OSTEOLOGY OF LARID^. 593 



lie rather in grooves along the edges of the orbital ridges than In true fossie, and con- 

 sequently project so as to form more of the roof of the orbit thin is afforded by its 

 bony pariot«s. Stercoranus is intermediate between the two, but much nearest to Larus. 

 In all three of the subfamilies, however, the internal or convex borders of these supra- 

 orbital fossa3 meet on the median line, where they are merely separated by a longitud- 

 inal ridge ; and the fossse are so deficient anteriorly that the ducts of the glands do 

 not pass through a foramen in the roof of the orbit as in some genera of brachypterous 

 JS'atatores, but pass directly beneath the anterior orbital processes. 



The iorehead is rather broad and flat, with a slight concavity both longitudinally 

 and transversely. It is bounded posteriorly by the broad and prominent anterior or- 

 bital processes ; anteriorly by the fronto-maxillary suture. This suture is only very 

 moderately moveable (very different in this respect fiom that of likyiiohojia), and is 

 somewhat concave from the projecting backward of the articulatioa of the nasal bones 

 behind that of the intermaxillary. In the most adult skulls the nasal bones may be 

 distinctly traced, lying on either side of and closely apposed to the intermaxillary bone 

 at its base, running forward for a short distance, and then becoming perfectly anchy- 

 losed with the latter. Just external to the nasal bones, on either side, is the "nasal 

 process of the superior maxillary," running upward and backward from the superior 

 maxillary, to join the frontal bone at its edge just anterior to the orbital process. This 

 bone, which forms the posterior boundary of the anterior nasal fossaj, is usually called 

 the " maxill.iry or descending ramus of the nasal bone ;" and such is its true character, 

 as appears even in such a nearly allied genus as Pvffinus. But here this bone seems 

 quite disconnected with the nasal bone proper, and at the same time to be completely 

 consolidated with the superior maxillary. The separation is well marked in the three 

 subfamilies, especially in Sterna. 



The chief difference in the intermaxillary bones of the three subfamilies is, that in 

 Larina; and Lesiridinw the apex is decurved at a considerable angle over the apex of 

 the inferior maxilla ; while in Sierniiin: there is no such deflection. The coalescence of 

 the two mandibular with the superior mesial process is also here more extended than 

 in the other subfamilies, and consequently the bony margin of the nasal aperture is of 

 less extent. In all, the mesial process is thin and depressed, flat beneath, trans- 

 versely convex above, rising gradually to the fronto-maxillary suture ; and there is 

 but a slight angle between the planes of the frontal and intermaxillary bone, very 

 different from what obtains in Shytichops. This mesial process always retains more or 

 less distinctly marked the longitudinal line denoting its original separation into two 

 halves. The mandibular processes, as usual, are weaker and slenderer thau the mesial, 

 though not greatly so ; they diverge from it, and from each other, at a very slight angle. 

 They are horizontal and quite straight, and lie nearly in a direct line with the zygo- 

 matic bones. 



The orbits, as usual among the Natatorea, are large and irregularly shaped, being de- 

 ficient as to their bony parietes. Beside the difference in their roofs, already detailed, 

 each of the subfamilies varies as to the character of the ethmoidal plate which forms 

 the inter-orbital septum. In Lams* this plate is perfect, and the two orbits are com- 

 pletely separated by bone. In Stercoranus the moderate-sized oval foramen exists in 

 the centre of the plate, while in Sterna there can hardly be said to be a bony septum 

 at all, beyond a slight backward projection of the ethmoidal plate from the middle of 

 the anterior walls of the orbit. The septum between the orbits and the interior of the 

 cranium is also exceedingly defective in Sterna, all the anterior nerves escaping by one 

 large orifice, which extends from, the roof to the floor of the cavity. There is seen, 

 however, an indication of the more complete foramina which exist ih the other sub- 

 families, in processes which project inward from the edges of this large orifice. They 

 do not, however, meet each other, nor yet meet the process of the ethmoidal plate 

 which extends backward. Even the olfactory nerves have not distinct foramina of 

 egress, but escape by the common orifice. In all the subfamilies the course of the ol- 

 factory nerve is indicated by a distinct deep groove along the roof of the orbit. In 

 the skulls of Larus and Stercoranus before me, the foramen, or rather canal, which con- 

 ducts this nerve from the orbit into the nasal passages, is large and conspicuous ; in 

 Ste>-na it is much more contracted. The anterior parietes of the orbits, which separate 

 them from the nasal passages, is very deficient below, but superiorly it is more com- 

 plete, and there pierced by two foramina ; one, just described, for the olfactory nerve; 

 the other for the duct of the nasal gland. The extremity of this bony septum is acute, 

 and extends downward nearly to the level of the zygoma in Larus and in Sterna ; in 

 Stercorarius it is much shorter and more obtuse. 



The skull presents on its inferior aspect the shape of an almost perfect isosceles tri- 

 angle. The base of this triangle is a little wider in the Lestridiua- than in the other 

 two subfamilies, in consequence of the greater divergence of the mandibular processes 

 of the intermaxillary bone, and, beyond them, of the malar bones. These last form, as 

 usual, with the thin flat rami of the inferior mandible the lateral boundaries of this 



'Skulls may vary in this respect. I describe the three subfamilies from the samples 

 I have before me. 



38 



