362 AXXALS yEY\' YORK ACADEMY OF SCIEXCE^ 



Fleshy lobe of fin Arm and hand 



Mesoptery?inm (single basal piece j Humerus 



Mesoptervgial axis Humerus, ulna, ulnare. 'iigit V (?) 



Preaxial parameres (radials) reduced Losr 



or absent 



Postaxial parameres Radius, carpus, digits I-IV (?) 



Dermal rays Lost 



The pectoral limb of Sauripterus differs from the tetrapod type in the 

 folloTring characters: (1) In Sauripterus, as in other Ehipidistia. the 

 chief ascending blade of the shonlder-girdle is formed by the cleithrum, 

 the scapulocoracoid is small and the articular region is promberant: 

 whereas even in the most ancient known Tetrapoda the cleithrum is of 

 relatively small size, the chief ascending blade is formed by the scapulo- 

 coracoid. and the articular region is sunk below the plane of the sur- 

 rounding parts. (2) The dermal rays (lepidotrichia ) are well developed 

 (lost in Tetrapoda: cf., modern Dipnoi). (3) The limb is articulated 

 more on the postero-interior border of the ascending blade than on its 

 outer face, and is directed backward rather than outward and downward. 

 (4) The supposed radius, ulna, carpals and digits differ widely in form 

 from these elements in the Amphibia, so that their supposed homologies 

 are only recognizable after close study. (5) The successive segments of 

 the limb apparently were not sharply inclined to each other, as at the 

 elbow and wrist of Tetrapoda. but were arranged radially more like the 

 axonosts and baseosts of ordinary fins. (6) While the evidence is not 

 positive, the distal rods seem to dichotomize and even the undivided 

 radials exceed in number the digits of the tetrapod manus. (T'i The fin 

 as a whole conforms to the imperfect archipterygial type seen in Mega- 

 Irclitliys and Eus^henopteron. rather than to the cheiropterygial type of 

 t^trapods. TThile all these differences may 1:>e simply primitive char- 

 acters, separating members of two distinct classes, nevertheless they 

 raise the question whether the tetrapod resemblances in the pectoral fin 

 of rhizodonts may not be fortuitous, and without phylogenetic signifi- 

 cance: but in view of the differences in function of the Sauripterus 

 paddle and a true cheiropterygium such underlying similarities as may 

 exist can hardly be attributed to convergence, and when taken in con- 

 nection with the resemblances ia the skull above noted (pp. 332-337) 

 they gain in importance. 



Dr. Broom's view that the tetrapod cheiropterygium developed only 

 from the anteroventral border of a ])Te-Sauripterus stasre appears to me 

 to lack adequate evidence. Xor do I consider the presence of dermal 

 rays as unfavorable to the development of a cheiropterygium. The re- 



