362 Mr. Hogg on the Branchias in young Caeciliae. 



Sub-class II. DiPLOPNEUMENA. Respiiing doubly, both by gills 

 and lungs. 

 Order II. Iniperfectibranchia. Gills imperfect. 

 Family I. Menopomalida>. Body loiifr, lizard-like ; or lengthened, snake- 

 like ; with a tail ; legs four ; gill-like organs internal. 

 Genera : Menopoma, Amphiuina. 



Order III. Manentibrancliia. G'lWs permanent. 

 Tribe I. Raniibranchia. G'lWs ramified iiv tufted. 

 Family I. Sirenidce. Body lengthened, snake-like, having a tail; legs two 



in front ; gills tufted, external. 

 Genera : Siren, Parvihranchus. 

 Family II. Proteidce. Body long, lizard-like, or fish-like, with a tail; 



legs four ; gills ramified, external. 

 Genera : Proteus, Menohranchiis, Siredon. 



Tribe II. P^imbribranchia. G\\\& fringed. 

 Family I. Amphihichthijidce. Body lengthened or long, fish-like, covered 

 with scales, having a tail ; dorsal and caudal membranes, resembling 

 fins, strengthened by soft rays ; legs four, rudimentary ; gills fimbri- 

 ated, internal. 

 Genus : Jmpliihichthys (Lepidosiren). 



It is worthy of remark, that in comparing the gradual mo- 

 difications in the organization presented by the different fami- 

 lies in this very natural class, there will be found many sin- 

 gular resemblances even between the two extreme groups, 

 the C(eciliad(S and the Amjihibichthyidie ; inasmuch as they 

 both possess scales, and the former seem to be furnished, 

 in their young state, with the same kind oi fringed gills, 

 concealed within the branchial cavity, as the latter retain 

 during the whole of life : and whilst, on the one hand, the 

 Cceciliada are snake-like in their form and habits, they con- 

 stitute the link between the class Reptilia and the class Am- 

 phibia-, so, on the other hand, the fish-like shape and characters 

 of the Amphibichthyidce as clearly and gradually connect the 

 class Amphibia with the class Pisces, both approximations 

 being carried on in a most extraordinary and beautiful man- 

 ner. And I have before noticed, that the order Abranchia, 

 which had been previously formed for a certain genus in this 

 class, cannot be applied to any Amphibian ; for it is now well 

 ascertained that every animal included in this class possesses, 

 during one period of its existence at least, some branchial ap- 

 paratus, which, with the retention of lungs, fully proves that 

 these animals ought, according to their natural conformation, 

 to be arranged in a distinct class, and not in a mere order of 

 the class Reptilia. Wherefore the principal characters of the 

 three latter classes of Animalia Ve7'tebrata, — Reptilia, Am- 

 phibia, Pisces, — taken from their organs of respiration, are. 



