THE ENDOSKELETON 



179 



facts which forbid us from imagining that it is really a primi- 

 tive condition. Thus the animal in which it comes to its most 

 perfect realization is the turtle, the carpus of which is almost 

 diagrammatic, while the salamanders, where a more primitive 

 type is to be expected, depart almost as widely from the dia- 

 grams as do the mammals. From certain indications it seems 

 probable that in the early carpus and tarsus there were two 

 centralia (Fig. 48, a), and that the separate bones were ar- 



in iv 



FIG. 48. Various forms of carpus. Figures (a)-(c) after ELISA 

 NORSA; figure (e), after FLOWER. 



(a) Sphenodon (Hatteria), a New Zealand lizard, (b) Chick embryo, early 

 stage. (c) Chick embryo, later stage, (d) Lacerta, a European lizard, (e) Talpa, 

 European mole, (f) Pig. 



R, radius; U, ulnar; r, radiale; u, ulnare; *, intermedium; c, centrale; 1-5, car- 

 pal ia; p, pisiforme; f. os falciforme; I-V digits. 



ranged, not symmetrically, but in oblique rows continued more 

 or less directly to the digits, and suggesting the derivation of 

 both carpus [and tarsus] and digits from long fin-rays, divided 

 into numerous joints. 



In the nomenclature of the carpal and tarsal bones em- 

 ployed in human anatomy we have an unusually good illustra- 



