PHYLOGENY OF THE PAL^EOGNATHJE AND NEOGNATELE. 227 



wards, this region of the pelvis in this respect resembling Struthio. In half-grown 

 specimens it is quite pre-caudad ; in adult birds it becomes fused with the extreme 

 postero-dorsal angle of the ischium. 



The ilium increases markedly in breadth during its growth. This increased breadth 

 accompanies a corresponding increase in the height of the neural spines of the sacral 

 and post-synsacral vertebrae, which, as has already been described (p. 208), in the 

 embryo are represented only by low cartilaginous transverse ridges. In the embryo the 

 pre- and post-acetabular ilia are slightly convex superficially, in the adult they become 

 slightly concave. 



The supra-trochanteric process marks the region, more or less, of the sacral vertebrae. 

 It lies immediately above the anti-trochanter, but does not project outwards as in 

 Rhea. One of the most remarkable features of the pelvis of Dromceus and Casuarius 

 is the close embrace with which the post-ilia receive the post-sacral vertebrae of the 

 synsacrum, almost entirely suppressing the para- and diapophysial processes. 



In Dromceus and Casuarius the sacral vertebrae extend outwards beneath the vertical 

 border of the ilium to articulate with the ascending process of the ischium at its 

 junction with the ilium. 



The pelvis of D. ater differs from that of D. novce-hollandice chiefly in that the 

 development of the pectineal process is but slight — wherein it approaches Casuarius, — 

 and in that the obturator fissure is wider, the rotation backwards of the pelvis being 

 less. The crest of the pre-ilium is less arched. 



The pelvis of Dromceus can scarcely be distinguished from that of Casuarius. 



In the young bird (half-grown) the neural spines of the sacral region of the syn- 

 sacrum appear to be more transversely expanded than in Casuarius. In the adult the 

 pre-ilium appears to be relatively of greater breadth in Casuarius and the post-ilium is 

 relatively longer and narrower. The strong -^ -shaped supra-trochanteric processes 

 and the width across from one side to the other are apparently rarely attained by the 

 Cassowary. 



Judging from the skeletons in the Museum collections it would seem that in the 

 width across the supra-trochanteric process and the relatively long post-ilium, the 

 pelvis of Dromceus is readily distinguishable from that of Casuarius. 



Unfortunately, however, these data do not appear to be really reliable, for in a pelvis 

 of C. casuarius the resemblance to these very characteristic points in Dromceus were so 

 close as to render it most difficult to distinguish between the two genera. 



In C. bennetti the width across the supra-trochanteric process is as conspicuous as in 

 Dromceus, but the whole pelvis is of course much smaller than in Dromceus. 



In the young (half-grown), both of Casuarius and Dromceus, the ischiadic fissure is 

 complete. Later in life it is closed by fusion of the extreme antero-dorsal angle of the 

 fore end of the ischium fusing with the ilium. In very old birds the width of this 

 fissure caudad becomes seriously encroached upon by the downgrowth from the post- 



