neal: nervous system in squalus acanthias. 213 



Trijumeau " (Mitrophanow, '93), and " erste periaxiale Strang" (Gorouo 

 witsch, '93). Its cells at this stage (15 or 16 somites) have already 

 migrated half way down the side of the neural tube (Fig, 7). In the 

 region of encephalomere V the disassociatiou of the cells of the neural 

 crest has begun, and the dorsal part of the encephalomere in consequence 

 appears enlarged. A ventral migration of its cells, however, does not 

 take place until a later stage. 



In an embryo of 18 or 19 somites (Plate 3, Fig. 8) two regions of cell 

 proliferation, separated sharply by encephalomere lY, are seen. Mi- 

 trophanow ('93) has stated that at the beginning the facialis is not 

 ■wholly separated from either the trigeminus or the vagus group. I find 

 on the contrary, as already stated, that no neural crest is found in the 

 region of encephalomere IV, and that consequently the " Trigeminus 

 Anlage " is separated by the space of this encephalomere from the pos- 

 terior portion of the neural crest. Apparently as a consequence of cell 

 proliferation and migration, the dorsal wall of encephalomere III is very 

 thin at this stage, while that of encephalomere IV is considerably thicker 

 and its cells are more compactly arranged. The cells of the neural ridge 

 which form the " Tx'igeminus Anlage " now extend ventrally as far as van 

 Wijhe's second somite. The second region of cell migration is at this 

 stage sharply confined to encephalomere V. Behind this a disassocia- 

 tiou of neural-crest cells has begun in the region of encephalomere VI, 

 but no migration has taken place. From an examination of later stages, 

 the cells proliferated from the region of encephalomere V are easily 

 proved to pass ventrally into the hyoid arch, and to form the gangli- 

 onic Anlage of the acustico-facialis. From a study of mitotic cells and 

 from the grouping of cells one is led to believe that the greatest cell 

 proliferation takes place in the posterior part of this neuromere. 



It is to be noticed that the advancing ventral end of the ganglionic 

 Anlage extends toward the cleft between van Wijhe's third and fourth 

 somite. Also that cell processes ft'om each of these somites now extend 

 toward the ganglionic Anlage. 



When the embryo possesses 19 or 20 somites (Plate 3, Fig. 9) the 

 " Trigeminus Anlage " shows a differentiation into an anterior smaller 

 portion, which passes in front of the midbrain vesicle toward the optic 

 evagination, and a posterior larger portion, which extends ventrally into 

 the mandibular arch, just beneath the superficial ectoderm and external 

 to the second somite. I am inclined to believe that this division of the 

 Anlage is partly due to the enlargement of the vesicle of the midbrain, 

 since frontal sections show that the lateral wall of the midbrain lies very 



