AMPHIPNEUSTA. 273 



animals to respire air without ribs or diaphragm, 

 and without the power possessed by the tortoises 

 and frogs to cause it to enter by the nostrils, in 

 order that, so to speak, it might be swallowed, be- 

 cause the nostrils of the Sirens do not lead into 

 the mouth, and the branchial apertures must let 

 it escape. But his o\mi observations, made upon 

 well-preserved individuals, showed Cuvier that 

 the nostrils in the Siren do communicate with 

 the mouth by a hole pierced as in the Proteus, 

 between the lip and the palatal bone which carries 

 the teeth. The membranous opercula of their 

 branchi^ are muscular internally, and capable 

 of hermetically sealing the apertures ; then it is 

 very easy for the Siren, by dilating its throat, to 

 introduce the air into the mouth, and to force 

 it afterwards, by contracting the throat, into its 

 larynx. Even without this structure of the 

 nostrils, the animal could produce the same 

 effect by opening its lips a little : a theory 

 which Cuvier applies to the Proteus as well as 

 the Siren. 



Professor Owen has contributed some interest- 

 ing observations to the Penny Cyclopaedia,* on 

 the size of the blood-disks (commonly called 

 globules) in the Amphibia as compared with other 

 animals. Their size in these Reptiles is very 

 great, and their magnitude seems to bear a pro- 

 portion to the permanency of the external gills, 

 or hranchice. In the double-breathing animals 

 before us, the Sirens and Protei, these disks are 

 so large as to be distinguishable even with the 

 unassisted eye, while their appearance under a 

 microscope is exhibited in the accompanying 



* Vol. xxii. p. 61. 



T 



