[27] THE EVOLUTION OF THE FINS OF FISHES. 1007 



is evident that various displacements have occurred, * due to growth, 

 accompanied by a coalescence in the peduncular region of the basal 

 actiuophores of the paired fins, which has obscured the homologies of 

 their parts and also developed the coalesced series of paired nerves so 

 as to develop the plexuses which innervate the limbs. 



The evidence of metameric or serial homology in the unpaired fins is 

 too palpable to need serious discussion, for we find one spine bearing 

 accessory radii to correspond with several somites (dorsal of Polypterus)^ 

 or one spine or ray to each somite, or two or even three or even more 

 to a single somite. This is the clearest possible proof, and goes far to 

 supplement the ontogenetic evidence respecting the metameric origin 

 of the basipterygial elements of the paired tins. 



The case of the dorsal of Polypterus is a remarkable one, and has puz- 

 zled me greatly until recently, when I noticed that the strong, bony, 

 anteriorly enamel-covered spines, with their posterior accessory rays, 

 were probably not homonomous, or all derived from a single segment^ 

 but probably homodynamous, as Gegenbaur would express it, or seri- 

 ally homologous, according to Owen. This becomes evident in the 

 light of the ontogeny of the basipterygium, as worked out by Balfour 

 and Dohrn, for we find' that the principal spines are homonomous with 

 about every fourth vertebral spine, so that about three intervening 

 spines do not apparently support rays or spines, and are thus without 

 homonyms ; but in view of the way in which the principal spines sup^ 

 port accessory jointed radii posteriorly, it is probably fair to conclude 

 that these accessory rays are the homonyms of the intervening appar- 

 ently spineless segments between the great spines, and that in the 

 course of development the bases of the accessory rays have been shoved 

 forward out of their original relations with these segments, as the pos- 

 terior actiuophores of the pectorals of Elasmobranchs have been, and 

 crowded up against the principal anterior ones and carried up and 

 away from the body as these grew in length. The process is therefore 

 one which in all probability is perfectly i^arallel with that involved in 

 the production of the so-called archipterygium, so that the primordial 

 continuous system of median dorsal homodynamous rays has been 

 interrui:»ted in Polypterus so as to develop discrete dorsal finlets by a 

 process of proximal concrescence of sixteen short series, consisting of 

 about four to five rays each. The foregoing hypothesis of the true 

 nature of the dorsal of Polypterus we may, I think it j^robable, confi- 

 dently expect to be confirmed by ontogenetic research. 



* See my remarks upon the rotation of the pectoral of Gadus upon its base (Con- 

 trib. Embryog. Osseous Fishes, pp. 06, 67). It is evident that such a rotation does 

 not take place so soon in Elasmobranchs. The base of the hn in them permanently 

 occupies a more primitive position or one which is much less different from the di- 

 rection of the original fold. This is especially noteworthy in the Rays, in which the 

 base of the jiectoral of the adult undergoes almost no alteration of position in rela- 

 tion to the position of the fold from which it is derived. (Sec Wyman's memoir on 

 the development of Baia batis.) 



