184 THE AMERICAN NATURALIST. (VoL. XXXV. 
The first of these is the obvious fact, to which my attention 
was first directed by Dr. H. S. Pratt, that lungless forms 
breathe in the same way as do other salamanders ; that is, 
they indulge in rhythmic respiratory movements during which 
the throat fluctuates in and out with considerable rapidity. 
A second closely related fact is that recorded by Bruner 
(1896), that, in common with salamanders with lungs, the sev- 
eral species of lungless forms investigated possess a muscular 
apparatus for opening and closing the anterior nares. Although 
in lunged forms this apparatus is directly connected with the 
respiratory act, Bruner, influenced by the general opinion that 
lungless forms could not really breathe, explained the presence 
of this apparatus in such genera as Desmognathus and 
Spelerpes as solely a protection against water. The third fact, 
which has remained entirely unexplained until now, is the 
presence in lungless salamanders of well-developed respiratory 
muscles identical with those possessed by other Urodeles and 
inserted into the walls of the pharynx and cesophagus (Wilder, 
1896). - > 
The meaning of these incongruous and apparently unrelated 
facts has at last been made clear through the careful investi- 
gations of one of my assistants, Miss Anne Ide Barrows, whose 
preliminary report on the subject has just appeared in the 
Anatomischer Anzeiger (Bd. XVIII, Nos. 18, 19) Miss 
Barrows has just concluded a long and exhaustive study of the 
entire circulatory system of Desmognathus fusca and has suc- 
ceeded in demonstrating that not only the pharynx, as shown by 
Bethge in Spelerpes, du¢ also the entire esophagus, is supplied 
with a dense capillary plexus, the vascular area of which is of 
Sufficient extent to transform the entire pharyngo-esophageal 
region into a functional lung of at least as great respiratory 
power as that of the paired lungs of normal salamanders. The 
plexus is formed mainly by four arteries, two external maxil- 
laries on the dorsal, and two pharyngeal on the ventral side. 
Posteriorly a few branches of the gastric artery contribute to 
the formation of the plexus. 
It is worthy of note that a large part of the plexus is formed 
by arteries from the fourth or respiratory arch, and that much 
