George L. Streeter 149 
by various investigators, notably by Krause, go, and 03, who approached 
the problem along the whole line of vertebrates. He demonstrated that 
the canals are formed one after the other in definite sequence, the superior 
first then the posterior and lastly the lateral. Our information con- 
cerning the human semicircular canals is based principally on the work 
of His, Jr., 89. 
An interesting interval, which was left open by His, Jr., between his 
stages shown in his Figs. 6 and 7, Plate I, is filled in by my models, 
made from 11 and 13 mm. embryos (Figs. g-k, Plate I). What is to 
be particularly noted is the change occurring in the structure of the 
vestibular wall which can be seen from a surface examination of the 
model. Those areas which are to persist stand out prominently ‘and 
present a fairly definite outline of the future labyrinth, while the inter- 
mediate areas, which are destined to be absorbed, collapse before the 
advancing mesoderm; this is well shown in Figs. 7 and k. It might 
be thought that the absorption of epithelium in Fig. 7 had been com- 
pleted as far as the superior canal is concerned, and that the remaining 
epithelium would go to make the canal wall, necessarily stretching out 
to obtain the diameter represented by the same canal in Fig. m. This, 
however, is not the case; it is only the thickened edge of the pockets of 
the vestibular pouch that becomes canal wall. In Fig. 7 there still 
remains a large area of epithelium that is to be absorbed before the inner 
rim of the superior canal is reached. 
The histogenesis of the semicircular canal is shown in the accompany- 
ing Text Fig. 3, in which A, B, C, and D represent transverse sections 
of the superior canal in four stages of differentiation, taken at corres- 
ponding points and magnified the same number of diameters., The 
striking feature of the process is the persistence in the canal anlage of 
the primitive epithelium of 2-3 layers until after the canal is closed 
off, evidently being a factor in its rapid growth. Section A is taken 
from the ear vesicle of a 9 mm. embryo, the same as shown in Figs. d, 
e, f, Plate I. Aa shows the entire section of which A is a portion, and 
Ab indicates the direction of the section as regards the ear vesicle. Sec- 
tion B is from a 11 mm. embryo. The entire section is represented by 
Ba, whose position as regards the ear vesicle is shown on Bb, which is 
from the same model shown in Figs. h, i, 7, Plate I. Sections through 
the vestibular region at this stage are very interesting, as they show 
by the thickness of the wall which are the persistent areas; section Ba 
is made in such a way as to include the anlages of two canals, the lateral 
wall of the crus commune and a part of the utricle and the ductus 
