RECAPITULA TION. 489 



not know how the paired Hmbs were transformed from the 

 radiate to the pentadactyle form. But we consider it most 

 probable that the Amphibia diverged at an early period 

 from the Dipnoan stem, and therefore arose from ancestors 

 which had fin rays in a continuous median fin, which 

 possessed paired limbs like those of the Dipnoans, and 

 scales or dermal ossifications. It is more probable that 

 the Amphibian larva has lost its paired limbs and fin rays 

 since the adult form was evolved, than that the adult 

 ancestor never jDOssessed paired limbs and that the penta- 

 dactyle limbs were evolved independently of the paired fins 

 of the Dipnoi. The Amphibian larva has also lost all trace 

 of scales or other dermal ossifications. The extinct Am- 

 phibia called Labyrinthodonts possessed dermal osseous 

 plates, and the existing- Coeciliidse possess scales. The 

 larva of the frog therefore, instead of recapitulating the 

 ancestral condition has lost nearly everything characteristic of 

 a fish except the pharyngeal branchiae, which it could not lose 

 without ceasing altogether to be an aquatic larva. The larva 

 therefore is much more of an adaptation than a recapitulation. 

 At a time when the doctrine of recapitulation had more 

 influence on phylogenetic inquiries than it has now, W. 

 K. Parker observed characters in the tadpole of the Anuran 

 Xenopus {^Dactylethra) which he considered more obviously 

 ancestral than those of the common frog. He regarded the 

 skull and the tail as remarkably chimseroid. The tail he 

 said terminated in a long pointed lash, the fin membrane 

 ceasing at some distance from its end. Mr. Beddard found 

 on examining specimens of this tadpole hatched in the Zoo- 

 logical Gardens, that the fin membrane really continued to 

 the extremity of the tail, but was very narrow, and the 

 description of Parker was a mistake due to the contraction 

 of the membrane in spirit specimens. There are other 

 peculiarities however in the Xenopus tadpole in which it 

 differs from the larva of the frog. The mouth is terminal, 

 there are no suckers on the chin, and the fore-limbs do not 

 develop beneath the opercular fold. These differences 

 show so many features in the frog larva which must be re- 

 garded as special modifications in that form. 



