ON THE VERTEBRATE SKELETON. 261 
_I am inclined to regard the ‘ odontoid process’ of the mammalian axis as the 
homologue of one of these subvertebral wedge-bones, or peripheral develop- 
ments of the outer layer of the notochordal capsule. It cannot be the pecu- 
liarly developed anterior articular epiphysis of the second vertebra, since this 
is represented by a distinct centre of ossification between the odontoid process 
and the body of that vertebra, according to Professor Miiller’s observation 
in a foetal foal *. 
The diverging appendages of the hemal arch in the abdominal vertebre of 
fishes present the form of long and slender spines (fig. 17, a a), usually at- 
tached to, or near the head of the ribs, and extending upwards, outwards 
and backwards, between the dorsal and lateral portions of the muscular 
segments, to which they afford a firmer fulerum or basis of attachment ; 
acting, therefore, asso many pairs of rudimental and concealed limbs. They 
are termed the ‘obere rippe’ by Meckel, and at the fore-part of the abdomen 
of the polypterus they are stronger than the pleurapophyses themselves. 
As the vertebre approach the tail these appendages are often transferred 
gradually, from the pleurapophysis to the parapophysis, or even to the cen- 
trum and neural arch. 
_ In the air-breathing vertebrata, in which the heart and breathing organs 
are transferred backwards to the trunk, the corresponding osseous segments 
of the skeleton are in most instances developed to their typical complete- 
hess, in order to encompass and protect those organs. The thoracic hemapo- 
physes in the crocodiles are partially ossified, and in birds (fig. 15, h, h) com- 
pletely so; in which class the hzemal spines of the thorax (As) coalesce together, 
become much expanded laterally, and usually develope a median crest down- 
wards to increase the surface of attachment for the great muscles of flight. 
This speciality is indicated by the name ‘sternum’ applied to the confluent 
elements in question. The abdominal hemapophyses and spines retain their 
primitive aponeurotic condition, though still preserving their characteristic 
expansion+. In the crocodiles and enaliosaurs the abdominal hemapophyses 
are also ossified; and, in the latter, they manifest the same composite character 
which has been noticed in the pleurapophyses of the sturgeon, consisting of 
three or more pieces, which overlap each other}. The abdominal hemal 
spines, in the Plesiosaurus Hawhinsii, are transversely extended, they are 
marked ¢ c in the figure quoted below: the compound hzemapophyses them- 
selves are marked 6 6 in the same figure. 
The typical thoracic vertebrze of most birds support diverging appendages 
(fig. 15, a, a), either anchylosed or articulated, as in the penguin and apte- 
ryx, to the posterior border of the pleurapophysis (pl). The function of the 
appendages in this form of typical vertebra is to connect one hemal arch 
with the next in succession, so as to associate the two in action, and to give 
firmness and strength to the whole thoracic cage. (A portion of the next 
rib so overlapped is shown at pl, a, fig. 15.) 
With regard to the connections of the pleurapophyses, we have seen that, 
in fishes, they may be directly attached to the centrum, or to the ends of the 
parapophyses (fig. 17,p),or they may be quite detached from their proper seg- 
ment, and suspended to the heemal arch of an antecedent vertebra, as in the 
case of the clavicle or epicoracoid, no. 2s. In batrachians, ophidians, and 
lacertians, the proximal end of the pleurapophysis is simple, as in fishes, 
but is articulated to an exogenous tubercle or transverse process from the 
is Vergleichende Anatomie der Myxinoiden. Abhand. Akad. der Wissensch. Berlin, 
1834, p. 105. 
T Myology of Apteryx, Zoological Transactions, vol. iii. pt. iv. pl. 35, g*, g*. 
} Buckland, Bridgewater Treatise, vol. ii. pl. 18, fig. 3. 
