632 



PROCEEDINGS OF THE NATIONAL MUSEUM 



(here included in tiie count of the lower limb) forward. At the at- 

 tached end 1-4 tubercles occur that are similar to those of the upper 

 limb, and the outgrowth behind the tubercles varies in development 

 in a similar manner as the corresponding one on the upper limb. 

 As the number of tubercles is readily determinable and no sharp line 

 can be drawn between them and the gill rakers, all are included in one 

 count in this comparative study. 



In one small specimen each of anatirostris and gobioides, 58 and 43.5 

 mm. standard length, respectively, the gill rakers near both ends of 

 the arch are relatively better developed and sharply delimitable from 

 the two tubercles at both ends. Therefore, the relative development 

 of the gill rakers as compared with the tubercles appears to change 

 with growth, in these two species anyway. 



Development in the two available specimens of magnisquamis (p. 633) 

 is somewhat different than in macromma, anatirostris, and gobioides. 

 Although an adequate sample is not available to determine growth 



Table 3. — Frequency distribution of the number of gill rakers, including tubercles, 

 in five Western Atlantic percophidid species 



changes satisfactorily, it seems that in the latter thi-ee species the 

 anterior gill rakers are not in the form of tubercles in small fish, and 

 one at both ends changes to a tubercle with growth. 



ExTRALiMiTAL SPECIMENS EXAMINED: Couuts of four extralimital 

 specimens in the U. S. National Museum, examined for comparative 

 purposes, are as follows. Taken off Kagoshima, Japan, now labeled 

 as Bembrops maculicauda (59647, 134 mm.): D VI; 14. A 16. P 24. Sc 

 52. GR 4 + 13. Holotype of B.filijera Gilbert from off Hawaii (51613) : 

 D VI; 14. A 18. P 27. Sc 63. GR 4+15. Holotype of B.filijer Fowler 

 from east of Masbate Island, Philippines (98868) : D VI ; 14. A 16. P 28. 

 Sc 51. GR 5+14. Paratype of B. nematopierus Norman from off 

 Zanzibar (109489): D VI; 14. A 16. P 25. SC 47. GR 5+15. While it 

 is necessary to study adequate samples to determine the intraspecific 

 variability of these species, the combinations of the foregoing counts 



