ORDERS OF AMPHIBIA. 4II 



attains a length of about a foot, and has two pairs of weak 

 limbs, of which the anterior have three toes, and the posterior 

 only two. From its habitat, the power of vision must be quite 

 unnecessary, and, as a matter of fact, the eyes are altogether 

 rudimentary, and are cbvered by the skin. Several varieties 

 oi Proteus are known, and the one figured above has been 

 described as a distinct species (P. xanthostidus). The blood- 

 corpuscles in Proteus are oval in shape, and are larger than 

 those of any other Vertebrate animal. 



Fig. 156. — Head and fore-part of the body q( Proteus a7iguiHUS, showing the 

 external branchiae and tridactylous fore-limb. 



Of the Sirenida, the most familiar are the Sirens and the 

 Axolotls. The Siren, or Mud-eel, is found abundantly in the 

 rice-swamps of South Carolina, and attains a length of three 

 feet. The branchiae are persistent, and the hinder pair of legs 

 wholly wanting. Two other species are known, but they are 

 likewise confined to North America. 



The Mexican Axolotl (Siredon pisdforme, fig. 157) is a native 

 of the Mexican lakes, and attains a length of about a foot or 

 fourteen inches. It possesses both pairs of Umbs, the anterior 

 pair having four toes and the hinder pair five toes. As ordi- 

 narily known in its native country, the Axolotl is certainly 

 perennibranchiate, and they breed in this condition freely. 

 There is no doubt, however, that individual specimens may 

 lose their gills, without thereby suffering any apparent change, 

 except it be one of colour. The Axolotl, therefore, is in the 

 singular position of being sometimes " caducibranchiate," 

 whilst it is ordinarily "perennibranchiate."* Nearly allied 



* Professor Marsh of New Haven has recently shown that the. Siredon 

 lichenoides of the western States of America, when kept in confinement, 

 loses its gills, and dorsal and caudal fins, whilst it changes much in colour, 

 and undergoes various minor modifications in structure. Its habits, also, 

 become less aquatic, and it becomes apparently absolutely identical with 

 AmUystoma mavortium, a Salamandroid. This discovery has thrown con- 

 siderable doubt upon the value of the distinction between perennibranchiate 

 and caducibranchiate Amphibians, and has rendered it probable that all 

 the species of Siredon are merely larval Salamanders, as long ago suspected 

 by Cuvier. At the same time, the Axolotls certainly breed freely whilst in 

 possession of their branchia;, and there is as yet no proof that they lose 

 their gills whilst in a state of nature. 

 19 



