Australian Journal ofEntornology (1999) ]8,247-256 Cephalosphaera Enderlein, a genus of Pipunculidae (Diptera) new for Australia, with descriptions of four new species Jeffrey H Skevington Department o/Zoology and Entonwlogy, The University o/Queensland, St. Lucia, Qld 4072, Australia (Email: j.skevington@ento.uq.edu.au). Abstract Cephalosphaera Enderlein is a nearly cosmopolitan genus of big-headed flies (Pipunculidae) which was previously unknown from Australia. Four species have now been found in Australia, all of them undescribed. New species in the subgenus Neocephalosphaera include: C. eukrenaina, C. parthenopipis and C. petUa. Only one new species occurs in the nominate subgenus: C. prionotaina. Descriptions and a key to these species are presented. Key words Aschiza, Australia, Cephalosphaera, Neocephalosphaera, Pipunculidae, taxonomy. INTRODUCTION Pipunculidae are distinctive, but inconspicuous, relatives of the Syrphidae. They can be differentiated from syrphids by the large compound eyes that occupy most of their hemi- spherical head, distinctive wing venation (no vena open discal cell), chitinised postspiracular plate found in the larvae and their unique life-history. Pipunculids are exclu- sively parasitoids of Auchenorrhyncha. Over 1200 species of Pipunculidae have been described worldwide and it is likely that well over 2000 species exist. The taxonomy of the Australian Pipunculidae is poorly understood. Perkins (1905, 1906) first treated the continent's pipunculid fauna and, since then, only four publications have supplemented our knowledge of these flies (Hardy 1964; De Meyer & Grootaert 1991, 1992; Kuznetzov 1993), bringing the total described species to 36. I am currently revising a large portion of the Australian pipunculid fauna, and have recognised several genera previously unrecorded in the country. Cephalosphaera Enderlein is one such genus. Forty-five species of Cephalosphaera are known, five of them Australasian (four from New Guinea, one from Ambon) (De Meyer 1996; Rafael 1996). Recent collecting efforts and acquisition of specimens from over 30 collections have resulted in the accumulation of over 6000 Australian pipun- culids. Among these are 83 specimens of Cephalosphaera representing four species, all of them undescribed. MATERIALS AND METHODS Specimens of Cephalosphaera were obtained from the fol- lowing collections: AMSA, Australian Museum, Sydney, New South Wales, M.S. Moulds, M. Humphrey; ANIC, Aus- tralian National Insect Collection, Canberra, Australian Capital Territory, P.S. Cranston; GDCB, Greg Daniels Col- lection, Brisbane, Queensland, G. Daniels; INHS, Illinois Natural History Survey, Champaign, Illinois, USA, M.E. Irwin, D. Webb; QDPC, Queensland Department of Primary Industries, Indooroopilly, Queensland, J.F. Donaldson; QMBA, Queensland Museum, South Brisbane, Queensland, C. Burwell, S. Evans; UQIC, University of Queensland Insect Collection, St Lucia, Queensland, G. Daniels. Collection abbreviations follow Evenhuis (1997). The following abbrevi- ations are used in the material examined sections: Biological Centre (BC); Brisbane Forest Park (BFP); Creek (Ck); Mount (Mt); National Park (NP); Road (Rd); State Forest (SF). Adult flies were preferentially preserved by pinning a minuten through their right pleuron into pith mounted on a number 3 stainless steel pin. Some specimens were mounted by pinning a minuten through the right-hand side of the scutum, while others were glued to paper points. These methods were found to be inferior to the fonner method. The former method allows easy viewing of the genitalia, easy dis- section and places the minuten and specimen away from the handler's fingers. Specimens stored in 70% alcohol were dehydrated to 100% alcohol by serial dehydration over a period of several days. Specimens were then critical-point dried and pointed or acetone dried and pointed (following the methods of van Noort (1996». Dissection of male genitalia was often necessary. The abdomen was removed and boiled in lactic acid (approxi- mately 85%) for about 30 min. Lactic acid was chosen as a clearing agent due to its many advantages over potassium hydroxide or sodium hydroxide (Skevington & Marshall 1998). The macerated abdomen was then removed from the acid and placed directly in clean glycerin. Dissected term- inalia were stored in glycerin in plastic microvials on the same pin as the specimen. Once the genitalia were cleared, dissection of Cephalo- sphaera involved separating syntergosternite 8 from the remainder of the abdomen. In almost all cases this was all that was necessary to identify the specimen. Further dis- section was necessary only when specimens bad to be disar- ticulated for drawing purposes. Genitalia thus prepared were examined in glycerine placed on depression slides. All specimens are labelled with a unique reference