DEVELOPMENT OF EAR AND ACCESSORY ORGANS IN FROG. 529 
The two vertical canals are at first entirely distinct, but as 
they lengthen they approach each other and meet. In this 
way a common length is formed, short at first, but gradually 
elongating, and through it both canals open into the utriculus. 
With this stage the adult relations of the canals are at- 
tained, further change being limited to their change in size, 
and the growth of mesoblast between them and the rest of 
the auditory vesicle, whereby they are removed farther and 
farther outwards and attain their characteristic curves. The 
mesoblast concerned consists at first of irregular scattered cells, 
but as the cartilaginous auditory capsule develops the cartil- 
age extends between the canals, and in tadpoles of about 20 
mm. it entirely surrounds them. 
With regard to the relative times of formation of the canals, 
the anterior vertical and horizontal are formed simultaneously, 
but the posterior vertical is only just commenced when the 
other two are nearly cut off, and there are stages in which the 
posterior vertical canal is not fully separated by the blending 
of the folds constituting its septum, whilst the other two have 
been formed some time. It is noteworthy that of the sensory 
epithelia of the ampullz, that of the posterior is the first to be 
specialised, though the corresponding canal is the last to 
develop. ‘This will be more fully described and an explana- 
tion offered in dealing with the differentiation of the sensory 
epithelium of the ear. 
It is usually stated that the two vertical canals in the 
higher Vertebrates are the first to develop, and that they cor- 
respond to the two canals of the Cyclostomata. From this it 
is argued that the Cyclostome ear is more primitive than 
others, a stage representing their condition being passed 
through in the development of higher forms. It will be seen 
from the above that this doctrine is very much shaken by the 
results arrived at in this paper. 
The ampullz, concerning the formation of which nothing 
has as yet been said, are developed by the constricting off of 
parts of the vesicle at the ends of the canals. The sensory 
epithelium of each ampulla is present some time before the 
