L24 WBALDEN AND PURBEOK FOSSIL FISHES. 



a thickened upper border. In three specimens the ceratohyal is shown as a simple 

 hour-glass-shaped bono, without any ossified filament connecting its ends. The 

 preoperculum, as usual, is sharply bent at its angle, with a relatively large lower 

 limb j and there seem to be two or three ridges radiating backwards from the 

 slime-canal on the slight expansion at the angle. There are not less than 10 

 branchiostegal rays, of which the foremost are relatively small, slender, and spaced. 



The exact number of the vertebras is uncertain, but there are about 22 in the 

 abdominal region and at least 20 in the tail. About four are comprised within 

 the branchial region, and there are five or six in the upturned end which supports 

 the caudal fin. As shown both by the type specimen and by others, the centra 

 are delicate constricted cylinders pierced by a considerable remnant of the noto- 

 chord, and their outer surface is smooth, not strengthened by longitudinal ridges. 

 The neural arches are longest in the anterior half of the abdominal region, where 

 they support loosely-apposed stout neural spines as far back as the origin of 

 the dorsal fin (PI. XXIII, fig. 3). Beneath the dorsal fin the neural arches are 

 much shortened, without separate neural spines, and in the caudal region they 

 are also short, sharply inclined backwards, and symmetrical with the haemal arches. 

 The slender arched ribs extend nearly to the ventral border, and the expanded 

 haemal arches supporting the caudal fin are all separate (PL XXIII, fig. 5). 

 Intermuscular bones, very delicate, are seen only lying across the neural spines in 

 the abdominal region. 



As shown in both the specimens figured by Brodie (re- figured in PI. XXIII, 

 figs. 3, 4), the pelvic fins are not much less elevated than the pectorals. They 

 comprise 7 or 8 rays, and seem to be inserted opposite the front half or the 

 middle of the dorsal, their position varying in the fossils according to the manner 

 in which they are crushed. The dorsal fin is especially well seen in the type 

 specimen, where its maximum depth somewhat exceeds that of the trunk (fig. 3). 

 Its foremost three rays are short and simple, gradually increasing in length ; next 

 there are 1 or 1 1 rays which bifurcate, and are distantly articulated in their 

 distal half and slightly diminish in size backwards. Of the fin-supports the 

 foremost two or three are fused together at an acute angle, while above the others 

 there are distinct traces of the short intercalary pieces which are well known in 

 Amia and several fossil Amioids. The remote anal fin, best seen in PI. XXIII, 

 fig. 6, is small, and comprises only about 7 rays, rapidly shortening backwards. 

 Its rays also bifurcate and are distantly articulated in their distal half. The forked 

 caudal fin is especially powerful (PI. XXIII, figs. 5, 6), with distantly-articulated 

 bifurcating rays, of which the five or six shortest in the middle of the fork are spaced. 

 Stout sigmoidally-bent fulcral rays are seen at its base both above and below (fig. 6). 



The thin cycloidal scales are often seen in the fossils, marked only by the 

 concentric lines of growth ; but no enlarged or thickened ridge-scales have been 

 observed. 



