232 



BULLETIN: MUSEUM OF COMPARATIVE ZOOLOGY. 



meut of the secondary roots is the same as that described for the pri- 

 mary one, viz. as processes from neuroblast cells in the ventral hona. 

 By following the fibres of the roots in the wall of the brain, it is 

 easily ascertained that the motor "nucleus" of the abducens is a 

 very elongated one, as is known to be the case in higher Vertebrates 

 (see Edinger, '96). 



The study of the development of the abducens is simpler than that of 

 the oculomotorius, since the nerve never comes into relation with a 

 ganglion, and thus resembles the primitive ventral nerves of Amphioxus 

 more than do the ventral spinal nerves. The gradual extension of its 









Figure J. 



fibres through the mesenchymatous tissue at the base of the medulla 

 mav therefore be easily followed. It is a matter of some morphological 

 importance, in my opinion, that not all the nerve fibrils extend anteriorly 

 toward the third somite (van Wijhe's), but that in later stages of develop- 

 ment, e. g. in embryos with 78-80 somites (Plate 4, Fig. 20), a nerve 

 fibril is seen to pass from the posterior root of the nerve in a posterior 

 direction toward the myotome of the sixth somite, which has at this stage, 

 however, only a few rudimentary muscle fibres. Miss Piatt ('91) like- 

 wise has mentioned the fact that this nerve also distributes fibres to 

 mesoderm posterior to the third somite (muse, rectus posterior). In the 

 abducens, therefore, we have to do with a post-otic ventral nerve, 



Fig. J. Parasagittal section of a Squalus embryo with 60 somites, showing the 

 abducens as a fibril formed by the processes of at least four neuroblast ceUs. 

 X 4-17. ahd., abducens ; cL n'hl., neuroblast cell. 



