370 Annals of the Carnegie Museum. 



turns outwards to fuse with the inner side of the superior mandible 



(fig- 4). 



The short, scroll-Hke maxillo-palatines do not meet in the middle 

 line, nor do they come in contact with the anterior free extremity of 

 the vomer. Such part of the roof of the mouth as lies beyond the 

 palatines and between the sides of the prcmaxillary is not spanned 

 by bone, thus allowing an uninterrupted view of the interior of the 

 fore part of the rhinal chamber. 



As one would naturally suppose, the medio-stapedial element of the 

 middle ear, and the sdcrotal plates of the eye, are, for very obvious 

 reasons, not to be found among the bones of these two collections. 



The mandible has the typical V-shape pattern, with the sides of the 

 anterior dentary portion low and thickish. The symphysis is very 

 short, while the apex of the enlarged distal extremity is acute and 

 decurved. It is longitudinally grooved dorsally. For the posterior 

 ramal portion, we find the side thin and lofty, with the superior 

 margin sharp and the lower one rounded. No " splenial vacuity" is 

 present in fully adult birds, while posteriorly, on the mesial aspect, 

 there is considerable excavation or concavity just anterior to the en- 

 larged articular posterior extremity. This latter is concaved and 

 truncate on its posterior aspect; double-concaved dorsad, to accom- 

 modate its surface to the quadratal articulation, and its inturned part, 

 on this upper side, showing the minute, pneumatic foramen usually 

 found in that situation. 



The Vertebral Column. — When Forbes came to study the Petrels 

 collected by the naturalists of the Challenger Expedition, as cited on 

 a previous page, he found, in a specimen of " CEstrelata grisea" that 

 there were fifteen (15) cervical vertebra:, seven (7) dorsals, eleven 

 (11) sacrals, and seven (7) caudals, making forty (40) in all. As 

 he does not mention the pygostylc, I presume he simply considered it 

 as the ultimate or terminal caudal. ^^ He also found that in " CEstre- 

 lata grisea" there were eight (8) pairs of vertebral ribs — that is, one 

 cervical pair and seven (7) dorsal, or ribs joining with the costal 

 ribs below. The dorsal ribs supported epiplenral appendages, with 



1* He also examined another Petrel, which I am inclined to believe was 

 not an CEstrelafa grisea (b), and he seemed to doubt the correctness of the 

 identification. It had forty-one (41) vertebrae in its spinal column {loc. cit., 

 p. 419). 



