CLASSIFICATION, AND PHTLOGENT OF THE DINORNITHID.E. 403 



plates are quite vestigial ' ; in the other genera they extend backwards and articulate 

 with the vomer : there is no vomerine process as in some of the Moas. In Stmthio, 

 Bhea, and Bromwus the body of the bone is very flat, but in Casuarius its median 

 region is elevated into a strong arched keel, from which the nasal process proceeds, and 

 there is a distinct prenarial septum. In A])teryx the form of the body is essentially 

 similar: its height is equal to its breadth, and there is a short, thick prenarial septum. 

 The unique form of the beak in this genus is due to the shortening of the body of the 

 premaxilla and to the elongation of the region between the prenarial septum in front 

 and the turbinals behind. 



The maxilla shows a wide range of variation. In the Ostrich (Plate LXII. fig. 59) it 

 is a flat bone divided posteriorly into palatine and jugal processes, and sending off" from 

 its mesial border an axe-head-shaped maxillo-palatine, which articulates by its thickened 

 inner edge with a facet on the side of the vomer. The lateral half of this process is 

 double, consisting of dorsal and ventral laminiE so arranged as to enclose a wedge-shaped 

 cavity, the antrum, open behind. In Bhea (fig. 60) the maxillo-palatine is a broad 

 flat plate which gives off" from its dorsal surface a nearly vertical, slender, ascending- 

 process, which is attached by an outer crus to the maxilla proper and by a long inner 

 crus to the maxillo-palatine. Between the two crura is a small cavity opening behind, 

 apparently the vestige of an antrum. In the Emu (fig. 61) the maxilla is narrow, 

 except at its anterior end, where it broadens out into a maxillo-palatine having the form 

 of a pocket, wide from side to side, narrow from above downvvards, and opening behind 

 along its whole width ; this cavity is obviously the antrum, resembling pretty closely 

 that of the majority of the Moas (fig. 63), but situated farther forwards. In Casuarius 

 galeatus (fig. 62) the maxilla proper is still narrower, but the maxillo-palatine has the 

 form of a long conical pouch, like a jelly-bag, its point directed forwards, dorsad of the 

 palatine process of the premaxilla, and its base widely open behind. Lastly, in Apteryx 

 the maxilla is a long flat bone and the maxillo-palatine is represented only by a narrow 

 seam-like projection of its mesial border : there is no dorsal prolongation of the 

 maxillo-palatine, the walls of the antrum being entirely membranous. 



It is obvious that in the structure of the maxillo-palatine, upon which Huxley largely 

 founded his classification of birds (lo), the Dinornithidse, with the exception of Emeus, 

 approach most nearly to the Australian Eatitae. Prof. Huxley makes no mention of 

 the antrum, describing the maxillo-palatine of the Emu, Cassowary, Moa, and Kiwi as 

 flat imperforate plates. 



In the structure of the vomer it is Apteryx which comes nearest to the Dinornithidse, 

 that genus having a single vomer, deeply cleft posteriorly and ankyl&sed with the 

 palatines and pterygoids. The maxillo-palatines touch it by a part of their thin mesial 

 edges, but do not articulate with it. In Casuarius it has the same general form and 



' In the skulls in the Otago University Museum, as i.Iso in Huxley's (lo) and Selenka's (27) figures ; but 

 iu my father's figures of advanced embryos (25) large palatine processes are shown. 



