MESSES. A. AND E. NEWTON ON THE OSTEOLOGY OE THE SOLITAIEE. 339 
pneumatic foramina of the bone, and in the largest examples it is, with them, altogether 
obliterated. In Pezophaps, however, it has a very distinct existence apart from the 
pneumaticity of the bone, and (as has just been said) it occurs conspicuously in the Paris 
specimen, which is equal in size to the largest Didus , where no trace of any depression, 
much less of a deep hollow, is present*. 
§ 6. Scapular Arch. 
Of this portion of the skeleton we have numerous scapulae and coracoids, but unfor- 
tunately not a fragment which we can identify as belonging to the furcula. None of 
the specimens offer any trace of the remarkable anchylosis, observed not unusually in 
Didus, of these three bones, and from the considerable series in our possession it is 
perhaps not too much to conclude that such anchylosis never took place in Pezophaps. 
The scapula (Plate XIX. figs. 84-86, 97-99, XX. figs. 132, 133), of which thirty-six 
specimens are contained in the collection, differs very greatly in its general form from 
that of Didus. Instead of being sabre-shaped and curved decidedly backward, it is 
rather straight, or, if anything, inclining generally somewhat forward — a character not 
known by us to exist in any other bird, and only pointing backward and inward at its 
extremity when it becomes spatulate in form. Considerable individual variety of shape, 
however, is found in this part in Pezophaps, as is also the case in Didus. As a whole 
the bone is comparatively stouter in Pezophaps, especially its proximal part, and the 
markings of muscular attachments are not so well defined, there being in this respect 
little difference between its inner and outer surface. The latter does not so often 
display towards the distal extremity the slight elongate cavity observable in Didus; 
but is, on the contrary, very frequently somewhat convex, while the inner surface is, 
almost without exception, decidedly concave, though to a variable extent. The process 
sent off from the lower and outer border, which is a marked and (so far as we can judge) 
an invariable feature in Didus , is represented merely by a slight notch formed by the 
abrupt termination of the ridge, which eventually becomes the edge of the thinner and 
more distal extremity of the bone. The articular end of the scapula is relatively thicker 
than in Didus, with the coracoidal surface occupying a proportionately larger space, 
while the humeral surface is not so~ extensive, and the acromial process is proportion- 
ately smaller, much less curved, and not containing in its hollow the pneumatic foramen 
which appears to be always present in Didus. 
* In Diclunculus there is at this part a good-sized hollow, at the bottom of which are some four or five pneu- 
matic foramina. In Goura the median hollow is very deep and elongate, constricted in the middle ; behind it 
lies a series of pneumatic orifices, and on each side a shallow depression. Leucosarcia is simpler, having a single 
deep median hollow and two lateral depressions. The same appearance is presented by Trocaza (a Mauritian 
form), Columba, and the more normal Doves. Geopelia and Pcitagicencis much resemble Goura in this part. 
In Chamcepelia there is little appearance of the deep hollow, and the lateral depressions are nearly obliterated. 
Of the normal Gallince, most forms, as PJiasianus, Perdix , and Coturnix, have no lateral depressions here ; but 
the genera Qaccabis, Franco'linus, Ortyx and LopTiortyx possess them, while Odontophorus does not. A single 
median depression, which is often thickly set with orifices, exists in most of the genera Tetrad, Bonasia, and, 
Ldgopus . 
2 Z 2 
