THE SHORE FISHES OF PERU 239 



the fifth ray (counting downward) wsually longest, reaching to or a 

 little beyond origin of anal, longer than head, 2.7 to 3.0 in length. 



Color grayish above; pale below, with a slight silvery sheen; sides 

 of the larger specimens with about 8 irregular and indefinite dark 

 vertical bars, missing in the smallest specimen (48 mm. long); fins 

 mostly pale, membranes between spines of dorsal and anal dusky, the 

 tip of spinous dorsal of the smallest specimen conspicuously black. 

 Largely silvery in life. 



Four specimens from Peru, respectively 48, 115, 120, and 125 mm. 

 long, form the bases for the description. The tlu-ee larger specimens 

 were secured in the Rio de Eten, at Eten, by R. E. Coker, and the 

 smallest one was taken in Chimbote Bay by the Mission. The species 

 apparently is not numerous on the coast of Peru. 



Regan (1907, p. 38) claimed that Pacific coast specimens have 

 "fewer scales, larger head, longer maxillary, and longer second anal 

 spine, whilst the bars on the sides are usually more numerous." In 

 his key to the species, Regan gave as a distinguishing character 

 "5 or BYi scales between the lateral line and middle of dorsal fin" for 

 the Atlantic specimens and only 4 for the Pacific. Regan, then, named 

 the Pacific coast representatives G. similhmus. Meek and Hildebrand 

 (1925, p. 589), who had 35 specimens from the Atlantic coast of 

 Panama, and 5 from the Pacific, who had studied additional specimens 

 from Panama and elsewhere contained in theU. S. National Museum 

 collection, and who had especially checked the supposed differences 

 mentioned by Regan, stated, "We are unable to detect any difference, 

 whatever, between the Atlantic and Pacific representatives." Ac- 

 cordingly, they synonymized G. simillimus with G. cinereus. How- 

 ever, Jordan, Evermann, and Clark (1930, p. 342) did not see fit to 

 follow Meek and Hildebrand but gave G. simillimus full specific rank. 



I have now rechecked Meek and Hildebrand's original data and 

 have analyzed further data based on recently collected specimens 

 from the coasts of Panama, and the Peruvian material described 

 above, and have found no reason for altering the earlier conclusion. 

 In 26 specimens from the Atlantic, 23 have 4 major rows of scales 

 between the lateral line and the middle of the dorsal, 2 have only 3, 

 and 1 has 5. In 22 specimens from the Pacific there is no variation in 

 the number of major rows, which is constantly 4, though often an 

 additional "half row" is present. In a lateral series, counted just 

 above the lateral line, 21 specimens from the Atlantic have 40 to 44, 

 the average being 42.1. Similarly, 15 specimens from the Pacific 

 have 39 to 44, the average number being 42.0. Neither is it possible 

 to demonstrate a difference in the length of the head. In 25 specimens 

 from the Atlantic, 55 to 260 mm. long, the head is contained 2.75 to 3.2 

 times in the length, and in 21 specimens, 50 to 225 mm. long from the 

 Pacific, 2.9 to 3.3 times. The maxillary extends slightly beyond the 



