
257 
shows the same general specific characters as the one above described, with individual 
differences due to degrees of ossification of the posterior half of the sternum. 
The lateral notch, measured along the border formed by the lateral process, is 5 inches 
3 lines; the length of the mesial process along the median line is 3 inches 3 lines, the 
breadth of that process across its middle part 2 inches 3 lines; it is thus shorter and 
broader than in the first described specimen (Pls. LX XIII. & LXXIV,). The lateral 
processes, about 53 inches in length, are rather broader than in figs. 1 & 3, but diverge 
at a corresponding angle. The more essential characters afforded by the costal tracts 
and processes, the coracoid depressions, and pneumatic fosse, with the general outward 
convexity and inward concavity of the entire part of the sternal body, exemplify the 
specific identity of these bones. 
Thus the specimens of sternum in Mr. Sumpter’s collection show two well-marked 
modifications of the Dinornithic or Apterygian type of sternum, which type may be 
characterized as “ subquadrate, keelless, more or less flattened, with a pair of deep and 
wide posterior notches, and with small and remote coracoid pits.” The characteristic 
which differentiates Apterya is the anterior emargination: in Dinornis the different 
degrees of divergence of the lateral processes, involving corresponding differences in the 
breadth of the sternum, appear to be the best-marked modifications, though not the 
only ones; but before referring to these I may note the concurrence of the broad 
modification with the peculiar robustness of the legs and breadth of pelvis in Dinornis 
elephantopus. ‘The chief minor modification, after the difference of divergence and 
breadth, is that shown by the costal tract in a fragmentary sternum obtained by 
Mr, Percy Earl from the turbary at Waikawaite, and referred by me to “ one of the 
larger, if not of the largest, species of Dinornis” (p. 124). The costal tract occupies a 
relatively greater extent of the lateral border of the sternum between the coracoid and 
lateral processes, and shows in that extent three articular transverse ridges (Pl. XXXYV, 
fig, 3,7 r” rrr’), ‘The anterior border (ib. fig. 2) exceeds in extent that in Dinornis 
élephantopus, but is less thick ; the lateral processes are more slender, with a narrower 
base and minor degree of divergence, so far as can be judged from the proportion of 
the process preserved and figured in pl. 43, fig. 1, p. 
The more perfect sternum of the smaller species of Dinornis described (p. 197) and 
figured in Pl. XLVIILI. figs. 1-4, is of the same type as that of Dinornis rheides. Its 
somewhat smaller size, with the more acute termination of the posterior notches, and 
greater prominence of the mid part of the outer surface of the sternum anteriorly, 
as shown in fig. 4 of the plate above cited, induced me to regard it as having come 
from a different species, which is very probably the Déinornis casuarinus, leg-bones ot 
which have also been obtained from deposits south of Otago, Middle Island, where the 
sternum (Pl. XLVIII. figs. 1-4) was found. 
At the beginning of the work of determination of the remains of wingless birds from 
New Zealand I found two generic types of skull, and referred in 1848 one of these 
