342 BULLETIN 189, UNITED STATES NATIONAL MUSEUM 



erations and proportions used in the description are based on 16 

 specimens, 70 to 265 mm. (50 to 208 m.m. to base of caudal) long. The 

 smaller specimens differ from the large ones, in addition to color 

 already mentioned, in having the fins less densely scaled, in having 

 the edges of the scales rough, and in having a proportionately larger 

 eye and narrower interorbital as usual among young fish. The 3 

 juveniles, respectively 18, 19, and 19 mm. long, described by Evermann 

 and Kadcliffe (see reference above) as having the preopercle "strongly 

 serrate" is scarcely correct, as the margin consists of a rough mem- 

 branous (not bony) edge. Neither is there any indication of a ser- 

 rated bony edge in specimens 70 to 73 mm. long, which are next small- 

 est in size among the specimens at hand. The condition as to the 

 adherence of the suborbital is not different in the young from that of 

 the adult. The number of dorsal spines, and the rows of scales are 

 remarkably constant among specimens. 



The Peruvian examples were compared with three large ones from 

 the Galapagos Islands, identified as Nexilosus albemarleus Heller and 

 Snodgrass, with which they quite certainly are identical. This con- 

 firms the view already reached by Evermann and Radcliffe (see 

 reference above) from the study of the description and figure of N. 

 albemarleus in connection with Peruvian specimens. 



The specimens collected by the Mission, taken mostly with trammel 

 nets set near rocks, are from Paita Harbor, and Bay ; Lobos de Tierra 

 Bay; Lobos de Afuera Bay; Guanape Island; Samanco Bay, off Mount 

 Campana; Pachacamac Island; and North Chincha Island. The 

 ones collected by R. E. Coker are from Lobos de Afuera, Guanape 

 Island, and from Mollendo. 



It is stated in the report of the Mission (1943, p. 278) that "casta- 

 neta" is the "commonest of the Peruvian rock fishes" yet of compara- 

 tively little economic importance, because it is not highly esteemed as 

 a food fish. The largest examples observed by the Mission were only 

 about 350 mm. long. 



Range. — Entire coast of Peru, and at least as far south as Iquique, 

 Chile. The species occurs also in the Galapagos Islands. 



Family LABRIDAE '' 



Body oblong or elongate, more or less compressed ; mouth moderate, 

 generally terminal; lips thick, the lower one usually with a lateral 

 flap; premaxillaries protractile; maxillary slipping under preorbital; 



" Jordan (1923, pp. 221, 222) divided the family Labridae, as formerly understood, into two families, 

 namely, Labridae and Coridae. The family Labridae was defined as including those fishes inhabiting 

 ' 'the colder seas" and having 27 or more vertebrae. The family Coridae was to include the "tropical forms" 

 having 25 or fewer vertebrae. As the vertebrae apparently have not been counted in many species, it is 

 not at all certain that the division based on the number of vertebrae is tenable. Furthermore, all the 

 species of the genera listed under Labridae by Jordan do not inhabit the colder seas, and on the other hand 

 some of the species of the genera listed under Coridae range well into temperate regions. Accordingly, 

 I am not ready to follow the division made by Jordan and have included the genera HalichOres and Pseudo- 

 julU, placed with the Coridae by Jordan, back with the Labridae. 



