neal: nekvous system in squalus acanthias. 171 



show the characteristics stated above in the most marked way. Only 

 a faint external constriction, without internal constriction or ridge, sep- 

 arates neuromere VII from the region of the spinal cord. 



At a stage with fifty somites (Balfour's stage K) the structure of the 

 neuromeres is slightly but not materially changed. In Figure 1 7, Plate 

 3, is represented an embryo of this stage, viewed as a ti-ansparent ob- 

 ject. Figures 26-29, Plate 5, show four frontal sections of such an em- 

 bryo, Figure 26 being the most dorsal, and Figure 29 the most ventral 

 of the series. Figure 26 shows that the most dorsal portion of the 

 Deckplatte has become very thin, being only one layer of cells thick. 

 The constrictions and dilatations are only faintly shown, the nuclear 

 arrangement being the same in the region of the constriction as in the 

 region of dilatation. Figure 27, more ventral than Figure 26, though 

 still in the region of the Deckplatte, shows the conditions, both nu- 

 clear and cellular, to be almost precisely the same as in Figure 36, 

 Plate 6. The internal ridges, or cusps, are sharp, and the cells in the 

 region between the internal ridge and external constriction are closely 

 crowded together. It is to be noted that the separation of the lateral 

 walls of the hindbrain is least marked in the region of neuromei'e VI, 

 opposite which the ear capsule lies (compare Fig. 17, Plate 3). Figure 

 28 seems to show that the neural walls have become considerably 

 thickened in the region of the lateral zones. There is no doubt that 

 the lateral zones are absolutely and relatively thicker than at the stage 

 last described, while the neuromeres have increased in length. It is to be 

 observed that this thickening is accompanied by a change in the outline 

 of the lumen of the tube, vertical grooves appearing in the place of the 

 vertical ridges of the more dorsal sections. In the most ventral of the 

 sections. Figure 29, the internal ridges appear again, though the con- 

 cavity of the inner surface of each neuromere in the antero-posterior 

 direction is only faintly indicated. 



During stage K, as the result of the great expansion and thinning of 

 the Deckplatte in the region of the medulla oblongata, the neuromeres 

 come to affect only the lateral zones. Locy ('95, pp. 524 and 525) 

 notes changes in the appearance of the " neural segments " at this stage, 

 the explanation of which he does not state with precision. His opinion 

 seems to be, however, that a union of part of each of the original seg- 

 ments with the segment lying just in front of it, accounts for this con- 

 dition. An examination of the series of Figures 7 to 21 of my Plates 

 3 and 4, and of the frontal sections of Plate 5, shows that no such 

 fusion of neuromeres takes place. The constrictions and ridges between 



