70 DR. W. G. RIDEWOOD ON THE CRANIAL [May 3, 



nearest approach to such a condition is found in Notopterus, in 

 which the two nasals are fairly large, and just meet in the middle 

 plane of the head, and are with some difficulty removable frora 

 the underlying mesethmoid and prefrontal bones. A very similar 

 condition obtains in Petrocephalus, in which the nasal bones are 

 exceptionally large. 



The postorbital cheek-plates are large in Arapaima, Osteo- 

 glossum, Hyodon, Elops, and Megalops. Both postorbital and 

 suborbital bones are large in Alhula. In Ghirocentrus the bone 

 lying antero-ventral to the orbit is large. There is one large 

 suborbital, extending also behind the eye, in Ghaiios, Engraulis, 

 and Clupea harengus, but in Clupea Jinta the postorbital part of 

 the cheek is protected by two large plates. In some forms, such 

 as Coilia, Albula, and Notopterus, in which the postorbital and 

 suborbital sensory canals are large, the bones have a scroll-like 

 form ; but in the other genera the sensory canal lies farther from the 

 orbital edge of the circumorbital bones (e. g., Clupea, Chatoessus), 

 or is more deeply embedded in the bone (e. g., Arapaima, Ghanos). 



In Notopter%is there are processes of the two hindermost of the 

 suborbital bones directed inwards below the eyeball towards the 

 ectopterygoid and entopterygoid, with which they are united by 

 lig-ament. There is also a process passing inwards from the 

 anterior end of the front suborbital bone, and entering into close 

 fibrous union with the under surface of the prefrontal. Boulenger 

 finds that in some famihes of the Perciform fishes the " subocular 

 shelf " is sufiiciently constant to be of taxonomic value (Ann. & 

 Mag. Nat. Hist. (7) xiii. 1904, p. 179). 



In this connection it is interesting to note that in Albula and 

 in Dnssumieria there is an outwardly-directed process of the 

 ectopterygoid, underlying the eyeball, which meets the edge of 

 the suborbital bones. TLiis process, like the subocular shelf, 

 serves the purpose either of supporting the eyeball, or of limiting 

 the lateral play of the series of suborbital bones. 



In Osteoglossu'rn there is a close fibrous connection between the 

 maxilla and the preorbital and suborbital bones, but this is not 

 the case in Heterotis and Arapaima, 



Maxillary Series. — The evidence of palseontology goes to show 

 that the most piimitive form of mouth is that bounded above by 

 a small premaxilla and a comparatively long maxilla ; at all 

 events, an enlarged premaxilla has not yet been noted in Iso- 

 spondylous fishes below the Cretaceous strata (Smith Woodward, 

 Vert. Palaeontology, 1898, p. 113). Most of the forms now under 

 examination have a premaxilla- maxillary gape ; but in Albida, 

 in which, according to Jordan and Gilbert (Bull. U.S. Nat. 

 Mus. No. 16, 1882, p. 258) and Smith Woodward (Brit. Mus. 

 Cat. Foss. Fish. iv. p. 59), the lateral margin of the upper jaw 

 is formed by the maxilla, the maxilla can be of service only 

 when the mouth is widely opened, and indeed, since it bears no 

 teeth, while the premaxilla does, it probably does not function in 



