shaped, areas. This upper series of ridges forms a line which extends from the nape, a little above the commencement of the lozenge- shaped, areas, and, passing backward, parallel to them and. about midway between them and. the dorsal margin, terminates immediately behind the anterior elongated, portion of the dorsal fin. This line of ridges resembles the upper line of mucus-tubes in Polypterus (' Poissons Fossiles,' t. ii. pt. 2. p. 50) and in Dapedius punctatus (ibid. t. ii. p. 192, pl. 25 a) ; and, indeed, in the latter, which in form closely resembles Dorypterus, this upper or second lateral line, according to Agassiz, holds relatively exactly the same position.
The marginal hour-glass-shaped plates have their sides abrupt and slightly elevated into ridges. A similar ridge passes along the centre ; and the most contracted part of the plate is thickened or elevated, the ends becoming depressed and thin. These peculiar plates, we have said, form a portion of a marginal series (fig. 1, k) that reaches to the root of the tail. Behind the anterior or elevated division of the dorsal fin they are much reduced in size, diminishing backwards in length in proportion to the reduced height of the fin, and are not connected with the sigmoidal extremities of the transverse series of plates and rods ; neither do their outer extremities appear to articulate with the fin-rays, though there are pretty regularly two rays to each plate. The largest of the hour-glass-shaped plates are 3/16 of an inch in length. These, in their arrangement and situation, resemble fin-supports.
A similar series of hour-glass -shaped plates extends along the ventral margin immediately within the base of the anal fin, and are large in front for some short distance backwards, corresponding to the space occupied by the enlarged anterior portion of the fin. These large plates and the large ones at the root of the dorsal seem to be articulated with the fin-rays.
The whole of the transverse plates, areas, and rods, as well as the ventral plates and columns and great posterior abdominal rods, seem as if covered with black enamel-like matter, having a semigloss similar in appearance to that which covers the head-bones and fin-rays. Indeed some of them seem as if composed of nothing else ; and such is the appearance of a few of the cranial bones themselves. The bony support, however, can be traced in some of them ; and a few of the lateral rods are hollow, the bony or cartilaginous support having apparently disappeared. But this enamel-like matter does not seem to have been confined to these parts ; it appears to have been continued as a thin film composed of granules between the series of plates, and was extended over the entire surface of the fish as a dermal envelope, the plates and rods being, as it were, immersed in it. Such is the appearance particularly in the region of the sigmoidal ridges, where there is a continuous darkish film of considerable thickness, having a granulated surface and giving the appearance to them of a series of broad continuous plates, which in all probability they are, the grooves in the ridges limiting the anterior and posterior margins of each plate. And there is usually, extending from the margin of the lateral plates, a broken fringe (figs. 2, 3, v) of
3 A 2
