THE CHEIROPTERYGIUM. 619 



In the Urodela (Gotte) the ulnar and fibular sides are 

 primitively dorsal, and the radial and tibial ventral : in_ Mam- 

 malia however Kolliker states that the radial and tibial edges 

 are from the first anterior. 



The exact changes of position undergone by the limbs in the 

 course of development are not fully understood. To suit a 

 terrestrial mode of life the flexures of the two limbs become 

 gradually more and more opposite, till in Mammalia the corre- 

 sponding joints of the two limbs are turned in completely 

 opposite directions. 



Within the mesoblast of the limbs a continuous blastema 

 becomes formed, which constitutes the first trace of the skeleton 

 of the limb. The corresponding elements of the two limbs, 

 viz. the humerus and femur, radius and tibia, ulna and fibula, 

 carpal and tarsal bones, metacarpals and metatarsals, and 

 digits, become differentiated within this, by the conversion 

 of definite regions into cartilage, which may either be completely 

 distinct or be at first united. These cartilaginous elements 

 subsequently ossify. 



The later development of the parts, more especially of the carpus and 

 tarsus, has been made the subject of considerable study ; and important 

 results have been thereby obtained as to the homology of the various 

 carpal and tarsal bones throughout the Vertebrata ; but this subject is too 

 special to be treated of here. The early development, including the suc- 

 cession of the growth of the different parts, and the extent of continuity 

 primitively obtaining between them, has on the other hand been but little 

 investigated ; recently however the development of the limbs in the Uro- 

 dela has been worked out in this way by two anatomists, Gotte (No. 482) 

 and Strasser (No. 487), and their results, though not on all points in com- 

 plete harmony, are of considerable interest, more especially in their bearing 

 on the derivation of the pentadactyloid limb from the piscine fin. Till 

 however further investigations of the same nature have been made upon 

 other types, the conclusions to be drawn from Gotte and Strasser's observa- 

 tions must be regarded as somewhat provisional, the actual interpretation 

 of various ontological processes being very uncertain. 



The forms investigated are Triton and Salamandra. We may remind 

 the reader that the hand of the Urodela has four digits, and the foot five, 

 the fifth digit being absent in the hand 1 . In Triton the proximal row of 

 carpal bones consists (using Gegenbaur's nomenclature) of (i) a radiale, and 

 (2 and 3) an intermedium and ulnare, partially united. The distal row is 

 formed of four carpals, of which the first often does not support the first 



1 This seems to me clearly to follow from Gotte and Strasser's observations. 



