417 
power of flight, that the sternum does not increase in size in the ratio of the increase of 
the legs and general stature of the species of Dinornis. 
The length of the present sternum at its mid line, and the breadth from tip to tip of 
the lateral processes, are less than that of the subject of Plate LX XII. But the 
length of the body of the sternum anterior to the lateral hind notches is greater 
in the present bone, whilst that of the part prolonged between those notches is less ; 
the costal border (Plate XCVIII. figs. 3 & 4) is nevertheless absolutely longer in 
the present sternum; and although in the angle of divergence of the lateral processes 
(i, h) it conforms to the type of the sternum of Dinornis elephantopus, and with it departs 
in a marked degree from that of Dinornis rheides, yet the differences noted are sufficient 
to indicate that the present perfect sternum accords more closely with that of the 
skeleton of Dinornis robustus, and has, at least, belonged to a distinct species of the 
eroup of Moas, which, on sternal characters, might be denoted by the generic term 
Palapteryz. 
There is, however, a gradational approach to the less divergent type of sternum 
exemplified in Dinornis rheides. In (Palapterya) elephantopus, e. g., the breadth of the 
fore border of the sternum being 8 inches, that of the bone at the ends of the divergent 
processes is 15 inches. 
In Dinornis robustus the breadth of the fore border being 8 inches 3 lines, that of 
the bone at the ends of the divergent process is 135 inches, 
In Dinornis maximus the breadth of the fore border of the sternum being 9 inches, 
that of the bone at the end of the divergent processes is but 12 inches; moreover, the 
length of the sternum anterior to the notches is relatively rather more than in D, ro- 
bustus, and makes a corresponding approach to the more elongate type of sternum 
represented by D. rhetdes. 
With these approaches, in sternal gradations, to that type the limb-bones coincide, 
showing longer and more slender proportions as compared with Dinornis elephantopus 
and PD, crassus; so that, even accepting, or resuming for convenience’ sake, my old 
subdivision of Moas into two subgenera, I cannot shut out the conviction of its essential 
artificiality. 
It is hardly probable that a nominal generic distinction will be ultimately accepted 
on the differences here pointed out between the sternums of the species of Dinornis 
outlined in cut, fig. 35, especially as they are associated with corresponding gradational 
differences of proportion in the bones of the hind limbs. 
Admitting such generic or subgeneric group for the species crassus and élephantopus, 
showing the extreme divergence of sternal processes with robustness of hind limbs, 
and if the term Palapteryx had not had priority, I must have adopted Reichenbach’s 
Emeus', of which my Dinornis (Palapteryx) crassus is the type. 
1 Das natiirliche System der Vogel, 4to, 1849-50, p. xxx. 
as 
=~ 
