FISHES OF NEW YORK 355 



not compressed to an edge; mouth small, the gape curved, very 

 oblique, usually not reaching the eye; lower jaw short and weak; 

 maxillary slipping entirely under preorbital; jaws each with a 

 band of simple, usually villiform teeth; premaxillaries very 

 freely protractile, their spines comparatively long, nearly equal 

 to the eye, extending backward beneath a fold of skin, which 

 connects the basis of the maxillaries; posterior part of the pre- 

 maxillaries broad; no teeth on vomer or palatines; both dorsals 

 short, the usual radial formula being D. V-1, 8, first dorsal 

 usually, but not always in front of anal; soft dorsal and anal 

 scaleless; scales rather large, entire. 



178 Menidia gracilis (Gunther) 



Slender Silversides 



AtherinichtJtys (gracilis Gunther, Cat. Fish. Brit. Mns. Ill, 405, 1861. 

 Menidia gracilis Jordan & Evermann, Bull. 47, U. S. Nat. Mus. 797, 1896, 

 Specimens from St George's Island, Potomac River. 



The origin of the anterior dorsal fin is opposite to the vent, 

 exactly in the middle of the distance between the end of the 

 snout and the base of the caudal. The distance between the 

 origins of the two dorsal fins is somewhat more than one half 

 of that between the origin of the posterior and the caudal. 

 The hight of the body is contained nine times in the total length, 

 the length of the head five and one half times. The silvery baud 

 is narrow, and occupies a part of the fourth series of scales. 

 Scales with the margin entire. Caudal lobes equal in length; 

 caudal somewhat longer than the pectoral, and rather shorter 

 than the head. 31 lines long. Probably young. Habitat 

 unknown. D. IV, I, 8; A. I, 19. Scales 9-40. (After Giinther) 



The specimens above referred to, from St George's island, 

 lower Potomac river, were obtained by Dr Hugh M. Smith, of 

 the U. S. Fish Commission, in the summer of 1890. 



The specimens were compared with the published descriptions 

 of M. b e r y 1 1 i n a (Cope) and were found to differ in some 

 minor details, the dorsal formula being V, I, 10 instead of V, I, 

 11, the anal rays averaging I, 16 or I, 17 instead of I, 18, and the 

 silvery stripe apparently taking a different course. 



