140 CALIFORNIA ACADEMY OF SCIENCES 



angulated. Both caudal lobes are produced into filaments, the upper the longer. 

 The fin is very deeply forked, the median rays but one-fourth the longest rays of the 

 upper lobe. The lateral line is discontinued at the anterior edge of the yellow 

 blotch below last dorsal rays. It contains 19 or 20 scales. Each of the scales of the 

 median series on caudal peduncle is distinctly pitted, and contains a small tube which 

 is aj^parently imperforate, and represents the vanishing stages of the lateral line 

 which formerly occupied this region. The median series of scales on the trunk con- 

 tains 28 to 30 scales. Between the lateral line and the dorsal sheath are 21 rows 

 of scales. 



The dorsal contains 12 spines and 13 rays; the anal fin 2 spines and 12 rays. 



C. atrilohatus resembles strikingly in coloration C. notalus from Japan; but 

 the latter is deeper, and has 13 dorsal spines, larger scales, and a black-edged 

 anal fin. 



262. Pomacentrus rectifraenum Gill. 



A large number of adult specimens, 15 to 18 cm. long, are referred to this 

 species. 



Two very closely related, yet distinct, species were found associated around rocky 

 islands in Panama Bay. They were separated in the field by their slightly diflferent 

 color and proportions, characters to which we can now add a slight but perfectly 

 constant difference in the fin-counts. Owing to the absence, among our Panama 

 material, of series illustrating changes which occur with age, and to the lack of adults 

 from any other region, our identification of one of these forms with the Mexican 

 species redifrcenum is subject to some uncertainty. The second form is described 

 below as a new species {P.gilli). Our specimens of rectifrcenum are almost uniformly 

 light brown, with darker edges to the scales. The head is darker than the trunk, 

 and the ventral and vertical fins are black. The jjectorals are distinctly blackish, 

 with a light blotch on the upper rays near the base. A small black spot occupies the 

 extreme base of the uppermost ray. In the majority of specimens, no traces persist 

 of blue spots, but in others of full size, there are very distinct blue spots on the sides 

 and top of head, and on the scales covering the anal fin. In one specimen 14 cm. 

 long, there are in addition distinctly visible the vertical -blue streaks on the scales of 

 the sides, which are characteristic of the young of rectifrcenum. 



There are constantly 15 soft dorsal rays, and 13 anal rays, the last split ray 

 being in each case reckoned as one. These numbers hold in our Panama specimens 

 (thirteen in number), and are also found in two young sjiecimens of redifranum (all 

 to which we have access) from Mazatlan. In Gill's description of the types of redi- 

 frcenum, he assigns to it IG dorsal rays and 15 anal rays. It is reasonable to supjiose 

 that the split ray at the end of each fin was by him reckoned as two rays. In giving 

 an account of some of the tyj^e material furnished him by Dr. Gill, Dr. Giinther 

 gives 15 dorsal and 14 anal rays. In their description of the species, Jordan and 

 Evermann assign 13 rays to each fin, but this is certainly an error. 



