in the thick peduncle forming the ecto-calcaneal process; in Pigeons and Rterocles, the ento-calcaneal
process is thicker, and the tube terminates on the outer side of the buttress, which is continuous with that
process. In Rterocles and in all the typical Gallina, the inner trochlea is somewhat more elevated than the
external; but in the Cracida, it is more depressed. In the breadth of the inferior extremity, in the
greater equality in size of the trochlese, and in the great depression of the internal, beneath the level
of the external one, the metatarsus of Megapodius approaches nearest to that in the terrestrial Pigeons.
The variation in form and relative size of the surfaces giving attachment to the inter-osseous muscles,
need not be dwelt upon here. The groove transmitting the tendon of the M. adductor annularis is in many
of the Gallina converted into a canal by an osseous floor, as in certain Pigeons, both terrestrial and arboreal.
The principal point, then, in which the metatarsus of the Dodo differs from that in the ordinary Gallma, is
the greater development of the calcaneal buttress, which terminates superiorly in the ento-calcaneal process.
In the American Vulture (Cathartes California/rius), the metatarsus is compressed in the antero-poste-
rior diameter, and deeply excavated in front, beneath the proximal extremity; the section of the shaft, both
above and below, is therefore transversely oblong. The outer segment of the posterior surface is transversely
subconcave; the calcaneal process is depressed, broad, and imperforate, presenting two broad
shallow grooves posteriorly; from its centre extends downwards a moderately developed ridge subsiding
towards the middle of the shaft; the external trochlea is more depressed than in the Dodo.
In the Vulture [Gypsfulvus) and Eagle [Haliaetus albicilla), the shaft of the metatarsus is sub-
triedal, and the broad posterior surface presents a shallow groove in its whole extent. The wide, vertically
concave, external border looks directly outwards in the Eagle, but in the Vulture has an inclination forwards
and inwards; and the anterior surface slopes rapidly backwards towards the internal edge. The surfaces
for the origin of the M. M. adductor indicis and extensor medii are very small, while that for the
Extensor pollicis occupies the upper half of the anterior surface. The various grooves and tubes for the
transmission of the flexor tendons in the Dodo and Pigeons, as well as in the Gallma, are represented by
a single, deep, semilunar notch extending from the subquadrate plate, representing the ento-calcaneal process,
to a prominent tubercle forming the outer and posterior angle of the proximal articular extremity;
this process is not supported by a ridge descending on the posterior surface of the shaft, it is only a little
more extended downwards basally, than at its free flattened extremity.
In the Vulture, the trochlese are more nearly on a level; the external, as in Catha/rtes, being much lower
than in the Dodo and terrestrial Pigeons. In the Eagle, the trochleas are relatively narrower; the inner is
placed on the same plane as the middle one, and its internal and posterior angle is more produced than in
the Dodo and other Pigeons, or than in Catha/rtes. The inter-condyloid tubercle is relatively very small in
the Vulture and Eagle. While the metatarsus in the Dodo is distinguished from that of the ordinary
Gallma, only by a few and comparatively slight peculiarities, in which it approaches the typical Pigeons, it
differs from the corresponding bone in the Vulture and Eagle, in the form of the shaft; in the presence
of the complex ecto-calcaneal process, and of the highly developed calcaneal buttress; and in the greater
elevation of the external trochlea: points of distinction so important as, even overlooking innumerable
minor dissimilarities, to preclude any idea of the affinity of the Dodo to these raptorial forms. The
metatarsus of the Dodo, in the presence of the calcaneal buttress, resembles that of Catha/rtes more than
its homologue in the less aberrant raptorial birds; but this will not outweigh other important differences,
and is such as not to indicate affinity, but simply a general resemblance in mechanical construction.
The posterior metatarsus in the Dodo (Plate XI. Fig. 7—10), is formed by a thick oblong plate,
twisted on itself from behind forwards, and from within outwards; the line of flexure corresponding to its
diagonal. The channel thus formed [ib. Fig. 9) lodges the flexor tendons of the hind toe. The lower
extremity supports the transversely elongated trochlea, which is very slightly concave in front, but behind,
it is grooved for the deep flexor tendon; it is broader internally than externally, and also projects backwards
beyond the plane of the stem to a greater extent internally.
The thin outer margin of the triangular posterior portion of the stem is concave: its lower angle forms
a subquadrate process projecting outwards beyond the trochlea [styloidprocess), which gives attachment to
the annular ligament; its anterior surface is covered by a thin layer of synovial cartilage, the superficial
flexor tendon of the inner toe gliding on it. The upper extremity, which articulates with the metatarsus,
viewed from before, consists of a semicircular plate, forming the anterior wall of the channel; its thick
convex, external border is roughened in front for the attachment of the strong inter-osseous ligament; the
upper part of its subconcave anterior aspect is covered by articular cartilage, and the concavity probably
gave origin to some fibres of the M. add/uctor indicis. The floor of the channel is wider below than above; and
the lower untwisted portion of the stem projects obliquely backwards from the articular plate, which is perpendicular
when in apposition with the metatarsus ; the rounded inner margin of the former expands above
into the triangular surface of the floor of the channel. The tendon of the M. extensor pollicis passes along
the posterior surface, and is bound down immediately above the trochlea by an annular ligament, attached
externally to a roughened portion of the outer edge, and internally to a narrow pit close to the inner
border.
In all Pigeons, the shape of the posterior metatarsus is precisely the same as in the Dodo;—the styloid
process exists in all (Plate XI. Fig. 16, 18, 25, 31, 37, 43). After observing that peculiar character, I
was kindly allowed to test the Colmibine affinity of. th e Dickmculus, by removing its accessory metatarsal
(Plate X. Fig. 10, 10 a), which proved to be a miniature of that in the Dodo. In the arboreal Pigeons, it
is relatively larger than in the Gov/ri/na, and attains its maximum in Lopholamus.
In the Gallma, the posterior metatarsus is relatively shorter, and the twist less distinctly marked than in
the Dodo and Pigeons. In the common Cock, the curved plate is much thicker, and hence the channel
more open, but the under or trochlear portion projects relatively farther back. In the Cracida, it is
thinner, and more distinctly twisted; and from its greater elongation, its articular surface is placed lower
down. In Megapodius, the outer margin of the curved plate is less concave, concealing from behind, the
expanded articular surface; but the essential distinction in all, lies in the absence of the styloid process.
In Cathartes, it is very small, subpyramidal, and not bent on itself; the anterior surface is nearly flat;
its lower extremity is elevated considerably above the internal trochlea, so that the hind toe is,, as in the
typical Gallma, &c., above the plane of the heel.
In the Vulture, it is narrow transversely, and slightly twisted, the anterior surface being broadly concave,
and the lower extremity placed nearly in contact with the inner margin of the metatarsus. In the
Eagle,1 also, the peculiar flexure of the posterior metatarsus almost disappears, the bone being nearly flat, and
consequently it is readily distinguished from the corresponding bone in the Dodo; it is also destitute of
the styloid process; and, as in Catha/rtes and the Vulture, the lower portion is shortened, and projects
backwards only to a very slight extent. We therefore find, that as in the metatarsus, so also in the accessory
metatarsal, the Dodo deviates more from the Raptores than from the Gallma, but is really distinct from both.
i The figure furnished by Mr. Owen, of this bone, in situ, is evidently taken from a badly mounted skeleton, in
which it is placed in an unnatural position, as theligaments would retain the accessory metatarsus close to the main one.