
389 
distinctly and separately articulate with the sternum; those of the fourth dorsal less 
distinctly than is shown in Plate XCVIL. As a rule, I believe their sternal portions 
join those of the third pair, as the sternal portions of the fifth pair are terminally 
attached to those of the fourth pair. 
The sternum (Plate XCVI. figs. 2 and 5) shows two well-defined ridges for the 
articulation of sternal ribs on each of the short costal borders. The angle in advance 
is feebly notched for the articulation of the coracoid, which, with the confluent scapula, 
is indicated in situ in the side view of the sternum in figure 1. The anterior surface 
of the sternum is shown in figure 2, and the under or outer surface in figure 3. 
The lateral processes are less slender and more divaricated than in Dinornis rheitdes 
(Plate LX-XIII.), more slender and less divaricated than in Dinornis elephantopus 
(Plate LXXII.). The lateral notches are shallower and wider than in D. rieides, 
deeper and narrower than in J. elephantopus. ‘The postmedial notch is well marked, 
but there is no foramen or vacuity in the body of the sternum of Dinornis robustus as 
in that of Dinornis rheides (at least in the subject figured in Plates LAXUT. and 
LXXIV,). On the whole, the sternum of D. robustus most resembles that of 
D. maximus ; but its ossified portion, or body, is longer in proportion to its breadth, as 
are the lateral and middle processes, if we may so call the portions bounding the two 
angular unossified tracts of the sternum. The notch at the apex of the mid-production 
is deeper and shallower than in D. maximus. The sternum of D. giganteus and 
D. ingens 1 have not seen, 
The sacrum of Dinornis robustus (Plate XCVI. fig. 4) includes, as in D. maximus 
(Plate CXTIV. fig. 2), seventeen vertebre. Of these the ribs of the first and second 
retain the movable joints with the centrum; those of the third sacral are anchylosed 
and shortened; those of the fourth are still shorter, and terminally coalesce with the 
antecedent ribs and with the ilium. After the eighth sacral the ribs are suppressed ; 
they are resumed (as parapophyses) in the eleventh and following sacrals. 
The portion of the Moa chick’s skeleton found with that of the (seeming) parent 
included, with some other parts, the pelvis, sternum, and scapular arch. 
The sacrum, of which the structure in the mature bird is described ina larger species 
(p, 420), is formed, not only by certain caudal vertebra, united with one another and 
with an antecedent sacral, but by anterior vertebra bearing, as in the ‘dorsal’ ones, free 
ribs, and by others haying shorter coalesced ribs, as in ‘lumbar vertebra.’ If a name 
should be required by anatomy for the vertebree massed together in the full-grown bird 
to form the ‘sacrum’ and anthropotomically ‘ sacral,’ by the character of such anchylosis, 
then ‘dorso-sacral,’ ‘lumbo-sacral,’ ‘ sacral proper,’ as well as * urosacral,’ are terms 
needed to express correctly the vertebrae which go “to form the ‘sacrum’ of a bird i" 
The ilium, ischium, and pubis of the right side of the immature pelvis are bgaret in 
Plate CXV. fig. 6. ‘The ischium and pubis have coalesced ; the ilium retains its distinet- 
L See Proc, Zool. Soc. April 11, 1867, p. 422, and the supposed ostedlogical characters of Dinornis, p. 423. 
