Page:Quarterly Journal of the Geological Society of London, vol. 29.djvu/595

1873.] reptilian one is grafted, shows profitable comparisons to be within the limits of the feathered class. The inference which has been drawn as to the length of the beak leads us first to compare Odontopteryx with those birds in which that part also exceeds in length the rest of the skull, which latter portion, bounded in front by the fronto-nasal suture, I shall speak of in the ensuing comparisons as the "cranium."

Such character is exceptional in the Aves aereæ and Aves terrestres of Nitzsch. The Hornbills, Toucans, a few Crows, certain Woodpeckers, Kingfishers, Cuckoos, Humming-birds, Kivis, Ostriches, manifest it, but with well-marked differential characters pointing to another road, for the closer affinity of which we are in quest.

A beak longer than the cranium is the rule in the Aves aquaticæ; but not any of the waders has the external nostrils so remote from the orbits as in Odontopteryx. This character of the fossil confines one to the Totipalmates and tubinarial Longipennates; but the Petrels, like the Albatrosses, Gulls, Terns, and Skimmers have other well-marked characters which remove them from the present extinct genus.

Indeed, the absence of the superorbital gland-pit in Odontopteryx limits the field of comparison to the Totipalmates and Lamellirostrals, in which, however, the Swan (Cygnus olor) and some Geese (Cereopsis) and Teal show traces, more or less definite, of the impression of such gland above and behind the rim of the orbit. There is no such trace in the Cormorants, Anhingas, and Gannets; and it is in these fish-eating sea-birds that an extent of upper beak-bone, free from narial vacuities, would be found corresponding with that which is preserved in the Sheppey fossil. But the Totipalmates have not the orbit bounded by a hind wall as in Odontopteryx; the superorbital border is abruptly truncate behind by a wide and deep crotaphyte fossa, which in the Cormorant and Gannet ascends so as almost to meet its fellow upon the parietal region of the cranium.

In Odontopteryx, the parietal region is broadly and smoothly arched (Pl. XVI. fig. 4, ); and the crotaphyte fossa (Pl. XVI. figs. 1, 2, ), very shallow, commences low down at the side of the arch (fig. 1, ), very little above the level of the foramen magnum. Now this is the character of the fossa in certain Anatidæ, the Goose (Anser palustris) e.g.; and in this family, also, the orbital wall is continued down the back part of the cavity as in Odontopteryx, but is there produced forward as a strong process, which seems not to have existed in the fossil. The hinder half, however, of the external nostril would have appeared in the base of the beak preserved in the fossil, if the bird it represents had partaken of the narial characters of the Lamellirostrals.

In most of these water-birds the coronoid border of the mandible is raised into a definable process; and where, as in Mergus, this is not the case, the outstanding tubercle is present, of which there is no trace in Odontopteryx, as there is none in the Totipalmates.

The hind half of the mandibular ramus resembles in its depth and thinness that part of the lower jaw of the Lamellirostrals more