HYOMANDIBULA OF THE GNATHOSTOME FISHES 617 
petroso-lateralis, and its distal end, which is said to be primarily 
connected with the ‘squamosum,’ secondarily has that connec- 
tion transferred to the underlying palatoquadrate. Thesecells 
thus have the relation to the vein that either the extrabranchial 
or the interarcual cartilage may have to the vein in the hyal 
arch of fishes, and to determine which one of them is represented 
the relations of the stylus to the nerve, artery and vein of the 
arch should be known. This relation of the stylus to the nerve 
is given by Kingsbury and Reed, and in this respect the Urodela 
examined by them present two distinctly different groups. In 
the larger group the nervus hyomandibularis facialis is said (1. 
c., p. 610) to lie ventral and anterior (‘cephalad’) to the stylus, 
but in Necturus, Proteus and Typhlomolge the two branches of 
the nerve straddle the stylus, the ramus mandibularis lying ven- 
tral and anterior to it and the ramus jugularis (hyoideus) dorsal 
and posterior to it. 
In both of these two groups of the Amphibia the basal plate 
of the columella is quite certainly, as just above stated, the 
pharyngeal element of the hyal arch. In the first mentioned 
group the stylus has the relations to the nervus hyomandibularis 
facialis that the posterior articular head of the teleostean hyo- 
mandibula has, and is probably the extrabranchial of the hyal 
arch. In the second group the stylus has the relations to the 
nerve that the hyomandibula of Polypterus has, and it is accord- 
ingly uncertain what element of the arch it represents. In the 
frog, the plectrum, which is said by Kingsbury and Reed to be the 
homologue of the columella of the Urodela, is said to lie below and 
anterior to the entire nervus hyomandibularis facialis, and accord- 
ingly has the relations to that nerve that the dorsal process of 
the hyomandibula of Ceratodus has, and is probably the inter- 
arcual cartilage that lies: between the hyal and mandibular 
arches. It accordingly seems quite certain that the columellae 
in these three groups of the Amphibia are not homologous struc- 
tures, and this offers a much more plausible explanation of the 
differing relations of the nervus hyomandibularis to the colu- 
mella than the one usually given that the nerve has radically 
changed, for some unknown reason, its relations to that skeletal 
