NO. 146S. FISHES FROM ECUADOR AND PERU—STARKS. 785 



above the lower edge of the eye. Narrowest part of preorbital includ- 

 ino- e3^elid one-half of diameter of pupil; the bone only one-fourth of 

 pupil. Posterior edge of preopercle concave; the lower edge and the 

 angle ])roadly rounded. Gill rakers moderately slender; the longest 

 scarcel}' one-half the diameter of eye; 15 of them on lower part of 

 arch. 



Scales cycloid and regularly arranged; those of lateral line scarcely 

 enlarged but raised to a slight ridge, especially on caudal peduncle. 

 Thin scales present on cheeks and opercles; the rest of head naked. 



Spinous dorsal low; closing into a groove; the longest spine not 

 exceeding diameter of eye in length. Soft dorsal and anal highest in 

 front; the longest ra3^s equal to length of snout. Anal spines very 

 small and not separated from the soft rays; the first spine directly 

 under middle of soft dorsal. Soft dorsal and anal coterminous; the 

 distance from ba,se of dorsal to upper caudal rays If times the diameter 

 of eye. Pectoral reaching to above front of anal; its length a little 



,..^: 

 -U^.. 



Fig. 8.— Neptomenus crassus. 



less than that of head. Ventrals adnate to the belly; their tips reach- 

 ing halfway from their base to the middle of vent. Caudal deeply 

 forked. 



Color, dusky above; black on top of head; sides and lower parts 

 silvery. Sides of head, and particularly mandible, set with small points 

 of dark ])rown. Vertical fins dusky ; the dorsals darker than anal; pec- 

 toral slightly dusky; darker on inner surface; axil dark brown. 



This species agrees with K Imrma in number of fin rays and scales, 

 but if current descriptions of the latter are dependable it is a more 

 slender species, with a larger head, and with the eye above the middle 

 of the height of the head. 



Glinther describes the type as having the depth 3i in the total length, 

 and the head 4i. He evidently includes the caudal in his measure- 

 ments, as his specimen was U| inches in total length and U inches 

 deep. 



The type, however, was a stuffed specimen, and these measurements 

 could not be depended upon did not a description by Hutton (presum- 

 ably from fresh or alcoholic specimens) agree very well on these points 

 Proc. N. M. vol. XXX— 06 50 



