24 Observations uu the Natural History 



a little from one side of the abdomen to the other. So, in the 

 siren, we see the trachea to open directly into the lungs, 

 which, as in the above-mentioned larvae, says M. Cuvier, " sent 

 deux longs sacs cylindriques, que s"'etendent jusqu"" a Textremite 

 posterieure de Fabdomen, et se replient meme alors en avant."" 

 But, in the pi'oteus, neither do the supposed lungs reach to the 

 pelvis, nor does the supposed glottis open into the air-bladders, 

 but issue in a cavity which communicates with the air-bladders 

 by two long conduits. Thus, then, the structure of the bran- 

 chial arches, the distribution of the bloodvessels, and the form 

 and size of the lungs in the proteus, differ entirely from the 

 corresponding organs in the siren and larvae of the salamander. 



If, farther, we consider the mode in which frogs and sala- 

 manders respire air, and compare it with that of the proteus, 

 we shall obtain still farther evidence of the differences sub- 

 sisting between them. All zoologists, including M. Cuvier, 

 now admit that frogs first receive air into the mouth through 

 the nostrils only, and from thence force it into the lungs by an 

 action resembling deglutition. But neither the proteus nor the 

 siren are able to respire in this manner ; for the nostrils in the 

 former do not open into the mouth, but beneath the upper lip ; 

 and in the siren, " les narines, simplement creusees sur les 

 cotes du museau, ne penetrent point dans la houclie^'' says M. 

 Cuvier. Neither do these animals respire air in the manner of 

 serpents, for they are both destitute of ribs. When also the 

 proteus takes air into the mouth, it escapes rapidly through the 

 branchial apertures : nor is there any ground for believing that 

 any portion of it enters the very narrow chink of the glottis to 

 pass into its cavity, and from thence through the two mem- 

 branous canals into the air-bladders. No muscular structure 

 suited to produce such effects exists, and the fine membranous 

 canals, subject to compression every instant from the stomach, 

 altogether unfit them for performing the office of air-tubes or 

 bronchi. In all reptiles that respire air, the structure of the 

 organs is such as to permit free inspiration and expiration, how- 

 ever different the form may be ; but in the proteus, the want 

 of ribs and diaphragm, the fact that the nostrils do not open 

 into the mouth, the extreme narrowness of the aperture termed 

 glottis, and the narrowness, length, and compressibility of the 



