5y4 Transactions of the Uoyal Society of South Africa. 
should be more carefully and extensively studied by ornithotomists. The 
few notes I have given here upon them should be compared with descriptions 
of mine in former papers for other species of the Psittaci, and more especially 
with Mivart's figures and descriptions for Psittacus erithacus and Lorius 
fiavopalliatus. 
When we come to compare the i^elves of different parrots, we soon 
observe that there is considerable uniformity in the matter of their morpho- 
logy, with respect to the various species mentioned in this paper. To this 
statement Palaeornis forms no exception : Gonurus carolinensis is not far oft", 
and this is probably true for a good many other species of parrots (see 
Plate XLI, fig. 7.) 
Viewed from above, the pelvis of Palaeornis docilis shows the anterior 
rounded and emarginated borders of the ilia, and the rather elongate, con- 
caved surfaces of these bones between these borders and the cotyloid cavity 
upon either side. These bones do not meet in the middle line in this 
region, while they do meet and form an osseous union with the crest of the 
sacrum. Posteriorly, the ilio-neural canals are entirely sealed over, and but 
few foramina are found to exist between the diapophyses of the uro-sacral 
vertebrae. All this formation gives a very smooth and unbroken appearance 
to the dorsal aspect of the pelvis of this and other parrots, which is quite 
characteristic of this bone in the case of other members of the group. 
On lateral view, it is seen that the obturator foramen is, as a vacuity, 
almost continuous with the obturator space ; the ischiadic foramen is rather 
large and subelliptical in outline. Another feature of note is the pointed- 
ness and elongation of the ischium in the posterior direction, the distal apex 
of this bone of the pelvis coming in contact with the upper margin of the 
long, slender pubic element behind. As a *' notch," the ilio-ischiadic one is 
hardly worthy of the name ; in other words, scarcely any mark of differen- 
tiation on the posterior pelvic border exists at all to define the meeting of 
the ilium and the ischium. We know this " notch " is very pronounced in 
the pelves of some birds. 
There are six free caudal vertebrae, the transverse processes of which, all 
to those of the ultimate one, are rather spreading, and not very markedly 
bent ventrad. To these is to be added a triangular pygostyle of correspond- 
ing proportions with a blunt-pointed apex. 
Amazona leucocephala possesses but five free caudal vertebrae, and the 
pygostyle differs in form entirely from that bone in Palaeornis, for it is of a 
quadrilateral outline in the former bird, with a thickened posterior border, 
and, moreover, it is pierced from side to side at its antero -inferior angle by 
a round foramen. In the specimen before me, there is also a deep little pit 
in the lower part of the posterior margin. 
Curiously enough, Ara severa has hut four free caudal vertebrae, and the 
rather large pygostyle is inclined to be somewhat pointed superiorly. Ara 
