a Sees ae ARG sig SMITHSONIAN INSTITUTION WASHINGTON, D.C. 1973 psf BISH FOX Ent: Bee Flies of the World The Genera of the Family Bombyliidae ry iA iy om ONT r q ~ By FRANK M, HULL Research Associate, Smithsonian Institution Smithsonian Institution Press City of Washington 1973 United States National Museum Bulletin 286 SI Press Number 4903 Library of Congress Cataloging in Publication Data Hull, Frank Montgomery, 1901- Bee Flies of the World. Bibliography: p. 1. Bombyliidae. 2. Parasites—Insects. I. Title. QL537.B75H85 595.071 73-1581 ISBN 0-87474-131-9 Two species of the remarkable wasp mimics of the Genus Systropus Wiedemann: S. arizonensis Banks (top) and S. elegans Hull. Artist: Arthur A. Cushman. Contents Page MD ERICA TION eters prot Sars tr See ops tay eeeaeureeN tnd Pay eplrey ea act tee eh Meigen ecuceets nets 1x Cn Owed oMeNtS yee eri a out ieucliplete saa tay eden cee Sans (diay se Rita ne al ec diaa try anus Xl TNGROMUCELOMBH 32 rau ee cite Pent ceyeyltes wsrcise se Meme mice asi rol ders ie el er Gaeoe, wera sites 3 Eastorysof Bombyludacystudiesi.-.) ea ee ee sy yo ns eee eee 4 Generalizations of the Host Relationships of Bee Flies During Immature Stages 6 Review of Published Information on Host Relationships. .......... 8 Ethology and Interrelationships of the Adult ..............22.. 33 Morphologysof the immatures Stages. a) esos ayo) 2 is oe Seeeetee e 36 Morphology of the Adult Bee Fly... ............+2....2.4. 47 Terminology Used in Structures of the Male Genitalia of Bee Flies... ... 52 @hromosomes yeas ck ears, eye teen reyes aay aye wth as tan «A AE AUR DD 56 Hossil@Bomlbylitdaeye is) ee eee eae Re ae NUE 2% 56 Zoogeography of the Bombyliidae...........2........2.. 59 Symopsis and MDistributvons es peel aan OUe le tay) ea Renee 61 Phylogeny and Subdivision of the Family Bombyliidae. ........... 63 Fomeo pli thialimae tesa yi) eee eine nse a Sh Bate souls g.hogl) oo RETO 68 Bomibyliumae aries Lee ne PR IGN Cie elke elo ucad taba g EON 68 Bomlbyluuniies rica Sates 2 Ree Memes oe ay sae es EE TS 69 Dischistiniyeayikaiey wae See a Sed Cee aes auc AERO. 69 Heterosty limit wayne ucuyeiat sa ee tk eoueNi came Gide, Randi ee ao ls REE AG 69 @y therein eine vei nr eabre sit iene ene ici Rel USS iie Sue E IR ERIL 22 69 Womop Orin ee ek ek hse ms eas orien tla cee) RDPB Ac 69 COrsOnmyZlMIe ccccket vee gow Yodo /20 sieves gui keoe Peer Pere alleela nee aries a REMAN PEE gh 70 IMiarTobezzvumnitys 2) Causes yey Woe netesetues Wey Naima eeueui sinite Deh NSH ER REL OR AREA DAVES tet 70 Acrophthalimy dint, ss) he tae menial pier ioe pephecc SOLER MERE aap 70 Paratoxophorini etre Mn hiid tame nas as cet histidine LOOM o> 70 Chimie ees e eet ee eee NO Satan. Men * AMARONE ean 2 70 OCTANE ANS Seah Ate RS ae RT oh ahs a Walaa sae a oh OAR ARRCP Cea POEL 70 Phthairinm adele. hse ila Ne caer aye) nV aeSO EMMA wo ugah uty cede dines Qe REALE a 194 (Geromtinaeiyy ee usury cciles eis iplsrive Uieymtee rainy Nevaeh earls Jasgre dg cca, wae NCD Ff 202 (WISIN Scheer i empenmnay wld fra can ep baumtanabeman ot deaee ei Hey ice Jie On MSEPORWUS ROOM 8 212 FVeterotropinae a pete we ob oats Aiur olge IME ie ono) weld alc rons SUR be atthe 229 FROXOPHOEIMACH rma ecw Meron uel ie tou uemep asa CU ey as Ucch a). ey ey cee RUMI anctetope s 1s 231 MOXOPNOLINIS Give) copia d 6 on ce ty eeemuusl hatin eal delnacs eho: L.A RACER Eo ars 232 Wepidophorimigmanes feugiat sii leerenys) os a elec sugiaiar' ayn T Pactabicinys 232 SV SULOPUN Arar eee etey cect coy .th eA eMC ambien Zaieencd wou relict vellke 242 Wolichomiymmigm evecare et ey rete emer meres sa wet nee tee fey leh est ler yey asl te 243 RS yASLAR OY ONUTEL SR SA et Sh coh Acc laa cae ps Se ene ecco ae Fa Pa 243 Xenoprosopinae Platypyginae . . Cyrtosiini Psiloderini . Platypygini Mythicomyiinae Mythicomyiini Empidideicini Tomophthalmae Cylleniinae . CyHeniini Henicini . Tomomyzini Peringueyimylini . . Neosardini . Womatimaew es wae. Lomatiini Antonini Aphoebantini. . . Prorostomatini Xeramoebini Comptosiini Exoprosopinae \Wanlbim 5 : Villoestrini . . Exoprosopini . . Anthracinae ..... Anthracini . . . Walkeromyini Bibliography . Index vi Myonematini . . . Supplemental Text-Figures Illustrations of Bombyliidae (figures 1—-1030) Text-Figures vii Page Eggs, larvae, and pupae of Anthrax limatulus Say 37 First-stage larvae .. . 38 Mouthparts of bee fly larva ae . Ihe Shea Se nie ie Reo eg ante 39 Pupae of bee flies... . . Dae ete tahoe rhe eet iat epee een ee OA Armament of pupae of three subfamilies . : : : 46 Mouthparts of adult bee flies of Lxoprosopa Macquart : 48 Thorax of a bombylid BS 49 Wing of a generalized Recent bee iy. te 51 4, Male genitalia of Poectlanthrax californicus C ally PRA St Lhe eine Ee Se OOD. Fossil bee flies . : . 57,58 Concentration areas of beet fires of the Ww orld) 60 Provisional phylogenetic chart of bee flies 65 Distribution pattern of the Bombyliinae 68 Habitus, Bombylius major Linné . Ut Habitus, Germinaria canulis Coquillett . 115 Habitus, Corsomyza simplex Wiedemann 164 . Eelimus gracilis Loew, after Austen . 178 Distribution pattern of Phthiriinae . : 194 Habitus and lateral aspect of PAthiria sremies 196 . Distribution pattern of Gerontinae . 202 Habitus, Geron senilis Fabricius . 204 Distribution pattern of Usiinae . 213 Habitus, sta aenea Rossi . NEN 214 Distribution pattern of Heterotropinae cond Remap: osopinae . 222 Habitus, Heterotropus aegyptiacus Paramonov 204 Distribution pattern of Toxophorinae . 231 Habitus, Zoxophora leucopyga Wiedemann 233 Distribution pattern of Systropinae . 242 Habitus, Systropus (Cephenius) maccus Ender fem 244 . Distribution pattern of Platypyginae . . 253 Habitus, Cyrtosia nitens Loew . 255 Habitus, Pstloderoides mansfieldi Eteses 258 . Distribution pattern of Mythicomyiinae . 264 Habitus, Wythicomyia monacha Melander . 266 Distribution pattern of Cylleniinae . 278 Habitus, Cyllenia maculata Latreille . 281 . Distribution pattern of Lomatiinae . . Habitus, Zomatia lateralis Meigen . . Distribution pattern of Exoprosopinae . Seven maps showing distribution patterns of areas @ Posciantinas, . Habitus, Zxvoprosopa rhea Osten Sacken . . Distribution pattern of Anthracinae . . Habitus, Anthrax tigrinus deGeer . Supplemental Text-Figures 7. Bee fly pupae . Heads of pupae . Bee fly larval head str ucture . Female genital structures . . Wings of fossil bee flies . . Map ‘of collections in 1962 . Habitus, Zinnomyia karooensis, after Hesse : Habitus, Dischistus pulchellus Austen, from Austen . . Habitus, Cytherea barbara Sack, from Austen Habitus, Callastoma fascipennis Macquart, from Austen . . Habitus, Legnotomyia trichorhoea Loew, from Austen . : Habitus, Heterotropus nigrithorax, new ispecies, after Fran¢ois . Habitus, Caenotus minutus Cole, from Cole . Habitus, Dolichomyia gracilis Williston, from Cole Habitus, Amictus minor Austen, from Austen Habitus, Lomatia lepida Austen, from Austen viii Page 302 308 364 393 410 432 434 Dedication To my late and beloved wife, Mary Marguerite Chappell Hull, I owe an immeasurable debt. She gave of her time patiently and untiringly to the typing and checking of thousands of pages of manuscript in its several revisions. To the compilation and checking of nearly three thousand bibliographic references she gave diligent and painstaking attention. At the time of her death she had also compiled a bibliography of the Syrphidae, comprising perhaps two thou- sand titles. Not the least of my debt to her lies in her encouragement and her faith in the worth and completion of these two world monographs on important groups of insects. Neither this present work, nor the earlier work, Robber Flies of the World, would have been possible without her devotion and help. All of these tasks were cheerfully undertaken in addition to caring for a family and a home. Her noble and saintly life is done. In God’s home she rests. ix Lage MTP LS We i Acknowledgments This work was begun in 1952 and was based upon a study of the extensive collections of the United States National Museum (now the National Museum of Na- tural History), the American Museum of Natural His- tory, the Museum of Comparative Zoology at Harvard University, and the British Museum (Natural History ) in London. The National Museum of Natural History is rich in Bombyliidae. It includes the extensive collec- tions made by D. W. Coquillett in the western United States and much other material added in recent years. This also includes the acquisition of the Melander Col- lection. The British Museum (Natural History) afforded material drawn from many quarters of the World. The collections in England were visited in 1952 and again in 1953 and 1954. The National Science Foundation gave three grants which made possible the study of museum collections in the United States, England, Europe, and Australia. To this Foundation I should like to express appreciation both for the support of this monograph and for the helpfulness of its officials. The University of Missis- sippi provided a temporary Research Professorship which permitted travel time and absence from teaching. I wish especially to thank the late Honorable Walter Sillers of Rosedale, Mississippi, and the Honorable Robert D. Morrow of Brandon, Mississippi, for encour- aging state support for this work. I am deeply indebted to Dr. M. B. Huneycutt, Chairman of the Department of Biology, University of Mississippi, for continuous help, support, and encouragement through the long task. Dr. Lewis Nobles, recent Dean of the University of Mississippi Graduate School, and his successor, Dr. Joseph Sam, have been most helpful. I wish to express my indebtedness particularly to the many persons in numerous countries who have gener- ously helped me in this work on the bee flies. The staff members at the United States National Museum, both of the Department of Agriculture and of the National Museum staffs, were most helpful. I want to thank Dr. Willis Wirth and Dr. Curtis Sabrosky for their interest, suggestions, and guidance concerning problems of nomenclature. Mr. George Steyskal was most helpful in recommending preferred nomenclature for the morphology of the male genitalia and for his remark- able knowledge of appropriate gender in specific names. I shall always have a deep feeling of gratitude toward the present Chairman of the Department of Entomol- ogy, Dr. Karl Y. Krombein, and his predecessor, Dr. J. F. Gates Clarke. I wish to thank the staff members of the Smithsonian Institution Press for their patience and skill in getting the manuscript and illustrations into print. My thanks also to Dr. L. V. Knutson. Many other persons have been most helpful. I am deeply indebted to Dr. A. J. Hesse of the South African Museum in Capetown for furnishing material which, owing to the illness of my late wife, I would not have been able to see except by travel and visitation. The late Dr. Reginald H. Painter of Kansas State Univer- sity lent several genera I had not found elsewhere. I wish to thank Dr. Frank Morton Carpenter of the Biological Laboratories, Harvard University, Dr. J. Bequaert, also others of the Museum of Comparative Zoology; also the late Dr. Howard Curran of the American Museum of Natural History; Dr. Edward Ross and P. H. Arnaud of the California Academy of Sciences; Dr. Frank Cole, Dean E. Gorton Linsley, Dr. P. H. Hurd, and Mr. A. T. McClay of the Uni- versity of California. I especially appreciate the help- fulness of Mr. Jack Hall of the Riverside branch of this University, also Carl Rettenmeyer and F. L. Truxen. I wish to thank Dr. W. G. Downs, then of the Rocke- feller Institute staff in Mexico City, for having been instrumental in 1951 in interesting me in both Bombyli- idae and Asilidae during my visit to Mexico City and to Cuernavaca. Professor Candide Bolivar of the Insti- tuto Politecnico was most helpful then and on later visits. Professor H. F. Strohecker, University of Miami, sent a few bee flies from Florida. Dr. Z. Kazab, Director General of the Hungarian Natural History Museum, very kindly sent some bee flies from Mon- golian expeditions; no new genera were found. I owe a debt of gratitude to the authorities at the British Museum (Natural History), especially to Mr. N. Riley, Dr. W. E. China, Mr. Harold Oldroyd, the late Mr. R. L. Coe, Mr. Clark, and the late Dr. Fritz van Emden. All of these gentlemen were extremely helpful. The late Professor J. Timon-David of Mar- seille, Dr. Willi Hennig and Dr. Erwin Lindner, and the late Professor Fred Keiser have all been most help- ful. I acknowledge also help from Mr. B. R. Stucken- berg of the Natal Museum and of Mr. John Bowden of the University of Ibadan, Nigeria. xi xii ACKNOWLEDGMENTS To the many in Australia who gave generously of time and helped in numerous ways, ‘T owe an untorget- table debt. These people include Dr. A. J. Nicholson, Director (when I was there) of the Entomology Divi- sion of the eae and his efficient staff as follows: the late Dr. S. J. Paramonov, Mr. Edgar Riek, Mr. Kk. R. Norris, ie item (0; ampbell, Dr. M. Je ce White, Mr. Frank Gay, Mr. Paul Wilkerson, Mr. Me Er Wallace, Mr. John Callaby, Mr. Keith Tay es and Mr. Don Wilson. Others in Australia who gave assistance of great value were Mr. R. T. M. Pescott and Mr. Alexander Burns of the National Museum in Melbourne; Dr. and Mrs. (Dr.) I. M. Mackerras of the Queensland Institute of Medical Research; Professor F. A. Perkins and Dr. Elizabeth Marks of the University of Queensland; Mr. George Mack of the Queensland Museum; Mr. Anthony Musgrave of the Australian Museum in Sydney; Dr. P. D. F. Murray, Dr. A. R. Woodhill, and Woe, ds ly Henry of the U niv ersity of Sydney; Mr. D. J. Lee of Public Health and Tropical Medicine at the University in Sydney; the late Mr. H. Womersley of the South Australian Museum; Mr. Harry Lower of the ee Agricultural Research Institute, Adelaide; Mr. G. H. Athol Hardy of Katoomba, New South Wales; Mr. Douglas of the Perth Museum; and Mr. E. F. H. Jenkins, the Government Entomologist at Perth. The fine and valuable drawings prepared by Mr. Arthur Smith deserve special mention. He is responsi- ble for most of the faces of the bee flies—identified by the initials A.S. His exceptional, fine drawings have added greatly to the project. Many of the remaining drawings and all of the drawings of the genitalia were done by Mr. Kenneth Weisman—bearing the initials K.E.W. I am greatly indebted to him for his fine and skillful work. Mr. Arthur Cushman prepared the fron- tispiece and the text-figures identified with initials A.D.C. Most other drawings were done by the author, a few were made by my graduate students, Clyde Sartor and William Martin. I wish to thank the many librarians at all institutions who have used their special talents to help locate obscure references. The staffs at the Smithsonian Insti- tution Library and at the University of Mississippi Library have been particularly helpful. I wish to especially thank Miss Mahala Saville. I owe a deep debt of gratitude to my principal typist, Mrs. Emily H. Peacock. My daughter, Cecil Hull, also was most helpful in typing manuscript. The imdex has been prepared by my youngest son, Clovis Hull. Among others who helped with the numerous minor tasks, assisted with manuscript and illustrations, I wish to thank my eldest son, Frank M. Hull and his wife; the late Professor Ralph B. Brundrett, Jr., for checking translations of German descriptions; and Rex Paul, a diligent graduate student in my classes. I especially wish to express my appreciation to my former graduate student, William C. Martin, a skilled insect illustrator and photographer, for his assistance in preparing several drawings and charts, and also fig- ures 1-9, 15, 16. Maps were taken from the thesis of my er aduate student, Miss Dee Sellers. It is with great happiness that I thank my dear pres- ent wife, Mrs. Laura Felton Hull, for her help in the heavy task of reading manuscript and of bookkeeping of the numerous descriptions and illustrations; also, I add my thanks to her for her patience and skill. Since the beginning of these monographs in 1952, the following great dipterists who collaborated in these works are now deceased: Dr. Reginald H. Painter, Dr. Charles Howard Curran, Dr. 8. J. Paramonov, Dr. Stanley Bromley, Dr. Fritz van Emden, Dr. Richard Frey, Prof. J. Timon David, Prof. Fred Keiser, Mr. Hassan C. Efflatoun, Mr. R. L. Coe, Mr. H. Womersley. Possibly there are others. I also acknowledge with deep gratitude the assistance and encouragement of eight great dipterists and entomologists of former years, now deceased: Dr. Raymond C. Osburn, Dr. J. M. Aldrich, Raymond Shannon, Prof. James Stewart Hine, Dr. P. P. Calvert, Prof. Robey Wentworth Harned, Mario Bezzi, and A. L. Melander. The author wishes it had been possible to incorporate all of the suggestions made by friends who read the final manuscript. Unfortunately, this has not been possible, because the three generous grants given for preparation were exhausted; also because of my regular teaching duties and of several years of illness in my family. What changes were possible have been made. I make these statements lest I be deemed unappreciative. I bespeak the tolerance of my readers in my efforts to produce a work that will be useful and provide a com- parative, worldwide view of a family of remarkable beauty and usefulness in man’s increasing battle against destructive insects. Bee Flies of the World The Genera of the Family Bombyliidae ot iv bares j aa pried t sles con ‘Saab ae c eS GPA Isy hon ur oe : ‘ FROWN Nh Le PO hid Ra Oe 7 ee ' a iy : a ieee a PO ihe oe : iV PN Gr cea tea Sie Siuk 7 "2 Oates Bato ceil rt niss, of. “5 Patty hte Introduction The bee flies, or Bombyliidae, constitute at the pres- ent time one of the large and diverse families of the order Diptera. This family is worldwide in distribution ; nearly 4,000 species are known, distributed among 194 genera and 24 subgenera. These flies are significantly more abundant in arid and semiarid regions, but even in such areas they are more numerous at well-watered spots, arroyos, or canyons, which support the rank vegetation necessary for the population of insects that serve as hosts for these flies. In numbers of present-day species, this family is smaller than the Asilidae, yet we know 46 species of fossil bee flies found in 31 genera, 20 of which are extinct genera. Of the 39 species of fossil asilids, only 3 fall into extinct genera. So far all fossil bee flies are from the Oligocene and the Miocene, with none known from the Eocene period. Wherever the habits of the larval stages of bee flies have become known, they have proved to be destroyers of locust eggs or parasitic or hyperparasitic upon a wide variety of other insects. The net effect of the many species which attack and destroy egg pods of hungry locusts is certainly of great benefit at times in controlling the numbers of these insects. Other species parasitize the extensive group of noctuid moths, whose larvae form the destructive cutworms and sod worms. Adult bee flies of many species visit flowers in search of nectar and pollen. Perhaps only three groups of these flies affect man’s economy adversely: (1) one species of the small genus Heterostylum Macquart is known to destroy the bees that pollinate forage crops; (2) the group of secondary parasites, which by destroying the scoliid enemies of white grubs, such as Anomala and Phyllophaga, oper- ates to maintain scarab populations, instead of dimin- ishing them. This second group includes Ligyra New- man and some species of Hxvoprosopa Macquart; (3) a third group is hurtful, like some species of Hemi- penthes Loew, because the flies destroy the tachinid parasites of caterpillars. Probably the useful species of Hemipenthes outnumber the hurtful species. Thirty-two genera of bombyliids in eleven sub- families have been reared from the hosts, and as time goes on, and more rearing data are accumulated, other genera and species will undoubtedly be added to the list. The hosts of bee flies are found within six orders of insects: Lepidoptera, Orthoptera, Hymenoptera, Dip- tera, Coleoptera, and Neuroptera. In any case it is clear that the Bombyliidae exert a considerable though an- nually variable effect upon the populations of insects. Occasionally bee fly populations are extensive and in such instances may have a key role in maintaining the natural balance of the components of faunas. Even where the percentage of parasitism remains at a rela- tively low level, perhaps 5 percent, bee flies are still a factor in insect control; nor do we know that low-level bee fly parasitism is not somehow due to competition by high-level tachinid or Hymenoptera parasitism. While the Bombyliidae stand fifth in number of species in the order, they really fall into what might be called the third level of speciation, for bee flies are exceeded in number of species only by the approxi- mately equal Tipulidae and Tachinidae, which two families are of nearly the same size, and the Asilidae and Syrphidae, which are also of nearly the same size. The latter two families have fewer species than either the Tipulidae or the Tachinidae but are considerably larger in number of species than the Bombyliidae. This leaves Bombyliidae in third place. The fossil record suggests that the Bombylidae, like the smaller family Nemestrinidae, are rather ancient, and further that today they meet with considerable competition from both the parasitic Tachinidae and the parasitic Hymenoptera. Bee flies generally are large and often beautiful and graceful insects with attractively patterned wings and brightly colored bodies; it is not surprising that many persons have given much time and study to them. Few other Diptera match these flies in powers of flight, and to watch a giant species of H’zoprosopa such as ingens Cresson—at times common in the Rio Grande area— with its long, delicately slender legs, as it comes to lightly rest upon a flower head is an attractive sight to students of insects. Austen (1937) states: “the charm of spring in Palestine is enhanced by visits to low- growing flowers of dainty little Bombyliinae like Anastoechus stramineus.” Certainly, this is true over a wide part of the globe. It has been 42 years since I had the opportunity to see large numbers of early bee flies in the New Mexico desert coming to verbenaceous 3 4 BEE FLIES OF THE WORLD blooms, but it is one of the sights I have long remem- bered. Many dipterists have contributed to our knowledge of the family Bombyliidae, and today a great deal of attention is being directed to these insects by a younger generation of entomologists. It is the hope of the author that this work will acquaint persons interested in the family with a knowledge of these insects in other regions and will generally stimulate interest in a most attractive family of Diptera. This work is designed to integrate all previous studies in order that we may have an effective world concept of this group. The phylog- eny, distribution, and life histories have been treated, in addition to the morphology and taxonomy of the group, and a world checklist of species is included. The accompanying bibliography covers a span of years and contains the work of more than 500 authors and nearly 1,600 titles; it ends with the year 1964, but a few later papers have been added. Complete bibliog- raphies in all groups of insects are much needed. The author takes full responsibility for the arrange- ment of genera and subgenera in this work. With many ardent workers in different parts of the world it is hardly to be expected that all will find themselves in full agreement with the disposition of genera made here. Nevertheless the present treatment may serve as a basis toward further studies of these flies. History of Bombyliidae Studies Linné (Linneaus), in 1758, in his tenth edition of “Systema Naturae” erected the genus Bombylius, in which he included three species (Bombylius major, B. medius, and B. minor) all of which are valid today. At the same time (1758) in his genus Musca he described three species of bombyliids (J/usca morio, M. maurus, and J. hottentotus). In his twelfth edition Linné, besides adding Bombylius capensis, described the species I/usca denigratus. This last species later proved to be the same as his earlier A/wsca morio. All four species of J/usca, three of them valid, were soon taken out of J/usca and placed in the new genus Anthrax, which had been created by Scopoli (1763). Scopoli (1763) based his genus Anthrax on a single species, at first believed to be Musca morio Linné (1758), but it was later shown by Bezzi (1902, 1908) and Aldrich (1926) that Scopoli’s type-species actually consisted of a 1746 unnamed species of Linné’s, and oddly a species which Linné himself still confused with his own J/usca morio in his 1761 paper. This unique type-species of Anthrax Scopoli, errone- ously believed by him to be MZusca morio Linné, proved to be the species that was described by Schrank (1781) as Musca anthrax Schrank. The critical character of separation of the genus consisted of a tuft or pencil of hairs at the apex of the antennal style, which Scopoli had penetratingly observed in setting forth the char- acters of his genus. All of these several species of bee flies left by Linné in his genus Musca, together with Musca anthrax Schrank and others added to this group of flies by later authors, came to be regarded as a separate family of flies known as the Anthracidae, or more commonly pub- lished as Anthracides. One of the Linnean species, Musca morio (1758), was indeed a mixture of three species—Musca morio Linné, Musca denigratus Linné (1767, 12th edition), and Musca anthrax Schrank (1781). It was not until many years later that Meigen and, still later, Macquart astutely recognized that these Musca species were indeed diverse members of the same family as Bombylius Linné, and were united with it. The student of bee flies is referred to Aldrich (1926). Linné’s species, Bombylius major, stands as the type of the genus and may be reckoned relatively generalized as far as the other members of the family are concerned. Moreover, Bombylius major Linné and Musca morio Linné (now Hemipenthes morio) are Bombyliidae which are holarctic in distribution and therefore occur in both Europe and America. It is interesting that they each belong to widely separated major subdivisions of the family. Out of the many species placed by later authors in the second division, typified by Linné’s species of Musca and in early days called Anthracides, the genus Anthraw itself is restricted to species having the tuft of hairs at the apex of the antenna. Later dipterists were able to discover important characters supplementary to the pencil of hairs and the group gradually became known as the Bombyliidae Tomophthalmae, while the group Bombylius and allied forms were grouped under what has been called the Bombyliidae Homeophthalmae. This nomenclature is discussed later in this work. The genus Anthrax Scopoli is restricted to those forms that do have the pencil of hairs at the apex of the antenna; all others in this group fall into one or another of many later genera. Shortly after Linné, Fabricius (1775-1805) in several publications described 63 species of bee flies, and 2 genera which are now the basis of tribes or subfamilies. Mikan (1796) must have been an ardent enthusiast of bee flies, publishing his monograph of the Bombyliidae of Bohemia with 4 color plates. Still later Wiedemann (1817-1830) described 113 species and erected 4 impor- tant genera, one of which is the basis of a subfamily. Meigen, in the same period, besides describing many species, erected 4 genera. Subfamilies are now based on three of these genera. About the middle of the nine- teenth century, a little before or after, many dipterists made significant contributions to our knowledge of the family. Among these were Loew, Macquart, Walker, Rondani, and Bigot. Macquart described 14 valid genera. Loew erected 18 genera that are valid today. In “Die Dipteren-Fauna Siidafrika” Loew gave us our first comprehensive study of the Diptera of this vast region. Most of Loew’s work was on the Palaearctic and Nearctic fauna. Other dipterists of this century who made contributions were Schiner, Philippi, Jaen- nicke, Roeder, Williston, Coquillett, and Osten Sacken. BOMBYLIIDAE 5 The works of the last three were particularly important in supplying us with knowledge of the fauna of the United States and Mexico. Williston erected 6 genera, Coquillett 8, and Osten Sacken 12 new genera that are now in use. The early years of the present century ushered in a period of greatly intensified work on bee flies. While both Becker and Bezzi began publishing on bee flies prior to 1900, almost all of their many publications fall in the period of 1900-1926. The untimely and tragic death of Bezzi ended a career of brilliant work in dipterology, much of which had gone to the Bombyl- idae. Bezzi’s 45 publications on bee flies of the Palae- arctic and Ethiopian faunas stand as a monument to his astuteness and penetration, as well as to his industry. Throughout his life Bombyliidae were favorite objects of study. Becker’s fine series of publications totaled more than 30 and is particularly important in relation to the fauna of eastern Europe and Asia Minor. Beginning in 1924, Paramonov made numerous, exhaustive studies of the bee flies, especially of the region that is now Soviet Republics. His publications total more than 60 on this family. Two additional monumental studies of the bee flies of the Palaearctic region have appeared: The fine, illustrated volume by Engel (1932-1937) as a part of Lindner’s Die Pliegen der palaearktischen Region, and the great work by the late Efflatoun Bey (1945), “A monograph of Egyptian Diptera. Part 6. Family Bombyliidae,” which represents only the first half of his treatment of the Bombyliidae. The second half les written but unpublished because of his untimely death. His fine color plates remain unpublished for lack of a sponsor. Efflatoun comments that as a result of his studies and those of his students, the number of known Egyptian Bombylidae has quadrupled since 1919. Ver- rall, Austen, Edwards, and Oldroyd have also worked with bee flies and each has given us several publications. In the United States in recent years several students have undertaken a serious study of this family. The late Dr. R. H. Painter became interested in this family i 1920. Since 1925 he has published 20 papers on bee flies. The untimely death of this able and enthusiastic worker, which occurred suddenly in Mexico City on December 23, 1968, cut short a brilliant career at its zenith, and it deprived me of a valued friendship which began when we were tent-mates Japanese beetle scout- ing in New Jersey in 1920, each turning to Diptera for life’s work. We were also roommates two years at Ohio State University. In recent years he was ably assisted by his wife, Mrs. Elizabeth M. Painter. Other recent workers include the late A. L. Melander, C. H. Curran, J. C. Hall, D. E. Johnson, and R. B. Priddy. In South America, Messias Carrera has given some attention to this family. In South Africa, A. J. Hesse has devoted more than a quarter of a century to intensive study of South African Bombyliidae, result- ing in three astounding and comprehensive volumes, illustrated with more than 800 figures. Recently John Bowden has completed a fine volume on the Bombyli- idae of Ghana. In Australia, Roberts made significant studies on bee flies of that continent, and G. H. Hardy also interested himself in this group. This work attempts to supply the minimum of ade- quate illustrations for each genus. Past publications of Diptera having to do with bee flies have been sparse in illustrations. The most notably illustrated works are those of Engel, with 239 text-figures and 15 plates with 198 figures, and of Efflatoun Bey, with 551 figures, all in black and white. Bezzi (1924) used 46 figures; Austen (1937) illustrated with 72 figures by Terzi and 1 color plate. Hesse’s South African work with 823 figures has already been mentioned. There is also a color plate in the initial volume of Die Fliegen der palaearktischen Region. Painter and Hall (1960) pro- vided color photographs of the species of Poecilanthrax Osten Sacken in their monograph of this genus. Effla- toun Bey, in a communication to this author, stated that he was unable to publish any of his beautiful color plates of Bombyliidae. This is a great loss to students of this family. Séguy and Becker have illustrated some of their publications. [very student of insects stands deeply indebted to the work of others before him. I wish to pay especial tribute to the fine work not only of Engel, but also to Bezzi, who assisted me in my early studies 40 years ago, and to Dr. Paramonoy with whom I have had pleasant excursions, and to the very generous assistance of Dr. A. J. Hesse of the South African Museum at Capetown. Certainly, the student of the Bombylidae is deeply indebted to the authors of the more than 200 papers that deal with the fascinating and little known life histories of this family. These are summarized here with the special interest and hope of drawing much increased interest to this important aspect of this family. The species list in this work, which has been included with the genera and subgenera, is offered as a practical device for the student, new and old, to check on the contents of each genus. Where a species was described in a different genus from which it now stands, this is indicated by the earliest generic assignment enclosed in parentheses. The dates accompanying each species as listed will readily give the bibliographic clue to the place of publication. In the preparation of the species list I have followed Kertész (1909) in his world catalog of Diptera, except for the Palaearctic region where I have turned to Engel and Paramonoy for the disposi- tion of these species. Of course Hesse (1938, 1956) has been followed for the assignment of South African species. Also the assignment of North American species reflects the thinking of recent students in this area and many of them have been checked by the author. Wherever I have been privileged to see types of species at various museums, I have used the information to reassign them where necessary, according to the con- cepts of this study. Many of the exotic species of older authors stand in need of additional study and especially 6 BEE FLIES OF THE WORLD of illustration. Kertész (1909) listed 84 genera and 1,696 species; the number of both genera and species has more than doubled since that time. This checklist ends with the year 1965. Generalizations of the Host Relationships of Bee Flies During Immature Stages The intricate relationships of bee flies with other insects, upon which they are parasitic, or more infre- quently hyperparasitic, provide deeply interesting and fascinating material for study by the future generations of entomologists. Hosts are known for comparatively few genera, and adequate life-history studies have been made for only a few species. The account by Shelford (1913a) of Anthrax analis Say (as Spogostylum), the work of Clausen (1928) in India on the hyperparasite Ligyra oenomaus Rondani (as Hyperalonia), and the work of Bohart, Stephen, and Eppley (1960) on Heterostylum robustum Osten Sacken, as well as Krom- bein (1967) on trap nesting stand as excellent and highly satisfactory examples of able life-histor 'y studies. Such painstaking studies require much time and effort. Some generalizations do emerge. We know hosts of a few of the genera of eleven subfamilies out of four- teen. Of five subfamilies we know the host of only one genus. We know that these hosts are derived from six orders of insects, but much more commonly from Lepi- doptera, Orthoptera, and Hymenoptera. The first group of bee fly genera—no less than nineteen, and probably the vast majority—seeks soil-hidden hosts; females of many genera have acanthophorites. These include Sparnopolius Loew, which attacks white grubs in the soil, and at least one species of Villa Lioy, which attacks the tenebrionid J/eracantha Kirby. Genera, like Poecilanthrax Osten Sacken, that seek sod-hidden cut- worms also fall into this group. Four genera attack mud cells of wasps or stem-mining wasps, and three genera presumably attack exposed caterpillars. No less than thirteen genera in seven tribes or subfamilies specialize in the consumption of egg pods of locusts; probably many other genera do likewise. We know some of the hosts for thirty-two genera. A second group of bee flies attacks subterranean in- sects which are miners, for example Heterostylum Macquart parasitizing the mining bees Voma Latreille and Anthrax analis Say, a parasite of the mining beetles of the genus Cicindela Linné; both of these flies flip their eggs directly into or near the mine or tunnel. Presumably, Dipalta Osten Sacken, which has been reared from ant lion larvae, deposits its eggs the same way, as ant lion pits are conspicuous objects. From this group it is perhaps but a small extension to the two genera Anthrax Scopoli and Toxrophora Meigen, of separate subfamilies, which attack the earthen nests of mud daubers, eumenids, or the cliff-bank nests of Anthophora Latreille and stem tunnels of Yylocopa Latreille. In all of these cases, the egg is flipped by a quick, sudden, forward jerk of the abdomen in the direction of the desired objective, and with surprising accuracy. Finally, a third group of genera seeks exposed hosts, such as lepidopterous larvae. Here belong the Systro- pinae, seeking larvae or pupae of the family Eucleidae. The Geroninae, which parasitize psychids and pyralids, are also properly placed in this group. Bee fly populations are supported by some families of Lepidoptera: in this order six families—Noctuidae, Psychidae, Eucleidae, Cossidae, Pyralidae, and Tortri- cidae—of widely div ergent habits are known to be parasitized by three subfamilies of bee flies. None of the diurnal Lepidoptera are included. 1. The family Noctuidae: cutworms, sod worms, and army worms. These species, the largely nocturnal larvae of the owlet moths, are exceptionally abundant in many parts of the world. They belong to the family Noctuidae of worldwide distribution and contain some 500 genera and more than 20,000 species (Essig, 1942). Maximum development appears in the Nearctic and Neotropical faunas (3,500 Nearctic species against 1,800 Palaearctic fauna (Imms, 1948). Walkden (1950), who reared several species of Poecilanthraz Osten Sacken from noctuid larva, obtained nearly 300 species of noctuids from light traps in Kansas. I have been astonished at the enormous numbers of these moths attracted to lights at wayside restaurants by night in August on Kansas highways. Walkden also reared large numbers of some 53 species of noctuids, and he suggests that powerful repressive factors evidently work against them since so few species are major pests. He found the egg count for some to range to more than 2,000, and many species with egg production of 600 to 800. Walkden classifies these into larval habit as: (a) subterranean, (b) tunnel makers, (c) surface feeders, (d) climbers, and (e) boring types. Of the species he reared, some in limited numbers, six were attacked by bombyliids; these bee flies consisted of three species of Poecilanthraz Osten Sacken and two unidentified bee flies. Moreover, all of the above eco- logical groups of cutworms were attacked by bee flies, except the boring types. Some interesting results appear from these studies. Of 53 species of noctuids he reared, many of them in considerable numbers, 27 species were attacked by hymenopterous parasites. Where these were identified, 18 species were attacked by an average of 3.5 species of Hymenoptera; 10 species were attacked by 2.7 species of tachinids; only 6 species of larvae were attacked by 1.3 species of bombyliids; 5 species of larvae were para- sitized by 1 species of bee fly. From 1 species, Choriza- grotis auxiliaris Grote, 3 species of bee flies were reared, which is perhaps not surprising, as this is a surface- feeding larva. The other genera attacked by bee flies consisted of 3 climbing species, Z'uawoa scandens Riley, Euxoa messoria Harris, and Peridroma margaritosa Haworth; the tunnel maker, Agrotis venerabilis Walker, and the subterranean A grotis orthogonia Mor- rison. The above data suggest that tachinids are more BOMBYLIIDAE / efficient parasites and Hymenoptera still more efficient than are the bombyliids, as far as the cutworm group of noctuids is considered. 2. The family Psychidae: caseworms and bagworms. Mik (1896) has recorded the parasitization of a Palae- arctic species of /wmea Stephens by bee flies of the genus Geron Meigen. /umea Stephens is rather excep- tional since in this species the female leaves the case during mating. The family Psychidae is large and dis- tributed throughout temperate and tropical parts of the world, the larvae feeding on many kinds of trees and shrubs. There are about 150 Palaearctic species. The species of Geron Meigen are numerous in the Nearctic region and so far scanty in the Old World and related genera extend down into South Africa. They are also present in Australia. One species of fossil bee fly from the Miocene has been tentatively assigned to Geron Meigen, and it is of interest perhaps that both the psychid moths and the eucleid moths, which are para- sitized by the almost worldwide Systropinae, are placed in the same superfamily, the Zygaenoidea of Graven- horst (Psychoidea of Imms). Both these bee flies have reduced venation. Until recently we did not know the host for any Nearctic species of Geron Meigen; Dona- hue (1968) notes that Solenobia walshella Clemens is a host of Geron calvus Loew. 3. The family Tortricidae. The bee fly Geron argen- tifrons Brunetti has also been recorded by Maxwell- Lefroy (1909) as attacking Laspeyresia Hiibner. 4. The family Pyralidae. Mik (1896) recorded the genus Geron Meigen as a parasite of the European pyralid Nephopteryx Hiibner. 5. The family Cossidae: carpenter moths or goat moths. A widely distributed, small, and apparently ancient family of wood-boring moths; both larvae and moths are usually of large or very large size. One genus and species of bee fly is known to live at the expense of these moths. Oldroyd (1951) has recorded, described, and illustrated a very large bee fly, Oestran- thrax goliath Oldroyd, from Malaya, which was reared from a cossid pupa. Many interesting questions might be raised as to how the fly reaches the larvae. He adds that the general appearance of the adult strongly sug- gests a nocturnal or crepuscular habit. 6. The family Eucleidae (Limacodidae of authors) : slug caterpillars. These makers of egglike cocoons rep- resent a family of some 850 or more species of wide- spread distribution, chiefly tropical, but found also in temperate areas of moderate rainfall. The bee flies of the genus Systropus Wiedemann, likewise found all over the world, are exceptionally beautiful insects and quite remarkable for their mimicry of the long, slender wasps of the genus Ammophila Kirby (Sphex Linné), and they have been taken from the genera Parasa Moore and Sibine Herrich-Schaeffer. It is a curious fact that the adults of Systropus Wiedemann, among the most beautiful of all bombyli- ids, seem to be rare. In 40 years of collecting at Oxford, Mississippi, I have taken but a single specimen of one species. In Mexico in 1962, however, with three pairs of eyes searching—my wife, my son, and myself—we found a single specimen at one locality, but before leaving Mexico, at milepost 1258 on the Mexican Pacific Highway, we found a large population, and each of us collected many mating pairs of two associated species. This was from shortly before noon to about 2:00 p.m.; Dr. R. H. Painter (1962) in the same summer also found two associated species in eastern Mexico, in smaller numbers. Schaffner (1959) in extensive rearing experiments with temperate Lepidoptera only obtained a single specimen of Systropus Wiedemann. Since Cockerell points to the marked scarcity of muscoid flies as fossils in the Colorado Miocene, it seems likely that tachinids have been steadily replacing bombyliids. Finally, I point to the total number of Diptera and Hymenoptera reared by Walkden from 19 species of cutworms. Of 10,108 larvae of these 19 species, mostly reared over a period of 12 to 15 years or more, approximately 523 were Diptera emergents, which were not broken down into respective number of tachinids and bombyliids, and approximately 788 were Hymenoptera of the several species. One species was attacked by as many as 6 species of tachinids; 1 was attacked by 13 species of Hymenoptera, and the maxi- mum number of bee flies from 1 species was 3. Four families of Coleoptera have been known to be attacked by bombylids. There is a single record by Hyslop (1915) of a bee fly, Villa alternata Say, which he obtained from a pupa of the tenebrionid Meracantha contracta Beauvois. Numerous larvae of this beetle had been obtained in rotting wood at the base of a stump at the top of South Mountain, Maryland; some of the larvae transformed to pupae but considerably more of them had been parasitized by a tachinid. Recently, Jack Hall recorded the parasitism of a cerambycid beetle by the bee flies of the genus Thevene- myta Bigot. Shelford (1913a) showed that tiger beetles of the genus Cicindela Linné serve as host for at least one species of bee fly, Anthrax analis Say. The soil-living scarabaeid population, however, sup- ports a greater number of bee flies. Forbes (1907) called attention to the larvae of Sparnopolius lhermi- nierti Macquart acting as an ectoparasite upon white grubs. Davis (1919) in an extensive study of enemies of white grubs (Phyllophaga spp.) noted that the scoliid wasp 7%phia Fabricius, abundant parasites of grubs of the genera Phyllophaga Harris, Ligyrus Bur- meister, Cyclocephala Latreille are in turn parasitized by the bee flies Hxoprosopa fascipennis Say and Exoprosopa pueblensis Jaennicke, as well as by Rhyn- chanthrax parvicornis Loew. Swezey (1915) found Chrysanthrax cypris Meigen attacking the scoliid wasp His Fabricius in Tlinois. Clausen (1928) in a remark- able study made in India found the closely related genus Ligyra Newman (as Hyperalonia Rondani), species oenamaus Rondani, parasitizing Tiphia Fab- 8 BEE FLIES OF THE WORLD ricius, Scolia Fabricius, and Campsomeris Lepeletier cocoons, Riley (1877, 1878) was the first to find bombylids living in egg cases of locusts. He published a series of seven papers on this subject from 1877 to 1881. In Russia, however, Stepanoy at very nearly the same time discov ered the relation of these flies to locusts and pub- lished his first paper in 1881 and two others in 1882. Also, Saunders, Waterhouse, and Fitch reported bee flies feeding on locust eggs in the Troad, specimens having been sent to London by Calvert. In the years immedi: itely following, several other European workers gave some attention to acridophile bombyliids, in par- ticular Kunckel in three works (1889, 1893-1905, 1894). After the turn of the century and recently still more attention has been paid to this group, because of the important effect of devastative plagues of locusts on man’s food crops. These findings will be reviewed later. No less than 13 genera in 5 subfamilies have the habit of consuming locust eggs. While some workers refer to this habit as predation, I here treat it as parasitism extended to the egg stage, inasmuch as only a single egg pod is consumed by a single fly; after the bee fly larvae has reached the second or third instar, it is too portly to go searching for a second meal ticket ; furthermore, bee flies have remarkable powers of maturing on small victims, as evidenced by the wide range in size of adults, a matter discussed elsewhere. The Diptera infrequently furnish hosts to this family. Austen (1914) noted that 7hyridanthrax Osten Sacken extends its host behavior to include tsetse flies, Glossina Wiedemann, in Africa, and because of its important significance in disease this relationship has been fol- lowed up with other studies. Several tachinids and even calliphorids furnish host relationship to bee flies, which thus become hyperpara- sites. Examples are: Zemipenthes morio Linné para- sitizng pupae of Puarasetigena Brauer and Bergen- stamm and “rnestia rudis (Fallén) (Baer, 1920), and according to Vassilew (1905) Hemipenthes morio Linné and Villa velutina Meigen (as Anthrax) para- sitizing Masicera silvatica (Fallén), a tachinid that attacks Lepidoptera. One Nearctic genus, Dipalta Osten Sacken, has been reared from the pupal cases of ant lions (Myrmeleon- idae). Four subfamilies and six genera of bee flies are known to depend upon Hymenoptera for hosts. These bombyli- ids attack a wide variety of solitary types of bees and wasps. Twenty-two hymenopterous genera have been noted to produce bee flies in rearing experiments. It should be further noticed that ZLigyra Newman and ELxoprosopa Macquart act as secondary or hyperpara- sites by attacking scoliids of the genera Tiphia Fab- ricius, Scolia Fabricius, Campsomeris Lepeletier, the sphecid Sceliphron Klug, and the psammocharid Pseudogenia. Also Chrysanthrax Osten Sacken has been reared from Z7is Fabricius. It is also most inter- esting that Finlayson and Finlayson (1958) reared Hemipenthes morio Linné from cocoons of the Euro- pean pine sawfly in Ontario, but again the bee flies, as parasites, were greatly outnumbered by tachinids and both of these dipterous families by Hymenoptera. The primary parasites in this group attack nests of mining bees or wasps in earth or cliff sides and also isolated nests of mud daubers and potter wasps. Hyperparasitism has developed, as far as is now known, only in the rather dominant subfamily Exo- prosopinae and to a lesser extent in the Anthracinae, which I believe should represent a tribe within the Exoprosopinae. Flies of the genera Hxoprosopa Mac- quart and Ligyra Newman attack scoliid cocoons of Tiphia Fabricius and other genera. Hemipenthes Loew is a hyperparasite of Ophion Fabricius and Banchus femoralis Thomson which themselves have attacked nocturnal Lepidoptera. Anthrax Scopoli and Hemipenthes Loew are hyperparasites of tachinid flies. Review of Published Information on Host Relationships (Abbreviations used for area: AU, Australian; ET, Ethiopian; HO, Holarctic; NE, Nearctic; NEOT, Neotropical; OR, Oriental; PA, Palaearctic.) BOMBYLIINAE: Bombylius Linné Species Hosts Authors boghariensis (PA) Not known canescens (PA) Odynerus spinipes Linne Chapman (1878) Halictus Latreille spp. Yerbury (1902) Pupa extruded from earth Imhoff (1834) *Andrena Dufour (1858) Andrena labialis (Kirby) Chapman (1878) minor (as pumilus Colletes daviesana Smith Neilsen (1903) Meigen) (PA) Epeolus productus Thomson (hyper- Lucas (1852) major (HO) parasite) vulpinus (as Panurgus dentipes Séguy & Baudot fugaxr) (PA) Latreille (1922) sp. Colletes fodiens Schmidt-Goebel (Fourcroy ) (1876) sp. Halictus farinosus Bohart (1960) in F. Smith Bohart, Stephen, & Eppley * Junior synonym of A. ovina Klug. Bombylius boghariensis Lucas. A pupa of this species was obtained in North Africa by Lucas (1852), who gave a fairly good, though diagrammatic figure of the stage, as well as the adult. He found these pupae emerging from the soil, but he was unable to learn the host. Bombylius canescens Mikan. Chapman (1878) ob- served a species, believed to be B. canescens, ovipositing freely in nests of Odynerus spinipes Linné upon a bank. Yerbury (1902) noted many individuals ovipositing in nests of several species of Halictus Latreille. This was done by sudden swoops or jerks of the abdomen; the egg was clearly seen as it left the insect but was never found. BOMBYLIIDAE 9 Bombylius major Linné. Though Imhoff (1834) de- scribed a pupa that he found extruded from the earth, Dufour (1858) was the first person to give much infor- mation concerning the life history of this species, the type-species of the family. Dufour found the species attacking Andrena nitidiventris Dufour. He described and figured both larvae and pupae. The nearly mature larva (third stage) still had thoracic bristles. Chapman (1878), during some excavation of earth in England, found a colony of Andrena labialis Kirby, among which he found a few overwintering pupae of this bee fly. The pupa was described but not illustrated. Chapman adds: “four present subject (B. major pupa) has to force a passage through the clay filling placed by the bee in its burrows, and to comb through 6 to 10 inches of these burrows. .. . though many pupae have to force their way through obstructions, this is the only one I know of actually provided with a mattock and shovel with which to do its own navigating.” In view of the abundance of this widespread, Hol- arctic species, which is extraordinarily common in Mississippi and other parts of the United States, as well as in Hurope, it is surprising that so little accurate information is known concerning the life history of Bombylius major Linné. Bombylius mexicanus Wiedemann. On May 28, 1964, the author found a considerable population of this large species ovipositing in dry soil about mesquite trees on the outer limits of Laredo, Texas. A sunken depression with a half-filled pond and old, undisturbed mesquite surrounding it had produced a wealth of both vegeta- tion and insects. This fly was found ovipositing both by jerky thrusts of the abdomen toward the soil near the base of mesquite and by distinctly inserting the abdomen into the somewhat friable and powdery soil. Bombylius minor Linné. J.C. Nielson (1903) gives a good account of the biology of this species as Bomby- lus pumilus Meigen. Austen (1937) in a footnote on p. 14 states: “Apud Engel, the subject of this paper was really 6. minor L.; B. pumilus does not occur in Den- mark, where Nielsen’s observations were made.” Neil- sen’s report, with the exception of Riley’s work on Aphoebantus Loew, was perhaps the best of its kind up until this date. He gives detailed figures from the larvae and pupae of the species, which are only slightly diagrammatic. These flies were found in the nest of Colletes daviesana Smith in Denmark. Neilsen also found them attacking H’peolus larvae, which had para- sitized Colletes, thus acting in this case as a secondary parasite. For a time the first-stage larvae feed upon the pollen stored up by the bee for its own brood, switching over later to a parasitic existence upon the bee grub. Bombylius vulpinus Wiedemann. Under the name of Bombylius fugaz Wiedemann, Séguy and Baudot (1922) describe the egg, larva, and pupa of this species, which they found parasitizing Panurgus dentipes Latreille. They give excellent figures of all of the Stages. Bombylius sp. Schmidt-Goebel (1876) found a species of Bombylius as a parasite of Colletes fodiens Fourcroy. Bombylius sp. Bohart (1960) in Bohart, Stephen, and Eppley notes that many two-year old larvae of Bombylius sp. have been observed parasitizing Halictus farinosus F. Smith. Allen and Underhill (1876) relate having found the young larvae of Apalus (as Sitaris) clinging to “hum- ble bee flies,” by which they assume the bee fly was in a definite way associated with nests of Anthophora Latreille; the species of bee fly was unfortunately not given, but the term used above suggests Bombylius Linné, but it may have been Anthrax Scopoli. BOMBYLIINAE: Systoechus Loew Species Hosts Authors Egg pods: acridophagus (as Locustana pardalina Potgieter (1929) albidus) (HTH) Walker Hesse (1938) aurifacies (ETH) Schistocerca gregaria Greathead (1958) Forskal autumnalis (PA) Dociostaurus (Staurono- tus) maroccanus Thunberg Several locust Stauroderus scalaris Stepanov (1881) Ingenitsky (1898) Troitsky (1914) (F.-W.) Gomphocerus sibiricus (L.) Pararcyptera microptera Vorontsovsky (F.-W.) (1926) Dociostaurus kraussi (Ingen. ) Dociostaurus albicornis ( Ey.) gradatus (PA) Stauronotum vastator Stepanov (1882b) Stev. Calliptamus italicus Linne Lindeman (1902) marshall (ETH) Acrotylus deustus Paramonov Thunberg (1931) Hesse (1938) Riley (1877, 1880) oreas (NE) Camnula pellucida (Seudder) Melanoplus spretus (Walsh) Melanoplus mexicanus Sauss., femur-rubrum De Geer, & differentialis Thomas Camnula pellucida (Seudder) & other Parker & Wake- land (1957) Treherne & Buckell (1924) grasshoppers socius (OR) Colemania sphenarioides Fletcher (1916) I. Bolivar somali (ETH) Schistocerca gregaria Hynes (1947) Forskal Greathead (1958) Dociostaurus maroccanus Thunberg, Gomphocerus sibiricus L. Gomphocerus sibiricus (L.) Stauroderus scalaris (F.-W.) Stauronotas vastator Stev. Stepanov Dociostaurus maroccanus (1882b ) Thunberg Paoli (1937) sulphureus (PA) Bezrukoy (1922) 10 BEE FLIES OF THE WORLD Species Hosts Authors Parker & Wake- land (1957) Melanoplus mevicanus Saussure Melanoplus femur-rubrum DeGeer Melanoplus differentialis Thomas Melanoplus differentialis Thomas Melanoplus bivittatus Say Melanoplus bilituratus Walker Syrbula admirabilis Uhler Melanoplus mexicanus vulgaris (NE) Painter (1962) Gilbertson & Saussure Horsfall (1940) verophilus (ETH) Locustana pardalina Potgieter (1929) Walker Hesse (1938) Research on Systoechus Loew and other genera that attack grasshopper egg pods have centered around ethological and quantitative studies by some workers, and taxonomic and morphological studies by others. This important aspect of bee fly behavior is usually listed as predatism, but I should prefer to think of it as ectoparasitic ovoparasitism, following Austen (1937). Some egg-destroying species of bee flies in other genera are apparently able to vary their host selection from locust eggs to caterpillars or to wasps. Systoechus acridophagus Hesse. This is one of the two species reared by Potgieter (1929), its biology is quite similar and its host species the same as that of Systoechus werophilus Hesse. Hesse (1938) illustrated and described the pupa. Some of the many South African species of Systoechus Loew are extremely large, with 35 to 40 mm. wing spread; their biology should be very interesting. Systoechus aurifacies Greathead. Greathead (1958) records this new species as attacking egg pods of Schistocerca gregaria Forskal in Eritrea. Systoechus aurifacies was found in a limited area of about 200 by 300 square yards. Of the 80 locust pods found, 22 were infested with a small number of larvae, the yield was 25 larvae of what proved to be this species. Greathead notes that the biology of S. aurifacies, which is appar- ently scarce, is very similar to that of Systoechus somali Oldroyd. Systoechus autumnalis Pallas. Stepanov (1881) bred this species from egg pods of Dociostaurus (Staurono- tus) maroccanus Thunberg in Russia. Ingenitsky (1898) bred it from several locusts in eastern Siberia; Troitsky (1914) records it from egg pods of Stawro- derus scalaris (F.-W.) and Gomphocerus sibiricus (1h.) in Siberia; Vorontsovsky (1926) records this species from Paracyptera microptera (¥.-W.), Dociostaurus kraussi (Ingen.), D. albicornis (Kv.) from western Siberia. Systoechus gradatus Wiedemann. Stepanov (1882) records this species under the name Systoechus leuco- phaeus Wiedemann with host egg pods from Stawrono- tum vastator Stev. and in 1881 from Dociostaurus maroccanus Thunberg in the Crimea. Lindeman (1902) reared it from Calliptamus italicus Linné in Russia. Systoechus marshalli Paramonov. This species was recorded from egg packets of Acrotylus deustus Thun- berg by Paramonoy in 1931. Systoechus oreas Osten Sacken. Riley (1877, 1880) was the first worker who discovered flies of this genus in their role of locust-egg destroyers. In his first annual report of the U.S. Entomological Commission he notes that numerous larvae, which he tentatively assigned to the Ichneumonidae but which later proved to be Systoe- chus Loew, were feeding within the egg pods. Riley states: “It exhausts the eggs, and leaves nothing but the shrunken and discolored shells.” Riley quotes a com- munication from S. D. Payne in Minnesota: “as I was strolling through the fields, I stopped to examine some eggs (locust). I found the ground in spots quite full of white grubs, worms or maggots. ... many of them were in the ege-pods, busy at work.” In 1880 Riley had realized the true identity of the maggots he had found in the locust-egg pods, and in an interesting report in the American Entomologist he gives the first published figure of the larva of Systoechus oreas Osten Sacken with details of head capsule, the pupa, and the imago. Further, he summarizes knowledge of bee fly histories at that time. In 1881 he gives a good description of the larva and pupa and more adequate figures of both this species and of Aphoebantus mus Osten Sacken (as Triodites). See also Parker and Wakeland (1957) under Systoechus vulgaris Loew. Systoechus somali Oldroyd. Considerable informa- tion was given on the bionomics of this species by Hynes (1947), a staff member of the Anti-Locust cam- paign in Italian Somaliland. In May and June of 1945 it was noticed that egg pods of locusts were being de- stroyed in many places by bee fly larvae, which proved to be this species. The host locust was Schistocerca gregaria Forskal. Egg packets of this species contain about 70 eges, and the number of larvae in each one ranged from 0 to 30, but averaged 10. Hynes states that they were all third-stage larvae and that no first- or second-stage larvae were found. The larvae in the pods at the time of collection ranged from 2 to 14 mm., which he was inclined to believe were not all of the same age. Growth was rapid; 10 or 11 days after the laying, a large proportion were full grown. Most of the larvae, he states, were found with their mouthparts applied to the middle of a locust egg with nearby eggs shrunken. He estimated that 5 or 6 larvae were necessary to destroy one locust-egg pod. He did not learn the fate of the larvae when excessive numbers were present in a single pod. The final size variation amounted to 6 to 16 mm. In a few days the translucent white of the larvae be- came opaque yellow; larvae burrowed a short distance from the egg packet, and came to rest in small ovoid cavities, similar to those described by Potgieter (1929) for Systoechus acridophaga Hesse (as albidus Loew) ; BOMBYLIIDAB iL they were found about 5 to 10 cm. below the soil level. A week after the locust had hatched, the soil was be- coming dry and caked and it remained so for months; the usual interval between rains is 4 to 5 months. An examination 2 months later showed the larvae slightly shrunken, active if disturbed, and resting in diapause. Although the irregular areas laid over by locusts amounted roughly to some 300 or 400 miles in each direction, the Systoechus were abundant only in a com- pact area about 120 by 160 miles, a situation maintained through two seasons. Hynes says that a staff member reported about 40 percent of the locust eggs were de- stroyed in soft soils and nearly 100 percent destroyed in hard soils. Hynes found that in very soft, sandy soils bee fly larvae were nearly absent. In late May the locusts had laid heavily, and in the harder soils he found 96 percent of the egg packets infested with an average of 4 larvae, but only 32 percent infestation in sandy soil, and these packets had only 1.6 larvae. In one place actually 100 percent of the eggs were infested with 18 larvae per packet. Hynes (1947) carried out laboratory experimentation which indicated the importance of rain as a factor con- cerned in the emergence of adults. Some 1,400 larvae in diapause were buried in dry soil within shallow metal pans with drainage. One pan was immediately watered heavily, and a few days later adults started to emerge. Another pan was watered lightly on the surface; an- other pan was kept air dry. The first pan was kept saturated. Emergences were recorded daily for over a year. Only 41 flies emerged, but at the end of the period many larvae were alive and in diapause; after 15 months 40 percent of the larvae were still alive in the dry cage, but no flies had emerged from it. Hynes con- cluded that ght rain was insufficient to break diapause. Oddly, Hynes, working with Systoechus somali Oldroyd, and Potgieter (1929), with S. acridophaga Hesse (as albidus Loew), and De Lepiney and Mimeur (1980), with Cytherea infuscata Meigen (as Glossista) , observe that even when conditions are favorable for emergence, some larvae remain in diapause. Larvae in diapause in dry soil, if wetted, swelled to original size, appeared to have a slightly reduced fat body, and if refusing to pupate, they could remain normal for a long time, move actively, and again form new soil cavities. Apparently emergences from pupae occurred only in the day time. Pupal period was short, but the larvae would not pupate in soil-filled tubes. Under laboratory conditions the pupal period was 9 to 15 days; for Systoechus acridophaga Hesse (as albidus Loew), Potgieter found the pupal period to be 7 to 9 days in the laboratory, and 14 to 23 days outdoors in the summer. De Lepiney and Mimeur record the pupal period of Cytherea infuscata (as Glossista) to be 10 days. Hynes quotes unpublished notes of Dr. D. L. Gunn on oviposition. Dr. Gunn states that the fly hovers over the hole in the sand left by a locust at a height of 1 to 3 cm.; an egg can be seen ejected from the extended ab- domen, which is curved forward, and a few grains of sand can be seen to move at the hole entrance; no eggs or first-stage larvae were recovered. The number of jerks of the bee fly ranged from 10 to 20 on an average, and to 40 in one case. Greathead (1958) adds considerable additional data on Systoechus somali Oldroyd beyond that supplied by Hynes (1947). He notes that Hynes’ larvae represented both second and third stages, and calls attention to the patchiness or irregularity of the distribution of the species. Greathead gives a table of the percentage parasitism for this as well as other parasites from 13 localities. He also gives a good figure of the larva of this species and a very valuable series of comparative figures of details on 4 species of Systoechus Loew larvae and pupae. Greathead notes that larval growth is quite rapid, in a few cases completed in about 4 days. More- over, he finds that there are usually not enough larvae in a pod to completely destroy all eggs; however, he quotes a personal communication from Ellis and Ashall who commonly found 20 to 40 larvae in an egg pod and who noted as many as 66 larvae in one egg pod. Great- head found that the larvae puncture an egg and suck out the contents, leaving the chorion dry and collapsed. This method does not result in the general putrefaction such as accompanies feeding of Stomorhina Rondani. Greathead states that Systoechus somali larvae sent to laboratories in England remained unchanged for as long as 3 years at temperatures of 25° to 30°C. Systoechus sulphureus Mikan. Recorded by Canizo (1944) from egg cases of Dociostawrus maroccanus Thunberg, Gomphocerus sibiricus (1), and Stawroderus scalaris (F.-W.). He gives a figure of the mature larva, and pupa with details of its armament. Paoli (1937) gives an excellent, illustrated account of the morphol- ogy of this species. Systoechus vulgaris Loew. Parker and Wakeland (1957) collected the data made in 1938, 1939, and 1940 from grasshopper-egg surveys. These studies were made in 16 study areas in the United States in the states of Arizona, California, Kansas, Minnesota, Montana, North Dakota, and South Dakota. This work assessed the average annual destruction of locust-egg pods by bee flies, as well as blister beetles and ground beetles. The bee flies included Systoechus oreas Osten Sacken, and S. vulgaris Loew, Aphoebantus hirsutus Coquillett, and A. barbatus Osten Sacken. Of these S. vulgaris was not only the most common, but also the most widely distributed species occurring in Michigan, Minnesota, Montana, Nebraska, North and South Dakota. A phoe- bantus hirsutus was next most frequently found, but limited to California and Oregon. These authors found that the average annual destruc- tion of egg pods amounted to 17.87 percent, of which bee flies accounted for 6.18 percent. The highest egg- pod destruction for a single year reached 77.52 percent in North Dakota. However, average egg-pod destruc- tion for many counties reached 50 percent or more, with bee flies taking a proportionately larger total. Criddle 12 BEE FLIES OF THE WORLD (1933) found that the combined efforts of Systoechus vulgaris Loew and other predators accounted for over 20 percent egg-pod destruction in the whole province of Manitoba in 1932. Shotwell (1939) in an account of grasshopper eggs destroyed in 6,277 fields in 11 states noted that bee flies per square foot of soil amounted to 0.31, which was higher than either blister beetles or carabids. Gilbertson and Horsfall (1940) in a study in South Dakota noted that of 59.3 percent destruction of Melanoplus mewicanus Saussure, bee flies accounted for 35.6 percent, again a greater proportion than meloid larvae. Parker and Wakeland (1957) discuss methods of taking data and results of findings by each state. Systoechus werophilus Hesse. This was one of two species of Systoechus Loew reared by Potgieter (1929) from Locustana pardalina Walker, the brown or Tek locust of South Africa. Its larva is figured by Hesse (19388). BOMBYLIINAE: Anastoechus Osten Sacken Species Hosts Authors baigakumensis Locusta migratoria Linné Zakhvatkin (PA) (1931) barbatus (NE) Melanoplus mexicanus Parker & Wake- Saussure land (1957) melanohalteralis Dissosteira longipennis Painter (1962) (NE) Thomas Locusta migratoria Linné Sacharov (1913) Locusta migratoria-gallica Roerich (1951) nitidulus (PA) Remaudiere Dociostaurus maroccanus Stepanov (1882) Thunberg Chinkevitsch (1884 ) De Lepiney & Mimeu (1930) Seabra (1901) Gomphocerus sibiricus Troitsky (1914) (L.) Stauroderus scalaris (F.-W.) Pararcyptera microptera (.-W.) Calliptamus italicus Linné Bezrukoy (1922) Moritz (1915) Anastoechus baigakumensis Paramonov. Zakhvatkin (1931) notes that he reared some individuals many years earlier from Locusta migratoria Linné. Anastoechus barbatus Osten Sacken. Parker and Wakeland (1957) give an extensive report covering the results of grasshopper-egg destruction from 7 states over a 3-year period. Their data include three other species of bee flies, besides meloids and carabids, with- out any separation of hosts attacked by the several predators. They list 13 species of grasshoppers, which were the dominant or second dominant species in the areas studied. They show very interesting maps of from 15 to 30 states for each year, giving the percentage of egg pods destroyed by all 3 major groups of predators. The greatest efficiency of destruction tended to lie in a north-south belt including Minnesota to Texas. An- astoechus barbatus was collected by the survey only in Montana. J/elanophus mexicanus Saussure was the dominant locust in all Montana areas. I have, however, found adults of A. barbatus exceptionally abundant in northern Arizona and in New Mexico west of Socorro. Painter (1962) found A. barbatus common in Wyoming; he gives a map of the distribution of this wide-ranging species. Parker and Wakeland conclude: “Without destruction of grasshopper egg pods by larvae of bee flies, blister beetles and ground beetles, and by other natural factors, the frequency of major outbreaks over extensive areas would be greatly increased.” Anastoechus melanohalteralis Tucker. Painter (1962) records having reared this species from egg pods of Dissosteria longipennis Thomas. Anastoechus nitidulus Fabricius. In a paper on the parasites and predators of Locusta migratoria Linné in Gascony, Roerich (1951) gives a brief discussion of this species with photographs of the larvae amid an egg case and of mature larva and pupa. He states that the larva consumes about 6 eggs. Stepanov (1882) and Chinkewitsch (1884) both record this species as prey- ing on the eggs of Dociostaurus maroccanus Thunberg in Transcaucasia. Zakhvatkin (1931) says that the species A. nitidulus is of little importance in Turkestan. Seabra (1901) found this bee fly attacking the same locust in Portugal, while Sacharov (1913) reared these bee flies from Locusta migratoria Linné in Astrachan. Troitsky (1914) reared A. nitidulus from Gomphocerus sibiricus (l.) and Stauroderus scalaris (F.-W.) in Siberia. Likewise in Siberia, Moritz (1915) found it attacking eggs of Pararcyptera microptera (¥.-W.). Finally, Bezrukov (1922) reared this species from the eggs of Calliptamus italicus Linné in Siberia. De Lepiney and Mimeur (1930) also recorded this species from Dociostaurus maroccanus Thunberg in the region of Guerrouaou and Djebel Guilliz (Rift). Thus, it may be clearly seen that A. nitédulus is a species of wide distribution, which feeds in the egg pods of at least 6 acridid hosts. BOMBYLIINAE: Sparnopolius Loew Species Hosts Authors Iherminierii (NE) Forbes (1907) Davis (1919) Malloch (1915) Phyllophaga Harris Sparnopolius lherminierii Macquart. Forbes (1907) found a larva of this bee fly attached to the back of a Phyllophaga grub. Davis (1919) believed that 8S. therminierti probably deposited its eggs in cracks within the ground on the chance of finding a scarabaeid grub. Although Davis found Sparnopolius Loew far from abundant, I have seen this species in profusion on a few occasions. One autumn I colleced about 150 flies in an hour upon a patch of Helianthus of about one-sixth of an acre in extent. Malloch (1915) described the pupa in detail. BOMBYLIIDAE 13 BOMBYLIINAE: Heterostylum Macquart Species Hosts Authors Bohart, Stephen, & Eppley (1960) Nomia melanderi Cockerell Nomia bakeri Cockerell Nomia triangulifera Vachal Nomadopsis anthidia Fowler Nomadopsis scutellaris Fowler Halictus rubicundus (Christ. ) Nomia melanderi Cockereli robustum (NE) Frick (1962) Heterostylum robustum Osten Sacken. An excellent account of the biology of this species was published by Bohart, Stephen, and Eppley (1960). This bombylid is a major enemy of the alkal bee Vomia melanderi Cockerell over a large part of its range, and the bee Nomia Latreille is considered an extremely valuable pollinator of alfalfa and in consequence responsible for the high yields of seed. They found that the percent- ages of host prepupae destroyed varied from 90 percent in some areas in Utah to as low as 5 percent in parts of Washington. In Oregon a nesting site with half a mil- lion bees had a parasite incidence of 91 percent in 1956. Before 1946 these authors state that several nesting sites contained over a hundred thousand nests; they quote Todd from personal communication, who de- scribed the ground as covered by thousands of empty bombyliid pupal skins. Two factors operate to maintain reduced bee populations; some of them nest away from congested areas; other late-emerging bees continue to lay eggs after most flies are gone. Other hosts included Nomia bakeri Cockerell, NV. triangulifera Vachal; Nomadopsis anthidius Fowler, N. scutellaris Fowler, and Halictus rubicundus (Christ.). The first bee flies appear 10 days earlier than the bees, but their emergence period overlaps the bee-emergence period by several weeks. Bee flies begin emerging at 8 a.m. and continue until 1 p.m, with the peak occurring between 9 and 10 a.m. After emergence the adults seek nectar before mating, and ovipositing then begins; by 2 p.m. most of the adults have sought shelter in sur- rounding vegetation in the deepest shade. Bohart, Stephen, and Eppley point out that dispersal must be great, for though several hundred flies may emerge from 2 or 3 square yards of soil, generally only 1 or 2 adults can be seen hovering over this area. Of 162 flies marked at emergence, only 2 were seen later. Caged flies were kept alive over a week. Blocks of soil with bees in them were moved a mile away from all nesting sites, the emerging bee flies killed, but new ones came in as soon as bee activity began. In oviposition, the female perceives a hole or crack in the ground, dips downward, gives the abdomen a down- ward flick, touches hind tarsi to the ground, hovers again for a second or two, releases another egg; after this has been repeated 5 or 6 times, H. robustwm is apt to rise up 5 or 6 inches and pursue her way, searching for more holes; the bee fly never alights during ovi- position. One female was observed to make 210 dips in 20 minutes, laymg an estimated 1,000 eggs during the day, even with occasional inactive periods. Twenty-five females dissected had from 44 to 424 eggs in the lower ovaries and oviducts, and a greater number undeveloped in the upper ovaries. Out of a series of vials, blackened and buried in the soil, 50 or more eggs were collected in a single day in each vial, these eggs were 1.2 mm. long, 0.7 mm. wide, oval, and tapered equally at each end; they had an ad- hesive coating. In the laboratory, incubation takes 8 to 11 days. First-stage larvae are planidial, 1.6 mm. long, nearly white, very active, and besides 8 pairs of thoracic bristles, they have a long, caudal pair also. Head and neck have 8 pairs of smaller bristles. These first-stage larvae progress much like those of Anthrax Scopoli, according to these authors, grasping the surface with their mouth hooks, making contact with posterior pseudopods, then pushing forward. Little is known of their early history, for they never made contact with the host until in its final mstar, and out of many cells examined, the few which contained planidia prior to such time showed that these minute larvae were either inside the pollen ball or on the wall of the cell against the ball. They surmise that the planidia can enter the cells through the entrance plug after it is sealed, and that this probably accounts for the occasional appearance of H. robustwm Osten Sacken larvae in early summer feeding on overwintered pre- pupae. Likewise, they suggest that early in the season the flies may flip eggs into reopened burrows of the pre- vious season, the planidia attacking hosts that are late in breaking diapause. They suggest that the planidia are able to survive for a considerable period and point to the existence of moderate parasitism in the bee Nomia triangulifera Vachal, which provisions its nests after the bee fly has virtually disappeared. These authors note that recently hatched bee fly larvae are not attracted to host larvae of any age, and confined with prepupae at room temperatures, they wander about and die in 24 hours, but live longer at lower temperatures. The question of what kind of food, if any, these planidia take in their long wait seems not to have been clearly settled. Bohart, Stephen, and Eppley noted that it was common to find 2 or 3 first-stage larvae on a single host larva, but only one made significant growth and moulted to the next stage; the fate of the other planidia is unknown, but there was no evidence of com- bat as in clerids or meloids. As in other bee flies, the first-instar larvae do not attach themselves firmly and often change position before starting to feed. Third-stage bee fly larvae of Heterostylum robustum Osten Sacken have the mouth tightly appressed against the host; this is believed to assist in the suction neces- sary to draw out the fluids of the host; the wound must 14 BEE FLIES OF THE WORLD be indeed minute. Fabre, with lens of limited powers, was unable to observe any mouth hooks and came to believe that suction alone was sufficient to withdraw the host fluids. He even tried to inflate the empty skins of host larvae, and, while he observed no leakage of air, we know now that the larvae do possess minute, knifelike mouth hooks as described under the morphology of larvae. It is a curious fact noted by both Fabre and Bohart, Stephen, and Eppley that the host larvae not only remain alive but maintain a fresh and healthy hue during the period of engorgement of the parasite. Eventually the host becomes shrunken and its final remains are a collapsed shell. For further comments see Anthrax trifasciatus Meigen. Bohart and his associates note that the fourth-instar larva will readily release its hold upon the host, if dis- turbed, but will soon again attach itself with the mavxillae until the mouth hooks have sunk into the host’s integument; they observed that the disturbed larva will just as readily attack a new host and will accept a wasp or hive bee larva with no hesitation. Still more interest- ing is the fact that several larvae placed together become cannibalistic, due doubtless to a very strong urge to feed. These fourth-stage larvae feed 3 or 4 days and almost double their length. It is peculiar and very interesting that in the Utah area the fourth-stage larva consumes all of one host and about half of another. As it leaves the first cell it plugs it tightly with dirt leaving host skin against the wall. It then burrows through the soil until it finds another cell with a prepupal bee; it requires about 2 days to consume half of this bee’s con- tents. In the Oregon area it appears that only one host larva was sufficient to bring the fly parasite to maturity ; as the parasite grub weighed 0.19 to 0.22 mg., and the host bee 0.126 to 0.198 mg., the margin is close. In over- wintering, Bohart, Stephen, and Eppley found that the full-grown larva burrows laterally or upward from the brood chamber, which may have lain buried some 5 to 10 inches, until it lies 2 or 3 inches below the soil sur- face; here it forms a large, oval, overwintering cell. This active behavior of the last instar of Hetero- stylum robustum Osten Sacken is very much in contrast to so many bee fly larvae, particularly the group attack- ing the Hymenoptera, which I have observed to be totally helpless and which never leave tight, tough co- coons of host. Bohart, Stephen, and Eppley note that a mature larva placed in plaster of Paris will hollow out a pupal cell; the Zeterostylum pupa is very active and restless, alternately extending and telescoping its ab- dominal segments. These authors found that the pupa will gyrate the abdomen, packing soil behind it, then gyrate the anterior end to loosen more soil in front of it, repeating the process until the surface is reached. Breaking the crust, the head is thrust free, and the ab- dominal segments are rotated rhythmically until both head and thorax are free. Breaking the pupal skin, it crawls out as the split extends backward; for a few minutes the abdomen and rear legs remain in the case; struggling free with its hind legs, it crawls to the near- est lump of dirt or other object and remains until wings expand and dry, releasing fluid from the anus. They state that in 10 minutes it is ready to fly. Bohart, Stephen, and Eppley have provided the interested stu- dent with 16 magnificent photographs depicting stages in the life history of Heterostylum robustum Osten Sacken. Frick (1962) made an extraordinarily fine study of ecological factors of Nomia melanderi Cockerell and Heterostylum robustum Osten Sacken in Washington State. Populations were sampled and percentage of parasitism of many lots was determined, the average percent being 12.1 in that area. Much seasonal emer- gence data were collected and a great deal of data ac- cumulated concerning the effect of temperature. High temperatures proved rather lethal to stages beyond the prepupa, and bee fly larvae had a threshold of develop- ment about 5°F. higher than bee prepupae. Soil mois- ture in the spring had a marked effect on both beginning date and rate of emergence. Efforts have been made by Bohart (1958) to reduce the number of bee flies with the use of methoxychlor. While bee flies average 169 mature ova, the alkali bee has only 25, but Frick (1962) observes that the shorter life span of the bee fly works in favor of the bee. Moreover, the bee fly uses only a small part of the day for oviposition, and indiserim- inate oviposition cancels out many bee flies, as Frick observes that they will even cast eggs at eyelets of shoes. Nye and Bohart (1959) present a remarkable photo- graphic record of the emergence stages of Heterostylum robustum Osten Sacken. BOMBYLIINAE: Callostoma Macquart Species Hosts Authors desertorum (PA) Dociostaurus maroccanus Portschinsky Thunberg (1894) Calliptamus turanicus Zakhvatkin Tarb. (1931) Dociostaurus kraussi (Ingen. ) Dociostaurus crucigerus- tartarus Stschelk. Dociostaurus albicornis turkemenus Uv., D. maroccanus Thunberg Schistocerca gregaria Forskal Ramburiella turemana Fr. W. fascipennes (PA) Dociostaurus maroccanus La Baume (1918) Thunberg Saunders, Water- house, & Fitch (1881) Austen (1937) Callostoma desertorum Loew. This species as re- ported by Portschinsky (1894) feeds on egg pods of locusts, probably Dociostaurus maroccanus Thunberg, in Turkestan. Zakhvatkin (1931) adds six species to the list of hosts. Zakhvatkin says of this species: “. . . this large, brightly colored fly is one of the most valuable BOMBYLIIDAE 15 insect enemies of D. maroccanus Thunberg throughout Turkestan.” He notes that it is a polyphagous species, feeding freely on eggs of several important xerophilus locusts. Egg laying occurs in the places where the locusts swarm; most of the eggs are laid in holes or fissures in the earth, with as many as 80 or 100 deposited simul- taneously, others are placed in pulverized soil by the abdomen of the female, with its excavating spines placed directly in the soil. Potential egg production ranges from 1,600 to 2,000. Primary larvae 2.5 to 2.8 mm. emerge in about a week. They hunt energetically through the soil in different directions, and on reaching an egg pod, they enter on the lower part. After 2 moults full growth ensues, and the larva leaves the egg pod for a narrow, elliptical hibernation cell. Pupation is in May, lasting 3 weeks. Average destruction of Docio- staurus maroccanus eggs is 20 percent, often much higher, and sometimes accomplishing almost complete destruction. Callostoma fascipennes Macquart. This fly has been known as an effective destroyer of locust eggs since 1881, when Saunders, Waterhouse, and Fitch recorded the species destroying locust eggs in great numbers in Cyprus. They indulge in some interesting, if perhaps estimated, statistics on the abundance of the species: “Biddulph’s dispatch informs us that from 5 to 8 per- cent of the locust eggs are this year devoured by these larvae. Since 800,000 okes of locust egg-cases have been destroyed in Cyprus this season to the end of October, it follows, from the lowest computation, that about eighty millions of our powerful natural allies—the bee flies— were associated with them, and must have been sacri- ficed if the destruction of the egg-cases took place before the larvae of the fly had left the cases.” In their Decem- ber 1881 statement they say: “Of the egg-cases received 1 0z., avoirdupois, contains 48 white-earth cases or 38 red-earth cases; say 45 cases average. An oke being 934 lbs. English (44 oz. avoirdupois), there would be thus about 2000 egg-cases to the oke, and if 5 percent were affected, one oke of these cases would contain about 100 Bombyliid larvae.” These authors give a good drawing of the imago, and for the times a good figure of the larva, head capsule, and pupa. La Baume (1918) discusses the biology of this species in Syria, and figures imago, larva, and pupa. The host for his material was Dociostaurus (as Stauronotus) maroccanus Thunberg. Austen (1937) gives a fine fig- ure of Callostoma fascipennes done by Terzi. He quotes Calvert (1881) : A grey pupa I was holding in my hand suddenly burst its envelope, and in half a minute on its legs stood a fly: thus identifying the perfect insect. ... I found the fly, now identified, sucking the nectar of flowers, especially of the pink scabious and thistle, plants common in the Troad. (Later on I counted as many as sixteen flies on a thistle-head.) The number of flies rapidly increased daily until the 13th, when the ground appeared pitted all over with small holes from whence the parasite had issued. A few pupae were then still to be found—a larva the rare exception. The pupal state thus appears to be of short duration. It was very interesting to watch the flies appearing above ground: first the head was pushed out; then with repeated efforts the body followed ; the whole operation was over in two or three minutes: the wings were expanded, but the colours did not brighten until some time after. Occasionally a pupa could not cast off its envelope and came wriggling out of the ground, when it was immediately captured by ants. Unfortunate flies that could not detach the covering membrane, adhering to the abdomen, also succumbed, as indeed many of the flies that could not get on their legs in time. BOMBYLIINAE: Cytherea Fabricius Species Hosts Authors holosericea (PA) Panurgus canescens Latreille (bees) Séguy (1930) Egg pods: infuscata (asa Dociostaurus maroccanus De Lepiney & Glossista (PA) Thunberg Mimeur (1930) obscura (PA) Dociostaurus maroccanus Stepanov (1881) Thunberg Stefani-Perez (1918) Paoli (1937) Calliptamus italicus Linné Zakhvatkin Calliptamus turanicus (1931) Tarbinsky transcaspia (as setosa) (PA) Cytherea holosericea Fabricius. As Chalcochiton holosericea, Séguy (1930) illustrates a pupa of this species taken from the nest of Panurgus canescens Latreille. This pupa is noteworthy for the fringe of exceptionally long hairs lying transversely along the margins of the abdominal segments. Cytherea infuscata Meigen. De Lepiney and Mimeur (1980) give a short account of the larval and pupal stages of this species, which they found destroying the ege pods of Dociostaurus maroccanus Thunberg in Guerouaou and Djebel Guilliz. They recorded this species as Glossista infuscata Meigen. Cytherea obscura Fabricius. Stepanov (1881) reared this species from locust eggs of Dociostaurus maroc- canus in the Crimea. Stefani-Perez (1913) likewise discusses the development of this species in Sicily and gives figures of the stages. Paoli (1937) in a fine series of three papers gives an account, not only of the biology of this species including all stages, but a very good study of the morphology of the larvae, its mouthparts, head capsule, sectional studies, and fine illustrations. Paoli’s third paper concerns the morphology of the female abdomen. Cytherea transcaspia Becker. As the synonym of this species (setosa Paramonov), Zakhvatkin (1931) states that this species is an important parasite on egg pods of Calliptamus italicus Linné in the Zeravshan Valley and likewise of Calliptamus turanicus Tarbin- sky, in nearby deserts. He finds that this bee fly de- stroys an average of 13 percent of the egg pods of both species and reaches a maximum at times of 40 percent destruction. He lists no less than four species of bee flies competing for the eggs of C. twranicus Tarbinsky ™ addition to a species of M/ylabris Muller. 16 BEE FLIES OF THE WORLD BOMBYLIINAE: Thevenemyia Bigot Species Hosts Authors sp. Phymatodes Mulsant unpublished In a personal communication, Mr. Jack Hall, Senior Museum Scientist, at the Riverside Banch of the Uni- versity of California, tells me that he has an unde- scribed species of Thevenemyia Bigot that was reared from the cerambycid Phymatodes Mulsant from a maple tree. This is a very important discovery. I have before me several pupae taken from the deadwood of Umbellularia californica Nuttall, at Los Gatos, Cali- fornia, from which Thevenemyia luctifer Osten Sacken (as auripilus Bigot) was reared. These were supplied by the courtesy of Dr. Krombein of the National Mu- seum of Natural History and Dr. Wirth of the U.S. Department of Agriculture. Hall (1954) obtained individuals of Thevenemyia Bigot which were reared from logs that contained anobiid beetle larvae, Ptilinus acuminatus Casey. Exact host was undetermined, but very possibly consisted of these beetle larvae. Hall (1969) notes that they have been reared on several occasions from wood, usually deadwood. He names also Ceanothus thyrsiflorus Eschscholtz and Pinus contorta Douglas, and also from “dry firewood” in Australia. Several persons, including myself, have noticed a propensity of the adults for fire- blackened trees, especially around Mt. Rainier. TOXOPHORINAE: Lepidophora Westwood Species Hosts Authors lepidocera (NE) Euodynerus foraminatus Krombein (1967) apopkensis Robertson Stenodynerus saecularis rufulus Bohart Trypargilum tridentatum archboldi Krombein Trypoxylon politum Say Spears, Sartor and Hull (in litt.) Podium rufipes Fabricius Unidentified spiders and cockroaches in wasp nests Krombein (1967) Lepidophora lepidocera Wiedemann. Krombein (1967) was able to rear four individuals of this remark- ably interesting bee fly in the Lake Placid and High- lands Ridge area of Florida. The host wasps proved to be Euodynerus foraminatus apopkensis Robertson and also Stenodynerus saecularis rufulus Bohart. The trap- nests were tied to the side of the dead trunk of a small tree, one placed on a dead stump, and others suspended beneath the limb of live scrub hickory and oak trees. The borings for these nests were 4.8 mm. and also 6.4 mm. Krombein found this genus to be unique because the bee fly larva fed only upon the olethreutid cater- pillars stored for the host wasp; it was also unique in that the content of several cells was necessary for the growth and maturation of the parasite. He noted that the adult fly eclosed a few minutes after one of the pupae wriggled free from the trap nest. He found a fifth nest, occupied by 7rypargilum tridentatum arch- boldi Krombein which was probably attacked by Lepi- dophora sp., but he was unable to rear the larva as far as the pupa; this larva had sucked dry all of the spiders that this wasp had stored in two cells. Krombein also reported (p. 255) the species lepido- cera Wiedemann, as appendiculata Macquart, parasitiz- ing Podium rufipes Fabricius at Lake Placid, Florida, and feeding upon the cockroaches stored by this wasp. Sartor, in Spears, Sartor, and Hull (in litt.), reports that he found a strange and curious larva in an un- plugged organ pipe mud nest of 7rypoxylon politum Say in Lafayette County, Mississippi; this nest was filled with caterpillars and was discovered beneath a bridge in a small valley known to have a high popula- tion of Lepidophora lepidocera Wiedemann (more than fifty adults have been taken in this restricted area). Without realizing the nature of his find and without illustrating its cephalic morphology he allowed it to pupate, and a very short time later a perfect adult eclosed from the pupa. TOXOPHORINAE: Toxophora Meigen Species Hosts Authors amphitea (NE) Pachodynerus erynnis Krombein (1967) Lepeletier Stenodynerus lineatifrons S. beameri Bohart S. saecularis rufulus Bohart Huodynerus megaera Lepeletier E. schwartzi Krombein Ancistrocerus campestris Saussure Eumenes fraternus Say Osten Sacken (1877) Osten Sacken (1877) Séguy (1926) leucopyga (NE) Eumenes fraternus Say maculata (PA) Humenes, Pelopaeus (=Sceliphron), Odynerus EHumenes pomiformis Fabricius Dianthidium dubium dilectum Timberlake Odynerus sp. pellucida (NE) Townsend (1898a ) Krombein (1967) virgata (NE) Stenodynerus toltecus Saussure Toxophora amphitea Walker. Krombein (1967) was able to rear more than twenty individuals of this species from vespid nests both in the Plummer’s Island, Mary- land, area and the Lake Placid, Florida, area. Most of his nests were suspended from limbs of living hickory in the sand-scrub areas of Florida; others on the side of standing dead tree trunks. The host wasps for these interesting flies proved to be: Pachodynerus erynnis BOMBYLIIDAE 17 Lepeletier, Stenodynerus lineatifrons Bohart, S. beam- eri Bohart, S. saecularis rufulus Bohart, E’wodynerus megaera Lepeletier, 2. schwartzi Krombein, and Anci- strocerus campestris Saussure. He estimated a life cycle of 31 to 35 days in Maryland. He obtained no informa- tion on duration of egg stage or elapsed time until the planidiiform larvae attack the host. His data suggest that amphitca males emerge 4 to 7 days earlier than the females. The planidia generally attached themselves transversely on the dorsum of the host prepupa on a segment near the head; they reached larval maturity in less than a week. The parasites were in random distri- bution within the cells of the host wasps. Dr. Krombein notes that Osten Sacken (1877) reported Glover had reared amphitea from a jug nest of Humenes fraternus Say feeding either on the caterpillars or the wasp larva. Toxophora leucopyga Wiedemann. Under the name T. fulva Gray, this species was reported by Osten Sacken as bred from a cell of Zwmenes fraterna Say. Toxophora maculata Rossi. Cros (1932) gives a very detailed description without illustration of the pupa of this species, which he reared from a larva taken in the nest of Humenes pomiformis F. Séguy (1926) illus- trates the immature stages. Toxophora pellucida Coquillett. Hurd and Linsley (1950) record rearing this species from a number of nests of Dianthidium dubium dilectum Timberlake, a megachilid bee. This bombyliid was one of five para- sites, and the material was collected in Santa Clara County, California. Toxophora virgata Osten Sacken. Townsend (1893) describes the pupal skin of this species in great detail; it was bred from a nest of Odynerus Latreille at Fort Collins, Colorado. Krombein (1967) records rearing this species from the nest of the vespid Stenodynerus toltecus Saussure in the Portal, Arizona, area, probably others of the same species and also several others from unidentified vespids in the same area. Nests were placed beneath a cedar limb, or on fence posts, or by bridges in this desert area. USIINAE: Usia Latreille Species Hosts Authors atrata (PA) Cataglyphis cursor (Fonscolombe) Xambeu (1898) Usia atrata Fabricius. The only record of this species concerns a larva and pupa which Xambeu (1898) said he had found in the vicinity of the nest of the ant Cataglyphis cursor (Fons.) upon which he believed the Usa larva had been feeding. He gives a brief de- scription of them. Some species of Usta are confined to the warmer parts of the Palaearctic region. It is difficult to understand how they could attack the ants themselves, but they might have a parasitic or scavenger role in connection with ant larvae as do Microdon Meigen species in the Syrphidae. P. du Merle (1971) reared Usta atrata Fabricius and U. ¢aenea Rossi from a tenebrionid beetle. CYRTOSIINAE: Psiloderoides Hesse Species Hosts Authors mansfeldi (ET) Locustana pardalina Walker Hesse (1967) Psiloderoides mansfeldi Hesse. Hesse (1967) states that this very remarkable fly was bred from the egg packets of the brown trek locust, Locustana pardalina Walker. These flies were obtained in the Kenhardt district of the North-Western Cape, South Africa. CYRTOSIINAE: Cyrtomorpha White Species Hosts Authors Egg pods: Austroicetes cruciata (Saussure) flaviscutellaris (AU) Fuller (1938b) Cyrtomorpha flaviscutellaris Roberts. Fuller (1938) described the mature larvae and pupae of this species but was unable to add any information on the younger stages or behavior of the adults. The larvae were found among empty egg pods of Austroicetes cruciata (Saus- sure) in a rather dense egg bed of this species in West Australia. In the laboratory she found that they re- mained as prepupa several months and sometimes for more than a year. She gives good illustrations of the larval mouthparts and the armature of the pupa, as well as the exceptionally minute spiracles of the larva. The larva and pupa are unusually strongly arched, a shape perhaps correlated with the unusually compact form of the adults. MYTHICOMYIINAE: Mythicomyia Coquillett Species Hosts Authors sp. Anthophora edwardsii Cresson unpublished In a recent communication Jack Hall, Senior Museum Scientist of the University of California, Riverside, notes that these little gnatlike flies have been reared from mud cells of the bee Anthophora edwardsii, and that numerous examples were recovered from a cell. This is a very important discovery. It is not difficult to see that such a large host could well provide food for a considerable number of these small flies. According to Hall the recovery of many individuals from a single bee cell suggests a gregarious condition of the larvae, a multiple parasitism. Such might result from these flies alighting directly upon the mud cells, or very close to them, and depositing a number of ova, resulting in many planidia. 18 BEE FLIES OF THE WORLD GERONTINAE: Geron Meigen Species Hosts Authors Maxwell-Lefroy (1909) Mik (1896) argentifrons (OR) Laspeyresia jacculatrix Meyer (Tortricidae) Nephopteryax subline- atella Strg. (Psychidae) Fumaria crassiorella Bruand (Psychidae) Solenobia walshella Clemens (Psychidae) gibbosus (PA) Mik (1896) calvus (NE) Donahue (1968) Geron argentifrons Brunetti. Maxwell-Lefroy (1909) reports the rearing of this species from the tortricid Laspeyresia jaculatrix Meyer in India. Geron gibbosus Olivier. Mik (1896) relates that O. Werner reared this species from caterpillars of the pyralid Nephopteryx in Dalmatia, and from /umea crassiorella Brd., a psychid, collected in Vienna. Geron calvus Loew. Donahue (1968) reared this species from a psychid, Solenobia walshella, in Mich- igan. W. C. Martin (1968) reared thousands of bag worms from North Mississippi without finding any Geron species. PHTHIRIGNAE: Phthiria Meigen Species Host Authors n sulphurea (NE) ? Romatlea Serville unpublished Data taken from flies reared from sand at Crescent City, Florida, from which Romalea Serville were emerging. From pupae supplied by Dr. Krombein of the National Museum of Natural History and Dr. Wirth of the U.S. Department of Agriculture. SYSTROPINAE: Systropus Wiedemann Species Hosts Authors barnardi (ET) bicuspis (ET) ? Parathosea sp. Stenomutilla beroe Peringuey Sibine bonaerensis Berg Hesse (1938) Bezzi (1924) conopoides (NEOT) crudelis (ET) Kunckel (1904) Westwood (1876b ) Hesse (1938) Eucleid caterpillar Coenobasis amoena Felder fumipennis Miresa clarissa (Stoll) (NEOT) macer (NE) Goncalves (1946) Prolimacodes scapha Harris Lithocodes fasciola H. 8. Limacodes sp. Adoneta spinuloides H. 8. Parasa urda Druce Sibine fusca Stoll undetermined eucleid on cherry in India Brooks (1952) Schaffner (1959) Walsh (1864) Lugger (1899) Seydel (1934) Dyar (1900) Clausen (1928) marshalli (ET) nitidus (NEOT) sp. (OR) Systropus barnardi Hesse. This author illustrates the pupa of this species, which was reared from cocoons of a eucleid, presumed to be Parathosea sp. Hesse describes and compares the pupa with that of Systropus crudelis Westwood; he notes that the cephalic ridge is much reduced and abdominal spines are longer, stronger, and reduced in number. Systropus bicuspis Bezzi. In his description of this Central African species Bezzi (1924) notes that the specimens are labeled: “Bred from cocoons of Steno- mutilla beroe Peringuey.” Systropus conopoides Winckel. This is a South American species from Argentina, which was made the subject of a short essay by Kiinckel (1904). The flies were reared from the eucleid (limacodid) Szbine bonaerensis Berg. Systropus crudelis Westwood. Westwood (1876), in his monograph of the eleven species of this genus known to him, describes the above species from reared material from Natal. While the host species, a caterpillar feed- ing on mimosa, was not named, the cocoons were of the eucleid type (limacodid), capsular in shape, and are illustrated by him, together with several views of the Systropus pupa. “No one,” says Westwood, looking at the robust pupa, “would have supposed that it could have produced such an elongated, slender imago as the Systropus.” Within each of the cocoons, a Systropus pupa was found, which was quite unlike that of other bombylid pupae. A strong conical projection was on its head, with which Westwood surmised the pupa was able to dislodge the operculum of the cocoon. On the underside of the head was a long appendage extending as far as the first ventral segment with the basal half grooved and with the remaining distal part jointed in the middle. Westwood suggested that the paired, basal parts were likely antennal cases and that the remainder represented the proboscis and sheath. On the robust, convex abdomen the first 7 segments bore a strong, short, curved bristle laterally, and the dorsal surface was provided with a transverse row of very short, fine spines. Westwood speculates as to the type of parasit- ism involved. No trace of the larval skin could be found in the cocoon : Was its larva an internal parasite like the larvae of Tachinae or was it external, like the larva of Scolia.... the latter seems to imply difficulties in the formation of a compact, oval cocoon, like that before us, by a caterpillar infested by an external parasite, unless we suppose that it was not until the cocoon had been formed, that the egg of the parasite hatched, so as to enable the parasitic larva to feed without hindrance upon its prey within the enclosed cell of the cocoon. Hesse (1938) records material reared from the co- coons of the eucleid Coenobasis amoena Felder feeding on Acacia. Systropus fumipennis Westwood. Goncalves (1946) recorded this South American species bred from J/zresa clarissa (Stoll). Systropus macer Loew. This species has been reared on several occasions. Walsh (1864) reared this species, which he mistook for a conopid, from undetermined eucleids found on oak. Lugger (1899) also reared the species. Schaffner (1959) in rearing many thousands BUMBYLITDAE 19 of Microlepidoptera only chanced upon a single individ- ual of this species, which came from the host Lithacodes fasciata H.-S. in Connecticut. He believed that there were at least two generations per year. He felt that the hibernation was probably in the puparium. Brooks (1952) reared one individual in Ontario from Prolimacodes scapha Harris, which emerged in April 1944, but this date probably bears no relation to the natural date of emergence. Brooks illustrates the ante- rior and lateral aspects of the head of the pupa and the last segment of the abdomen; these figures show it to be very different from the African species such as Systropus crudelis Westwood. Systropus marshalli Bezzi. Seydel (1934) records a long series of 65 individuals of this species reared from cocoons of Parasa urda Druce. These he compared with Belonogaster griseus Fabricius as a probable mimetic model. LOMATIINAE: Aphoebantus Loew Species Hosts Authors hirsutus (NE) Camnula pellucida Wilson (1936) Scudder mus (NE) Camnula pellucida Riley (1880, Scudder 1881, 1882) Lemmon (1879) sp. (OR) Tiphia grubs (secondary Clausen (1928) parasite) Aphoebantus hirsutus Coquillett. A remarkable study of this species was made by Wilson (1936). He re- corded results of control measures appled against an outbreak of Camnula pellucida Scudder in northern California in 1927; by using poisoned bran plus the natural enemies, the breeding ground in the Tule Lake area was reduced from 5,000 acres of dense egg beds to local and sparse beds in 1929. Wilson describes his method of sampling the Aphoebantus Loew population in some detail. The method of mating is especially interesting; he states that the male inserts its abdomen into the soil beside that of the ovipositing female—as many as 10 males may attend one female. The congre- gation of mating flies helped locate egg beds. A count of 104 locust capsules showed 2,886 eggs, an average of 27.75. Heavier populations of both were encountered in moist peat or sandy loam soils with sparse vegetation. A tendency of migration by the locusts to and from selected oviposition beds was noted. Oviposition beds, once established, become permanent. The A phoebantus Loew attacked from 0.7 to 62.0 percent of the grass- hopper-egg capsules in the egg beds of the Tule Lake district and constituted an important control factor. Aphoebantus mus Osten Sacken. Riley was the first to discover that these small bee flies live in the egg pods of locusts during their larval life. In 1881 he describes the large, fat, white larva that J. G. Lemmon sent to him from the Pacific Coast. He quotes Lemmon (1879) : “We don’t know certainly what this larva becomes, but at a venture he is hailed with great joy. The ground that was first filled with locust eggs by the end of September looked as if scattered with loose shells, so thorough was the work of destruction.” Riley (1881) gives a description of the stages and illustrates details of the species, which used the locust Camnula pellucida Scudder for its host. Lemmon (1879) recorded his experiences with the locust plague and its predators in the Sacramento, California, Weekly Record-Union, for November 29: “the grubs ate out and destroyed thou- sands of eggs last fall, ... having lain dormant all winter, and being now found still among the eggs, which are fast hatching out (June).” Aphoebantus sp. Clausen (1928) records for the first time an undetermined species of this genus attacking Tiphia Fabricius in Assam, India, in the role of a secondary parasite. LOMATIINAE: Lomatia Meigen Species Hosts Authors hamifera (PA) Schistocerca gregaria Séguy (1932) Forskal quoting Régnier et al. Régnier, Lespes, and Rungs (1931) record the ob- taining of Lomatia hamifera Becker from egg pods of Schistocerca gregaria Forsk. (From Séguy 1932). ANTHRACINAE: Anthrax Scopoli Species Hosts Authors albofasciatus (NE) Tachysphez terminatus analis (NE) anthrax (PA) argyropyga (NE) Smith Cicindela scutellaris Say subsp. lecontei Hald. Hoplomerus spinipes Linné Osmia rufa Linné Eumenes unguiculus Villeneuve Megachile sp. Odynerus sp., and Osmia sp. Chalicodoma muraria Fabricius Odynerus spinipes Linné Chalicodoma muraria Fabricius Ophion sp., and Banchus compressus (Fabricius) Trypargilum tridentatum archboldi Krombein Trypargilum c. rubro- cinctum Packard Trypargilum striatum Provancher Ancistrocerus c. catskill Saussure Stenodynerus f. fulvipes Saussure S. pulvinatus surrufus Krombein Marston (1964) Shelford (1918a) Verhoeff (1891) du Buysson (1888) Bugnion (1886) Giraud. See Brauer (1883) Giraud. See Brauer (1888) Jacquelin du Val (1851) Laboulbéne (1858a ) Lampert (1886) Lassman (1912) Krombein (1967) 20 BEE FLIES OF THE WORLD Species Hosts Authors Species Hosts Authors S. saecularis rufulus Sceliphron caementarium Bohart Drury S. beameri Bohart Chalybion californicum Pachodynerus erynnis Saussure Lepeletier Anthophora sp. aterrimus (NE) Trypargilum striatum Krombein (1967) limatulus Dianthidium curvatum Custer (1928) atripler (NE) binotatus (PA) caffer (ET) cintalapa (NE) diffusus (BT) distigmus (PA) fur (NE) irroratus (NE) isis (PA) jazykovi (PA) limatulus s. str. (NE) Provancher T. clavatum Say T. collinum rubrocinctum Packard Tsodontia auripes Fernald Euodynerus megaera Lepeletier Monobia quadridens Linné Ancistrocerus spinolae Saussure Bembiz sp. Trypoxrylon politum Say Sceliphron caementarium Drury Megachile gentilis Cresson Chalicodoma muraria Fabricius Ceratina nasalis Fr. Megachile concinna Smith Megachile sp. Megachile nipponica Cockerell Odynerus mikado Karsch Rhychiwm mandarineum Saussure Sceliphron tubifer Latreille Trypoxylon obsonator Smith Solenius spp. (7 unnamed spp.) Megachile sculpturalis Smith Sceliphron caementarium Drury Chalybion californicum Saussure Rygchium sp., Osmia sp. Rau, 1946; Spears, Sartor, and Hull (in litt.) Krombein (1967) Schaffer (1764) Frauenfeld (1861) Hesse (1956) Butler and Ritchie (1965) Hesse (1956) Iwata (1933) Marston (1964) Rau (1940) Osten Sacken (1887) Megachile gentilis Cresson Krombein (1967) Megachile mendica Cresson Dianthidium heterulkei fraternum Timberlake Ashmeadiella bucconis denticulata Cresson Hylaeus asininus Cockerell and Casad Egg pods of Dociostaurus maroccanus Thunberg Prepupal larvae of Epicauta erythrocephala Pallas feeding on Calliptamus italicus Linné Anthophora abrupta Say Trypoxylon politum Say La Baume (1918) Zakhvatkin (1981) Frison (1922) Spears, Sartor, and Hull (in litt.) artemesia (NE) limatulus columbiensis (NE) limatulus larrea (NE) limatulus vallicola (NE) monachus (PA) nidicola (NE) oophagus (PA) tigrinus (NE) trifasciatus (PA) trifasciatus leucogaster (PA) trimaculatus (NB) zonabriphagus (PA) sayi Cockerell Sceliphron caementarium Drury Chalybion californicum Saussure Trypoxryton texrense Saussure Anthophora occidentalis Cresson also reared from cells of Antho- phora parasitized by Melecta californica miranda Fox Anthophora sp. Anthophora sp. Anthophora flexipes Cresson Anthophora linsleyi Timberlake Mylabris scabiosae scabiosae in egg pods of Ramburiella turcomana F. W. Diadasia consociata Timberlake Diadasia enavata Cresson Dociostaurus maroccanus Thunberg Dociostaurus kraussi Ingenetisky Dociostaurus crucigerus tartarus Stechelk Calliptamus turanicus Tarbinskii Ramburiella turcomana Fischer von Waldheim Xylocopa virginica virginica Linné Xylocopa californica arizonensis Cresson Xylocopa tabaniformis orpifex Smith Xylocopa augustii Lepeletier Chalicoderma sp. Cemonus sp. Diadasia sp. In locust egg pods feeding on other bombyliids and on Mylabris sp. Marston (1964) Marston (1964) Marston (1964) Marston (1964) Torchio and Youssef (1968) Marston (1964) Zakhvatkin (1931) Linsley, Mac Swain, and Smith (1952, 1957) Cole (1952) Zakbvatkin (1931) Angus (1868) ; Rau (1926) Hurd (1959) Davidson (1893) Nininger (1916) Hurd (1959) Fabre (1879) Frauenfeld (1864) Marston (1964) Portschinsky (1895) Troitsky (1914) Anthrax albofasciatus Macquart. Marston (1964) figures a pupae of this species sent to him from New England, which had been taken from cells of Vachys- phex terminatus Smith; females were seen ovipositing in the depressed areas resulting from the closure of BOMBYLIIDAE a1 cells; and pupae, as well as shed pupal skins, were found on the sand of the same spot in the spring. Anthrax analis Say. This species selects a host quite unlike that of all other known members of the genus. Shelford (1913) has given a fine account of the biology of the species and its relation to tiger beetles of the genus Cécindela Linné. Oviposition flight is described as taking an irregular zigzag flight about 2 inches above the sand until the fly passes above a hole in the sand; as this happens the fly suddenly halts, moves backward and downward so as to touch the sand 5-10 mm. from the hole; some sand moves into the burrow; such thrusts are repeated a number of times. Shelford notes that the host larvae may appear at the surface during this time, and on two occasions the fly stopped ovipositing when it did so. By using gentle pressure on a female bee fly Shelford was able to squeeze out a large number of eggs; they were light brown ellipsoids, 0.28 by 0.12 mm. in size and not adhesive. Young larvae were most commonly found singly on the ventral side of the third instar of the host larva, where they clung between the legs. He found no second-instar larvae with parasites; here they cannot be reached by the host and do not easily come into contact with the burrow. He noted that not infrequently hosts had more than one parasite be- tween the legs, or on other parts of the body. Shelford (1918) was unable to learn how the bee fly larvae reach their host; he did note that host larvae, dug from a spot where a fly was seen depositing eggs on July 16, had parasites of first and second instars when removed from the burrows on September 23. He noted that an average of about 7 percent of the host larvae had parasites; some collections showed a para- sitization as high as 16 percent; in pine areas only one out of several hundred larvae had a bee fly parasite upon it. The head segment of the larvae bears long, curved mandibles according to Shelford; the smallest larvae found were 0.5 to 0.6 mm. in length and were taken in late summer and autumn, occasionally in spring. Most of these larvae moulted in the fall, passing the winter attached to the host; he was not sure when the second moult took place, but believed the host had fed about a month in early June in the Chicago area. A third moult took place about the time the host stopped feeding, and in the observed cases before the pupal cell was constructed. Host pupation, Shelford found, was delayed for about a month after the cell was constructed. The parasite did not grow rapidly until the host had been in its cell about 3 weeks; by this time the old organs of the host had broken down and the internal parts were in a semifluid condition. At this time the host larva is torpid and helpless; the parasite shifts its position to the middle of the ventral side of the host; tapping food here, it grows very rapidly, increasing in length from 4.5 mm. to 10 mm. in 48 hours. Parasite growth is completed in 144 hours, its length becoming 1.8 cm., all this later growth is without fur- ther moults; it remains 6 or 7 days in a quiescent state before pupating. Shelford (1913) describes the pupa and gives an excellent figure; he describes the digging motion of the larvae that he confined to a wide glass tube filled with sand. The average rate of progression was 1 cm. per hour; the path traveled may take the pupa an extensive distance as it works its way from the pupal chamber in a devious path to the surface. Shelford found the para- sitism of this species almost confined to Cicindela scutellaris lecontei Haldeman. Anthrax anthrax Schrank. This European species, the type-species of Anthrax Scopol, was reared from a nest of Humenes unguiculus Villeneuve, by Bugnion (1886), who reported it as Anthrax sinuata Fallén (lapsus for Meigen). Percheron, as early as 1835, com- mented briefly upon a pupa of this species, also under the name A. s¢nwata. About the same time de Buysson (1888) reported that he found the same species (again reported as A. stnuata Fallén) in the nest of bees of Osmia rufa Linné. Von Roser (1840) obtained the species from nests of Anthophora Latreille. Laboul- béne (1858) obtained this species from nests of Ody- nerus spinipes Linné. In 1857 he describes a pupa and reviews knowledge of bee fly life histories up to this date, pointing out that several dipterists( Zetterstedt, 1842; Walker, 1851) considered them as living in roots of plants. Lampert (1886), in a review of parasites of the wall bee Chalicoderma muraria Fabricius, states that he has reared many Anthrax anthrax (as A. sinuata Meigen) from the larvae of this bee. Lampert also reared no less than nine parasites from these wall bee nests, which shows that the bombyliid has consider- able competition. Verhoeff (1891) found the young larvae of this species. Anthrax argyropyga Wiedemann. Krombein (1967) in a unique study of trap nesting of bees and wasps carried out over a period of twelve years reports finding Anthrax argyropyga Wiedemann and three other species of Anthrax Scopoli as parasites of several species of wasps and bees. The species argyropyga Wiedemann was obtained from the following species of wasps: Zrypargilum collinum rubrocinctum Packard, T. striatum Provancher, 7’. collinum Smith, 7. triden- tatum archboldi Krombein, Stenodynerus fulvipes Saussure (parasitism ranging as high as 40 percent of available cells), S. pulvinatus surrufus Krombein, S. saecularis rufulus Bohart, S. beameri Bohart, Pacho- dynerus erynnis Lepeletier, and Ancistrocerus collinum catskill Saussure. Krombein found that about 28 days were required between attachment of the first-instar larva and the emergence of the adult fly. He believed that about 2 weeks represented the elapsed time between hatching of the egg and attachment of the young larva. He notes that it is entirely possible for the first-stage larvae to remain in this instar as long as 2 months if a suitable host is not available. This species attacks the host pre- pupa or pupa, and 8 days were required to exhaust a host. This species occurred in the company of Anthrax 22 BEE FLIES OF THE WORLD aterrimus Bigot in one instance. They were likely to occupy any of the cells in series in a nest. His research supplies considerable additional seasonal data. Schmidt and Hull obtained this bee fly from a trap nest containing an unidentified wasp in Mississippi. Anthrax aterrimus Bigot. Krombein (1967) in his remarkable studies of trap-nested bees and wasps found this species living in the nests of the following wasps: Trypargilum striatum Provancher, 7’. clavatum Say, T. collinum rubrocinctum Packard, Huodynerus mega- era Lepeletier, Ancistrocerus spinolae Saussure, and Monobia quadridens Linné, and [sodontia auripes Fern- ald. They were all from shaded settings in open woods. In many of the nests the bombyliids overwintered as diapausing larvae, which coincides with the findings of Spears, Sartor, and Hull (in litt.) in Mississippi. Krombein supplies much additional data on the ethol- ogy of this species and notes that an example has been also reared from /sodontia philadelphica Lepeletier. Rau (1946) records this species as Anthraa slossonae Johnson, as a parasite of Z’rypoxylon politum Say. Anthrax atriplex Marston. Krombein (1967) reports that he reared this species from the cocoons of the bee Megachile gentilis Cresson from Arizona. The bee fly larva fed on the bee prepupa in the cocoon. In another case it fed upon the resting bee larva. The trap nests were placed beneath dead limbs of mesquite in open desert. Anthrax binotatus Wiedemann. Under the names Argyramoeba subnotata and A. binotata this species has been bred by several persons from larvae of the Euro- pean wall bee Chalicodoma muraria Fabricius, of which the earliest were Schaffer (1764) and Frauenfeld (1861). Anthrax caffer Hesse. This species was bred from the nests of the bee Ceratina nasalis Fr. and reported by Hesse (1956), who figured the anterior horns of the pupa. Anthrax cintalapa Cole. Butler and Ritchie (1965) found that the bee fly Anthraa cintalapa Cole para- sitized the bee Megachile concinna Smith in both Ari- zona and California. Anthrax diffusus Wiedemann. Hesse (1956) figures the pupa and caudal spines of this species, the material being bred from clay nests of a Megachile sp. nesting in sand. Anthrax distigmus Wiedemann. Iwata (1933) pro- vided nesting tubes on a shelf in his house in Japan; he states that Anthrax distigmus visited the shelf every clear day, especially around noon in the summer of 1931. The bombyliid alighted on the nest tube, touched the entrance floor with the tip of the abdomen, and flew away, but soon returned to repeat the same procedure. Although Iwata could not find eggs, his nests proved to be heavily parasitized. He states that the active, hatched larva, finding a host larval cell enters it, at- taches itself onto the larva, and remains “inoffensive” until the larva finishes the spinning of its cocoon and becomes dormant; the bombyliid then begins to suck in the body fluid of the host larva. Iwata (1933) noted that if the host was small, the bee fly larva consumed all, leaving only the skin; if large, as in the case of Megachile sculpturalis Smith, it reached maturity, leaving half of the prey unconsumed. When Anthrax distigmus attacked MW. sculpturalis there was only one brood a year, but if it attacked wasps and bees with two or three generations, then the bee fly like- wise went through two or three generations. It is espe- cially interesting that Iwata found the bee fly larva overwintered generally in the mature larval state, but not infrequently overwintered in the first larval instar. He noted that the pupa uses its strong spiny cephalic and caudal armature to pierce the partitions of resin or mud and works its way even from as many as nine cells to the entrance. The bands of hooks and setae on the abdomen support the pupa securely at the entrance as it emerges. Iwata found that the larva does not excrete in the larval period, the meconia is discharged for the first time after emergence, the first is white, the second black. Iwata lists other hosts from which he reared Anthrax distigmus. Anthrax fur Osten Sacken. Marston (1964) found that Anthraw fur overwintered mostly as full-grown larvae; he found that in some cases the first-stage larvae remained as such through the winter; he found two such larvae on Rygchium larvae on April 19, and an- other on a Sceliphron Klug larva. Marston introduces the interesting speculation that the host larvae produces some substance in its body that is voided about the time it deposits its fecal pellet, and that such substance is necessary to stimulate the moulting of the parasite larva to the second and actively feeding stage; it could also operate as a retained substance in the host larvae, which no longer effects retarded feeding, once it is eliminated. Marston (1964) found that about 7 days before pupation, the bee fly larvae elongates and begins to “arch” itself in the characteristic shape of the pupa; pupal spines can be seen through the hyaline integument of the larva, and the pupa, at first transparent, as it emerges by contractions of the body, quickly becomes opaque yellow. This yellow color changes and darkens some 6 or 8 days prior to emergence or even sooner, de- pending on temperature. Its final preemergence color is quite black. On emergence, the wings expand in 2 min- utes. Marston found that it required about 2 hours before it was ready to fly. It is of particular interest that the pupa must be anchored for successful emer- gence; this accounts for the fact that pupal exuviae are so frequently seen attached to and extending half- way out of nest entrances with the thoracic and anterior abdominal segments free. Under the name Argyramoeba fur, Rau (1940) re- ports a bee fly found heavily infesting nests of Mexican mud daubers in Mexico, 40 km. south of Victoria. He also reports that he found these same bee flies in mud dauber nests from Franklin, Texas. In 1964 Hull reared BOMBYLIIDAE 23 this species from a mud dauber nest beneath a bridge north of Laredo, Texas. Anthrax irroratus Say. Rau (1940) records having taken a dead bee fly from a mud nest of a wasp in February from Canyon de Galeana, Mexico. He called this bee fly Spogostylum sp. near oedipus Fabricius. Townsend (1893) gives a very full description of the pupa of Anthrax irroratus under the name of Argyra- moeba oedipus Fabricius. This species had been bred * from a nest of Odynerus sp. Brooks (1952) reared in- dividuals from Megachile nivalis Fries in Saskatche- wan. He describes the pupa and illustrates details of the armament. Krombein (1967) provides considerable data on the ethology and hosts of this widespread species. At Scottsdale, Arizona, he reared it from the following hosts: the megachilids, Megachile gentilis Cresson, Dianthidium heterulket fraternum Timberlake, Ash- meadiella bucconis denticulata Cresson ; the colletid bee Hylaeus asininus Cockerell and Casad and an unidenti- fied vespid wasp also in Arizona. He gives data on the pupal period and notes that this bee fly may overwinter either as an egg or an unfed planidial larva in the nests of Hylaeus and Ashmeadiella or as a diapausing larva in the Deanthidium nests. Other records for this species consist of Megachile nivalis Friese listed by Brooks (1952) in Saskatchewan, Baker (1895) for an odynerid type wasp in Colorado, and Cooper (1954) reared this species from H'wodynerus foraminatus Saussure (under the name Rygchium ru- gosum Saussure). Anthrax isis Meigen. In 1918 La Baume recorded this species as destroying the eggs of Dociostaurus maroccanus Thunberg in Syria, in the vicinity of the Euphrates in Wilajet Aleppo. La Baume further states that two examples of another, undetermined, species of Anthrax were also taken from similar egg pods. Anthraxw jazykovi Paramonov. Zakhvatkin (19381) says that the larvae of this species inhabit the egg pods of Calliptamus italicus Linné and are in these egg pods parasitic upon prepupal larvae of H’picauta erythroce- phala Pallas. Second-stage larvae were found feeding upon them in May, the full-grown larvae not uncommon in June and July. Pupation occurred in July and emergence in August in the Zaravashan Valley. Be- tween 800 and 1,000 eggs were laid. Hibernation is presumed to be as first-stage larvae. Anthrax limatulus Say. Marston (1964) has given an extended account of the biology of Anthrax limatulus Say and its several subspecies. This author has reduced Anthrax fur Osten Sacken to subspecies status of Anthrax limatulus Say, but in spite of the fact that the pupal spines are alike, I am not satisfied with this status. I therefore leave Anthraw fur Osten Sacken as a distinct species characterized by its definite range, unique pattern and coloration, and other particulars. The author has observed that Anthrax fur extends far- ther west, and rearings show that quite distinct species of Anthrax do overlap and do compete for available wasp’s nests for parasitism, in which numerous com- plex factors doubtless operate. Marston states that his reared individuals of Anthrax fur presumably para- sitized Sceliphron caementarium Drury and Chalybion californicum Saussure (Specidae). He also reared this species from the vespid Rygchium sp. and the mega- chilid bee Osmza sp. Females of Anthrax limatulus were observed to ovi- posit August 18 and 19; one of these began ovipositing in an old abandoned nest. This bee fly, close to the bridge ceiling, kept the body at an angle and when about 3 inches away from a site flipped the egg with a quick jerk of the abdomen. Marston points out that the egg has an adhesive coating that adheres to whatever it touches. Many of the oviposition sites were small holes in the concrete. Marston found that the egg- laying stimulus seems to be black or dark colored spots and noted that the bee fly even pelted eggs against the black spots on the wing tip of a resting moth. Egg- laying periods lasted from 1 to 4 minutes, the fly going away to rest nearby between layings. Marston collected some 400 eggs from six sites by means of an aspirator; 100 were taken on September 7. Marston (1964) points to the evident high mortality of the young larvae, which are seemingly unable to use the 3 pairs of thoracic hairs with the complete efficiency of legs; he introduced as many as 5 young planidial bee fly larvae onto a single bee grub. Only 1 larva survived and Marston conjectures that this may be due to canni- balism. Incubation period of the eggs ranged upward to 18 days or more; eclosion is by means of mouth hooks from the end of the egg, the hole enlarging as the larva forces its way through, dragging the shell, until, be- cause of wedging, it is forced off. First-instar locomo- tion was found by Marston to be caterpillar-like, an alternate extension of thoracic segments, attachment with mandibles, drawing up the remainder of the body in an undulatory fashion, and taking hold with proleg- like pseudopods. The first-instar larva remains within the cell for about 4 days, then attaches itself between first, second, or third thoracic segments, feeding irregu- larly ; with restless periods it moves about over the hosts back; it moulted to the second instar about 9 days after the host larvae began its cocoon, or some 20 days after the host began to feed. In order to moult, the bee fly larva attaches to its host, leaves its cast skin holding on by the mandibles. In the second instar the bombylid larva loses its thoracic and anal setae and abdominal pseudopods. It at once becomes more sluggish, feeding rapidly and growing about 0.5 mm. a day; about 7 days after the first moult, the second occurs and the larva still feeding rapidly grows about 2.5 mm. a day, Mars- ton found. Host punctures apparently close up without exudate; the parasite frequently shifts its position and feeds from new spots. Spears, Sartor, and Hull (in litt.) made extensive studies of Anthrax limatulus Say in northern Missis- sippl. The work of Spears dealt with the attack of this species and also of Anthrax aterrimus Bigot on mud 24 BEE FLIES OF THE WORLD dauber nests, of which she collected 2,638 nests. Of these nests 1,070 were 7rypoxylon politum Say and the remainder were all Sceliphron caementariwum Drury. Her nests came from 14 sites in 6 counties. From the nests of 7'’rypoxylon politum Say she obtained 165 males and 180 females of Anthrax limatulus Say, but of Anthrax aterrimus Bigot, only 28 males and 22 females. From the 1,568 nests of Sceliphron caementarium Drury she obtained 19 males and 27 females of lima- tulus Say and no males and a single female of aterrimus Bigot. The percentage parasitism was found to be: Parasitism Parasitism by Host Parasite by nest individuals T. politum A. limatulus Say 32.24 5.27 A. aterrimus Bigot 4.67 0.76 S.caementarium A, limatulus Say 2.93 6.97 A. aterrimus Bigot 0.06 0.15 It is very interesting that although Rau (1916, 1946) recorded no less than six species of bee flies as parasites of Sceliphron caementarium Drury, Spears secured only two species in northern Mississippi. The abundance of mud dauber nests of species of Sceliphron Wlug, and of the organ pipe mud wasp Trypoxylon texense Saussure is truly remarkable at times. Beneath the bridge across the Tallahatchie River in Lafayette County, Mississippi, a bridge four-tenths of a mile long, I found the mud nests approximated one hundred thousand. Many cells contained parasitized grubs. Sartor’s work concerned entirely the attack by Anthrax limatulus Say wpon the nests of the cliff bee Anthophora abrupta Say. This was the only species of bee fly which he found attacking this bee. He found that under laboratory conditions no more than one larva ever developed from a host grub; when more than one was placed upon a host grub, all but one bee fly. larva would leave the brood cell within 3 days. He was unable to decide positively whether cannibalism takes place under conditions of crowding, but he noted that on three occasions dead larvae were found in the same brood cells with actively feeding larvae. He noted that feeding occurs irregularly in the first instar; at each feeding it appears to make a different puncture at new sites on the dorsal side of one of the thoracic segments. The first- instar larvae moulted to the second instar 12 to 18 days after hatching. Seven days later it moulted to the third stage. Sartor found that an Anthrax limatulus Say larva would consume fully the host tissues of a grub within 24 days. Pupation took place around the first of May. He comments on the wear and tear consequent to the pupal armature. Oviposition may begin as early as 8:00 a.m. on clear days and may continue for 4 hours, the active periods lasting 5 to 10 minutes followed by rest periods of 5 to 20 minutes. Oviposition was ob- served as early as June 10 and as late as September 15, but it is most intense in July and August. Twenty-one nesting sites in 4 counties in northern Mississippi were studied. His work extended over 1964 and 1965 and involved a study of about 6,000 brood cells and revealed a bee fly parasitism ranging from 1.6 percent to 35.4 percent. Clifford Osborn, a former student of the author, in some uncompleted studies of bee fly parasites reared Anthrax limatulus Say from cocoons of a cuckoo wasp, Chrysis sp., where it was acting in the role of a hyper- parasite. Anthrax limatulus artemesia Marston. Marston reared this subspecies from nests of several hosts as Sceliphron caementarium Drury, Chalybion califor- nicum Saussure, and Trypoxylon texense Saussure. Custer (1928) recorded this subspecies, A. limatulus artemesia, from nests of the leaf-cutting bee Dianthi- dium curvatum sayi Cockerell. Linsley and MacSwain (1942) pointed out that the parasitic anthophorid Melecta californica miranca Fox is subject to secondary parasitism by a species of Anthrax, which Marston identified as A. Limatulus artemesia. Not only does the subspecies Anthrax limatulus artemesia attack the Melecta, but it also attacks the bee Anthophora occi- dentalis Cresson, which is host of the Melecta. Anthrax limatulus columbiensis Marston. The only three known individuals of this subspecies were reared from nests of Anthophora Latreille. Anthrax limatulus larrea Marston. One individual of this subspecies was reared from a cell of a species of Anthophora Latreille. Anthrax limatulus vallicola Marston. This sub- species was first reported by Linsley and MacSwain (1942) as Anthrax sp. near fur Osten Sacken and at- tacking Anthophora linsleyi Timberlake. Anthrax monachus Sack. Zahkvatskin (1931) found these larvae in the egg pods of Ramburiella turcomana F. W. In these egg pods they were hyperparasitic on the pseudopupae of J/ylabris scabiosae scabiosae Olivier. Anthrax nidicola Cole. In a remarkable study of the biology of Diadasia consociata Timberlake, an emphor- ine bee, which nests in colonies upon the ground, Linsley, MacSwain, and Smith (1952) discovered two bee fly parasites, which belonged to undescribed species in different genera. They found that fungi at times destroy 50 percent or more of the cells in those ground- nesting, turret-making species, but they were able to secure a good assortment of parasites by covering square-yard areas with screens placed in position in early morning; at the end of the nesting season each burrow was excavated and contents recorded; field samples of parasites were collected and were reared in depressions upon a coat of paraffin, overlaying damp (sterilized) sand in petrie dishes. The bombyliids re- mained as larvae in the cells until May or June, appear- ing then as pupae, emerging as adult flies in late May until early July. These authors state that oviposition occurs by the female hovering over the nesting area, throwing the eggs into cracks, crevices, or burrows by a rapid movement of the abdomen. BOMBYLIIDAE 25 They observed that when a first-instar bombyliid was introduced into a bee burrow of the current season, it moved to an open cell by using the six long thoracic setae, which seemed to serve as legs. It remained in- active within this cell until the host larva had com- pleted feeding and then oriented itself along one of the intersegmental membranes; here it started feeding through small, integumental punctures and continued until the host was reduced to a mere shell. Linsley, MacSwain, and Smith further noted that feeding might be delayed until the next spring. When both bombylid and rhipiphorid larvae were introduced into the same cell the bee fly larva was the one that survived. In one case a rhipiphorid larva had almost entirely consumed the bee larva and was in consequence able to destroy the bee fly larva. These authors also concluded that bee fly parasitism was materially aided in areas where many bee turrets had been injured or destroyed. In seven samples containing 1,181 cells, the percentage of bee flies ranged from 1.2 percent to 29.3 percent. These data apply both to the Anthrax nidicola Cole and the Para- villa apicola Cole. Linsley and MacSwain (1957) also record Anthrax nidicola from nests of Diadasia enavata Cresson. Anthrax oophagus Paramonov. Zakhvatkin (1931) found this species to live both as a parasite and as a hyperparasite within the egg pods of several species of locusts in Turkestan, such as Dociostaurus maroccanus Thunberg, D. crucigerus tartarus Stechelk, Ramburiella turcomana F.-W., and Calliptamus turanicus 'Tarbin- sky. In this bee fly the larvae were also living hyper- parasitically, both on the bee fly Callostoma Macquart, and the meloid, A/ylabris atrata (Gebler). Percentage infestation of pods never rose above 6 to 9 percent. Zakhvatkin stated further that the cycle includes a par- tial second generation. Flies of the first generation emerge in early June and oviposit in the same places as Callostoma Macquart and the Moroccan locust. Eggs are placed in the powderlike, thin, upper layer of soil or thrown into fissures. After incubation of 5 to 9 days the young larvae emerge, and, while much like those of Callostoma Macquart, are only half as large; after undergoing two moults, they rapidly mature; some remain in the pod without change until the next spring, then pupate for 16 to 34 days; others pupate in July, the flies emerge in August and September and oviposit in the same places. Anthrax tigrinus De Geer. This large and con- spicuous Nearctic and Neotropical species of wide dis- tribution has been observed by a number of entomolo- gists, and the best studies so far were made by Nininger (1916) and Hurd (1959). Four species of Yylocopa Latreille in three subgenera of this bee have been found to harbor this bee fly. Hurd (1959) points out that many of the remaining 150 species of American car- penter bees will be found to be hosts. Interested stu- dents are especially referred to Hurd’s excellent sum- mary of this bee fly, and to his monograph on carpenter bees, for the distribution of these prospective hosts. Synonymy of Anthrax tigrinus De Geer is listed under the description of the genus Anthrax Scopoli, but in various studies on Xylocopa Latreille parasites the bombylid species has been referred to as Anthrax simson Fabricius, and as Argyramoeba simson Fab- ricius, or Argyramoeba tigrina De Geer. Angus (1868) was the first person to find the pupa of Anthrax associated with Xylocopa Latreille. In this and in another brief paper (1868) Angus refers to this fly as Anthrax sinuosa Wiedemann, and while we sus- pect that he actually had Anthrax tigrinus De Geer, we cannot be certain. At any rate he observed the fly ovi- positing and casting eggs toward the openings of Xylo- copa Latreille nests “in the same manner as a bot fiy depositing its eggs on the horse.” The eggs he found “quite numerously around the openings of the cells of the insects, and also to extend some distance from them.” In pricking some of these freshly deposited eggs, he noted that small maggots made their appearance. Angus sent his specimens of larvae to Parkard, who acknowledged them (1868) and later described them (1897). Balduf (1962) in his comments merely repeats the notes of Angus. Davidson (1893) comments on Anthrax tigrinus (as Argyramoeba simson) and says: “it was interesting to observe this pupa with its rings of hooked hairs on its body, preventing it from going backward as it gradually wriggled itself through the partitions to the external opening leaving its case hang- ing to the edge of the opening.” Out of 6 cells, David- son found 3 with bee fly pupae, 1 with a chalcid pupa, and 2 live bees. Nininger (1916) studied behavior of Anthrax Scopoli in XYylocopa cells during approximately a year in the San Dimas Mts., California. He points out that the bee fly first appeared upon stored pollen as a very minute, but extremely active, larva that “restlessly creeps about over food-mass, egg and larva, feeding promiscuously, then finally settles down and fastening itself by means of its hooked beak to the sixth or seventh segment of the Xylocopa larva.” The first-instar larva appeared before the bee egg had hatched and was 4 weeks old or more when it attached itself, at which time its length was 3-5 mm. Growth was very slow, increasing in 2 weeks to only 4 or 5 mm., after which growth increased explo- sively, doubling the size of the parasite in 24 hours. From this point on, the bee grub rapidly shriveled. Nininger comments upon the advantage of retarded growth rate, which enabled the bee larva to become large enough to serve for its complete development. The full-grown fly larva remains motionless for 10 to 12 days, becomes active, moults 2 days later, and pupates; it remains as a pupa for 15 to 20 days before emergence. He found parasitism amounted to about 10 percent, and was well distributed, with usually only one parasitized larva in the same brood. Hurd (1959) collected Xylocopa Latreille nests from Yucca stalks near Rodeo, New Mexico, in September. He speculated upon an apparent lack of synchrony with 26 BEE FLIES OF THE WORLD respect to bee fly emergence at this late season and avail- able pollen filled bee nests; had the pupae he found in the cells been left there, it is possible that they would have remained in this state until the following spring. I certainly agree that the adult bee fly can hardly sur- vive a winter of hibernation, although I have known emerged Anthrax Scopoli species to live in a bottle 6 or 7 days without food at reom temperatures. Much more likely, it seems that either eggs or planidia overwinter, instead of adults. Anthrax trifasciatus Meigen. Fabre (1908), in a truly delightful style mixed with curious teleological interpretations, recounts his observations upon the larvae of this species, which were plentiful in the nests of the European wall bee, Chalicodorma muraria Fab- ricius. It isa meticulous account by a masterly observer. Anthrax trifasciatus leucogaster Wiedemann. Engel (1937) makes Anthraaw leucogaster a subspecies of tri- fasciatus. Under the name Argyramoeba leucogaster, Frauenfeld (1864) reared the species from the wasp Cemonus sp. and described the pupa. Anthrax trimaculatus Macquart. Marston (1964) illustrates a pupa of this species obtained in southern Brazil. Numerous individuals were ovipositing there in the openings of Diadasia nests in a forest path. Anthrax zonabriphagus Portschinsky. Both Port- schinsky (1895) and Troitsky (1914) record this species as found within the egg pods of grasshoppers, but feed- ing upon other bombyliid larvae, and upon I/ylabris sp. larvae instead of the locust eggs. The genus Anthrax Scopoli. Present information” shows that different species of Anthrax have been reared from no fewer than 24 genera and 47 species of bees, wasps, locust egg pods, and beetles as parasites and sometimes as hyperparasites. According to Clausen (1928) a species has been reared from Schistocera. Another species of Anthrax attacks beetles of the genus Cicindela Linné. Many species attack mud nests of wasps or bees. These are easily found and brought into laboratories for observation. Krombein (1967) in his very remarkable work on trap-nesting wasps and bees records three species of Anthrax as parasites of no less than twenty species and subspecies of bees and wasps; this does not take into account the host species discovered by other authors. ANTHRACINAE: Walkeromyia Paramonov Species Hosts Authors lurida Walker Xylocopa submordaxr unpublished (NEOT) Cockerell Data taken from flies reared from pupae supplied by Dr. Krombein of the National Museum of Natural His- tory and Dr. Wirth of the U.S. Department of Agricul- ture. These pupae were quite different from those of other members of the Anthracinae. They were almost iden- tical with the pupae of Anthrax tigrinus Fabricius. From these studies of pupae I conclude that these very large anthracine flies should fall within a separate tribe, the Walkeromyini. EXOPROSOPINAE: Villa Lioy Species Hosts Authors Cut worms: alternata (NE) Cut worm larvae Riley and (also as scro- biculata Loew ) fulviana (NE) handfordi (NE) hottentota (also as flava Meigen) (PA) hypomelaena (NE) molitor (NE) paniscus (PA) Agrotis orthogonia Morrison Eucoa flaviocollis Smith Euxoa ochrogaster Guenee Eucoa tesselata Harris Feltia ducens Walker Hucxoa sp. pupae Agrotis vetusta Walker Lycophotia porphyrea (Denis & Schiff ) Euxoa segetum (Denis & Schiff) Eucoa forcipula (Denis & Schiff) Dichronia aprilina (Linné) pupa Barathra brassicae (Linné) Panolis piniperda pupa Feltia herilis (Grote) Similar to Taeniocampa rutula Grote Agrotis orthogonia Morrison Lepidopterous pupa Howard (1890) Brooks (1952) Brooks (1952) Brooks (1952) De Geer (1776) Walker (1851) Westmaas (1861) Ritsema (1868) Rogenhofer in Brauer (1883) Mulsant (1852) Wahlberg (1838) Vassiliew (1905) Riley and Howard (1890) Riley and Howard (1890) Brooks (1952) Yerbury (1900) quinquefasciata Pine processionary moth _ Biliotti, Demolin, (PA) pupa and Du Merle (1965 ) sexrfasciata (ET) Spodoptera exempta Hesse (1956) (Walker ) vitripennis (ET) Noctuidae pupa Hesse (1956) Beetles : alternata (NE) pygarga (PA) Meracantha contracta Beauvois Podonta nigrita Fabricius Hyslop (1915) Portschinsky (1915) Villa alternata Say. Hyslop (1915) recorded this species as bred from one or two of the numerous larvae of the tenebrionid, Meracantha contracta Beauvois, which he found buried a few inches about the base of an old tree stump. The single bombyliid pupa emerged from the pupa of the beetle and was figured by Hyslop. The Meracantha Kirby larvae were quite numerous and the bee fly suffered competition from tachinid parasites of the genus Veopales. Brooks (1952), in a fine research report, records this species from several caterpillars and pupae of the genus Agrotis. He gives a key to 12 species of bombyliid pupae and to 10 species of larvae and illustrates many details; this is a very important con- tribution to the morphology of this family. BOMBYLIIDAE 27 Villa fulviana Say. Brooks (1952) reports Huwoa sp. as host to V. fulviana, and host pupae were collected from June 30 to July 12. Villa handfordi Curran. This species is reported by Brooks (1952) to parasitize Agrotis vetusta Walker. Villa hottentota Linné. Under the name Anthrax flava Meigen, Wahlberg (1838) recorded rearing this species from caterpillars of the noctuid, Barathra bras- sicae (Linné). This species has been bred from a wide variety of hosts. Mulsant (1882) reported it from Dichronia aprilina (Linné) pupae; Vassiliew (1905) gives Panolis piniperda pupa as host. No less than five authors have reported Agrotis as host to hottentota. According to Biliotti, Demolin, and Du Merle (1965), d’Androic (1956) has recorded this species as a parasite of Thaumatopoea (as Cnethocampa) pityocampa Schiff. I have not seen d’Androic’s paper. Villa hypomelaena Macquart. Riley and Howard (1890) record having received four individuals bred in Indiana by F. M. Webster from cutworm pupae of Feltia herilis (Grote). Villa molitor Loew. Riley and Howard (1890) re- cord a pupa resembling Zaeniocampa rufula Grote sent to them. They give a few details of the pupal armature. The lepidopterous pupa was consumed, the bee fly issu- ing from a hole in one end, presumably in the larval state. Brooks (1952) lists Agrotis orthogonia Morr. as host of this bee fly. Villa paniscus Rossi. Yerbury (1900) reported this species bred from a lepidopterous pupa found in sand at St. Helen’s, Isle of Wight, which pupa was found July 7 and the fly emerged July 12. Villa pygarga Loew. This species attacks tenebrionid beetle larvae, according to Portschinsky (1915); the host was Podonta nigrita Fabricius. According to this author, parasitism reached as high as 50 percent in some localities. Villa quinquefasciata Wiedemann. Biliotti, Demolin, and du Merle (1965) made a very important study of the ethology of this species. They found it to be an im- portant parasite of the pine processionary moth Thaumetopoea pityocampa Schiff. Egg-laying behavior was divided into two phases: In the first phase the females alight on ground composed of very fine sand where it fills up the perivaginal pouch with such sand. In phase two, it flies away, and the egg now covered with adhesive substance is released into the perivaginal pouch, there it becomes covered completely with a fine earthen film. It is then ejected into situations that avoid direct exposure to the sun, such as at the base of stones or cracks in the earth. It continues to do so until all the sand in the perivaginal pouch is used up, where- upon it refills the pouch for a new egg-laying sequence. The authors note the importance of soil types in meet- ing satisfactory conditions for this species. Villa sexfasciata Wiedemann. Hesse (1956) reared this species from the noctuid Spodoptera exempta (Walker), both in Pretoria and Southern Rhodesia. The pupa is described, but not illustrated. Villa vitripennis Loew. Hesse (1965) notes that a large individual of this species was reared from a moth pupa of the family Noctuidae in 1941. Hesse figures parts of the pupal skin. EXOPROSOPINAE: Poecilanthrax Osten Sacken Species Hosts Authors Wasps: Hyperparasite of Elis haemorrhoidalis Fabricius lucifer (NE) Box (1925) Cut worms: Chorizagrotis thanatologia 5 Dyar Euzoa ochrogaster Guenee Eucoa flavicollis Smith Pseudaletia unipuncta alcyon (NE) Brooks (1952) Haw. Peridroma margaritosa Walkden (1950) Haw. fasciatus (NE) Chorizagrotis auxiliaris | Walkden (1950) Grote flawiceps Agrotis subterranea Hall (Painter fuliginosus Fabricius and Hall, 1960) (NE) lucifer (NE) Laphygma frugiperda Allen (1921) A.&S. sackeni monticola Orymodes devastator (NE) Brace Agrotis orthogonia Brooks (1952) Morrison Euzxoa flavicollis Smith Agrotis c-nigrum L. Agroperina dubitans Walker Augoa ochrogaster Guenee Eucoa flavicollis Smith Euczoa tesselata Harris Feltia ducens Walker Crymodes devastator Brace Chorizagrotis thanatologia Brooks (1952) Dyar Euaxoa scandens Riley Chorizagrotis auviliaris Grote tegminipennis (NE) willistoni (NE) Brooks (1952) Walkden (1950) Painter and Hall (1960) All except one of the known hosts of Poecilanthrax Osten Sacken are noctuids. Box (1925) reported the rearing of P. lucifer Fabricius from Elis haemorrho- idalis Fabricius in Puerto Rico. Seven species of this Nearctic genus have been recorded as endoparasites of noctuids of the sod worm or cutworm type. Very little information is available on the earlier phases of these life histories, and we are left to conjecture with respect to certain details. Brooks (1952) thought the flies over- wintered as eggs in soil or vegetation. Painter and Hall (1960), while admitting that this may be true in the north, surmise that in southern areas the eggs may hatch soon after oviposition and the planidia search around for any small caterpillars lurking among grass or vegetation roots. Another factor, the number of generations, plays a part here. 28 BEE FLIES OF THE WORLD In Mississippi the present writer found P. lucifer Fabricius apparently bivoltine. On May 1, 1963, extra- ordinary numbers of this species were present on a smal] plot beside the University of Mississippi campus high- way, over a hundred were collected in a little more than 2 hours; about 60 were collected near 10:00 a.m. from abundant flowers of Coreopsis, until scarcely any were left. On return an hour later, almost as many more were collected. With P. lucifer were a few Parabombylius sp. In Mississippi P. lucifer almost disappears in late May and June, reappearing in late July or August. They again become quite abundant in September and October. H. W. Allen (1921) reared numerous individ- uals of P. lucifier from the southern grass worm Laphygma frugiperda A. & S. Parasitized larvae pupated, but shortly thereafter the bee fly larvae ma- tured and pupated within the pupa of the grass worm, from which it emerged by twisting and wriggling until free. Harned (1921) also comments on the abundance of this parasite. Painter and Hall (1960) suggest that the planidial life in this genus is short, perhaps no more than a week, and that the remaining larval instars are of brief dura- tion. The flies emerge soon after the host pupates. Emergence from the host pupal skin is in the usual way, the parasite cutting its way out by means of its cephalic spines, resting a short time, then wriggling its way to the surface; the pupal case projects partly from the soil. The extent of parasitism by Poecitlanthrax Osten Sacken in cut worm populations varies widely. Walk- den (1950) found it relatively low in Kansas. The sev- eral species known to Brooks (1952) gave parasitism of 2 to 5 percent. P. flaviceps fuliginosus had a low inci- dence, only 1.3 percent. This species was reared by Hall (Painter and Hall, 1960) from Agrotis subterraneum Fabricius in Calexico. Clausen (1940) gave 18 to 25 percent for the range of parasitism in P. lucifer Fab- ricius. Almost certainly the parasitism of this species may rise to 50 percent in some areas at certain times. Most species of this genus was univoltine, but Painter and Hall (1960) note that several species appear to be bivoltine, at least at times, these are: Poecilanthrax flaviceps flaviceps Loew, P. flaviceps fuliginosus Loew, P. effrenus Coquillett, P. lucifer Fabricius, and more rarely P. poecilogaster poecilogaster Osten Sacken. EXOPROSOPINAE: Paravilla Painter Species Hosts Authors Hymenoptera : apicola (NE) Diadasia consociata Timberlake, D. bituberculata Linsley & Mac- Cresson Swain (1952) gorgon (NEOT) Elis sp. Wolcott (1928- 1924) perplera (NE) Diadasia diminuta Linsley & Mac- Cresson Swain (1957) tricellula (NE) Diadasia bituberculata Linsley & Mac- Cresson Swain (1952) sp. (near P. Diadasia vallicola Linsley & Mac- flavipilosa) (NE) Timberlake Swain (1957) Paravilla apicola Cole. This was one of two species of bee flies found parasitizng Diadasia consociata Timberlake, in San Joaquin County, California. The discussion under Anthrax nidicola Cole applies to this species as well. P. apicola also parasitizes Diadasia bituberculata Cresson. Paravilla gorgon Fabricius was reared by Wolcott (1923-1924) from cocoons of HLlis haemorrhoidalis Fabricius in Puerto Rico, behaving in this case as a secondary parasite. Paravilla perplexa Coquillett. According to Linsley and MacSwain (1957), this species was seen commonly ovipositing in nesting sites of Diadasia diminuta Cres- son near Carlsbad, New Mexico. They state, “The fe- male hovers over a turret and drops lower to oviposit, repeating this activity several times at each burrow. After ovipositing in several burrows, the bee fly rests on the ground a few feet from the nesting site for sev- eral minutes. Later she resumes oviposition and fre- quently includes some of the same burrows in the second visit.” Still another species of Paravilla, near P. flavi- pilosa Cole, was found ovipositing within the nests of D. vallicola Timberlake. While they suggest that cer- tain species of bee flies may be regularly associated with Diadasia, they note that some species at least are not host specific, and host size is not a limiting factor. P. apicola was reared from both the small cells of Diadasta consociata and the large cells of D. bituberculata. Paravilla tricellula Cole. Linsley and MacSwain (1952) obtained this species from ground-nesting bees of Diadasia bituberculata Cresson in San Diego County, California. This fly, one of five parasites of this bee, was obtained from 102 cells, or a total percentage of 27.1. Linsley (1958) evaluates this and other species of bombyliids that attack ground-nesting bees. Writing of bees, he states, “potentially, the most effective insect parasites of gregarious, ground-nestinz species are probably bombyliid flies, which in a Momia Latreille colony are capable of infesting almost 100 per cent of the burrows, which are open at a given time.” EXOPROSOPINAE: Rhynchanthrax Painter Species Hosts Authors parvicornis (NE) Tiphia sp. (on Phyllophaga spp.) Davis (1919) Rhynchanthrax parvicornis Loew. Davis (1919) ob- tained this species from 7iphia Fabricius cocoons; the exit hole in the cocoon, always made at the end, is circular and sharp, as if made with a sharp knife. Tiphia emergence holes are ragged and made a little to one side of the tip. EXOPROSOPINAE: Chrysanthrax Osten Sacken Authors Swezey (1915) Species Hosts cypris (NE) Myzinum ephippium (Fabricius) Davis (1919) edititia (NE) Anthophora montana Davidson (1900) Cresson BOMBYLIIDAE 29 Chrysanthrax cypris Meigen. Swezey (1915) reared this species from cocoons of Myzinum ephippium (Fab- ricius), at Urbana, Illinois. Chrysanthrax edititia Say. Davidson (1900) relates the curiously long-drawn-out larval history of 10 indi- viduals of bee fly larvae he took, which were parasitiz- ing Anthophora montana Cresson. He got them in July 1895; one bee fly hatched out in the same season ; 2 years later another pupated, but died; in July 1899, after a short pupation period, four others succeeded in emerg- ing as adult flies. EXOPROSOPINAE: Oestranthrax Bezzi Species Hosts Authors Cossid moth: goliath (OR) Pupa of a cossid moth Oldroyd (1951) (indet. ) Oestranthrax goliath Oldroyd. In 1951 Oldroyd re- corded, described, and illustrated this very large bee fly from Malaya, which was reared from a cossid pupa. Many interesting questions might be raised as to how the fly reaches the larvae. Oldroyd says that the general appearance of the adult strongly suggests a nocturnal or crepuscular habit. EXOPROSOPINAE: Dipalta Osten Sacken Species Hosts Authors serpentina (NE) Myrmeleon immaculatus De Geer Smith, Roger C. (1934) Dipalta serpentina Osten Sacken. Roger C. Smith (1934) relates that he obtained individuals of this species, which he reared from cocoons of the above ant lion collected in Kansas. This record extends the hosts of bee flies into the order Neuroptera. EXOPROSOPINAE: Thyridanthrax Osten Sacken Species Hosts Authors Egg pods: fenestratus (PA) Pararcyptera microptera Portschinsky (Fischer-Waldheim ) (1895 ) Bezrukoy (1922) Dociostaurus maroccanus Séguy (1930) Thunberg Ocneridia volxemii Kunckel (1893, (1. Bolivar) 1894) pallidipennis (PA) Dociostaurus maroccanus Zakhvatkin Thunberg (1931) perspicillaris (PA) Paracyptera microptera Troitsky (1914) (F.-W.) Tsetse flies : abruptus ; argenti- Glossina morsitan West- frons ; alliop- wood, brevipalpis terus ; beneficus; Newstead, austeni brevifacies ; Newstead, pallidipes burtii ; lloydi; Austen, tachinoides lugens ; salu- Westwood taris ; transiens (All ET) Austen (1914, 1922, 1929) Lloyd (1916) Hegh (1929) Lamborn (1915) Fiedler and Kluge (1954) Species Hosts Authors Cutworms: leucoproctus (BT) Lowxostege frustalis (Zeller) (Pyralidae) Hesse (1956) lugens (ET) atratus (NE) Wasps: Bembiz occidentalis beutenmullerit Fox Bohart and Mac- Swain (1939) lugens (ET) Muscid fly : Muscid fly found in a nest of Odontotermes badius (Haviland ) Hesse (1956) velutinus (PA) Thaumatopoea (as Cnetho- Biliotti, Demolin, campa) pityocampa and du Merle Schiff (1965) Some of the Palaearctic species of Thyridanthrax Osten Sacken attack the egg cases of locusts. A Nearctic species attacks bembecids, while South African species utilize tsetse flies (Glossina) as hosts. One South African species has been reared from a muscid fly. Still others act as both primary parasites, and as secondary parasites of braconids, in noctuid caterpillars of the army worm and cut worm type. Thyridanthrax fenestratus Fallén. Bezrukov (1922) found this species living in the egg pods of the locust Pararcyptera microptera (F.-W.) in Siberia. Kunckel (1894) records it from eggs of Ocheridia (Ocneridia) at Biska, and Stauronotus at R’hiras. Séguy (1930) lists the larva of this species as found in the egg pods of Dociostaurus maroccanus Thunberg, Ocneridia vol- xemii (I. Bolivar), and Pararcyptera microptera (F.-W.) Thyridanthrax pallidipennis Paramonov also de- stroys locust egg pods. Thyridanthrax perspicillaris Loew. 'Troitsky (1914) records this species as feeding in the egg pods of Par- arcyptera microptera (F.-W.) in Siberia. Thyridanthraw Uoydi Austen. This species was de- scribed by Austen (1914); he noted it to be the first species of Diptera recorded as a parasite of tsetse flies, the genus Glossina Wiedemann. The original discovery was made by Lloyd, who comments upon the relation- ship in 1916. He states that the pupa of the bee fly bursts forth from the Glossina pupa anteriorly and works its way to the surface of the ground by means of the long fringe of hairs which are alternately raised and depressed, with the peglike legs acting as a fulcrum. On the surface of the ground it becomes quiescent, elon- gates, splits along the dorsal surface, and the time from emergence to flight occupies only 2 or 3 minutes. Austen (1914, 1929) and Hesse (1956) have shown that 10 species of Thyridanthraw Osten Sacken are parasites of 5 species of Glossina Wiedemann. The additional species of Thyridanthrax known to use Glossina Wiedemann as hosts are: 7. abruptus Loew, 30 BEE FLIES OF THE WORLD argentifrons Austin, alliopterus Hesse, beneficus Austen, brevifacies Hesse, burtii Hesse, lugens Loew, salutaris Austen, transiens Bezzi. Records available to the author show the following : (1) Of these species, 5 have been bred from only a single species of Glossina. From Glossina morsitans were bred the 4 species of 7’. beneficus, lloydi, salutaris, and transiens, and from Glossina austeni was bred the species 7. burtii. (2) Two species of bombyli- ids were bred from 2 species of Glossina. T. argenti- frons was bred from Glossina morsitans and G. tachino- ides. The species 7’. ugens was reared in small numbers from both Glossina morsitans and G. austent. (3) How- ever, Hesse (1956) points out that the remaining 8 species of Thyridanthrax were each reared from 3 species of Glossina. Twenty individuals of 7’. alliop- terus were secured from Glossina austent, brevipalpis, and pallidipes. And 98 specimens of 7’. brevifacies were obtained from these same 3 species of Glossina. No less than 371 individuals of 7. abruptus were reared from Glossina morsitans, brevipalpis, and pallidipes. No less than 359 of the 7. abruptus species were bred from G@. morsitans alone. Thus, it will be seen that Glossina morsitans is parasitized by no less than 7 species of Thyridanthrae and is the most heavily parasitized species. Hesse (1956) notes that the extension of sev- eral of these species of Thyridanthrax, namely 7. abruptus, brevifacies, lugens, and transiens, into areas where tsetse flies do not occur indicates still other hosts for these species of Thyridanthrax. Thyridanthrax leucoproctus Loew. Hesse (1956) calls attention to the fact that this species was bred by S. J. S. Marias from caterpillars of the Karoo army worm Lowostege frustalis (Zeller). He notes further that Marias was able to show that this bee fly was actually a secondary parasite of the army worm, since the bred individuals emerged from pupae lodging within 2 braconid cocoons of the genus Macrocentrus. Thyridanthrax leucoproctus then becomes a harmful species of bombyliid. Thyridanthraw lugens Loew was also bred from the Karoo army worm, the noctuid Lowxostege frustalis (Zeller). Thyridanthrax atratus Coquillett. This large, black, American species was found by Bohart and MacSwain (1939) to be parasitizing Bembix occidentalis beuten- mulleri Fox associated with Hxoprosopa eremita Osten Sacken as a competitor. The area studied was Contra Costa County, California. See comments under F'xo- prosopa applying to both species. Thyridanthrax lugens Loew. Hesse (1956) reports this species as parasitizing a muscoid fly found in a nest of Odontotermes badius (Haviland). Thyridanthrax velutinus Meigen. Biliotti, Demolin, and du Merle (1965) point out that d’Androic (1956) records 7’. velutinus as a parasite of Thawmatopoea (as Cnethocampa) pityocampa Schiff. I have not seen d’Androic’s paper. EXOPROSOPINAE: Hemipenthes Loew Species Hosts Authors catulina (NE) Bessa harveyi Townsend, parasite of the sawfly Pristiphora sp. Brooks (1952) maura (PA) Cocoons of Ophion sp.and Portschinsky Banchus sp. as hyper- (1895 ) parasite of Panolis piniperda Panzer morio (PA) Ophion sp., Banchus com- Lassmann (1912) pressus F. Nemoraea sp. on Panolis Sack (1899) piniperda Panzer Masicera silwatica (Fallén) Vassiliew (1905) Parasetigena sp., and Baer (1920) Ernestia rudis (Fallén) Dendrolinus sp. Gonia spp. and Bonnetia comta Fallén out of noctuid pupae Neodiprion sertifer Geoff. Séguy (1930) Brooks (1952) sinuosa (NE) Finlayson and Finlayson (1958) Neodiprion sertifer Geoff. Griffiths (1959) Masicera silvatica (Fallén) Vassiliew (1905) as hyperparasite of Dendrolinus pini L. Winthemia sp. Paraphyto opaca Coquillett velutina (PA) sp. (NE) sp. (NE) Clausen (1928) Brooks (1952) Hemipenthes catulina Coquillett. This species acts as a hyperparasite. Brooks (1952) obtained flies from the tachinid Bessa harveyi Townsend, which is a parasite of the sawfly Pristiphora sp. in British Columbia. Hemipenthes maura Linné. In 1895, Portschinsky obtained this species from the cocoons of Ophion sp. and Banchus sp., where it assumed the role of a hyper- parasite of Panolis piniperda Panzer. Hemipenthes morio Linné. This species, recorded as Anthrax morio Linné, was reported by Lassmann (1912) as a hyperparasite, that is a secondary parasite, from a species of the genus Ophion Fabricius and also from Banchus compressus F. In 1899, Sack recorded this bee fly from the tachinid Nemoraea, which had been parasitizing Panolis piniperda Panzer. Baer (1920) reared this species from the tachinids Parasetina sp., and Ernestia rudis (Fallén) where Hemipenthes morio was acting as a secondary parasite to them. Vassiliew (1905) had previously reared this bee fly from Masicera silvatica (Fallén), a tachinid parasite of other Lepidoptera. It seems likely that its chief role, perhaps its only role, is that of a hyper- parasite, in which case H. morio must be classed as a harmful species in man’s economy. Baer raises some interesting questions. Does the Hemipenthes larva enter the tachinid larva before or after it leaves the body of the caterpillar? Does it penetrate the pupar- ium, or does it reach the tachinid host while the latter is crawling about the ground seeking a place to pupate? Baer felt the latter possibility was more probable. If true, bee fly eggs must be laid before pupation of the _ tachinid, and after hatching, the bee fly larvae must BOMBYLIIDAE 31 await the maturing of its host, which might take place after attachment. There seems to be no evidence, as far as Hemipenthes Loew is concerned, whether the bee fly is endoparasitic, or feeds externally upon the host within its puparium, which it could certainly do. It is believed that those bee flies which parasitize noctuid pupae do lead an endoparasitic existence. Brooks (1952) recorded the rearing of this species (as H. moriodes Say) from the tachinids Gonza spp. and Bonnetia comta Fallén out of noctuid pupae in Canada. Hemipenthes sinuosa Wiedemann. Finlayson and Finlayson (1958) reared this species from large num- bers of cocoons of the European pine sawfly Veodiprion sertifer Geoftroy in Ontario during 5 of the years be- tween 1941-1949. The percentage of parasitism by Hemipenthes Loew was given as 0.3 in 1941 on 626 cocoons; 0.1 in 1943 on 2,308 cocoons; 4.5 in 1946 on 1,892 cocoons; 0.4 in 1947 on 3,225 cocoons; and 12.0 in 1949 on 275 cocoons. Their study of the parasites of this sawfly recorded 23 species; 13 species in 9 genera were Ichneumonidae; 6 species in 5 genera were chalcids; and 4 species in 4 genera were Diptera. The other flies included tachinids Neophorocera hamata A. & W. and Spathimeigenia erecta Ald. They give a description of the last instar larva of the bee fly, a figure of the buccopharyngeal apparatus, and an arched, dorsal view of a pupal exu- vium, which did not show the details of the coronal armament. The exit hole on the sawfly cocoon is directly on the end of the cocoon with the cap attached. Brooks (1952) also reared this bee fly from Veodiprion sp. in British Columbia. Hemipenthes velutina Meigen. Vassiliew (1905) reared this species from pupae of M/asicera silvatica (Fallén), where it was acting as a hyperparasite of Dendrolinus pint L. EXOPROSOPINAE: Exoprosopa Macquart Species Hosts Authors apicalis (ET) Tachypompilus ignitus Hesse (1956) (Smith ) Parasitic Hymenoptera Bembiz occidentalis beutenmulleri Fox Phyllophaga spp. (pupae) caliptera (NE) eremita (NE) Brooks (1952) Bohart and Mac- Swain (1919) Richter and Fluke (1935) Davis (1919), fasciata (NE) fascipennis (NE) Tiphia Fabricius on Phyllophaga Kirby Richter and Fluke (1935) pueblensis (NE) Tiphia Fabricius on Davis (1919) Phyllophaga Kirby EHxoprosopa apicalis Wiedemann. Hesse (1956) noted that an example of this species had been reared from the pupal cocoons of the large pompilid wasp Tachy- pompilus tgnitus (Smith), which in turn preys upon the large theraphosid spider Harpactira. Exoprosopa caliptera Say. Brooks (1952) reared an individual from an unidentified parasitic hymenopteron in Alberta. He describes the pupa and illustrates the cephalic spines. Exoprosopa eremita Osten Sacken. This species was given considerable study by Bohart and MacSwain (1989) in a sand-dune area in Contra Costa County, California. It was accompanied by Thyridanthrax atrata Coquillett, which behaved quite similarly. Both species were parasitizing Bembix occidentalis beuten- mullert Fox. Flies became active only after dunes were warmed by the late morning sun, and they became in- active as early as 3:00 p.m. In ovipositing, the flies hovered a few inches from a Bembiz nest hole, flipping the tip of the abdomen forward; as this is done several. times at each burrow, it seems clear that several eggs are cast toward each tunnel. After such behavior the female rests on sand a few inches from the tunnel before searching for another burrow. Perhaps this rest is needed to allow additional eggs to move downward to the exit of the ovipositors. Bohart and MacSwain found that only prepupae were attacked; the first moult left the larvae maggot- like, without its bristles, and with head region retracted. The bee fly overwinters as a mature larva; pupation takes place in the summer within the Bembix cocoon. The combined percentage of parasitism of Bembix Fabricius by the two species amounted to about 1 per- cent. These authors surmise that the habit of this species of Bembix of keeping its entrance covered while hunting and after complete provisioning accounts for the low degree of parasitism, inasmuch as the bombyli- ids will only oviposit at an opened burrow while the wasp is inside. Bohart and MacSwain figure the adults and larvae in situ in cocoons. Exoprosopa fasciata Macquart. Richter and Fluke (1935), in a brief but remarkable study, record finding 27 bombyliid larvae of this species attacking exclusively the pupae of Phyllophaga sp. in 4 localities of Wis- consin pasture land in 1933 and 1934. Most of them were found at an average depth of 15.5 inches, while 237 unparasitized pupae and prepupae were arranged at an average depth of 16.2 inches. Parasitism was slightly below 10 percent at the principal locality. These authors further state that the small larvae of the bee fly feed on the ventral surface of the white grub pupae from the middle of July and increase rapidly in size, leaving nothing but shriveled grub pupal skins; these mature larvae then remain in old pupal grub cells during the following winter. They note further that 3 full-grown bee fly larvae collected in April were still larvae in December of the same year. One larva taken in 1933 became full grown that season, was kept over winter at room temperature, pupated on September 30 of the following year, and emerged in November. Exoprosopa fascipennis Say. This species was first reared from Tiphia Fabricius cocoons by Forbes (1907). Davis (1919), in his excellent report on the natural enemies of Phyllophaga, notes that F. fasc?- 32 BEE FLIES OF THE WORLD pennis had been reared from such cocoons in several parts of Indiana and Illinois; he did not see any adults ovipositing and surmised that the flies might lay their eggs on flowers or directly on Tiphia Fabricius. Exoprosopa pueblensis Jaennicke. Davis (1919) fig- ures a pupa and adult of this species, which had been reared from a 7%iphia cocoon collected in Kansas. EXOPROSOPINAE: Litorrhynchus Macquart Species Hosts Authors dilatatum (BT) Bezzi (1921) Hesse (1956) Sceliphron spirifer (Linné) Litorrhynchus dilatatum Bezzi. In 1921, Bezzi relates that a specimen of Litorrhynchus tollint Loew was bred by Dr. Peringuey from the mud nest of the ubiquitous South African sphecid Sceliphron quartinae (Gribodo) together with a large ichneumonid, Osprynchotus capensis Spinola, and a mutillid, Dolichomutilla syco- vax (Smith). However, Hesse (1956) states that the bee fly is properly the species Litorrhynchus dilatatum Bezzi, and the wasp properly Sceliphron spirifex (Linné). Actual host of the fly within the nest was not determined. EXOPROSOPINAE: Ligyra Newman Species Hosts Authors Ruiz Pereira (1929, 1930) Copello (1938) Sorensen (1884) Clausen (1928) morio (as ery- throcephalus ) (NEOT) Monedula sp. Pompilus Fabricius oenomaus (OR) Campsomeris Lepeletier Scolia Fabricius Tiphia Fabricius Ligyra morio Fabricius. In 1929 and 1930, Flaminio Ruiz Pereira published some notes on this species under the name Lxoprosopa erythrocephala Fabricius. These bee flies were found in the nest of Bembix spp. in the Chilean province of Atacama and were utilizing the wasps as hosts. Copello (1933) recorded the same species as a parasite of I/onedula surinamensis De Geer from the barancas of San Isidro. He described egg and larval stages and gave rather crudely executed fig- ures of several stages; these suffice to show that the planidium is of the usual type; the spines as shown upon the pupa are unusually stout and heavy. Ligyra oenomaus Rondani. Under the old genus name Hyperalonia Rondani, Clausen (1928) presented a remarkable ethological and statistical study of the relationship of this species of bee fly to its hosts, 77phia sp. This 7iphia sp. was attacking the grubs of Anomala dimidiata Hope. Field collections of cocoons made by a large species of Tiphia at Shillong, India, in the summer of 1925 showed parasitism of more than 60 percent by this bee fly, consequently causing a sharp reduction in the effec- tive parasitism of Anomala dimidiata by scoliids. Clausen’s data showed that Z7iphia adults began to emerge May 25 and ended July 20. 7iphia larvae that were feeding were found beginning June 16 and ended August 17. Tiphia larvae in the cocoons were first found June 26 and continued beyond September; whereas, Ligyra adults first emerged July 8 and ended August 29. Parasitism of the 7iphia cocoons by Ligyra Newman began August 1, mounting slowly at first and then sharply after September 2, to a peak of 55 percent on September 26. It is noteworthy that the adult popu- lation of Ligyra began as much as 5 to 6 weeks after that of its host and the greater part of the 77phia larvae had spun cocoons before any adult Ligyra appeared. In investigations extended over two summers, 1925 and 1926, more than 1,400 cocoons of 7iphia were collected. In the first summer the parasitism ranged from 57.6 to 65.3 percent. In 1926, beginning with zero parasitism until July 15, the percentage gradually rose from 1.9 on August 2 to 56.3 on September 30. No eggs or egg laying by Ligyra Newman was ob- served, and only two first-instar larvae were found, these in September. A dissected, gravid female showed 537 mature, slightly brownish, ellipsoidal eggs, each 0.5 mm. long. As the life of the female fly is long, the author pointed out that the ultimate total number of eggs per female may be much greater. Clausen (1928) concluded that the period of the first instar must be a very short one, but that the time spent by them in search of cocoons was considerable. The 7iphia cocoons lay in sandy, light soil at a depth of 1 to 3 inches. Clausen surmised that the planidial larvae must be obliged to seek out the 7iphia cocoons. He felt that it was more likely that the planidium would be able to penetrate the cocoon than that the adult Zigyra would be able to de- tect the presence of the 77phia cocoons beneath the soil, but this must remain an unproved point. Clausen also concluded that the planidium must pene- trate the cocoons of 7%phia Fabricius, since such a large part of the 7%phia population is so encased by the time the bee flies are about. The planidium is at first 0.9 mm. long and reaches 2.0 mm. before moulting. Three pairs of long, thoracic setae and a pair of caudal setae are lost at the first moult. Clausen states that feeding begins almost immediately when they enter the cocoon and is largely in the thoracic region. The second-instar larva measures from 2.7 to 3.5 mm., is capable of free move- ment, and the host larva begins to show signs of numer- ous feeding punctures in its integument on both thorax and abdomen. In the third and final instar the thoracic segments are yellowish, the abdomen is milky white, due to the numerous fat bodies beneath. Winter and spring are passed in this state and pupation is late in June. Emergence is by means of a circular cut at the head end of the cocoon, in contrast to an irregular hole made at one side of the anterior end by Tiphia Fabricius. This author found that the Ligyra Newman normally attacks the larva of the host and rather uncommonly the pupa. A few individuals of this bee fly were reared from cocoons of Scolia sp. and Campsomeris sp. BOMBYLIIDAE 33 Clausen (1928) reports that Ligyra sp. has been reared from the psammocharid Psewdagenia Kohl, but I have not seen the reference. Ethology and Interrelationships of the Adult Bee flies are characteristically sun-loving insects. They are expert in flight, and in most species the wings are large in proportion to the body, rendering them highly adapted to hovering and poising, whether in search of hosts or in the process of taking nectar from flowers. This they do as beautifully as nemestrinids or syrphid flies. Since the rather long and slender legs are weak, particularly in the tarsi, these flies generally confine themselves to brief rest stops upon the ground, stones, or tops of flowers, and sometimes on leaves. From such perches they often spring upward in rapid, power- ful flight, usually coming back down within a short distance. The larger species of Hxoprosopa Macquart may not alight within fifty or a hundred feet, after being disturbed. Some species of Geron Meigen and of Thevenemyia Bigot have a way of bobbing up and down above flower heads before alighting. Bee flies are more active during the late morning and early after- noon hours. During a bee fly hunting trip to Mexico and the southwestern United States in 1962, it was noted that these flies seldom appeared before 8 o’clock in the morning; by 10 o’clock peak populations could be ex- pected in any locality with activity falling off toward the middle of the afternoon. No nocturnal activity has been reported for any species, but the comment of Oldroyd (1951) on a possible crepuscular species is interesting. Males of Bombylius Linné may spend several days hovering during much of the daylight hours beneath trees in forest glades; poising at one spot with very little change in position, but so sensitive are they to motion or disturbance that I was once obliged to make three tries with a net before I secured a male of Bomby- lius major Linné. Hardy (1920) comments on the extreme rapidity with which Systoechus crassus Walker darts away from its hover position 8 or 10 feet above the ground. Holmes (1913) made some interesting observations on numerous individuals of a species of Bombylius Linné that he found in the hills east of Berkeley, Cali- fornia. All of the hovering individuals had their heads turned away from the sun, and all were hovering in the sunlight; when a shadow was thrown upon a fly it immediately darted elsewhere; moreover, he noted the habit of darting toward objects that approached, for example, several times after honey bees and twice after yellow jackets, a reaction doubtless associated with mating; he further noted that when a fly met a member of its own species of the same sex, they would spin around in a whirl. We can therefore, distinguish three separate occasions when bee flies hover or poise: in a mating or premating exhibition; hovering in search of hosts or the nesting sites of hosts together with egg deposition or egg thrusting; and hovering before flowers in seeking nectar. In this connection I should like to take note of the marked attention that several of the larger species of Anthrax Scopoli pay to people. On many occasions they have come within inches of my bare arm or neck and often refuse to be intimidated, and I have had them light upon my arm for brief periods. They will also hover about automobiles regardless of their color. I have noted some evidence that the large Anthrax tigrina Fabricius and similar species seem to have a route over which they hover and flutter in search of ovi- position sites, although to positively demonstrate this would require marked individuals. Some bee flies evince a territorial distaste for other species. In September recently, about the few wide clumps of desert blossoms along the highway southwest of Riverside, California, I watched a series of Villa crocina Coquillett chase away individuals of a species of Lepidanthrax Osten Sacken of about half the size of the Villa. Washburn (1906) personally claims to have been “quite severely bitten” by bombyliids. He quotes Lug- ger (1898, p. 48) as follows: “This proboscis can be used for other purposes besides sipping nectar, as the writer found out to his sorrow, when he attempted to catch some of them in his hand and succeeded; violent pain, a swollen finger and added knowledge were other results of the catch not bargained for.” Both authors appar- ently refer to Systoechus oreas Osten Sacken as the vicious one. While bee flies do exhibit a brief, artificially stimu- lated catalepsy or death-feigning, it is finished within seconds. The longevity of bee flies varies. Species that seldom or never feed, such as those in the genus Anthrax Scopoli, have a more limited life. However, I have kept captured individuals of both sexes of Anthrax limatulus Say alive up to 6 days in a jar with food and water. I have induced them to accept a mixture of water and honey from the end of a pipette. While I have never seen the flies of this genus feeding from flowers, they have occasionally been seen to feed according to the records of other observers. Robertson (1928) recorded two small species of Anthrax Scopoli on 11 flowers, and feeding on 6 species of flowers. With the species of genera that have long mouthparts adapted for probing flowers, the longevity may be greatly extended and may in large measure depend upon the length of the bloom- ing season of their favorite food plants. Some bee flies are bivoltine, especially in the southern ranges, but. cer- tainly the great majority of all members of the family are univoltine. Painter and Hall (1960) list 4 out of the 33 known species of Poecilanthraw Osten Sacken that are known to be bivoltine in some areas; among these 4 is Poecit- anthrax lucifer Fabricius, very common in Mississippi. One brood emerges in late April, but the principal brood emerges in midsummer and lingers until frost. 34 BEE FLIES OF THE WORLD Many bee flies, especially those of arid regions with little rainfall, are able to resist adverse conditions and remain in diapanse for long periods. Davidson (1900) records how several bee fly larvae of Thyridanthrax edititia Say, taken from cells of the turret-making ground bee Anthophora montana Cresson, survived in the larv al state for four years before pupating and emerg- ing as adults; one of these appeared to have been in the ground at least a year earlier. Other writers have occa- sionally noted a larval residence extending into three years. The author has maintained numerous larvae of Anthrax limatulus Say in enforced diapause up to three years by keeping them on damp sand in petri dishes in refrigerators. These were all from mud dauber wasp nests; at the time of writing I still have many unopened refrigerated cells of the cliff-dw elling bee Anthophora abrupta Cresson that contain bee fly larvae. Bee flies appear to produce two types of sounds, although both kinds may possibly have the same basic origin. Some of the larger species of L'voprosopa when frightened suddenly from a flower head will rise up in the air with a loud whir of wings. Some species pro- duce a distinct hamming sound at “different pitches. The shrillest bee fly with which I am acquainted is Hetero- stylum robustum Osten Sacken; in Mississippi, at least, this species has a loud, shrill hum pitched almost as high as nemestrinids and audible at some distance. Those I collected in western Kansas were scarcely noticeable. It is possible there is a sex difference in the hum. On the other hand the lowest notes I have en- countered come from a species of Gevon Meigen which are only audible when the fly is placed within a tube. The hum of Anastoechus barbatus Osten Sacken is pitched quite low; several years ago Marguerite Hull, my late wife, and I found a population of over a hun- dred ina very small patch of blooming Helianthus in the New Mexico mountains; their combined sound was very audible indeed. Mimiery is highly developed within the subfamily Systropinae, where it is indeed remarkable; the flies in this worldwide subfamily resemble wasps of the slender, thread-waisted type. Many American species mimic the species of Sphex (Ammophila). Brues (1939) points out that many of the Oriental species of Systropus mimic vespoid types of wasps which have conspicuous yellow spotting on the thorax and abdomen. He notes that this is not universally true and finds that the ammophiloid pattern is more widespread and therefore apparently older. Bezzi (1924) calls attention to the beautiful bee flies of the genus Antonia Loew; these he considers to be mimics of vespids and crabronids and to resemble strongly such syrphid flies as Sphaerophoria Lepeletier and Serville and Yanthogramma Schiner; he notes that Kneucker captured individuals of Antonia suavissima Loew upon flowers of Zygophyllum coccineum 1.., which strongly matched the coloration of the fly, a cryptic coloration situation. IT am able to add another interesting example of mimicry. The very large robust, stout bodied species Bryodemina valida Wiedemann closely resembles a large species of Bombus found associated in the same area and same time sequence in the area of Guadalajara, Mexico; this species is abundant at least as far south as Cuernavaca. Cryptic coloration also occurs rather frequently among those species which rest upon sand of different colors; some species of Paravilla Painter have become conspicuously whitish, a close match of unusually whitish desert situations. Very few observations have been published concern- ing the courtship and mating of bee flies. In south Texas, south of San Antonio, June of 1964, I had an opportunity to witness the nuptial activity of a pair of small individuals of xoprosopa fascipennis Say ; they pursued each other in small circles among the low vegetation 5 or 6 inches above the ground, facing each other; this continued several minutes; mating was un- successful, perhaps the flies were frightened away. In July of 1964 in Minnows Valley, Lafayette County, Mississippi, at 2 p.m., I found a mated pair of Towxo- phora amphitea Walker, attached end to end and resting upon leaves of low-growing sneezeweed, Heleniwm amarum (Raf.) Rock. After watching them some time I tried to catch them, missed them; they flew away, but when I returned to almost the same spot 15 minutes later I found them still mated. Linsley and MacSwain (1942) comment on the mating of Anthrax limatulus vallicola Marston (reported as Anthrax sp., near fur Osten Sacken) which took place during the warmest part of the day. The pairs remained in copulation for a considerable period of time and when disturbed flew away without separating. Sartor (in litt.), i extended studies of Anthrax limatulus and its relationship to Anthophora abrupta Cresson, observed mated pairs only twice, one of these at midmorning. He noted that the male in resting on a cliff was oriented upward, the female downward. The author has found paired Lepi- dophora lutea Painter resting on sneezeweed flowers in July, and has seen others attempting to mate. Du Merle (1966) has recently published an excep- tional study of the mating and related behavior of Villa quinquefasciata Wiedemann. He was able to contrive a remarkably successful breeding cage and was able to induce the species to mate easily in captivity. He found that the individuals refuse to feed and their longevity averaged 4 days with a maximum of 6 days, which com- pares closely with our own findings for Anthrax Scopoli species. The females fill up their perivaginal pouches with sand provided for them in sand boxes within the cage and then eject their eggs toward the base of smooth white stones resting on mee wire. Using this technique he was, in 1965, able to secure 42,000 eggs from 503 adults of this species. Eggs were collected with a vacuum cleaner. See figures of cages in this paper. BOMBYLIIDAE 35 Generally speaking bee flies oviposit by flicking the eggs toward the oviposition site or in many species by inserting the abdomen within the soil, as do certain species of asilid flies. In May 1964, on the outskirts of San Antonio, I watched a large, orange-red species of Bombylius Linné use both methods of oviposition in dense, old-mesquite forest. It is of interest that the color of this species is very impermanent, the bright red fading to a dirty brownish yellow. The population of bee flies will vary and will depend in any area on the extent to which the host species are available and also upon the abundance of flowers to which the species is attracted. At times they are extra- ordinarily abundant. I have found two hundred or more Sparnopolius lherminierii Macquart concentrated in less than an acre of Helianthus sp., and even greater swarms of western species about clumps of flowers in the highlands of New Mexico. In the genus Poecil- anthrax Osten Sacken, which attacks sod worms and cutworms, a few species are quite abundant, others dis- tinctly rare. Van Duzee (1931) comments on swarms of a species of Conophorus Meigen along the Salinas River of California. Probably no instance of greater con- centration of bee flies has ever been noticed than that recorded by Calvert (1881) having to do with the abun- dance of Callostoma fasciatum Fabricius, a locust-egg predator. In eastern New Mexico, in 1965, my wife and I found a species of Phthiria Meigen so abundant that as many as 25 individuals crowded in upon each Helianthus flower head; this occurred over a wide area. One of the factors concerned in bee fly populations must certainly be competition for available hosts. Walkden (1950), in studying life histories and in rear- ing large numbers of cutworms and sodworms from western prairies, shows that besides being attacked by bee flies these worms have many tachinid and hymenop- terous enemies as well. Most areas have meloid com- petitors. Many bee fly eggs must be lost by indiscrimi- nate scattering. Concentration of bee flies in terms of genera and species is especially interesting when considered in terms of small areas or specific localities. Such species con- centration depends largely upon the complexity of the local environment. It also varies from month to month. Jack Hall (personal communication) informed me that a year of continuous intensive collecting in a single long canyon in the Colorado desert yielded nearly 50 species of bee flies. No such richness of bee fly fauna can be found in the eastern United States. I have collected 28 of the 38 species of bee flies so far known from Mississippi in one ‘single small valley, lying between wooded hills and cleared for the development of more than a hundred small minnow-rearing ponds. This could hardly be matched by any other spot in northern Mississippi. However, the whole fauna of Mississippi 1s depauperate with respect to this family. Numerous trips have been made by the author, accompanied by his late wife, into the southwestern United States, southern Texas, and one extensive trip into Mexico for the purpose of col- lecting and studying these flies. In this extensive trip into Mexico, in August of 1962, about 75 collecting sites were sampled by the three collectors on the trip. The total bee fly fauna for any one spot did not exceed 22 species and only one such spot yielded this many bee flies; this was approximately 40 miles east of Guadala- jara. Similarly in south Texas in May of 1964, only one spot, a growth of virgin mesquite together with dense undergrowth and thickets, yielded 23 species. It would require year-round collecting to learn what the total fauna of these two bee fly-rich spots might be. The enemies of bee flies are numerous, and there are at least 11 kinds. For soil-living bee flies there is prob- ably no greater enemy than mold, which in hot, wet seasons undoubtedly destroys both larvae and pupae. Next in importance in all probability are the birds. Frick (1962) obtained some interesting data; he found that birds were the most important enemy of emerging bee fly pupae of Heterostylum robustum Osten Sacken. Blackbirds, larks, sparrows, and magpies all fed upon them at the nesting sites; while these birds destroyed bees also, they appeared to remove a higher percentage of the flies because the wriggling of the pupae at or near the surface tended to attract attention. In 1956 Frick found 29 bees and 48 bee flies in the stomachs of 24 birds; in 1959 he found 85 birds contained 119 bees and 44 bee flies; none of these were blackbirds. Almost certainly the flycatchers and the vireos catch some bee flies; these types of birds become more prominent in the tropics. Hardy (1920) noted an insectivorous bird in the act of catching a species of Systoechus Loew. Krombein (1967) comments upon the prevalence of eulophid parasites in trap nests placed by him for the attraction of bees and wasps. He noted one species of the bee fly was attacked by these chalcidoid parasites. Shrews and mice may be expected to consume some larvae and pupae, especially of species like some Villa Lioy which live in a forested cover of leaf mold. Frick (1962) found that mice destroyed some bee fly prepupae which were removed, eaten, and the skeletons dropped in runways; he also noted the skunks destroyed large numbers of pupae, and he believed that they ate the larvae as well since none could be found where they had dug. Robber flies have been recorded as including bee flies in their prey. Fattig (1945) and Linsley (1960) both record such instances. Frick (1962) found only one instance of a robber fly having seized an individual of Heterostylum robustum Osten Sacken, but he noted that tiger beetles and the ant Formica fusca Latreille were minor predators. I have captured a number of asilids that were holding bee flies on which they were feeding. I have found numerous instances of where crab spiders (AZiswmena) and the green lynx spider Peucetia viridans Walckenaer, which hide in flower heads, have captured bee flies. These lynx spiders, family Oxyop- idae, occur from coast to coast, but the group is more 36 BEE FLIES OF THE WORLD common southward. This is also true of the ambush bugs of the family Phymatidae; they catch bee flies. Another enemy is represented by certain species of bembicid wasps which favor dipterous insects as a source of prey. Presumably some lizards, more especially the anolids, capture some bee flies, although I have seen no records of such captures. Anolis carolinensis Voigt is not often seen but is likely more abundant than realized. Anolids increase in number in the warmer latitudes. Studies by the author of stomach contents of Sceleporus species and of H’wmeces species show that their food consists almost totally of spiders, at least in MississippI. I have seen no record of either mantids or reduviids capturing bee flies, yet in all probability they do some- times clutch these flies when such predators are lurking in flower heads to which bee flies come. The relationship of bee flies to flowers is a matter of considerable interest and one to which collectors and students are just beginning to pay attention. While I have listed most recorded flower preferences under each genus, some generalizations are pertinent here. Robertson (1928) made an extended study of insect visitors to flowers in the vicinity of Carlinsville, Tlinois. Out of 396 species of flowers in 62 plant families, a total of 169 flower species in 41 plant families were visited by bee flies. A total of 29 species of bee flies were found in the area. There were 8 species of Bombyliinae in 5 genera and 13 species of Exoprosopinae in 3 genera; the remainder consisted of 1 species of Lomatiinae, 2 species of Phthiriinae, 2 of Toxophorinae, 1 of Systrop- inae, and 2 of Geroninae. Thirteen species of bee flies came to Rudbeckia triloba Linnaeus, the most attractive flower. Of the flowers, 86 species were visited by only 1 species of bee fly, 53 of them by Bombylius spp. only. The Bombyliinae visited 109 species of flowers in 11 families, of which 41 were composites, and the Exo- prosopinae visited 89 flower species in 19 families, of which many were composites. Seasonal factors were important: there were only 38 records of the genus Bombylius having visited a member of the Compositae, all the other Bombyliinae were of the genus Systoechus Loew or Sparnopolius Loew. The genus Z'oxophora Meigen visited 18 species of flowers in 7 families; Lepi- dophora Westwood appeared on 3 species of Com- positae, and Systropus Wiedemann visited 10 species of flowers in 3 families. See these genera in text. Langhoffer (1902) discusses for a few European species of bee flies their color preferences and the length of time spent on each floret. Knuth (1898-1909) gives some interesting but scattered comments on the rela- tion of Bombylidae to flowers in his extensive hand- book. He notes that the species of Bombylius Linné can bore into succulent tissues as well as extract nectar, and he had seen them probing into nectarless flowers. Mueller (1881) in his work “Alpenblumen” has a few comments on the bee flies that visit flowers. Morphology of the Immature Stages EGGS Bee fly eggs are produced in large numbers. Apart from watching various bee flies thrust their eggs toward the oviposition site, few writers have handled or ob- served the eggs. Eggs of known species vary from a length of 0.28 mm. and a width of 0.12 mm. in the case of Anthrax analis Say (Shelford, 1913) to 1.2 mm. in length and 0.7 mm. in width in the eggs of Hetero- stylum robustwm Osten Sacken collected by Bohart et al. (1960). It is clear that the eggs of many of the smaller species of bee flies such as Phthiria Meigen and Aphoebantus Loew, and the still smaller Wythicomyia Coquillett must be quite minute. Heterostylum eggs have a smooth, pearly white chorion with a mucilagin- ous coating. Marston (1964) found the eggs of Anthrax limatulus Say to be 0.5 mm. in length and 0.29 mm. in width. Sartor (in litt.) collected 850 eggs of the same species but stated that the eggs averaged 0.2 mm. in width; they were oval, deposited in two-days time by two ovipositing females. This bee fly was attacking cliff nests of Anthophora abrupta Say and as many as 97 eges were thrust into a single vial embedded in the bank. They were densely coated with mucilaginous material, obscuring the iridescence, and completely covered with minute, soil particles, leaving the color varying from white to tawny. According to Shelford (1913) the eggs of Anthrax analis Say are brown and not adhesive. Du Merle (1966) describes the methods by which he was able to obtain many thousands of eggs of Villa quinquefasciata Wiedemann; these eggs meas- ured 0.7 mm. in length, 0.5 mm. in width. FIRST-INSTAR LARVA These larvae are planidial. Sartor (in litt.) found this larva in Anthrax limatulus Say to be about 1 mm. to 1.2mm. in length. It is slender, elongate, cylindrical. The coloration is white, the internal tissues revealed by the transparent integument. He found 3 thoracic and 9 abdominal segments; each segment was approximately equal in length and similar in shape except the terminal segment which was conical and smaller in size. There is a pair of long stiff bristles on each thoracic segment, which Clausen calls heavy spines, and an additional pair of caudal setae extending backward from the terminal segment. Bohart et al. (1960) found that the planidium of Heterostylum robustum Osten Sacken has large mouthparts, with a median pair of tonglike hooks, flanked by long, slender, maxillae that bear club-shaped palpi tipped with long bristles; moreover he found that the head and neck have 8 pairs of small bristles. Sartor found a pair of pseudopods located on the anterior margin of abdominal segments 2 to 6 and on the pos- terior margin of the eighth segment. Bohart described these on the eighth segment as double, and he considered the respiratory system as peripneustic, although the BOMBYLIIDAE 37 Text-Ficure 1.—a-c: Eggs, larvae, and pupae of Anthrax limatulus Say; p, larvae, “in situ,” within wasp nest. Pp Pp Y3 ? > ? p 38 BEE FLIES OF THE WORLD \s\ we oe - oleae Vv ER & an KR ae ee ela es (yy i fa oe, A iat (lean es: Zaye z % (4 eee WG B | } yf fn H ve a) Ww 2 - A Bc D E Text-Ficure 2.—Planidial or first-stage larvae of bee flies: a, Systoechus species, after Brooks; B, B. pumilus Meigen; c, B. oulpinus Wiedemann, after Engel; p, late-stage larva, B. vulpinus Wiedemann; r, Heterosty- lum robustum Osten Sacken, after Bohart. abdominal spiracles anterior to the penultimate ones were poorly developed and possibly vestigial. Sartor described the first-instar larva as highly active and hav- ing a humped appearance in the thoracic region when it is moving forward. Berg (1940) presented a remarkable morphological study of the larval stages of Systoechus vulgaris Loew ; see figures reproduced here. SECOND-INSTAR LARVA Working with Anthrax limatulus Say, Sartor found that these larvae were 3 mm. long after the first moult, but Marston (1964) found them only 1.5 to 2.0 mm. in length at the time of this first moult. Sartor also states that the milky white tissues are much in evidence in this stage, and that 12 body segments are easily counted. The larva is compressed dorsoventrally, with broad pleural protuberances on the third thoracic segment and on the first to fifth abdominal segments; each seg- ment is well marked by segmental lines. He found the head shallowly invaginated with only the dark brown mouth hooks discernible. The spiracles were visible on the eighth segment and partly hidden dorsally by the overhanging seventh segment, but project over the ninth segment. The prothoracic spiracle according to Sartor is crescentic and opens posteriorly with 11 barlike sclerotic spots. The caudal spiracle is round with 8 barlike schlerotic spots. Berg (1940) gives an excellent description of this larval stage as well as the first and third stage in his work on Systoechus vulgaris Loew. Both Sartor and Bohart comment on the difficulty of finding second-instar larvae. Sartor states that the second-instar larvae of Anthrax limatulus Say reached the third stage in 20 hours and was then 5 mm. long. Bohart considered that this stage in Heterostylum robustum Osten Sacken lasted only about 12 hours. THIRD-INSTAR LARVA In this stage the bee fly larva increases greatly in bulk and length and also becomes characteristically crescentic or C-shaped. There are no bristles or tuber- cles for locomotion; it is now a grub with very little movement. The tight application of the concave portion about the mouth with doubtless some vacuum developed helps to hold it attached to the host. Perhaps the arched shape of the larva doubly accommodates the parasite to the surface of the host and the surface of the cell. It is mi‘ky white due to a transparent integument and white internal tissue; numerous oval, opaque white fat bodies make their appearance in the abdomen. This instar is amphineustic and cylindrical, with the bend in the body located in the anterior half. Segmental lines are emarginate, says Sartor, and are visible on segments 2 to 5; pleural protuberances are present on the first through the seventh abdominal segments. The eighth abdominal segment is small and forms a ledge between the seventh and ninth segments upon which the poste- rior spiracle is located. This spiracle is circular with usually 8 to 9 sections. The anterior spiracle is cres- centic, open posteriorly and has 11 sections. See accom- panying figures of the third-instar larva. Not all bee fly larvae become C-shaped in the final stages. Those mature larvae of Lepidophora Westwood which we have obtained from caterpillar-filled nests of species of wasps were straight and linear in form. FOURTH-INSTAR LARVA Bohart et al. (1960) describes a fourth-instar larva for Heterostylum robustum Osten Sacken, but this is the only work in which I have found reference to such an instar and the only authors as far as I am aware who refer to such a stage. Berg (1940) refers to none, and Sartor, working with Anthrax limatulus Say, and Marston (1964), with the same species, found none. PUPAE The pupal stages of bee flies are remarkably interest- ing. Pupae of this family are better known than larvae of bee flies, because so often only the pupal skins remain in chance rearing discoveries. These pupae are free and mobile and possess distinct sheaths which encase the mouthparts, the antennae, the wings, and the legs. When the fourth-instar larvae changes over to a prepupa, or pupa, it is at first milky white and gradually turns pale yellow; it gradually darkens to brown and may become almost black before emergence and tends to become very active in the late stages of pupal life, wriggling about and extending or telescoping the abdominal segments. We have kept many hundreds of both larvae and pupae of Anthraa aterrimus Bigot and Anthrax “imatulus Say alive on damp sand. They may be kept in either state for long periods on sand in a refrigerator. As this is written I have plump, white larvae that have remained in a petri dish within a refrigerator for several years. The pupae of bee flies resemble those of asilids in many respects. Asilid pupae bear thornlike spines upon the dorsum of the abdominal segments which are sharp or blunt, but usually sharp, and are vertically erect even if curled backward. Bee fly pupae may in rare instances as in BOMBYLIIDAE 39 Text-Ficure 3.—Mouthparts of bee fly larvae of the two principal divisions of the family: a and , front and lateral aspects of larva of Systoechus vulgaris Loew; c and p, front and lateral aspects of larva of Anastoechus barbatus Osten Sacken; § and F, front and lateral aspects of Anthrax limatulus Say. Figures A-p redrawn from Painter (1962); Figures E-r drawn by William Martin. E F Phthiria spp. show similar, sharply pointed erect thorns, Bee fly pupae rather characteristically show a trans- but in such cases they arise from the anterior part of — verse band of close-set chitinous rods attached across what might possibly be called the anlage of the thorns, | many of the abdominal segments, from which at either whereas in asilids they arise from the posterior edge. _ posterior or anterior end or from both ends, the apices 40 BEE FLIES OF THE WORLD Text-Ficure 4.—Pupa of a bee fly, Anthrax limatulus Say. of the rod are apt to be sharply erected into a sharp hooklike spine or thorn. From a study of pupae of 8 subfamilies, 18 genera, and 26 species that are before me, this distinction holds good. Occasionally, as in Aphoebantus (Tvriodites) mus Osten Sacken, these longitudinally placed chitinous rods are so short that the posteriorly erect spine is suggestive of the spine armor of an asilid. Malloch (1915) states that long, slender, curled hairs alternate with the chitinous rods and spines; while such hairs may be irregularly present there are none in Systropus Wiedemann. A second character of importance in distinguishing bee fly pupae from other families lies in the presence and character of strong, chitinized thorns upon the head capsule. In asilids this cephalic armament as noted by Malloch (1917) consists of a unified, fused group of 3 very strong spines present on each antennal sheath and lying below the anterior, apical, cephalic thorn. He notes that such are absent from the pupae of Leptogaster Meigen. In the opinion of this author this is indeed not an argument for separate family status of these obviously asilid flies. There are bee fly genera that lack the usual cephalic spines, as in the case of Systropus Wiedemann and Towxophora Meigen, yet these are in every respect bombyhids; it does not make sense to create separate families for these two bee flies any more than it does to put the leptogastrine asilids into a separate family. Characteristically asilid pupae have the postcephalic thorn group therefore consisting of 3 well-developed clustered thorns; whereas, there are never more than 2 in this position in the bee flies, and these two are often separated at the base. It should be here noted that other bee fly pupae vary from the above characterization, such as the curious Pstloderoides m the Cyrtosiinae and Glabellula Bezzi (as Pachyneres Greene), and for which figures are here reproduced. All bee fly pupae and pupal exuviae tend to show some curvature. The four most important areas from the standpoint of generic and subfamily characters lie in the head armament, that of the dorsum of the prin- cipal abdominal segments, and the apical, terminal processes of the pupa together in some instances with the armament of the abdominal pleurites. The head armament usually consists of four groups of strongly developed heavily chitinized thorns, each paired, so that there are 8 thorns. The posteriormost pair may be very much reduced, as in ' Thevenemyia auripila Osten Sacken. In Neodiplocampta Curran the outer spines of the lateral pairs are much reduced, but there is an additional small pair below the lowermost of the principal spine groups, making 10 im all. In Hemipenthes Loew all the spines are reduced except the anterior pair, which are approximated, semitrun- cate, and shovellike. In 7'riodites Osten Sacken there are 2 additional unpaired spines, one on front and one behind the lower pair, and moreover the anterior spine BOMBYLIIDAE Al d << war \ 2 Text-Ficure 5.—Pupae of bee flies: a, Thevenemyia species; B. Systropus macer Loew. and both members of the lateral pair are fused into groups of 3. In Anthracinae each group of 3 becomes fused into a hemicircular arch of 6 spines. The base of the wing sheath in Heterostylwm Macquart bears 2 spines on each side. The dorsal abdominal armament varies from genera in which neither end of the longitu- dinal chitinous rods is raised into erect thorns or spines, as in Hemipenthes Loew; here each end is about equally well developed. In Hwoprosopa Macquart usually, but with exceptions, each end is raised into remarkably stout, erect, sharply thinned but bluntly rounded, shovellike plates. In Walkeromyia Paramonov the dorsal rods end in sharper spikes and give way laterally to curious, con- spicuous, very wide, quite long, flattened, closely ap- pressed swordlike outgrowths. In Phthiria Meigen the few central rods end in long, strong, strongly curved, erect, sharp spikes or spines, which arise from the pos- terior ends of the rods. Terminally the bee fly pupae vary from the Anthracinae where the paired processes are heavily chitinized, elongate, narrowly separated, bifid or fin-shaped and accompanied above and below on each side by short, small, tubercular spines, and another, paired or unpaired, placed medially upon the preceding segment, to the majority of genera in which these terminal processes are small, slender, consisting of one slender, sharply pointed protuberance with or with- out an additional shorter one below. In Walkeromyia 42 BEE FLIES OF THE WORLD yy x, ) vit oe Wy > oe S—e — A ——— aos Sy NS 1K Wee Sih LA SS Text-Ficure 6.—Pupae of genera of three subfamilies of bee flies: a. Anthrax tigrinus De Geer; 3. Phthiria sulphurea Loew; c. Toxophora amphitea Walker. BOMBYLIIDAE 43 Paramonoy and related forms the pupa ends in a curi- ous, long, widely separated, spine tipped, fleshy protu- berance. Finally in some genera there are important differences on the abdominal pleurites. They may have a tuft of 3 very long, very stiff, uniformly thick hairs, as in Zoxophora Meigen, to as many as 9 very long, flat- tened curved epikolites or swordlike outgrowths, or as in most genera only fine, long hairs. In most genera the anterior, circular, flattened spiracle is composed of cor- rugated ridges which exactly resemble a coiled milli- pede. But there may be curious specific differences in the number of lobes; they are reduced to 5 in Anthrax irroratus Say. T have prepared a list of the subfamily characteristics of all those subfamilies for which I have pupae. To aid others in the fascinating study of the pupae I have illustrated all those for which I have material. To further help students of these flies I have included Malloch’s key (1917) to 16 species, 9 of which are not among the 26 species before me. In bee flies the last larval stage is usually followed by a prepupa (pronymph of authors): in this stage the head thorns are partially developed and the longitu- dinal plates of chitin are already laid down upon the dorsum of the abdomen. Séguy (1926) who illustrates the pupae of 11 species, all figures rather small, also illustrates the pronymph of several species. The presence of these stout spines is associated with the difficulties the pupa meets in emerging to the surface from mud nests of wasps, which do not usually offer much obstruction, to the deeper lying layers of clay and earth under which many subterranean forms find them- selves. If the comparative morphology of the pupae is valid, as I believe it is, we must conclude that the nearest relative to the subfamily Systropinae is the genus Toxophora Meigen, and that these two subfamilies are relatively closely related. Also it is at once apparent that Lepidophora Westwood does not find its relation- ship with Zoxophora Meigen but rather is related to the Bombyliinae; while I suspect it should be regarded as a tribe within the Bombyliinae, I place it with some reluctance in the Toxophorinae. Also it is clear from these studies that Anthrax tigrinus Fabricius should be placed in a genus entirely separate from Anthrax Scopoli, Argyramoeba Schiner, or Spogostylum Mac- quart; it should be placed within the genus Walkero- myia Paramonov or a new genus should be erected for it. The former course is preferable. KEY TO THE SUBFAMILIES OF BEE FLIES AS KNOWN FROM PUPAE 1. Antennal sheaths present and prominent, distinct and fleshy, and arising from and proceeding backward from anterocephalic thorns which may be almost completely fused, side by side. Dorsum of pupae with 7 rows of close-set, longitudinal, slightly curved, posteriorly distinct but also appressed spines which are not raised above the surface at apices. Cephalic thorns range from 1 fused pair to 2 distinct pairs... . 6 3 No distinct antennal sheaths present! Usually a pairs of cephalic thorns present; some of which may be reduced, semifused or vestigial. Dorsum almost always with chitinous longitudinal ridges, one or both ends of which are raised into sharp, erect spines. .........8 2. Cephalic thorns reduced to a single, almost completely fused pair, with however a minute stublike tubercle adjacent at the level of the second antennal segment. Apex of pupa with 1 or 2 pairs of mammillate nodules, the more poste- rior pair semifused. Dorsal rows of curved spines indis- _tinct anteriorly. Lower lateral edge of first 6 abdominal pleurites with curved, attenuate, sharp-tipped, fleshy pro- trusions. Some species characters afforded by the flutings on the thorax. Proboscis sheath fused. Based on different species of Systropus Wiedemann . . SYSTROPINAE Second pair of cephalic thorns small, borne on the middle of the divergent antennal sheaths. Anterior thorns fused except the divergent apices. Proboscis sheath divided longitudinally. Posterior apex of pupa with 2 pairs of widely separated, short, triangular thorns; lower pair curved. Upper and lower pair as far apart as thorns of each set are separated. First 6 abdominal pleurites each with 3 extraordinarily long, uniformly stout, finely pointed bristles. Based on different species of Toxrophora Meigen . . TOXOPHORINAE 38. Chitinized spinous longitudinal ridges restricted to first two abdominal segments and these two segments have only 3 ridges, each raised anteriorly into strong, erect, long, backwardly curved spines. Posterior apex of pupa with a pair of widely separated, thin, posteriorly attenuate and pointed, bladelike protrusion. Thorn bearing part of cephalic capsule circular, with 3 pairs of strong, sharp thorns, the anterior ae especially sharp, widely sepa- rated Bite suede . PHTHIRIINAE Chitinized ridges present on ieccona to eighth segments and usually with the posterior part ending as an erect, ver- tical, sharp or blunt spine, sometimes both ends with erect spines. Sometimes first segment also with a few spines. 4 4. Each anterior spine and each lateral pair gathered together into a unit of 8. Units of each side may be separated or fused together infront ..... Naeem) uteri ey All of the cephalic spines are separate and Gistince even if somevofathem) aressmally wae Mon einen ron end wi Both units fused together at base transversely so that all 6 spines though distinct, stout and long, are gathered to- gether into a hemicircular unit remote from the posterior pair of spines; all 6 arise from an elevated common base. Apex of pupa with a large, chitinous, elongate, black, bifid or finlike process, narrowly but distinctly separated. Also the base of the tenth segment has a pair of smaller, black spines above and below; also a paired or unpaired small spine dorsally in the middle of the ninth segment. Based on many species of Anthrax Scopoli. ANTHRACINAE Each anterior unit of three spines on the head capsule SEDAT ATS MANE LAN sites wierane eal eter fal Na taby Nualicante tye aah urn LG 6. The spines of the head capsule are large, blunt, or rounded or spadelike. Chitinous ridges of the dorsum with sharp erect spines both anteriorly and posteriorly at least on the middle segments. The spines give way laterally to long, thick, appressed, flattened, swordlike hairs and on the sides of the first segment to many, long, curled hairs. Pleurites with from 4 to 8 or more long, flattened, con- spicuous, swordlike hairs or outgrowths. Apex of pupa BEE FLIES OF THE WORLD 44 nd po of two subfamilies of bee flies together with dorsal views of anterior a 7.—Pupae of genera T-FIGURE TEx of Villa lateralis Say. armament 10. 11. 12. Cotte BOMBYLIIDAE with a pair of widely separated, fleshy, attenuated proc- esses ending in a spine. Very large flies. Tribe Walkero- myini . Balti . ANTHRACINAE The spines of the ead eapeniel are long and sharp. Poste- rior pair of spines strong with an unpaired additional spine in front and behind. Apex of pupa with a single pair of very small spines, widely separated, arising from a tubercular swelling. Longitudinal ridges quite short ending in an erect spine only posteriorly. No spectacular hairs, swordlike outgrowths, etc. Aphoebantus mus Osten Sacken . LOMATIINAE Posterior pair of spines of the head capsule weak or much reduced... . AVP AENeRi pst. ct PataeHed AS Posterior pair of eines (ell developed A TAL gt oan ee htt eg ante) Apex of pupa with a pair of small, widely separated, attenu- ate spines borne on a basal protuberant swelling. The- venemyia auripilus Osten Sacken, Tribe Eclimini. BOMBYLIINAE Apex of pupa with a large, triangular, heavily chitinized more or less triangular, widely separated plate on each side. Lepidophora lepidocera Wiedemann. Tribe Lepido- phorini . BOMBYLIINAE Chitinous ridges of dorsum with spines in front as well as behind... . Be ree ey ba aU Chitinous ridges of aon sum apie nik? beni or not at all. 10 Chitinous ridges without spines either in front or behind. All the cephalic spines much reduced, except the anterior pair which are blunt, obliquely truncate, narrowly sepa- rated. Spines at apex minute, slender, widely separated arising from a vertical ridge which bears below two other even smaller, tubercular spines. Hemipenthes Loew. EXXOPROSOPIN AE Chitinous ridges with a row of erect spines on their pos- terior ends. Villini; Paravilla Painter, Neodiplocampta Curran, ete. Hxroprosopa Macquart, spp. EXOPROSOPIN AB Apex of pupa with stout, heavily chitinized, upward thrust spines. Chitinous ridges of dorsum very strong, the ante- rior and posterior erect spines sharp, long and conspicu- ous. Hxoprosopa Macquart . . EXOPROSOPINAE Apex of pupa with a pair of slender, weak, upward thrust Spinessonkeachysidey as cy seh. whee edn cael, Gaede Apex of pupa with a pair of additional, stout, shovellike plates or spines in the middle, dorsally, semifused, above the slender apical spines. Base of wing with a pair of small spines on each side. Tribe Heterostylini. BoMBYLIINAE Apex of pupa with only a pair of weak, slender, upward turned, slender, lateral spines. No spines at base of wing. Bombylius Linné, Systoechus Loew, ete. . . BOMBYLIINAE KEY TO PUPAE (FROM MALLOCH, 1917) Upper central pair of cephalic processes thorn-like, widely separated from their entire length; lateral cephalic proc- ess or processes thorn-like, but little if any shorter than the central pair... SERS Katee Upper central pair of Eepnalich arOCORECS miout not thorn- like, contiguous for the greater portion of their length; lateral cephalic processes tubercle-like, much shorter than the central pair... . aipetie ie vege ty Apical abdominal segment feominativen ina aie of long, tapering, backwardly directed thorns; first abdominal segment with the postspiracular hairs as ae as head and thorax combined (Spogostylum) .. . ‘ 5183 Apical abdominal segment usually more or ies imeneated and with an upwardly and backwardly directed upper OO a | 10. 11. 13. 45 process and one or two smaller protuberances below it; first abdominal segment with the postspiracular hairs much shorter than head and thorax combined .....5 Head with 4 long thorns on upper anterior margin, the lower one on each side with a small protuberance at base on under side; the pair of thorns on lower portion of central line of face large, their bases contiguous; hairs on head and thorax very long; laterad of the short thorns the transverse armature of dorsal abdominal segments 2-6 consists of 2-8 long, widely placed rounded hairs ite . Spogostylum anale. Head w vith 6 RROLt ero fahaseng on upper anterior margin 4 The pair of thorns on lower portion of central line of face small, their bases subcontiguous; hairs of abdomen, ex- cept those of verse armature of dorsal abdominal seg- ments 2-6 consists of 12-20 long, closely placed, flattened hairs . 3 5 . Spogostylum simson (p. 393). The pair of thors on lower portion of central line of face large, their bases, subcontiguous; hairs of abdomen, ex- cept those of basal segment, normal. . . . Spogostylum albofasciatum “(p: 395). puritan. lateral margins of head each with 1 strong thorn, upper anterior margin with 2 such thorns, making 4 in all; labrum with a bifid thorn . Chrysanthraz fulvohirta. Antero-lateral margins of head each with 2 strong thorns, upper anterior margin with 2 such thorns, making 6 TD EMS og ee) os Bane syd Asai atte ena Labrum with a Btrone) bifid nore wing with a median subcostal protuberance . . Aphoebantus mus. Labrum unarmed .... . Be Ree SASEL ( The stout thorns.on dorsal ahaounall Bcaenante turned up at bases and apices... . Bh ari ee ite} The stout thorns on dorsal abdominal lennents turned up at apices only. fy... SA iced a tenii L b Lower lateral cephalic aon oan a asin nike organ pro- jecting on its under surface at base, the apex of which is armed with several hairs . . Bombylius. Lower lateral cephalic thorn without a palp-like organ on under surface... . 5 : atviohernregecculte, Apical 3 segments Sithont the dorsal aan series of short thorns, armed only with slender hairs; no slender hairs interspersed between the short thorns of median portion of series on other segments . Hf 5 Anguramoeba Oedipus. At ona ine menuiein ate inl antepenultimate segments with dorsal transverse series of short thorns; slender hairs interspersed between the short thorns onallsegments . 10 Wings extending to apex of third abdominal segment, their color pale . Systoechus oreas. Wings extending short of ts apex of second abdominal seg- ment, fuscous apically . . Hxoprosopa fasciata? Transverse armature of first abdominal dorsal segment con- sisting of a series of short, stout thorns on middle portion, and a number of long, slender, closely placed hairs on each side . 3 4 . Exoprosopa fascipennis. Transverse armature of first abdominal dorsal segment con- sisting of a few widely placed hairs, the middle portion either entirely bare or with very slight indications of small tubercles which do not appear as distinct thorns . 12 Lower one of the pair of lateral cephalic thorns simple apically, but with a small wart-like protuberance at base on lower surface, the small wart bearing 2 distinct hairs ; Pee without discal Ia ech oe 6 3 é Spormonolus Gonna. Taweon one er ae. pair ae lateral cephalic thorns with a short subapical protuberance, the apex of thorn turned upward, base simple; wings each with a pair of protuber- ances, one about one fourth from base and the other near middle . . Anastoechus nitidulus. No well-developed pair as thorns on lower median portion of face . . Toxophora virgata. 46 BEE FLIES OF THE WORLD Text-Ficure 8.—Anterior and posterior armament of the pupae of three subfamilies of bee flies: a, cephalic spines of Toxophora virgata Osten Sacken; 8, cephalic spines of 7. amphitea Walker; c, caudal spines of T. virgata Osten Sacken; p, caudal spines of Aphoebantus mus Osten Sacken; £, caudal spines of Toxophora amphitea Walker; F, cephalic spines of Aphoebantus mus Osten Sacken; c, caudal spines of Phthiria sulphurea Loew; H, cephalic spines of P. sulphurea Loew. BOMBYLIIDAE AT — A well-developed pair of thorns on lower median portion of TEV EO} oi eee | open Scr cen eat eee oni EN Alp AUe: co eilicy mallee 14 14. Eighth ventral abdominal segment without hairs on disc . . EELS AAS Ua AN MMe aL ee Ba Hyalanthraz hypomelas. — Eighth ventral abdominal segment with hairs on dise . 15 15. Eighth ventral abdominal segment with 2 hairs on each side of disc; distance from the pair of thorns on lower central portion of head to apex of basal portion of sheath of mouth-parts about 4 times as great as distance from the latter to apex of proboscis . . Hyalanthraz lateralis. — Wighth ventral abdominal segment with 10-12 long hairs on disc; distance from the pair of thorns on lower central portion of head to apex of basal portion of sheath of mouth-parts about twice as great as distance from the latter to apex of proboscis . . . Hyalanthrazx alternata.” Morphology of the Adult Bee Fly Bee flies range in size from the large and conspicuous species, such as members of the genus H’xoprosopa Mac- quart which may reach a wing span of 64 mm., down to very minute, deserticolous floriphiles, such as Mythi- comyia Coquillett, which have much reduced wing venation; many of these are not more than 1.2 mm. in length, a few range down to 0.9 mim. Most species have a compact form, both in abdomen and thorax, the former being broad and short. In the remarkable subfamily Systropinae and to a lesser extent in a few other genera, the form is exceptionally long and slender and wasplike; these flies are rather bare in appearance, the microsetae scarcely noticeable. Most bee flies have fine, rather long vestiture, sometimes with coarser elements intermixed, sometimes composed entirely of dense plushlike pile. Some bee flies such as the genus Towxwophora Meigen have well-developed macrochaetae. A few species have closely adpressed scales, either fine and narrow or wide. A few bombyli- ids resemble syrphids, therevids, or empidids. While the ground color of the integument is more often black or dark brown, in many groups it is pale yellow, whit- ish, or red, and still other species show a pronounced brassy or bluish metallic sheen. Wings of bee flies show a highly variable venation and may be hyaline or may show very beautiful patterns of spots; iridescence varies. The head is usually approximately hemiglobular, but in the Tomophthalmae it may be almost completely globular due to the extensive cuplike development of the occiput. The vertex and frontal area are nearly plane with the eye; the eye is only rarely raised a trifle ; the frons may be slightly higher than the adjacent eye margins. The head is usually of nearly the same width as the thorax but may be more narrow and not infre- quently is much wider. The frons may show a trans- verse depression or fovea immediately in front of the antenna; it is rarely inflated or tumid. The eyes are always large and extensive, almost always occupying much the greater part of the head; they are convex in front, with the upper facets enlarged in the male, and in the subfamily Heterotropinae a dis- tinct dividing line separates an area of larger facets above, smaller ones below; this is true in males of Geminaria Coquillett. While generally bare, Bowden (1964) has pointed out that a few bee flies (Exoproso- pinae) do show fine, scattered ocular pile. In the male the eyes are usually holoptic or at least nearly touching, but in many groups they are dichoptic, or there are sometimes exceptions within the same genus (Hury- carenus Loew). In the genus Systropus Wiedemann the eyes are widely holoptic in both sexes. The post- ocular margin is variable and important second only to the character of the occiput in separating the two main divisions of the family. It may be nearly or quite plane, throughout its vertical length in the more gen- eralized bee flies, or sinuous, slightly concave, and in the higher forms it is characteristically indented to a varying extent and with or without a short anteriorly directed bisecting line. Three ocelli are present and they are very rarely reduced in size. The occiput varies from flattened to convex, and in the higher members of the family is characteristically tumid, greatly expanded posteriorly, with a dorsal, vertical, bilobate, foveate depression and a deep central sunken cavity. Many Usiinae have ventral lateral bullae upon the occiput. The character of the occiput is of primary importance in separating the major divi- sions of the family and is discussed further under the phylogeny and evolution of these flies. The oral opening or buccal cavity is always large but does vary some in size and still more in position in re- spect to the principal axis of the head; it is generally deeper below and at the point of emergence of the pro- boseis and the opening may slope almost or quite to the base of the antenna, eliminating what would prop- erly constitute the face as separated from the clypeal space. I restrict the face to the nonsunken part of the front aspect of the head, below the antenna, and before the clypeal area begins; below is the clypeal space which T call the oragenal cavity, or oral cup; its sides are usually sharp, or as Hesse (1938) says, carinate, and with a vertical wall. This whole area of the bee fly head from proboscis to antenna is most variable and impor- tant. The face as here defined may be almost entirely absent, or restricted to a small triangle extending later- ally below the front, ending with the gena, or again the face be extensive and even conical, or also tumidly swollen. At the apex of the labium are to be found the paired labellae, which are separable by the bee fly. Dimmock notes that on the inner side of each labellum there are 3 longitudinal grooves or channels held open by semi- rings of chitin at right angles to their axes and together on each side; they have been termed pseudotracheae but are not likely comparable to muscoid pseudotracheae. Curiously, in a few bee flies the mouthparts are greatly reduced or even totally absent, with a mere slit remaining in the oral space. The gena I define as the lateral-ventral space outside the oragenal cup or clypeal space and only rarely extending below the eyes; poste- riorly it would merge with the lower occiput; genae 48 BEE FLIES OF THE WORLD WCM Text-Ficure 9.—Mouthparts of adult bee flies of Exoprosopa sp. Expianation: 1, labrum, with labellum at apex; 2, mandible; 3, labrum-epipharynx; 4, maxilla; 5, maxillary palpus. may be absent; it is bare, pubescent, or scaled. The tentorial fissure lies outside the oragenal cup and is often knife-thin, or it may constitute a more con- spicous fissure. Medially it encloses the clypeal space. The proboscis of the bee fly varies from stout and short, as in the Anthracinae and Lomatiini, or even absent (Villoestrini) to very long and slender as in Bombylius Linné and Phthiria Meigen; it is sometimes 4 times as long as the head. It is entirely absent in a few bee flies. According to Dimmock (1881) and Peter- son (1916), it is composed of 5 parts. At rest the hypo- pharynx lies within a groove on the dorsal aspect of the labium and is covered by the labrum (labrum-epipharynx) Beneath and at either side are to be found the maxillae, which are very fine and slender, shorter than the other parts; all of these may be disclosed if the mouthparts are teased apart with a very fine-pointed needle or insect pin, the maxillae being the most troublesome to sepa- rate. The maxillary palpi lie at the base on each side and consist usually of 2 segments; there may, occasion- ally however be 3 segments, sometimes only 1, and the palpus is fused and semivestigial in Comptosia Mac- quart; the apical segment may be somewhat dilated and clavate, one or both segments may be covered with prominent hairs or scales; the trend is toward reduc- tion in number of segments. Besides the long, slender type of proboscis seen in Phthiriinae and most Bombyliinae there is the shorter, stouter type with the more or less expanded, almost muscoid-like labellum seen in Anthracinae which flies, when they feed at all, probably utilize pollen; Bezzi (1924) speculates upon the possible use of the curious facial brush in Corsomyza Wiedemann in flower polli- nation. Finally, in 4 genera the mouthparts range from vestigial to a mere slit, as in Villoestrus Paramonov; this trend in reduction was compared by Bezzi (1924) to similar phenomena in other Diptera, such as the oestrids and the tabanid Adersia Austen. In the bee fly Euanthobates Hesse there are curious coecalike proc- esses below the head (Hesse, 1965), which may be involved in pollination. The antennae of bee flies are usually set close to one another but may be widely separated in the Cythereni and a few genera of other groups. In the Bombyliinae BOMBYLIIDAE 49 Text-Ficurr 10.—Lateral aspect of the thorax of a bombyliid, Antonia suavissima Loew. Exp.anation: 1, mesonotum; 2, scutellum; 3, postalar callosity; 4, humerus or humeral callosity; 5, anterior spiracle; 6, mesopleuron; 7, pteropleuron; 8, metapleuron; 9, posterior spiracle; 10, metasternum, 11, posterior metasternum; 12, first tergite; 13, second tergite; 14, halter; 15, metanotum; 16, hypopleuron; 17, mesosternum or sternopleuron; 18, prosternum; 19, notopleuron; 20, anterior metapleuron. the antenna tends to be rather long and slender with the first and third segment of nearly equal length. The second segment is always small and beadlike in nearly every genus; Veosardus Roberts is an exception. The third segment may end without a style, or have a conspicuous one as in Othniomyia Hesse. Moreover, as a marked exception, Prorachthes Loew has the third segment rather widely dilated. The third segment fre- quently has one or two small microsegments at the apex, and sometimes an apical, or subapical spine set within a cup. The third segment is occasionally more or less plumose. The style has a whorl of hairs at the apex in the Anthracinae. The thorax in this family is usually quadrate, and widest posteriorly, but may be a trifle longer or shorter. Except in the Toxophorinae, where it forms a conspicu- ous collar, the pronotum is lower than the level of the prominent mesonotum which abuts against the upper occiput. The mesonotum itself is usually only slightly convex, but may be strongly arched anteriorly, and in some genera it is conspicuously humped; in such species the pleuron tends to be compacted and high vertically. Surface of mesonotum with usually only simple pile, often long and dense but looser, more scattered and thinned when scales are also present as in Anthracinae and various other genera. Bristles often present on prealar, supraalar, and postalar areas, and in the Toxophorinae there are large, long, conspicuous, arched bristles. There may be also acrostical, dorsocentral, and anteromarginal bristles. The scutellum is large and semicircular, subtriangular, either flattened or slightly tumid, the margin entire or rarely bilobate or emargi- nate. The mesopleuron is large, and frequently bears a dense tuft of upturned stiffened hair with sometimes additional bristles intermixed. The metapleuron may be quite bare in some forms but usually has a distinct tuft of stiffened or bristly hairs which may take the form of a vertical fanlike row. The hypopleuron is bare and the pteropleuron usually bare. The metanotum is variable, 50 BEE FLIES OF THE WORLD usually well developed and lying beneath the scutellum. The propleuron usually bears conspicuous pile and sometimes bristles. In Systropinae the strigula is prominent; as Painter and Painter (1963) point out, this structure has been variously named the membra- nous tubercle by Verrall (1909), the scutellar callosity by Bezzi (1924), and the foliate scutellar callosity by Hesse (1938). The term strigula was adopted by Willis- ton (1901) and reused by Carrera and D’Andretta (1950), and by Painter. It has the double advantage of priority and brevity. The prosternum varies from widely dissociated from the propleuron, as in Comptosia Macquart and many other genera, to barely touching on the mesal aspect of the latter, as in Zoxophora Meigen. There is also much variation in the form of the propleuron, and in many aspects of the posterior pleuron. The metasternum is sometimes very prominent and strongly developed so much as to compact the anterior sternum and coxae and crowd them forward; this is the situation in the Systropinae. On the metapleuron two divisions can be recognized corresponding to the episternum (or mesopleuron) and the epimeron (or combined pteropleuron, hypopleuron, etc.) of ‘the middle pair of legs; in this work these two divisions are merely called the anterior and posterior metapleuron ; the line that divides them reaches the top of the meta- coxa. The legs of bee flies with few exceptions are weak; they are adapted for alighting and resting in the often brief rest stops these insects make upon flowers, leaves, soil, rocks, logs, etc. In most genera the hind legs are much longer than the others and in some bee flies the anterior pair are greatly reduced. Middle and posterior coxae are short, except in those forms with a highly arched mesonotum; in Toxophorinae they are all three of equal length. In some bee flies there are moderately strong bristles beneath the femora but they are usually restricted to the hind pair and even there they may be lacking. The tibiae bear distinct rows of spicules or straight spinelike bristles of varying degrees of thickness, but here again there is often much variation and they may be absent even upon the anterior tibia; their presence or absence upon the front tibia has been used in classifica- tion, but it is a weak character; in a tibia which other- wise appears to be devoid of bristles very close inspec- tion will usually reveal minute, slender, bristly hairs erected obliquely above the basal coat of microsetae. Most tibiae have a well-developed apical circlet of spines as is again seen in asilids. In males of Walkero- myta Paramonov and in both sexes of Hxoprosopa Mac- quart, subgenus Pterobates, the hind tibiae are feathered with long scales. The lower surface of the front tarsi may have the typical, oblique, sharply pointed bristles modified into a fine, erect, fuzzy pile, and in a few genera the tarsal spiculae are thickened and rather blunt. The claws vary from almost straight, slightly curved to sickle-shaped. Sometimes the apical tarsal segment is broadened. The claws may be of about the same size throughout or the anterior pair may be quite minute; in some Exoprosopinae the claws have a long, sharp, basal tooth, and in the Bombyliinae the genus Zinnomyia Hesse is so characterized. Pulvilli are usually well developed, wide or narrow, sometimes re- duced or vestigial or absent; some genera have a trace of an empodial or more probably an aroliar structure. The halteres vary and more extended study might be profitable; usually with slender stalks the knob varies from oval to various shapes, apically truncate or excavated. The wing is well developed in most subfamilies of bee flies. It is large and extensive, as would be expected in insects which have such well-developed powers of flight. Posterior cells range in number from 5 in rare instances, as for example in three subgenera of H'’xoprosopa Mac- quart to 4, the usual number, down to 3 in the sub- families Usiinae, Phthiriinae and Gerontinae. The remigial area, or radial field of these flies, often shows complicated venation reminiscent of ancient and archaic types but yet greatly reduced in the subfamily Cyrtosi- inae, still more in the Mythicomyiinae. The genus E'mpidideicus Becker, as Melander (1946b) has re- marked, shows the greatest reduction of neuration in the entire family; here the auxiliary vein is vestigial, the second vein fused with the first, the third vein un- branched, the discal cell open on account of the loss of the posterior crossvein, and finally the ambient vein is absent. In certain African elements in the family, such as Tomomyza pictipennis Bezzi, and related species, and also in Bombylius namaquensis Hesse, the radial field is unstable and, though normally with 3 submarginal cells, may show 4 or more such cells because of the pres- ence of stump veins and supernumerary crossveins which may be remnants of a much more primitive venation; Hennig (1954) calls these accessory veins. Some species of Comptosia Macquart from Australia show an even more erratic radial field. Erratic venation often appears in the form of spur veins below the discal cell. The position of the anterior (radiomedial) crossvein is quite variable; it may le near the radial sector or base of third vein, at the middle of the discal cell, or quite distally placed near the outer third or fourth or sixth of this cell. The first posterior cell while com- monly open may be closed and stalked; the same is true of the anal cell. The alula may be broad and lobate, reduced, almost vestigial, or even absent and scales if present upon the thoracic squama are likely to border the alula as well. The ambient vein is commonly pres- ent but may extend only part way around the wing or be wholly absent. The wing of bee flies often shows beautiful patterns, which are especially characteristic within the subfamily Exoprosopinae. In many genera the wing is character- istically hyaline or even vitreous hyaline with at most the costal or costal and auxiliary cells tinged; there are numerous genera and species where the basal part of the BOMBYLIIDAE 51 Text-Ficure 11.—The wing of a generalized Recent bee fly, Dischistus mystax Wiedemann. Expianation: 1, costa; 2, subcosta or auxiliary vein; 3, first branch of the radius, or first longitudinal vien (Ri); 4, second longitudinal vein (R2 and R;); 5, third longitudinal vein (Ra and Rs); 6, anterior branch of third vein (Ri); 7, posterior branch of third vein (Rs); 8, first branch of medius or fourth vein (Mj); 9, second branch of fourth vein (M2); 10, third branch of fourth vein, or posterior intercalary vein (M3), with anterior branch of the cubitus; 11, ambient vein; 12, second branch of fifth vein (Cuz); 13, sec- ond anal vein; 14, main stem of the radius (R); 15, main stem of the medius (M); 16, main stem of the cubitus (Cu); 17, first wing is tinged with gray or brown or yellow; many species in various genera have the wings spotted, banded, or fenestrate with odd windowlike spots, or are mottled and infuscated. The apex of the wing in certain species of Comptosia Macquart may be milky white. The costal border at the base may or may not have a widened margin, may be with or without a bristly comb, and may or may not be preceded by a curious hook, called variously the costal hook, prealar hook, basal hook. Efflatoun (1945) points out that in Exo- prosopinae there is often a kind of fan formed of long scales on the base of the wing, actually arising from the mesonotum and which may conceal the hook; he notes that this was called the patagium erroneously by Lund- beck (1908) and erroneously the wing scale by Becker (1916) ; the term “wing scale” he reserves for a minute hook at the base of the alula, resting on the alar squama. Efflatoun calls the frenulum—again an apparent misuse of terms—a narrow skinlike band joining the inner base of the squama with the lateral base of the scutellum, and he further states that this so-called frenulum may bear on its margins the plumula composed of a small tuft of conspicuous and colored hairs or scales, espe- cially in Anthracinae and some Hxoprosopa Macquart anal vein; 18, remnant of third anal vein; 19, anterior branch of cubitus; 20, radial sector; 21, arculus; 22, humeral crossvein; 23, anterior, small, or middle crossvein; 24, anterior intercalary vein; 25, lower branch of the medius; 26, first or basal costal cell; 27, costal cell; 28, subcostal or mediastinal cell; 29, marginal cell (Ri); 30, first submarginal cell (Re and R;); 31, second submarginal cell; 32, first posterior cell; 33, second posterior cell; 34, third posterior cell; 35, fourth posterior cell; 36, third basal, anal or lower basal cell; 37, axillary cell, anal angle; 38, discal cell (1st Me); 39, second basal cell; 40, first basal cell; 41, alula, axillary or posterior lobe; 42, posterior, or medial crossvein. species; this may be the structure from which, by modi- fication, the strigula is developed in the Systropinae. Near the upper border distally within the second basal cell there is very often a small, pale, unpigmented spot, called the prediscoidal spot; it is much in evidence in the Villini. In a few genera there are scales upon the wings. In this family the abdomen is generally obtuse, stout, robust, scarcely as long as wide, or even more or less globular in genera in which the whole body is thick dorsoventrally, as in Platypygus Loew, Cyrtomorpha White, and a few other genera. In many genera, how- ever, the abdomen is moderately lengthened, short to long-conical or tubular; in the subfamily Systropinae the abdomen is remarkable for the long, slender, petio- late, wasplike form with the last segments a trifle widened. The vestiture of the abdomen varies from dense, erect, very fine, plushlike pile concealing the ground color, to modifications in which the lateral mar- gins bear appressed, extended, fanlike, radiating exten- sions of the pile or scales. Also the surface may be nearly bare, except for dense, minute, appressed micro- pubescence or microscalation, or there is very com- monly a mixture of loose, long hairs together with micropubescence or scales. Not infrequently the poste- 52 BEE FLIES OF THE WORLD rior margin bears a row of conspicuous, stout bristles, as in Cytherea Fabricius and many other genera. Lepidanthrax Osten Sacken is characterized by patches of bright scales on lateral margins and terminal seg- ments, and it is sometimes more glittering in the male sex; it is also true of some Anthracinae that the apex of the male abdomen is covered with brilliant silvery scales, which perhaps are important in courtship. As Hesse (1938) points out, the abdomen shows a minimum of 6 to 9 segments, males with usually one less visible segment and even in females the last seg- ment may be recessed beneath the one above, as in Heterostylum Macquart and many others. The last segments in the male form a hypopygium sometimes small, sometimes as in Usia Fabricius or Phthiria Meigen large and terminal, or again in many genera the hypopygium, which is composed of the associated parts of the male genitalia, may be reflexed ventrally, incon- spicuous, almost hidden, twisted, or turned to the right side (if viewed dorsally), as in the case of L'xoprosopa Macquart. Young (1921) in his study of the dipterous thorax found 7 abdominal spiracles, all lying in mem- brane. He noted the prominent first tergum and sternum inflated, large, and jammed up against the metathorax beneath the scutellum. I have not adopted his terminology for the thorax as I think a simpler one is better, so I have adopted most of Williston (1908) and Curran (1934). The female genital system often involves a series of lateral spines placed and attached just beyond the lateral valves of the seventh sternite. These are shown in the diagram here reproduced from the remarkable studies of Biliotti, Demolin, and du Merle (1965). These spines, characteristic of species of Villa Lioy and E’'xo- prosopa Macquart and other genera, are often utilized in deposition of eggs, although the eggs are sometimes ejected directly from hovering bee flies instead of being placed directly by the bee fly alight. I have seen the same species of Bombylius Linné in southern Texas utilize both methods of oviposition. Many other genera of bee flies have the female terminalia enclosed apically by an extraordinarily dense cloak of long, fine hairs; du Merle notes that the females of the species of Villa Lioy have a special perivaginal cavity. Hesse (1938) was the first to point out how truly indispensable are the male genitalia for the final eluci- dation of species. In his remarkable and incomparable works (1938, 1956), he illustrated dissections of 519 species in many genera of South African bee flies and has done the world a great service. Hesse devised his own terminology, overlooking the excellent work of Maughan (1935) and of Cole (1927) who dissected and illustrated and labeled the genitalia of 5 bee fly genera. Bowden (1964) has brought the terminology of male genitalic parts in line with those used by Snodgrass (1957) for the Tabanidae; Snodgrass did not, as far as I am aware, ever make any study of the Bombyliidae. I prefer the terminology of Cole and of Maughan, and I have used it on the accompanying series of text figures showing four levels of bee fly male genitalia. Following is a list of terms used herein compared to my under- standing of equivalent terms as used by several researchers. TERMINOLOGY USED IN STRUCTURES OF THE MALE GENITALIA OF BEE FLIES Terms used in this work 1. Basistylus : Bowden, Snod- grass. This is the large, basal, ventral structure, to which the dististyli are attached. 2. Dististylus: Cole. 3. Epandrium: Cole. This is the large, dorsal structure (tergum 9) to which the cereus is at- tached. 4. Axial system: Metcalf. 5. Basal part of aedeagus: This is the ejaculatory sac. 6. Ejaculatory duct and proc- ess. This is the apical part of aedeagus. Lateral and basal (ven- tral) apodemes. This, to- gether with ventral, or basal apodeme, forms the sperm pump. -l Terms of other authors believed equivalent Penis sheath: Metcalf. Tenth segment: Cole. Basal parts: Hesse. Basimere: Snodgrass. Gonocoxite. Clasper : Metealf, Telomere: Bowden, Snodgress. Beaked apical joint : Hesse. Epandrium: Van Emden, Hen- nig. Ninth (IX) segment. Tergum IX. Aedeagal complex: Hesse. Aedeagal sheath: Bowden. Middle part of aedeagal com- plex : Hesse. Chitinous box: Metcalf. Aedeagus of authors. Lateral ejaculatory apodeme. Basal and lateral keels of authors. Basal and lateral struts of Hesse. Paraphyses : Bowden. Sustentacular apodemes_ of authors. 8. Ramus. Basal margin of basimere (gonocoxite). Ramus: Hesse. Dorsal rami: Bowden. Anterior, or basal margin of ninth sternum ; apex may be split, or may be medially sutured. Hypandrium. 9. Epiphallus. 10. Sternum IX. Segment lying below at base of basistylus. 11. Cercus. Cercus of authors. Any male of a species of Bombylius Linné makes a good starting place for the study of the male genitalia. The hypopygium consists of two, symmetrical, capsular, shell-like structures, which are not independently mov- able. They are usually more or less recessed within the last visible tergum and sternum, but there are excep- tions; they may be visible only from below, and more or less reflexed backward, as in E'zoprosopa Macquart, but in Bombylius Linné, although directed downward, are visible apically. The double structure comprising the hypopygium consists of the large ninth and much reduced following somites. The morphologically dorsal part is the ninth tergum or epandrium, which bears the remains of the following somites, ending in the cerci and the anal BOMBYLIIDAE TERGUM 8 EPANDRIUM ~ CERCUS EPIPH BASISTYLUS US STERNUM8 DISTISTYL' STERNUM 6 A TERGUM STERNUM TERGUM 6 TERGUM STERNUM 6 STERNUM TERGUM TERGUM 7 STERNUM 8 DISTISTYLUS TERGUM 6 Soren ogee BASISTYLUS CERCUS TERGUM 7 7 EPANDRIUM CERCI BASISTYLUS Text-Ficure 12.—Genitalia of a male bee fly, Poecilanthrax californicus Cole. ExpLANATION: A, sternum of abdomen; B, parts “‘in situ’’; c, lateral aspect of male genitalia exserted. 53 54 BEE FLIES OF THE WORLD BASAL EJAC.APOD. CERCUS SE 3 of ‘a = BASAL PT.AED. EPIPHALLUS AEDEAGUS DISTISTYLUS \ BASISTYLUS GYOL77 MN) LAT. EJAC.APOD. BASAL EJAC.APOD. BAS.PT.AED. BASISTYLUS A EPIPHALLUS AEDEAGUS DISTISTYLUS Text-Ficure 13.—Male genitalia of Poecilanthrax californicus Cole: a, surface levels; B, second level, with the epandrium removed. orifice. The ventral part most in evidence consists of a pair of basistyli between which ventrobasally may some- times be seen a more or less rudimentary ninth sternum. In most bombyliids there is a movable structure at- tached at the apex of the basistylus, which is the telo- mere of Snodgrass, the beaked apical part of Hesse, and the dististylus of Maughan and Cole; the latter term is adopted here. It may be noted that the base of the basi- stylus may itself be produced into a lobe or process. In the Gerontinae this dististylus is more or less im- movable and lobelike. The dististylus takes on almost an endless variety of complex shapes and forms, short and wide, or oval, elongate, etc., and as Hesse points out, flattened, hollowed, or convex. Found within the pair of basistyli is the aedeagal apparatus or aedeagal complex of Hesse; it is attached to the basistylar shell by membranes and also by the ramus, the ramal element from each side generally is BOMBYLIIDAE A, BASAL EJAC.apop. BASAL EJAC.APOD. LAT. EJAC.APOD. EJAC.DUCT AEDEAGUS BASISTYLUS EPIPHA EPIPHALLUS \ oS LAT. EJAC.APOD BASAL PT.AED. EPIPHALLUS 3 AEDEAGUS BASAL EJAC.APOD. LAT.EJAC.APOD. ii BAS.PT.AED. EPIPHALLUS AEDEAGUS ©? D BASAL EJAC.APOD. Text-Figure 14.—Male genitalia of Poecilanthrax californicus Cole. Third level: a and B, dorsal aspects; c, the basistylus removed; pb, Fourth level, the axial system. 55 56 BEB FLIES OF THE WORLD fused or coalesced to form an outer rim from which proceeds the aedeagus. The aedeagus is very variable: short or long, slender, straight or curved, or even tubular. Bowden (1964) considers that the ventral aedeagal processes of Hesse are not part of the aedeagus but belong entirely to the last abdominal seg- ment and really form an aedeagal sheath which is con- nected by strong lateral processes, the dorsal rami of Hesse. He considers that these connect the whole to the base of the basistyle. According to Bowden, the aedeagus is easily dissectable from the sheath and con- sists of the aedeagus, and in addition a basal strut (Hesse, 1938) or apodeme, and lateral struts or para- physes. We consider the aedeagus to be overlaid and often enclosed in a dorsal, wider sheath which we call the epiphallus. Hesse observes that the aedeagus is the intromittent organ of the bee fly, serving as a guide for the penis itself and the seminal duct within. All of these structures are especially complex within the Systropodinae and Gerontinae, with additional spines, prongs, hooks, ete. For the study of the male genitalia, it was possible with the aid of the National Science Foundation grant to secure the services of Mr. Kenneth Weisman as a technician and delineator. In preparing the male genitalia for the purpose of illustrating, for the most part, the technique of Metcalf (1921) was adopted. The dried specimens were placed in a relaxing jar containing a solution of approximately 3 percent car- bolic acid for 12 to 16 hours, after which the genitalia were removed. They were then cleared from 24 to 36 hours in a 10 percent solution of potassium hydroxide. This step of the procedure necessitates frequent obser- vation in order to make sure the genitalia do not be- come overcleared. After clearing for the appropriate time the genitalia were removed to a neutralizing solu- tion of 5 percent acetic acid. While in the neutralizing solution adhering muscle and other tissues that still remained were removed, after which the structures were deposited in glycerine-filled microvials and sub- sequently fixed on the pin that held the dried specimen. Study and drawings were made while the genitalia were supported by glycerine in a spot dish. The primary illustration for each genus consists of the complete genitalia in lateral aspect. After this figure was completed the epandrium was removed and the dorsal aspect of the basistylus (penis sheath) and its accompanying structures were illustrated. The final illustration, that of the lateral aspect of the axial sys- tem, necessitated this structure to be dissected from the basistylus. This view, called the “axial system,” has been prepared for the more important genera. After com- pleticn of the illustrations, all the genital structures were again placed in the individual microvials and affixed to the pin carrying the dissected example of the genus. Chromosomes Chromosome studies have shed much light upon the interrelationships of the groups in the family Syrphi- dae. Similar studies in the Bombyliidae should be very fruitful, but according to personal communication from J. W. Boyes there are very few records in the literature, possibly not more than five or six. Metz (1916), in a very interesting paper entitled “Chromosome studies on the Diptera: II. Referring to Villa lateralis Say (as Anthrax),” states: “no more conspicuous cases of chromosome pairing have come to my attention than those exhibited by this and other species of the Bomby- lhidae.” For this species he shows 5 large pairs of chromosomes and 1 small pair (Figures 129-133) ; the diploid number is 12. He was not able to identify the sex chromosomes, but believed them to be the very small pair that was present. For Hemipenthes sinuosa Wiede- mann (as Anthrax), he found 9 pairs of chromosomes, all of different sizes, with the « and y chromosomes identified and the diploid number is 18 (Figures 134- 137). For Anthrax tigrinus De Geer (as Spogostylum simson Fabricius) he finds the diploid number is 12, with a less decided variation in size (Figures 141-142). Fossil Bombyliidae Forty species of fossil bee flies have been given names. These species are distributed within 30 genera, of which 20 are extinct genera. This compares with the Asilidae which at present have the same number of known fossil species but only 18 genera, 3 of which are extinct. There are 3 genera of asilids known from an Eocene formation, 2 of which are extinct genera; no Eocene bombylids are known. Four entomologists have con- cerned themselves extensively with fossil bee flies. Cockerell in a series of papers in the early part of the present century described numerous genera and species but with perhaps an inadequate background knowledge of the family. Many of his determinative decisions were based upon fragmentary portions of the bee fly wing, as illustrated in his work of 1914. Paramonov, with more exhaustive knowledge of the family, published a good summary of known species of bee fly fossils in 1939. Melander (1949) added several species which he described from material in the American Museum of Natural History. Hennig (1966) has provided a fine illustrated study of the bee flies known from the Baltic amber; I am able to add another from drawings I made in 1936, at which time I searched the collection at Konigsberg for Syrphid fossils. I present a list of the bee flies that are known from fossil forms with comment where I am able to make such. I have not been able to examine the type material of many species. BOMBYLIIDAE 57 ye G65 OD) 7 (ere Mir Xe Be /, a S \ AAS ANN \\ Hy) Text-Ficure 15.—Fossil bee flies: a, Proglabellula electrica Hennig; 8, Paracorsomyza crassirostris Loew. Redrawn after Hennig. 58 BEE FLIES OF THE WORLD Text-Ficure 16.—Fossil bee flies:a, Amictites regiomontana Hennig; 8, Glaesamictus hafniensis; c, Glaesamictus hafnie nsis Hennig, face. Redrawn after Hennig. BOMBYLIIDAE 59 FOSSIL BOMBYLIIDAE ACCORDING TO SUBFAMILIES BOMBYLIINAE Fossil Genera Praecythera sardii Theobald, 1937b, Oligocene (Aix-en-Prov- ence) Paracorsomyza crassirostris Loew, 1850, Meunier, 1910, 1915, 1916, Oligocene (Baltic Amber) Recent Genera Bombylius depereti Meunier, 1915, Oligocene (Aix-en-Provence) Bombylius sp. Berendt, 1830, Lower Oligocene (Amber of Sam- land, Poland) Bombylius sp. Schlotheim, 1820, Upper Miocene (Oeningen Shales, Baden) TOXOPHORIN AE Fossil Genera Alepidophora pealei Cockerell, 1909, Miocene (Florissant) Alepidophora minor Melander, 1949, Miocene (Florissant) Alepidophora cockerelli Melander, 1948, Miocene (Florissant) PHTHIRIINAE Fossil Genera Protophthiria palpalis Cockerell, 1914, Miocene (Florissant) Protophthiria atra Melander, 1949, Miocene (Florissant) Acreotrichites scopulicornis Cockerell, 1916, Miocene (Floris- sant) Recent Genera Phthiria oligocenica Timon-David, 1944, Lower Oligocene (Camoins) Apolysis magister Melander, 1946a, Miocene (Florissant) GERONTIN AE Fossil Genera Palaeogeron vetustus Meunier, 1914, Oligocene (Aix-en-Prov- ence) Geronites stigmalis Cockerell, 1914, Miocene (Florissant) Recent Genera Geron? platysome Cockerell, 1914, Miocene (Florissant) USIINAE Fossil Genera ; Lithocosmus coquilletti Cockerell, 1909, Miocene (Florissant) Recent Genera Usia atra Statz, 1940, Upper Oligocene (Upper Oligocene, from Rott) SYSTROPINAE Fossil Genera Melanderella glossalis Cockerell, 1909, Miocene (Florissant) Melanderella testea Melander, 1949, Miocene (Florissant) Pachysystropus rohweri Cockerell, 1909, Miocene (Florissant) Pachysystropus condemnatus Cockerell, 1910, Miocene (Floris- sant) Recent Genera Systropus rottensis Meunier, 1917, Oligocene (1’Aquitanien de Rott, Siebengebirge, Rhineland) Systropus acourti Cockerell, 1921, Upper Oligocene (Isle of Wight) : Dolichomyia tertiaria Cockerell, 1917, Miocene (Florissant) MYTHICOMYIINAE Fossil Genera Proglabellula electrica Hennig, 1966, Lower Oligocene (Baltic Amber) LOMATIINAE Fossil Genera Protolomatia antiqua Cockerell, 1914, Miocene (Florissant) Protolomatia recurrens Cockerell, 1916, Miocene (Florissant) Alumatia fusca Cockerell, 1914, Miocene (Florissant) Recent Genera Xeramoeba gracilis Giebel, 1862, Lower Oligocene (Copal) CYLLENIINAE Fossil Genera Megacosmus mirandus Cockerell, 1909, Miocene (Florissant) Megacosmus secundus Cockerell, 1911, Miocene (Florissant) Palaeoamictus spinosus Meunier, 1916, Upper Oligocene (Bal- tic Amber) Protepacmus setosus Cockerell, 1916, Miocene (Florissant) Verrallites cladurus Cockerell, 1913, Miocene (Florissant) Amictites regiomontana Hennig, 1966, Upper Oligocene (Baltic Amber) Glaesamictus hafniensis Hennig, 1966, Upper Oligocene (Bal- tic Amber) Recent Genera Amphicosmus delicatulus Melander, 1949, Miocene (Florissant) - ANTHRACINAE (Tribe ANTHRACINI) Fossil Genera Anthracida zylotona Germar, 1849, Upper Oligocene (Orsberg bei Rott im Siebengebirge, Rheinlande) EXOPROSOPINAE (Tribe VILLINI) Recent Genera Hemipenthes s. l. sp. Goldfuss, 1831, Upper Oligocene (Rhein- lande [Aquitain?] ) Hemipenthes s. l. sp. Burmeister, 1832, Lower Oligocene (Bal- tic Amber) Hemipenthes s. 1. sp. Keferstein, 1834, Upper Miocene (Oenin- gen, Baden) Hemipenthes s. l. provincialis Handlirsch, 1908, Lower Oligo- ecene (Aix-en-Provence) Hemipenthes s. l. tertiarius Handlirsch, 1908, Upper Miocene (Gabbro, Italy) Hemipenthes s. l. gabbroensis Handlirsch, 1908, Upper Miocene (Gabbro, Italy) Zoogeography of the Bombyliidae Bee flies, while worldwide in distribution, are far more abundant in arid lands. They are not absent from rainy temperate zones or tropical rain forest but in the latter they may be quite scarce. In several weeks of col- lecting in the Panama rain forest in August 1938 1 saw only one bombyliid. In a study of bombyliids in 1962 over a five-week period, in Mexico, ranging from arid eastern country to the upper edges of the rain forest near Tamazunchale, to the western desert part of the country, the number of bombyliids varied significantly according to the rainfall and area visited. Four or five areas were sampled each day by three collectors spend- ing about an hour at each locality. Generally it was found that an hour of intensive collecting by three per- sons was sufficient to obtain practically all the Bombyli- idae that were present. More time given to the same place seldom resulted in additional species. These areas and the number of bombyliids found at each are shown on the accompanying map. While the total bombyliid fauna must be significantly greater for each of these localities, if accumulated on a year around basis, it is nevertheless possible to gauge the relative richness of the fauna from this sampling.* Of course, there is an additional factor which is important too; the amount of rainfall at any one time has a distinct bearing on the * Hall found 50 species in one canyon in California on a year basis. 60 BEE FLIES OF THE WORLD bee fly fauna as indeed upon all of the insect populations of an area. When bee flies are mapped to the point of location from which the type was described, it is at once apparent that the annual rainfall has some peculiar relationship to bee fly populations. Desert rains often “trigger” the emergence of bee fly populations which appear seven days or a week or more later. The peculiarities in the distribution of bee flies throughout the world are most interesting. Since, as far as is known, all bee flies are parasites or hyperparasites, it is not to be expected that species density centers would be the same as they are for the syrphid flies and the asilid flies, in both of which families the Neotropical region has much the greater number followed by the Palaearctic region. Even in the Syrphidae the special conditions affect these densities; for example, the wet, New World rain forests, so extensive, lend themselves to the great development of the flies in the Syrphidae, which often are aphidophagous and coccidophagous. In the Bombyliidae the development of the several sub- families must be tied in very closely with the develop- ment and abundance of species and genera of the two great sources of hosts, the fossorial bees and wasps and the egg masses and packets of the locustids. A recent review of the Orthoptera of South Africa shows a total of about 2,000 species of locustids; it is not surprising then to find such an immense development of the genera Systoechus Loew and allies and of Lomatia Meigen. As of the end of 1965 the world count of bee flies stood at about 3,924 species, distributed through 194 genera and 27 subgenera. The total species and genera for the Homeophthalmae division, 1,884 mn 109 genera, is not greatly different from the 2,012 species in 85 genera within the Tomophthalmae; there are more species in the latter division but fewer genera. Only genera are considered in these totals, the subgenera are omitted. Now if we look at the separate world regions we find 625 species in the first division from the Palaéarctic, and 559 species from the Ethiopian area, against 494 Palaearctic species and 641 Ethiopian species in the second division. Note that these two world regions lead very considerably; almost 60 percent of the known bee flies are described from these two regions. We must not lose sight of the fact, however, that many arid lands have not been adequately explored for bee flies; such certainly are the southern areas of South America, still parts of the western United States, and still more parts of Australia and Central Asia. The bee fly fauna of Asia, apart from Asia Minor and the Transcaucasus, is depauperate indeed upon the basis of present-day rec- -ords, with only 169 species listed from there; even more species—263—are known from Australia. Only 9 species have been described from what is usually called Oceania, or the South Pacific Islands. There is a single unique genus from New Zealand. After an inspection of the accompanying tabulation, if we look more closely at some of the subfamilies, we will note that the subfamilies Bombyliinae, and the Lomatiinae, Anthracinae, and Exoprosopinae as well A / Fae pate ee / ~' / / / i 7 Z / Text-Ficure 17.—Concentration areas of the bee flies of the world. BOMBYLIIDAE as the small subfamily Cylleniinae are all dominant in what may be called the Afro-Eurasian landmass—the more easterly parts of Asia omitted. The home of the Heterotropinae is unmistakably in southern Europe and Asia Minor; this is largely true of the Cyrtosiinae. Other peculiarities of distribution will be noted; such is the great speciation of a single genus of the Mythico- mylinae, Mythicomyia Coquillett, in the western United States; the confinement of the moderately large genus Usia Latreille to the southern and western Palaearctic region. The smaller subfamilies tend to be absent or much reduced in one or more world regions, as the Cyrtosiinae in the Neotropical area, the Heterotropinae from both Australia and the Neotropical area. There are odd phenomena with respect to genera. Species of Anastoechus Osten Sacken are more numer- ous in the Northern Hemisphere, and species of Systoe- chus Loew are more abundant in the southern half of the world. Both appear to be locust-egg pod consumers. One of the outstanding features of the zoogeography of this family is the large number of endemic and restric- tive or peculiar genera of certain regions. For example, of the 50 genera endemic to South Africa, 20 are mono- typic. Only 16 genera are endemic in the Palaearctic region and 16 in the Neotropical region. There are 18 genera peculiar to the Nearctic region. Some of these, like Lordotus Loew from the first division of the family and Poecilanthrax Osten Sacken from the second divi- sion, are well developed and have prominent features. Not surprising are the 12 genera peculiar to Australia, including unique genera such as Meosardus Roberts, Myonema Roberts, and Comptosia Macquart. The virtual restriction of the relict genera of the Comptosini to Australia and the Chilean subregion will be noted with interest by everyone concerned with these flies. Synopsis and Distribution DIVISION HOMEOPHTHALMAE BOMBYLIINAE, Latreille, 1802 Eclimini, new: Eclimus Thevenemyia Tillyardomyia Paratoxophorini, new: Paratoxophora Dischistini, new : Dischistus Bombylosoma Chasmoneura Doliogethes Gonarthrus Legnotomyia Lordotus Geminaria Othniomyia Lepidochlanus Prorachthes Adelidea Sosiomyia Paramonovella, new Pilosia, new Sparnopolius Cacoplox Sericusia Platamomyia Bromoglycis Cryomyia Hallidia Neodischistus Conophorina Euprepina Conophorini, Becker, 1912 Conophorus Aldrichia Corsomyzini, new: Corsomyza Callynthrophora Gnumyia Hyperusia Megapalpus Zyemyia Mariobezziini, Becker, 1912 Mariobezzi Cythereini, Becker, 1912 Bombyliini: Ce NENEOT PA ET OR AU OC Total ‘ Pantarbes Acanthogeron 14 #4 15 Sericosoma Systoechus 5 2282) 120 Gyrocraspedum Lissomerus 1 1 Oniromyia Anastoechus 3 50 21 76 Bombylius 28 28 139 137 8 395 Heterostylini, new : Bombylodes 3 3 Heterostylum Tsocnemus 1 1 Eurycarenus Brychosoma i Triploechus Zinnomyia 2 2 Efflatounia Parabombylius 9 9 Karakumia Sisyromyia 3 Nectaropota 1 1 Crocidiini, new: Eusurbus 2 Crocidium Sisyrophanes 7 ai Mallophthiria Eucharimyia 1 1 Desmatomyia Staurostichus 1 Adelogenys Apatomyza Acrophthalmydini, new: Semiramis Acrophthalmyda 1 1 Tamerlania 61 NENEOT PA ET OR AU OC Total 28 19 2 mee oO ol ot oO 2 3 a Be ee Bae EE rFPwWOaAnNnKS 13 20 Re bh oO oe NORrFNORF NOE HE NWWOMDUNDH& b = St ll a Bee wre 62 PHTHIRIINAE, Becker, 1912 Acreotrichus Phthiria GERONTINAE, Hesse, 1988 Amictogeron Geron Pseudoamictus Us1in ak, Becker, 1912 Apolysis Oligodranes Dagestania Usia BEE FLIES OF THE WORLD NE NEOT PA ET OR AU OC Total ol 2 OO HETEROTROPIN AE, Becker, 1912 Apystomyia Caenotus Heterotropus Prorates Pe Re TOXOPHORINAE, Schiner, 1868 Toxophorini : Toxrophora Lepidophorini, new : Lepidophora Cyrtomyia Marmasoma Palintonus SYSTROPIN AE, Brauer, 1880 Dolichomyiini, new: Dolichomyia Systropodini: Systropus a bo 4 XENOPROSOPINAE, Hesse, 1956 Xenoprosopa 10 10 1S Lz, 33 31 PLATYPYGINAE (Cyrtosiinae, of authors), Verrall, 1909 Cyrtosiini, Becker, 1912 Cyrtosia 31 Ceratolaemus Onchopelma Psiloderoidini, new : Psiloderoides Cyrtomorpha Platypygini : Platypygus at 15 Cyrtisiopsis MYTHICOMYINAE, Melander, 1902 Mythicomyiini : Mythicomyia 130 al Aetheoptilus Doliopteryz Glabellula 6 6 Pseudoglabellula Empidideicini, new : Empidideicus 4 9 Anomaloptilus il Euanthobates Leylaiya 1 16 39 we OL mb he bore w bo 17 He A 2 126 DIVISION TOMOPHTHALMAE NE NEOT PA ET OR AU OC Total CYLLENIINAE, Becker, 1912 Cylleniini : Cyllenia 1 7 8 Sphenoidoptera 1 1 Amictus ea 26: 27 Sinaia 2 2 Henicini, new: Henica 1 1 Nomalonia 6 6 Tomomyzini, new: Tomomyza 11 11 Pantostomus 9 9 Amphicosmus 4 4 Metacosmus 3 3 Paracosmus 6 6 Peringueyimyini, new: Peringueyimyia 1 1 Neosardini, new Neosardus 4 4 LOMATIINAE, Schiner, 1868 Lomatiini : Lomatia 1 34 97 4 136 Anisotamia 1 Ogcodocera 1 1 Bryodemina 3 EHdmundiella 1 Hohe Myonematini, new: Myonema 1 Docidomyia 5 or Antoniini, new : Antonia Gs 76 4 = ee Aphoebantini, Becker, 1912 Aphoebantus Pteraulaxr Epacmus Prorostomatini, new : Prorostoma Coryprosopa Epacmoides Stomylomyia Plesiocera Conomyza Exepacmus Eucessia Xeramoebini, new : Xeramoeba Chionamoeba Petrorossia Desmatneura Chiasmella Comptosiini, new : Comptosia Lyophlaeba Ylasoia Doddosia Ulosoma Oncodosia 52 6 13 ro EXOPROSOPINAE, Becker, 1912 21 bt = CO OD pee o S = wo _ Bee Oe Ree BOMBYLIIDAE 63 NE NEOT PA ET OR AU OC Total Villini, new: Villa 50 152 80 53 34 35 29 483 Astrophanes ail 1 2 Chrysanthraxr 22 22 Cyananthrax 1 1 Deusopora 1 1 Dipalta 2 2 Diplocampta 3 3 Hemipenthes 24 1 16 41 Lepidanthraz 9 3 PNP 16 Mancia 1 1 Neodiplocampta 3 3 Paravilla 26 26 Poecilanthraz 33 33 Prothaplocnemis 1 1 Pseudopenthes 1 al Stonyx il 3 4 Synthesia 1 1 Thyridanthraxr 9 38 64 44 120 Rhynchanthraxr 3 3 Oestrimyiini, new : Oestranthrazr 1 5. 4 10 Oestrimyza 1 1 Marleyimyia 1 1 Villoestrus aby al 2 Exoprosopini : Atrichochira 1 1 Diatropomma 2 2 Exoprosopa 41 39 74 236 40 138 443 Heteralonia 2 2 Tsotamia 1 al Ligyra 2 13 11 #16 «17 «22 81 Litorhynchus pag) aL 30 Micomitra Colossoptera 1 1 ANTHRACINAE, Latreille, 1804 Anthracini: Anthrazr PAS) Psy aL at yee 6 278 Dicranoclista 2 1 3 Walkeromyiini, new : Walkeromyia 2 2 Coniomastix 1 1 DIVISION HOMEOPHTHALMAE 110 Genera BOMBYLIINAE 14 18 238 30 4 16 1 106 PHTHIRIIN AE 1 1 aed edo), 7 GERONTIN AE 1 1 a Ut ay a bel vega LN a 9 USIINAE 2 By ey al 10 HETEROTROPIN AE 4 2 dls ia 8 TOXOPHORIN AE 2 3 nD eile etd: 11 SYSTROPIN AE 2 2 a Least Kapaa Wee 9 XENOPROSOPIN AE 1 1 CYRTOSIIN AE 1 Bo (6 2 11 MYTHICOMYIINAE 3 1 4 8 1 17 DIVISION TOMOPHTHALMAE 81 Genera CYLLENIIN AE 5 2 3. 5 1 16 LOMATIIN AE 6 HA ie ale bye eis 48 EXOPROSOPIN AE 15 14 8 1 6 5 63 ANTHRACINAE 2 2 aera nc went ce k 10 Totals from each region 58 51 63 91 20 41 2 8326 DIVISION HOMEOPHTHALMAE NE NEOT PA ET OR AU OC Total No. of Species BOMBYLIINAE 1387 «66 381 421 19 75 8 1107 PHTHIRIINAE OA) PAL P43 ales; al 99 GERONTIIN AE 212 10) Oo 2G On USIINAE 61 62) dr 2 136 HETEROTROPIN AE 7 Pal AILS PD 37 TOXOPHORIN AE aly abe alpen vali 23 59 SYSTROPIN AE By.) By y Sal GD ale} 132 XENOPROSOPIN AE 1 1 CYRTOSIIN AE 1 46 16 3 66 MyYTHICOMYIINAE 140 ay ab nea ts) 1 174 DIVISION TOMOPHTHALMAE No. of Species CYLLENIINAE 14 2 35 28 4 83 LOMATIINAE 69 38 86148 2 65 408 EXOPROSOPIN AE 225 224 252 395 95 73 29 1293 ANTHRACINAE 381 27 121 70 28 7 284 Totals from each region 749 442 1106 1226169 250 38 3980 Phylogeny and Subdivision of the Family Bombyliidae One of the principal attractions of this ancient and fascinating family lies in the remarkable diversity of the many suprageneric basitypes which it contains. I define the word basitype as being a uniquely different genus within a higher category, such as a subfamily, which sets it at once apart; such genera have often been used as an excuse for the creation of new taxons within a family; sometimes such treatment is justified, more often not. An example would be the rather unique genus Antonia Loew; I have made this genus the basis of a new tribe, and a full dissection of its genitalia seems to warrant and support such treatment. Universally, previous students of this family have expressed dissatisfaction and uncertainty with regard to the relationships of several groups. The taxonomy of the family has been overburdened and today is top- heavy with subfamilies, of which no less than 23 have been proposed by various authors; certainly these are not all of equal value. In my classification of the family I have found it necessary to rectify this situation by reducing many of these to a tribal status, which I be- lieve expresses more accurately the basic relationships within the family. I have named no new subfamilies, although I have added new tribes and suggested other new tribes. Distinguished students have even suggested separate subfamily status for several more “groups,” such as Tomomyza Wiedemann, Pantostomus Bezzi, Plesiocera Macquart, and Antonia Loew. These groups do deserve recognition. Yet it is my firm conviction that all values are better preserved if they are placed within tribes of the most appropriate subfamily. This I have done. There are twelve subfamilies of scarcely disputable 64 BEE FLIES OF THE WORLD validity; eight are in the Homeophthalmae—Bombyli- inae, Phthiriinae, Toxophorinae, Systropinae, Ger- ontinae, Usiinae, Platypyginae, Mythicomyinae—and four are in the Tomophthalmae—Cylleniinae, Lomati- inae, Exoprosopinae, and Anthracinae. It is likely that the Heterotropinae should be left as a tribe within the Bombyliinae. While I believe that Xenoprosopinae should be placed as a tribe within the Bombyliinae (even as Oestranthrax Bezzi is placed in Exoprosopinae as an aberrant specialized offshoot), I have refrained from doing so until such time as Yenoprosopa Hesse is better known. The classification in this work recognizes 11 tribes within the Bombyliinae, 7 within the Lomati- inae, 5 within the Cylleniinae, and 3 within the Exo- prosopinae. The evolution of the Bombyliidae has proceeded along two main lines of development. On the one hand, the division or section (or series) Homeophthalmae has the posterior eye margin almost always entire, except for the small group of genera centered around Hetero- stylum Macquart, and has the occiput flat, or even slightly concave, not expanding posteriorly and not bilobate above; this is the older group; Cockerell (1914) shows it to comprise the most ancient known forms (in his study of the fossil bee flies) and Cockerell also con- sidered this division to be of Old World origin. More- over, in this first division the cervix of the thorax is short, the head less freely movable, its mobility still further reduced by a general occurrence of a long, slender proboscis, more hovering and poising genera, and perhaps a greater dependence upon nectar. This division in large measure tends toward seasonal differ- ence and dependence upon different kinds of flowers, but of course with exceptions. This division, the Homoeophthalmae, contains the Bombyliinae itself with 11 tribes, and also 9 additional, highly modified, subfamilies, which are specialized and frequently show a much reduced venation. Included within this division there are 110 genera of which the greater number are Palaearctic; only 15 are Holarctic; only 6 are world- wide. All subgenera are omitted, the division contains approximately 1,885 species. Known hosts seem to favor the solitary wasps and bees but also egg pods of locusts; three subfamilies have turned to families of Lepidoptera for hosts. Hosts are known for 11 genera. On the other hand we have those flies in which the occiput becomes more and-more extensively developed, thickened, and with a bilobed aspect dorsally, and hol- lowed parts with extensive deep central cup; also they have an increasingly conspicuous indentation in the pos- terior eye margin which may or may not have an accompanying, horizontal, linear, bisecting line. There is a tendency in this division for both the antenna and the proboscis to become shorter, and in fact rather con- spicuously shorter, with a fleshy, muscoidlike labellum or even for the complete reduction of the parts. The cervix becomes longer, the head more freely movable, a trend toward different kinds of flowers, especially the clustered, late season composites. This division, the Tomophthalmae, contains especially the Exoprosopinae, the Anthracinae, the Lomatiinae, and the Cylleniinae, together having a total of 17 tribes. But the characterization of the family is not quite this simple. There are still other characters which trend in certain directions but are quite variable. In a sense all of these subdivisions fail to be clear-cut and all of them right down to genera have to be characterized collectively with an ensemble of characters because of overlapping features. For this reason I have tried to select what seems to be the most important single fea- ture for each group to which I add additional distin- guishing characteristics. Each of the several divisions that have been proposed in the past, regardless of the rank assigned them, together with new ones here recog- nized, seem to me to constitute what might be called so many basitypes. In the Tomophthalmae we find 81 genera. The Palaearctic region contains 27 genera; there are 9 Holarctic and 4 worldwide genera. There is a total of 2,068 species from the world. Hosts are known for 16 genera. The host preference is known for comparatively few genera but many species seek out subterranean wasp grubs, another tribe seeks almost exclusively mud-nest- building wasps, still another seeks out nocturnal Lepi- doptera; others parasitize Glossina or Calliphora pupae or myrmeleonids. Some attack cossid larvae, beetle larvae, or egg pods of locusts; as at present known their larval host range is much more diverse than in the first division of the family. There are numerous examples of hyperparasitism. My studies convince me that of still greater im- portance in the segregation of second subdivisions of the family is the deeply hollowed and conspicuous cup- like bowl of the central occiput; for actually this deep hollowed cup is present in all four of the subfamilies in the second division, whereas, in many of the tribes or genera of the Cylleniinae there is scarcely any back- ward overgrowth or overdevelopment of the occiput, or forward recession of the eye, and no indentation or bisecting line such as seen in the Lomatiinae and Exo- prosopinae. The Cylleniinae falls into this second division because they have the hollowed, cuplike occi- put; they are the most primitive or generalized mem- bers of this division and while the extent of the occiput varies from little in Paracosmus Osten Sacken, to much in Pantostomus Bezzi, there is at most only a faint indi- cation of the bilobed indentation. They also show the characteristic modification of the antenna, which is a feature of this division. In order to clarify the mor- phology of the posterior occiput I have prepared a series of figures illustrating the conditions of the occiput in each tribe or major subdivision; I refer to Figures 745, 748, 750, 752, and 775. While Bezzi (1924) took note of the hollowed occi- put, he placed first emphasis upon the thickened occiput and the bisected and indented posterior eye margin. Phylogenetically I believe there is no question but that BOMBYLIIDAB 65 the hollowing of the occiput is the primary divisionary character. It began in the Cylleniinae, and hence I place these flies as the first subfamily within the Tomophthalmae (section [2]). With it and accompany- ing it there follows a profound anterior recession of the eye (as begun in Zomomyza Wiedemann and ending in Oestranthrax Bezzi, etc.), the consequence of which is the greatly thickened occiput. The end result is a short, occipital fringe around the rim of the cup, a long cervix, complete freedom of head movement, generally shortened proboscis, more and more muscoidlike label- PLEISTOCENE PLIOCENE Alomatia, Protolomatia-- - > MIOCENE Lithocosmus _--~_ OLIGOCENE OCCIPUT EXCAVATED | lum, change of flower preference, and perhaps a greater consumption of pollen. The more time and study I expend upon this family the more apparent it seems to me that many characters represent a situation of what must be thought of as duplication or as parallel development or convergence, with the reappearance of certain characteristics inde- pendently in different parts of the family. These I shall indicate below. Cellular content of wings: from a condition where there are commonly as many as 4, and sometimes even MYTHicoMV\NNt ~~ Alepidophora Mie ecar ees - Protophthiria, Acreotrichites Bombylius OCCIPUT GLOBULATE Text-Ficure 18.—A provisional phylogenetic chart of bee flies and the relationship of subfamilies and tribes. 66 BEE FLIES OF THE WORLD 5, submarginal cells (some of which result from super- numerary crossveins), and as many as 5 posterior cells in occasional instances, the trend has been toward a gradual reduction in the number of veins and cells, to such genera as H'mpidideicus Becker, which has a simple, unbranched third vein, open discal cell with loss of the posterior crossvein, the second vein fused with the first vein, the auxiliary vein vestigial and has no ambient vein. As far as present time and presently known bee flies, we find that L’mpidideicus Becker shows the greatest reduction and the greatest simplifi- cation. The most unstable field in the bee fly wing is first the radial field, and second the posterior field as centered around the discal cell and the third and fourth posterior cells. We find: a. The radial field with from 5 submarginal cells, as in Bombylius namaquensis Hesse, to one, as in Cyrtosia Perris. b. From a 3-branched radial sector in many genera (4 in Adelidea Macquart) to 1 branch in H'uantho- bates Hesse. c. From a free second vein, as in most genera, to one ending in the costa, as in Glabellula Bezzi. d. From a complete auxiliary vein, the usual state, to a vestigial one, as in Mythicomyia Coquillett. e. In the postremigial field, from 4 posterior cells, the usual state, and occasionally 5, to 3 posterior cells in no less than five subfamilies. It is notable that the most complex venation ap- pears to be found within Zigyra Newman in the tribe Exoprosopini and again in the Comptosini within the subfamily Lomatiinae. The Comptosini are today almost completely restricted to the Australian and Chilean regions. Another example of survival of a complex bee fly is the Australian Veosardus Roberts, with its 4 submarginal cells. In other parts of the world the flies within these subfamilies have rather simplified venation, and, except for the close relation of Ligyra Newman to Hxoprosopa Macquart, they seem to be rather distantly related to the other members of their respective groups. The Comptosini and Neo- sardini seem indeed to be relict genera in the sense that they tend to have an extreme southern distribution much as TZvrichophthalma Schiner in the family Nemestrinidae. Other peculiarities of venation are noted under subfamilies and genera. Among the bee flies the subfamily Bombyliinae with some 74 genera and subgenera and approximately 1,107 species is quite large, and we believe these to be the most generalized. The next largest subfamily is the Exoprosopinae, which has 33 genera (all subgenera omitted) and has approximately 1,293 species. Though much more specialized in some respects this subfamily has several very large and seemingly active and evolv- ing or differentiating groups. Such are Villa Lioy, used in the wide sense, with about 433 species. Haxoprosopa Macquart has about 443 species. The widespread distribution of such semiphylogeront groups as Zowophora Meigen and Systropus Wiede- mann seems to indicate an ancient origin. Since both of these groups parasitize highly successful types of host insects, they seem to be in no way threatened with extinction. The genus Lomatia Meigen appears to have arisen in the Palaearctic but has become remarkably successful in South Africa. Ideal conditions and isola- tion have apparently been responsible for the enormous number of species of the gnatlike M/ythicomyia Coquil- lett which have developed in southern California and to a lesser extent in nearby states. Gigantism appears within a few genera. Some species of H'xoprosopa Macquart attain a wing spread of 64 mm.; perhaps the bulkiest bee fly is Bryodemina valida Wiedemann, from Mexico. Great size reduction or dwarfing (nanism) has appeared in several genera and subfamilies but none more so than in the minute flies of the genus Mythicomyia Coquillett. KEY TO THE SUBFAMILIES OF THE BOMBYLIIDAE 1. Third longitudinal vein with two branches .......8 Third longitudinal vein with only one branch. .... . 2 2. Second longitudinal vein vestigial or absent, or if present, and complete, it ends in the first vein and not in the costa and results in a short, triangular marginal cell. MYTHICOMYINAE Second longitudinal vein normal and well formed and al- ways ending separately and independently in the costa, beyond the end of the first vein. Second basal cell gen- erally shorter than the first . . PLATYPYGINAE 3. Wing with 4 or rarely with 5 posterior cells. Includes species with posterior veins partly atrophied .... . 6 Wing with only 3 posterior cells .... . ; 4 4. Abdomen long and slender, many times loner than idee laterally compressed throughout its length, or more or less attenuate on the basal half or more. Metasternum oblique and often greatly and ee drawn out and extended posteriorly . . SYSTROPINAE Abdomen robust, oval or subspherical, or abdomen short eonate, attenuate apically but never extensively lengthened. Metasternum not drawn out extensively posteriorly ; never more than normal in development . 5 5. Third antennal segment long but gradually attenuate, with a short, still more narrowed microsegment or style. GERONTIN AE Third antennal segment blunt and obtuse at apex, with a minute, subapical, subdorsal spine . . USIINAE 6. Flies without a large, conspicuous cavity in the occiput. Occiput usually not extensive, or if thickened, the outer part slopes gradually away from the eye margin. Occiput not bilobate above, with deep, vertical, postocellar fissure. Middle of posterior eye margin almost always smooth, entire and continuous, without indentation; an indenta- tion is very rarely present in this group. . . Bidets) Flies with a large, central, deeply sunken cami in the posterior occiput of the head. Posterior occiput extensive and tumid posteriorly beyond the lateral dorsal eye margin. Occiput always bilobate above and with resultant BOMBYLIIDAE deep, vertical postocellar fissure. Middle of lateral pos- terior eye margin with or without, more often with, an indentation, which occasions even greater exposure of the occiput. Indentation, when present, generally conspicu- ous, more rarely the eye margin is recessed forward on thexentirehuppershalte- eee 2 BRC GENUS MERN HL 7. Lateral eye margin neither indented nor recessed forward above ... . . . . . CYLLENIINAE Lateral eye mareint with Giga indentation or at least the upper half recessed forward ..... Pine AEE 8. Second longitudinal vein arising before the middle crossvein and at an acute angle, rarely with a bluntly rounded ANGIE Sie Me . . . . . LOMATIINAE Second vein arising Gonosite fo middle crossvein or nearly so, and always at right angle . . . . . . EXOPROSOPINAE 9. Prothorax prominent, bearing remarkably long, stout, straight or curved macrochaetae on lateral, dorsal or both portions. The anterior margin of the mesonotum and post- margin of the scutellum bear long, slender or stout, macrochaetae or bristles. Femora and tibiae with ap- pressed pile and scales. The tibiae often have numerous, long, spiculate spines or bristles. Thorax gibbous and often strongly and conspicuously arched. Abdomen arched and decumbent, or straight, cylindroid and elongate. I have treated Lepidophora Westwood as a separate tribe under this subfamily, although my later studies convince me it should be regarded as a tribe under the Bombyliinae. TOXOPHORINAE Prothorax short and normal. Macrochaetae if present never exaggerated. Thorax and abdomen not exceptionally arched, gibbous, and decumbent ..........10 10. Small, compact flies of 2 to 5 millimeters length, in which the third antennal segment is blunt and obtuse at the apex, and bears a short, thornlike spine, concealed or ex- posed, lying in a recess which may be either lateral and subapical, or dorsal and subapical, or apical and situated between two short protuberances above and below the spine. Anterior margin of oral opening always knife- sharp, the inner walls vertical and the oral cavity deep. Proboscis long and slender and always longer than the head ; palpus likewise slender, usually long. Wings broad at the base and upon the axillary portion. Never more than 2 marginal cells present. Discal cell always much wider apically and first posterior cell widest at the wing margin. Males holoptic and head of male flattened across 67 the top of eyes. Head of female with the front flattened and bearing a transverse groove or fovea over the de- pressed middle portion. Tibiae and femora with fine pile only ; the hairs sometimes stiffened; bristles and spicules absent... - . . . . PHTHIRIINAE Flies not with these Characters feollectivels dive Woes epee a EL “Facial and buccal part of head remarkably transformed and aberrant, depressed or excavated; mouth parts very aberrant, represented by a slight, central, boss-like ele- vation from the lower part of which projects a short, central, blunt, downwardly directed, spline-like process (medial anterior part of buccal rim), bounded on each side below by an oval, inflated lobe; antenna somewhat close together, with hairs on all the segments, segment 1 produced below into a large, conspicuous, densely haired, bladder-like or lobe-like extension, and segment three stoutish, bluntly tapering; front remarkably broad; occi- put not deeply excavated behind, only slightly concave, very broadly and shallowly depressed groove-like behind and below ocellar tubercle, bounded on each side by a tumid, lobe-like prominence; ocellar tubercle remarkably broad, slightly elevated boss-like, its posterior ocelli wide apart and reniform and its anterior one much reduced; hind margin of eyes only feebly sinuous, not indented and not bisected ; third posterior cell markedly narrowed and converging apically ; legs more or less shortened, without spines on femora below; spicules and spurs on tibiae and tarsi feebly developed ; vestiture in form of fine hairs and scaling not densely developed, without stiff, bristly hairs or bristles on any part of body, and hairs on abdomen markedly short and poorly developed.” XENOPROSOPINAE Mouthparts functional and otherwise not as above. BoMBYLIINAE 12. Discal cell (discoidal cell) broadened at the end (before the ultimate narrowing) and much of the widening due to the shortening of the anterior crossvein. Discal cell broader than the second posterior cell. Eyes of male sepa- rated into two distinct size groups of facets, the small ones below, the large ones above, but without dividing line. Inner margin of eye often indented near antennal base. Relatively bare flies . . . . . . . HETEROTROPINAE Discal cell not broadened at the end and not broader than the second posterior cell. Anterior crossvein long. Eyes of male not divided into two sizes of facets. PHTHIRIINAE Division HOMEOPHTHALMAE Subfamily Bombyliinae Latreille, 1802 This is a large and important subfamily. The adults are primarily nectar feeders and the proboscis as a rule is long and slender, yet there are exceptions. The hosts as far as known are usually solitary bees, but Sys- toechus Loew destroys locust egg pods and two genera attack beetle larvae. For a great many genera, however, the hosts are unknown. The following characters will serve to define this sub- family. A discal cell is always present. There are never less than 2 submarginal cells, and 3 submarginal cells are frequently seen in the Corsomyzini, Conophorini and Cythereini, as well as in some other genera. The posterior cells are never less than 4. The posterior eye margin is very rarely indented; the only exception is the Heterostylini. The occiput is not prominent and bilobed, but not concave either. The antennae are al- Bombyliinae ways subadjacent, except in the Cythereini. I do not consider the interesting group of genera around Cytherea Fabricius comparable to the other subfamilies within the Bombyliidae, and because I regard the family classification as top-heavy with subfamilies, I reduce these to tribal status; I have done the same with the #climus Loew group of genera. The antennae are generally long and slender. The wings tend to be either hyaline or at most smoky and basally tinged with brown, but there are exceptions. This is in strong con- trast to the frequency of distinct and characteristic pat- terns in the Exoprosopinae and Lomatiinae of the second great division of the family. I have treated Lepidophora Westwood as a separate tribe under Toxophorinae, although my later studies convince me it should be regarded as a tribe under the Bombyliinae. a of a aren Text-Ficurr 19.—Pattern of the approximate world distribution of the species of the subfamily Bombyliinae. 68 BOMBYLIINAE 69 Tribe Bombyliini In addition to the genus Bombylius Linné, on which the family is founded, those other genera in which the first posterior cell is closed and stalked fall here; such are Systoechus Loew, Anastoechus Loew, Sisyrophanes Karsch, Acanthogeron Bezzi, etc. Tribe Dischistini These flies have the first posterior cell open and usually widely open. They may be regarded as the most generalized of the Bombyliidae and contain much the greater number of genera within the subfamily. While it is true that this cell at the wing margin narrows in some genera and this division is probably an unnatural one; it is nevertheless highly useful. Probably there is a slow trend in certain genera in the direction of the closure of this cell. Tribe Heterostylini Here I place the 4 genera that have a distinct inden- tation near the middle of the posterior eye margin. It includes, besides the Nearctic Heterostylwm Macquart, the genus 7’riploechus Edwards, H'urycarenus Loew, an Ethiopian genus, and H7latownta Bezzi and Karakumia Paramonov. Again in this indentation of the posterior eye margin, we may possibly see a development parallel to that so commonly found in the Tomophthalmae. Tribe Cythereini Because of the great variability in the width of the face and still greater variability in the degree of sepa- ration of the antennae, I can see little justification for placing Cytherea Fabricius and Callostoma Macquart in a separate subfamily. It would be just as reasonable to place the odd Corsomyza Wiedemann and _ the Eclimus Loew group in separate subfamilies. Cytherea Fabricius and allies are separable on only three par- ticulars; in Cytherea the division or separation of the antennae reaches its greatest extent. It is somewhat less in Callostoma Macquart. I note much variation even within the species of a genus; Pantarbes willistoni Osten Sacken has the antennae but little separated, but the distance is twice as great in certain other species. I have followed Becker (1913) in placing the rather striking Gyrocraspedum Becker in this tribal group. I have done so in spite of the nearness of the antennae, because the base of the second longitudinal vein arises from the third longitudinal unusually close to the small crossvein and also abruptly and is in consequence sug- gestive of the second great division of the Bombylidae, the so-called Tomophthalmae. This character is also variable; it is not present for instance in Pantarbes Osten Sacken where this vein arises normally even though the face is very wide. Sericosoma Macquart has the antennae very widely separated but lke Pantarbes Osten Sacken the second longitudinal vein arises nor- mally. I cannot escape the conclusion that we are here likely dealing with a case of parallel development, or convergence of characters, in this tendency of the point of origin in the Cythereini to edge outward distally as is present in so many Tomophthalmae genera. The genus Oniromyia Bezzi from South Africa, an odd, aberrant bee fly, belongs rather doubtfully to this tribe. It has the antennae quite close together, the head not especially wide, and the second vein arises quite nor- mally. Perhaps the only character that suggests a rela- tionship to the Cythereini lies in the widely bossed or inflated frontovertical area. Its other characteristics are considered below. I reduce Glossista Rondani and Chalcochiton Loew to the status of subgenera where they still serve a useful purpose. Thus, there are 8 genera and subgenera within the tribe as represented here. Tribe Conophorini Flies with an exceptionally tumid, swollen occiput, in which the posterior expansion and swelling is quite unlike the Tomophthalmae division; in Conophorus Meigen the central part of the occiput is most promi- nent, rounding gradually off toward the eye margin, increasing or contributing toward a subglobular effect ; coincidentally the first two segments of the antennae tend to be swollen, often much swollen, reminding one of the condition often seen in therevids or in Symphoro- myia Frauenfeld; finally the venation, especially in Conophorina Hesse, is remarkably generalized with all cells widely open; in others there are 3 submarginal cells with a tendency to minor expansion of the apical remigial portion of the wing. I place in this tribe Conophorus Meigen, Conophorina Hesse, Aldrichia Coquillett, and the more specialized Prorachthes Loew and Legnotomyia Bezzi, and follow- ing Becker I include Codionus Rondani, with sub- generic status only; Codionus has only 2 submarginal cells, and the apical venation is rather different in all those species that have only 2 submarginal cells, as these cells are broadly rounded instead of being blocked out rectangularly. Several European species would fall in Codionus, as I have shown in the discussion of Conophorus,; also Calopelta fallax Greene. 70 BEE FLIES OF THE WORLD Tribes Corsomyzini and Mariobezziini These are rather unique flies, but I cannot see any justification for placing them in a separate subfamily when they are much better left in a tribe where they are obviously related. Probably these two tribes must be united. I have not had males of M/ariobezzia Becker to dissect, though Dr. A. J. Hesse has generously fur- nished abundant material of Corsomyza Wiedemann. The facial, antennal, and venational characters are unique. Tribe Acrophthalmydini Because this Chilean genus has such a_ peculiar pygidium with its peculiar sleevelike collar, very dif- ferent from the condition which obtains in Bombylius Linné where the genitalia are small, recessed, and turned downward and sideways beneath the apex of the abdomen, I believe these flies deserve to be placed in a separate tribe. Tribe Paratoxophorini Here I place the peculiar single genus Paratoxophora Engel; it bears no relationship whatever to Toxvophora Meigen, and the name is an unhappy one. It may later prove to be related to Amictus Wiedemann since its genitalia are similar in one respect. There is only one species of this genus known and while there are several species of Amictus Wiedemann, I have so far been able to dissect the male of only one species. At present I do not believe these genera are related. Tribe Eclimini Here I place 2 unusually interesting genera. H'climus Loew, which as Jack Hall (1969) has shown is re- stricted to southern Europe, and the more numerous species of Thevenemyia Bigot, the center for which lies in the southwestern United States. There is no justifi- cation for placing these flies in a separate subfamily. If every unique bee fly is placed in a separate subfamily there would be no less than 40 subfamilies; the true relationships of the bee flies and the phylogeny of the family is better served by placing these numerous peculiar groups in tribes within one of the giant sub- families. Tribe Crocidiini I place here a group of bee flies that clearly fall within the definition of the Bombyliinae but which are notable for several characteristics which preclude them being placed in other tribes. Crocidiwm Loew is found in Egypt and more abundantly in South Africa. Mallophthiria Edwards is from southern South America. Desmatomyia Williston with its peculiar an- tennae is known only from two species from the south- western United States. Apatomyza Wiedemann is a unique genus from South Africa and so is Adelogenys Hesse. The remaining genera are from the Trans- caucasus region. All the flies in this tribe have 4 pos- terior cells and only 2 submarginal cells. They can be recognized by the straight condition of the anterior branch of the third vein which reaches the margin of the wing at or very close indeed to the apex of the wing. The antennae show some peculiarities in some genera. The male terminalia in the Bombyliinae are vari- able in some respects but comparatively uniform in a number of features. The dististylus is almost always stout basally, narrowing apically into a hornlike point of variable sharpness, usually convergent at the apex. It may be also directed outward. Only in rare instances such as Heterostylum Macquart is the dististylus sup- pled with an additional or accessory lobe or prong. Nevertheless, it would be a very rash move to segregate Heterostylum Macquart and its related forms such as Triploechus Edwards and EHurycarenus Loew into a separate subfamily. Rather I have placed this group within a tribe within the Bombyliinae. Several of these have the aedeagal complex, which includes the epiphal- lus and the ejaculatory process, divided by a sinuslike space, resulting in a clasper dorsally. In general the entire structure is much shorter and more compact than the majority of genitalia such as in the Lomatiinae or the Anthracinae. The aedeagal complex is shorter and less spoutlike. KEY TO GENERA OF BOMBYLIINAE 1. Middle of lateral posterior eye margin distinctly indented, but central occiput normal and without enlarged cavity. 2 Lateral posterior eye margin continuous and entire, without indentation. Central occiput likewise normal, with at most a small central depression |... :... 2%. .6 2. Third antennal segment with a pencil of hairs at the apex. Karakumia Paramonov Apex of third antennal segment, bare, without a cluster of VTS 5) Fe yry 5 ovis ddl a pupa ibiewlewinad soa iohepn tern a SO ket ee 3. First posterior cell open or closed in the margin. Pulvilli quite minute and vestigial. Male claw toothed. Alula 6. 7. 10. BOMBYLIINAE almost or quite absent . 4 . Efflatounia Bezzi First posterior cell closed and petiolate; alula_ well developed .. . Be ae ee 4 Three submarginal ails, End of mareinel ceil recurrent but plane, leaving the adjacent submarginal cell very broad against the costa, vertical in position and greatly nar- rowed below. Anterior crossvein at middle of discal cell. Abdomen with scanty appressed pile upon the tergites, the dorsum with very scanty long, fine, erect, tiny hairs, but the side margin with rather dense tufts of similar pile . , . Triploechus Edwards Two suimmenetinll colic aul, End of marginal cell gently CUIVed ceca em 4 6 Go Se aan peta) End of first posterior call peteninees ana Ronminate! the stalk comparatively short. Upper anterior intercalary vein present though short. Anterior crossvein lies before the middle of the discal cell. Sides of mesonotum and pos- terior margins of abdominal tergites with strong bristles. Eyes in male usually holoptic or approximate, more rarely dichoptic. Male terminalia at apex with strong toothlike comb. Genae wide below, opposite the proboscis, and continuously densely pilose in contrast to the related genera . : sede . Eurycarenus Loew End of first moeteTiOn an not poaminntes the anterior branch of the fourth vein ends close to the fork of the third vein. Upper anterior intercalary vein usually quite absent, the medial crossvein and fourth vein cross at this point. Anterior crossvein lies at or beyond the middle of the discal cell. Bristles on mesonotum weak, wanting on the abdominal tergites. Males holoptic. The bare oragenal cup margin unusually wide, obliterating the gena in some species . Heterostylum Macquart Antennae widely Separated at the base. Second vein usually arising more or less abruptly close to the anterior cross- vein (Cythereini) .. . PANES eh ie xe} Antennae closely adjacent, serch if. ever aeepamated by more than the thickness of one segment ......... 138 Posterior branch of the third vein drops downward to meet the fourth vein; it closes the first posterior cell and be- cause of its shortness it simulates a crossvein. Second vein arises abruptly, before and quite near the anterior erossvein. First vein turned upward beside the auxiliary vein, both ending closely together. Proboscis 2 or more times as long as the head. Face bluntly conical, the face and front with short, fine, loose pile. Large flies with elongate wings and short, stout, robust abdomen and reddish-brown wings . . Gyrocraspedum Becker First posterior cell open or closed; if closed it ends rather acutely. Face and front often widely covered with dense, appressed scales which, however, may be restricted to the eye margins. Groups with dense facial and frontal scales have scanty pile; if the pile is dense the scales are more restricted ..... pig awics tonics: Second vein arises at or perore fe bags ae fhe discal cell and more or less acutely. No bristles anywhere upon the abdominal tergites; sides of first three tergies with a few stiff hairs. Pulvilli well developed ........9 Second vein arises well beyond the base of the discal cell and more or less abruptly. Face with scanty pile but widely and densely covered with appressed scales. Pos- terior tergomargins of the abdomen beset with numerous long, stiff, more or less erect, conspicuous bristles. Pul- villi usually rudimentary ..... Ooi sah arcana ak First posterior cell closed with a stalk O10 ono id . 10 First posterior cell open maximally. Antennae eridely: sepa- rated. First antennal segment not swollen ventromedially. Metapleuron in front of haltere bare; metapleuron in front of spiracle with a tuft of long pile. Sericosoma Macquart Stalk of first posterior cell quite short. Wing with 2 sub- marginal cells; slender and long, much narrowed and 11. 12. 138. 14. 15. 16. 71 reduced in size. Antennae separated at base by a little less than twice the thickness of one segment. First an- tennal segment swollen and expanded ventromedially somewhat as in the Lomatiinae. Antennae situated low upon the head, the extensive front covered by loose, very long, erect hairs and scattered appressed scales between the pile. Abdomen moderately broad at base, elongate, and tapered toward apex. Surface covered with small, slender, appressed scales; no bristles present along the posterior margins of abdominal terga . Oniromyia Bezzi Stalk of first posterior cell long. Wing with 3 submarginal cells. Wings of normal width. Antennae widely sepa- rated. First antennal segment simple, the third segment long, slender on the basal half, widened beyond. Face with dense, long pile and dense scales restricted to the eye margins. Abdomen with dense, long, erect, fine pile only ; similar pile on the thorax. Metapleuron in front of haltere with dense tufts of pile; metapleuron in front of haltere bare, overshadowed by the multiple fringes of the squama . . Pantarbes Osten Sacken First posterior cell closed, usually with a rather long stalk, or rarely closed in the margin. Rather large species with elongate, regularly tapered and banded abdomen with tergomarginal bristles posteriorly. Wings often whitish in tint, generally with a broad band of pale brown across the middle of the wing. Third antennal segment nar- rowed near base then tapered to the apex. Callostoma Macquart First posterior cell open. Somewhat more robust flies, the abdomen tapered but shorter, not or inconspicuously banded. Wings with basal half or two-thirds tinged or marked or clouded with brown. Tergomarginal bristles THRESEINE GUI) Broce gae bab oo) 6) a eilo- old. ope 4 Face extended forward only bluntly. Whole face, front, and genae densely covered with pale, or often silvery scales, wings with basal half or two-thirds strongly tinged with brown. Male eyes very widely separated. Proboscis long and slender, usually quite long, always extended beyond the oral cup. Usually with 3 submarginal cells (Cytherea, sensu stricto); if only 2 submarginal cells (subgenus Glossista Rondani) . Cytherea Fabricius Small flies. Pulvilli absent. Face and front with sparse, fine, erect hair in place of appressed scales. Oragenal cup and face extended conically forward, separated by a sharp, distinct crease. Wing clear hyaline, faintly tinged basally with brown. Male eyes less widely separated; 2 submarginal cells . . Cytherea (Chalcochiton) Loew First posterior cell open, even if narrowly open... . 14 First posterior cell always closed and usually with a long stalk, rarely closed in the margin. ....... . 63 First antennal segment cylindrical and extraordinarily stout and wide, 2 or 3 times as wide as the second seg- ment, and 4 or 5 times as wide as the third segment. First antennal segment as long or nearly as long as the outer two combined. Occiput very strongly developed from both dorsal and lateral aspect; marginal cell strongly widened apically, the ending rounded or plane, or even slightly recurrent. Metapleuron pilose, or base; 3 submarginal cells .... Staemot oul A) First antennal segment not orion sively Wiened and at the same time elongate ................ 16 Three submarginal cells ............... 79 Two submarginal cells . Conophorus (Codionus) Rondani Second antennal segment as long as the first or longer, and both cylindrical, stout, much thickened, and bristly. Marginal cell not usually widened apically. Third seg- ment flat but much widened in the middle, the apex with some long, stiff hairs. Three submarginal cells. Aldrichia Coquillett 12 17. 18. 19. 20. bo bo BEE FLIES OF THE WORLD Second segment of antenna shorter than the first, usually much shorter; neither segment extensively thickened. 17 Both marginal and first submarginal cells greatly and exceptionally widened in the outer half. Second sub- marginal cell as wide as long or wider. Three sub- marginal cells present ..... ak} Marginal and submarginal cells araually not Sidaned® only the marginal cell widened or the first submarginal cell, or neither, but not both. Second submarginal cell long or short, but longer than wide; usually with 2 submarginal cells; rarely with 3 submarginal cells ....... . 20 Head as wide as high and the face wide; third antennal segment without a long style... . Seeties LY) Head higher than wide, the face narrow; ihird antennal segment with a long style . hOiinionuyia Hesse Anterior crossvein enters discal cell at the outer fifth. Eyes with or without a complete dividing line from middle of posterior margin to middle of anterior margin. Ab- domen long, conical and tapered, the pile mostly quite short; face with scanty pile and a few bristles. Small AiCSisca Abe .. . . . » Geminaria Coquillett Anterior crossvein enters discal cell near the outer fourth. Large flies with dense, fine, conspicuous, long, erect pile, plushlike on thorax and abdomen and also on the face. Males holoptic; females with some appressed pile. Lordotus Loew Face very broad, with dense, erect, but rather short pile in a hemicircular band around the shallow oragenal cup. Head, thorax, and abdomen clothed primarily with dense, usually dark, flat-appressed, glittering scales. All anten- nal segments relatively short, the second especially so; third segment flattened laterally, greatly expanded dorsoventrally and bearing a tuft of spines dorsally. Wings dark colored and characteristically bear a mottled, marbled, or irrorate pattern . . Prorachthes Loew INOteSUCH PLCS aac -pasyeapiomesan bem iakclais SM iereg ter Kno: First basal cell distinctly longer than the ona basal cell, the anterior crossvein situated close to the middle of the discal cell or more Soe much beyond the middle of the discal-cell |. 3. 0. : Cee oy eT Anterior crossvein situated near ithe ace of ine discal cell and at least before the middle; consequently the first basal cell is approximately of the same length as the second basal cell, or at most very slightly longer . . . 60 Second antennal segment more than half as long as the first segment, and 2 or 3 times as long as wide and swollen through the middle; first segment also a little swollen and both bear numerous, long, coarse, bristly hairs, especially below. Rather small flies with semidecumbent abdomen with scattered, very long, erect, bristly hairs along the postmargins of the tergites. Conophorina Hesse INOPESUCHEATESH in ce csi Rede scite lacs eter aun Sar MP eC Euan an fea ie3| Head very broad and sometimes remarkably broadened ; as broad or broader than the thorax. Eyes in male always separated at least as much as width of ocellar tubercle and often very much more widely separated. Both front and facial region broad. Face not only broad but ex- tensive downward, genae likewise very wide; face may be much inflated and tumid, sometimes with character- istic dense brush of long, erect hair; third antennal seg- ment elongate but thickened or thinned basally only to become clavate apically and apex sometimes excavated, and microsegments reduced or absent. Apex of last male sternite notched medially; mesonotum low and _ not humped; thorax and especially abdomen with dense, long, fine erect, conspicuous pile. Wings with 2 or 3 sub- marginal cells. Dististyli generally compressed, claw- shaped or hook-shaped. Wing short, only twice as long as ot wide, wide basally, usually widest at base. Tribes Mario- bezziinisand! Corsomyziniuesrs,