336 MESSES. A. AND E. NEWTON ON THE OSTEOLOGY OE THE SOLITAIEE. 
The next succeeding vertebra occasionally developes an imperfect pleurapophysis, but 
the remainder are attached to the ilia only by diapophyses. These are stouter and less 
lamelliform than in Didus, and, owing to the more convex form of the posterior portion 
of the pelvis, the cavities they intercept are shorter. The whole of these vertebrae are 
broader than in Didus , and this comparative widening of the centrum becomes more 
obvious towards the posterior extremity. 
The ilium above (Plates XVI. fig. 65, XVII. fig. 68, XVIII. fig. 70) is not so di- 
stinctly divided into two parts as in Didus, owing to the obsoleteness of the gluteal ridge 
already mentioned. The anterior part is less concave, the posterior more convex, and 
both are narrower than in Didus. The anterior border of the ilium is thickened and 
curved with the convexity forward, but the most convex part of the curve is nearer the 
lower margin than in Didus , where it is situated about the middle. The form of the 
upper margin of the ilium has been previously described. The orifices of the ilioneural 
canals are very much smaller than in Didus, and often more circular. The acetabular 
region is much as in Didus, but the trochanterian surface is narrower, more prominent, 
and directed more forward. Immediately above this surface, where the ilia have diverged 
furthest from each other, they rise, and without showing any angular ridge, bend over 
convexly. Here they slope downward till about as far as the fourteenth sacral vertebra, 
approaching each other slightly, but not so much as in Didus. They then again diverge. 
The inner surface of the ilium is also thickened behind the ischiadic foramen, but in 
Pezophaps to a less degree. The outer border of the posterior part of the ilium pro- 
jects pretty equally in the two forms, and the coalescence of the ilium, ischium, and 
pubis is much the same in each. The pubis, of which in Didus only a small portion has 
been described, is present in one specimen of Pezophaps (Plates XVII. figs. 66, 68, 
XVIII. fig. 70) in what appears to be a nearly perfect condition. It originates below 
rather than behind the acetabulum, and diverges, as soon as it is free, both vertically and 
horizontally from the ischium. It is at first trihedral with the addition of a sharp project- 
ing ridge on its upper and outer margin. This ridge disappears as the bone passes back- 
ward, and broadens into a lamina with the upper edge thickened, and bevelled on the 
outer side. After passing beyond the extremity of the sacrum the pubis bends upward, 
slanting slightly inward, and the bevel turning to the inner side disappears. The ischium 
of Pezophaps greatly resembles that of Didus, so far as can be judged from the imperfect 
examples of the latter which exist. The ischiadic foramen of the former, however, is as 
broad behind as before. The posterior portion of the ischium, at the distance from the 
ischiadic foramen of about its own length, becomes sensibly convex. It then narrows 
and becomes concave, its lower margin sloping downward and outward, as if to pass and 
avoid the pubic style, which is here directed upward and inward in a different plane; but 
the extremities of both pubis and ischium are broken off, and it is not perfectly certain that 
they may not have met eventually, though it is nut easy to suppose that they have done 
so. Quite enough, however, remains to show that in the posterior portions of both these 
bones Pezophaps differs entirely from Didunculus, or indeed any Columbine form, and 
we must go far to seek a similar case. The conjectural restoration of the ischium and 
