628 PROCEEDINGS OF THE NATIONAL MUSEUM vol. m 



The six or seven genera discussed above, which evidently are of 

 close relationship, are placed in three families and the content of two of 

 these families differs in the two classifications. If this process of 

 splitting the families is carried to its logical conclusion, four families 

 should be recognized for the following four diverging groups of genera: 

 Percophis, Bembrops and Chrionema, Acanthaphritis and Pterosaron, 

 and Hemerocoetes. Such excessive splintering of the families seems out 

 of line with family divisions in other groups of fishes and in zoology in 

 general and serves no useful purpose. Therefore, it would seem best to 

 place all these genera in one family. At any rate, the terminology 

 used in this paper is in line with this viewpoint. Perhaps other related 

 genera should be included in the Percophididae, 



Key to five Western Atlantic percophidid species 



la. Maxillary without a tentacle. Anterior part of space between mandibular 

 rami scaled. Caudal without a definite spot behind its base. Gill rakers 

 7-9 + 20-21, those at the ends slender, moderately developed. Pectoral rays 

 21-22. Dorsal rays usually 15, sometimes 16. Scales 55-61. 



Chriomystax squamentum (p. 629) 



1&. Maxillary with a broad tapering tentacle. Space between mandibular rami 



scaleless. The smaller specimens with a caudal spot disappearing with 



growth (unknown for macromma). Gill rakers 4-6+14-16, including 1-4 



at both ends of arch in form of low tubercles. 



2a. Scales 47. Pectoral rays 24. Dorsal rays 15. Eye larger than snout. 



Bembrops magnisquamis (p. 633) 



26. Scales 55-59. Pectoral rays 25. Dorsal rays 14. Eye subequal to snout, 



varying slightly both ways (well divergent in this character from anati- 



rostris and approaching gobioides, see table 4). Without a prolonged 



dorsal spine in three males examined (185-202 mm.). A lateral series 



of four spots Bembrops macromma (p. 634) 



2c. Scales 61-71. Pectoral rays 26-28. Eye smaller than snout. 



3a. Dorsal rays usually 15, sometimes 14. The larger males having the sec- 

 ond dorsal spine filamentous. With a lateral row of about 10 spots 

 better marked in the smaller specimens. 



Bembrops anatirostris (p. 635) 

 3b. Dorsal rays usually 17, sometimes 18. Without a filamentous dorsal 

 spine. Without a definite lateral row of spots. 



Bembrops gobioides (p. 637) 



Chriomystax, new genus 



Genotype : Chriomystax squamentum, new species. 



Comparison: Chriomystax agrees with Chrionema Gilbert (genotype 

 Chrionema chryseres Gilbert (1905, p. 645) by original designation) in 

 lacking a maxillary tentacle, the presence of which is a constant 

 feature of Bembrops. Chriomystax differs from Chrionema in having 

 the ventral aspect of the head, that is, the mandibular rami and the 

 space between them, scaled instead of naked. (In the species of 



