1898.] OSTEOLOGY OF BIEDS. 85 



height of the upper jaw in fact, and backwards into the lachrymo- 

 nasal fossa, and are of com-se fused in the middle line ventrally. 

 Seen from behind (PL VIII. fig. 6), they are quite distinct one 

 from another. In some specimens, what I take to be traces of an 

 osseous septum nasi are found. As in the Storks, the maxilla and 

 maxillo-palatine processes make up the greater part of the upper 

 jaw. The nasal hinge is generally well marked. 



The palatines are completely fused from the posterior narial 

 aperture backwards, and, further, are provided with an enormous 

 dorsal and ventral median keel. A trace of this, as we shall see, 

 is found in Sula. 



The skull of Sula, in the obliteration of the anterior narial 

 apertures, the form of the maxillo-palatine processes and of the 

 palatines, closely resembles that of Phalacrocorcw. These are 

 some of the latest acquirements of the group, and tend, amongst 

 other things, to single out the two families which they represent as 

 conspicuously " Steganopodous." They may perhaps be regarded 

 as the most intensely modified members of the suborder. 



The maxillo-palatine processes of Sula difi'er from those of 

 Phalacfocorax in that their coalescence is more complete. Seen 

 from behind, they present an obliquely truncated surface of 

 cancellated or lattice-like tissue, which ventrally does not even 

 extend as far backwards as the posterior end of the maxilla itself. 

 Moreover, a closer examination shows that the bony tissue of the 

 interior of these processes has been more or less completely 

 absorbed, so that the truncated posterior end just described is 

 practically a mere shell or screen concealing the hollow space 

 within. There is no trace of an osseous septum nasi. The 

 palatines are completely fused in the middle line from the posterior 

 narial aperture backwards, and there is a slight median dorsal and 

 ventral keel, just as in Peleccmus, but less developed. In Phala- 

 crocorax this region of the palatines is rarely, if ever, fused 

 throughout its whole length. An open suture is generally visible. 

 In Plotus it appears to be constantly fused. The fronto-nasal 

 hinge is strongly marked. 



The basitemporal plate and basisphenoid appear to be an ex- 

 tremely modified form of that obtaining in Fregata. The former 

 was very small, not more than half covering the latter, with 

 which it had so completely fused that what should be its free 

 edge is only traceable as a thin faint line. The apertures of the 

 Eustachian tubes appear to have become completely obliterated, 

 leaving only a faint scar on the basisphenoidal rostrum. 



Sula and Phalacrocorax are the only Steganopodes in which the 

 postorbital process is emarginate. This is a common feature 

 amongst the Ciconise and Ardese, e. g. Pseudotantalus, Nycticorax, 

 Cancroma. 



In the skull of a very young nestling Sula I found the palate 

 to be scliizognaihous, as in PhaetJion, which it closely resembled 

 in the triradiate form of its maxillo-palatine processes. Another 

 point of very considerable significance was the fact that the 



