dipnoi) ; B, heterocercal type (Selachii) ; C, modified diphycercal (Coelacanthini) ; D, 

 heterocercal (Chondrostei) ; E, homocercal type (Teleostei) ; P, abbreviate heterocercal type 

 (Amioidei). a./, anal tin ; ax.l, axial lobe ; c.r, epichordal radial ; def, dorsal lobe partly 

 formed by epichordal lobe ; d.f, dorsal tin ; ef, epichordal lobe ; h.a, haemal arch ; Jif, 

 hypochordal lobe ; hy, hypural arch ; n.a, neural arch ; nt, notochord ; r, radial. The 

 endoskeleton is black. 



continuous skeletal rod with several joints. We should not, then, 

 believe that the fin -skeleton has developed independently at the 

 base of the fin itself, and has grown inwards centripetally ; but, 

 on the contrary, we should consider the wide separation of the 

 radials from the vertebral column in some fish (Elasmobranchs) as 

 secondary. Thus the discrepancies in number, and the want of 

 concordance in position, of the arches and fin-radials would be due 

 to concentration and modification of the originally corresponding 

 elements (Fig. 46) ; the nonconformity of the two sets of supports 

 may be either the cause or the effect of their separation. 



