332 MESSES. A. AND E. NEWTON ON THE OSTEOLOGY OP THE SOLITAIEE. 
their Museum, when informed by their Curator, Mr. Flower, of our wish to compare its 
structure with that of Didus and Pezophaps. Finally, our thanks are in a special manner 
due to Mr. John Willis Clark, the Superintendent of our University’s Museums of 
Zoology and Comparative Anatomy, for the trouble he has taken in mounting for us 
some of the bones both of the Dodo and the Solitaire with the success commonly 
attained by him in reconstructing the skeletons of animals. 
§ 2. Vertebrae. 
The hundred and sixty-one vertebrae contained in the collection, few of which pre- 
serve all their processes in a perfect state, form a subject of no small perplexity, and it is 
beyond our power to determine precisely the number which the skeleton contained. 
On the presumption that Pezophaps possessed the same number of dorsal (8) and cer- 
vical (12) vertebrae as is assigned to Pidus by Professor Owen’s draughtsman *, there 
is such a great numerical preponderance of specimens of the eighth, ninth, and tenth 
cervicals (Plate XV. tigs. 24-34), counting backwards, that one is at first led to suppose 
that Pezophaps must have been endowed with more, till a diligent study of each example 
compels the belief that no other presumption will meet the case equally well, and an 
explanation of the curious fact that the specimens of each of these three vertebrae are 
nearly twice as numerous as most of those of the rest must be sought elsewhere. And 
in support of this view it may be stated that of the three anchylosed dorsal vertebrae 
there are present in the collection but five examples. 
The penultimate free dorsal (Plate XV. figs. 56-59) intervening between that which 
coalesces with the pelvis and the three f anchylosed vertebrae just mentioned, is repre- 
sented in the collection by nine specimens, none of which is entirely perfect. The 
hypapophysis is aborted ; the centrum posteriorly resembles that of the next anterior 
vertebra; the neural canal is elongated vertically and contracted in the midst; the 
neural spine is directed forwards, but its outline is not to be completely traced in any of 
the examples we possess ; near the anterior articular surface on either side is an articular 
somewhat reniform costal cavity ; the postzygapophyses have comparatively small arti- 
cular surfaces, oval, and slightly concave. 
The three anchylosed dorsal vertebrae (Plate XV. figs. 51-55), the last of which is the 
antepenultimate dorsal, bear a great general resemblance to the same bones in Pidus. 
Taking a specimen of each species of nearly the same absolute size, the greater extent 
of the articular surfaces, and especially the lateral expansion of the anterior centrum, in 
Pidus is plainly perceptible. But though these relative proportions are preserved, some 
* Professor Owex himself does not offer any opinion as to the number of vertebrae in Didus. We prefer 
citing his draughtsman as the authority for the number assigned to the bird in the large plate illustrating his 
paper, because on the same plate Didunculus appears to be represented as possessing fourteen cervical and seven 
dorsal vertebras, being altogether two more than we are able to count in the very specimen, now in the Museum 
of the Eoyal College of Surgeons, which served as the subject for his pencil. 
t Three is the most usual number in the Columbce., but in Goura two only are anchylosed. In one example 
(Plate XYI. fig. 60) of Pezophaps the anchylosis is incomplete. 
