than it does the same part in the Totipalmates, where it is thicker and shallower.
The outer surface of the dentary is divided into an upper and lower tract in Swans and Geese by a groove which, beginning near the trace of the suture with the angular and surangular elements, curves feebly downwards as it advances forward: Cygnus Ruppelii, in this character, nearly repeats that in Odontopteryx.
The upper beak-bone in Anatidæ does not show the longitudinal groove which impresses it in Odontopteryx. But this groove is present in Sula and Phalacrocorax. It commences behind, in these sea-birds, a little in advance of the outer end of the naso-frontal suture, and extends straight forward, about midway between the upper and lower borders of the upper beak, to near its pointed termination. The groove (P1. XVI. figs. 1 & 2, g) has the same relative position on the sides of the upper beak in Odontopteryx; but it begins below the fore part of the zygoma, and rises with a curve convex upward, to midway between the upper and lower borders of the maxilla, along which it then runs straight as far as that bone is preserved.
The upper part of the upper beak in Sula is broad and arched at its base, the transverse convexity being more marked as the beak narrows and advances. In Odontopteryx an upper tract is pinched off, so to speak, from the sides, flattened above at first, and becoming transversely convex as it narrows and advances, the sides of the beak below this tract being transversely concave in a feeble degree before attaining the groove (g). This upper median raised tract recalls the more strongly developed one in Procellaria, and suggests the possibility of its having been prolonged, in Odontopteryx, to terminate forward, as in Petrels, in the outer opening of the tubular nostrils; but the mutilation of the beak in the fossil leaves this point purely conjectural; and in all other comparable characters of the skull the resemblances are found with the Lamellirostrals and Totipalmates, not with the Longipennate sea-birds.
Another character approximates the fossil to Sula; there is no trace of a mid notch at the fore part of the frontal, into which, in Anser palustris, the end of the nasal branch of the premaxillary is produced; the transverse fronto-nasal suture abruptly defines the cranium from the beak in Odontopteryx, as in the Totipalmates. But the transverse contraction of the interorbital part of the frontal is more considerable in the fossil, and the hind part of the naso-premaxillary tract is flatter, with other differences from the Gannets and Cormorants already noticed.
Thus Odontopteryx, independently of its teeth, shows, in the unique fossil representing the genus, its distinctness from all known existing genera of birds.
Of the species which have the bill armed with tooth-like processes, the enumeration is easy. The true Falcons have the single "tooth" on each side of the upper jaw; a like armature of the beak of the Butcher birds has suggested the term "dentirostres" for the tribe of Passerines including the Laniidoæ. The male of one genus
