"My present task is to demonstrate
that there is no part of the bony framework of fishes that cannot find
its analogue in the other vertebrated animals."[89] It seems at first
sight that many bones are peculiar to fish, formed expressly for
performing the functions which fish do not share with higher animals.
These are the bones connected with respiration--the operculum, the
branchiostegal rays, the branchial arches, and others. That the peculiar
bones should be connected with the respiratory functions is only
natural, for the contrast between fish and higher Vertebrates is
essentially a contrast between water-breathing and air-breathing
animals. Considering first the general form of the skeleton in fish, we
are met at once with a difficulty; there is no obvious homologue in
fishes of the neck, the trunk, and the abdomen of higher animals. What
apparently corresponds to the trunk is in fishes crowded close up under
the head. But, after all, it is not of the essence of the vertebrate
type to have the trunk and the abdomen attached at definite and
invariable distances along the vertebral column--that is a notion
surviving from the anatomy which made man its type. The "trunk" differs
in position according to the class, in quadrupeds, birds, and fishes (p.
9). Now, says Geoffroy, allow me this one hypothesis, that the trunk
with its organs can, as it were, move bodily along the vertebral column,
so as to be found in one class near the front end of the vertebral
column, in another about the middle, and in a third near the end, then I
can show you in detail that the constituent parts of this trunk are
found in all classes to be invariably in the same positions relatively
to one another (p. 10). It is important to note this hypothesis of a
"metastasis" which Geoffroy makes, for it is the key to the
understanding of many of the far-fetched homologies which he tries to
establish. It is, of course, clear that this hypothesis is in formal
contradiction with his principal hypothesis of the invariability of
connections, and that he, so to speak, gets a hold on his fish to apply
his principle of connections only by admitting at the very outset an
exception to his primary principle. A further application of the
hypothesis of metastasis will be noticed below in connection with the
determination of the sternum of fishes. We note here an interpretation
of the first metastasis in terms of functional adaptation. "The constant
and violent a