NEAL: NERVOUS SYSTEM IN SQUALUS ACANTHIAS. 201 
connection with the ectoderm immediately posterior to the infundibulum. 
Still later (40 somites) cavities appear both in the median connecting 
stalk and in the lateral meso- 
derm, and these by their fusion 
form the continuous cavity in 
the manner already described 
by Miss Platt (91°). It follows “4? = =- 
therefore that the premandib- “6”... LA 
ular cavities comprise dorsal Sa 
and only dorsal mesoderm,! 
1 Hoffmann (’94, p. 648), how- 
ever, finds evidence of a splanchnic 
portion of the premandibular so- I 
mite in a “ Zellstrang, welcher Titos D. 
dem Mandibularbogen parallel ver- 
läuft und der Vorderfläche deises Bogens unmittelbar aufliegt” (his Fig. 4 x, 
p. 648). He adds, “Ein Lumen dieses Bogens habe ich im diesem Strange nie 
gesehen,” and he uses this evidence to support his conclusion that the mandibular 
arch is double. I can confirm Hoffmann’s statement as to the presence of this 
“ Zellstrang ” in the anterior portion of the mandibular arch; but there is another 
cord, not mentioned by Hoffmann, which is in every respect similar to this one and 
extends parallel and close to the posterior wall of the arch. I hold Hoffmann’s in- 
terpretation, however, to be incorrect, since, according to my determination, the cells 
of these strands are in large part if not entirely ectodermal in origin, i.e derivatives’ 
of the neural crest. The cells of the Anlage of the Trigeminus may be followed in 
closely connected stages as they migrate ventrad until they enter the mandibular 
arch, where they come to surround the mesothelium as a ring of loose cells between 
the mesothelium and the superficial ectoderm. This evidence confirms the pre- 
vious results of Kastschenko, Platt, and Goronowitsch (°93). While the fate of 
these cells is not clear to me, Miss Platt (’94 and ’97) finds that in Necturus they 
contribute in large part to the formation of the cartilage of the mandibular arch. 
Considering the similarity in the origin of the anterior and posterior cell strands, 
as seen in parasagittal sections through the mandibular arch, it becomes note- 
worthy that Hoffmann (’94) in his preliminary paper failed entirely to reproduce in 
his figures the posterior, while in his later paper (96, Taf. III. Fig. 22), he figures 
two cell strands as histologically quite different from each other. This appears to 
me a notable illustration of the prejudicial influence of a theory. Although I 
Fie. D. Cross section of a Squalus embryo with 11 or 12 somites in a plane cor- 
responding with that of the line a 8 of Fig. C. x 50. The section shows clearly 
the dorsal position of the connecting stalk of van Wijhe's first somite (1) in rela- 
tion to the pre-oral pouch (ar’ent.). 
I, mesoderm of the connecting stalk of van Wijhe’s first somite; ar'ent., arch- 
enteron = pre-oral pouch; cl. ers. n., neural-crest cells; ec’drm., ectoderm; ent., 
entoderm; 2f0,, infundibulum; tb. n., neural tube. 
