aoe FRANCIS VILLY. 
of these are situated in the ampulle, three in the cochlea, and 
of the remaining two one is on the floor of the anterior part of 
the utriculus, and one on the inner and lower part of the 
sacculus. 
The first division of the columnar epithelium takes place at 
the hinder end of the vesicle in tadpoles of about 11 mm. It 
occurs on the formation of the oblique septum, which I have 
described as marking the first trace of division of the vesicle 
into utriculus and sacculus. The small sensory patch thus 
separated lies in the hinder end of the utriculus, that is, where 
the posterior ampulla is formed later, so that the ampulla as it 
develops includes this epithelial patch. 
The fact that this epithelial tract is the first to develop might 
be taken to show that though the posterior vertical canal 
actually forms later than the others, nevertheless it is indicated 
earlier than they are. This early specialisation may, however, 
possibly be explained as a convenience due to the out-of-the- 
way position occupied by the organ concerned and to the early 
formation of the septum between the utriculus and sacculus, 
which in its growth plays an important part in separating this 
patch from the rest. Besides, I do not think that the time of 
specialisation of the sensory tracts of the ampulle is a safe 
guide to the relative times of evolution of the canals, as there 
is no rule of development common to all three. If the ampulle 
were exactly homologous structures, and their development had 
not been interfered with, they would form in the same way, but 
as they do not do so it is safer to assume that the original course 
of development has been modified than that the ampulle are 
not homologous. This would seem to show that the order 
of the evolution of these epithelial patches has not been 
preserved. 
It is interesting to note that just as the anterior vertical and 
horizontal canals form close together, and are even united from 
the first, so the epithelium of their ampulle develops from a 
common rudiment, which becomes separated from the main mass 
soon after the posterior tract already mentioned. This rudiment, 
common to the two ampulle, is situated on the outer wall at 
