72 FOSSIL FISHES OF THE ENGLISH CHALK. 



At least fifty vertebrae are indicated in the type specimen (figs. 1, ~\a) and in 

 another coiled-np example from the zone of Micraster coranguinum at Charlton, Kent 

 (B. M. no. 41198). Both these, however, are incomplete at the caudal extremity; 

 and a third fossil (B. M. no. 43512), with remains of this extremity, suggests that 

 the total number of vertebra? was about seventy. As shown by many specimens, 

 the centra are delicate, smooth, constricted cylinders, all originally pierced by a 

 thin persistent strand of the notochord. Those of the abdominal region, from 

 35 to 40 in number, are almost always exposed from above or below, on 

 account of the great development of their transverse processes ; while the caudal 

 vertebrae are usually seen in side view. As shown in section in a continuous series 

 in B. M. no. 41198, the twelve foremost vertebral centra are remarkably elongated, 

 the length of each being about three times as great as its maximum diameter. In 

 these the transverse process is a delicate narrow ledge of bone, extending from end 

 to end of the centrum on each side, and widest in its anterior half (as indicated in 

 the type specimen, fig. la). Some delicate rod-like fragments probably represent 

 ribs. Beyond the twelfth vertebra the remainder of those of the abdominal region 

 are larger and stouter, with the centra only about twice as long as deep. The 

 peculiar transverse processes also increase in stoutness and prominence until, at 

 about the thirty-second vertebra, the length of the laterally directed anterior 

 process equals that of the centrum. Each of these processes is still laminar, but 

 strengthened by being bent upwards on its long axis. Some anterior caudal 

 vertebras are seen in the type specimen (fig. la), and still better in a fragment of a 

 larger individual (fig. 7), which has been unsatisfactorily figured by Agassiz (torn. 

 cit., pi. lxvia, tig. 5). In these vertebras the transverse process has already 

 disappeared or been displaced downwards to form the base of the haemal arch. 

 The neural arch (it.) forms a low straight ridge, extending the whole length of the 

 centrum, and marked with some very fine pittings. The short, slender, back- 

 wardly-inclined neural spine is fixed at its hinder end; while a small prominence 

 like ;i zygapophysis occurs at its anterior end. The haemal arch (It.) is also a low 

 ridge, but it extends only along the anterior two thirds of the centrum, while it is 

 abruptly truncated in front and behind. Its lower margin is somewhat excavated, 

 and its bunder produced angle bears a slender haemal spine. 



The post-temporals (fig. 2a, ptt.) are a pair of triangular plates, longer than 

 broad, Impinging on the supraoccipital. The external-lateral margin of each is 

 slightly thickened and produced at the hinder angle; the upper face of the bone is 

 smooth. The pectoral arch and fins arc only known by unsatisfactory fragments. 

 In one specimen in the Willett Collection (no. 113), from the Upper Chalk of 

 Brighton, the pectoral tin is shown to be rather large and inserted close to the 

 ventral border. The anterior pelvic fin-rays are remarkably stout (B. M. no. P. 

 L905a). A Pragmenl of the caudal tin in 15. M. no. 43512 shows that it was large 

 and powerful. The other tins are only known by scattered pieces of their rays. 



