PROFESSOR OWEN ON THE GENUS DINORNIS. 121 
surfaces are undivided, the lower one being more extensive and broader than in D. rheides, 
not projecting as a ridge. The depression between the surfaces is much less extensive 
and less deep than in D. rheides; the costal process is broader but thinner, directed 
more outwardly than in D. rheides, from which it shows the more important difference 
of a deeper and much better-defined coracoid depression, of which the anterior boun- 
dary is partially produced as a ridge, and the posterior or upper one is plainly defined. 
The articular surface, broadest externally, is concave from before backward, slightly 
convex transversely, but with a small special depression at the narrower median end ; 
there is also a pneumatic hole near the front ridge. The costal process has an obtuse, 
convex termination. The pneumatic depression on the anterior inner surface of the 
sternal body is more definite than ‘in D. crassus, but is wider and less deep than in 
D. rheides. This portion of sternum shows characters specifically distinct from those 
in the two species named. 
A third fragment of sternum in the lot, no. 16, includes the costal process and border 
of the right side of that bone. ‘The process is entire, of a different shape from that in 
either D. rheides or D. crassus, the anterior margin being obliquely truncate toward 
the narrower but obtuse apex. There is a well-marked distinction from no. 1 in the 
absence of any defined coracoid depression; and sufficient of the body of the bone is 
preserved to show the absence of a pneumatic depression, such as exists in D. rheides. 
The external convexity leading to the anterior border is less abruptly defined than in 
D. rheides. The two articular tracts on the costal border are narrow, oblique, and 
continuous as in D. crassus; but the shape of the tract is like that in no. 1 and in 
D. rheides. 
A portion of the right side of a sternum, which appears to be part of the same bone 
as no. 1, shows a continuation of the lateral process in the same direction relative to 
the contour of the costal border as in D. rheides. A corresponding portion of another 
sternum, no. 6 in lot 16, shows a divergence of the lateral process from the line of the 
costal border like that in the sternum of D. robustus, D. elephantopus, and probably 
also D. crassus, to which nos. 2 & 5 may have belonged. Assuming that nos. 1 & 3 are 
parts of the same sternum, it is certainly of another species, which probably may be 
D. casuarinus ; the portion no. 4 belongs to a different species, and the portion no. 6 
to a different individual of, perhaps, D. crassus; but four distinct birds, at least, must 
have contributed the fragments of sternum ascribed to the skeleton of D. crassus in 
Dr. Haast’s list. 
A more important contribution to the reconstruction of the extinct wingless birds of 
New Zealand has been made by the eminent State Geologist of Canterbury Settlement, 
by a series of photographs of skeletons obtained from the Glenmark Marsh, and pre- 
served articulated in the Museum at Christchurch. From the front view of that of 
the Palapteryx robustus I infer that the remark, that “the attempted restoration of the 
sternum of a large species referred to Dinornis (Vol. III. p. 516, pl. 43. figs. 1, 2, 3) may, 
