The Development of the Cartilaginous Skull ete. in Necturus. 443 
thelial tissue which is morphologically part of the wall of the coe- 
lom. The origin of these muscles from a composite tissue such as 
the »periaxiale Stränge« which Goronowrrscu (92, '93) describes as 
taking part in the formation of the branchial muscles in the bird, 
is out of the question. 
The hypoglossus musculature. 
In the embryo of 15 mm, the somitic divisions of the mesoderm 
of the head are limited to two somites, which lie respectively lateral 
to the two vagus roots. Divisions of the axial mesoderm anterior 
to the first vagus segment have been lost in scattered cells of the 
mesenchyma which surround the auditory vesicles and bound the 
walls of the brain, occupying much of the space that separates the 
neural tube from the external ectoderm. 
Although in younger embryos, the mesodermie divisions in the 
head anterior to the vagus segments are not as distinct in Necturus 
as in the Selachii, they nevertheless are constant, and correspond 
to the divisions determined by v. WisuHE (’82), except that the prae- 
mandibular mesoderm is not divided from the mandibular. v. WHE 
tells us regarding the divisions of the mesoderm in the head of the 
Selachii, that the first head-somite, the praemandibular, lies anterior 
to the mouth. The second head-somite is continued in the mesoderm 
of the mandibular arch. The third lies above the hyomandibular 
cleft. The fourth, immediately anterior to the auditory invagination, 
is widely connected with the mesoderm of the hyoid arch, while 
the fifth somite, the anterior part of which lies opposite the An- 
lage of the glossopharyngeal nerve, is connected with the mesoderm 
of the glossopharyngeal arch. From the sixth somite a myotome 
arises, and from each of the following somites of the head and 
trunk myotomes are formed. 
Save that there appears to be no early division in Necturus 
between the mesoderm of the praemandibular and mandibular so- 
mites, the above description of the divisions of the mesoderm in 
the Selachii in equally applicable to Necturus at the time when 
there are but three branchial clefts. Indications of further divisions 
of the praeotic mesoderm such as are described by DoHrx ('90) 
and Kinuian (91) in Torpedo, or by myself in Acanthias (91), 
do not appear in Necturus. I therefore felt warranted in stating 
in my first study (94, page 928) that the segmentation of the meso- 
derm in Necturus is the same as that described by v. Wısux. To 
