xeal: xeeyous system in squalus acanthias. 173 



actually obliterated all surface evidence of such a condition. Each hind- 

 brain neuromere, therefore, consists of a lateral thickening and a dorsal 

 and ventral dilatation of the wall of the neural tube. The constrictions 

 are exactly opposite on the two sides of the brain. The elongated cells 

 are placed radially to an imaginary point situated in the middle of the 

 thickening of the wall opposite. The nuclei are generally nearer the 

 outer than the inner surface, and approach the latter only in the region 

 of the constriction between the neuromeres. In this region the cells 

 are more crowded, but the cells of one neuromere do not extend into 

 the adjacent neuromeres. 



The hindbrain neuromeres, being structural differentiations of the 

 ■walls of the neural tube, are not to be explained as the result of a 

 simple mechanical process. The essential similarity of these serial 

 groupings of nerve cells to the metameric ganglia of Annelids will, I 

 believe, impress others as well as myself. A reconstruction of the neu- 

 romeres as they appear in this typical condition is shown in Figure 40, 

 Plate 6. 



III. The Neuromeres in the Trunk Region. 



a. Development of Mtelomeres. 



It might seem that a more natural sequence in the study of neuro- 

 meres than the one here followed would be to pass from the simpler 

 conditions which obtain in the trunk to the more complicated ones in 

 the head region. Instead of this, I follow the historical sequence, hav- 

 ing begun with the " Krauselungen," or foldings, first seen by observers 

 in the region of the hindbrain, and now pass to the study of the con- 

 ditions in the spinal cord. That "hindbrain neuromeres" could be 

 compared with segments of the spinal cord was an aftei'thought on 

 the part of embryologists, evidently bom of the conception that the 

 head has a segmentation comparable with that in the trunk. 



"While the neural plate in the trunk region is still widely open, its 

 dorsal surface exhibits cross furrows, which are proved by longitudinal 

 sections to correspond with the interspaces, or clefts, between the meso- 

 dermic somites. The number of the cross furrows exactly equals that oi 

 the interspaces, increasing in number as the constrictions between the 

 somites do. They do not, however, extend to the edges of the neural 

 plate, but are restricted to the region where the plate rests upon the 

 somites. In these cross farrows we have the first indications of those 

 structures which were called by McClure ('89) " myelomeres," and were 



