SKULL OF THE ^GITHOGNATHOUS BIRDS. 299 



behind they send upwards and inwards a long epipterygoid hook {e.pg). The palatines 

 are extremely delicate, and, for the length of the head, not long; for their frail prse- 

 palatal bar (pr.pa) is soon lost in the base of the large rostrum. Passing backwards, 

 we see how wide they become in the middle to form the transpalatine lozenge (t.pa) 

 which is relatively about the largest ever seen. If the zygomata had kept parallel, 

 instead of diverging to articulate with the large quadrate bones, they would have 

 passed close to the edge of the transpalatine process on each side, as in those Rep- 

 tilia which possess an " os transversum " '. 



The broad part of these palatines (figs. 1 & 2) is formed of very fine, papyraceous, 

 wavy plates, whose strength is increased by a selvage or ribbed edge on the outer 

 margin of the transpalatine, the other free edges being irregularly toothed. The 

 oblique and somewhat steep transpalatines send backwards and inwards a narrow but 

 widening isthmus, which grows into two plates — the lower, or interpalatine, being 

 small and but little spiked, whilst the upper, or ethmo-palatine, is broad, half coiled, 

 and united to the vomer (figs. 1, 2, d, i.pa, e.pa). The postpalatine laminae (ptpa) are 

 deep and steep, and end suddenly, with a high emarginate edge : between these parts 

 the parasphenoid is bare. The rostrum, as seen from below, is deeply sulcate ; late- 

 rally there is a small submarginal sulcus behind ; the whole structure is solid and gently 

 arched. The palatal part of the prsemaxillaiy sends a median tongue of bone beneath 

 the front part of the nasal floor ; the sublateral processes are united to the prsepala- 

 tine bar ; and the dentary edge ends in a free angular process, which overlaps the 

 descending crus of the nasal and the wide part of the praemaxillary ; but all these 

 bones are soldered together. The j ugal (j) is continuous with the maxillary (mx) ; and the 

 maxillo-palatine processes (mx.p) have a long inbent stalk, and a pedate blade, which is 

 two-toed, behind. These processes are the frail homologues of the dense bony fore 

 palate of the Mammal. 



The two bacilliform vomers are united by a commissural plate, whose hinder margin 

 reaches but little beyond the middle of the bars ; and the free gently diverging " horns " 

 are of greater extent than the uniting plate. That plate is a little convex below and 

 concave above ; and the whole bone is such as appears to belong only to the Nectari- 

 niidse. No septo-maxillary appears on it ; but the anterior bones are grafted on the 

 inturned alinasal lamina as usual {v, i. al). The whole alinasal part of the labyrinth 

 (indicated by faint drawing) is unossified, as well as tlie inferior turbinals. It corre- 

 sponds very closely with what is seen in Ruticilla (Part I. pi. Iv. fig. 13) ; but I have 

 not been able to find alee to the septum, either in this type or the next. On the whole, 

 the huge spongy ecto-ethmoidal mass {f.p, e.eth) is like what is seen in the noblest 



' The setting free of the complex palate, the attenuation of the zygoma into a fine " spring," the mobility of 

 the quadrate, the gliding of the palate on the basicranial axis, and the formation of a cranio-facial hinge — all 

 these things are ornithic specializations of equal interest to the student of fitness and the student of form. 



VOL. X. — PART VI. No. 7. — June \st, 1878. 2t 



