S. Mahunka szerk.: Folia Entomologica Hungarica 52. (Budapest, 1992)

The subgenus is strictly Oceanic: C. auripennis Saunders and C. longueti Théry occur on the Marianne Islands, while "C.obsoleta Blair" and C. bedoci Théry are inhabitants of the remote archipelago of Marquesas (where they reach - together with Pleiona tayauti Guérin, the closest relative of the genus Cyphogastra Deyrolle - the extreme limits of bupreslid distribution); C. taitina Kerremans and C.similis Kerremans have been described from Tahiti. "C. obsoleta Blair" and C. bedoci Théry seem to be allopatric within the Marquesas, the former inhabiting Fatu Hva, the latter Nuku Hiva and Ua Pou (I have not been successful in trying to identify two other locality (?) of C. bedoci Théry from label in the British Museum: "Upola"); probably each of them occurs also on other islands of the respective (southeastern for "C. obsoleta Blair", northwestern for C. bedoci Théry) part of the archipelago, but 1 have seen specimens from the above-mentioned ones only. Instead, the situation on the Mariannes remains unclear: C. auripennis Saunders, and C. picota Kerremans var. guamensis Kerremans have been described from Guam, and my collection contains one so labelled specimen of the Saunders' species, but none of the remaining representatives of C. auripennis Saunders; or C. longueti seen by me bears any information as to the definite locality within the archipelago. Anyway, the obviously relict distribution on the remote peripheries of the geographical area of the genus, together with the incipient stage of development of the abdominal plaque and morphological similarities to Pleiona Guérin or even Paracupta Deyrolle, make it evident, that Guamia Théry represents an ancient group, probably ancestral to Cyphogastra Deyrolle s.str. The synonymy of C. auricollis Saunders and C. longueti Théry is somewhat unclear, mainly due to Kerremans' (1910) misinterpretation, resulting in applying the Saunders' name to what we now know as C. longueti Théry. Théry (1926), examining Saunders' (1867) figures, disagreed, and pertinence of his objection has been fully confirmed in my study of the type specimen in the British Museum (Natural History): its yellow tarsi, coppery-red colouration of the greater part of elytra and abdomen, ad well as coarser elytral sculpture, leave no doubt that Kerremans (1910) was in error. The type of C. latro Kerremans (also in the British Museum) is simply an aberrant (probably young: brown with coppery-red pronotum) individual of C. auripennis Saunders. The situation with C.picata Kerremans is similar: Kerremans (1910) himself made it a synonym of his "C. auripennis Saunders (=C. longueti Théry), but the specimen in the British Museum, marked as the type, as well as the paratype in the Natural History Museum in Budapest, belong to the genuine C. auripennis Saunders (even if both show much more green on the elytra than is usual in that species)! Some doubts still remain - the "type"-label attached to the London specimen is provided with a question-mark ... - yet, until serious evidence to the contrary comes to light, treating C. picata Kerremans, as a green variety of C. auripennis Saunders seems most reasonable. I have not seen the type of C. guamensis Kerremans, but it was described (Kerremans 1911) as a variety of C.picata Kerremans, and is probably also synonymous with the Saunders' species. In the British Museum there were - under the collection-label "C.obscura Blair (unpubl.?) nec Kerr. 1895" - 8 ex. of a species closely related to C. bedoci Théry, but markedly different in its oily shine, very fine sculpture of the dorsal side, broad and clearly delimited pronotal foveae, and elytral sides evenly rounded to the very apices (Fig. 1); also the dorsal line is regularly rounded in profile (Fig. 2). In fact, the description of "C. bedoci Théry v. obscura Blair" had, indeed, been published (Blair 1932), but its author apparently soon took notice of the name being preoccupied: the only two determination labels in this series (those attached to the specimen marked "Type") have been written by himself and identify the beetle as C. bedoci Théry resp. C. bedoci Théry ab. obsoleta Blair. This latter has, to my knowledge, never been formally described (and, as an infrasubspecific name, is in any case invalid), yet I propose to retain it - at least until the identity of C. taitina Kerremans is fully cleared up - for this species, and to accept the