THE ENDOSKELETON 185 



and position recalling the claws or nails of the hand-form, al- 

 though placed on both sides instead of one. Even the usually 

 excessive number of the fin-rays is no real objection, since in 

 some fishes this number is a very limited one, and the theories 

 of prse-pollex and post-minimus point to a previous larger 

 number. Thus in a way the derivation of hand or foot from 

 the fish-fin may be accounted for ; but an insuperable difficulty 

 lies in the presence of the two lengths of limb bones inter- 

 posed between the girdle and the distal complex, which seem 

 to correspond with nothing found in the fin. It is to be re- 

 membered, however, that these are not especially long in the 

 more primitive forms, like salamanders, so that the main 

 obstacle lies not so much in their length as in their very 

 existence, which has never received a satisfactory explanation. 



An early suggestion along this line was based upon the 

 anterior fin of selachians with its three basal pieces (Fig. 

 51, a). The mesopterygium is usually much the largest and 

 shows a tendency to form the sole connection with the shoul- 

 der-girdle. If this becomes established, the mesopterygium be- 

 comes pJtfji*La*ue i and the pro- and metafirerpquun. slipping 

 away from the girdle, and bearing the free rays, would become 

 respectively radius and ulna (Fig. 51, b and c). This theory 

 seemed for a time to receive especial corroboration from the 

 structure of the paddle of the extinct sea-lizards, Ichthyosaurus 

 and Pleisiosaurus, but the time is now passed for drawing 

 such broad conclusions from a chance resemblance in some 

 highly specialized form, aside from which it must be noted 

 that the theory is based upon the anterior fin alone, leaving 

 the more primitive posterior one out of the question. 



When, a little after this, the biserial dipnoid fin was heralded 

 as the primitive type, and named in consequence the " archi- 

 pterygium," it turned thought in a new direction, and an effort 

 was made to seek in the more primitive cases of the hand-form 

 a central axis with lateral rays proceeding from it. In one 

 such attempt the central axis was formed by humerus, ulna, 

 ulnare, carpale V and the fifth digit, to which the remaining 

 bones served as lateral rays upon the inner or radial side, 



