MESSES. A. AND E. NEWTON ON THE OSTEOLOGY OE THE SOLITAIRE. 335 
is generally vertically ovate and sometimes constricted midway. The buttresses of bone 
supporting the prsezygapophyses are more slender than in Didus, and the prsezygapo- 
physes themselves are smaller ; with their facets inclined more forward and less verti- 
cally sloped. The fore part of the neural spine, which is nearly as strong as in Didus , 
has a similar rough syndesmotic surface and a curved anterior outline, but it is not con- 
stantly confluent with the ilia, the thickened margins* of which (Plates XVI. fig. 65, 
XVII. fig. 68, XVIII. fig. 70), rising behind it and meeting together, appose, more or 
less, their curved inner (and under) edges to one another till nearly over the anterior 
renal cavity, when, instead of sweeping backward and forming a prominent gluteal ridge 
as in Didus, they suddenly turn downwards, then, in some specimens, as sharply forwards, 
and finally trend backwards towards the part over the acetabulum, soon becoming obso- 
lete. The coalesced neural spines of the remaining sacral vertebrae, after dipping down 
to allow of the meeting together of the margins of the ilia as just described, rise again 
and form a continuous, elevated and slightly curved ridge (Plates XVI. fig. 65, XVII. 
figs. 68, 69, XVIII. fig. 70), which, sloping slightly downwards towards its termina- 
tion, extends to the posterior extremity of the sacral region, where it presents a small 
lozenge-shaped,, partly articular and partly syndesmotic surface to the anterior caudal 
vertebra, as in Didus, receding around the neural canal, here reduced to a very small 
foramen, but with the centrum projecting backwards. 
The abutments of the first sacral vertebra (Plate XVII. fig. 66), which seems to be 
somewhat more elongated, are generally almost exactly as in Didus. One specimen, 
however, shows a remarkable deviation from the ordinary form, in having the pleur- 
apophysis, usually a slender straight filament, very greatly thickened, and curving down- 
wards outwardly to meet the ilia, with which it is confluent. 
In the next three vertebrae there is considerable individual variation : sometimes all 
have their pleura pophyses thickened, sometimes the first two only, and in one specimen 
(that which has just been mentioned as having the first sacral abnormal) the last two. 
The majority of specimens show the thickening of the first two. On the whole, how- 
ever, the general aspect of this part is very like what it is in Didus. 
The fourth and fifth sacrals have their pleurapophyses occasionally developed, but 
though in one specimen those of the fourth are very much thickened, in none are they 
sufficiently produced to coalesce with the ilia. In another specimen the diapophysis of 
the fifth is greatly thickened on the right side so as to present the appearance of a 
thickened pleurapophysis, while on the left it is of the ordinary character. 
The next succeeding sacral vertebrae closely resemble those of Didus, and are attached 
to the ilia by their diapophyses only. 
In the seventh, eighth or ninth sacrals, ho wever, and most generally in the eighth, the 
pleurapophysis reappears, just as it does in Didus; extending backwards and rapidly 
widening, it abuts against the underside of the ilium, immediately above the acetabulum, 
but not, as in Didus, behind it. 
* These margins show no disposition to coalesce. In one example (Plate XYI. fig. 65), apparently of a very 
old hird, they are developed to excess, and then they turn down outwardly on themselves. 
