136 



TRANSACTIONS OF THE AMERICAN PHILOSOPHICAL SOCIETY 



Pholidophoridae, at the summit stand all the Jurassic and Cretaceous Leptolepidje, while 

 the Jurassic and Cretaceous Oligopleuridae are a side branch. 



The skull of Pholidophorus (Fig. 29) agrees with that of the holostean Macrosemius in 

 the marked forward inclination of the suspensorium of the lower jaw, but the mouth is 

 directed partly upward instead of forward, the jaws are longer, teeth more numerous and 



Pholidophorus 



Fig. 29. Pholidophorus macrocephalus. After Zittel, from Smith Woodward. 



less styloid, the eye much larger; the body as a whole is "elegantly fusifprm," while that 

 of Macrosemius is more robust and deep. All these features contribute to the herring-like 

 appearance of Pholidophorus and mark its contrast with the far more primitive Macro- 

 semius. A further conspicuous agreement with the primitive isospondylous teleosts is seen 

 in the small size of the premaxilla and the convex oral margin of the maxilla, which is 

 loosely attached and bears typically two well developed supramaxillary plates (A. S. 

 Woodward, 1895, p. 446) as in the Elopidse, Albulidse, Clupeidae, etc. The supramaxillae 

 appear to support Tate Regan's suggestion that these elements in teleosts originally belonged 

 to the suborbital series, from which they were abstracted by the maxillas. Inspection of 

 A. S. Woodward's figure of Pholidophorus limbatus (1895, Pt. Ill, PI. XII, Fig. 7) suggests 

 that the maxilla was already free at its posterior end and that the three supramaxills 

 were derived from the second or postorbital row of plates rather than from the inner circum- 

 orbital series. The skull top appears to be primitive. According to Watson (1925, p. 866) 

 there were a pair of large tabular bones. The supraoccipital had not yet begun to move 

 forward between the parietals. 



Classification of the Isospondyli. — The existing families of isospondylous fishes have 

 branched in many directions and the problem of their phylogenetic relationships with 

 each other and with their Cretaceous forerunners, although essayed by many authors, is 

 still only partly solved. 



