86 MR. W. p. PYCEAFT ON THE [feb. l6, 



anterior narial apertures extended almost the whole length of the 

 upper jaw, while, as has already been remarked, in the adult they 

 are quite obliterated (PL YIII. lig. 2). 



In the skull of Fregata (PI, VII. fig. 1) the free posterior extre- 

 mities of the maxillo-palatine processes closely resemble those of 

 Phaethon. They differ, however, markedly in the fact that they 

 extend vertically as well also as horizontally, reaching upwards to 

 the roof of the upper jaw. They are divided by a distinct and 

 osseous septum nasi, swollen dorsally. Vestiges of a precisely 

 similar septum are found in Fhalacivcorcuv and Pdecanus. There is 

 a long, slender, curved vomer, anteriorly resting upon the posterior 

 ends of the fused maxillo-palatines and posteriorly fused with the 

 palatines. The pterygoid ends of these last, again, are perfectly 

 ankylosed, just as they are in many Ciconice. 



The basitemporal plate in Fregata, like that of Sula^ does not 

 cover the whole basisphenoidal surface. Unlike that of Sula, 

 however, it still preserves a free edge, though this is very slight. 

 A short distance in front of the anterior border of this plate, in a 

 slit-like depression of the basisphenoidal rostrum, lie the Eustachian 

 apertures, though so small that only a very slender bristle can be 

 passed through them. If this skull be compared with that of 

 Sula, faint traces of the Eustachian tubes in this latter genus 

 will be found, as well as a faint ridge, representing the once free 

 edge of the basitemporal plate. 



The uatui-e of the mandibular articular surfaces of the quadrate 

 deserves some notice, since, if this had been adopted instead of the 

 form of the basitemporal plate, and basisphenoid, for the purposes 

 of systematic arrangement, the results would have been almost 

 identical. These surfaces are two in number, the qnadrato-jugal 

 and the pterygoid. 



In Phaethon and Pelecanus the quadrato-jugal surface is directed 

 obliquely outwards and forwards ; that of Pelecanus being much 

 broader in proportion to its length than that of Phaethon. The 

 pterygoid surface is placed almost at right angles to the long axis 

 of the skull. The two surfaces are divided by a groove, which is 

 most marked in Pelecanus. The general impression of the articular 

 end of this bone as a whole is that of a narrow bar continued 

 backward from the pterygoid to the quadrato-jugal bar, which it 

 joins almost at right angles. 



In Fregata, Sula, and Phalacrocorax the surfaces have a V-shaped 

 arrangement. In Phalacrocora.v the^ has almost become U-shaped. 

 The pterygoid surface is subcircular, the quadrato-jugal hour- 

 glass-shaped. 



Though the articular surfaces of the quadrate in Sula and 

 Fregata closely resemble one another, that of Fregata can at once 

 be distinguished by the form of the orbital process, which is verj' 

 large, with a broadly expanded free end resembling that of the 

 Heron. In Sula, as in Phalacrocorar and Plotus, the orbital 

 process is reduced to a small spur standing out at i-ight angles to 

 the long axis of the bone and about halfway down. 



