MESSES. A. AND E. NEWTON ON THE OSTEOLOGY OE THE SOLITAIRE. 333 
specimens of these bones in Pezophaps are in every way very much bigger, and longi- 
tudinally bear to the corresponding part of their ally the proportion of ten to nine. In 
these larger (male ?) specimens, the hypapophyses of the anterior and middle vertebrae 
of the anchylosed three are not generally confluent so as to inclose a full elliptical 
space, and in none of the specimens, whether large or small, is there any trace of such 
a second foramen as seems to he invariably the case in Didus. 
The specimens of anterior dorsal vertebrae (Plate XV. figs. 42— 50) are in a very muti- 
lated condition, and it is almost impossible to institute such a comparison between their 
remains and the corresponding bones in Didus as would serve any useful purpose. 
Further onward (Plate XY. figs. 24-41) a sensible though gradual elongation and 
attenuation of the vertebrae is manifested, which becomes perhaps the greatest about the 
region of the fifth, sixth, and seventh cervicals (Plate XY. figs. 13—23) on the presump- 
tion here followed. The postzygapophyses are considerably elongated as compared with 
Didus in proportion to the breadth of the vertebrae. The neural spine of the anterior 
cervicals in that bird is much more prominent but less expanded than in Pezophaps 
(Plate XY. figs. 6-12). From the gradual elongation just mentioned the neck of this 
last, when living, must have been absolutely longer than that of Didus , a conclusion 
which testifies in another point to the fidelity of Leguat’s figure, of which more will be 
said in our concluding observations. 
Of the axis the collection contains six specimens, to one of which the atlas is united, 
possibly by partial anchylosis (Plate XY. figs. 1-5). The former closely resembles the 
axis of Didus as figured by Professor Owen {op. cit. pi. xvii. figs. 12, 13) ; but the odon- 
toid process projects much less, and indeed hardly at all beyond the anterior upper por- 
tion, while on the other hand the neural spine and (as in the case already mentioned of 
the succeeding vertebrae) postzygapophyses project considerably further backward. The 
same is to be observed in the postzygapophyses of the atlas as compared with Didun- 
culus ; and in this respect Pezophaps resembles Goura. The articulating cavity for the 
reception of the occipital condyles is deep, and its lateral margins approach each other 
inferiorly somewhat more closely than appears to he the case in either Goura or Didun- 
culus. 
§ 3. Bibs. 
Of ribs, the collection contains (besides a vast number of fragments) nearly seventy 
examples, scarcely any of which are perfect, and a large portion consists of merely the 
upper extremities of the dorsal ribs. Of these there are about fifty-three specimens, 
and of the sternal ribs some fourteen. Yery few of the former have their epipleural 
appendage preserved (Plate XVI. fig. 63), and as this seems to be the most (if not the 
only very) characteristic feature possessed by the ribs in birds, not much of importance 
can be determined in its absence. It would therefore be unnecessary to dwell at any 
great length on these bones, which present no very apparent distinctive characters from 
those of Didus. Perhaps on the whole the ribs of Pezophaps have their heads and 
articular tubercles proportionally somewhat thicker. This is certainly the case in the 
