NO. 3 TELEOSTEAN FISHES — GOSLINE 31 



antorbital and the one remaining supraorbital bones on their way out. 

 Before disappearing, however, there is a transformation of function 

 of these bones into a pumping mechanism for the supraorbital di- 

 verticulum of the nasal capsule. This new function has probably 

 prolonged the evolutionary life of the antorbital and the supraorbital. 



The higher teleosts usually have nasal diverticula (cf., Eaton, 

 1956) ; however, the supraorbital diverticulum has disappeared. Per- 

 haps it is no coincidence that the supraorbital sac and supraorbital 

 bones drop out just below the level of protrusile premaxillary de- 

 velopment (see section IV of this paper), for a protrusile upper jaw 

 would certainly affect the arrangements of any pumping systems in 

 the snout region. 



From a phylogenetic standpoint, the antorbital-supraorbital pump- 

 ing system would seem to be the sort of mechanism that would only 

 have evolved once, for it is made up of rather heterogeneous parts. 

 It seems improbable that the two elements in this link were originally 

 more than casually in contact with one another. If, then, this ant- 

 orbital-supraorbital link evolved only once, it is evidence that at least 

 the fishes that possess it are of monophyletic origin. 



IV. UPPER JAW PROTRUSION IN TELEOSTEAN FISHES 



The freeing of the maxillary from the cheek is generally considered 

 a milestone in the evolution of the actinopterygian fishes. The de- 

 velopment of a protrusile upper jaw in the teleosts is a further, 

 though probably less fundamental, step in the same general direction. 



Probably the principal method of feeding in fishes is to suck the 

 food in with the surrounding water by expansion of the oral and gill 

 cavities. If some method is evolved of shooting the mouth opening 

 forward at the same time the suction is developed, the chances of 

 catching any moving prey should be increased. This, in its barest 

 terms, would seem to be the basic advantage of a protrusile upper 

 jaw. 



There are, of course, a number of supplementary considerations. 

 For example, fishes have on the one hand found it possible to adapt 

 such a protrusile jaw for other types of feeding, e.g., the nipping 

 structure of the parrot fishes. On the other, there are some types of 

 feeding in which a protrusile upper jaw may be disadvantageous. 

 Thus, many fishes feeding on large prey have redeveloped fixed pre- 

 maxillaries, e.g., barracudas, gempylids, tunas, most carangids. Fur- 

 thermore, even those fishes in which the pipette system of feeding is 

 developed to its greatest extent, i.e., the pipefishes and sea horses, do 



