1550 DuUctin //, Ujiitcd States National l\hisciim. 



of spinous dorsal, usually not deeply incised nor lobed, the soft part more 

 or less elevated, its last rays gradually shortened ; lower limb of preoperde 

 usually more or less scaly; preorbifcal narrow without deep notch; anal 

 fin similar to soft dorsal, with 2 spines, of which the second is much the 

 larger; soft rays 12 to 16; dorsal spines with a slieath of large scales, the 

 membranes of both dorsal and anal covered high uji with small scales; 

 caudal fin more or less forked, the lobes rounded; lower pharyngeals 

 triangular; branchiostegals 5 or 6. Species numerous, in the tropical 

 seas. Species chiefly American, extremely variable in form and color, the 

 brilliant coloration apparently dependent on surroundings. Thi^ species 

 are little known and the classification of those found in the West Indies 

 is, in default of material, largely guesswork. 



Eupomacenirus, as understood by us, differs from Pomacentrus chiefly in 

 having but 1 series of teeth in each jaw. In Fomacentrus an inner series 

 of a few teeth is present. Parapomacentrus and AniMypomacei)trtis differ 

 from Eupomacentrns chiefly in the different scaling of the head. This is a 

 minor character, and perhaps all 3 should be united under the oldest name, 

 Parapomacentrus. 



The following analysis of species has very little value. Perhaps all our 

 Atlantic species (omitting j;?aMi/ro«s) are reducible to 2,fuscu8 and leiteo- 

 siictus, and these 2, if really distinct, can not always be separated with cer- 

 tainty; leucoriis nml Jtavilatus are certainly distinct from fiisciis, and thus 

 fiir we have found rertifra'tinm also distiuguishabh^. Our species, so far as 

 known, all belong to the section Eupomacentrns, which is regarded by 

 Bleeker as a genus distinct from Parapomacentrus, being chiefly distin- 

 guished by the naked lower jaw. Perhaps the 2 genera should be united; 

 but not having seen specimens of Parapomacentrus, we hesitate to place 

 our species in it, though they evidently should not be left in Pomacentrus. 

 (ev, genuine; /'omaceH^rus, which is from 7f«>//a, opercle; Msvrpov, spine.) 

 a. Upper anterior profile of head arclied. 



b. Depth of body moderate, 1| to 2J in length, without caudal. 



c. Lower posterior half of body dark, like the anterior half; caudal fin mostly 

 dusky. 

 d. Pectoral flu with a broad white edge. leucoeus, 1960. 



(Id. Pectoral fin not edged with white. 



e. Depth of bodj' IJ in length of body (without caudal) ; sides with 

 faint cross streaks. auustus, 1901. 



ee. Depth of body 2 to 2i in length. 



/. Opercle without distiuct dark spot; caudal not tipped with 

 orange. 

 (/. Anal without distinct blue spot in its posterior axil, 

 except in young. 

 /(. Head with few, if any, small accessory scales. 

 i. Base of j)ectoral with one black spot. 



Fuscus, 1962. 

 ii. Base of pectoral with two black spots. 



DIENC^US, 1963. 

 hh. Head with very many small accessory scales, 

 especially on forehead and front of opercle. 



RECTI PR.«NUM, 1964. 



gg. Anal with a bluish spot at base of last ray; head and 



fins mucli spotted with blue. ANALIS, 1965. 



