ON THE VERTEBRATE SKELETON. 311 
Cuvier seeks to obscure by the normal absence of its proper transverse pro- 
cesses in man, and the assumed transference of them to another part of the 
skull. 
_ Cuvier in the next place objects to the comparison of the supraoccipital 
with the neural spine of a trunk-vertebra, ‘“ because of its vast difference of 
structure and function.” He does not specify the nature of the difference : 
he admits that the neural spines have distinct centres of ossification in certain 
animals; and all will allow that, in most of the trunk-vertebre of such, the 
neural canal is closed by the coadapted ends of the neurapophyses to which 
the spine articulates or becomes anchylosed : that therefore such spine does not 
directly cover the neural axis, but, retaining the shape signified by its name, 
performs exclusively the function in relation to muscular attachments. At 
first view the contrast seems conclusive against all homology between such 
‘mere intermuscular spine and the broad thin convex plate applied over the 
cerebellum and posterior cerebral lobes in man. And it must be confessed 
that the determination of their general homological relations could not have 
been satisfactorily demonstrated by the mere relations of the parts to the 
laminz supporting them, in so limited a range of comparison. - But, if we 
descend to the fish, we shall find the supraoccipital equally excluded from 
the neural canal by the meeting of the exoccipitals beneath its base; we 
shall, also, see it still retaining the spinous figure, indicating its function in 
relation to muscular attachments to predominate over that in subserviency 
to the protection of the epencephalon. If we next ascend to the crocodile, 
we shall find the neural spine of the atlas to be one of those examples alluded 
to by Cuvier, where the ossification proceeds from an independent centre: 
and it not only thus manifests its essential character as an autogenous ver- 
tebral element, but maintains its permanent separation from the neurapo- 
physes: and it further indicates the modifications of form to which the cor-— 
responding elements will be subject in the more expanded neural arches of 
the antecedent cranial segments by having already exchanged its compressed 
spinous for a depressed lamellar form. Here indeed Cuvier might not only 
have objected to recognise it as a vertebral spine by reason of its change of 
form and function, but also by its continuing a distinct bone, which is 
not the case with the expanded ‘spine’ of the mammalian occipital vertebra. 
But returning to the crocodile, we observe in the segment anterior to the atlas 
that both the form and connections of the supraoccipital (fig. 22, 3) are 
so closely similar to those of the neural spine of the atlas that the recog- 
nition of their serial homology is unavoidable; and we have a repetition 
of the same characters of the vertebral element in question in the small and 
undivided parietal (ib.7). Now Cuvier makes no difficulty in admitting the 
‘occipital supérieur’ in the crocodile to be the homologous bone with its 
more expanded namesake in the bird; or this with the still more expanded 
‘partie grande et mince de l’occipital’ in mammals and man: he is also 
disposed to admit the special homology of the supraoccipital under all 
its variations of form and function in the above-cited air-breathing animals 
with the bone 3 in fishes, which he sometimes calls ‘ occipital supérieur,’ 
sometimes ‘interpariétal.’ If then the special homology be admitted on the 
‘ground of the constancy of the connections of the part, with what show of 
reason can its general homology be rejected which forms.the very basis or 
condition of the characters determinative of such admitted special homology ? 
But Cuvier is not consistent with himself in his grounds of objection to the 
essential nature of the human supraoccipital as the neural spine of its seg- 
ment ; for he does not hesitate to call the atlas of the crocodile a vertebra, 
