422 CHARLES HILL, 



the most anterior segments have not progressed far enough to make 

 them different in any marked way from those of the medulla. Seg- 

 ment 1 is elliptical, and represents the anterior end of the neural 

 axis. Segment 2 is wedge-shaped and has its broad end extended 

 ventrally to form the iufundil)ulum (inf). The dividing lines between 

 .'i, 4 and 5, could be detected in this section in the ventral region 

 and the series of sections confirm the description given of these seg- 

 ments in surface studies. The 5 segments of the medulla (7 — 11) 

 and the cerebellum (segment 6), are well defined. Just ventrad to 

 the latter is the deep dorsal groove (f) observed in both the divided 

 and the dissected specimens. 



While the segments of the primary fore-brain and the mid-brain 

 are detected with difficulty in sagittal sections, they are more clearly 

 defined in horizontal sections. Fig. 43 represents a section nearly 

 horizontal through this region , the section passing at an angle such 

 as to intersect the right eye and a point just dorsal to the left eye. 

 This section is made from a trout embryo 26 days old and should 

 be compared with the divided embryo of the same age (Fig. 13). In 

 this as in the sagittal section the evidence of the jointed character 

 of the fore- and mid-brain is very clear. Since the section passes 

 through the dorsal region segment 2 is very. narrow and segment 3 

 very broad. A section through the ventral region would just reverse 

 this order, as these segments are reciprocally wedged-shaped. It is 

 observed that a secondary depression is present in segment 5 and 

 also a median depression in segment 3. These depressions, if sections 

 alone were studied, would lead to error in counting the number of 

 encephalic segments, a mistake that is checked by the study of a 

 series of divided embryos. 



Fig. 44 represents a horizontal section through the anterior por- 

 tion of a chick embryo with 4 somites and with open neural groove. 

 The encephalic mesoblast is very loose and presents no satisfactory 

 evidence of segmentation. There is an irregular grouping of its cells 

 which makes loose clusters of cells that agree in number with the 

 segments. The neural groove, however, is distinctly divided into seg- 

 ments by a series of external and internal constrictions. These seg- 

 ments, on the right and left sides, are directly opposite each other 

 and furthermore each external constriction has a corresponding internal 

 depression that lies in the same transverse plane. No "internal crests" 

 or "ridges" are present as described in older stages of the chick embryo. 



As soon as the neural groove closes, internal transverse crests 



