GILBERT AND HTARKS— FISHES OF PANAMA HAV 185 



Maiidiblu antLiiorly u iih ;ui DUlcr series of small canines, lichiiid which is a narrow band of 

 short cardiform teeth. The latter pass toward sides of jaw into a single series of canines, enlarging 

 toward angle of mouth and hooked backward and laterally. Prema.xiliarics with a single close-set 

 series of slender conical teeth, a median pair slightly larger than the others, but scarcely canine-like. 

 Vomer with a pair of widely separated canines, about equaling the large teeth in sides of mandible, and 

 much larger than any of the palatine teeth. The latter are of small size, nearly uniform, ten to 

 twelve in number. The eye is small, its diameter about equaling the bony interorbital width. 



The lines of sense organs and phosphorescent organs of this species have t>een described by 

 Dr. C. W. Greene (1899, p. 676), to whose paper we refer. As there stated, the rows agree very 

 closely with those in P. notatus and /*. margaritatus. The phosphorescent organs arc, however, 

 much smaller, proportionally, than in these species, and are le.ss developed in the dorsal region and 

 on to[) and sides of head. The following account gives the princii)al differences l)etween P. grecnci 

 and the other species mentioned. In some minor details, it differs from the account given by Dr. 

 Greene. 



Ill the (/('/'.frt/ series, no phosphorescent spots are present. In the lateral seucs, there is no 

 upper row of phosphorescent organs. In xhn filiiiral vow, both sets of organs are constantly present; 

 the sense organs are difficult to distinguish, as they are not accompanied by dermal filaments; the 

 line is discontinued at a point opposite the twenty-first anal ray. Concerning the phosphorescent 



organs in this series in /'. 7iotatiis, Greene says (1. c, p. 671): "The organs of this line have 



no relation to the body segments." We find that the typical arrangement, in the three species 

 known to us, gives two spots to each anal ray, along that [lart of the line which is parallel with the 

 anal tin. The spots are not equally spaced, but are rather ob\ iously arranged in pairs, though one 

 member of a pair is occasionally undeveloi^ed. Hotli caudal rows are well developed as in other 

 species. The ffastroffular row is always complete, but the sense organs are again difficult to detect, 

 owing to the absence of filaments. In the tipper opercular series, and in a row along the lower edge of 

 the black subocular blotch, the phosphorescent organs are well developed; they are absent, however, 

 or very sparingly developed on all other series on to[) and sides of head. 



The dorsal and anal tins are longer than in other species, the dorsal containing 2 spines 

 and 38 or 39 rays, the anal 35 or 36 rays. The membrane of the last dorsal spine joins base of first 

 soft ray. The last two or three dors;il rays are shortened, and the last ray is joined for its full length 

 with the upper ray of the caudal; there is thus a notch jjetween the two tins. The anal is continuous 

 witii tiie caudal, without notch, as the last rays are not shortened. 



As in other species, a gland is i)resent in the a.xil of the pectorals; also a .series of canals pene- 

 trating the base of the tin on the a.xillary side, as in Balrachoides, one for each interradial membrane. 

 The color is translucent grayish or olive, with five broad dark cross-bars on back along base 

 of dorsal fin, and a sixth on nape (including base of spinous dorsal); there are also one on occiput, 

 one behind orbits and one on snout, the last three less intense and more or less joined. The basal 

 portion of the pectoral fin is more or less dusky. Fins otherwise translucent, unmarked. 



Twelve specimens were obtaincil. The species is named for Dr. Charles 

 Wesley Greene. 



