-14- Raiatea: Forster?; Wodehouse; Garrett; Wilson. Wilson says that on Bolabola and Raiatea he found a form of P. purpuratus "with more yellow" on the underparts. He probably referred to this species. Some of the earlier writers (Latham; Forster) refer to a bird from Raiatea that had a darker purple cap. (Cf. Wiglesworth, Ibis, 1891, 571; Salvadori, Cat. Birds Brit. Mus., xxi, 1893, 108, note.) ?Serresius R. forsteri (Wagler). (Type locality: Tahiti.) Tahiti: Forster. Forster's description agrees fairly well with S. galeatus from the Marquesas Ids. (cf. Salvadori, Cat. Birds, xxi, 173, note), and Wiglesworth recognizes it as the same. More material is required to settle this case, which may resolve itself into a stage of plumage of Globicera "wilkesi"- aurorae. The alleged species S.? R. forsteri is said to have ferruginous under tail-coverts, but there is no mention of the peculiar saddle-shaped feature (feathers projecting over the culmen) that pertains in the Marquesas bird. Immature birds of the Globicera aurora may, in some stage of plumage, agree with the bird described by Forster, in which case the name R. forsteri (named for Johann Reinhold Forster) may have precedence over aurorae (there are earlier Columba forsteri's, but no earlier G. R. forsteri). Globicera aurora (Peale). (Type locality: "Aurora or Maitia Island, one of the Society Group." = Makatea or Aurora of