GREGORY: FISH SKULLS 233 



Regan notes that it has been suggested by Jordan and Evermann that these fishes are re- 

 lated to the percoid families Percidae and Centrarchidse; Mr. Regan adds that this is not 

 confirmed by a study of the anatomy, and the evidence of the otoliths supports his opinion. 

 On the other hand, there is a strong resemblance of the otoliths to those of the percoid genus 

 Apogon, which differ from the remainder of the Percoids in the sulcus. . . ." Thus it 

 would seem that the otoliths, like the branchiostegal rays and other characters, lend support 

 to the conclusion that the Salmopercse are a distinct group intermediate between some 

 ancient soft-rayed forms and the primitive percoids. 



Berycoidei 



Many of the lower teleosts already noted have foreshadowed the perch type, the " ideally 

 perfect fish," in one or more characters: several of the anterior rays of the dorsal and anal 

 fins may have become more or less spiny, the air-bladder may have lost its duct to the 

 oesophagus, the pelvic fins have perhaps lost the mesocoracoid arch and have moved forward 

 beneath the pectoral fins, the body may have become more or less short and deep {Cteno- 

 thrissa), the scales ctenoid, the maxillae often have withdrawn from the gape before the 

 advancing premaxillae, the orbitosphenoid may have disappeared, the supraoccipital has 

 often gained contact with the frontals and pushed aside the parietals, and so forth. But 

 except perhaps in the case of Ctenothrissa, the few spiny-finned characters thus acquired 

 have been associated with predominant tendencies that led toward other goals and it has 

 gradually become evident that we were not dealing with true ancestors of the spiny-finned 

 fishes but only with originally soft-rayed stocks which had progressed in certain features 

 toward the spiny-finned stage of evolution, the parallelism being on a grand scale and af- 

 fecting many different lines; but until the true percomorphs themselves appeared in the 

 Upper Cretaceous it never resulted in the complete combination of spiny-finned characters. 



In the Cretaceous and recent berycoids, however, we find a group which has almost 

 attained the spiny-finned status and to which Tate Regan (1929, p. 319) has awarded the 

 palm of being an order Berycomorphi, "directly intermediate between the clupeoid Isospon- 

 dyli and the Percomorphi." 



The restoration of the Cretaceous Hoplopteryx lewesiensis by Smith Woodward (1901, 

 p. 398) shows a fish (Fig. 110) that in many features recalls the Cretaceous deep-bodied 

 clupeoid Ctenothrissa, but which in skull structure is almost a percomorph. The head is 

 short and deep in correlation with the deep compressed form of the body. The premaxillae 

 have prominent ascending processes and are evidently more or less protrusile; the long 

 alveolar process (which is provided with minute teeth) excludes the stout maxilla from the 

 upper border of the mouth. The maxilla is deepened posteriorly and bears a broad supra- 

 maxilla, in front of which is a small triangular anterior 'supramaxilla (A. S. Woodward, 

 1902, p. 18). As the suspensorium is nearly vertical and quite long, the articulation of the 

 mandible is brought almost directly beneath the posterior border of the orbit and at a low 

 level. As the snout is also short, the result is that the mouth opening slopes gently upward. 



The circumorbital bones have their orbital margins everted, their peripheral areas de- 

 pressed for the reception of large slime cavities connected with the lateral line system. 

 Similar raised ridges and depressed areas are present on the preopercular and on the lateral 

 surface of the mandible, as in many acanthopts. The relatively large size of the orbits 

 shows that even in Cretaceous times the eyes had already assumed the dominant role in 



