panded surface so overhung the fossil is remarkable. The albatross, again, most nearly resembles it in this character, but without the overhanging production of the head (compare figs. 1 & 5). The smooth outer crust of this surface is impressed in the fossils with minute points, and shows very faint longitudinal striæ.
There is a trace of the foramen pneumaticum (fig. 2, f) at the bottom of the fossa beneath the ulnar tuberosity, the walls of the fossa being broken away.
The thinness of the compact outer wall of the shaft (with the concomitant wide air-cavity), and the size and conformation of this part of the wing-bone, bespeak a bird of flight, and a species as large, at least, as any of the existing birds which enjoy the characteristic locomotion of the class.
The portion of the right humerus includes, with the head (fig. 3) and indications of proximal structures above described, five inches of the shaft. From a preserved breadth of 2 inches it diminishes to that of 10 lines at the fractured end. The smooth ridge noted in the left humerus (fig. 2, h) is continued distad for 212 inches before subsiding.
With the head of the left humerus were brought three portions of a long limb-bone with the pneumatic structure of the humerus in longipennate birds of flight (Pl. VI. figs. 9 & 10), and corresponding in size with the portion of the shaft of the right humerus, but less crushed and distorted.
The first of these portions I infer to be from the proximal half of the shaft; it is three-sided, with the angles broadly rounded off, but least so at that angle which shows the base of an outstanding ridge. An inch and a half of this base or origin is preserved (figs. 7 & 8, i), continued from the proximal fractured end of the bone. I infer it to be part of the origin of the "pectoral crest." Anconad of this origin, and about 3 lines distant therefrom, is a linear ridge (fig. 7, k) running longitudinally parallel therewith, and bending slightly more anconad before terminating at the distal fractured end of the portion. This ridge denotes the insertion of the ancono-deltoideus, or "posterior deltoid muscle."
Now, as to the trihedral shape of this proximal portion of the humerus, which may include one fourth or one third of the shaft, such shape is not shown by the humerus in the majority of birds: I find it in certain Raptores, Longipennates, and Totipalmates. In all that do show it one of the sides is palmad, the other two sides are anconad (as in figs. 7, 13). In the eagles and vultures (Vultur monachus, monogr. cit. pl. iii. fig. 7) the angle dividing the radial from the ulnar sides of the anconal surface is sharper than in Argillornis.
Pelecanus and Diomedea come nearest to the fossil in the obtuseness or degree of rounding-off of that angle, but fall short of that degree shown in the fossil. In all the characters above compared, the portion of shaft (fig. 7), like the head of the bone (Pl. VI. fig. 1), is from a left humerus.
The next character which I find available in steering toward
