CHEILODIPTERID FISHES SCHULTZ 417 



Ih. Anterior edge of pelvic spines smooth. 



Sff/. One obviously strong humeral spine ; margin of antei'ior ridge of 



preopercle smooth grossidens (Smith and Radcliffe) 



5b. No humeral spines or at most only a rudiment of one; anterior 



ridge of preoperculum with a few serrations at lower angle. 



6a. No keels on outer surface of posterior portion of preoperculum 



near lower angle; dorsal fins joined or practically so; Anal 



II, 9. 



7a.° About 40 scales in lateral line, depth 3% to SV2 in length ; eye 



3i'3 to 3% in head; gill rakers on lower half of first arch 



13 or 14. formula 5+1+14 microlepis Norman 



76. About 35 scales in lateral line ; depth 3.3 in length ; eye 2.9 in 

 head; gill rakers on lower half of first arch 17 (total 



rakers 25) pseudomicrolepis, new species 



66. Two or three small keels on outer surface of posterior portions of 

 preoperculum near lower angle; first and second dorsal fins 

 definitely separated; about 30 scales in lateral line; depth 3% 

 to more than 4 in length, eye 3 to 3% in head ; anal II, 7. 

 8a.^ Maxillary to below center of eye, 2Vi to 2% in head. 



bella (Goode and Bean) 

 8b.^ INIaxillary to below anterior margin of pupil, 2^^ in head, 

 japonica (Steindachner and Doderlein) 



SYNAGROPS SERRATOSPINOSA Smith and Radcliffe 



Si/nagrops serratospinosa* Smith and Radcliffe, in Radcliffe, Proc. U. S. Nat. 

 Mus., vol. 41, p. 444, pi. 38, fig. 2, 1912. (U.S.N.M. No. 70254, type.) 



SYNAGROPS SPINOSA, new species 



Ilolotype. — A male, 105 mm. in standard lengtli, taken at Albatross 

 station 2401, latitude 28° 38' 30" N., longitude 85° 52' 30" W., depth 

 142 fathoms, March 14, 1885 (Gulf of Mexico). This specimen was 

 one of the cotypes of Hypodydonia heUa Goode and Bean. It bears 

 U.S.N.M. No. 74324. 



Description. — Body compressed, elongate with large cycloid scales, 

 all except a few of which have been lost; lateral line continuous but 

 not extending on caudal fin rays, except one or two pores; interor- 

 bital space slightly convex; premaxillaries protractile, lower jaw 

 projecting but not entering profile, a small symphyseal knob on lower 

 jaw fits into a concavity at symphysis of upper jaw; each side of tip 

 of lower jaw is a concavity with narrow band of villiform teeth, 

 upper edge of maxillary fitting into a groove formed by the sub- 

 orbital and preorbital bones anteriorly, but mostly exposed pos- 

 teriorly; near symphysis of premaxillaries occur a pair of canine 

 teeth followed posteriorly by a band of villiform teeth ; at symphysis 

 of lower jaw is a pair of canine teeth that fit between those of upper 



s These portions of the key were taken from Norman, Discovery Reports, vol 12, p. 11, 

 1935. Our limited material does not indicate these differences when specimens of the same 

 size are compared, and I am unable to find any character that will separate our Japanese 

 and Atlantic material of heUa and japonica with certainty. 



