168 BULLETIN : MUSEUM OF COMPARATIVE ZOOLOGY. 



later stages, whereas it will be shown iu Squalus that this is not the 

 case. 



The want of abundant material of early stages felt by investigators in 

 most cases is not a hindrance in the case of Squalus, for the early stages 

 are as easily obtained as the later ones. In the study of the develop- 

 ment of neuromeres, I have made use, first, of specimens very li^htlv 

 stained in hasmatoxylin and mounted in toto in balsam, and secondly 

 of the usual cross, frontal, and sagittal sections. The series of embryos 

 represented in Plate 3 is chiefly of value in showing the neuromeres in 

 successive stages, and the relations of the masses of cells composing 

 the neural crest, or ganglionic Anlagen (colored blue in the figures), as 

 seen in cleared specimens. The neural tube is represented as seen in 

 optical section, while the other structures of the right half of the embryo 

 are projected upon the median plane. 



The earliest evidence of hindbrain neuromeres which I have found is 

 seen iu embryos of 14 or 15 somites in which the cephalic plate has not 

 closed in the hindbrain region. In most embzyos with that number of 

 somites the plate is already closed, but in cases where it has not, neuro- 

 meres IV, V, and YI are seen as thickenings of the lateral walls of the 

 hindbrain before its closure. Usually closure takes place, as in the 

 chick, first in the region of the so called trigeminus Anlage, and later in 

 the region of neuromere V, the most anterior portion of the cephalic 

 plate remaining open as the neuropore until considerably later stages. 

 Figure 7, Plate 3, shows that iu embryos of 14 to 16 somites (in the 

 specimen figured, after the closure of the cephalic plate) four expansions 

 of the neural tube in the hindbrain region are difl^erentiated (neuromeres 

 III, IV, V, VI). The hinder boimdary of neuromere VI marks the for- 

 mer posterior boundary of the cephalic plate. The figures show (and this 

 is a point of considerable importance in considering the morphological 

 value of neuromeres) that each neuromere corresponds to the region of 

 a dorsal as well as a ventral expansion of the neural tube, and that the 

 neuromeres are separated from one another by both dorsal and ventral 

 constrictions, which are to be seen both in sagittal sections and in 

 cleared specimens. 



Frontal sections at this stage give additional evidence concerning the 

 structure of hindbrain neuromeres. A frontal section just below the 

 axis of the neural tube is shown in Figure 22, Plate 5. The section 

 shows that the cephalic plate is still open in the region of the forebrain. 

 The dorsal portion of the mesoderm in the region of van Wijhe's 2d and 

 3d head somites (2 and 3) is cut on the right side only, the sections not 



