pper Silurian rocks.

Cuttle-fishes (_Cephalopoda_) are animals belonging to the molluscous
primary division of the animal kingdom, which division contains animals
formed upon a type of structure utterly remote from that on which the
animals of the higher division provided with a spinal column are
constructed. And indeed no transitional form (tending even to bridge over
the chasm between these two groups) has ever yet been discovered, either
living or in a fossilized condition.[58]

Nevertheless, in the two-gilled Cephalopods (_Dibranchiata_) we find the
brain supported and protected by a cartilaginous cranium. In the base of
this cranium are two cartilaginous chambers. In each chamber is a
membranous sac containing an otolith, and the auditory nerves pass from the
cerebral ganglia into the cartilaginous chambers to reach the auditory
sacs. Moreover, it has been suggested by Professor Owen that {75}
sinuosities between processes projecting from the inner wall of each
chamber "seem to be the first rudiments of those which, in the higher
classes (_i.e._ in animals with a spinal column), are extended in the form
of canals and spiral chambers, within the substance of the dense nidus of
the labyrinth."[59]

[Illustration: CUTTLE-FISH.
A. Ventral aspect. B. Dorsal aspect.]

Here, then, we have a wonderful coincidence indeed; two highly complex
auditory organs, marvellously similar in structure, but which must
nevertheless have been developed in entire and complete independence one of
the other! It would be difficult to calculate the odds against the
independent occurrence and conservation of two such complex series of
merely accidental and minute haphazard variations. And it can never be {76}
maintained that the sense of hearing could not be efficiently subserved
otherwise than by such sacs, in cranial cartilaginous capsules so situated
in relation to the brain, &c.

Our wonder, moreover, may be increased when we recollect that the
two-gilled cephalopods have not yet been found below the lias, where they
at once abound; whereas the four-gilled cephalopods are Silurian forms.
Moreover, the absence is in this case significant in spite of the
imperfection of the geological record, because when we consider how many
individuals of various kinds of four-gilled cephalopods have been found, it
is fair to infer that at the least a certain small percentage of dibranchs
would also have left traces of their presence had th