386 SNODGRASS AND HELLER 



In 1862, Gill {Proc. Ac. Nat. Sci. Phila. 1862, 148) described 

 2 species of Pojiiacentrus from Cape San Lucas — P. rectifrcemim 

 and P. Jlavilat7is., the type of each being an immature specimen. 



Later in the same year, Gvinther {Cat.., IV, 27, 1862) published, 

 from Gill's manuscript, the description of a third species, Pomoccntrus 

 analiguttata Gill. This species Gill himself in the following year 

 retracted. 



In 1863, Gill {Proc. Acad. Nat. Sci. Phila. 1863, 215) redescribed 

 P. rectifrcenum and P. Jlavilatus. In these descriptions, based on a 

 greater number of specimens, the difference described between the 

 types of the 2 species is much less prominent. 



Giinther, in 1866 {Fishes of Central America, 445) regarded 

 Gill's 2 types, Pomacentrus rectifrcenum and P. fiavilatus, as simply 

 2 color forms of the same species, P. rectifrcemun. 



In 1 89 1, Gilbert {Proc. U. S. Nat. Mus. 1891, 554) described the 

 species P. leucorus, distinguished from the other species by having 

 the pectorals tipped with orange (white in alcohol) and posteriorly 

 bordered with white. The type specimens of this species are from 

 the Revillagigedo Archipelago, but the species is abundant also at 

 both Cocos and the Galapagos islands. 



Jordan, 1896 {Fishes of Sinaloa, Proc. Cal. Acad. Sci., 2d Series, 

 Vol. V, 1895 (1896), 474, desci"ibed as Fzipomacentrus favilatus, 

 a very young example from Mazatlan, about |- inch in length. 



In 1899 Jordan and McGregor {Fishes of Revillagigedo Archi- 

 pelago and neighboring Islands, Rep. U. S. Fish Comtn.for i8g8, 

 282, 1899) recorded from Clarion and Socorro islands, 6 specimens 

 of a Pomacentrus that they identified as P. rectrifr(E?zum, and 12 

 others that they identified as adults of P. flavilatus. 



We have examined this Revillagigedo material and find that, 

 excluding P. leucorus with orange-tipped pectorals, it includes only 

 one species — not 2 as identified by Jordan and McGregor. Further- 

 more, we have examined both young and adult examples from 

 Panama which are certainly P. rectifrccmi??t Gill. The Revillagigedo 

 adults are not the same species as the Panama adults, and the young 

 are neither P. rectifra^mim Gill nor P. favilatns Gill. Hence, we 

 have described the Revillagigedo form as a distinct species — Poma- 

 centrus redemptus {Proc. Wash. Acad. Sci., v, 1903, 200, pi. vi). 



At Cocos and the Galapagos islands there occur 2 species of 

 Pomacefitrus — one is P. leucorus, the other we have described as P. 

 arcifrons {ibid, 202, pi. vii). 



In 1904, Gilbert and Starks (Mem. Cal. Acad. Sci., iv, 141, pi. xxi) 

 described the species P. gilli hom Panama. 



