No. 393-] AUTODAX LUGUBRIS HALLOW. 703 
In concluding this fragmentary contribution to the life his- 
tory of this interesting salamander, we may remark that we are 
now directing our efforts to the securing of material, particu- 
larly embryonic material, that will enable us to complete the 
story. In the meantime we may call attention to some of 
the exceedingly interesting, but equally difficult problems of 
adaptation that are presented by our California salamanders. 
The species of Autodax appear to be among the most terres- 
trial of American Urodela.! They are, nevertheless, lungless, 
and are possessed throughout life of a delicate, 
smooth, moist skin. Déemyctylus torosus, on 
the other hand, which has practically the same 
geographical range, and hence is subject to & ; 
the same climatic conditions, while being 
thoroughly aquatic for a considerable portion 
of its life, has developed an exceedingly hard 
and rough epiderm well fitted to resist desicca- 
tion ; and, besides, it possesses well-developed 
lungs. In other words, the adaptive modifica- 
tions in the two forms have gone in directions 
opposite to what might have been expected. 
We should have supposed that whatever the 
immediate influences may have been which 
caused the disappearance of the lungs, these 
would have been most potent in animals that 
passed the most of their time in water; 
and we should likewise have supposed that it would be just 
the animals that lived exclusively on land, in the air, that 
would have taken on the dry, hard, rough epiderm. And 
the case is the more puzzling from the certainty which we 
have, Wilder (94), that in some species at least, and hence 
inferentially in all, the lungless condition is secondary and 
Fic. 7. 
1 Wilder’s ('99) recent interesting paper on Desmognathus fusca and Spelerpes 
bilineatus shows that the former species approaches Autodax in this regard, but 
here there is a true aquatic larval period despite the fact that the eggs are laid on 
land. Plethodon oregonensis and Batrachoseps attenuatus, two geographical neigh- 
bors to Autodax, will be found, we are quite convinced, to be entirely terrestrial ; 
but full evidence on the point is not yet at hand. And Wilder ('94) gives reason 
for supposing that Plethodon erythronotus is entirely terrestrial. 
