359 
Fig. 4 represents an embryo 36 days old, just before the neural 
segments disappear. The segments, both caudad and cephalad to the 
cerebellum, are covered by a thin, transparent, unsegmented roof. 
The latter develops gradually by a process that accompanies the 
lateral expansion of the neural tube. As soon as the neural canal, 
or cleft, appears (Fig. 2), the encephalon begins to widen dorsally and 
the neural segments in this manner are carried laterally and ventrally 
till they occupy the position represented in Fig. 4. It will be observed 
that the olfactory pit is applied against segment 1. Segment 2 is 
wedged-shape with its narrow end turned dorsad. Segments 3, 4 and 
5, like those of the medulla, can still be counted in the ventral zones. 
The origin and developmental history of the five anterior seg- 
ments, constituting fore- and mid-brains, differ in no essential point 
from that of the five segments of the medulla. It seems to me strain- 
ing a point to say that the two sets have different morphological 
values. 
Neural Segments of Salmo. Dissected Specimens. 
Our knowledge of neural segments has been almost exclusively 
based upon the study of sections. Locy ’95 diverged from this con- 
ventional way, removing completely the mesoderm in very young shark 
embryos and united studies of dissected specimens with the study of 
spections. It is my experience that segmental grooves can be traced in 
dissected specimens with greater satisfaction than in sections. 
Figs. 5, 6 and 7 represent the external surface of the Trout en- 
cephalon from which the mesoderm has been removed. The five an- 
i af 
Fig. 5. Salmo purpuratus, Encephalon of a dissected embryo with 10 somites, 
19 days old. Right surface view. inf, infundibulum. Other letters and figures the 
same as in Fig. 1. >< 60 diameters. 
