NO. 3582 GENUS ENTOMACRODUS — SPRINGER 147 



Entomaciodus macrospilus, new species 



Plate 30 



Description. — (Character for holotype in parentheses.) Segmented 

 dorsal fin rays 15 or 16 (16); segmented anal rays 17 or 18 (18); 

 posteriormost anal pterygiophore supporting 1 or 2 external elements 

 (2); total gill-rakers on first arch 13 or 14; pseudobranchial filaments 

 5; vertebrae 35 or 36 (35); 1 supraorbital cirrus on each eye; nape 

 without cirri; predorsal commissural pores 9-18 (13) (too few speci- 

 mens available to determine if pore number increases with increase in 

 size) ; preopercular series of pores with all positions with simple 

 pores; 1 pore before each anterior nostril; lateral line pores termi- 

 nating on side in area below and between dorsal spines 10 and 11 (10) ; 

 ventral margin of upper lip entire (without crenulations) . 



Proportional measurements (as percent SL, except as noted; only 

 the holotype was in sufficiently good condition to afford measurement; 

 for abbreviations, see "Methods"): SL 18.7, HL 24.6, OL 8.0, OCL 

 5.3, DS3 10.2, DRl 10.7, PECL 21.4, PELL 14.4, CL 18.2. 



No mature males were recognizable. The largest specimen was 

 a male, approximately 21.9 mm SL; the smallest, a male, approxi- 

 mately 14.0 mm SL, was not an ophioblennius stage. 



Color pattern. — The color pattern is best discerned from plate 

 30, the holotype. The paratypes are essentially similar except that 

 the smallest specimen lacked the large dark spot on the head. 



Relationships. — Entomacrodus macrospilus is a member of the E. 

 thalassinus species group. It differs from E. thalassinus , the only 

 other member, in having higher average dorsal ray, anal ray, and 

 vertebral counts; in having a single cirrus above each eye (apparently 

 constant with size), and in having on the head a large dark spot that 

 does not encroach on the circumorbital pores. 



The Marquesas Islands, only locality from where E. macrospilus is 

 known (fig. 10), are probably of more recent origin than the low 

 Tuamotus, nearest island group to the Marquesas. As E. thalassinus 

 is widely distributed in the central Pacific and occurs in the Tuamotus, 

 it seems possible that E. macrospilus is a derivative of E. thalassinus. 



Remarks. — Entomacrodus macrospilus has been collected with E. 

 corneliae and E. randalli, both endemic to the Marquesas. E. macro- 

 spilus differs most prominently from both these species in having only 

 simple pores in the preopercular series, no crenulae on the upper lip, 

 no nuchal cirri, and probably more predorsal conmiissural pores at 

 comparable sizes. 



Holotype.— USNM 200279, an immature male, 18.7 mm SL, from 

 east side of Anaho Bay, near chffs at shore, coll. J. E. Randall, July 17, 

 1957. 



