DEVELOPMENT OF THE NERVOUS SYSTEM 533 



much greater regularity than in the anterior portions. In this 

 latter situation the variation in the observations upon different 

 species and upon individuals of the same species throws some 

 doubt in my mind upon the segmental value of the structures 

 described. It seems possible or even probable that these struc- 

 tures are not segmental but due to growth processes entirely 

 distinct from segmentation. It has been suggested that where 

 there is a rapid growth and migration there is also, usually, an 

 irregularity of the surface. Thus the ridges observed by Smith 

 ('12) lying external and parallel to the medullary folds are ex- 

 plained by him as due to the rapid shifting of material in this 

 region, concomitant with the closing of the neural tube. In the 

 medullary plate there is a rapid lengthening, consequently a 

 rapid migration of material, this migration taking place more 

 longitudinally than laterally. In this way the oblique and 

 transverse ridges can well be explained. Further than this, I 

 believe that phylogenetically we can consider this as more prob- 

 able than that they are segmental structures. The view generally 

 held that segmentation should appear first upon the open medul- 

 lary plate and thus precede mesodermal segmentation, appears 

 to me as incorrect. Rather a segmentation of the neural plate or 

 tube would follow mesodermal segmentation and be due to the 

 development of the centers necessary to control the actions of the 

 various body segments. If this be true, we would not expect to 

 find neuromeres in the cephalic portion of the medullary plate. 



For the closed neural tube there has been commonly described 

 three neuromeres in the primitive fore-brain, two in the mid- 

 brain and five to seven in the hind-brain. To the hind-brain 

 neuromeres segmental value has been accorded by most workers. 

 This should be supplemented by demonstrating the continuity 

 of these structures (or lack of it) with the neuromeres of the medul- 

 lary plate stage. In the fore- and mid-brain a second complica- 

 tion is met, namely, whether the smaller divisions (neuromeres) 

 appearing in these segments are primary, or the primitive fore- 

 and mid-brain segments primary, and the 'neuromeres' second- 

 ary. Upon this point there is a division of opinion. Hill and • 

 Locy, for example, believing that the so-called 'neuromeres' ante- 



