246 PROFESSOR OWEN ON THE GENUS DINORNIS. 
of the outlet of 1 inch 3 lines, and one of the inlet of 10 lines; through the latter the 
bodies of the tenth, eleventh, and parts of the ninth and twelfth vertebral centrums 
are visible. 
Thus the pelvis of the smallest known species of Dinornis repeats, in the main, the 
characters of that part of the skeleton in the larger species above referred to. It 
includes the same number of vertebre as that of Dinornis giganteus'. As in this 
species and D. robustus, the pelvis is broader in proportion to its length than in the 
Apteryx, but in most other respects resembles the pelvis in that bird more than it 
does that of any other known avian genus. ‘The pelvis is equally divided lengthwise 
at or by the coalesced pleurapophyses of the seventh and eighth vertebre, giving origin 
to the ischia, whereas in Struthio? the part of the pelvis behind this origin is twice as 
long as the part in front. The terminal junction of the ischium with the pubis, which 
does not take place in Dinornis parvus,is more extensive and complete in Struthio than 
in Apteryx*; and the pubis in Struthio is extended much beyond that junction, where 
the ischium terminates. It need hardly be remarked that the produced pubes bend 
downward and inward, expanding to a symphysial union in the Ostrich, since this 
mammalian character is peculiar to it among birds. 
It appears that in Dinornis robustus the terminal junction of ischium and pubis is 
as well marked as in the Apteryx, and that the distance between the ischium and 
ilium, posteriorly, is less. In Dinornis parvus, as in D. robustus, the terminal expanse 
of the ischium is relatively greater than in Apteryax, and the interspace between it and 
the ilium is less; and this narrowing is increased in D. robustus as compared with. 
D. parvus. 
If the anterior end of the pelvis of D. parvus (Pl. LIII. fig. 3) be compared with that 
of Aptornis, the preaxial surface of the first sacral is less extended transversely, the hypa- 
pophysis is wanting, and the neural spine springs more abruptly from between the 
prezygapophyses in the present Moa. 
If the same view be compared with that of Cnemiornis, the vertical contraction of 
the preaxial surface in the latter is still more marked; the difference in the origin of 
the long neural spine is the same as in the comparison with Aptornis. In this genus, 
as in Cnemiornis, the ischiadic notch is converted into a foramen by confluence of the 
ischium with the ilium; and such confluence is greater in Aptornis than in 
Cnemiornis®. 
As the osteology of Didus is recorded and illustrated in the 6th and 7th volumes of 
the ‘Transactions’ of the Zoological Society, I may here add to the comparison of 
Dinornis with other genera of extinct birds deprived of the power of flight, that of 
Didus in relation to the pelvis. It includes sixteen coalesced vertebre, with which the 
1 Trans. Zool, Soc. vol. iii. 1839, pl. xix. fig. 4. ? Op. cit. vol. iii. pl. xix. 
* Op. cit. vol, ii. pls, liv, & lv. 
