36 SMITHSONIAN MISCELLANEOUS COLLECTIONS VOL. I42 



Halosauriformes. — The palatine-maxillary articulation and the 

 sensory canals of the head of halosaurids suggest a relationship with 

 Alhula and Pterothrissiis. The pelvic organization is typical of the 

 Clupei formes. 



Notacanthi formes. — Though highly specialized there seems to be no 

 reason to believe that this group is not most closely related to the 

 halosaurids, even though the pelvic structure of notacanthids is 

 unique. 



Beloniformes. — This group is so specialized that few inferences 

 concerning relationship can apparently be drawn from the structures 

 treated above. The pelvic osteology does suggest that of the iniomous 

 fishes. 



Gadiformes and Macruriformes. — The protrusile upper jaw of 

 these fishes seems to be essentially of perciform type. 



Gasterosteiformes. — Another group with a typical protrusile perci- 

 form upper jaw. 



Syngnathiformes. — This group seems to differ from the Gastero- 

 steiformes about as widely as possible in upper jaw structure. 



Lampridiformes. — Nothing to be added here. 



Cyprinodontiformes. — The protrusile upper jaw, where it occurs 

 in these fishes, is different from that of the percoids and, judging 

 from Amblyopsis and Chologaster which have fixed upper jaws, a 

 protrusile jaw mechanism has been independently developed within 

 the group. The caudal skeleton of the basal members Amblyopsis and 

 Chologaster is highly peculiar but bears some resemblance to that of 

 the Percopsi formes. 



Phallostethiformes. — A typical perciform upper jaw. 



Percopsiformes. — Though the caudal skeleton is specialized in a 

 direction peculiar to this order and apparently the Cyprinodonti- 

 formes, the pelvic structure and antorbital bone indicate a "lower" 

 teleostean condition. The ensemble of characters suggests that the 

 Percopsiformes may be an offshoot of a primitive scopeliform or 

 protoscopeliform stock. 



Stephanoheryciformes. — Nothing to be added here. 



Beryciformes. — In pelvic structure a scopeliform derivation is 

 suggested. The protrusile jaw seems to be typically perciform. So far 

 as the living teleosts are concerned a "true," percoid-type pelvic spine 

 seems to be present for the first time here. 



Zeiformes. — The protrusile upper jaw, pelvic structure, and caudal 

 skeleton all appear to be percoid. 



Mugiliformes, Polynemiformes, Ophiocephaliformes, and Sym- 

 branchiformes. — Nothing to be added. 



