1899.] OSTEOLOGY OP THE TUBTN"AEES. 383 



sagittal crest ; in others this crest is very broad. In Procellaria, 

 Pelar/odroma, Oceanites, and Cijmodroma these fossae can hardl}'- be 

 said to exist. 



The interorbital region of the frontals is, hke the region 

 posteriorly, gently furrowed and moderately wide. This feature, 

 however, does not obtain throughout the group, but varies 

 according to the form and depth of the supra-orbital grooves 

 on either side. Thus in ProceUaria the iuterorbital region is very 

 broad, relatively broader than in any other Petrel. In this case 

 the supra-orbital grooves look outwards and not upwards as usual, 

 being only narrow and shallow depressions scooped out of the 

 free edge of the frontal. In Cymodroma these grooves are very 

 short antero-posteriorly, their length being less than the width 

 across the frontals between the lachrymals. This occurs in no 

 other Petrel. In Oceanites and Pelagodroma the grooves of either 

 side are practically confluent, reducing the interorbital region to 

 a faint and barely perceptible ridge. The grooves are wider 

 posteriorly than in any other forms, crossing the skull almost 

 transversely in this region. In Pelecanoules, Priocella, and 

 Plicehetna the grooves are divided by a high, narrow ridge, which 

 in Priofinus becomes wide enough to permit the existence of the 

 median groove previously referred to, whilst in Pelecanoides it has 

 acquired a knife-like edge. In none of the Procellariidae is there 

 ever any pronounced supra-orbital ledge such as obtains in the 

 Penguins. In some, as in Priocella, this is feebly developed, but 

 it is never conspicuous. In the Diomedeidae the case is otherwise. 

 In this group, and especially in Diomedea exalans, it may be as 

 well developed as in Catorr/tacte amougst the Periguius, and, as in 

 this genus and that of Pijgoscelis, the free edge of this ledge is 

 greatly flattened. 



The Base of the Skull. — The typical Procellarian form of the 

 basitemporal plate of the pai'asphenoid is triangular, with a free 

 anterior border. Prom this it follows that the Eustachian passages 

 are represented by grooves intead of tubes. In ProceUaria, 

 Oceanites, Cymodroma, Pelagodroma, Bidweria (occasionally), 

 Pirffiiius (P. l-uhli), and Pelecanoides a tube is more or less 

 perfectly formed, by the downgrowth of a thin plate of bone from 

 the alisphenoidal wings of the parasphenoid. In all the Procel- 

 lariidse except Pelecanoides there is a more or less conspicuous 

 aperture, receiving numerous pneumatic foramina, opening down- 

 wards immediately above the pneumatic grooves, to the inner side 

 and a little in front of the articular surface of the quadrate. This 

 aperture is in some cases of very considerable size, e. g. Fulmanis 

 glacialis i a probe passed down it, in a forward direction, leads 

 into the parasphenoidal rostrum. In the Diomedeidse this aperture 

 is smaller and opens directly backwards rather than downwards ; 

 furthermore, it is situated much nearer the middle line than in 

 the Procellariidae, inasmuch as it does not pass the level of a line 

 drawn through the mammillary processes, whilst in the latter, 

 as just stated, it opens near the quadrate articular surface. In 



