THE SEGMENTATION OF THE HEAD. 147 



Exactly the same conditions occur in the formation of the optic 

 nerve. It arises as a hollow outgrowth from the dorsal 1 part 

 of the wall of the brain, which reaches the ectoderm, if it does 

 not actually fuse with it. From the distal portion of this out- 

 growth is formed the ganglionic layer of the retina, correspond- 

 ing exactly .to the Anlage of the gasserian ganglion, and from 

 this ganglionic layer the observations of Keibel, Froriep, and 

 Assheton have shown that the permanent optic nerve grows 

 back to the brain, the primitive optic stalk becoming aborted. 

 So that, then, there appears no escape from the view that the 

 optic nerve, instead of being a structure sui generis, is clearly 

 homologous with the other admittedly segmental cranial nerves, 

 trigeminal, glossopharyngeal, etc. The same discovery also 

 opens up other possibilities ; besides offering a possible explan- 

 ation of the mooted question of the origin of the vertebrate eye, 

 it renders it necessary to take into account in considering the 

 neural segmentation of the head the little understood pinealis 

 and the secondary epiphysial structures, for, as Locy has shown, 

 there are two paired outgrowths from the brain walls behind, 

 and apparently serially homologous with the optic outgrowths, 

 and these become developed into what we may collectively 

 term the epiphysial structures. It is hence possible that we 

 are to look in this region for some of the long-sought dorsal 

 roots. 



Lastly to be mentioned is the earlier segmentation of the 

 neural structures discovered by Locy. He points out that in 

 the very early stages of the embryos of elasmobranchs, batra- 

 chians, and the chick, before the medullary plate has begun to 

 close, its margins are ornamented with a series of bead-like 

 prominences, and that the series of these are continued back 

 into the trunk-region, and in elasmobranchs they can even be 

 traced into the embryonic rim (affording some puzzling prob- 

 lems to those who deny concrescence in the formation of the 

 vertebrate embryo). In the expanded head-plate of the dog-fish 

 he finds eleven of these beads on either side, and, following the 

 history of these through until other landmarks come into view, 



1 Morphologically dorsal ; the later hypertrophy of the posterior part of the 

 dorsal surface forces it to an apparently ventral position. 



