PROFESSOR OWEN ON THE GENUS DINORNIS. 147 
is evidence of the two outer and larger curves (3, d, fig. 3), convex forward, continued as 
in Dinornis struthoides (op. & tab. cit.) to the paroccipital ridges. These were inclined 
backward, as in 2hea and Dinornis ; but to what degree, or how far the ridges descended, 
the broken specimen gives no information. 
Against an indication of a short pterapophysis, on the right side, part of what is 
plainly a pterygoid abuts by its hinder end; this lamelliform bone extends forward and, 
as in Rhea, slightly outward, and joins a similar fragment of a lamelliform palatine 
which has been pressed upward into the orbit, above the level of the presphenoidal 
rostrum (fig. 2,9). Of this rostrum, a length of nearly two inches is continued forward 
from the basisphenoid (ib. 5); its wide-celled pneumatic structure is exposed, as one 
sees in similarly abraded Dinornis-skulls. To the left of the anterior broken end of the 
rostrum, in the same relative position as in Dinornis robustus, is a portion of the hind 
part of a broad palatal plate of the premaxillary (fig. 2, 22); and suturally connected 
therewith is the palatal process of the maxillary (ib. 21), fractured across where it was 
contracting and thickening to join the palatine bone (compare Pl. XVI. fig. 2, 22” 21”, 
with Pl. XV. and Trans. Zool. Soc. vol. v. pl. 56. fig. 1, 22” 21” ). 
On the left side of the cranium, part of the smooth upper surface is continued upon a 
process arching downward (Pl. XVI. fig. 4,12), which I regard as homologous with the 
postfrontal in Dinornis; the broken termination shows a fore-and-aft breadth of 5 lines, 
a transverse thickness of 2 lines; and the fracture exposes the same open pneumatic 
diploé as in Dinornis. This process is distant from the back part of what remains of 
the paroccipital process 1 inch 3 lines. It is consequently nearer that process, being 
more backwardly situated, than in Dinornis robustus or D. elephantopus. But the Moas 
differ among themselves in this respect, according to, or with concomitant differences 
in, the antero-posterior extent of the temporal fosse. Thus Dinornis rheides more 
resembles Dasornis in this respect. But in the proximity of the postfrontals to the 
occiput Dasornis still more nearly resembles Struthio; and the resemblance extends to 
a concomitant large expanse of the superorbital arch. 
Again, we find in what is preserved of the fore part of the cranium a marked departure 
from the dinornithic type, and an adhesion as well marked to that of existing Struthio- 
nide. The fore half of the interorbital part of the frontals (fig. 1, 11) is contracted, as 
in Rhea and Dromaius, and is concave transversely, as in Rhea. ‘To its sides articulate 
the broad hind parts of a pair of bones (ib. 15,15) which I regard as homologous with 
the two distinct nasals in Rhea and Struthio. These parts of the nasals, beginning 
narrow, or by a point, behind, rapidly expand and meet as they advance, so as to give 
a pointed form to the included part of the calvarium. Whether this part be the 
frontal (11), or an exposed surface of the connate prefrontals (athmoid, 14), the abraded 
surface of the bone does not permit to be defined with certainty. 
The structure of this interesting fossil, as far as it can be defined, shows it to be of 
a bird; its configuration and proportions exemplify combinations of dinornithic and 
VOL. ViI.—PART I. Jan. 1870. x 
