340 MESSES. A. AND E. NEWTON ON THE OSTEOLOGY OF THE SOLITAIEE. 
Of the coracoid (Plates XIX. figs. 76-83, XX. fig. 136) the collection contains 
twenty-seven more or less perfect specimens, besides the anterior extremity of eleven 
others, thirty-eight in all. This most significant bone differs very greatly from the cor- 
responding part in Pidus — a fact not perhaps so surprising when the exceedingly abnor- 
mal form which it assumes there is taken into consideration. It is in Pezophaps pro- 
portionately much more bulky, and generally smoother. At its sternal end it differs 
from Pidus, and indeed from all Columbine birds, in the extension and rounding off of 
the outer border so as entirely to mask the more or less elongate process which in some 
of the Columbidce (Golumba livia for instance) is turned distinctly forwards. Corre- 
sponding with this the inner border is produced into a sharper angle, though blunt at 
the tip, than it is in Pidus , and much more so than in most other Pigeons* * * § . Making 
allowance for the extreme attenuation of the coracoid in Pidus , which is its especial 
characteristic, the surfaces of the coracoid in Pezophaps present, on closer study, much 
the same aspect. The upper surface at the sternal end (Plate XIX. figs. 78, 82) is 
more convex, and the lower one is less flat, but generally perforated with similar pneu- 
matic foramina f, though these are variable in shape and extent, and sometimes, though 
seldom, entirely disappear. The muscular ridge and rough surface which mark the back 
part of this bone, below the middle of the shaft, in Pidus are not usually so distinct in 
Pezophaps ; the former, however, can be traced in all the specimens. At the upper end 
of the coracoid the articular surfaces (Plate XIX. figs. 77, 81) have much the same 
aspect as in Pidus , an allowance being made for its increased relative thickness. The 
surface for the articulation of the scapula is, however, proportionately larger. It is very 
variable in form, being in most of the larger (male?) examples roughly quadrate, in the 
smaller (female?) triangular J. In these last there commonly springs from the inner 
corner a somewhat elongate process, directed inward and slightly downward, and then 
curving backward. The humeral articular surface is not nearly so well defined, and is 
not oval and smooth as in Pidus , but generally somewhat semilunar in shape, with 
two or three depressions, caused (as some of the specimens show) by minute foramina 
(which in others are entirely obliterated) §. The surface for the articulation of the 
* In Diclunculus the coracoid exhibits certain well-marked characters ; whether they are diagnostic we will 
not say. Chief among them is the increased development of a muscnlar ridge along the inner edge of the sternal 
end. Arising at the front part of the sternal articulation it passes obliquely upward and backward till it ceases 
abruptly on meeting the usual ridge running along the internal side of the bone. At the other end of this 
bone also the usual internal curved process which serves as the basis of attachment of the upper side of the 
scapula is more largely developed than seems to be usual with the Doves, being extended downward so as most 
nearly to meet the furcula. 
t The absence of pneumaticity in the coracoids of the Ardeidce, and their consequent difference of colour, is a 
striking feature when the specimens have not been macerated so much as to remove all the grease from the bones. 
+ An additional proof of the value of the characters furnished almost invariably by this bone in birds, on 
which I have elsewhere slightly touched (Ibis, 1868, p. 95). — A. N. 
§ In one specimen of the coracoid of Didus in our possession there exists a good-sized deep circular hollow 
on the lower side of the bone, immediately opposite to the humeral articular surface. There is no trace of this 
hollow in any of the other six coracoids we have. This same specimen presents on its upper aspect a singularly 
