The superior surface of this specimen does not present the complete pair of grooves which run along the superior surface in Belemnoziphius, enclosing between them a central vomerine area ; nor, like typical Choneziphius planirostris, does it present a bifid structure in the presence of a central groove stretching from the orifice of the vomerine canal ; on the contrary, the curious projection of the anterior end of the rostrum (which is paralleled in some specimens of Belemnoziphius) overhangs and conceals the vomerine canal, and the surface is perfectly smooth, neither indicating the junction of the intermaxillaries by a median fissure, as in Choneziphius planirostris, nor allowing the vomer to appear between those bones, as in Belemnoziphius.
The large unsymmetrical fossse excavated in the expanded portion of the intermaxillaries at the posterior part of the specimen lead into short grooves, which run forward on the surface and soon dwindle away (Plate XXXIII. fig. 2, l g and r g). These are identical with the grooves demarcating the vomerine tract in Belemnoziphius ; but here they terminate rapidly, as in Choneziphius, by the junction of the intermaxillaries in the middle line. They are present in Choneziphius planirostris, which has similar fossae to those seen in this specimen ; but in Cuvier's first specimen the grooves are for a short space converted into canals, whilst in another specimen of Choneziphius planirostris, of which there is a cast in the British Museum, the canals are as open as in this specimen.
Generic position and species. — The rostrum under description clearly does not belong to Professor Huxley's genus Belemnoziphius, which is remarkably well characterized not only by the solidity of the rostrum, the complete exposure of the central vomerine area (which Professor Owen, differing from Cuvier, Duvernoy, Gervais, and Van Beneden, terms prefrontal), but also by the two obvious perforations placed near the middle line in the intermaxillaries on the anterior wall of the nasal fossa marked a a by Professor Huxley in his figure of Belemnoziphius compressus.
The series of canals in Belemnoziphius and Choneziphius differ in a very marked manner ; but it is not possible to determine their exact relations without detailed comparison, and perhaps cutting specimens. Recently in Paris I had the opportunity, by the kindness of M. Gervais, of examining carefully the rostra of Cuvier's types of Z. planirostris and Z. longirostris. Z. longirostris (Belemnoziphius) comes very near to the Seychelles Ziphius (M. densirostris), and differs remarkably, as do all our Crag Belemnoziphii, from Choneziphius in the absence of the large unsymmetrical fossae seen in the latter genus, and in the presence of the sharply marked orifices leading into a canal (a a of Huxley's paper), which is absent in the other. Professor Owen marks these structures as identical in his recent monograph ; but they have a distinct character, though possibly related in origin. It is not possible to fully compare the canals of Belemnoziphius with those of Choneziphius, on account of the loss of the expanded portion in all specimens of the former genus. The specimen which best shows the proximal part of the rostrum of
vol. xxvi. — part i. 2 o
