PHYLOGENY OF THE PALZOGNATHA AND NEOGNATHA, 237 
extends proximad on to the shaft, terminating rather behind its middle. There are 
no conspicuous pneumatic foramina. The great trochanter is very large; it is pro- 
duced upwards above the head, and not backwards caudad of the head, being an 
exaggeration of what obtains in Dromeus and Casuarius. The popliteal fossa is well 
defined, but not deep. 
The tibio-tarsus is much flattened from front to back, as in Apteryx and Apyornis ; 
the ectocnemial crest is large, and gives the anterior view of the proximal end of the 
bone a flabellate form; the procnemial crest is feebly developed. The shaft may or 
may not be inflected—according to the genus. ‘There is a strong extensor bridge and an 
intercondylar tubercle. 
In £pyornis the femur is also very short. The great trochanter is very lofty and 
produced caudad. The intercondylar gorge is wide and deep, and the popliteal fossa 
shallow. 
The tibio-tarsus has its proximal end much flattened from front to back; a feeble 
procnemial and large laterally directed ectocnemial crest. Distally the shaft is 
inflected. There is no extensor bridge, and the groove is shallow. 
The tarso-metatarsus is short, wide, and grooved anteriorly, but the groove is shallow. 
The ectotrochlea is widely separated from the mesotrochlea. There is a low, median, 
hypotarsal ridge. 
Phalanx 1 of digit u. is moderately long, the 2nd is short; phalanges 2, 3 of 
digit 1v. are conspicuously short, the 4th is shorter. The proportions of the phalangeals 
agree closely with those of Dromeus. 
In working out the development of Apterya, Parker [71] found, in the earlier stages, 
all five digits present—as in some Neognathw. The fifth digit was a short conical 
cartilage attached by its proximal end to the fibulare, and by its pre-axial border to the 
distale. 
The fibula was of the same length as the tibia. 
The tarsus contained the usual elements—a tibiale, fibulare, and distale. The 
ascending process of the tibiale at no stage showed any sign of a separate origin, 
comparable to an intermedium. Later, after hatching, there appears in the mesotarsal 
articular pad a pair of centralia. Sometimes only one is present. 
The procnemial crest ossifies, as usual, from a separate centre. Concerning this 
ossification I would remark that it is doubtful whether it has any phylogenetic signi- 
ficance. Rather it would seem to be comparable to the separate ossification-centre of 
the great trochanter of the mammalian femur, and to indicate a position of great 
strain. In a recent paper I, however, described it as an epiphysis, as also, though 
unknown to me at the time, did the late W. K. Parker [79]. In my own case I may 
claim some justification, since the base of this procnemial ossification—that of a 
young Grebe—was sufficiently large to form a complete tibial cap, divided from the 
shaft by cartilage, as is a true epiphysis. Moreover, the resemblance to a true 
VoL. xv.—Part v. No. 17.—December, 1900. 2 
