No. 522] A CARBONIFEROUS SALAMANDER 375 



Caudata arc forms which have evolved by a process of 

 degenerative evolution and lie has supported his con- 

 tention by a direct comparison of the skeletal anatomy 

 of the Branchiosauria and the Caudata. So far as the 

 skeleton is concerned the caudate organization is less 

 high than that of the Branchiosauria. Such can not be 

 said of the alimentary canal, for it at least lias progressed 

 in its development in many forms, although, as has been 

 stated above, some of the more slender species are ap- 

 parently restricted in their development. The fact that 

 a young form of Diemycti/lus torosus shows a closer ap- 

 proach to the fossil condition than any other is of ex- 

 treme interest. It can be said with certainty that the 

 Divmyctylus torosus goes through the phylogenetif 

 stages in the development of its alimentary canal and in 

 my opinion we have here a representative in. the fossil 

 condition of one of these phylogenetic stages. But too 

 much importance can easily be attached to the. close re- 

 semblances outlined above. It is to be remembered in 

 this connection that the form of the alimentary canal is 

 so highly modified by the food habits and by the form of 

 the body of the individual animal that it will be difficult 

 to draw safe conclusions. On the other hand, so far as 

 we have ascertained, the form of the body of the sala- 

 manders has not changed essentially from that of their 

 forebears, the Branchiosauria, so there is no really good 

 reason why the form of the alimentary canal should have 

 undergone much change. 



