1879.] On the Morphology of the Semicirceular Canals. 371 
tion in its own peculiar direction.! The fact that the involution of 
the side organs in Teleosts does not take place until the organism 
is already far advanced towards the completion of its develop- 
ment, and the fact that the mucous canal in the embryo of Elas- 
mobranchs does not appear until the epiblast has been differen- 
tiated into its two layers, both imply that the side organs had for 
ages existed as mere “ nerve-hills,” or nerve epithelium, projecting 
into the water, and this renders our idea not only possible but 
probable. In the course of the evolution of the ear the first 
differentiation of structure that occurred may have been as fol- 
lows: Certain of these areas of nerve-epithelium or sensory 
maculz, may, for functional specialization, have become enclosed 
in canals, (The great functional advantage of this form of involu- 
tion over the closed vesicle, we shall presently see.) Now, for still 
greater specialization, the whoie area containing the canals with 
perhaps one or two maculz not yet enclosed, may have under- 
gone a general process of involution and thus given rise to the 
vestibule and semicircular canals. 
This view of the evolution of the ear is favored by the fact that 
the semicircular canals are constant structures in all vertebrates! 
Again, their number is always the same, namely, three, with but 
two exceptions, which may be easily explained. I refer to the 
Myxines which have but one, and the lampreys which have but 
two. Probably no naturalist who accepts the theory of evolution, 
looks upon the selachians or any other existing order of fishes 
as ancestrally related to the other vertebrates, and even the 
structure of the Marsipobranchs indicates that they belong to an 
early differentiated and highly aberrant type. 
In studying the gradual development of the ear, palæontologi- 
cal evidence, unfortunately, cannot come into play. It is too 
fragmentary. The very first vertebrates that we find are already 
fishes of a high degree of development. ee 
Let us now see to what views of the functional value of the 
Semicircular canals, the above discussion leads. The first ques- 
tion that arises in our minds is, what is gained by the involution — 
of the sensory macula? And secondly, why should this process © 2 
have resulted in the formation of canals? Why did not each : 
macula become enclosed in a separate cavity? Probably the . 
benefit derived by a delicate sensory structure from the formation 
1 In this category it would be no more than just to inclads cick: a as the 
= sensory ampulle of sharks and rays and the Savian vesicles. 
