1904.1 OSTEOLOGY OF THE ELOPID^ AND ALBULIDiE. 75 



In Ghirocentrus, although the motxth is fairly large, the quad- 

 rate is advanced by a forward rotation of the hyomandibular. 

 The symplectic in this case is in a direct line with the axis of 

 the hyomandibular. In Arapaima the mouth can hardly be 

 described as small, yet the forward slope of the hyomandibular is 

 excessive. The explanation of this is most probably to be sought 

 in the great length of the postorbital portion of the head. 



In Engrcmlis and Coilia alone of the forms under consideration 

 does the quadrate slope backward. 



Opercular Series.— The reasons for excluding the preopercular 

 and interopercular bones from this series are given on p. 68. 

 The bones considered under the present heading are the opercular 

 and subopercular bones, the branchiostegal rays and the jugular 



plate. 



As is well known, the subopercular bone is wanting m JSoto- 

 pterus. This condition is unparalleled among the other forms 

 under consideration, although the subopercular is distinctly small 

 in Osteoglossum and Arapaima, and very small in Heterotis.^ It 

 is comparatively large in Alhula and Gonorhynchus ; it is said to 

 be wanting in Pantodon. 



The lower Malacopterygian fishes are well adapted for demon- 

 strating the continuity of the opercular and branchiostegal 

 systems. The opercular and branchiostegal bones are functionally 

 similar, serving to support the gill -cover, and there seems to be 

 wood reason for regarding thoBi as morphologically similar also. 

 The view is by no means a new one, for Traquair ("Ganoid 

 Fishes Brit. Carb. Form.," Paleeontogr. Soc.) demonstrated it in the 

 case of the Palfeoniscidaj in 1877, and Shufeldt (Rep. U.S. Com. 

 Fish. 1883 (85), pp. 818 & 820) mentioned it in 1883 ; and 

 although the latter does not in his paper give references to 

 previous expressions of such opinion, he does not himself claim 

 the idea as original. The only recent reference to such a view 

 appears to be that by Cole and Johnstone (Proc. & Trans. Liverp. 

 Biol. Soc. xvi. 1902, p. 175), who mention the opercular bones 

 (excepting the preopercular) as " modified branchiostegal rays." 



In Chanos, Alhula, and Elops the transition from the upper- 

 most branchiostegal rays to the subopercular and opercular bones 

 is very evenly graduated, and in Osteoglossum it becomes a matter 

 of some difficulty to decide whether the bone lying antero- 

 A^entrally to the opercular is a reduced and displaced subopercular, 

 or the uppermost branchiostegal ray which has lost its connection 

 with the epihyal. The difficulty of deciding whether a bone lying 

 below the opercular bone is a subopercular or a free branchi- 

 ostegal ray, the true subopercular being wanting, must frequently 

 have occurred to the systematic ichthyologist. Boulenger dis- 

 tinctly admits the difficulty when dealing with the Mormyroid 

 fishes (' Poissons du Bassin du Congo,' 1901, p. 50, footnote). In 

 preparing the skull of EngrauUs the subopercular comes away 

 readily with the epihyal, and is very liable to be mistaken for a 



