THE DIGESTIVE AND RESPIRATORY SYSTEM 307 



three, in connection with which certain integumental struc- 

 tures arise which are gills physiologically, but are unrelated to 

 the former. The most widely distributed form of these is 

 that of the external branchic?, three in number upon each side, 

 and attached to the cartilaginous gill-arches. In structure 

 they are usually plumose or dendritic (Fig. 85, e and f), 

 although in a few cases they are thin and leaf-like. The slits 

 appear between these, with occasionally an additional one in 

 front of the first, and the animals obtain fresh water for res- 

 piration in part by forcing a current through these slits in the 

 manner of fishes, and in part by waving the branchiae up and 

 down by means of special muscles with which these organs 

 are furnished. As stated above, external branchiae are char- | 

 acteristic of the larvae of all amphibians, and are found per- j 

 manently in a few aquatic salamanders, which are either more 

 primitive than the rest, or are paedogenetic, that is, they retain 

 the larval form while becoming sexually mature. These sala- 

 manders are called perennibranchiate in distinction from those 

 in which the branchiae become lost, the caducibranchiate sala- 

 manders. A second form of gills which are external in origin 

 but become internal in position, occurs in frog larvae, where 

 they replace the former, which appear at first. As these are 

 plate-like and are attached to the gill-arches, they have often 

 been considered exactly homologous with the gills of fishes, 

 but their ectodermic origin renders such a conclusion impos- 

 sible. 



Aside from the two sets of branchiae most amphibians pos- 

 sess definite lungs, which arise in the larvae and exist for a 

 time side by side with the external branchiae, usually replacing 

 them in later life. These are often in the form of simple 

 sacs, without any formation of internal partitions, and even 

 when in their highest development, as in frogs, are far from 

 complex. It thus seems probable that, although they are true 

 lungs anatomically, they play a subordinate role in respiration, 

 and are perhaps primarily used either for regulating the spe- 

 cific gravity of the animal in the water, or in the production 

 of the voice, since the larynx is often very large and curiously 



