NO. 13fi6. 



OSTEOLOGY OF BERYCOID FISHES— STARKS. 



603 



Fig. 4.— Hyoid of polymixia japonica. 6, barbel; 6r, branxh 



lOSTEGAL RAY; ch, CERATOHYAL; Ml, HYPOHYAL. 



REJECTED FA:MILIES. 



Sev(M"iil form.s huvo been oxuiiiined which have been plaood witli the 

 Bervcoids at (lirt'erent times, luit which are not related, or at lea.st not 

 closeh' enoui'h to be re- 

 tained in the <>roup. ""-^ tk 



The family Mullida> 

 has been associated with 

 the Pol ymixidfv because 

 ])oth families have bar- 

 bels at the anterior end 

 of the hyoids, thouoh 

 they differ in most other 

 respects. 



Though apparenth' 

 similar the l)ar])els are 

 not exacth" of the same 

 character. In the Poly- 

 mixiida? each barbel is 

 suspended from the lower edge of the h3"poh3'al, where three modified 

 branchiostegal ra^'S curve around its thickened base (as described else- 

 where in this paper). In the iVIullidti? the barbel is suspended from 

 the extreme tip of a slender, nearly straight ray of bone attached to 



the end of the ceratohyal, 

 ^^ which is exserted l)eyond 



the hypohyals. This l)one 

 forms an integral part of 

 the barbel, appearing as its 

 ossified base. 



The family Aphredoder- 

 idiv has little in common 

 with the Berycoid fishes, 

 except the increased num- 

 ber of ventral rays. The 

 orbitosphenoids are absent. 

 There are two interorbital 

 septa, which are widely 

 separated. The exoccipi- 

 tals are rather l)roadly connected, but the atlas is unmodified. The 

 myodome, basisphenoid, suborbital shelf, and supplemental bone to 

 the maxillary are all absent. The pelvic bones are forked." and only 



"It appears pro1)able that parts of some skeleton other than that of AphredoderuK 

 have l)eeonie mixed with Doctor Boulenger's speeimen. He states tliat the pelvie 

 bones are not forked, as described by Cope and Woodward, and that there is a snlior- 

 bital shelf. In my specimens the pelvic bones are distinctly and widely forked, and, 

 thonjrh I have carefully examined three specimens, I have found no jiroccss on the 

 inner edjre of the suborl)ital rino:. 



Fig. 5.— Hyoid of Mri,LOiDES .wriflamma. h barbel: h 



BRA.N(HIOSTE(iAL RAY; ch, CERATOHYAL; /(/(. HYP0:'YALS. 



