234 PROFESSOR OWEN ON THE GENUS DINORNIS. 
D. parvus, D. crassus* differs in the minor breadth across the zygomatic arches (ib. 27,27) 
in proportion to the length of the skull, which exceeds that of the present species, 
and differs more markedly in the broader and more obtuse premaxillary, and in the 
minor breadth and capacity of the temporal fossee; the coronoid process also rises 
higher, but has less basal extent. Dinornis gravis*, with equal length of skull, has 
greater breadth across the postfrontals, with less breadth of the temporal fosse. 
Dinornis casuarinus*, with equal length, has a shorter extent of the premaxillary part 
of the upper beak, and a narrower nasal process of the same bone. 
In the occipital region (Pl. LIl. fig. 4) the paroccipitals (4,4) are shorter and 
narrower, or less obtuse, than in D. rheides*; the occipital tubercle (1) is larger in 
D. parvus ; the ectotympanic processes of the mastoid (8') are rather shorter in D. parvus. 
The upper transverse superoccipital ridge, to which the parietal (7) and mastoid (s) 
also contribute, is more strongly marked than in D. rheides. The masto-paroccipital 
wall (ib. fig. 1,4, 8) has a less arched border than in D. rheides; the basioccipital ma- 
milloid tuberosities (fig. 4, a, a) are relatively larger than in that species, indicative of 
greater power in downward strokes of the bill. The posterior wall of the Eustachian 
canal (ib. fig. 3, e) is less developed than in D. rheides; this canal shallows as it 
advances, and is lost on the plane of the basis cranii opposite the pterapophyses (<7). 
The chief character of the basisphenoid, and to me the most interesting, is the well- 
marked longitudinal channel and foramen (fig. 3, 5,5; see also D. ingens* and D. ele- 
phantopus®), which I take to be residuary indications of the upward embryonal produc- 
tion of the primary mouth, subsequently becoming, or communicating with the conario- 
hypophysial tract, which crosses vertically the brain, by the ‘third ventricle,’ between 
the pros- and mesencephalon. The pterapophyses (fig. 3, 5’) are relatively larger 
(longer in proportion to their breadth) than in Dinornis ingens and D. elephantopus. 
The presphenoid is not carinate below, or is less so at its mid extent, than in D. gravis, 
D. elephantopus (ib. ib.), and D. ingens, where the keeled part is longer. The presphe- 
noid in D. parvus is not compressed at its mid third, as in D. gravis’, and is not 
expanded there as in D. robustus*. 
The alisphenoid is indicated, as usual, by the ‘foramen ovale,’ as is the orbito- 
sphenoid by the larger ‘foramen opticum:’ the original divisions of these cranial 
neurapophyses are soon obliterated, and their bases become confluent with their 
1 Trans. Zool. Soc. vol. vii. pl. xi. figs. 2, 3; Memoirs, vol. i. p. 266. 
? Tbid. vol. vii. pl. xiv, fig. 3; Memoirs, vol. i. p. 280, pl. Ixxxi. 
* Ibid. vol. vii. pl. xiii. fig. 1; Memoirs, vol. i. p. 278, pl. Ixxvii. 
4 Ibid. vol. vii. pl. xii. fig. 2; Memoirs, vol. i. p. 271, pl. Ixxy. 
§ Thid. vol. vii. pl. xv. fig. 3, 5’; Memoirs, vol. i. p. 281. 
6 Ibid. vol. vii. pl. x. fig. 4, 5’; Memoirs, vol. i. p. 262. 
7 Ibid. vol. vii. pl. xiv. fig. 4. 
* Ibid. vol. y. pl. lxiy., pp. 344-355. 
