[25] THE EVOLUTION OF THE FINS OF FISHES. 1005 



phoreSj which correspond to the segaients whence the mu&cles are de- 

 rived, coalesce at their proximal ends to form the basipterygial carti- 

 laginous plate, Fig. 8, PL X, which subsequently segments to form the 

 pro- meso- and meta pterygium. The antero-posterior constriction of the 

 horizontal peduncles of the pectoral and pelvic fins is greatest from be- 

 hind, so that a fold of skin is pushed in behind the basipterygial plate, 

 shoving it outwards and freeing it posteriorly from the sides of the 

 body to thus give rise to the uniserial limb of these forms. The inner- 

 vation of these fins is also segmental. The pectoral and pelvic girdles, 

 according to Balfour, are at first integral anterior parts of the basip- 

 terygial plate and become segmented off afterwards from the basiptery- 

 gium. The coracoid and pubic are then developed still later in a proximal 

 direction as continuations of the scapula and ilium. Dohrn's researches 

 therefore enable us to complete the theory of the paired limbs of Elasmo- 

 branchs, and lead us to the conclusion that such a biserial limb as that of 

 Ceratodus is merely the result of the concrescence of j^rimitive segmental 

 rays or embryonic actinophores from behind and from before instead of 

 from behind only, or possibly the middle or axial actiuophore is composed 

 of the bases of a lateral series fused together and secondarily segmented. 

 At any rate, it would not be surprising to find that when the develop- 

 ment of Ceratodus is known that it had almost continuous lateral fin- 

 folds, along the whole length of which actinophores were developed, 

 one actinophore of the series in each paired fin being accelerated in 

 growth longitudinally to form the axis of the limb, and, as the fin grad- 

 ually became pedunculate, carried its anterior and posterior serial fel- 

 lows outwards along with it. The primitive character of its skeleton, 

 the coelacanthous structure of its median arches, interspinous bones, its 

 cartilaginous biserial and multiradiate paired tins, and its simple skull 

 are remarkably embryonic in every j>articular, and contrast in the most 

 remarkable way with its specialized breathing apparatus. Some such 

 manner of development as described above, if it somewhat exaggerated 

 that which is figured by Balfour (Comp. EmbryoL, II, 505, Fig. 348) as 

 obtaining for the pectoral skeleton of an embryo of Scyllium stellare^ 

 might lead to the evolution of a fin like that in Ceratodus^ without any 

 very extensive anterior coalescence of the segmentally-arranged actino- 

 phores of the fin-fold and lead to the formation of a biserial limb. At 

 any rate, I would expect to find the lateral fin-folds of greater relative 

 length in Ceratodus embryos than in most of those of the /Squal% Cross- 

 opterygians, or Chimseroids, thus contrasting, but not quite so decidedly, 

 as the fin-folds of these latter contrast in length with the short, paired 

 fin-folds of Teleosts. This last contrast is in accord with the conclu- 

 sions to be drawn from the facts of embryology generally, as to the 

 highly specialized character of the Teleosts when compared with other 

 groups of fishes. Pronounced reduction in the length of the lateral fin- 

 folds of Teleosts has occurred, so that we should not expect to find in 



