George L. Streeter 147 
The surface markings of the vesicle during this stage assume a signifi- 
cant character. In the first place the vestibular pouch at once takes 
on a triangular shape with the apex toward the appendage. The three 
borders of this triangle form the anlages of the semicircular canals 
(see Fig. d, Plate 1), which bear the same inter-relation as the canals 
in later stages. A second feature which is apparently constant and 
important is the sharp, vertical groove, which cuts in between the anlage 
of the posterior canal and the posterior end of the lateral canal. This 
we may call the lateral groove. It was not represented by His, Jr., 89, 
but can be seen in the model from the 8 mm. rabbit of Krause, go, p. 296, 
and still better in models 3, 4, and 5 of Denis, 02, which were taken 
from the bat. The latter author mentions it in his text. 
Ventral to the anlage of the lateral canal, on the lateral surface of 
the vesicle there is a rather large depression or fossa, which becomes 
more marked in proportion to the increasing projection of the lateral 
canal, which overhangs it like a shelf. This fossa forms the lateral wall 
of the atrium from which the utricle and saccule are to develop. The 
cochlear portion of the vesicle is limited to its ventral tip and extends 
up along the rounded posterior border nearly to the prominent anlage of 
the posterior canal. There intervenes between them that portion of 
the wall that is to become the posterior ampulla. The tip of the 
cochlea begins to bend forward practically as soon as the cochlear pouch 
can be distinguished as such. 
The changes in the structure of the wall of the ear vesicle which ac- 
company the pouch formation are limited to the thinning out of certain 
areas on the dorso-lateral surface of the vestibular pouch, and the lateral 
surface of the appendage as has already been referred to. The remainder 
of the vesicle wall is of the primitive type; there were no areas that could 
be recognized as nerve endings. In embryo No. 163, 9 mm. long, how- 
ever, protoplasmic nerve processes extend from the ganglion and lose 
themselves in the vesicle epithelium. The branch destined to become 
the posterior ampulla nerve could be seen with great distinctness; but 
where it ended there was no reaction to be seen on the part of the 
epithelium. 
The period of semicircular canal formation is shown in Figs. g-k, 
Plate I. The process consists in the expansion of the edges of the ves- 
tibular pouch, 7. e., the canal anlages, and the coincident absorption of 
the intermediate vestibular walls, as was essentially described by Boéttcher 
in his monumental work of 1869, and to some extent by other observers 
even previous to that. Since then further details have been worked out 
