190 BULLETIN: MUSEUM OF COMPARATIVE ZOOLOGY. 
median wall migrate into the centre of the cavity, and cells bounding 
the inner wall above and below assume the elongated contour of muscle 
cells. 
Hoffmann (’94, p. 649) also, while not able to state with definiteness 
that the anterior cavity is a dorsal or lateral diverticulum from the 
alimentary canal, i.e. whether it represent a mesoderm segment or a 
visceral pouch, considers it probable that it represents the former, since 
it is very similar to the succeeding head cavities of van Wijhe. Hoff- 
mann mentions the migration of cells into the cavity of the somite, but 
does not specify from which wall they are proliferated. He also states 
that from the walls of the somite “entstehen keine Muskelfasern ” (’96, 
p. 256). Against these views of Miss Platt and Hoffmann no special 
arguments have as yet been raised. 
The first somite of van Wijhe possesses the peculiarity of a median 
stalk connecting the somites of the opposite sides of the body.* The re- 
lations of this stalk to the dorsal wall of the alimentary canal, to chorda, 
and to dorsal aorta have been used as the chief criteria in contending 
for its dorsal or its ventral nature. The evidences that the first somite 
represents somatic (dorsal) mesoderm are as follows: (1) Its cells are 
proliferated from the dorsal wall of the alimentary canal (Platt, Br) 
1 Such a median connection, however, also appears in the early stages of develop- 
ment of the “anterior cavities.” The connecting stalk of the “anterior cavities,” 
however, as stated by Hoffmann, never possesses a lumen, as does the median con- 
necting stalk of the premandibular cavities. 
2 Miss Platt, in her earlier paper (’91, p. 81), states that the mesoderm of the 
premandibular cavity is formed, at least in part, by a proliferation of cells from 
the mandibular cavity, while in her later paper (918, p. 256) she writes, “ The 
most anterior mesoderm of the head does not take its origin from the mesodermic 
plates, but from the dorsal wall of the alimentary canal, The mesodermic plates end 
with the mandibular cavities.” The lumen of the connecting stalk, according to Miss 
Platt, is, as stated by Marshall (’81), formed secondarily by the fusion of a median 
with the two lateral cavities. This evidence is interesting, since it bears on the 
question whether the cavity of the connecting stalk is to be regarded as a part of 
the archenteron, and would seem to answer this question in the negative. Killian 
C91, p. 102), however, finds the connecting stalk a “Sklerotomkommissur,” and 
thus, it is to be inferred, the lumen of the stalk, which according to his account is 
formed secondarily, not a part of the archenteron. He states (p. 102): “Erwáhnt 
sei noch, dass zwischen den beiden ersten Mandibularsomiten vor dem vorderen 
Chordaende und úber dem Aortensinus ein Mesodermzellenhaufen liegt, der die 
Sklerotomanteile beides Somiten in Verbindung setzt (Sklerotomkommissur). 
Was.nun der Oralzone angeht, so entsteht sie dadurch, dass die vordersten Zipfel 
des ursprünglich schwalbenschwanzförmig endenden dorsalen Mesoderms die vor- 
dere Darmkuppe (vordere Ektodermtasche von Petromyzon nach Kupffer) über- 
und umwachsen, und so einen medianen Zellkomplex bilden, aus dessen hinterer 
