ART. 20 NEW AND IMPERFECTLY KNOWN FISHES GINSBURG 5 



lower eye somewhat behind upper. Anterior nostril with a compar- 

 atively short flap on hind edge; posterior nostril larger, its rim not 

 raised. Preanal spine not evident. Length of a chord subtending 

 the curve in the lateral Une 3.4 in straight part, to end of hypoural, 

 a vertical from a chord to apex of curve 2.7 in the chord. Specimen 

 faded and color can not be described. 



Measurements. — Total length, 321. Standard length, 264. Depth, 

 116 (44); head (to end of scaled part, not including the soft border), 

 74 (28); maxillary, 36 (13.6); interorbital, 5.8 (2.2); pectoral of eyed 

 side, 41 (15.5) ; pectoral of blind side, 32.5 (12.3); caudal, 56 (21.2) ; 

 ventral of eyed side, 25.8 (9.8) ; ventral of blind side, 24.3 (9.2) ; 

 depth of caudal peduncle, 27.5 (10.4) ; snout, 18 (6.8); eyeball, 12.8 

 (4.8); straight part of lateral line, 150 (56.8). 



Holotype.—U.S.IS^M. No. 34919. Latitude 10° 37' 40" N., longi- 

 tude 61° 42' 40" W. (off Trinidad, West Indies); February 3, 1884; 

 31 fathoms; beam trawl; Albatross. 



Remarks. — This species is evidently very close to Paralichthys squa- 

 milentus Jordan and Gilbert,^ but it differs in a number of characters. 

 The vertebrae in the type specimen are 10 + 26, while in one specimen 

 of squamilentus that has been dissected they are 10 + 28. The most 

 striking difference on superficial examination is the presence of numer- 

 ous accessory scales in tropicus, while of all the specimens of squami- 

 lentus now loiown but a single accessory scale was found on one after 

 a prolonged search with a binocular microscope. P. squamilentus has 

 a very deep body, in seven specimens measured, 96 to 120 mm in 

 total length, the depth varied 46.6 to 52.3 per cent of the length with- 

 out caudal, while in the type of tropicus it is 44 per cent. The num- 

 bers of fin rays in the dorsal and anal of the type fall just outside the 

 frequency distributions for these characters in 12 specimens of squa- 

 milentus. Of the characters enumerated, the depth, and the profuse- 

 ness of accessory scales in those species of Paralichthys in which they 

 are present, are dependent on age to a certain extent, and since the 

 largest known example of squamilentus is but 120 mm long, the avail- 

 able specimens of the latter species are consequently not fairly com- 

 parable with the larger specimen here described. Nevertheless, the 

 differences are too pronounced to be caused wholly by the variation 

 in the size of the specimens compared. When the sum total of dif- 

 ferences is considered it becomes evident that the specimen here de- 

 scribed represents a hitherto unknown species, although the precise 

 degree of divergence between squamilentus and tropicus remains to be 

 elaborated by frequency distribution studies of the meristic differen- 

 tiating characters, as it is, indeed, necessary to do in nearly all species 

 of Paralichthys. A consideration of the remote location of the two 

 species, taken by itself, presents the probability strongly in favor 



• Proc. U. S Nat. Mus., vol. 5, p. 303, 1S82. 



