[69] THE EVOLUTION OF THE FINS OF FISHES. 1049 



biserial repetition of parts correlatively, >siicli as is apparent in the jVec- 

 toral and ventral fins, then I think it is fair to regard this as the oldest, 

 most unmodified, and nearest the condition of an archetypal c;audal 

 from which all the other modifications of the tail tin have been derived 

 by descent with modification. 



The Dipnoan Ceratodus, with its powerful pectoral and ventral limbs, 

 which exhibit such a singular biserial and segmented arrangement 

 of the elements of the limbs — even the muscles partaking of the last 

 character, so that they are in part segmented, according to Davidofl", 

 in a manner somewhat similar to the myotomes of the body — have had 

 a powerful influence in retarding the modification of the tail; for in no 

 Teleosts, with the exception, perhaps, of the Pedicnlati, do the muscles 

 of the appendicular skeleton extend so far outwards as to form a pe- 

 dunculate limb. In Ceratodus^ however, we must not be too hasty in 

 concluding that the segmented musculature of the paired fins is cor- 

 related with the segmented musculature or myotomes of the tail or 

 urosome. Since we know that metamerism begins to be developed 

 from within outAvard in the body, if we examine this condition as to 

 its origin in the limbs of Ceratodus, we will reacli a similar conclusion. 

 Embryology has proved that both the muscular and skeletal tissues 

 of the limbs are at first uusegmented ; from the mode of difierentiation of 

 the muscular buds thrust into the limb folds from the jnimitive somites 

 (gut-pouches), we may get an inkling of how the segmental arrange- 

 ment of the flexor and extensor muscles of the limbs of Ceratodus arose. 

 Our view is necessarily only tentative, because the development of the 

 Ibrm is not known, but it is certainly not venturing too much upon hy- 

 pothetical ground to suppose that the mode of outgrowth and concres- 

 cence of the muscular buds of the somites which entered into the forma- 

 tion of the musculature of the limbs gave rise to the segmentation of 

 the axial cartilage of the limb, which, as in other forms, must have been 

 a continuum at an early stage, just as the chorda remains unsegmented 

 so long as no osseous or distinctly chondrified tissues invade or replace 

 it, so as to make it rigid, and thus give rise to the necessity for the exist- 

 ence of segments or vertebral centra, in order to make the axis of the 

 body flexible and under tlie control of the forces exerted by the muscles. 



XI. — The tendency of HETEROCfEECY TOWARDS CiEPHYROCERCY. 



The general principle that the changes wrought by organic evolution 

 are progressively developed, is illustrated in the most forcible way by 

 the raor])hological history of the caudal fin of fishes. In the tail of 3IoIn 

 we have the extremest expression of gephyrocercy, while at the oppo- 

 site extreme of undifferentiated, jirimordial, embryonic caudal develop- 

 ment we must place the Dipnoans, Marsipobrauchs, and Leptocardians. 

 Taking a glance at our figures, we may readily verify the fact that het- 

 erocercy seems to be tending' towards gephyrocercy as the final term of 

 caudal differentiation. 



