Charles R. Stockard 499 
instead of upward, and to fuse at their margins with the ectoderm, thus 
suggesting the gill slit mode of origin of the mouth so elaborately de- 
fended by Dohrn. This statement is correct except for the very young 
stages in which, as I have shown, the mouth diverticula also bend 
upward, Fig. 1, and suggest, therefore, much more forcibly the gill 
slit origin of this organ. 
The hyomandibular pouch follows close after the mandibular, which 
in all stages is the most anterior, and there are no premandibular diver- 
ticula of the gut, which Dean indicated from his study of surface prepa- 
rations. 
he very young embryo described as having the nose tube leading 
directly into the gut, and shown in Dean’s Figs. 37 and 101, presents 
the hyomandibular pouch in the following condition. It is situated 
well in front of the auditory vesicles and is enormous in extent, as is 
shown in Fig. 15, a section of the same embryo from which Fig. 1, of 
the mandibular arch, was taken. In this section it will be seen that well- 
marked ectodermal pockets are present and a corresponding thickening 
of the endodermal gut wall is in close contact with them. It thus ap- 
pears as if this gill is about to establish an opening to the exterior, but 
such is not the case since from this stage onward there sets in a steady 
retrogressive development. 
Fig. 16 shows a section through an embryo in which the mandibular 
arch has flattened out or become horizontal, no longer curving upward, 
the condition shown in Fig. 2. Fig. 16 is through that part of the 
hyomandibular pouch where it comes in closest approximation to the 
corresponding ectoderm pocket. The gut pouch, therefore, is decreasing 
in its lateral extent, while in cross section the lateral diverticula become 
retort-shaped, large bulb-like chambers forming their extremities. The 
endodermal thickenings in the present stage are not so marked as in Fig. 
15. As the embryo continues to develop the ectodermal plates or 
thickenings become gradually less pronounced until in embryos of about 
15 mm. in length there appears no ectodermal evidence of the hyomandi- 
bular, Fig. 17. The bulb-shaped extremities of the gut diverticula in 
Fig. 17 are more definitely marked off than in the younger stage of 
Fig. 16, here, also, the hyomandibular cleft is behind the auditory vesicles 
instead of being anterior to them as in the younger embryos. In the 
matter of position the relation of the hyomandibular to the auditory 
vesicles is slightly variable in embryos of the same developmental stage, 
but as development progresses the ear-vesicles always attain a more 
anterior position. This fact is probably due to the forward shortening or 
