106 PISCES 



either articulate with (caudals of Eusthenopteron, Coelacanthus, 



etc.) or alternate with the spines (majority of Teleostei). In the 

 Teleostomi, then, the division between the axial and the fin skeleton 

 takes place at the distal end of the spines (p. 69). 



The skeleton of the paired fins diners considerably, not only 

 among the larger groups of the Pisces, but even among closely 

 allied families and genera. That the various t3 T pes have been 

 derived from some common ancestral form, by the fusion and 

 modification of originally separate and uniform elements, is most 

 probable (p. 73) ; but what that original type may have been it is 

 at present impossible to say for certain. Palaeontological evidence 

 alone could enable us to decide this question. 



According to Gegenbaur [157, 162], the original type resembled 

 the skeleton of the pectoral and pelvic fins of Ceratodus (Fig. 213). 

 Such an 'archipterygium' consisted of a median axis, articulating 

 with the girdle, and provided with an anterior preaxial and a 

 posterior postaxial series of radials. The radials were arranged in 

 pairs, diminishing in size towards the tapering distal extremity. 

 The archipterygium may be described as ' mesorachic ' and 

 ' rachiostichous ' (Lankester). That such a ' biserial ' fin skeleton 

 is a very ancient type can hardly be doubted. It is found not only 

 in archaic living fish (Ceratodus, Howes [218], Braus [48]), hut also 

 there is evidence that it was possessed by the Dipnoi and the 

 Osteolepidoti (Crossopterygii) of the Devonian epoch (p. 282). 

 Moreover, it is also found in the pectoral fin of the Carboniferous 

 Pleuracanthodii (p. 181, Fig. 148). 



As Gegenbaur has shown {158, 162], the various types of the 

 paired-fin skeleton found in the Teleostomi may be derived from the 

 archipterygium on the supposition that the axis has become much 

 shortened, that the radials have been reduced in number, and that 

 the postaxial radials have almost or entirely disappeared (Fig. QS), 

 The ' rhipidostichous ' type of skeleton of the paired fins of Elasmo- 

 branchs, in which the radials have a fanlike arrangement, may be 

 deduced in much the same way from the archipterygium. But here 

 the reduction has been less complete (Figs. 96, 134); and it is 

 unnecessary to assume that the pelvic fins of the Chondrichthyes 

 have ever fully conformed to the rachiostichous type, since even in 

 the Pleuracanthodii the pelvics are monostichous (Fig. 150), i.e. have 

 oidy one series of radials. 



The origin of the archipterygium itself remains to be explained. 

 The objections to Gcgenbaur's theory of its derivation from gill- 

 rays has been dealt with elsewhere (p. 74). Following Haswell 

 [198«] and Mollier [301], we may suppose that it has been formed 

 by the great concentration of a large number of radials to a very 

 narrow base, giving rise to a central axis, and leaving their distal 



