30 SMITHSONIAN MISCELLANEOUS COLLECTIONS VOL. I42 



With regard to the orders above the isospondylous fishes, I have 

 been able to discover a supraorbital and/or antorbital in only four : 

 the iniomous, ostariophysine, heteromous, and salmopercoid fishes.^ 

 In no members of these examined was there a well-developed supra- 

 orbital nasal sac. 



In the iniomous genus Aidopus there is a well-developed supra- 

 orbital bone. It is, however, rigidly attached to the frontal and seems 

 to function as a protecting bone for the upper portion of the eye. The 

 nasal capsule extends back very slightly below its anterior margin. 

 The antorbital forms a long, somewhat curved strut running forward 

 from the front of the supraorbital bone below the nostrils. In Chlor- 

 opthalmus both supraorbital and antorbital bones are again present. 

 The supraorbital has a somewhat more anterior position than in 

 Aulopus. It is less rigidly attached to the frontals, and dips down 

 below the rear border of the nasal capsule. The antorbital is a small 

 ossicle lying above the forward portion of the lacrimal. Neoscopelus 

 \\2£, a small antorbital but apparently no supraorbital. In the more 

 specialized iniomous fishes examined, e.g., Parasudis, neither an- 

 torbital nor supraorbital are present, at least as separate elements. 



Among the ostariophysine fishes, the antorbital and a somewhat 

 movable supraorbital are present in Brycon. The latter bone extends 

 slightly over the nasal cavity. In Astyanax there is a small antorbital, 

 but I have been unable to find any supraorbital. In the cyprinids, by 

 contrast, there is usually a supraorbital, but the antorbital generally 

 disappears. Presumably it fuses with the lacrimal. 



Among the heteromous fishes there is a very slender antorbital in 

 "Halosauropsis" and a somewhat stronger one that bears the anterior 

 end of the infraorbital canal in Notacanthus. In neither genus is there 

 any trace of a supraorbital. 



Finally in Percopsis the infraorbital channel runs forward through 

 the lacrimal, then turns upward through a terminal hoop in that bone 

 and doubles back to end over an ossicle which is undoubtedly an 

 antorbital. The whole arrangement is somewhat reminiscent of Al- 

 bida. In Aphredodorus there is no antorbital; the infraorbital canal 

 runs straight forward in the lacrimal and ends there without doubling 

 upward and backward. 



In general, it may be said that lower teleostean evolution shows the 



2 In a stained and cleared specimen of the eel Ariosoma there is a minute 

 ossicle above and in front of the lacrimal that may represent the antorbital. Since 

 the above was written an antorbital has also been located in the holocentrid 

 fishes. 



