276 BULLETIN: MUSEUM OF COMPARATIVE ZOOLOGY. 
stages of development of S. acanthias. In the early stages of this 
animal, the neuromeres are local thickenings of the lateral zones, as well 
as dilatations of all of the zones of the medulla. As paired glangionic 
enlargements of the central nervous system, they obviously resemble, 
except in position, the ventral chain of ganglia of Annelids, Therefore 
they cannot be explained as the passive result of mechanical shoving or 
bending. The constrictions between the neuromeres, as well as the 
crowding of nuclei in the regions of constriction, may however be, 
and most probably are, intensified by shoving or bending of the neural 
tube. 
No structural conditions are presented by the myelomeres which are 
not reconcilable with the hypothesis that their existence is dependent 
upon the presence of the mesodermal somites. If they ever possessed 
a dorsal segmentation like that of the “hindbrain neuromeres,” — and 
there is no evidence to show that they ever did, —it has been lost. 
But, though they appear of doubtful morphological value, their numerical 
correspondence with nerves and somites attests their metameric value. 
The so called neuromeres of the forebrain and midbrain (encepha- 
lomeres of Zimmermann) are not morphologically comparable with “hind- 
brain neuromeres,” since they are simply dorsal or ventral expansions 
which are secondary in the time of their appearance. I hold that there 
are much better reasons — viz. on the grounds of time of appearance, 
of structure, and of relation to nerves and somites — for regarding each 
of the primary forebrain and midbrain vesicles (neuromeres I and II) as 
serially homologous with hindbrain neuromeres (neuromeres III to VII), 
than for so regarding their later subdivisions. The latter are comoge- 
netic vesiculations of the neural tube, and not of metameric value. 
3oth dorsal ganglia and ventral nerves in the trunk develop in the 
regions of constriction between myelomeres. A comparison with the 
conditions in Amphioxus and Petromyzon shows that this condition is 
not to be regarded as primitive, but that previously dorsal and ventral 
nerves alternated, the former being intersomitic in position. Such 
topographical relation is retained by some cranial nerves, viz. V, VII, 
IX, and X (Urvagus). 
The ganglionic Anlagen of four cranial nerves, viz. V, VII, 1X, and X, 
are proliferated from four encephalomeres, viz. III, V, VI, and VII. 
Chiefly for this reason, but also because of the clear connection of two 
splanchnic motor roots, viz. V and VII, with two of the encephalomeres, 
I conclude that the primitive metameric relations of the latter were with 
the visceral arches. The local thickenings of the hindbrain neuromeres 
