200 Mr. G. A. Boulenser on the 



o" 



popliyses first appcariiijT on the fifth, the praimaxiHarles are 

 feebly but distinctly protractile; the second suborbital emits 

 an internal lamina for the support of the eyeball, and this 

 lamina is triangular as in Trdvhichthys] the pelvic bones are 

 quite similar to those of the latter genus, not forked, as stated 

 by Cope and after him by Woodward*, and attached to the 

 clavicular symphysis. The forward position of the vent, so 

 exceptional a character among Acanthopterygians, is found 

 likewise, though to a somewhat less degree, in the Berycid 

 Trachichthys TrailUi. The Eocene freshwater genera 

 Ai)ij>/n'pla(ja, Tn'cho/>/ianes, and Asineops should, perhaps, 

 also enter the Berycidse, but from the descriptions and figures 

 given by Cojje f I have failed to grasp the near affinity which 

 is supposed to exist between them and the Aphredoderidaj. 

 As an example of the uncertainty in which we still are re- 

 specting the exact systematic position of these fossils, I would 

 point out that Fygaiis oi' Agassiz, which Cope was inclined to 

 regard as nearly allied to and possibly identical with Asineops, 

 is placed by Woodward among the Cluvitodontida^. 



One thing is certain, the Berycidi« are a very ancient and 

 generalized group of Acanthopterygians, and were richly 

 represented in the Upper Cretaceous by several genera 

 which are identical with or closely related to the existing 

 forms. In those days, however, the Serranidai had already 

 dawned {Prolates), and it is probable that the connexion 

 between the two families was as close as it is at present. No 

 better evidence of this near affinity can be adduced than a 

 comparison of Beryx proper with Pempheris, the two genera 

 agreeing so completely in structure, both external and in- 

 ternal I, with the sole exception of the rays in the ventral 

 fins, that 1 am much inclined to doubt whether the difference 

 between them should be regarded as greater than that between 

 the former and Trachichthys. The relation between the 

 Berycidaj and the Pempheridie has already been recognized, 



of Pirate Perches, wliicli fiuds its natural position between the Percopsidae 

 and the Percoid forms. Structure of mouth and skeleton, so far as 

 iinown, essentially that of the Percoid Hshes. Uorsal lin single, with 

 few small sjdnes; ventrals thoracic, with a small s})ine, and more tiiiin 

 five soft rays. Air-duct not examined, probably obsolete, the air-bladder 

 large and adlierent. Intestinal canal ending at the tliroat in the adult, 

 the vent vaiiously posterior in the young. Vertebrfe 2'J." 



* A cliaracter 'taken from the Eocene genus IJrismatoptcnis, wliich, 

 having the pelvis suspended from the postclavicular buues, probably bears 

 no nnar relation to the Bervcidae. 



t Vert. Tert. Form. West, i. p. 80 (1883). 



X The number of vertebrse assigned to Peinpheris by Jordan and Evor- 

 mann, viz. 10-}- 24, is obviously a misprint for IO-f-14. 



