Аннотация:In her paper 'The Atoconeura Problem' Longfield (1953) sketched the taxonomic disorder in the genus, caused by considerable variation in the sparse material available.Her solution was the recognition of two species based on epiproct morphology, A. eudoxia and A. biordinata, the latter with four subspecies differing mainly in markings: A. b. biordinata from highlands in Tanzania, Malawi and Katanga; A. b. chirinda in Zimbabwe; A. b. kenya in Kenya; and A. b. pseudeudoxia in W Uganda. Kimmins (1958) added A. b. aethiopica from Ethiopia, stating apologetically that 'one might question the wisdom of adding yet another subspecies'.Pinhey (1961a) remarked that 'Longfield has divided this [genus] into two species, but I think kenya may eventually prove to be a third species.' Later Pinhey (1982) suggested that 'it is possibly more correct to regard them [the subspecies] all as biordinata, with variable forms rather than subspecies'.Bridges (1991) listed A. kenya as a good species with the subspecies A. k. aethiopica, but made chirinda and pseudeudoxia synonyms of A. biordinata.The problem demands a more objective analysis of variation, as well as the study of new material and characters.This paper presents the outcome of that study, with a revised classification of the taxa and a reassessment of their biogeography. Methods and materialTo validate the views of Longfield (1953) and her contemporaries, the analysis was designed to test how well their characters applied to their classification.Most of Longfield's specimens (bmnh) were restudied, besides new material including West 121
