OP THE COLD-BLOODED VEBTEBRATA. 223 



9. The enclosure of the gills and branchial arches in a branchial 

 chamber. Professor Owen reminds us that "the larva of the 

 tailless Batrachian presents at one period of its existence a 

 similar structure." 



10. The external appearance and habits of the living animal. 

 The mode in which the Zepidosircn (several healthy specimens 

 of which have, within the last few years, been received from 

 Africa) comes to the surface of the water to breathe, and the 

 use which the animal makes of its anomalous limbs, are indeed 

 suggestive of its amphibian affinities, though very insufficient 

 to prove them. 



On the other hand, the structure of the brain in Lejndosireny 

 its external branchise, the presence of a distinct left auricle, and 

 of posterior nares, must be viewed as strongly indicative, to say 

 the least, of its amphibian nature. 



Supposing Lepidosiren to be an Amphibian, few would hesitate 

 to place it among the Perennibranchiaus ; but there is no single 

 group of the class of Pishes with which it manifests an equal rela- 

 tionship. 



It is further to be observed, that when Professor Owen \ATote 

 his "Description of the Zepidosircn anneetens,'' he was \inac- 

 quainted with the subsequently constituted genus Archegosmmis, 

 in which a persistent notochord and the absence of occipital con- 

 dyles are associated with unquestionable amphibian characters. 

 Moreover, the heart of Lepidosiren has since been proved to possess 

 two auricles, and its nasal sacs have ceased to be accounted blind. 



Although it may fairly be inferred that Lepidosiren is more like 

 the Perennibranchiate Batrachians than any known group of Pishes, 

 yet its strongly marked ichthyic characters forbid our receiving this 

 conclusion without considerable qualification. Had Lepidosiren 

 been only known in a fossil state, what anatomist would have 

 scrupled to place it in the class of Fishes ? 



If, then, we admit the amphibian affinities of Lepidosiren, and 

 acknowledge also its evident iclithyic relations, the conclusion 

 deducible from the foregoing considerations may be stated tlius : — 



It is possible to define the limits between two great groups or 

 classes of Branchiate Vertebrates. These classes are confessedly 

 artificial. Lepidosir^en may have its place on the Amphibian side 

 of the boundary. 



If Lepidosiren be not an Amphibian, the union of Pisces and 

 Amphibia into one class becomes inevitable. Such a class might 

 receive the name of Anallantoidea. 



