VERTEBRATES. 31 



lii), on the corresponfling part of the brain-floor ; tliis structnre, when present, marks 

 off also the floor of the epencephalon (]ions)from that of the nietencephalon (medulla 

 oblonijata), bnt the latter name is frequently employed for the floor of both regions in 

 the absence of the pons. 



The regions of the brain named above are as little ' segmental ' in their nature as 

 the reo-ions of the skull already described. It is possible that all of the brain in front 

 of the notochord (that is the thalam- and pros-encephalon) is merely an outgrowth 

 of the epichordal part, and that consequently any trace of segmentation ought to be 

 looked for in the latter. We have seen how the segmentation of the spinal cord is 

 expressed by the origin of the spinal nerves, but the cranial nerves, although they 

 originate in a similar manner from the neural tube, do not retain the primitive seg- 

 mental arrangement, but acquire new connections with each other and with the brain, 

 which renders the study of their segmental arrangement possible only in the embryo. 

 It is not known how far the olfactory and optic (first and second) nerves, which 



Fig. 35. — 1 - 8, frontal sections of brain of yuiin^ gar-pike; 9, of spinal cord, through the planes indicated in Fig. 31. 



originate from the prechordal brain (Fig. 37)' agree with spinal nerves, but all the 

 other cranial nerves are certainly segmental. This has been determined with most 

 accuracy in shark-embryos, which exhibit more primitive conditions than those of 

 higher vertebrates. In the notochordal part of the shark's he.ad nine mcsoblastic 

 somites are present, the first three of which give rise to the muscles of the eye, the 

 List three to certain muscles connecting the skull and shoulder-girdle, while the middle 

 three do not develop any muscles. The cranial nerves have certain definite relations 

 to these somites, which explains their distribution in the adult, but their relation- 

 ship to the brain is somewhat different fi'om that of the spinal nerves to the 

 s]iinal cord. As with the s]nnal nerves, the ventral root is motor, but it pro- 

 ceeds merely to the muscles derived from the corresponding somite, not to those 

 from the lateral plate below. The dorsal ganglionated root, on the other hand, is 

 mixed, and not only contains sensory fibres, but motor fibres for the muscles derived 

 from the lateral plate. It alone forms dorsal and ventral branches similar to those 

 met with in the spinal region. 



The destiny of the ventral and dorsal roots belonging to the different segments will 

 now be intelligiljle. The ventral roots of the first three, forming res]iectively 

 the third, fourth, and sixth cranial nerves, are distributed to the muscles of the eye, 

 the ventral roots of the last three coalesice to form a nerve (known as the liypoglossns 

 in sharks), which goes to certain muscles of the shoulder girdle, while there are no 



