64 THE COMMON FROG. [chap. 



with the ventral plates save by means of soft struc- 

 tures. In these latter Chelonians then we have 

 in reptiles an interesting approximation to the con- 

 dition we have seen to exist in those exceptional 

 Anourans, Ceratophrys and Ephippifer. Moreover 

 this resemblance is still further increased by the 

 fact that in Trionyx the bony plates are not covered 

 with any tortoise-shell, but are merely invested 

 by soft skin as in the genera of dorsally-shielded 

 Batrachians. 



Have we then here a true sign of genetic affinity ? 

 Are these tortoises to be deemed the more specially 

 modified, descendants of shielded frogs or of some, as 

 yet unknown, slightly-shielded animals which were the 

 common ancestors both of frogs and tortoises ? 



Certainly tortoises cannot be the direct descendants 

 of frogs, they agree with all reptiles in characters 

 which are both too numerous and too important to 

 allow such an opinion to be entertained for a 

 moment. 



The other opinion is hardly less untenable ; for if 

 all the multitudinous species of frogs (together with 

 a number of reptilian forms more closely allied to the 

 tortoise than any frogs are) descended from slightly 

 shielded animals, how comes it that all frogs and 

 toads, save one or two species in no other way 

 peculiar, have every one of them lost every trace of 

 such shielded structure, which nevertheless cannot 

 easily be conceived to have been in any way pi'eju- 

 dicidl to their existence and survival } 



On the other hand, it cannot but strike us with 

 surprise that structures so similar — extending even to 



