428 
The presence of a back toe, inferred from the structure of the metatarsus, in Dinornis 
ingens and D. dromioides (p. 135), was subsequently confirmed by the acquisition of the 
proximal phalanx of that toe in D. robustus (Plate XLIX. figs. 1,7, 4 & 5), in D, erassus 
(Plate CXIII. fig. 2), and in D. rheides (Plate CLX.). 
The indication, it is true, on which the inference was hazarded is of the faintest 
character ; and the diminutive size and functionless condition of the high-placed ‘ hallux’ 
accord therewith. The attachment of this rudimental toe was merely ligamentous. 
The probability, however, I think, is, that, im the species of Dinornis in which the 
little bones have not been picked up with the larger parts of the skeleton, the absence 
of the hallux is due to that accident, rather than to non-development of the toe in 
such species, Its presence in the foot of species with the integument of that part 
preserved, as in Dinornis ingens, the metatarsus of which shows no conspicuous trace 
of such attachment, exposes the weakness of any conclusion from that character. 
I therefore abandoned the generic character founded on the inference and sub- 
sequent discovery of the hallux shortly after the suggestion of such taxonomic value. 
One could not place Dinornis giganteus and its probably local variety, Dinornis robustus, 
in distinct genera, because no complete series of bones, including those of the hallux, 
had been found in the North Island, where D. giganteus seems to represent the D. ro- 
hustus of the South Island; neither could the species didiformis be retained as a 
Dinornis, and the species dromioides be distinguished as a Palapteryx, on like grounds ; 
nor D, elephantopus be similarly separated from D. crassus?, 
‘The scapular arch, like the hallux, was originally inferred (p, 124) from an articular 
fossa of the sternum (Plate XX XV, fig. 2,¢, ¢c); and the inference as to the small pro- 
portional size of such arch, in comparison with that of the Ostrich and other existing 
wingless birds, was confirmed by the determination of the confluent coracoid and sca- 
pula in the Dinornis robustus (p. 170, Plate LXTY. figs. 2, 3 & 4). 
Small and shallow as were the depressions in the sternum of that species to which 
the coracoid had been ligamentously attached, these indications are less definite in the 
sternum of some other species; and with regard to the scapular arch, as to the hallus, 
it has been surmised that such arch was naturally absent in some kinds of Dinornis, as, 
for example, in the skeleton of D. crassus (Plate CVIII.), D. rheides (Plate CLX.), 
D. gravis (Plate CX.), D. didiformis (Plate CXI,), D. gracilis and D. casuarinus (Plate 
CXII,). Lincline to the belief, however, that the skeletons in the Museum of Natural 
History, Christchurch, Canterbury Province, from which the photographs of those 
species were taken, are deficient in respect of the scapular arch, through the accident 
of the non-discovery of that small and slender rib-like bone, rather than that it was not 
originally present, 
' Trans. New-Zealand Institute, vol. vii. 1875, p. 266, pl, xix, 
* Ina letter from Capt. Hutton, dated ‘* Otago Museum, May 6, 1876,” he writes: “* We have in the Museum 
legs, with the hind toe belonging to them, of D. ingens, D. cusuarinus, and D, gravis,” 
