296 On the Osteology of Caturus and Osteorachis. 



than one third that of the ceratohyal [c.hy.) , which shows no 

 evidence of twisting, is twice as deep behind as in front, and 

 is thickened at the anterior extremity for articulation with 

 tlie hypohyal. The liypohyal (PI. IX. fig. 2) is especially 

 robustj narrowed and turned inwards at its anterior end. 



Opercular, Branchiostegal, and Branchial Apparatus. 



The gill-covers form a complete series of plates, Tlie 

 preoperculura is large and much expanded at its angle, where 

 the exposed surface is rugose, and one specimen (PL IX. 

 fig. I, p.op.) exhibits a facette-like excavation of its lower 

 extremity, as if it articulated with the inferior prominence at 

 the hinder end of the mandible. The operculum, subjper- 

 culum, and interoperculum are large, and do not merit special 

 descrijjtion ; the suboperculum has a large ascending process 

 at its antero-superior angle (no. 29049). The branchiostegal 

 rays (PI. VIII. fig. 3, br.), slightly over twenty in number, 

 are all broad, but the uppermost especially so, and tiieir free 

 ends are sometimes {e. g. no. P. 6904) shown to be pectinated. 

 In advance of the branchiostegal rays there is a very large 

 gular plate (PI. VIII. fig. 3, gu.) extending between tlie 

 mandibular rami for more than half their length. A detached 

 example (PI. IX. fig. 3), wanting its hinder portion, exhibits 

 a slight longitudinal median keel in its anterior half. 



The brancliial arches are only known by fragments 

 (PI. VIII. fig. 5), which display the ordinary cliannelled bone, 

 mixed with small tooth-like gill-rakers (<7.^.) and slender 

 calcified gill-filaments {jiL). One specimen (PI. IX. fig. 1, 

 g.r^ shows that some of the gill-rakers at least were fixed on 

 the edge of little plates of bone, resembling those on the gill- 

 arches of the modern Amia. 



Axial Skeleton of Trunk. 



The examples of Caturus from the Oxford Clay of Peter- 

 borough do not afibrd any additional information as to the 

 axial skeleton of the trunk. The comjilete skeletons from 

 the Lithographic Stone of Bavaria and France liave already 

 })roved that the notochord was persistent and that the ossifica- 

 tions in its sheath were confined to separate hypocentra and 

 plcurocentra. A detached abdominal hypocentrum from 

 Peterborough, however, shown from three points of view in 

 PI. IX. figs. 5, 5 o, 5 h, is interesting for comparison, and 

 bears the characteristic small lateral processes for the support 

 of the ribs. Whether any vertebral element was fused with 

 the basioccipital remains unknown. 



