from the Shales of the Northumberland Coal-field, 365 



The remains in our possession of such fishes are divisible 

 into two species by the characters of the spines, scales, and 

 teeth. One of these is very much larger than the other. It 

 is the smaller of the two that appears to be identical with A, 

 Wardi. The larger species is probably the same as that of 

 which Sir P. Egerton had obtained the head and anterior 

 parts, and which is supposed by him to " have measm-ed two 

 feet six inches in length "*. A pectoral spine of this is stated 

 to have been three and a half inches long. 



The mandibular ramus of A. Wardi (PI. XV. %. 6) is 

 about one inch and a half long and a quarter of an inch 

 wide at the broadest part, which is near the proximal extre- 

 mity, whence it tapers gradually to the distal end, which 

 is rounded ; the proximal end turns upwards, and presents a 

 well-defined concave articular surface. The dentigerous bone 

 is very thin, and its walls are usually pressed close together ; 

 the outer wall is irregularly striated longitudinally, the inner 

 wall is smooth ; the lower margin is strengthened by a stout 

 styliform process, c, which is very liable to detach itself, when 

 it assumes the appearance of a cylindrical spine graduating to 

 a point in front ; it is united behind to the articular process, 

 and is probably nothing more than a prolongation of the 

 angular bone. 



This styliform process has been described as the entire 

 mandibular ramus in some of the Acanthodei, and is seen oc- 

 casionally attached to the head, — the dentigerous bone, with 

 the teeth, having been detached. In Sir P. Egerton's figure 

 oi A. TFar•c?^ these styliform bones, so denuded, are seen still 

 articulated to the head and thrown backwards. The teeth 

 are frequently found attached to the thin-walled dentigerous 

 bone, the styliform process having probably been left so at- 

 tached to the head. 



The teeth are never found separated from the bone. There 

 are five or six in each ramus, two of the larger being in the 

 centre, the smaller ones in front and behind ; they are com- 

 pressed in the direction of the jaw, and when seen in this po- 

 sition they have the shape of as many equilateral triangles 

 with the lateral margins a little hollowed towards the apices, 

 which are recurved ; they are expanded at the base, where 

 they become confluent, and are coarsely and irregularly stri- 

 ated from one extremity to the other ; and the surface being- 

 liable to erosion, the striation is frequently exaggerated. 



The upper jaw is coextensive with the mandible, and is ap- 

 parently formed of one piece. The teeth are like those of the 

 under jaw, and lock very accurately into them ; they are of 

 * Quart. Journ. Geol. Soc. vol, xxii. p. 470. 



