D. L. DEONU A Systematic and Ecological Study of Nearctic Hydrellia Diptera: Ephydridae) SMITHSONIAN CONTRIBUTIONS TO ZOOLOGY NUMBER 68 SERIAL PUBLICATIONS OF THE SMITHSONIAN INSTITUTION The emphasis upon publications as a means of diffusing knowledge was expressed by the first Secretary of the Smithsonian Institution. In his formal plan for the Insti- tution, Joseph Henry articulated a program that included the following statement: "It is proposed to publish a series of reports, giving an account of the new discoveries in science, and of the changes made from year to year in all branches of knowledge not strictly professional." This keynote of basic research has been adhered to over the years in the issuance of thousands of titles in serial publications under the Smithsonian imprint, commencing with Smithsonian Contributions to Knowledge in 1848 and continuing with the following active series: Smithsonian Annals of Flight Smithsonian Contributions to Anthropology Smithsonian Contributions to Astrophysics Smithsonian Contributions to Botany Smithsonian Contributions to the Earth Sciences Smithsonian Contributions to Paleobiology Smithsonian Contributions to ^oology Smithsonian Studies in History and Technology In these series, the Institution publishes original articles and monographs dealing with the research and collections of its several museums and offices and of professional colleagues at other institutions of learning. These papers report newly acquired facts, synoptic interpretations of data, or original theory in specialized fields. Each publica- tion is distributed by mailing lists to libraries, laboratories, institutes, and interested specialists throughout the world. Individual copies may be obtained from the Smith- sonian Institution Press as long as stocks are available. S. DILLON RIPLEY Secretary Smithsonian Institution SMITHSONIAN CONTRIBUTIONS TO ZOOLOGY NUMBER 68 A Systematic and Ecological Study of Nearctic Hydrellia (Diptera: Ephydridae) SMITHSONIAN INSTITUTION PRESS CITY OF WASHINGTON 1971 ABSTRACT Deonier, D. L. A Systematic and Ecological Study of Nearctic Hydrellia (Diptera: Ephydridae). Smithsonian Contributions to Zoology, 68:1-147. 1971.?The adults of 57 species of Hydrellia are described for the Nearctic Region. Adults of Hydrellia are semiaquatic and the larvae are leafminers in aquatic and semiaquatic plants. Twenty-one of the described species are new, and one represents a new geographic distribution record. Some or all of the immature instars of 18 species are described. The male terminalia and certain other critical characters are figured for the adults of 52 species. The larval feeding apparatus of 16 species and the puparia of 17 species are illustrated. Geographic distribution data are given for all species. New and previously recorded host plants are listed for several species, and other biolog- ical information such as habitats and behavior is given for most of the species. The general morphology of the genus, including morphology of the adult feeding appara- tus and gut and internal genitalia, is discussed and illustrated. The general ecology of Hydrellia is discussed in regard to ecological role and distribution, parasitological data, dispersal and zoogeography, behavior, and environmental tolerance. Addi- tional ecological data are included in checklists of known host plants and known hymenopterous parasites of Hydrellia. Literature pertinent to the genus is reviewed. Official publication date is handstamped in a limited number of initial copies and is recorded in the Institution's annual report, Smithsonian Year. UNITED STATES GOVERNMENT PRINTING OFFICE WASHINGTON : 1971 For sale by the Superintendent of Documents, U.S. Government Printing Office Washington. D.C. 20402 - Price $1.75 (paper cover) Contents Page Introduction 1 Review of Literature 1 Acknowledgments and Repositories 3 Methods and Materials 4 Morphology 6 Glossary 6 External Morphology of Adult 8 Internal Morphology of Adult 13 Morphology of Immatures 15 Ecology 18 Ecological Role and Distribution 18 Parasitological Data 19 Dispersal and Zoogeography 20 Behavior 20 Environmental Tolerance 23 Host Plants and Parasitic Wasps 24 Systematic Treatment 24 Classification 24 Key to Adult Hydrellia 26 Male Terminalia Key to Hydrellia 29 Key to Third-Instar Larvae of some Hydrellia 33 Key to Puparia of some Hydrellia 34 Species Descriptions 36 Checklist of Known Host Plants 106 Checklist of Known Parasitic Wasps 113 Literature Cited 118 Localities of Specimens Illustrated 125 Abbreviations Used on Illustrations 126 Illustrations 128 D. L. Deonier A Systematic and Ecological Study of Nearctic Hydrellia (Diptera: Ephydridae) Introduction The Ephydridae constitute a medium-size family of acalyptrate Diptera. They lead several different modes of life, mostly in aquatic habitats. Hydrellia and Lemnaphila are the only known genera of aquatic leaf-mining ephydrids. At least two species of Hydrellia?H. griseola and H. ischiaca?are economically important. In 1953, H. griseola destroyed from 10 to 20 percent of Cali- fornia's rice crop, with an estimated loss of $16 million. The same species caused heavy losses of rice in California in 1922. It has damaged rice in Japan in the last two decades. Ulljeborg (1861) first re- corded the pest status of H. griseola. He reported the fairly widespread damage of barley, oats, and timothy grass by that species in southern and south- eastern Sweden during the summer of 1860. After this, several authors reported outbreaks of H. griseola in various places in the Palaearctic Region, including Egypt. Balachowsky and Mesnil (1935) reported 50 percent infestation of barley by H. griseola in northern Europe. Hydrellia ischiaca attacks wild rice, which is now a minor crop in Minnesota. Most of the known larvae of other Hydrellia species feed mainly in plants of Potamogetonaceae, Alismataceae, and Hydrocharitaceae. There are 185 available specific names in Hydrellia other than my new species. Of these, perhaps 133 have existing holotypes, and perhaps 132 of the 185 have Palaearctic type-localities. Possibly 113 of the D. L. Deonier, Department of Zoology and Physiology, Miami University, Oxford, Ohio 45056. available names are valid. These 113 species are dis- tributed as follows: 55 Palaearctic, 35 Nearctic, 1 Holarctic, 2 Oriental, 12 Australian, 2 Ethiopian, and 6 Neotropical. This essentially cosmopolitan generic distribution has caused some interest in the dispersal center. Some data possibly indicate this was in the northern temperate zone. I started this research with the following objec- tives: (1) to describe, redescribe, and construct keys to the adults and as many immature stages as pos- sible of Nearctic species; (2) to describe the life cycles and behavior of as many as possible of the Nearctic species; (3) to present a brief morphology of the genus. Review of Literature The taxonomic literature on Hydrellia dates from 1813, when Fallen described Notiphila griseola and several congeneric species from southern Sweden. In 1830, Robineau-Desvoidy erected the new genus Hydrellia. Macquart (1835), Zetterstedt (1846), Walker (1856), Loew (1860), Schiner (1864), Brischke (1883), Gobert (1887), Kowarz (1894), and Becker (18%, 1903) made many of the initial contributions to the taxonomy of Palaearctic species. Strobl (1904), Griinberg (1910), Becker (1919), Collin (1928), Frey (1933), de Meijere (1939), Goetghebuer (1942), Grensted (1944), Kloet and Hincks (1945), and Tsacas (1959, 1960) added some descriptions of new species, but primarily they pre- sented reviews and new distribution records. Dahl (1967) presented a zoogeographical synopsis of 1 SMITHSONIAN CONTRIBUTIONS TO ZOOLOGY eleven European species and (1968) described three new species along with new records from eastern Siberia. Becker's (1896, 1926) synoptic keys to and descriptions of Palaearctic adult Hydrellia and Hen- nig's (1943) specific key to larval Hydrellia including Palaearctic species contributed most to the taxonomy of Palaearctic species of Hydrellia. The recent re- vision by Dahl (1964) of the Stenhammar and Zetterstedt collections eliminated much confusion and synonymy in Palaearctic Hydrellia. Loew (1861, 1862) described the first new species of Nearctic Hydrellia. Loew (1872), Osten Sacken (1878), Becker (1896), Aldrich (1905), Jones (1906), and Coquillett (1910a) presented short sec- tions on Nearctic Hydrellia. Very little additional taxonomic study of the genus was made until 1915, when Cresson started describing new species. Johann- sen (1935) has reviewed and presented a key to the known immature instars of Nearctic species. Hennig (1943) has summarized the literature on immature Hydrellia and presented a specific key to all known immature instars, including the few in the Nearctic. Cresson's studies culminated in a key (Cresson, 1944b) to the adults of the 35 Nearctic species then known. Subsequent to this, Berg (1949, 1950) con- tributed to the biology and taxonomy, including a key, of the immature instars of six species; Hennig (1952) presented morphological interpretations of immature Hydrellia; Wirth and Stone (1956) in- cluded a key to California species and a Nearctic generic key; Grigarick (1959) redescribed the life- cycle stages of H. griseola; Deonier (1964) presented keys to the adults of species of Iowa and adjacent states; and Wirth (1965) cataloged many Nearctic species. Cresson (1918, 1947a) described and presented keys to the six species of Hydrellia known from the Neotropical Region. Wirth (1968) cataloged these, and included data on their distribution. Wirth (1969) reported H. vulgaris Cresson from the Gala- pagos Islands. Little study has been made of Hydrellia in the Oriental Region. Cresson (1948) listed only two species, H. latipalpis Cresson and H. luteipes Cresson, from the Oriental Region in his Indo-Australian synopsis. The latter species is known only from Formosa, which is more or less transitional between the Oriental and Palaearctic Regions. Of Hydrellia in the Australian Region, Coquillett (1903), Tonnoir and Malloch (1926), Cresson (1948), and Harrison (1959) published on some or all of the twelve known indigenous species. The relatively small amount of attention given to Orien- tal and Australian Hydrellia is matched only for the Ethiopian Hydrellia. Cresson (1932, 1947b) listed two species, both from the Cape of Good Hope. The works of the following authors provide some framework for a more searching morphological study of adult Hydrellia. Becker (1896, 1926), Griinberg (1910), Cresson (1918), and Wilke (1924) dealt mainly with chaetotaxy, but Cresson did briefly dis- cuss facial contour and the supposed absence of vibrissae in Ephydridae. Frey (1921) discussed the mouthparts of H. obscura and the cibaria of other ephydrids. Seguy (1934) illustrated the head and wing of H. griseola, and Scotland (1940) illustrated by photographs the proboscis, antenna, and wing of the closely related genus Lemnaphila. Hering (1950) briefly described and illustrated the excised male terminalia of H. xenophaga and H. nigricans. Kato (1955)'and Kuwayama (1955) de- scribed and illustrated several aspects of the external morphology of adult H. griseola. Hennig (1958) discussed the external morphology of the head and the thoracic chaetotaxy of several ephydrid repre- sentatives including H. griseola. With phylogenetic interpretations as a primary objective, Hennig made an important contribution here, especially in propos- ing the presence of vibrissae in Ephydridae. Dahl (1959) illustrated the adult mouthparts, female abdomen, and hind tarsus of H. griseola. Grigarick (1959) illustrated the male and female abdomina, the wing, and the mesonotum of H. griseola. Harrison (1959) illustrated the head, wing, and portions of the male terminalia of five New Zealand species of Hydrellia. Dahl (1964) illustrated the male ter- minalia of 17 species of Hydrellia from the Sten- hammar and Zetterstedt collections. Deonier (1964) illustrated the head of H. harti and the chaetotaxy and sclerite nomenclature of the head and thorax exemplified in Ephydra riparia. Dahl (1968) illus- trated parts of the male and female genitalia for three new species of Hydrellia in eastern Siberia. Sturtevant (1925, 1926) made the only contribu- tions to the internal morphology of Hydrellia in his survey of spermathecae in Acalyptratae. Bolwig (1940, 1941) described and illustrated the internal genitalia and mouthparts of Scatophila unicornis. NUMBER 6 8 These studies helped in understanding the mor- phology of these structures in Hydrellia. Several authors contributed morphological data on the immatures of Hydrellia: von Frauenfeld (1866) described the gross aspects of the metamor- phosis of H. albilabris; Stein (1867) summarized briefly the life cycle of H. griseola; Gercke (1879, 1882, 1889) briefly described and illustrated the puparia and feeding apparatus of H. mutata and H. fulviceps; and Marchal (1903) figured in gross aspect the third-instar larva of H. ranunculi. Though concerned with Ephydra riparia, an in- vestigation by Tragardh (1903) so lucidly illustrated the larval feeding apparatus, gut, musculature, and tracheal system that it formed a basis for anatomical study of larval Hydrellia. Brocher (1910) examined the gross morphological, ecological, and physiolog- ical aspects of the tracheal system' of H. mutata. Keilin (1915) also studied the metapneustic tracheal system of larval Hydrellia and illustrated the larval feeding apparatus. Malloch (1915) briefly described the larva and illustrated the puparium of H. griseola (as H. scapularis). Ping (1921), with his morpho- logical descriptions and illustrations, especially of the feeding-apparatus musculature of the larvae of Ephydra riparia (as E. subopaca), provided a very good basis for such work in Hydrellia. Likewise, Schiitte (1921), in discovering the phenomenon of seasonal (summer and winter) forms of the puparia of Hydromyza livens, provided the starting point for a similar investigation in Hydrellia which could answer several ecological questions. Wilke (1924) and Schoyen (1930) described the gross morphology of the immature instars of H. griseola. Collin (1928) illustrated the habitus of the third-instar larva of H. nasturtii. Johannsen (1935) briefly described the gross morphology of immature Hydrellia. Hennig (1943) summarized the known morphology of immature Hydrellia. Berg (1950) contributed much to the external morphological data on immature Hydrellia by describing and illus- trating most immature instars of six species. Seguy (1950) referred briefly to some aspects of morpholog- ical interest in immature Hydrellia. Hering (1951) commented generally on larval respiration and the puparial operculum. In 1952, Hennig discussed and illustrated much of the morphological data on im- mature Hydrellia. Lange et al. (1953) presented some significant photographs of the life-cycle stages of H. griseola. Kato (1955), Kuwayama (1955), and Grigarick (1959) showed detailed figures of the egg, feeding apparatus, larval body, puparium, spiracular peritremes, spinulosity, and setulosity of H. griseola. Many authors have contributed ecological data on Hydrellia, and the following have made major con- tributions: Stormer and Kleine (1911), Sorauer and Reh (1913), Linnaniemi (1913), Hendel (1926), and Kreuter (1927) reported some host plants of Hydrellia, principally of H. griseola; DeOng (1922) investigated the phenology of H. griseola and its damage to domestic rice; Hering (1924, 1937, 1951, 1957) contributed much to host-plant data and host damage of Palaearctic species of Hydrellia; Bala- chowsky and Mesnil (1935) reported a host-plant list and host damage for H. griseola; Thompson (1943) listed some hymenopterans parasitic on a few species of Hydrellia; Wahlgren (1947) published on the species of Hydrellia mining in Stratiotes aloides; Berg (1949, 1950) recorded some larval behavior, oviposition behavior, and some host-plant species of six species of Hydrellia; Laurence (1952) described some entomophagous behavior of H. griseola; Lange et al. (1953) discussed the biology and control of H. griseola in California; Grigarick (1959) studied the ecology of H. griseola in California rice fields; and Burghele (1959a, 1959b) and Fulmek (1962) recorded several hymenopterans parasitic in species of Hydrellia in the Palaearctic Region. Acknowledgments and Repositories I acknowledge the following sources of financial assistance for this research project: General Biolog- ical Supply House, Inc. (Turtox Scholarship) ; the National Science Foundation (summer grants at the Gulf Coast Research Laboratory, Ocean Springs, Mississippi, 1962; the University of Minnesota Bio- logical Station, Lake Itasca, Minnesota, 1963; and the Highlands Biological Station, Highlands, North Carolina, 1968); the Tennessee Academy of Sciences (summer grant at the Reelfoot Lake Biological Sta- tion, Tennessee, 1962); the University of Minnesota Agricultural Experiment Station (grant for research on wild rice pests, 1967) ; and Miami University (grant within United States College Work-Study Program). I am grateful to the following individuals: Dr. C. F. W. Muesebeck of the United States National SMITHSONIAN CONTRIBUTIONS TO ZOOLOGY Museum of Natural History for determination of the parasitic Hymenoptera; Dr. Jean Laffoon of Iowa State University for initial supervision of the research; Mr. E. A. Fiser of the Miami University Computer Center for analyzing certain data; my wife Mary Deonier, Miss Mary Wolf, and Miss Donna Mutchler for manuscript typing and proofreading. The following institutions and individuals loaned material to me for this research (abbreviations of those serving as paratype repositories are included in parentheses) : Academy of Natural Sciences of Philadelphia (ANSP), American Museum of Nat- ural History (AMNH), California Academy of Sciences (CAS), University of California at Davis (UCD), Canadian National Collection (CNC), Car- negie Museum, Chicago Natural History Museum, Cornell University (CU), Illinois State Natural History Survey, Iowa State University (ISC), Uni- versity of Kansas (UKC), Kansas State University, University of Massachusetts (UMM), University of Michigan, University of Minnesota (UMC), Uni- versity of Missouri, Museum of Comparative Zoology, Harvard University (MCZ), University of Nebraska, Oregon State University, United States National Museum of Natural History (USNM), Utah State University, Washington State University (WSC), Dr. C. O. Berg of Cornell University, Mr. Kenneth Goeden of the Oregon State Department of Agricul- ture (KG), Dr. W. W. Judd of the University of Western Ontario, Mr. George Steyskal of the United States Department of Agriculture, and Dr. M. R. Wheeler of the University of Texas (MRW). My own paratype holdings are labeled (DLD) under the pertinent new species. Methods and Materials COLLECTING OF ADULTS.?I collected adults by sweeping emergent vegetation, by lowering a killing tube over them on floating vegetation, and by a lighted-receptacle method. In the last method, I inserted the bottom of an open jar into the recessed lens of a flashlight and beamed the light to floating leaves. Adults of Hydrellia, Lispe, Hydromyza, Donaciinae, and a few others flew readily into the lighted jar. After several adults were in the jar, I closed it and replaced it with another. This method failed with H. bilobijera, H. trichaeta, and H. cras- sipes in North Carolina. Altogether, it was tried on about ten species. As a survey method for adults, I used a device invented and described by Grigarick (1959). As modified for this project, the device consisted of a circular aluminum pan 3 cm deep and 25 cm in diameter inserted in a hole of similar dimensions cut centrally in a square piece of styrofoam 50 by 3 cm. Holes in the corners of the styrofoam float accom- modated dowels run vertically through them and into the lake bottom to anchor the device. A deter- gent solution in the pan constituted the actual trap mechanism. The detergent destroyed the effective- ness of the tarsal hydrofuge setae so that flies landing on the solution sank immediately. In these traps I caught specimens of Donaciinae, Gerridae, Veliidae, Hydrometridae, Hydrophilidae, Tridactylidae, Col- lembola, Hymenoptera, and several dipterous fami- lies in addition to those of Hydrellia. I made most of my behavioral observations while collecting. After observing the behavior of the indi- vidual, I collected it with a killing tube or live- capture tube. I tried to keep some captured adults alive in the laboratory long enough to make addi- tional observations or for inseminated females to oviposit. PREPARATION OF ADULTS.?I killed and preserved in 80 percent ethanol specimens intended for anatom- ical and food-habits studies. I used a hydration series on these specimens before dissecting or section- ing them. I hydrated some specimens just before dissection and some just after fixation in Bouin's solution or dioxane in preparation for serial sec- tioning. To clear terminalia, I placed them in a hot, 10 percent potassium hydroxide solution for about 3 minutes and then added two to ten drops of 30 percent hydrogen peroxide over and around the floating terminalia. By closely watching the termi- nalia, I could ascertain when they were sufficiently bleached and desclerotized. The time required for this combined clearing and bleaching process varied with the condition of the terminalia and with the species. The average time was about 3.5 minutes? 3 minutes in hot caustic followed by a few seconds in the mixture of peroxide and caustic. Terminalia not quickly removed after the critical time limit often were overcleared and not usable. Apparently, the oxygen released from the hydrogen peroxide acts as the bleaching agent. After passing the terminalia through glacial acetic acid, distilled water, and NUMBER 6 8 absolute ethanol, I placed them in a small drop of glycerol in a depression slide for microscopic study. I used an ocular grid and squared paper to draw to scale the terminalia and other adult structures. Except for the structures that require a bilateral or whole view for proper interpretation, I show only the left half of bilaterally symmetrical structures in most drawings. I drew most of the adult structures with the aid of a compound microscope. COLLECTING OF IMMATURES.?In collecting the immatures of Hydrellia, I first searched for eggs on or near potential host plants in the habitats of the adults. When I discovered eggs I placed the entire plant on which they were laid into a plastic bag. After searching for eggs, I sampled the potential host plants in the locality for larvae and puparia. This was a random process in regard to larvae, for they can be found in situ most readily in the laboratory with strong direct illumination and a stereomicro- scope. In sampling some submergent plants in depths greater than 1.3 meters, I used a pike pole or employed a scuba diver. In studying the over- wintering of some Hydrellia, I used an ice chisel to cut through the ice to obtain samples of some host plants. In the laboratory, I examined all parts of each plant for immature Hydrellia by holding the plant part before a strong light. When there was evidence of mining, but no larvae apparent, I examined the tissue with a stereomicroscope. REARING.?When I discovered an egg or larva, I removed the plant part in or on which it was situated and placed the part in a small culture dish with tap water. I isolated each puparium together with some of the surrounding plant part in a 75-ml test tube containing a small amount of tap water and loosely plugged with cotton. I observed and recorded behavior (larval eclosion, molting, and ecdysis, mining activity, pupariation, adult emergence, etc.) and associated morphological changes daily. During these microscopic examina- tions, I would occasionally find a first- or second- instar larva previously undetected. In almost all cases I could distinguish between it and the original. This difficulty in detecting first-instar and some second-instar larvae existed because of their small size, translucency, the greater depth and shorter length of the mine. Because of this difficulty, I kept all plant material examined in screen-covered aquaria. I examined all of this plant material for immature Hydrellia periodically for two weeks. I prepared voucher specimens of each plant spe- cies examined for Hydrellia and also of some species collected for ecological indicators. Because of the large number of plants of many species examined, I could not preserve each plant found to be a host of Hydrellia. These specimens are on deposit at the Iowa State University Herbarium. Since temperature control was unavailable for the rearing work, I attempted only to record the tem- perature of the water in which specimens were reared. During the summer of 1963 at the University of Minnesota Biological Station and at most times thereafter to the present, I used a circular thermo- graph with sensor element kept in a beaker of tap water for recording daily temperature fluctuations in the laboratory. I changed the tap water around the sensor each time I changed the tap water in the rearing vessels. PREPARATION OF IMMATURES.?For species of Hydrellia for which I had collected sufficient speci- mens of each of the immature instars, I preserved some of each instar, but for several species I had so few specimens that I allowed all to develop to the adult instar. I preserved some Hydrellia eggs in 80 percent ethanol and some between filter-paper strips moist- ened with AFA and placed in small shell vials (about 5-ml capacity). I plugged the vials with cotton and immersed themf in either ethanol or AFA in large museum jars. I preserved the larvae and puparia by the same method, after killing them in hot water. To insure good specimens for anatomical and food-habits studies, I used AFA, but it had the one deleterious effect of bleaching the chlorophylls and carotenoids in the larval gut contents. For stereomicroscopic study, I placed larvae in 80 percent ethanol in depression slides. To study the same material with a compound microscope, I had only to add a few drops of glycerol to the ethanol in the concavity. I used this method for studying, drawing, and photographing eggs, larvae, and puparia. To store egg choria and puparia from which adults had been reared, I placed them in small drops of glycerol in the bottom of cork-stoppered, glass micro- vials and placed them on the pin holding the cor- responding adult. This method afforded a flexibility SMITHSONIAN CONTRIBUTIONS TO ZOOLOGY in studying specimens that was not afforded in the use of permanent glass-slide mounts; also, it kept most of the life-cycle stages together. Because of the microscopic size of some structures, e.g., setulae, spinules, and some other cuticular proc- esses, I did not risk clearing immature specimens with potassium hydroxide. Clearing was necessary only infrequently, and then it usually involved parts of puparia, for which a very mild agent such as methyl salicylate sufficed. I used the same procedures and materials for drawing structures of immatures as I did for adults, except that I made the habitus drawings of puparia initially with a stereomicroscope and then filled in some details under a compound microscope. Morphology Glossary New structures, configurations, and indices encoun- tered in a taxonomic study must be named. Since zoological terminology is already burdened with much ambiguity, I constructed the following defini- tions and explanations of structural and index names. In these, all distance measurements are straight-line and uniplanar on preserved specimens unless otherwise specified. TERMINOLOGY FOR ADULTS A-index: The quotient of the subcranial breadth divided by the anteclypeal breadth. (Figure 11.) Anteclypeal breadth (acb): The maximum transverse distance between the outer edges of the paraclypeal phragmata of the anteclypeus. (Figure 11.) Anteocellar distance (aod) : The distance between the anterior margin of the median ocellus and the upper edge of the ptilinal fissure along the frontal midline. (Figure 11.) Anterior pronto-orbital (afr): The anterior seta of the two setae commonly prominent in each fronto-orbital area. (Figure 1.) Apicodorsal antennal (a3ap): A spinous seta on the dorsal apex of antennal segment 2. Aristal rays (arr): All of the trichoid projections of each arista including the apical one. (Figure 1.) B-index: The quotient of the maximum anteroposterior extent of the fused surstyli divided by the length of the sclerotized portion of the cercus as measured in ventral view. (Figure 4.) Basal coxal (be) : A seta inserted laterally on the mid coxa. (Figure 6.) Basal end of costa (bee): The enlarged basal portion of the costa proper adjacent to the humeral plate. (Figure 135.) Body length: The distance between the most prominent part of the face and the posterior end of the abdomen. It is measured in lateral view and as if the head and abdomen were aligned horizontally. C-index: The quotient of the midline anteroposterior extent of sternum 5 divided by the projected length of the postgonite uncus. Color: The descriptions of color apply to views perpendi- cular to the sclerite concerned unless otherwise stated. Color designations follow the ISCC-NBS method (Kelly and Judd, 1955). Copulobus (cl): One of a pair of posterior lobes, or projections, of male sternum 5. (Figures 4, 136?138.) Costal section I: The distance between the distal edges of crossvein h and Rt apex. (Figure 135.) Costal section II: The distance between the distal edges of Rj and R2 + 3 apices. (Figure 135.) Costal section III: The distance between the distal edges of R2 + 3 and R4 + 5 apices. (Figure 135.) Costal section IV: The distance between the distal edges of R4 + 5 and M t + 2 apices. (Figure 135.) Costal section V: The distance between the distal edge of Mj + 2 apex and the proximal edge of M3+CU1 apex. (Figure 135.) Epistomal breadth: The transverse distance between the pair of primary facial rows at the level of the epistoma, measured from the inner edge of the setal sockets. (Figure 1.) Epistomal index: The quotient of the epistomal breadth divided by the minimum interocular distance .(Figure 1.) Frontal vitta {mj): The median quadrangular area of the frons on which are situated the ocellar triangle and the ocellar and postocellar setae. (Figure 1.) Fronto-orbital area: The narrow lateral section of each parafrontale parallel and contiguous to the compound eye on which the fronto-orbital setae are inserted. This usage differs from that of some other authors. (Figure 1.) Interfractural costal: A seta inserted on the costa between the two costal "fractures." (Figure 135.) Laterotergite (It): The metapleuron of other terminology. (Figure 6.) M1 + g index: The quotient of the distance between the distal edge of the junction of crossvein m and Mj + 2 divided by the distance between the distal edge of the m and M t + 2 junction and the distal edge of the r-m and M 1 + 2 junction. (Figure 135.) Mesanepimeron (aem^) : Part of the pteropleuron of other terminology. (Figure 6.) Mesanepisternum (aes2): The mesopleuron of other ter- minology. (Figure 6.) Mesofacial height (mfh): The distance between the outer edge of the subcranial cavity and the lower edge of the ptilinal fissure along the midline of the face. (Figure 11.) Mesofacial index: The quotient of the mesofacial height divided by the minimum interocular distance. Mesokatepisternum (kes2): The sternopleuron of other terminology. (Figure 6.) NUMBER 6 8 Metapleuron {mp): The hypopleuron of other terminol- ogy. (Figure 6.) Minimum interocular distance: The minimum transverse distance between the compound eyes in the area of the face. (Figures 1, 11, 139.) Ocular height (vod): The maximum distance between the upper and lower edges of the compound eye. The line of measurement is not quite vertical in most species. (Figure 9.) Ocular index: The quotient of the nearly vertical ocular height divided by the subocular height. (Figure 9.) Parafrontale (If): The region of the frons between the frontal vitta and the upper edge of the compound eye. (Figure 11.) Postdorsocentral (pdc): A seta inserted in the dorsocentral line posterior to the transverse sulcus. (Figure 6.) Posterior fronto-orbital (pfr) : The posterior seta of the two setae commonly prominent in each fronto-orbital area. (Figure 1.) Postgonite (pog): One of a pair of curved-, fingerlike projections of the gonal arch on each side of the disti- phallus. In terminalia preparations, the postgonite usually appears to be anterior to the pregonite. (Figures 4, 10, 13, 136, 137.) Postgonite uncus (pu) : The apical section of the postgo- nite which is often hooklike or clawlike and usually dis- tinctly more heavily sclerotized than the remainder of the postgonite. (Figures 10, 13.) Postocular: One of the setae inserted in a row posterior to and more or less parallel with the posterior edge of each compound eye. (Figure 9.) Pre dor so central (adc): A seta inserted in the dorsocentral line anterior to the transverse sulcus. (Figure 6.) Pregonite (prg): One of a pair of bifurcate setose pro- jections of the gonal arch on each side of the basiphallus. In terminalia preparations, the pregonite usually appears to be posterior to the postgonite. (Figures 4, 10, 13.) Primary facial (pfa): One of the longer facial setae in- serted in a row on each side of the face parallel and medial to the ptilinal fissure. The primary facial row is parallel to the outer edge of the obscured epistomal sulcus. (Figure 1.) Secondary facial (sfa): One of the shorter facial setae that are often in a row parallel and lateral to the primary facial row. Subcranial breadth (scb): The maximum transverse dis- tance between the outer lateral edges of the subcranial cavity. (Figure 11.) Subocular height (soh): The minimum distance between the lower edge of each compound eye and the outer lateral edge of the subcranial cavity. (Figure 9.) Vertex breadth (vb): The distance between the upper edges of the compound eyes at the level of the lateral ocelli. (Figure 11.) Vertex index: The quotient of the vertex breadth divided by the anteocellar distance. (Figure 11.) Wing length: The distance between the apex of the tegula and the wing tip. (Figure 135.) TERMINOLOGY FOR IMMATURES Anal-plate index: The quotient of the transverse extent of the anal plate divided by the midline anteroposterior extent of the anal plate. (Figure 107.) Bifurcation index: The quotient of the longitudinal dis- tance between the level of the phragmatal bifurcation and the posterior end of the ventral phragmatal ramus divided by the minimum distance between the end of the dorsal phragmatal ramus and the upper edge of the ventral phragmatal ramus. (Figure 85.) Clypeal arch: The area of inclination in the frontoclypeus just anterior to the cheliform spot. (Figure 85.) Cly'peal-arch index: The dorsoventral extent of the fronto- clypeus at the level of the anterior edge of the cheliform spot divided by the dorsoventral extent of the fronto- clypeus at the level of the anterior edge of the labial gland orifice. (Figure 85.) Early pupa: The pupa as it appears prior to the time when its compound eyes become apparent. Egg length: The distance between the ends of the egg as measured in dorsal view. (Figures 75, 77, 78, 125.) Frontoclypeal length: The distance between the anterior edge of the frontoclypeus and the posterior end of the ventral phragmatal ramus as measured in lateral view. (Figures 84, 75.) Late pupa: The pupa as it appears after the time when its compound eyes become apparent. Larval length: The distance between the anterior edge of the head lobe and the posterior end of the spiracular peritreme measured with the larva outstretched. (Figure 124.) Maximum egg breadth: The maximum transverse extent of the egg as measured in dorsal view. (Figures 75?78.) Maximum larval breadth: The maximum transverse ex- tent of the larva as measured in dorsal view. Maximum mouth-hook base thickness: The maximum thickness of the articulated end of the mouth-hook as measured in lateral view. (Figure 85.) Maximum mouth-hook beak thickness: The maximum thickness of the free (or beak) end of the mouth-hook just distal to the enlarged base as measured in lateral view. (Figure 85.) Minimum puparial breadth: The minimum transverse extent of abdominal segment 8 as measured in dorsal view anterior to the tracheospiracular siphon. (Figure 107.) Phragmatal index: The quotient of the longitudinal dis- tance between the anterior edge of the frontoclypeus and the level of the phragmatal bifurcation divided by the longitudinal distance between the level of the phragmatal bifurcation and the posterior end of the ventral phrag- matal ramus. (Figure 85.) 8 SMITHSONIAN CONTRIBUTIONS TO ZOOLOGY Puparial length: The distance between the anterior pro- thoracic margin of the puparium and the posterior end of the spiracular peritreme measured as if the puparium were outstretched. (Figure 107.) Ventral frontoclypeal index: The quotient of the distance between the anterior edge of the frontoclypeus and the anterior edge of the labial gland orifice divided by the dorsoventral extent of the frontoclypeus midway between the anterior edge of the frontoclypeus and the anterior edge of the labial gland orifice. (Figure 85.) External Morphology of Adult HEAD.?There are several controversies concerning the morphology of the schizophoran head. The major of these is centered on the clypeus and frons. Snodgrass (1935, pp. 317, 322, 323), in elucidating the morphology of the clypeus, stated: In the higher Diptera, the median part of the clypeus becomes an independent sclerite, but the dilator muscles of the pump retain their attachments upon it. . . . The inverted V?shaped plate of the anterior wall of the rostrum (Fig. 174, C, D, dp) bears upon its lateral arms the origins of the dilator muscles of the cibarial pump (D, 3). There can be little question, therefore, that this sclerite represents at least the median part of the clypeus in the head of Tabanus (Fig. 171 B, dp). . . . The attachment of the dilator muscles of the cibarial pump on the arms of the V-shaped rostral plate, however, clearly demonstrates the clypeal origin of this sclerite, and confirmatory evidence of its homology with the median clypeal region in Tabanus is seen in the fact that a pair of labral muscle? (Fig 174 D, 2) take their origin on its dorsal part. The smaller sclerite above the V-shaped clypeal plate (C, e) is either a part of the clypeus or a secondary sclerotization hinging the latter to the lower margin of the face. In his discussion, Snodgrass did not commit himself on the morphology of the region of the head capsule between the subcranial margin, antennal sockets, and the compound eyes, but by emphasizing that the hinged anterior plate of the basiproboscis is homolo- gous with the median clypeal region in Tabanus he inferred that some part of the clypeus remains above the subcranial margin. In a later paper Snodgrass (1944, p. 70) withdraw this inference by stating that "In the Cyclorrhapha and some of the Brachycera, the median, muscle-bearing plate of the clypeus be- comes isolated by a membranization of the surround- ing clypeal area, and is thus flexible on its hinge with the frons." Between 1935 and 1944, Snodgrass decided that the muscoid homologue of the tabanid median clypeal plate was hinged "with the frons" and not "to the lower margin of the face." In 1953 he omitted any discussion of the partial membran- ization of the clypeus and the boundaries of the frons in explanations of the evolution of the dip- terous cibarium and proboscis. Snodgrass never explained the fate of the epis- tomal (frontoclypeal) sulcus, but since, by the accepted definition, the epistomal sulcus is the ex- ternal cuticular furrow, or groove, between the ante- rior tentorial pits which delimits clypeus and frons, I must assume that in the Snodgrass view the ante- rior tentorial pits and epistomal sulcus disappeared completely by membranization in the evolution of the schizophoran head capsule. Kim and Cook (1966, p. 79) stated that "The clypeus (tormae of Peterson (1916) ) lies below the ventral margin of the prefrons (Figures 1, 2, 10, 11) between the labrum and the frons and is anteriorly delimited by the membrane above the labrum and posteriorly by the membranous frontoclypeal suture." These authors apparently failed to answer the ques- tion of how the frontoclypeal (epistomal) suture or sulcus, which they interpreted to be the basiproboscis membrane between anteclypeus and head capsule proper, became entirely disconnected from its origin in the anterior tentorial pits. They stated: "The anterior tentorial pit is indistinct externally in sphaerocerids but is situated at the upper lateral angle of the prefrons below the eye and above the frontal ridge" (Kim and Cook, 1966, p. 79). Frick (1952) and Downes (1958) showed some evidence of the epistomal sulcus in Agromyzidae and Sarcophagidae respectively. Frick (1952) labeled what he considered the anterior tentorial pits in the apparent epistomal sulcus in Agromyzidae. My in- vestigation of the problem in Hydrellia indicated the validity of the interpretations of Frick and Downes. As shown in Figures 1, 11, 139, the epistomal, or frontoclypeal, sulcus extends from the lateral sub- cranial margins dorsad along the primary facial rows as faint internal cuticular thickenings. These thick- enings were visible only after clearing, and they were not connected above in the two species studied closely. The epistomal sulcus is incomplete in several species of insects. The ptilinal fissure apparently extends ventrolaterad from its conspicuous supra- antennal arc to the subocular genal regions. These paraocular extensions delimit the lateral facial areas next to the orbits as the parafacialia. Downes (1958) labeled the facial areas between the extensions of the NUMBER 6 8 ptilinal fissure and the epistomal sulcus as facial ridges, while, for some reason, Frick (1952) labeled the same areas parafacial regions. Frick did not label the areas contiguous to the eyes and set off by the ptilinal fissure extensions. Bolwig (1941) illustrated the upper extremities of the epistomal sulcus in Scatophila unicornis Czerny (Ephydridae), but he interpreted them as dorsal tentorial pits. It is possible that Bolwig (1941, p. 3) was nearer to the truth than Snodgrass, Frick, or Downes when he stated "that those extending from the impression beneath the antennae to the mouth opening [subcranial cavity] (d.t.) are homologous with the dorsal arms of the tentorium. The lateral thickenings (ant.t.) of the edge of the mouth open- ing are then supposed to be homologous with the anterior arms of the tentorium, while the thicken- ings stretching from the occipital foramen down- wards (pt.) to the mouth opening are supposed to be homologous with the posterior arms of the ten- torium." This concept of continuous sulci from dorsal tentorial pits through anterior tentorial pits to posterior tentorial pits can be visualized in illus- trations by Bonhag (1951, figs. 1-4). For Bolwig's interpretation to obtain, the anterior tentorial pits need have shifted posteroventrad only a short dis- tance. It is perhaps impractical with our present knowledge to attempt to distinguish pit and sulcus. In the classical view, the thickening between dorsal and anterior tentorial pits would be the epistomal sulcus, while the thickening between anterior and posterior tentorial pits would be the subgenal sulcus (in part, hypostomal sulcus). In my synthesis of several concepts, I have at- tempted to present a working scheme very similar to that used in modern calypterate taxonomy. In this collation, the question of the limits of postclypeus and the frons was considered moot and relatively unimportant because tradition has favored such terms as mesofacial plate, medifacies, medifacial plate, and facial plate and because specialization in cibarial muscles has obscured ordinary divisions. I have called the anterior edge of the subcranial cavity between the extremities of the epistomal sulcus the epistoma (Figures 1, 11, 139). There is considerable precedent for this designation, and the term is exten- sively used in chaetotaxy. The area superior to this epistoma between the indistinct arms of the epis- tomal sulcus I have called the facial plate (Figures 1, 11, 139). The ptilinal fissure forms the dorsal and lateral limits of the face. Frick (1952) called this area the mesofacial plate and placed the antennal sockets as the upper limit. Downes (1958) termed a similar area in Sarcophagidae the facial plate. The antennal sockets may be considered to demarcate a facial subregion above and between them and the ptilinal fissure called the frontal lunule. The cuticular area bounded by the lower edge of the compound eye, the lateral subcranial margin, the lower angle of the face, and the posterolateral post- genal inflection is commonly called the gena. Since, however, the gena is rather ill-defined, I have used the term subocular height in defining the ocular index (Figure 9). The term defined is identical to the genal height illustrated by Dahl (1959), but it avoids the ambiguity of the gena concept, for the- oretically much of the immediate postocular area is also part of the gena. Above and posterior to the ptilinal fissure is the frontal vitta, bordered on each side by an indistinct parafrontale, or genovertical plate. It often is conven- ient to distinguish a narrow paraocular section of the parafrontale as the fronto-orbital area. The frontal vitta and the parafrontalia constitute the frons. Frick (1952) distinguished only the frontal vitta, and lateral to it a genovertical plate, which is apparently synonymous with the parafrontale. At the vertex in the frontal vitta, the ocellar triangle protrudes only slightly in Hydrellia. Posteroventrad of the ocellar triangle, the frons transcends with the occiput, i.e., the occipital sulcus is absent. The incomplete post- occipital sulcus only partially defines the postocciput with its inconspicuous occipital condyles. One can distinguish three major divisions of the proboscis: the basiproboscis extending from the sub- cranial margin to and including the bases of the maxillary palpi, the mediproboscis extending from the bases of the maxillary palpi to the distal end of the hypopharynx, and the distiproboscis extending from the distal end of the hypopharynx to the ex- tremities of the labella (Figures 1, 134, 139). In the basiproboscis, the conjunctiva (membrane) encloses the cibarium, the paraclypeal phragmata, and the stipes. Only the anteclypeus, or rostral plate, and the maxillary palpi are exposed (Figures 134, 139, 140). In the mediproboscis, the sclerotized prementum is the prominent structure. Anteromedially from the prementum is the labial gutter containing the hypo- 10 SMITHSONIAN CONTRIBUTIONS TO ZOOLOGY pharynx and the labrum. Enclosed within the con- junctiva basal to the labrum are the labral apodemes and the trachea-like siphon connecting the labial gutter with the cibarium. The distiproboscis con- sists essentially of the labella. Each labellum is par- tially supported above by a labellar sclerite and con- sists of seven canaliculi which radiate from the margin of the prestomum. Apparently, the canaliculi terminate in the prestomum posterior to the apparent sclerotized prestomal margin. Each canaliculus has a row of several canalicular teeth. The schizophoran cibarium is homologous with the generalized orthopteroid cibarium, i.e., it is the chamber between the connected epipharyngeal clypeal wall and sitophore (anterior and posterior cibarial walls). Since in Diptera the chamber has a pumping function, it is stabilized either by antagon- istic muscles as in Nematocera and many Brachycera or by inflections of the lateral borders of the ante- clypeus, called paraclypeal phragmata. Within the cibarium, inserted on the anterior, or epipharyngeal, wall are small trichoid sensilla: paired clusters of three near the distal end of the cibarium and a biseriate row running dorsolaterally from each of these initial clusters to the end of the anteclypeus (Figures 1, 140-142). Frey (1921) apparently first illustrated these cibarial sensilla in Hydrellia and other ephydrid genera, but he simply called them setae as did Bolwig (1941). I closely examined and counted these cibarial sensilla only in several speci- mens of H. griseola, but I confirmed their presence in several congeners and in representatives of all dipterous suborders. The number and arrangement of the cibarial sensilla apparently varies somewhat in Hydrellia, for Frey (1921, p. 137) stated: "Die obere Pharynxwand hat 2 Reihen von 5-6 Borsten, wozu vorn jederseits eine Gruppe von 3 kiirzeren, dichter gestellten kommt." This concerned H. obscura Meigen. In H. griseola, I found four pairs of sensilla in the two long rows. Regarding the function of the cibarial sensilla, I offer the following hypothesis. The cibarium is the first and primary ingestion pump, and among the Schizophora the sole pump. This being the case, proprioceptors are necessary to maintain cibarial rhythm. The cibarial pump must perform the initial ingestive suction and in Schizophora the subsequent expulsion from the cibarium into the esophagus. To perform these two progressive functions, the shape of the cibarium must change rhythmically, for no valve has been demonstrated distal to the cibarium to prevent backflow from the contracting cibarium. My hypothesis is that during the initial ingestive phase the cibarium dilates progressively proximad, and during the subsequent emptying it constricts progressively proximad, and that the cibarial sensilla coordinate or govern these rhythmic changes. As proprioceptors, these sensilla enable the cibarium to act as a valve as well as a pumping chamber. As they touch the opposite wall, impulses pass from their neurones. In this way, regurgitative loss of cibarial contents is reduced. Before conducting this investigation, I thought the high rates of proboscis protraction and retraction observed in Parydra (Ephydridae) was evidence that pinching of the siphon connecting the food meatus and cibarium prevented regurgitative cibarial loss. The absence of such repeated flexion in many Schizophora, including Hydrellia, controverted this interpretation. Bonhag (1951) illustrated and described a func- tional mouth (and valve) just distal to the cibarium operated by muscles termed the labral compressors. If this functional mouth exists in the Schizophora, it can only be the trachea-like siphon of the food meatus. I did not find the unpaired labral com- pressors in the preparations of H. griseola nor did I find the median labral process on which they would be inserted. The cranial chaetotaxy differs very little from that common in Acalyptratae. The conspicuous macro- chaetae are the genal, inner and outer verticals, and postocellar. The ocellar, anterior and posterior fronto-orbitals, apicodorsals of antennal segment 2, postocular, and primary facial setae are less con- spicuous. In most species, smaller secondary facials occur, usually lateral to the primary facial row. According to Hennig (1958), the vibrissa is pres- ent but indistinguishable in Ephydridae. Also, according to Hennig, the postocellar has replaced the postvertical at least in prominence and approxi- mate location. THORAX.?My presentation of the gross external morphology of the thorax proper of Hydrellia is essentially an interpretative synthesis of the contri- butions of Osten Sacken (1881), Curran (1934), Snodgrass (1935), Comstock (1940), Crampton (1942), Bonhag (1949), and Downes (1955). NUMBER 6 8 11 The prothorax consists of the greatly reduced antepronotum as part of the border of the pro- thoracic foramen, the postpronotum, the propleura, the prosternum, and the fore legs (Figure 6). The prosternum consists of a small presternum and a considerably larger probasisternum with a median sulcus and a prosternellum, the defining sternacostal sulcus of which is continuous with the median probasisternal sulcus. I am uncertain about the nature of an apparent sclerite just above and pos- terior to pleurocoxal condyle 1 (Figure 6) , but it is similar to the proepimeron illustrated by Crampton (1942) andBonhag (1949). The anterior spiracle?or, as I have called it, the prothoracic spiracle according to Keilin's (1944) analysis?is situated in a small fossa inferior to the postpronotum and posterior to the superior termina- tion of the propleural sulcus. The mesonotum is divided by the transverse and scutoscutellar sulci and the postscutellar suture into the notopleuron, mesoscutum, mesoscutellum, and mesopostscutellum respectively. The mediotergite and the laterotergites are the ordinary discernible sclerites of the mesopostscutellum, but often an anatergite and katatergite can be distinguished as constituents of the laterotergite. The expanded mesopleura, each divided by the mesopleural sulcus into the mesepisternum and the mesepimeron (Fig- ure 6) , the wings, and the mid legs are the remaining main constituents of the mesothorax. The dorsopleural (or notopleural) sulcus sepa- rates the notopleuron and the mesopleuron. At the anterior end of this sulcus is a slightly distinguish- able sclerotization, which may represent part of the mesoprescutum. Bonhag (1949) illustrated the mesoprescutum as a pair of small triangular lobes interposed between the evident notopleura and mesopleura; however, Snodgrass (1935) illustrated the notopleura as the prescutal lobes. The large mesepisternum is partially divided by the incomplete mesanepistemal sulcus into the mesanepisternum and the mesokatepisternum. The posterior part of the mesanepisternum is traversed by a secondary suture, or membranous cleft. A con- spicuous sclerite occupies the upper corner of this cleft. Crampton (1942) showed a sclerite which resembles this one except for its apparent fusion with the posterior mesanepistemal lobe isolated by the secondary cleft. Crampton called this sclerite the anterior basalare. Downes (1955) showed two sclerites, basalarites A and B situated in the upper part of this cleft. Basalarite A is separated and occu- pies the uppermost part of the cleft. Basalarite B is connected by a narrow neck to the isolated mesanepistemal lobe. Bonhag (1949) showed two separate sclerites occupying this cleft in Tabanus. The anterior sclerite called the basalare extends down to the mesokatepisternum. Bonhag did not name the small, second sclerite in the uppermost part of the cleft. The arrangement of sclerites in the secon- dary cleft and the course of the mesopleural sulcus in Hydrellia approximates most closely Downes' illustration except that the basalar ampulla has evidently been displaced posteriad. The subalar ampulla is a distinct crescentic sclerotization in the lower wing area. The mesepimeron is distinctly represented only by the mesanepimeron. The mesokatepirneron is prob- ably represented by the small sclerite contiguous to the anterior edge of the metathoracic spiracular peritreme. It is possible that this small sclerite is part of the metapleuron, as may be inferred from Crampton's illustration. According to Crampton, the mesokatepimeron is usually indistinguishably fused with the large meron (meropleurite). Downes showed the mesokatepimeron to be slightly distin- guished by the coxopleural streak from the upper meral margin. The laterotergite extends down farther in Hydrellia than shown by Crampton, Bonhag, and Downes. This fact and the partial demarcation of a small triangular lobe anterior to the metathoracic peritreme illustrated by Bonhag led me to emphasize the first-mentioned interpretation of the location of the mesokatepimeron. The precoxale (anterior to and above the mid coxa) of the mesokatepisternum is distinguished as a narrow glossy sclerotization. The posterior meral margin is similarly distinguished. The apparent ventral extension of the meso- katepisternum is considered by Snodgrass (1935) a composite of the sternum, precoxale, and episternum. Basically, Bonhag differed little or not at all from this view. Of the metathorax, only thoracic phragma 2, the greatly narrowed metapleura, the halteres, and the hind legs remain distinct (Figure 6) . The meta- pleuron consists of a metathoracic precoxale, appar- ently defined by the lower part of the sclerotized 12 SMITHSONIAN CONTRIBUTIONS TO ZOOLOGY posterior meral margin and a continuation of this heavy sclerotization posteriad of coxa 3, the metepis- ternum, which is sharply narrowed and apparently divided by a transverse sulcus near the posterior meral lobe, and the very linear metepimeron behind the indistinct metapleural sulcus. The wing-vein nomenclature is modified from the Comstock-Needham system as illustrated by Downes (1955). Because of taxonomic importance, I added the designation "basal end of the costa" (Figure 135). The halter has three apparent divisions: the knoblike capitellum, the pedicel, and the basal scabellum. My thoracic chaetotactic nomenclature is modi- fied from Osten Sacken (1881), Curran (1934), and Cornstock (1940). The main modifications are in the pleural setae, which are named according to the sclerite on which they are inserted, e.g., post- pronotals, propleural, mesanepisternals, mesokatepis- ternal(s), and basicoxal(s). On the mesonotum, the dorsocentrals and acrostichals anterior to the incomplete transverse sulcus are termed predor- socentrals and preacrostichals, while those posterior to this point are called the postdorsocentrals and postacrostichals. The prescutellar macrochaeta is often considered a dorsocentral, though it is inserted between the dorsocentral and acrostichal lines. There is one large postalar macrochaeta and a small one inserted midway on the postalar bridge. More or less in line with the large postalar and the prescutel- lar is one large interalar (corrected from "intraalar" by Sturtevant and Wheeler, 1954). Above the meso- pleural wing process are two small supra-alar setae. There are two notopleural macrochaetae and one lat- eral macrochaeta just anterior to the notopleural apex. Inserted on the scutellar margin are the basal scutellar, intermediate scutellar, and the apical scu- tellar. The predominant features of the wing are the two costal fractures or discontinuities, one just proximal to Ri apex and one slightly distal to crossvein h. Although the whole costa is setose, the setae on the enlarged basal end of the costa and the setae between the costal fractures, the anterior and the dorsal inter- fractural costals, are of main taxonomic importance (Figure 135). ABDOMEN.?Except for dimensional and color characters, secondary sex characters in Hydrellia appear to be limited to the abdomen, especially the postabdomen. Only Hering (1950), Dahl (1959, 1964), Grigarick (1959), and Harrison (1959) have studied the abdomen other than superficially. It is convenient and traditional to distinguish two abdominal regions?the preabdomen and the post- abdomen. According to Steyskal's (1957) hypothet- ical illustration, the division in the male abdomen should be between terga 6 and 7. Unfortunately, these two terga are obliterated in male Hydrellia so that, from a practical view, I placed the division just posterior to tergum 5 in both sexes. Thus, in my discussion, the first five segments (except sternum 5) constitute the preabdomen and the remainder con- stitute the postabdomen. The male sternum 5 must also be considered part of the terminalia (Figures 4,5,136-138) in Hydrellia. Tergum 1 is partially fused with tergum 2 and is so short that it is very much obscured in dorsal view in its normal relation to the thorax. This condition led some earlier workers to assume there were four instead of five preabdominal terga. This assumption was enhanced by the slight displacement posteriad of sternum 1 and the highly modified form of sternum 5 in the male (Figure 4) . Although there is some assumption in simple sequential identification of the sterna, the condition of the female sternum in H. griseola as shown by Grigarick (1959, pi. 1, fig. 3) seems to support this course. I have illustrated the male postabdomen in Fig- ures 4, 5, 10, 12, 13, 136-138. Snodgrass (1957) believed that his analysis of the evolution and homologies of external male genitalia could be applied to all insect groups, but he failed to do this for the higher Diptera. Insufficient morphological knowledge has precluded the homologizing of the parts of the terminalia with paramere or mesomere. For this reason, I selected a system of terminalia nomenclature synthesized from Crampton (1944), van Emden and Hennig (1956), and Tuxen (1956). I think this system is more appropriate to the situa- tion in Hydrellia than that of Crampton (1944), Wirth (1948),orSteyskal (1957). I have considered the terminal tergum a syntergum of segments 9 and 10. Steyskal (1957) and previous authors have called this tergum the epandrium. There is little need for this name in Hydrellia because the tergum is only occasionally important taxonom- ically, and it seems to be only slightly involved in copulation. The surstyli have fused to a variable NUMBER 6 8 13 extent, forming a ventral flap partially covering the phallus (often most of the basal half). Crampton (1944) and Steyskal (1957) used the term surstyli, while Dahl (1959) called the two structures together the hypandrium. I have called the entire intromit- tent structure the phallus, partly because I was uncertain how much of it represented the true aedeagus and partly because phallus was a con- venient suffix (basiphallus and distiphallus). I needed these terms because of the taxonomic impor- tance of the common division of the phallus into a somewhat bulbous, heavily sclerotized basal part (basiphallus) and a usually narrower, lightly sclero- tized distal part (distiphallus). The basiphallus may be the actual phallobase and the distiphallus?the aedeagus of Snodgrass (1935, 1959). Phase micro- scopy confirms the presence of a duct (presumably the ejaculatory duct) passing through the length of the distiphallus. The phallus is suspended from the posterior ends of a structure I have called the gonal arch (after Tuxen, 1956, and Phillip Clausen, of the University of Minnesota, in litt., 1964). Crampton (1944) and Steyskal (1957) did not discuss or illustrate any comparable structure. Tuxen (1956) illustrated a structure in Aedes called the proctiger, which is somewhat similar to the gonal arch. The anterior ends of the gonal arch articulate with or are con- tinuous with a median phallapodeme, or aedeagal apodeme, and with sternum 5. Projecting from the gonal arch at sternum 5 are two pairs of sclerites or lobes called gonites (after Tuxen, 1956). The larger pair are considered the postgonites. The smaller, often inconspicuous pair with setose, bifurcated tips are considered the pregonites despite their usual pos- terior termination. In many species, the pregonites appear to have a more definite articulation with the phallapodeme. I have called the more heavily sclero- tized, clawlike end of the postgonite the postgonite uncus. This specific designation was needed because of its taxonomic significance. Sternum 5 varies specifically in shape. Usually, the general impression is a horseshoe shape. The posterior lobes are so significant taxonomically that I have designated them copulobi after Crampton (1944). Although almost certainly additional unillustrated muscles function in phallic movement, I have anal- yzed the process in the following manner. The phallapodeme and the gonal arch are pulled anteriad by contraction of phallic depressors inserted at their anterior ends and originating on or near sternum 4 (Figure 12). This contraction and phallapodeme movement has two effects: (1) it lowers, or de- presses, the phallus as a result of leverage between the superior condyle of the basiphallic socket and the posterior end of the phallapodeme; and (2) it lowers the postgonites, which in most cases are attached or closely appressed to the anterior end of the phallapodeme. (I surmise that the copulobi are simultaneously depressed by the same or associated muscles. When depressed, the bilobate sternum 5 would fit the female postabdomen just anterior to the cerci like a saddle). The postgonites either titillate some female terminalia or hold the cerci erect to permit coitus. This hypothesis is supported by the fact that in several species the distiphallus in repose projects anteriorly over the free posterior margin of sternum 5, and by the angle, or attitude, of the mounted male in several species. Phallic elevation is effected by the phallic levators levering the pos- terior end of the phallapodeme against the inferior condyle of the basiphallic socket and by the phallic depressors and associated muscles relaxing synchron- ously. Once elevated into the genital pouch, or cubiculum (after Crampton, 1944), the phallus is apparently retained by the gonites, sternum 5, or by both of these structures. I have illustrated parts of the female terminalia of a few species in Figures 2, 67-72. Grigarick (1959) accurately illustrated the female abdomen in ventro- lateral view. Dahl (1959) illustrated only a portion of the female abdomen of H. griseola. Wirth (1948) and Tuxen (1956) referred to the cerci as anal lobes and valvulae mediales respectively. This is under- standable, since in female Hydrellia and many other Diptera segments 9 and 10 have been obliterated so that tergum and sternum 8 are immediately anterior to the cerci. Sternum 8 is often called the subgenital, or pregenital, plate. Internal Morphology of Adult GUT.?I have illustrated most of the gut of H. griseola in Figures 1, 3. These illustrations seem also to apply fairly well for H. tibialis, H. bilobifera, H. harti, and H. columbata. In gross anatomy, the foregut, or stomodeum, is a uniform tube from the 14 SMITHSONIAN CONTRIBUTIONS TO ZOOLOGY dorsal cornua of the paraclypeal phragmata to the vicinity of the cardia near the level of thoracic phragma 2. Here there is a diverticulum from the esophagus, the crop meatus, leading to the crop proper. Distended, the crop occupies nearly the ven- tral half of the abdominal cavity. The midgut starts with the cardia, which contains the stomodeal valve. From this structure to a level about midlength of the crop, the midgut is a linear tube with the labial gland ducts and labial, or salivary, glands parallel and contiguous. At about midlength of the crop the midgut is bent sharply dorsad and coiled three or four times. The pyloric valve and most of the ileum are similarly coiled. The site of evagination of the Malpighian tubules is obscured by the gut coils. In H. griseola and H. tibialis the Malpighian tubules branch from two stems, or bases, arising near the pyloric valve. Two tubules run anteriad and two posteriad, both pairs parallel and close to the gut. The rectal valve and rectum with its rectal glands are the conspicuous parts of the hindgut, or proc- todeum. The anus is located between the cerci. The inner lines apparently representing the intima of the midgut and hindgut in Figure 3 may actually represent the peritrophic membrane. INTERNAL GENITALIA.?The internal genitalia of H. bilobifera illustrated in Figures 2, 14 are similar in outline to those of H. griseola, H. tibialis, and H. columbata. The internal male genitalia consist of testes, vasa deferentia, pouchlike enlargements of the vasa deferentia called seminal vesicles, a median, unpaired ejaculatory duct, and accessory glands. The testis and the initial portion of the vas deferens are suspended and covered by a thin peritoneal sheath. Although this sheath partially obscured the testicular components, I observed a germarium of spermatogonia in the apex anterior to the compactly coiled spermatic tubule. Snodgrass (1959, p. 78) stated that "Each testis, however, appears in its entirety to be a single testicular tube. The same is true of the testes in other Diptera." In the posterior end of the testis and in the seminal vesicle I saw what appeared to be packets of spermatids or spermatozoa. The accessory gland appeared dense and had a reniform contour. The accessory gland ducts joined the ejaculatory duct where it is convoluted anteriad between the testes. Posteriorly the ejaculatory duct is looped over the rectum as a result of pupal circum- version. Between the rectum and the phallapodeme, the ejaculatory duct is noticeably dilated for some distance before it passes into the posterodorsal region of the basiphallus. I located the apparent gonopore but, as Snodgrass (1935) inferred, the gonoduct probably opens into an endophallus, the external opening of which is the apical phallotreme instead of the true gonopore. The internal female genitalia (Figures 2, 133) consist of ovaries, lateral oviducts, common oviduct, genital chamber, median spermatheca, two lateral spermathecae, and two accessory glands. Each ovary is composed of several polytrophic ovarioles. Each of these consists of a short terminal filament, a long egg tube, and a very short pedicel connecting with the lateral oviduct. Histologically, the egg tube is com- posed of a very short germarium of oogonia and a long vitellarium of follicles. Each immature follicle contains cystocytes, or follicle cells, trophocytes, and an oocyte. The basal follicle contains a large oocyte, complete with yolk and chorion. Covering each ovariole is an epithelial sheath which seems to be continuous with the terminal filament. The expanded receptacular area of fusion of pedicels and lateral oviduct is the calyx. I found a peritoneal sheath over the ovary as shown by Snodgrass (1935) in Rhagoletis pomonella and by Miller (1950) in Drosophila. It is very delicate and is apparently the ovarian suspensorium, or suspensory ligament. Sturtevant (1925, 1926) first illustrated and de- scribed the spermathecae and accessory glands of Hydrellia in his comparative morphological study of these structures in Aschiza and Schizophora Acalyp- tratae. To the lateral spermathecae, which arise dorsally from the junction of common oviduct and genital chamber, Sturtevant was inclined to ascribe a glandular function. He illustrated the accessory glands, or parovaria, as arising just posterior to the lateral spermathecae and showed them subequal to the lateral spermathecae in H. griseola. Sturtevant (1926), unlike Snodgrass (1935) and Imms (1957), distinguished the median spermatheca as basically different from the lateral spermathecae and called it the ventral receptacle. Regarding the function of the ventral receptacle, Sturtevant (1926, p. 11) stated: "Sperm have been found in this organ [median spermatheca] in Dimecoenia, Discocerina, Hydrellia, llythea, and Philygria. In no case in this family [Ephydridae] NUMBER 6 8 15 have any sperm been found in any other part of the female reproductive system." He did not ascribe a specific function to the accessory glands, but I sur- mise that they or the lateral spermathecae secrete a lubricant and adhesive for the egg as it passes from the genital chamber through the vulva. The bases for this supposition are several observations of eggs protruding from the vulva ventral to the cerci and the fact that eggs are cemented to the oviposition substrate. Morphology of Immatures EGO.?Hydrellia eggs exhibit some variability in shape, condition of the chorion, and in size. The size range of known eggs is 0.40 by 0.12 to 0.71 by 0.22 mm. In dorsal view the eggs are usually sub- fusiform, or cucumiform, with fairly symmetric con- tours except on the ends (Figures 75, 77, 78). In lateral view, their shape is often boatlike, with the ends upcurved almost symmetrically (Figures 73, 74, 125). The micropylar end is often somewhat more acute than the opposite end. In all known eggs of Hydrellia the chorion is corrugated or rugulose, with alternate longitudinal ridges and channels. It is not known if these ridges, or cristulae, have an adaptive value or are simply incidental follicular impressions. It is possible that the ridges serve as egg guides to maintain some necessary orientation of the egg in the common oviduct and genital chamber or that they help maintain chorion shape. The identification of the micropylar protuberance is presumptive, since Berg (1950), Kato (1955), and Grigarick (1959) offered no conclusive proof such as direct observa- tion of sperm entry or sperm tracing. The nature of the globose, lacunose, pluglike struc- ture on the end opposite the micropylar protuberance is still unknown. Whether the chorion is unilaminate or multilaminate is also unknown. LARVA.?Basically, the three mobile larval instars differ little morphologically. The most conspicuous developmental changes occur in the feeding appara- tus and the spiracular peritreme. Von Frauenfeld (1866), Stein (1867), Gercke (1879, 1882, 1889), Brocher (1910), Keilin (1915, 1944), Hennig (1943, 1952), Berg (1950), Kato (1955), Nye (1958), and Grigarick (1959) contributed to the description and analysis of these changes. The first instar appears setulose and warty and is about 0.35-0.75 by 0.10-0.15 mm when newly eclosed. It changes rapidly so that at the end of the stadium it is similar to depictions in Figures 124 and 127 and measures about 1.00-2.25 by 0.18-0.25 mm. I have illustrated the changes in the feeding apparatus in Figures 93, 103, and 105 for the three mobile larval instars of H. spinicornis Cresson. The conspicuous differences are in size, color, and clypeal arch contour. In Figures 81, 84, and 85 I have illus- trated the morphological nomenclature of the feed- ing apparatus, or the so-called cephalopharyngeal skeleton. I collated the nomenclature of the feeding apparatus from Berg (1950), Hollande et al. (1951), Frick (1952), Snodgrass (1953), Downes (1955), and Allen (1957a, 1957b). From paired symmetrical parts as shown by Berg (1950) in Notiphila loewi Cresson, the feeding apparatus has evolved in Hydrellia to the single, partially fused state. The two mouth- hooks have become entirely fused, the posterior or proximal ends of the H-shaped sclerite (called hypostomal sclerite by Berg, 1950) and the anterior ends of the paraclypeal phragmata have become fused, and the cross piece lost in all species studied. The paraclypeal phragmata are closely apposed and are united dorsally by the frontoclypeal plate anteriad to the bottom of the clypeal arch (area of inclination near the usually distinct cheliform spot). Ventrally, they are united by the sitophore, or hypopharyngeal plate. Each paraclypeal phragma is bifurcated pos- teriorly into dorsal and ventral phragmatal rami between which the cibarial dilator muscles can be seen stretching from the dorsal rami to the epipharyn- geal clypeal wall (dorsal cibarial wall). Anterior to each phragmatal bifurcation is a protuberance, which, because of its shape, Berg (1950) termed the cheliform spot. From my dissections, it is evidently a process for muscle insertion. Supported by unillus- trated observations of feeding mechanics and some dissections, I hypothesize that the fossae of the cheli- form spots serve as insertions for protractor muscles of the feeding apparatus, as does the prominent clypeal arch, or dorsal angle of Allen (1957a), illus- trated by Berg (1950) in Notiphila, Hollande et al. (1951), Frick (1952), Snodgrass (1953), and Downes (1955). Retractor muscles insert on the rami of the paraclypeal phragmata and rotators on the phrag- mata below the cheliform spots. The primary factor underlying the reduction of 16 SMITHSONIAN CONTRIBUTIONS TO ZOOLOGY the clypeal arch and probably the entire stream- lining of the feeding apparatus in Hydrellia has been selective pressure for manipulative flexibility and for an anterior end capable of extreme flattening during feeding. The selecting factor, or at least the factor compatible with this anatomical modification, has been leaf mining?feeding internally in very thin submergent leaves such as in most Potamogeton. Metapneustism and perhaps some kind of cuticular respiration opened the way to obtaining oxygen from the aerenchyma of certain plants, and this in turn provided the opportunity to move deeper into plant tissues. But before this opportunity could be utilized, manipulative flexibility of the feeding mechanism had to be increased greatly to allow for feeding in very tight places. Much of the increase in rotatability of "head" and feeding apparatus has been due to the reductive fusion of the two mouth-hooks and of the sclerites of the paraclypeal phragmata and to the overall compression of the feeding mechanism. All of these modifications have enabled the larva to rotate the head lobe and feeding apparatus 90 de- grees either way from the vertical attitude. Anteroventrally on the feeding mechanism are two projections which Berg (1950) called ventral projections. Hollande et al. (1951), Frick (1952), Snodgrass (1953), and Allen (1957a) showed the larger, posterior one of these to be the aperture of the common labial gland (salivary) duct. After several dissections, I could concur with the findings of these authors. I have called the anterior projection the labial sclerite, although there is some reason to think it is a remnant of the H-shaped sclerite and is involved with the mouth-hook depressor apodeme in depression of the mouth-hook. It became obvious after making several dissections that the labrum, contrary to Snodgrass's (1953) illustration, is noticeably anterior to the hase of the mouth-hook. This is obvious from Figures 81 and 84, where the fusion of sclerites can be seen to elimi- nate any space for the labrum in the position shown by Snodgrass. Nye (1958) showed the labrum as bilobate, with a lobe on each side and above the beak of the mouth-hook in H. incana. The tracheospiracular system of Hydrellia is meta- pneustic, but Keilin (1915, 1944) demonstrated a vestigial, closed prothoracic spiracular atrium. Ac- cording to Hennig (1943, 1952) there appears to be a trend toward functional metapneustism in the Ephydridae. Developmentally, the caudal tracheo- spiracular siphon changes considerably between the first two larval instars of several species at least. I have illustrated (Figure 80) a spiracular condition which may be representative of the first-instar larvae of many species of Hydrellia. The peritreme is con- ical with a terminal aciculous spine, the spiracular atrium narrow, and the secondary atrial orifice medi- ally located. The spiracular peritreme is probably nonretractile. On abdominal segment 8 are supra- spiracular and subspiracular protuberances: the supraspiracular protuberance has one long spinous seta, and the subspiracular one several shorter setae. I did not discover the function of these setae and spines, but they may possibly aid in locomotion and anchorage. Although Figure 79 is of a different species than Figure 80, it is representative of the change after the first molt. The conical peritreme is larger and less aciculate, and the secondary atrial orifice is dorsally situated. The spiracular atrium and the primary tracheal orifice at the junction of atrium and the main longitudinal trachea are discernible (Figures 79, 83, 127). In species where supraspiracular pro- tuberances and conspicuous spinous processes occur in the second-instar larvae, these are lost at the second molt. According to Hennig (1943, 1952), the third- instar spiracular peritreme has three secondary atrial orifices; however, since he did not make an extensive survey, this may not be a generic character. Grigarick (1959) illustrated one secondary atrial orifice per peritreme in the second instar as in Figures 79, 83. Keilin (1915) did not specify the instar but he illus- trated a single, dorsal secondary atrial orifice in each peritreme in H. modesta Loew. Kato (1955) also showed a single, dorsal secondary orifice. In the third-instar larvae, the increased promi- nence of the creeping welts makes it possible to discern the eleven apparent segments and a head lobe. I have shown the head lobe bearing apparently three-segmented antennae (Figure 81). Immediately posterior to the head lobe, the dorsal and ventral head folds of Snodgrass (1953) can be seen. I have followed Hollande et al. (1951) in calling the setu- lose ventral fold the postlabial pad. Berg (1950) called it the postoral tuft. Although Nye (1958) gave it no specific designation, he showed i* as the anteromedial part of the prothoracic venter, as did NUMBER 6 8 17 Hollande et al. (1951) in a nonephydrid species. Nye considered the postlabial pad a constituent of the facial mask, along with the pair of two-segmented antennae, anterior cephalic papillae, a pair of frontal papillae, maxillary palps, and two labral lobes. Also Nye definitely showed two-segmented antennae, whereas Berg (1950) and Kato (1955) illustrated three-segmented antennae. Grigarick (1959, p. 10) stated, "Antennae (Fig. 11) two-segmented, set on a small enlargement which may be a third seg- ment . . . ." Obviously, the question will be difficult to resolve even with dissection because of the small size of the structure. Keilin's (1944) concept that all dipterous larvae have three thoracic and eight apparent abdominal segments can be applied to Hydrellia, but with some difficulty, for all known larvae exhibit a postanal extension of abdominal segment 8 in the third instar and puparium, which appears to be a ninth segment. According to Nye (1958), tracheation and innerva- tion prove it is not a true segment. As inferred by Hinton (1955), the creeping welts are homologous with prolegs. These creeping welts and specifically variable portions of the segments are covered by microcuticular processes called spinules by Berg (1950) and Hollande et al. (1951), loco- motory spinules by Hinton (1955), tubercles by Allen (1957b), and spiculi by Nye (1958) and Grigarick (1959). Because the term is more descriptively flex- ible than either tubercles or spiculi, I have called these microcuticular processes spinules. They are seldom truly needle-shaped, or spiculiform, nor are they often truly tuberculiform in Hydrellia. Accord- ing to Hinton (1955), the large locomotory spinules are similar in origin, structure, and function to the crochets of most dipterous prolegs. Allen (1957b) and Nye (1958) placed consider- able significance on tubercular (or spicular) zones and the number of processes and rows in each. But as Berg (1950) illustrated, and as I have shown in Figures 107-123, in the puparia the shape and size of creeping-welt spinular zone in relation to the lateral spinular pattern is the more readily usable taxonomic character in Hydrellia. Anterior to each creeping welt, Berg (1950), Grigarick (1959), and I found a transverse row of six setulae. Keilin (1915) and Nye (1958) showed only four setulae in this location in H. incana. Both of the latter also showed no cheliform spot on the feeding apparatus. Perhaps it was rudimentary or so uniformly sclerotized as to be indistinguishable. The spinulosity and setulosity of the third-instar larva often are less distinct than in the puparium because the cuticle of the former is more translucent than that of the latter. The trans- lucency is such that the living larva in situ appears green or yellow, depending upon the combined effect of labial gland color and aliment color. In some species, the fat body appears to retain some plant pigments. PUPARIUM.?Although it is actually part of the third instar, the puparium is a distinct morpholog- ical phase. The changes involved in the formation of this uniquely economic "cocoon" are physiolog- ically complex. Keilin (1944) used the term pupario- genesis to describe this formation. An equally satis- factory term and one more readily usable verbally is pupariation. In Hydrellia, the third-instar larva becomes quiescent and opaque just as at the first two molts, and within a few hours the cuticle has contracted lengthwise, expanded transversely, and hardened. For two species I found dimensional changes from third-instar larva to puparium to be from 5.53 by 0.60 mm to 4.00 by 1.06 mm and 6.50 by 1.22 mm to 4.75 by 1.60 mm. The puparial shape varies specifically from subcylindrical to fusiform. Some are noticeably attenuated posteriad and the puparia of four species have almost symmetrical ends in ventral view. In lateral view, the puparia are cyphosomatic. Conspicuous structures other than the creeping welts are the head-lobe scar and the anal plate (both ventral) and the dorsocephalothoracic cap (Figure 107). A better name for this latter structure is oper- culum, for it is simply the dorsal thoracic section of the puparium delimited by a continuous ecdysial cleavage line. There is nothing cephalic about it. Regarding the head-lobe scar, some authors such as Brocher (1910) called it the mouth scar, but TragSrdh (1903, p. 33), in discussing pupariation in Ephydra riparia Fallen, stated: "Das Verschliessen der vorderen Offnung erfolgt in der Weise dass die Larve den Kopfabschnitt so tief hineinzieht, dass eine tiefe trichterformige Einstiilpung der ventralen Seite des Prothoracalsegmentes gebildet wird. Dieser Trichter zieht sich im proximalen Teil zusammen und erhartet stark (Fig. 6, Taf. 1) zu einem schwar- zen, in das Puparium hineinragenden zapfenformigen Gilde." This complete retraction of the head lobe 18 SMITHSONIAN CONTRIBUTIONS TO ZOOLOGY upon pupariation seems to occur also in Hydrellia; thus, I have referred to the anterior end of the puparium as prothoracic. According to Wigglesworth (1956), in third-instar larvae of higher Diptera the epidermis initiates sclerotization of the soft endocuticle, making it chemically identical to the exocuticle in the pupa- rium, except in aquatic larvae where lime deposition is substituted for this endocuticular sclerotization. Because of their endophytic environment and per- haps other factors, Hydrellia puparia do not exhibit any apparent endocuticular lime deposition (Figures 126, 128, 131, 132). The translucency of the puparia of some species, e.g., those in Figure 131, seems to make the term "tanning" seem somewhat inappro- priate. The puparial cuticle is sclerotized, but dark- ening seems to vary ecologically and perhaps genet- ically. Hering (1950) observed that the prevalent translucency of Hydrellia puparia makes them some of the most suitable of all Acalyptratae for inves- tigating pupation. Ecology Ecological Role and Distribution Adult Hydrellia are polyphagous, but since the larvae are endophytophagous and since they consume much more food than the adults the species can be classified trophically as primary consumers. It is probable that this classification will be found to apply to all Notiphilinae. No species of Hydrellia has been studied suffi- ciently to integrate it completely in a food web, but food chains involving some species are known. Larvae of H. griseola parasitize hydrophytes of many species, primarily grasses. Adults of this fly prey upon very small insects and upon insects trapped in the surface film. Among these are specimens of Hydrellia (in- cluding Hydrellia griseola), Psyllidae, early-instar Ephemeroptera and Odonata, Collembola, Braconi- dae, several species of nematocerous Diptera, Musca autumnalis, and Sarcophaga sp. Adults of some spe- cies prey actively on Collembola. Yeasts and similar fungi, Cyanophyta, Chrysophyta, nectars, and leaf epidermis are other dietary items of adults. Sever- al species of lycosid and ctenid spiders, species of Ochthera (Ephydridae), and species of Lispe and Hydromyza (Muscidae) prey upon adults of Hydrel- lia. The chain continues through fishes, e.g., bluegills and sunfishes, to ichthyophagous reptiles, birds, and mammals. A side chain exists for the larvae. Several hymenopterous species parasitize the larvae. From these parasites, the energy flows through most of the same species as it does from the adult flies. Similar food chains exist for H. cruralis and H. pulla. Host-plant preferences of the larvae likely influ- ence the ecological distribution of Hydrellia species. Predominantly, these preferences center on species of Potamogetonaceae and Gramineae. Other prefer- ences include species of Alismataceae, Cyperaceae, Hydrocharitaceae, Lemnaceae, and Liliaceae. Mines of Hydrellia in dicotyledonous plants of Caryophyl- laceae, Labiatae, Chenopodiaceae, Scrophulariaceae, and Compositae apparently serve only for puparia- tion, and as such they are relatively short and simple. Apparently, larvae that leave their old mines just before pupariating have little host preference. I found H. ischiaca puparia in the floating leaves of Potamogeton natans growing near Zizania aquatica, the preferred host plant of the locality, and H. bergi puparia in leaves of Z. aquatica growing near a preferred species, Potamogeton richardsonii. Nastur- tium officinale (Cruciferae) is the one established exception to the general absence of feeding mines in dicotyledons. European authors reported certain Hydrellia species parasitizing N. officinale, and I have confirmed H. griseola as a fairly common miner feeding in plants of this species in some localities in the United States. The known host plants, including those serving only for pupariation, are listed in Table 1. Hydrellia play an important role in many aquatic ecosystems, especially eutrophic ones. Lange et al. (1953) estimated that H. griseola destroyed from 10 to 20 percent of the California rice crop in 1959. At Lake Itasca, Minnesota, I found that the infestation of several stands of wild rice, Z. aquatica, by H. ischiaca ranged from 33 to 89 percent in 1963. No one has measured the effects of the mining of many Hydrellia species in several Potamogeton species, but they are obviously considerable judging from observed leaf damage in many stands at several localities. The same can be said of many other host plants. Hydrellia larvae function, together with some Aphididae, Delphacidae, donaciine Chrysomelidae, Trichoptera, Lepidoptera (especially Nymphula spp.), Chironomidae (Cricotopus, Polypedilum, NUMBER 6 8 19 Glyptotendipes, and Tanytarsus spp.), Hydromyza confluens, and numerous other phytophagous insects in producing subtle, and sometimes conspicuous, changes in the littoral macroflora. Temperature and humidity requirements of adult and immature Hydrellia as well as requirements of host plants influence the effect and extent of mining. Wind and wave action and temperature and humidity tolerance tend to restrict oviposition to sheltered, eutrophic habitats. In such habitats, low mineral availability, low temperatures, wind, extensive algal and Lemna mats, and excessive water depth make some species of plants more susceptible to heavy infestation. The newest plant growth is generally most susceptible to infestation and is most severely damaged. Berg (1949) found 26 insect species in addition to those of Hydrellia living upon or in 17 species of Potamogeton. Most of these insects may not have to compete extensively with one another because of the abundance of the plants and of narrowly defined ecological roles. There is obviously interspecific com- petition with aquatic caterpillars, snails, phytopha- gous fishes, waterfowl, muskrats, beavers, deer, and moose. According to Fassett (1960), bluegills eat the leaves of several Potamogeton species and Vallisneria americana; trout eat parts of Nasturtium officinale and Zannichellia palustris; waterfowl and muskrats use several Potamogeton species and Sagittaria latifolia tubers as a staple; waterfowl, muskrats, deer, and moose feed extensively on Zizania aquatica; and ducks use Echinochloa species as a staple. Parasitological Data Burghele (1959a) stated: "It is a well established fact that all aquatic Hymenoptera are parasites, the female laying her eggs in the eggs or larvae of aquatic insects." This cannot be entirely true, for I observed a specimen of Chorebus aquaticus oviposit- ing repeatedly for 5 minutes in a Z. aquatica leaf harboring neither eggs nor larvae of Hydrellia. Also, I watched a submerged specimen of Trichopria columbiana oviposit in leaf tissue around an empty puparium of H. pulla. One of my observations, in which Chaenusa sp. parasitized three or four puparia of H. ischiaca reared from eggs, indicates that some eggs or larvae of parasites may enter the host larva through the gut, for adult hymenopterans had no access to these larvae. Grigarick (1959) believed Pnigalio sp., Sympiesis sp., and Solenotus inter- medius to be external larval feeders because they pupated free of the host. He discovered abundant eggs of S. intermedius and many larvae in the mines of H. griseola. Some larvae fed externally on H. griseola larvae. Parasitic Hymenoptera undoubtedly exert consid- erable control on population densities of Hydrellia, especially in certain marginal habitats and when population densities are very high. Grigarick found parasitism of H. griseola by external parasites re- peatedly higher in pools that were drying or very low. In one sample the parasitism was 60 percent. He found hymenopterous parasitism low on the first host generation mining rice but rising rapidly on succeeding generations to nearly 90 percent by July. Opius hydrelliae and Chorebus aquaticus (Braconi- dae) were most abundant. After July, parasitism declined and remained low for the rest of the year. This seemed to be correlated both with a great decrease in host density and with less vulnerability of Hydrellia larvae to parasitism when in certain restricted survival habitats, e.g., fall rain pools. I found 38 percent parasitism in 132 puparia of H. ischiaca and 63 percent in 61 puparia of H. pulla collected through one summer. Hormone concentration and balance and perhaps other physiological factors may stimulate develop- ment of endoparasitic larvae, for in several observa- tions they began to feed actively on host tissues only after host pupariation. I have photographed an opiine larva situated in the host larva (Figure 124). After one host pupariated, I saw the leechlike larva of Ademon niger devour the host tissues in a matter of hours. When the parasite pupates, it usually orients its head toward the prothoracic end of the puparium (as in Figure 131). At emergence, it usually leaves a dark meconium in its pupal exuviae located in the puparium (as in Figure 128). I observed several escape exits cut anteroventrally in the puparium (as in Figure 132) and a few postero- ventrally. I have observed only one case in which the hymenopteron used the puparial operculum and did not cut its own exit. According to Burghele (1959a), the Dacnusinae cut the escape exit with their mandibles, while the Chalcidoidea make the exit by shaving off tiny pieces of the puparium. Inflow of water would not necessarily kill the emerg- ing adult hymenopteron for Ademon niger and 20 SMITHSONIAN CONTRIBUTIONS TO ZOOLOGY Trichopria columbiana, at least, can remain sub- merged for several hours. I noted four adults of the latter species to live submerged for 24 hours. Two of these had already emerged and then submerged again. The known hymenopterous parasites of Hydrellia are listed in Table 2. In addition to these parasites, species of Stigmatomyces (Laboulbeniales) parasitize Hydrellia adults and larvae. This fungus apparently killed several larvae of H. bilobifera during labora- tory rearings. Dispersal and Zoogeography Most adults of Hydrellia locomote by a combination of walking and short, hopping flight. Flight through a distance of a few meters occurs in a zigzag pattern. Of all the species, H. griseola most probably has the longest flight range because it has the widest dis- tribution and the largest wings. Perhaps its large wings and flight behavior often subject it to upward wind currents, for it is the most common species of Hydrellia captured in the few aerial surveys per- formed thus far in North America. Dispersal via passive dissemination by high-altitude atmospheric currents is probably efficient in most cases only over a few hundred miles because of body water loss. Strong wind often causes adults to congregate at the downwind end of lakes and pools. Some dispersal in most species probably occurs over relatively short distances in or on plants caught in water currents. Immature instars are particularly susceptible to this passive dissemination. The phenomenon probably accounts for the presence of larvae in plants at depths of from 3.0 to 6.0 meters in some lakes. Wind action and currents may carry floating puparia for variable distances. I have collated the following data for zoogeo- graphic consideration: (1) the known host prefer- ences center on plant species having their greatest densities mostly between latitudes 25? and 65? north; (2) H. griseola and probably many other species have a temperature tolerance of 50?-90?F for adults and larvae; (3) high temperatures lower the surface ten- sion of water and thus probably adversely affect locomotion of adults and migrating larvae; (4) most of the known species are Holarctic, the greatest num- ber being in the Nearctic Region; (5) H. griseola has the most extensive distribution, being reported from all zoogeographic regions except the Oriental and Australian Regions; (6) H. griseola has the greatest host-plant range within the genus. Behavior EMERGENCE.?The pharate adult must do more to emerge than simply shed the pupal cuticle; it must escape from the puparium and from the plant tissue holding the puparium. Pharate adults start making slight movements as early as twelve hours before emergence. They use their ptilina to open the oper- culum of the puparium. Initially, the operculum remains attached posteriorly because the ecdysial cleavage line extends only around the front and sides. Once the fly gets its head through the opening it pushes with its proboscis against the tip of the puparium until its fore legs are free. After this, it rapidly leaves the puparium. To emerge completely, adults of most species have only to force their way through thin, and often partially decomposed, plant epidermis and then rise in an air bubble if the puparium was submerged. However, adults of spe- cies pupariating in stems and rhizomes sometimes fail to escape and die. This can occur also in H. griseola and other species mining grasses if the leaves dry and become impliable. According to Grigarick (1959), flies emerged several times during the daylight and darkness at between 50? and 90?F. After emergence, flies of at least several species walk around slowly for several minutes, stopping period- ically to clean their bodies and evert and then with- draw the ptilinum. The wings remain folded until about 15-30 minutes after emergence, when they rapidly expand. Often, they cannot fly until 45 min- utes after emergence. FEEDING.?Adults often search over the surface film for entrapped insects and sometimes mistakenly pounce on inanimate floating objects and manipu- late them with their fore tarsi as they do with small insects. They often protect their catches by making short rushes toward intruders. Adults have labellar canalicular teeth, which apparently function carnas- sially in lacerating conjunctival membranes of trapped insects. According to Berg (1950), adult H. cruralis chew small holes in floating leaves of several Potamogeton species. They probably use their labellar teeth for this chewing as well as for loosening yeast, Chrysophyta, and other periphytic micro- NUMBER 68 21 organisms. Adults of several species often congregate within corollae of flowers of certain aquatic plants, e.g., Nymphaea odorata, Nuphar advena, and Ranunculus longirostris. Some species exhibit pecu- liar behavior while feeding, e.g., H. biloxiae rhyth- mically pushes its body up and down and H. ischiaca sometimes touches its food-insects with its abdomen. Hydrellia biloxiae and H. pulla protect their catches of dead or dying arthropods by partially extending their wings and making short rushes to drive away intruders. MATING.?Dahl (1959) distinguished six phases in the mating process: (1) initiation, in which the male approaches the female from the front or side; (2) posturing, in which the male scissors his wings, curves his abdomen, or assumes some other distinc- tive posture, often while moving back and forth before the female; (3) restimulation, in which the male repeats his posturing to secure the female's full attention; (4) mounting, in which the male climbs on the female and grasps her partially extended wings; (5) insemination, in which the male inserts his phallus and ejaculates, often while titillating the female's abdomen with his hind legs; (6) dismount- ing, in which the male disengages and the female pushes at him with her hind legs. Although I have observed all of these phases in the mating of Hydrel- lia, I found the terms epigamic and gamic more flexible in describing mating behavior because of variations in different species. One pair of H. nobilis copulated without exhibiting any apparent epigamic behavior. Another male H. nobilis repeatedly climbed on this copulating pair and exerted his phallus each time. Epigamic behavior in many species consists of wing scissoring and walking to and fro before the female, but males of H. definita rhythmically push their bodies up and down, and H. bergi, H. surata, and H. columbata touch their faces or antennae before mounting. The mechanisms of species and sex recognition remain obscure, but surfaces reflecting ultraviolet rays may constitute at least one mechan- ism. Highly reflective objects such as automobiles, clean water surfaces, and polished aluminum pans attract several insect species, especially Diptera and Hymenoptera. One could hypothesize that the pre- dominant silvery or yellow facial pruinosity in Hydrellia (and other schizophorans) reflect consid- erable ultraviolet radiation. Nickel and silver reflect about 40 percent of radiation with wavelengths of 251 millimicra, and the percentage reflected should be greater in the normal ultraviolet band of 292-400 millimicra. Density patterns of the pruinosity may be important if this hypothesis is valid. Several authors, e.g., Milne and Milne (1959), have emphasized the importance of ultraviolet radiation in insect optics. Mechanisms of sex and species recognition some- times fail, as in cases where a male H. nobilis tried to mount a male H. ischiaca after repeatedly scissor- ing his wings, or a male H. nobilis tried to mount a female H. ischiaca, or a female H. amnicola at- tempted to mount another female of the same species. Some cases of agonistic behavior are difficult to distinguish from those involving failure in sex or species recognition. In H. bergi, small males scissored before large males in two instances, and the large males grasped them by the thorax as if to mount. However, most observed agonistic behavior was interspecific, especially between Hydrellia species and predators such as spiders, Lispe spp., and Ochthera mantis. Adults of several Hydrellia species made short rushes toward these predators, often with their wings partially extended. Copulation time varies specifically and with sev- eral other conditions. A pair of H. ischiaca copulated for 10 minutes and a pair of H. nobilis for 19 min- utes. Hydrellia griseola have copulated from 1 to 50 minutes and at various times during daylight be- tween 55? and 90?F. Copulation between the same pair may occur repeatedly in one day and, in H. griseola, for five consecutive days (according to Grigarick, 1959). However, continuous mating is unnecessary, for a female H. griseola has laid viable eggs 93 days after being isolated. Grigarick (1959) found the shortest time between emergence and copulation for H. griseola of the same age was three days, but Berg (1950) reported that H. cruralis copulated 24 hours after they emerged. In the lab- oratory, pairs of H. griseola have copulated as late as 70 days after emergence (Grigarick, 1959). Pairs may copulate in various types of microhabitats, but most frequently on pleustonic vegetation. One pair of H. cruralis copulated while skimming over the surface film, and (according to Grigarick) pairs of H. griseola copulated on emergent vegetation. In all species observed, the female continues to walk and occasionally to feed during copulation. OVIPOSITION.?When forced by environmental conditions, females oviposit on nearly any surface. 22 SMITHSONIAN CONTRIBUTIONS TO ZOOLOGY In the laboratory, they will oviposit on glass, and in the field they will oviposit on dead stems, twigs of shrubs, and several hydrophytes that are not their hosts, e.g., Nelumbo lutea, Polygonum scabum, Equisetum sp., and Sagittaria sp. Most Hydrellia females can apparently distinguish preferred hosts when they are available. Usually, they prefer to lay their eggs in at least partially concealed places, such as beneath folded leaf margins and on the adaxial surfaces of sepals and stipules, but females of the H. griseola group lay their eggs on the exposed leaf blades of grasses close to the water surface. Most species deposit eggs in irregular masses, but some, notably H. spinicornis, scatter their eggs over the leaf. It is not uncommon for females to lay eggs on top of another's eggs. If females oviposit on linear objects, e.g., leaf blades of grasses and various stems, they align their eggs with the long axis of the object. Hering (1951) reported that females of some Hydrellia species submerge to oviposit. I have not observed this nor have I seen any other report of it. ECLOSION.?After about two to six days of incuba- tion, the pharate larvae often begin to make slight movements several hours before eclosion. They slit the micropylar end of the chorion with their mouth- hook and either crawl entirely out of the chorion or only protrude their head lobe sufficiently to begin excavating a mine. Sometimes the larvae remain in the ruptured chorion for several hours. Mortality is probably high among newly eclosed larvae judging from oviposition sites quite removed from the host plants and from the number of recently eclosed larvae that fall into deeper water. Many larvae must eclose while submerged because of water level fluctuations. Those eclosing from eggs on the adaxial surface of floating leaves usually avoid direct sunlight by crawl- ing onto the abaxial surface. Larvae locomote on the surface by anchoring their mouth-hook and con- tracting the body segments and by pushing with their spiracular peritremes and creeping welts. Par- tial desiccation stimulates larvae, especially newly eclosed ones, to twist and roll. This emergency re- action often carries them into water, but once there they may sink into an area devoid of plants. MINING.?When the newly eclosed larvae find an area they can penetrate because of either weakened or thinner epidermis or some other condition, they first break the epidermis by striking it perpendicu- larly with their mouth-hook and then extend the opening by rotating the head lobe and mouth-hook and slicing the epidermis. The larvae ingest the loosened tissues. The prothoracic end flattens down to little more thickness than that of the feeding apparatus to project the mouth-hook into the ex- tremely thin mesophyll. By continuing this excava- tion process, the first-instar larvae completely enter their mines in 2 to 2.5 hours. Third-instar larvae require only about half this time. The initial opening does not have to be as broad as the larva because of segmental contraction. Inside the mine, larvae move forward by peristalsis, in which contraction waves pass anteriad through the creeping welts, and occasionally by pushing with the spiracular peritremes and pulling with the mouth-hook. They can move slowly backward by reversing the peri- stalsis. Also, even late third-instar larvae can turn about within the mine. During endophyllous loco- motion, the spinules of the dorsum and creeping welts press against the upper and lower epidermis and sometimes penetrate it. Mining larvae apparently depend on the host plant for oxygen. According to Hering (1951), frass and other debris seal the mine entrance and prevent the entrance of water. Air then enters through the plant tissue and accumulates in the mine around the larva. This does not always happen, for I found larvae in old mines partially filled with either water or tissue fluid, and I observed larvae of H. griseola, H. ischiaca, and H. bilobifera pierce several different tissue areas in their mines with their spiracular peritremes until they found an apparent oxygen source. Larvae crawling on the underside of the surface film or starting a mine just below the surface film were observed periodically raising their spirac- ular peritremes above the surface film. Observations by Aldrich (1912) and Nemenz (1960) indicate a strong possibility that some kind of plastron or cuticular respiration operates in Hydrellia larvae. Nemenz (I960, p. 222) showed in Ephydra that "die Larven gelegentlich eine Luftblase aus den Abdominalstigmen auspressen und wieder einziehen, die offenbar als 'physiologische Kieme' wirkt." Mines of first-instar larvae often are inconspicuous because of the small larvae and also, perhaps, be- cause of partial plant tissue repair. As the larva grows, the mine is enlarged, but it may remain inconspicuous if it is in the middle of a relatively thick leaf, e.g., floating leaves of several Potamogeton NUMBER 6 8 23 species. Feeding mines of first- and second-instar larvae in narrow leaves such as those of grasses and some narrow-leafed Potamogeton species are fairly straight, but those of third-instar larvae are usually tortuous and have blotch-like chambers. In broad leaves, the feeding mines of even first-instar larvae may be tortuous. Larvae eat the mesophyll and, since various conditions will cause several larvae to congregate in a single leaf, skeletonization often results. If this occurs before the larvae are ready to pupariate, they migrate to fresh leaves. Some species of Hydrellia overwinter as larvae. Larvae of H. bilobifera can survive in Potamogeton nodosus encased in ice. PUPARIATION.?Pupariation is a fairly new term for the distinct process of the formation of a puparium. It is not equivalent to pupation, since it is essentially reformation of the third-instar cutide and since a short fourth larval stadium occurs within the puparium before pupation. This fourth instar is very abortive, for a new feeding apparatus is not formed. When larvae are ready to pupariate, they become quiescent, with a greatly lowered heart rate. Puparia- tion occurs within a period of eight hours in some species. Larvae usually pupariate just under the epidermis of thick leaves. This improves the emerging adult's chances of escaping from the leaf. In thin leaves, such as those of grasses, position of the pu- parium seems unimportant. Third-instar larvae of H. bergi and H. biloxiae make escape slits for the adults in the stems and stolons that they mine. Third- instar larvae of several species often migrate to a new plant part for pupariation. This behavior probably aids survival by providing a more stable anchorage than the old, mined plant part. Also, this behavior partly accounts for the broad range of host plants of H. griseola. Apparently the larvae are nonspecific for these pupariation plants. According to Berg (1950), several species always pupariate with their spiracular peritremes anchored in the midribs of Potamogeton leaves. This fact seems to support the hypothesis that puparia must take air from the plant, but other data seem to refute this hypothesis. For instance, larvae of H. ischiaca often pupariate in skeletonized or wilted grass leaves, with their spi- racular peritremes embedded only in the epider- mis, and pupate successfully. Larvae of H. griseola usually make a blotch mine and pupariate in its center. Grigarick (1959) reported that H. griseola larvae pupariated on glass, in sand, and in blotting paper in the laboratory. Burghele (1959a) reported finding many H. griseola puparia on the bottoms of ice-covered pools. This may indicate overwintering capacity of puparia. Schiitte (1921) reported the existence of summer and winter puparia in Hydro- myza livens (Muscidae). The winter puparium has a wall eight times as thick as that of the summer one. Such seasonal forms may exist in Hydrellia. Environmental Tolerance Hydrellia are fairly stenotopic and stenohygric. The adult tolerance range for wind and temperature is relatively small. Wind often precludes oviposition on suitable host plants located in the limnetic zone of ponds and lakes, and it often causes host plants to break loose and pile up on the shore where the mining larvae have no opportunity to escape. Wind is probably one limitation to the height at which oviposition can occur on emergent vegetation. Grigarick (1959) presented the following tempera- ture limits (in degrees Fahrenheit) to adult and larval activities of H. griseola: Adult Heat death: Heat paralysis: Dropping: Heat rest: Normal locomotion: Slow walking: Heat death: Heat paralysis: Agitated movement: Normal mining: Quiescence: 110-114 107-110 104-107 98-104 52- 92 40- 50 Larva 112-114 104-111 95-103 50- 90 36- 50 In one of Grigarick's tests, 50 percent of a sample of twelve wild adults lived 34 days at 29?F, and, in another one, 20 percent of 80 eggs hatched after 120 hours of exposure to 29?F. The principal extrinsic factors affecting the length of stadia of Hydrellia are temperature, atmospheric saturation deficit, and food (composition and quantity). The last factor is also the principal extrinsic one affecting fecun- dity. 24 SMITHSONIAN CONTRIBUTIONS TO ZOOLOGY Host Plants and Parasitic Wasps The known host plants and parasitic wasps of Hydrellia are listed in tables starting on page 106. Systematic Treatment Classification The type-species of Hydrellia Robineau-Desvoidy (1830) is Hydrellia aurifacies Robineau-Desvoidy, the first known valid designation of which was made by Coquillett (1910b). Coquillett cited Notiphila flaviceps Meigen (with H. aurifacies as a synonym) as the type. According to Article 69a (iv) of the International Code of Zoological Nomenclature (1961), Coquillett effectively designated H. aurif- acies as the type-species. According to Wirth (1965), Westwood (1840) first designated H. aurifacies as the type-species, but Westwood actually designated H. flaviceps Meigen as the type-series. Since H. flaviceps was not an originally included nominal species and since Westwood did not synonymize it with H. aurifacies, his designation is invalid. Ever since 1896 (and perhaps before), Notiphila griseola Fallen has generally been considered the type-species of Hydrellia Robineau-Desvoidy (1830). Macquart (1835) first reported the synonymy of Hydrellia communis Robineau-Desvoidy (1830) and Notiphila griseola Fallen (1813). Macquart also transferred several other species from Notiphila to Hydrellia. Kloet and Hincks (1945) rejected the name Hydrellia as a homonym of Hydrelia Huebner and replaced it with Hydropota Rondani. Article 56a of the International Code (1961) specifically prohibits such rejection in this statement: "Even if the difference between two genus-group names is due to only one letter, these two names are not to be considered homonyms." The adults of Hydrellia can be distinguished from those of other Ephydridae by the following char- acters: ocular pubescence moderately dense; face usually slightly convex (distinctly receding and planate in one group) ; postocellar as long or longer than ocellar; two or three dorsocentrals present, at least one of which is as long or longer than longest facial; median spermatheca cupuliform, heavily sclerotized. Some additional characters present in adults of most species are as follows: about three palpal and five labellar setae; one very fine, ven- trally directed, secondary facial setula inserted above primary facial row; two fronto-orbitals (anteriorly and posteriorly inclined) ; apicodorsal antennals not projecting noticeably; three scutellars (basal, in- termediate, and apical); one mesanepisternal, mesokatepisternal, and basal coxal; ground color of cuticle usually dark brown or black. The larvae can be distinguished from those, of other ephydrid genera by these characters: meta- pneustic; spiracular peritreme spinous; mouth-hook single. Cresson (1944b) erected the tribe Hydrelliini solely for Hydrellia and indicated its close relation- ship to the tribes Hydrinini and Ilytheini. As char- acters for Hydrelliini, he listed essentially the generic characters. No author has disputed this classification. However, some controversy has existed on the sub- familial classification. Loew (1860) placed Hydrellia with Atissa, Philygria (as Hydrina), Hyadina, and Axysta in Hydrellina, which corresponded to a sub- family. Becker (1926) grouped Hydrellia, Philygria, and several parydrine genera in his subfamily "Hy- drellinae." Cresson (1930, 1942, 1944b), Wirth and Stone (1956), Dahl (1959), and Deonier (1964) grouped Hydrellia with Notiphila, Dichaeta, llythea, Philygria, Nostima, Oedenops, and several other genera in the Notiphilinae. I found 57 species of Hydrellia in the Nearctic Region, of which 21 are new and one is a new Nearctic record. I discovered and described im- mature instars of twelve species and redescribed those of seven species previously known. The phylogeny of this genus remains obscure con- cerning most details. The entire genus seems to be in a very active speciation phase. The absence of noticeable larval divergence within certain species groups is very likely an indication of relatively recent speciation. In the H. bilobifera species group, for example, H. bilobifera and H. trichaeta can be dis- tinguished by male and female terminalia, but the larvae are very difficult to distinguish. Where the species occur together, the larvae of each are often found in the same host-plant species and even in the same leaf of a plant. Berg (1950) and others have reported this food (host) overlap in some other Hydrellia species. The adults, however, are different enough that they do not attempt to interbreed in a laboratory colony nor, as so far observed, in natural NUMBER 6 8 25 habitats. Most of the species groups of Nearctic Hydrellia may well have in them similar situations regarding speciation. This state of the genus in the Nearctic Region has prohibited any objective segregation of its consti- tuents into morphologically discrete species groups. A tentative, subjective grouping based upon overall impressions of similarity in habitus is presented below. This corresponds somewhat with Cresson's impressions. A preliminary attempt to obtain a grouping based upon computer analysis of ratios of various measurements of male terminalia structures proved inconclusive. In this attempt, 25 ratios were established between the following characters: fused surstyli length and breadth, lateral and median lengths of fused surstyli, cercus length and breadth, copulobus length and breadth, sterna 4 and 5 breadths, lateral and median lengths of sternum 5, postgonite uncus length and thickness, cerci breadth (spread), total phallus length, length and breadth of phallus (including, in some cases, part of basi- phallus) anterior to fused surstyli, and breadth of distiphallus at midlength. The ratios represented the mean of usually ten specimens (in a very few species only one). These ratios for 53 of the species were subjected to stepwise discriminant analysis and then to clustering by the unweighted pair-group method in a computer. Each species was categorized as uncertain, present, or absent for each ratio char- acter depending upon whether the actual ratio was above or below the all-species mean of that ratio character. Eleven species groups ranging from one to ten species resulted from this program. All of these groups had a similarity coefficient (SSm) of 0.62 or greater (where a coefficient of 1 is identity). Only four of these correlated much with any of my 14 habitus groups. One group, bilobifera, was iden- tical in both groupings. Another computer group correlated fairly well with two habitus groups? morrisoni and platygastra. A computer program integrating all quantitative and qualitative charac- ters gathered for the species is needed to obtain a better generic analysis. SPECIES GROUPS.?The following represent habitus or physiognomic groups of Nearctic Hydrellia: griseola group: (1) griseola, (2) valida, (3) flanicoxalis, (4) spinicornis, (5) rixator, (6) ischiaca. notiphiloides group: (7) notiphiloides, (8) deceptor, (9) caliginosa, (10) pulla. subnitens group: (11) subnitens, (12) suspecta, (13) definita, (14) manitobae, (15) itascae, (16) amnicola, (17) agitator, morrisoni group: (18) morrisoni, (19) borealis, (20) cessator, (21) serena, (22) lata, (23) penicilli. cruralis group: (24) cruralis. tibialis group: (25) tibialis, (26) americana, (27) biloxae. formosa group: (28) formosa, (29) notata, (30) ains- worthi. crassipes group: (31) crassipes, (32) saltator, (33) proc- teri, (34) advenae. nobilis group: (35) nobilis, (36) idolator, (37) atro- glauca, (38) cavator. proclinata group: (39) proclinata, (40) melanderi, (41) dec ens. platygastra group: (42) platygastra, (43) wilburi. bergi group: (44) bergi, (45) insulata, (46) per sonata. prudens group: (47) prudens, (48) surata, (49) colum- bata, (50) luctuosa, (51) floridana. bilobifera group: (52) bilobifera, (53) gladiator, (54) discursa, (55) ascita, (56) harti, (57) trichaeta. REMARKS ON KEYS AND DESCRIPTIONS.?The high degree of homogeneity in this genus has severely hampered my construction of utilitarian keys to the adults, larvae, and puparia. The extensive use of various indices and the need to refer to male termi- nalia limit the utility of the keys. I have not always been able to group apparently closely related species in the keys. Initial use of the keys will require refer- ence to the section on definitions in the chapter on morphology. Descriptions of male terminalia pertain to the ventral view of the structures. Because of the possibility that a specimen may represent a new sibling species, one should consult the descriptions after using the keys. I studied the holotype or a syntype of each species unless stated otherwise under the taxonomic remarks. 2 6 SMITHSONIAN CONTRIBUTIONS TO ZOOLOGY Key to Adult Hydrellia 1. Ocellar present 2 Ocellar absent 50 2. Two subequal fronto-orbitals present 3 Three subequal fronto-orbitals present. (Male terminalia as in Figure 32.) H. ainsworthi, new species 3. Anterior fronto-orbital anteriorly inclined, posterior one posteriorly inclined 4 Both fronto-orbitals anteriorly inclined 51 4. Face slightly convex or protruding in profile 5 Face planate and receding at constant angle in profile 53 5. Palpus moderate yellow or moderate orange-yellow 6 Palpus at least partly dark brown or black 35 6. Fore tarsus dark brown or gray in dorsal view 7 Fore tarsus moderate yellow or moderate orange-yellow 24 7. Some tibiae moderate brown or dark gray; mesokatepisternals variable 8 All tibiae moderate yellow; 1 large and 3 small mesokatepisternals. (Male terminalia as in Figure 63.) H. cruralis Coquillett 8. Lower half of face not conically prominent; palpus moderate yellow; no sfa or 1-5 seriated sfa 9 Lower half of face conically prominent; palpus moderate orange-yellow; 4-7 scattered sfa. (Male terminalia as in Figure 61.) H. pulla Cresson 9. Antennal segment 3 dark brown, or if partly yellow then hind basitarsus without or with usually 2 black anterior basal setae (except H. cessator, with 4-7.) 10 Antennal segment 3 partly moderate yellow; hind basitarsus with 3?5 black, anterior basal setae. (Male terminalia as in Figure 66.) H. penicilli Cresson 10. Fore coxa and tibia mostly dark gray or moderate brown; hind tibia without or with usually 3-5 black, anterior apical setae 11 Fore coxa and tibia moderate yellow; hind tibia with 2 or 3 black, anterior apical setae. (Male terminalia as in Figure 44.) H. ischiaca Loew 11. One large mesokatepisternal; basal coxals variable 12 One large and 3 or 4 small mesokatepisternals; 2 basal coxals. (Male terminalia as in Figure 62.) H. jlavicoxalu Cresson 12. Thoracic pleuron and side of abdomen mostly light gray or light bluish gray, strongly contrasting with mesonotum and abdominal dorsum (except in nearly uniformly light gray H. valida) 13 Thoracic pleuron or side of abdomen mostly moderate brown or yellowish gray, not strongly contrasting with mesonotum or abdominal dorsum 19 13. Female fore femur without distinct, black, anteroventral spines on distal half; costal section I I : I 2.8 or less; mesofacial index 2.5 or less 14 Female fore femur with distinct, black, anteroventral spines on distal half; costal section II : I 3.0 or more; mesofacial index 2.6 or more. (Male terminalia as in Figure 40.) H. spinicornis Cresson 14. Large areas of body brown; usually 12 or fewer anterior interfractural costals 15 Body uniformly light gray (except mesonQtum and thoracic pleuron infrequently with sparse light yellowish brown pruinosity in small areas) ; 12?15 anterior interfractural costals. (Male terminalia as in Figure 22.) H. valida Loew 15. Antenna not velvety dark brown; frontal vitta and parafrontalia of different hues; prementum usually glossy brownish black 16 Antenna velvety dark brown; frontal vitta and parafrontalia uniformly semiglossy dark brown; prementum light gray. (Male terminalia as in Figure 42.) H. deceptor, new species 16. Body length:wing length 0.9 or more; costal section I I : I 2.6 or less; vertex index usually under 7.0 (except 7.5 in H. rixator) 17 Body length:wing length 0.8; costal section I I : I 2.6 or more; vertex index 7.0 or more. (Male terminalia as in Figure 26.) H. griseola (Fallen) 17. Frontal vitta and parafrontalia of different hues; vertex index 5.5 or less; male length usually over 1.8 mm; wing length usually over 2.0 mm 18 Frontal vitta and parafrontalia uniformly velvety dark brown; vertex index 5.5 or more; NUMBER 6 8 27 male length usually 1.7 mm or less; wing length usually under 2.0 mm. (Male terminalia as in Figure 29.) H. rixator, new species 18. Parafrontalia velvety black, contrasting with frontal vitta; 8-10 aristal rays; costal section II I : IV 2.5 or less. (Male terminalia as in Figure 19.) H. caliginosa Cresson Parafrontalia not velvety black; 6-8 aristal rays; costal section I I I : IV 2.5 or more. (Male terminalia as in Figure 65.) H. notiphUoides Cresson 19. Male and female lengths under 2.3 and 2.7 mm respectively; fewer than 12 dorsal and 14 anterior interfractural costals; costal section I I : I usually over 2.0 20 Male and female lengths over 2.8 and 2.9 mm respectively; 12-14 dorsal and 14-16 anterior interfractural costals; costal section I I : I usually under 2.0. (Male terminalia as in Figure 28.) H. itascae, new species 20. Face not silky light yellowish brown, without median crease or with one distinct in other views; apicodorsal antennal prominent only in H. lata; most of fused surstyli concealed in repose or if exposed, then not moderate yellow 21 Face silky light yellowish brown, with median crease distinct only in anteroventral view; apicodorsal antennal prominent; most of fused surstyli always exposed and moderate yellow. (Male terminalia as in Figure 17.) H. bergi Cresson 21. Prementum not glossy brownish black; antenna not velvety dark brown; costal section V:IV usually 3.5 or less 22 Prementum glossy brownish black; antenna velvety dark brown; costal section V:IV 3.5 or more. (Male terminalia as in Figure 46.) H. insulata, new species 22. Prementum pale yellow; apicodorsal antennal not prominent; costal section I I : I usually 2.4 or less 23 Prementum light gray; apicodorsal antennal prominent; costal section I I : I 2.4 or more. (Male unknown.) H. lata Cresson 23. Parafrontalia velvety black, contrasting with frontal vitta; 2 or 3 basal coxals; mesane- pisternum bronzed in posterolateral view. (Male terminalia as in Figure 35.) H. serena Cresson Parafrontalia not velvety black; 1 basal coxal (infrequently 2 ) ; mesanepistemum not bronzed in posterolateral view. (Male terminalia as in Figure 59.) H. cessator, new species 24. Abdominal terga 2-4 not velvety purplish black medially; frontal vitta, parafrontalia, occiput, and antennal segment 2 not uniformly velvety black 25 Abdominal terga 2?4 velvety purplish black medially; frontal vitta, parafrontalia, occiput, and antennal segment 2 uniformly velvety black. (Male terminalia as in Figure 34.) H. biloxiae, new species 25. Tibiae moderate brown with light gray pruinosity 26 Tibiae moderate yellow on one-third or more of their length 33 26. Antennal segment 3 at least partly moderate yellow 27 Antennal segment 3 dark brown or black 29 27. Parafrontalia not velvety black; wing length 2.4 mm or more; 6?9 aristal rays 28 Parafrontalia velvety black; wing length 2.1 mm or less; 4?6 aristal rays. (Male terminalia as in Figure 54.) H. cavator, new species 28. Ocular index 13.0 or less; 8 or 9 aristal rays; costal section I I : I 2.4 or more. (Male terminalia as in Figure 45.) H. subnitens Cresson Ocular index 15.0 or more; 6-8 aristal rays; costal section I I : I 2.3. (Male terminalia not figured.) H. suspecta Cresson 29. Tarsi moderate yellow; male hind femur without posteroventral flange and anteroventral rows of close-set, short setae 31 Tarsi moderate orange-yellow; male hind femur with posteroventral flange and 2 anteroventral rows of close-set, short setae 30 30. Antenna dark brown; ocular index 8.0 or more; 1 basal coxal. (Male terminalia as in Figure 33.) H- crassipes Cresson Antenna velvety black; ocular index 7.0 or less; 2 basal coxals. (Male terminalia as in Figure 56.) " . saltator, new species 31. Fore and mid coxae light gray anteriorly; 7 or more dorsal and 9 or more anterior interfractural costals; ocular index usually more than 9.0 32 Fore and mid coxae moderate yellow anteriorly; 7 or fewer anterior interfractural costals; ocular index usually under 9.0. (Male terminalia as in Figure 39.) H. advenae Cresson 28 SMITHSONIAN CONTRIBUTIONS TO ZOOLOGY 32. Wing length 2.1 mm or less; epistomal index 1.4 or less; male hind femur very stout and hind tibia expanded distally. (Male terminalia as in Figure 20.) H. procteri Cresson Wing length 2.4 mm or more; epistomal index 1.4 or more; male hind femur not stout and hind tibia not expanded distally. (Male terminalia as in Figure 45.) H. subnitens Cresson 33. Face moderate or pale yellow; ocular index 13.0 or more; abdominal terga without distinct bluish gray posterolateral wedges 34 Face light gray; ocular index 12.0 or less; abdominal terga with distinct light bluish gray posterolateral wedges. (Male terminalia not figured.) H. atroglauca Coquillett 34. Wing length 2.3 mm or more; vertex index 6.0 or more; ocular index usually over 17.0. (Male terminalia as in Figure 36.) H. nobilis (Loew) Wing length 2.1 mm or less; vertex index 5.5 or less; ocular index usually under 17.0. (Male terminalia as in Figure 21.) H. idolater, new species 35. Mid and hind tarsi moderate orange-yellow; posterior part of male abdomen compressed 36 Mid and hind tarsi dark brown; posterior part of male abdomen not compressed 37 36. Face velvety dark brown; antenna dark brown; 10-14 setae on basal end of costa. (Male terminalia as in Figure 38.) ?? platygastra Cresson Face white; antenna velvety black; 9-11 setae on basal end of costa. (Male terminalia as in Figure 38.) H. wilburi Cresson 37. One basal coxal; 12 or fewer anterior interfractural costals; wing length 3.0 mm or less 38 Two basal coxals; 11-14 anterior interfractural costals; wing length 3.0 mm or more. (Male terminalia as in Figure 48.) H. manitobae, new species 38. Thoracic pleuron dark yellowish gray, brown, or light gray; wing length 2.7 mm or less; 9 or fewer aristal rays; male mid and hind tibiae not both expanded 39 Thoracic pleuron light yellowish gray; wing length 2.7 mm or more; 9-11 aristal rays; male mid and hind tibiae expanded (Figure 8) . (Male terminalia as in Figure 16.) H. morrisoni Cresson 39. At least one adc much longer than others; face not light yellowish brown; apicodorsal antennal not prominent (except in H. personate); male hind tibia not expanded 40 All adc more or less uniordinate; face light yellowish brown; apicodorsal antennal prominent; male hind tibia expanded as in Figure 8. (Male terminalia as in Figure 53.) H. bore alts Cresson 40. Dorsal and anterior interfractural costals subequal; 2 or more pdc; male mid tibia variable 41 Dorsal interfractural costals twice size of anterior ones; 1 pdc; male mid tibia not expanded. (Male terminalia as in Figure 49.) H. personata, new species 41. Face not white; frontal vitta and parafrontalia not uniformly velvety black; 1 or more sfa; male mid tibia variable 42 Face white; frontal vitta and parafrontalia (except light-brown ocellar triangle) uniformly velvety black; no sfa; male mid tibia expanded. (Male terminalia as in Figure 27.) H. arnerienna Cresson 42. Mesonotum and abdominal dorsum not glossy dark grayish green; face not both light gray and slightly carinate (except in H. columbata) ; male mid tibia variable 43 Mesonotum and abdominal dorsum glossy dark grayish green; face light gray and slightly carinate; male mid tibia expanded. (Male terminalia as in Figure 41.) H. tibialis Cresson 43. Thoracic pleuron mostly light gray or yellowish gray; wing length 1.8 mm or more; male mid tibia not expanded 44 Thoracic pleuron mostly moderate olive-brown (body mostly with dense moderate or dark olive-brown pruinosity) ; wing length 1.8 mm or less (except 2.1 mm in H. luctuosa) ; male mid tibia expanded 46 44. Ocular index 7.0 or less; mesofacial index 2.0 or less; costal section V:IV 3.5 or less 45 Ocular index 8.0 or more; mesofacial index 2.0 or more; costal section V:IV 3.6 or more. (Male terminalia as in Figure 60.) H. dejinita Cresson 45. Antenna and parafrontalia velvety dark brown; 6-8 aristal rays; M1 + 2 index 1.5 or less. (Male terminalia as in Figure 47.) H. amnicola, new species NUMBER 6 8 29 Antenna and parafrontalia not velvety dark brown; 8 or 9 aristal rays; M 1 + 2 index 1.6 or more. (Male terminalia as in Figure 58.) B. agitator, new species 46. Face not entirely dark brown; usually 9 or fewer anterior interfractural costals 47 Face entirely dark brown; usually 9 or more anterior interfractural costals. (Male terminalia as in Figure 64.) B. luctuosa Cresson 47. Ocular index 8.0 or less; face partly brown (except in H. columbata), protuberant or carinate 48 Ocular index 9.0 or more; face entirely light gray or yellowish gray, slightly convex. (Male terminalia as in Figure 31.) B. floridana, new species 48. Face centrally protuberant, light gray only on lower third; vertex index 4.5 or less; mesofacial index usually 1.4 or more 49 Face slightly carinate on upper two-thirds, usually entirely light gray; vertex index 4.6 or more; mesofacial index 1.4 or less. (Male terminalia as in Figure 55.) B. columbata, new species 49. Palpus, antenna, and parafrontalia velvety black; 4-7 sfa; lower half of mesokatepisternum light gray. (Male terminalia as in Figure 25.) B. surata, new species Palpus, antenna, and parafrontalia dark brown, not velvety; 1-3 sfa; thoracic pleuron moderate olive-brown. (Male terminalia as in Figure 18.) B. prudens Curran 50. Thoracic pleuron nearly uniformly light gray; 3 fronto-orbitals. (Male terminalia as in Figure 52.) B. notata, new species Thoracic pleuron velvety brownish black and light gray; 2 fronto-orbitals. (Male terminalia as in Figure 43.) B. formosa Loew 51. Thoracic pleuron with 1 continuous light-gray area 52 Thoracic pleuron with 2 distinct light-gray areas. (Male terminalia as in Figure 57.) B. proclinata Cresson 52. Arista with 6-8 rays; 8-11 dorsal and 9-12 anterior interfractural costals. (Male terminalia as in Figure 51.) B. melanderi, new species Arista with 9-11 rays; 6-8 dorsal and 8-9 anterior interfractural costals. (Male unknown.) B. decent Cresson 53. Male abdominal syntergum 9+10 rounded or only slightly bilobate posteriorly; female cercus less than twice as long as wide 54 Male abdominal syntergum 9+10 prominently bilobate posteriorly; female cercus about three times as long as wide (Figure 72). (Male terminalia as in Figure 50.) B. bilobifera Cresson 54. Antennal segment 3 at least partly moderate yellow 55 Antennal segment 3 dark brown. (Female terminalia as in Figure 69; male as in Figure 15.) B. harti Cresson 55. Ocular index over 6.0 56 Ocular index under 6.0 57 56. Male abdominal syntergum 9+10 rounded posteriorly; female cercus mucronate apically (Figure 68). (Male terminalia as in Figure 37.) B. trichaeta Cresson Male abdominal syntergum 9+10 truncate posteriorly; female cercus acute apically (Figure 67). (Male terminalia as in Figure 23.) B. ascita Cresson 57. Wing length under 2.0 mm; female cercus ovoid apically, truncate basally (Figure 70). (Male terminalia as in Figure 30.) B. discursa, new species Wing length over 2.2 m m ; female cercus acute apically, rounded basally (Figure 7 1 ) . (Male terminalia as in Figure 24.) B. gladiator, new species Male Terminalia Key to Hydrellia [The descriptions herein apply only to the ventral view and to the sclerotized portions of structures unless otherwise stated. Only maximum dimensions are used unless otherwise stated. The copulobus length is measured from a level perpendicular through the base of the copulobus to the median point on the posterior margin of sternum 5.] 1. Phallus and cerci bilaterally symmetrical 2 Phallus and cerci bilaterally asymmetrical; copulobus about 25.0 times as long as median length of sternum 5 (Figure 17) B. bergi Cresson 3 0 SMITHSONIAN CONTRIBUTIONS TO ZOOLOGY 2. Copulobus length about 4.5 times (or more) as great as cercus breadth 47 Copulobus length about 3.2 times (or less) as great as cercus breadth 3 3. Basiphallus with black spinous process anterolaterally 4 Basiphallus without such process 5 4. Sternum 5 distinctly angular and prominent anterolaterally; posterolateral corner of copulobus acutangular (Figure 26) H. griseola (Fallen) Sternum 5 rounded anterolaterally; posterolateral corner of copulobus nearly 90* angular (Figure 22) "? valida Loew 5. Sternum 5 with rounded prominence anterolaterally; posteromedial margin of sternum 5 and its copulobi forming squared recess; distiphallus not lamellate 6 Sternum 5 without rounded prominence anterolaterally, or if so then posteromedial margin of sternum 5 and its copulobi not forming a squared recess; distiphallus variable 7 6. Fused surstyli narrowly and deeply cleft anteromedially, with anterolateral lobe, and without posterolateral spicate projection; cercus about 1.5 times as long as broad (Figure 62) H. flauicoxalis Cresson Fused surstyli narrowly and shallowly notched anteromedially, without anterolateral lobe, but with posterolateral spicate projection; cercus about 2.0 times as long as broad (Figure 44) H. ischiaca Loew 7. Posteromedial part of sternum 5 with paired, long bispinate processes directed toward anterior margin of fused surstyli; fused surstyli narrowly and deeply cleft anteromedially, with anterolateral bladelike lobe and central gibba (Figure 40) H. spinicornis Cresson Sternum 5 without such processes; fused surstyli variable, but without anterolateral bladelike lobe and central gibba 8 8. Fused surstyli with 3 narrow, deep clefts anteriorly; median third of anterior margin of sternum 5 broadly notched; copulobus with incurved posterior arm and shorter medial arm, both with macrochaetae on ends (Figure 29) H. rixator, new species Fused surstyli without or with only 1 cleft in anterior margin; sternum 5 not notched anteromedially; copulobus without such arms 9 9. Postgonite uncus truncate apically and about 4.0 times as long as thick; cerci longer than fused surstyli; copulobus sharply tapered to small tip nearly touching fused surstyli (Figure 65) H. notiphiloides Cresson Postgonite uncus acute or rounded apically, or if truncate then cerci shorter than fused surstyli; copulobus not sharply tapered, or if so then not nearly touching fused surstyli 10 10. Sternum 5 acutangular, hooklike anterolaterally; fused surstyli convex anteriorly with narrow, shallow anteromedial notch; distiphallus longitudinally lamellate (Figure 42). H. deceptor, new species Sternum 5 not hooklike anterolaterally; fused surstyli not so formed anteriorly, or if so then longer than broad; distiphallus variable 11 11. Fused surstyli about 2.0 times as broad as long and about 0.6?0.8 as long as cercus; copulobus angularly notched medially and nearly touching fused surstyli (Figure 19). H. caliginosa Cresson Without such combination of characters 12 12. Pregonite 2.5-3.0 times as thick as postgonite; postgonite uncus buttonlike, about 1.5 times as long as thick; sternum 5 noticeably broader than fused surstyli (Figure 61). H. pulla Cresson Pregonite about 2.0 times (or less) as thick as postgonite; postgonite uncus not buttonlike; sternum 5 variable 13 13. Cerci together as broad or broader than broadest part of fused surstyli and as long or longer than median part of fused surstyli 14 Cerci not nearly as broad as broadest part of fused surstyli, or if so not as long as median part of fused surstyli 22 14. Postgonite uncus projecting anteriad to or beyond distiphallus apex 15 Postgonite uncus not projecting much beyond midlength of distiphallus, or if so then it is directed laterad 17 15. Sternum 5 noticeably broader than fused surstyli; postgonite uncus 12.0-15.0 times as long as thick and hooked apically; copulobus about 4.0 times as long as broad (Figure 55) H. columbata, new species Sternum 5 not noticeably broader than fused surstyli; postgonite uncus about 4.0-6.0 NUMBER 6 8 31 times as long as thick, not distinctly hooked apically; copulobus about 2.0 times as long as broad or broader than long 16 16. Copulobus about 2.0 times as long as broad; distiphallus broadest at midlength; postgonite uncus nearly straight (Figure 25) H. surata, new species Copulobus not prominent, broader than long; distiphallus broadest proximally; postgonite uncus tip curved laterad (Figure 18) H. prudent Curran 17. Fused surstyli narrowly and deeply cleft medially along 0.9 of the length; uncus not noticeably darker than rest of postgonite; copulobus about 4.0-5.0 times as long as broad (Figure 28) H. itascae, new species Fused surstyli cleft no deeper than about 0.7 of the length; uncus noticeably darker than rest of postgonite; copulobus about 2.0-3.0 times as long as broad 18 18. Copulobus with sharp projection at midlength of medial margin; distiphallus longitudinally lamellate (Figure 58) H. agitator, new species Copulobus without such projection; distiphallus not longitudinally lamellate 19 19. Fused surstyli slightly concave anteriorly (nearly truncate); exposed part of basiphallus near fused surstyli about 4.0 times as broad as distiphallus at midlength (Figure 59). H. cessator, new species Fused surstyli indented or cleft anteromedially; exposed part of basiphallus near fused surstyli about 3.0 times (or less) as broad as distiphallus at midlength 20 20. Copulobus about 2.0 times as long as postgonite uncus; fused surstyli widely V-cleft anteromedially (Figure 16) H. morrisoni Cresson Copulobus about 5.0 times as long as postgonite uncus; fused surstyli narrowly cleft anteromedially 21 21. Fused surstyli about 3.5 times as broad as copulobus and narrowly and deeply cleft medially along about 0.7 of the length (Figure 66) H. penicilli Cresson Fused surstyli about 6.0 times as broad as copulobus and narrowly cleft medially along 0.5 (or less) of the length (Figure 53) H. borealis Cresson 22. Cercus only about 0.6 as long as broad; distiphallus nearly uniform in breadth with mucronate apex; fused surstyli broadly V-cleft anteriorly (Figure 60). H. defbuta Cresson Cercus more than 0.6 as long as broad, or if not then distiphallus broadly tapered to apex that is not mucronate; fused surstyli variable 23 23. Fused surstyli about 1.4 times as long as broad, smoothly convex anterolaterally to narrow shallow anteromedial indentation; distiphallus carinate and constricted at midlength (Figure 48) H. mamtobae, new species Fused surstyli usually more than or less than 1.4 times as long as broad, or if 1.4 then concave or deeply cleft anteriorly; distiphallus not both carinate and constricted (except in H. ainsworthi) 24 24. Anterolateral margin of sternum 5 a bare acutangular prominence; distiphallus longitudinally lamellate; copulobus about 6.0 times as long as broad (Figure 45). H. subnitens Cresson Anterolateral margin of sternum 5 not a bare acutangular prominence, or if so then copulobus only about 2.0 times as long as broad; distiphallus variable 25 25. Fused surstyli broadly and deeply V-cleft anteromedially; distiphallus not projecting much beyond anterior margin of fused surstyli; postgonite above surstyli, not the copulobus (Figure 47) B. amnicola, new species Fused surstyli not broadly and deeply V-cleft anteromedially, or if so then distiphallus projecting nearly to sternum 5; postgonite above copulobus (except in H. americana and H. floridana) 26 26. Fused surstyli papilliform or triangular anteriorly (about 12.0-15.0 times as broad posteriorly as anteriorly) 27 Fused surstyli not papilliform or triangular anteriorly 28 27. Fused surstyli 2.0-3.0 times as long as copulobus; pregonites and postgonites ending near each other (Figure 63) H. cruraUs Coquillet Fused surstyli about as long as copulobus; pregonite much larger and projecting more anteriorly than postgonite (Figure 49) B. per sonata, new species 28. Sternum 5 with broad, flat, shallow recess anteriorly; copulobus concave posteromedially and ending close to fused surstyli; fused surstyli about as long as broad, narrowly notched anteromedially, and lobate anterolaterally (Figure 46) B. insulata, new species 32 SMITHSONIAN CONTRIBUTIONS TO ZOOLOGY Sternum 5 convex or straight-edged anteriorly; copulobus without such posteromedial concavity, or if so then not ending close to surstyli; fused surstyli not so formed 29 29. Fused surstyli slightly mucronate anteromedially, about 1.6 times as long as cercus, and about 2.0 times as broad as sternum 4; copulobus papilliform posteriorly (Figure 41). H. tibialis Cresson Fused surstyli not mucronate anteromedially, usually more than or less than 1.6 times cercus length, and less than 1.6 times as broad as sternum 4; copulobus variable 30 30. Fused surstyli, except for posterior end, evenly semicircular or ovoid and covering basiphallus laterally 31 Fused surstyli cleft, indented, or some form other than semicircular or ovoid 32 31. Fused surstyli about 1.5 times as long as cercus; copulobus with 2 prominent macrochaetae on tip; distiphallus apex not lamellate (Figure 64) H. luctuosa Cresson Fused surstyli about 3.0 times as long as cercus; copulobus without macrochaetae; distiphallus apex transversely lamellate (Figure 34) H. biloxiae, new species 32. Distiphallus distinctly expanded proximally and distally, entirely lamellate longitudinally and bicarinate distally; fused surstyli 2.0 times (or more) as broad as long, with 3 broad anterior lobes (Figure 31) H. floridana, new species Distiphallus not so expanded nor bicarinate, or if so then not lamellate; fused surstyli 2.0 times (or more) as broad as long in only one species (H. americana) 33 33. Postgonite absent; copulobus much broader than long; fused surstyli broadly and deeply concave anteriorly and about 2.0 times as broad as long (Figure 27). H. americana Cresson Postgonite present; copulobus longer than broad; fused surstyli deeply cleft anteromedially, or if broadly and deeply concave then not over 1.3 times as broad as long 34 34. Sternum 5 convex posteromedially, projecting as far or farther posteriad than papilliform copulobus; fused surstyli broadly and deeply V-cleft anteromedially along 0.6 of the length (Figure 52) H. notata, new species Sternum 5 concave posteromedially; copulobus, if papilliform, much longer than median length of sternum 5; fused surstyli not V-cleft along 0.6 of the length 35 35. Fused surstyli truncate anteriorly, about as long as broad and about 4.0?6.0 times as long as cercus; copulobus notched medially (Figure 32) H. ainsworthi, new species Fused surstyli concave, convex, cleft, or indented and variable in proportions; copulobus variable 36 36. Fused surstyli broadly and shallowly concave anteriorly and slightly broader than long; cercus about as long as broad; distiphallus constricted along middle portion and not lamellate (Figure 43) H. jormosa Cresson Fused surstyli deeply concave, V-cleft, notched or indented anteriorly, or if broadly and shallowly concave then longer than broad or cercus only about 0.5 times as long as broad; distiphallus lamellate if constricted along middle portion 37 37. Copulobus with 1 or 2 posterolateral winglike projections; basiphallus not covered laterally by fused surstyli; distiphallus longitudinally lamellate 38 Copulobus without such projections; basiphallus covered laterally, or if not then distiphallus not longitudinally lamellate throughout 39 38. Fused surstyli about 2.5 times as long as broad; copulobus with 1 posterolateral projection; basiphallus broadest near anterior end of surstyli (Figure 36) H. nobilis (Loew) Fused surstyli about 1.3 times as long as broad; copulobus with 2 posterolateral projections; basiphallus broadest at about middle third of surstyli length (Figure 21). H. idolator, new species 39. Fused surstyli convex anteriorly; copulobus with 2 setae on posterior tip about 0.5 times as long as copulobus; distiphallus longitudinally lamellate distally (Figure 54). H. cavator, new species Fused surstyli concave, cleft, or notched anteriorly; copulobus without such long setae; distiphallus variable 40 40. Fused surstyli shallowly concave anteriorly and about 1.4 times as long as broad; copulobus verrucate on middle portion and acutangular posterolaterally (Figure 39). H. advenae Cresson Fused surstyli broader than long if shallowly concave anteriorly; copulobus not verrucate on middle portion, tip variable 41 NUMBER 6 8 33 41. Distiphallus longitudinally lamellate from apex to anterior margin of fused surstyli 42 Distiphallus not longitudinally lamellate, or if so then only distally 44 42. Fused surstyli broadly and shallowly concave anteriorly and broader than long; median length of sternum 5 about 1.3 times greater than copulobus breadth; distiphallus about 2.5 times longer than copulobus breadth (Figure 20) H. procteri Cresson Fused surstyli U- or V-cleft anteriorly and longer than broad; median length of sternum 5 about 0.9 of copulobus breadth; distiphallus length about 5.0 times greater than copulobus breadth 43 43. Postgonite uncus about 2.0 times as long as thick; sternum 5 nearly right-angled anterolaterally; cercus about 1.4 times as long as broad; fused surstyli broadly U-cleft or concave anteriorly (Figure 33) H. crassipes Cresson Postgonite about 3.0 times as long as thick; sternum 5 broadly rounded anterolaterally; cercus about as long as broad; fused surstyli V-cleft anteriorly (Figure 56). H. saltator, new species 44. Fused surstyli about 1.2 times as long as broad and V- or somewhat U-cleft anteriorly 45 Fused surstyli about 1.2 times as broad as long and shallowly indented or notched anteromedially 46 45. Sternum 5 acutangular and prominent anterolaterally; distiphallus acute apically, carinate distally, but not longitudinally lamellate; cercus broader than long (Figure 35). H. serena Cresson Sternum 5 smoothly rounded anterolaterally; distiphallus rounded apically and longitudinally lamellate on distal half; cercus longer than broad (Figure 38) H. platygastra group 46. Sternum 5 acutangular and prominent anterolaterally; copulobus and fused surstyli separated by about 1.0 times copulobus breadth; postgonite uncus about 5.0 times (or more) longer than thick (Figure 57) H. proclinata Cresson Sternum 5 obtusangular anterolaterally; copulobus and fused surstyli separated by about 3.0 times copulobus breadth; postgonite uncus about 2.0 times longer than thick (Figure 51) H. melanderi, new species 47. Posterior end of copulobus with fascicula of long macrochaetae 48 Posterior end of copulobus without such fascicula 51 48. Copulobus noticeably concave posteromedially just anterior to fascicula; sternum 5 with small acutangular prominence anterolaterally; fused surstyli shallowly concave anteromedially (Figure 50) H. bilobifera Cresson Copulobus not concave posteromedially; sternum 5 without small acutangular prominence anterolaterally; fused surstyli concave or not 49 49. Sternum 5 with anterolateral right-angled projection from above; fused surstyli truncate anteriorly; postgonite uncus about 3.0 times as long as thick (Figure 23). H. ascita Cresson Sternum 5 without such projection; fused surstyli shallowly or moderately concave anteriorly; postgonite uncus about 5.0 times (or more) as long as thick 50 50. Copulobus with posteromedial digitiform lobe projecting posteriad beyond postgonite uncus and without medial verruca; postgonite uncus bent (Figure 24). H. gladiator, new species Copulobus without such lobe, but with slight medial verruca; postgonite uncus nearly straight (Figure 30) H- dhcursa, new species 51. Posterior margin of fused surstyli truncate; copulobus without noticeable convexity posteromedially just anterior to postgonite uncus; basiphallus not covered laterally by fused surstyli (Figure 15) # ? harti Cresson Posterior margin of fused surstyli sinuous; copulobus with noticeable convexity posteromedially just anterior to postgonite uncus; basiphallus covered laterally by fused surstyli (Figure 37) ? ? trichaeta Cresson Key to Third-Instar Larvae of some Hydrellia 1. Ventral frontoclypeal index 2.0 or less 2 Ventral frontoclypeal index over 2.0 3 2. Bifurcation index 2.5-2.7; cheliform spot touching clypeal arch margin in lateral view; clypeal arch distinctly angular (Figure 101) H. biloxiae, new species 3 4 SMITHSONIAN CONTRIBUTIONS TO ZOOLOGY Bifurcation index 4.8-5.4; cheliform spot not touching clypeal arch margin; clypeal arch gradually sloping (Figure 89) ? ? bergi Cresson 3. Ventral frontoclypeal index under 4.0 4 Ventral frontoclypeal index 4.0 or more 14 4. Bifurcation index 4.0 or less 5 Bifurcation index over 4.0 10 5. Ventral frontoclypeal index 3.0 or more 6 Ventral frontoclypeal index under 3.0 9 6. Bifurcation index 3.4 or more; part of dorsal phragmatal ramus dark 7 Bifurcation index 3.0 or less; dorsal phragmatal ramus hyaline (Figure 90). H. ainsworthi, new species 7. Mouth-hook, in lateral view, with distinct, thick base and thin beak; mouth-hook light spot ovoid, less than 2.0 times as long as wide; cheliform spot arms connected 8 Mouth-hook, in lateral view, without distinct base and beak; mouth-hook light spot narrow, triangular, and 3.0 times as long as wide; cheliform spot arms separate (Figure 103). H. spinicornis Cresson 8. Dorsal phragmatal ramus all dark H. griseola (Fallen) Dorsal phragmatal ramus partly hyaline (Figure 100) H. ischiaca Loew 9. Mouth-hook beak and base lengths subequal; mouth-hook light spot discal, encompassed (Figure 97) " ? discursa, new species Mouth-hook beak only 0.7-0.8 as long as base; mouth-hook light spot touching top margin of base (Figure 88) H- trichaeta Cresson 10. Bifurcation index 5.5 or less; cheliform spot arms not removed from clypeal arch margin by their length; clypeal arch not concave in lateral view 11 Bifurcation index 5.8 or more; cheliform spot arms removed from clypeal arch margin by their length; clypeal arch slightly concave between 2 prominences in lateral view (Figure 102) H. itascae, new species 11. Feeding apparatus partly hyaline 12 Feeding apparatus entirely dark H. morrhoni Cresson 12. Cheliform spot nearly parallel with dorsal phragmatal ramus; bifurcation index usually under 4.5; ventral frontoclypeal index usually 3.0 or less 13 Cheliform spot distinctly oblique to dorsal phragmatal ramus (Figure 95); bifurcation index usually over 4.5; ventral frontoclypeal index 3.0 or more....W. notiphiloides Cresson 13. Ventral frontoclypeal index 2.7 or less; mouth-hook beak distinctly longer than base; ventral phragmatal ramus moderately dark (Figure 86) H. ascita Cresson Ventral frontoclypeal index 2.8 or more; mouth-hook beak and base lengths subequal; ventral phragmatal ramus mostly hyaline (Figure 91) H. bilobifera Cresson 14. Ventral frontoclypeal index over 4.5; bifurcation index 3.5 or more; mouth-hook light spot small, elliptical; dorsal phragmatal ramus variable 15 Ventral frontoclypeal index 4.5 or less; bifurcation index 2.7 or less; mouth-hook light spot large, ovoid; dorsal phragmatal ramus mostly hyaline (Figure 96). H. cruralis Coquillett 15. Mouth-hook with light spot; clypeal-arch index 1.8 or more; mouth-hook base about twice as thick as beak in lateral view 16 Mouth-hook without light spot; clypeal-arch index 1.7 or less; mouth-hook base about 1.5 times as thick as beak in lateral view H. caliginosa Cresson 16. Feeding apparatus hyaline except dark mouth-hook and cheliform spot (Figure 106); mouth-hook beak distinctly longer than base H. pulla Cresson Feeding apparatus mostly dark except partly hyaline ventral phragmatal ramus (Figure 92); mouth-hook beak only 0.9 as long as base H. luctuosa Cresson Key to Puparia of some Hydrellia 1. Prothoracic end describing over half of a circle in ventral view; maximum breadth of prothorax at midlength 2 Prothoracic end variable in shape, if describing more than half of a circle, then maximum breadth of prothorax at or near its posterior margin 4 NUMBER 68 35 2. Head-lobe scar much longer than wide, nearly as long as prothorax 3 Head-lobe scar, nearly circular, its diameter about half of prothorax length (Figure 112). H. ascita Cresson 3. Lengthrminimum breadth about 20.0; maximum breadth:minimum breadth about 4.2; anal-plate index about 2.6 (Figure 115) H. trichaeta Cresson Length:minimum breadth 12.0-15.0; maximum breadth:minimum breadth 2.4-3.2; anal-plate 1.8-2.4 (Figure 109) H. bilobifera Cresson 4. Intersegmental constrictions inconspicuous; prothorax variable posterolaterally; bifurcate supraspiracular spinules present or absent 5 Intersegmental constrictions extensive, making puparium distinctly scalloped laterally; prothorax constricted posterolaterally; distinct bifurcate supraspiracular spinules present (Figure 108) H. cruralis Coquillett 5. Maximum puparial breadth in abdomen 6 Maximum puparial breadth in metathorax 15 6. Maximum puparial breadth 5.0 times or less than maximum breadth of head-lobe scar 7 Maximum puparial breadth 5.5 times or more than maximum breadth of head-lobe scar 13 7. Maximum breadth:minimum breadth 6.2 or less; length:minimum breadth 24.0 or less; anal-plate variable 8 Maximum breadth:minimum breadth 7.0 or more; length:minimum breadth usually 24.0 or more; anal-plate index 2.8-3.2 (Figure 123) H. bergi Cresson 8. Maximum breadth of head-lobe scar subequal to minimum puparial breadth; length: minimum breadth 10.0 or more; maximum breadth:minimum breadth variable 9 Maximum breadth of head-lobe scar distinctly less than minimum puparial breadth; length: minimum breadth 10.0 or less; maximum breadth:minimum 2.6?3.6 (Figure 107). H. luctuosa Cresson 9. Distinctly bifurcate supra- and subspiracular spinules absent 10 Distinctly bifurcate supra- and subspiracular spinules present H. mnsworthi, new species 10. Length:minimum breadth about 20.0 or less; spiracular peritremes terminal 11 Length:minimum breadth about 24.0; spiracular peritremes subterminal (Figure 121). H. morrisoni Cresson 11. Anal-plate index 2.6 or more; anal plate ovoid or elliptical, definitely not nearly rec- tangular 12 Anal-plate index 2.2 or less; anal plate subrectangular (Figure 118) H. tibialis Cresson 12. Length:minimum breadth 18.0 or more; anal plate reniform, anterior margin convex; maximum puparial breadth 1.7 times or less than maximum prothoracic breadth (Figure 117) H. spinicornis Cresson Length:minimum breadth 18.0 or less; anal plate subelliptical, anterior margin straight or slightly concave; maximum puparial breadth 1.8 times or more than maximum pro- thoracic breadth (Figure 122) H. griseola (Fallen) and H. ischiaca Loew 13. Maximum puparial breadth 6.0 times or less than maximum breadth of head-lobe scar; maximum puparial breadth 2.3 times or less than maximum prothoracic breadth 14 Maximum puparial breadth 7.5 times or more than maximum breadth of head-lobe scar; maximum puparial breadth 2.5 times maximum prothoracic breadth (Figure 116). H. notiphiloieUs Cresson 14. Anal-plate index 3.8 or more; maximum breadth:minimum breadth 5.5 or less; length:minimum breadth 15.0 or less (Figure 110) H. itascae, new species Anal-plate index 3.0 or less; maximum breadth:minimum breadth 6.0 or more; length:minimum breadth 18.0 or more (Figure 111) H. caliginosa Cresson 15. Prothorax constricted posterolaterally; anal-plate index 3.0 or less 16 Prothorax not constricted posterolaterally (Figure 113); anal-plate index 3.5 or more. H. discursa, new species 16. Prothorax convex or somewhat triangular anteriorly, angular laterally; length:minimum breadth 16.0 or less; abdomen nearly uniform in breadth posteriad to segment 7 (Figure 114) ?? ??**? Cresson Prothorax semicircular anteriorly, rounded laterally; length:minimum breadth 16.0 or more; abdomen tapering posteriad from segment 4 (Figure 119). H. biloxiae, new species 36 SMITHSONIAN CONTRIBUTIONS TO ZOOLOGY Hydrellia advenae Cresson FIGURE 39 Hydrellia advenae Cresson, 1934, p. 236; 1936, p. 257; 1944, pp. 165, 174.?Procter, 1938, p. 351.?Wirth, 1965, p. 743 [catalog listing]. DIAGNOSIS.?Palpus moderate yellow; 5-8 aristal rays; vertex index 3.6-4.5; ocular index 6.0-9.3; thoracic pleuron indistinctly splotched light gray and moderate olive-brown; fore and mid coxae moderate yellow anteriorly. Male length 1.29-1.58 mm; female 1.62-1.99 mm. Male terminalia as in Figure 39. HEAD.?Face yellowish gray; 4?5 pfa; epistomal index 1.2-1.7; mesofacial index 1.8-2.2; vertex index 3.6-4.5; A-index 1.6-1.9; ocular index 6.0- 9.3. Palpus moderate yellow; 5-8 aristal rays; antenna dark brown; frontal vitta dark gray; para- frontale strong yellowish brown; 10-14 postoculars. THORAX.?Ppn and mesonotum moderate brown; 3-4 adc and 2 pdc; pleuron indistinctly splotched light gray and moderate olive-brown; fore and mid coxae moderate yellow anteriorly, light gray laterally; rest of legs light-gray pruinose except moderate- yellow tarsi and mid and hind tibial apices. Wing length 1.63-1.90 mm; veins light yellowish brown or moderate brown; 5-7 setae on basal end of costa; 5-7 dorsal and 6-8 anterior interfractural costals; costal-section ratios: I I : I 2.0-2.3; III:IV 2.6-3.3; V: IV 2.9-3.5; Mi ? 2 index 1.5-1.8. ABDOMEN.?Terga moderate brown medially, light gray laterally and ventrally. Male terminalia: median third of posterior margin of sternum 5 deeply con- cave; anterolateral margin of sternum 5 smoothly rounded; posterolateral angle of copulobus about 35? from lateral corner; copulobus irregularly setose and slightly verrucate at midlength laterally. Postgon- ite bent anteriad and postgonite uncus slightly laterad; distiphallus constricted near basiphallusr apex acute. Anterior margin of fused surstyli concave and narrow (only about half the midbreadth of structure) ; B-index about 5.2; C-index 1.7-2.0. TYPE.?Holotype male, ANSP 6513. TYPE-LOCALITY.?Bar Harbor, Mount Desert Island, Maine (VII-15-1933, W. Procter). SPECIMENS EXAMINED.?9 (4o*d\ 5 $ $ ) from 2 localities: Maine: Bar Harbor, Mount Desert Island (VII-15-1933, W. Procter), 2?$ combined second and third ra- dial vein R4 + 5, combined fourth and fifth ra- dial vein RG, rectal gland r-m, radiomedial crossvein RT, rectum RV, rectal valve SI?5, sterna 1?5 S8, sternum 8 SA, subalare SAO, secondary atrial orifice SAT, spiracular atrium SB, halter scabellum SBP, subgenital plate SCB, subcranial breadth SI, food-neatus siphon SL, supra-alar NUMBER 6 8 127 Abbreviations Used on Illustrations?Continued SO, spermatogonia SOH, subocular height SP, lateral spermatheca SP1, prothoracic spiracle SP3, metathoracic spiracle SPP, supraspiracular protuberance SSP, supraspiracular spinous seta SPT, subspiracular protuberance SS, fused surstyli ST, spermatic tubule SV, stomodeal valve SVS, seminal vesicle SZ, spermatozoa T4, tergum 4 T5, tergum 5 T9 ? io, syntergum 9-(-10 TB, tibia TF, terminal filament TS, testis VB, vertex breadth VD, vas deferens VOD, vertical ocular height VPR, ventral phragmatal ramus VR, median spermatheca WA, wing area WL, wing length WVR, wall of median spermatheca 128 SMITHSONIAN CONTRIBUTIONS TO ZOOLOGY LO MO PO IV OV CM SV 0.10MM FIGURES 1-3.? 1, Hydrellia griseola (Fallen), male: head, cleared, frontal view. 2, H. bilobifera Cresson, female: internal genitalia, ventral view. 3, H. griseola (Fallen), female: gut, dorsal view. NUMBER 6 8 129 PDC FIGURES 4-6.?Hydrellia gnseola (Fallen): 4, abdomen, male, ventral view; 5, posterior half of abdomen, male, lateral view; 6, thorax proper, male, lateral view. 130 SMITHSONIAN CONTRIBUTIONS TO ZOOLOGY 13 FIGURES 7-14.?7, Hydrellia crassipes Cresson, male: left hind femur and tibia, anteroventral view. 8, H. morrisoni Cresson, male: left hind tibia, anterior view. 9, H. harti Cresson, male: head, lateral view. 10, 11, H. griseola (Fallen): 10, male, terminalia, lateral view; 11, male, head, cleared, frontal view. 12, H. bilobifera Cresson, male: musculature of gonal arch and phallapodeme, dorsal view. 13, H. griseola (Fallen) : male terminalia, dorsal view. 14. H. bilobifera Cresson: male internal genitalia, ventral view. NUMBER 6 8 131 22 FIGURES 15?22.?Male terminalia of adult Hydrellia species (ventral view of left half unless otherwise specified): 15, H. harti Cresson; 16, H. morrisoni; 17, whole structure for H. bergi Cresson; 18, H. prudens Curran; 19, H. caliginosa Cresson; 20, H. procteri Cresson; 21, H. idolator, new species; 22, H. valida Loew. 132 SMITHSONIAN CONTRIBUTIONS TO ZOOLOGY 27 28 FIGURES 23?30.?Male terminalia of adult Hydrellia species (ventral view of left half unless otherwise specified): 23, H. ascita Cresson; 24, H. gladiator, new species; 25, H. surata, new species; 26, H. griseola (Fallen); 27, H. americana Cresson; 28, H. itascae, new species; 29, H. rixator, new species; 30, H. discursa, new species. NUMBER 6 8 133 'l '? * * * ? ? ' tot. 35 1 36 37 FIGURES 31?38.?Male terminalia of adult Hydrellia species (ventral view of left half unless otherwise specified): 31, H. floridana, new species; 32, H. ainsworthi, new species; 33, H. crassipes Cresson; 34, H. biloxiae, new species; 35, H. serena Cresson; 36, H. nobilis (Loew) ; 37, H. trichaeta Cresson; 38, H. platygastra Cresson. 134 SMITHSONIAN CONTRIBUTIONS TO ZOOLOGY fV'/l 43 ?.?? FIGURES 39?46.?Male terminalia of adult Hydrellia species (ventral view of left half unless otherwise specified) : 39, H. advenae Cresson; 40, H. spinicornis Cresson; 41, H. tibialis Cresson; 42, H. deceptor, new species; 43, H. formosa Loew; 44, H. ischiaca Loew; 45, H. subnitens Cresson; 46, H. insulata, new species. NUMBER 6 8 135 vr.J -i 53 FIGURES 47?54.?Male terminalia of adult Hydrellia species (ventral view of left half unless otherwise specified): 47, H. amnicola, new species; 48, H. manitobae, new species; 49, H. personata, new species; 50, H. bilobifera Cresson; 51, H. melanderi, new species; 52, H. notata, new species; 53, H. borealis Cresson; 54, H. cavator, new species. 136 SMITHSONIAN CONTRIBUTIONS TO ZOOLOGY 59 62 FIGURES 55-62.?Male terminalia of adult Hydrellia species (ventral view of left half unless otherwise specified): 55, H. columbata, new species; 56, H. saltator, new species; 57, H. proclinata Cresson; 58, H. agitator, new species; 59, H. cessator, new species; 60, H. definita Cresson; 61, H. pulla Cresson; 62, H. flavicoxalis Cresson. NUMBER 6 8 137 72 FIGURES 63?72.?63?66, Male terminalia of adult Hydrellia species (ventral view of left half unless otherwise specified): 63, H. cruralis Coquillett; 64, H. luctuosa Cresson; 65, H. notiphiloides Cresson; 66, H. penicilli Cresson. 67-72, Female terminalia of adult Hydrellia species (lateral view of left side): 67, H. ascita Cresson; 68, H. trichaeta Cresson; 69, H. harti Cresson; 70, H. discursa, new species; 71, H. gladiator, new species; 72, H. bilobifera Cresson. 138 SMITHSONIAN CONTRIBUTIONS TO ZOOLOGY 83 FIGURES 73?83.?73, Hydrellia bilobifera Cresson: egg chorion, parasagittal section. 74, H. bergi Cresson: egg, lateral view. 75, H. bilobifera Cresson: egg, dorsal view. 76, H. ischiaca Loew: egg, ventral view. 77, H. discursa, new species: egg, dorsal view. 78, H. spinicornis Cresson: egg, dorsal view, shading indicates numerous punctulae or pits. 79, H. spinicornis Cresson: posterior part of abdominal segment 8 and tracheospiracular siphon of second-instar larva, dorsal view. 80 H. discursa, new species: abdominal segment 8 and spiracular peritreme of first-instar larva, lateral view of left side. 81, H. biloxiae, new species: anterior part of third-instar larva, lateral view of left side. 82, H. ischiaca Loew: posterior part of abdominal segment 8 and tracheospiracular siphon of second-instar larva, dorsal view. 83, H. ainsworthi, new species: left spiracular peritreme and ramus of third-instar larva, lateral view. NUMBER 6 8 139 84 87 89 FIGURES 84-95.?Feeding apparatus of larval Hydrellia (third-instar and lateral view unless otherwise specified): 84, H. notiphiloides Cresson: dorsal view; 85, H. bilobifera Cresson; 86, H. ascita Cresson; 87, H. tibialis Cresson: mouth-hook, lateral view; 88, H. trichaeta Cresson; 89, H. bergi; 90, H. ainsworthi, new species: second instar; 91, H. bilobifera Cresson; 92, H. luctuosa Cresson; 93, H. spinicornis Cresson: second instar; 94, H. deceptor, new species; 95, H. notiphiloides Cresson. 140 SMITHSONIAN CONTRIBUTIONS TO ZOOLOGY 96 101 103 010MM 104 105 FIGURES 96-106.?Feeding apparatus of larval Hydrellia (third instar and lateral view unless otherwise specified) : 96, H. cruralis Coquillett; 97, H. discursa, new species; 98, H. caliginosa Cresson; 99, H. ainsworthi, new species; 100, H. ischiaca Loew; 101, H. biloxiae, new species; 102, H. itascae, new species; 103, H. spinicornis Cresson; 104, H. morrisoni Cresson; 105, H. spinicornis Cresson: first instar; 106, H. pulla Cresson. NUMBER 6 8 141 - - H I S 107 109 0.5MM 110 108 O.SMM 112 O.SMM O.SMM 114 O.SMM FIGURES 107-114.?Puparia of Hydrellia (ventral view; setae and creeping welts on middle part omitted): 107, H. luctuosa Cresson; 108, H. cruralis Coquillett; 109, H. bilobifera Cresson; 110, H. itascae, new species; 111, H. caliginosa Cresson; 112, H. ascita Cresson; 113, H. discursa, new species; 114, H. pv.Ua Cresson. 142 SMITHSONIAN CONTRIBUTIONS TO ZOOLOGY 115 0.5MM O.SMM 119 117 118 120 121 122 O.SMM 123 O.SMM FIGURES 115-123.?Puparia of Hydrellia (ventral view; setae and creeping welts on middle part omitted): 115, H. trichaeta Cresson; 116 H. notiphiloides Cresson; 117, H. spinicornis Cresson; 118, H. tibialis Cresson; 119, H. biloxiae, new species; 120, H. ainsworthi, new species; 121, H. morrisoni Cresson; 122, H. ischiaca Loew; 123, H. bergi Cresson. NUMBER 6 8 143 FIGURES 124?129.?124, Hydrellia spinicornis Cresson: late second-instar larva with opiine Hymenoptera parasite, lateral view ( X 16). 125, H. discursa, new species: egg, lateral view ^X 48). 126, H. griseola (Fallen) and H. ischiaca Loew: left puparium with early pupa of H. griseola; right one with late pupa of H. ischiaca, lateral view ( x 4). 127, H. discursa, new species: posterior third of third-instar larva, ventrolateral view ( x 17). 128, H. cruralis Coquillett: puparia on left void, two on right with hymenopterous parasites, dorsal and ventral views ( x 4.5). 129, H. itascae, new species: puparium in situ in submergent Potamogeton leaf, dorsal view ( X 5.2). 144 SMITHSONIAN CONTRIBUTIONS TO ZOOLOGY 132 FIGURES 130-132.?130, Hydrellia griseola (Fallen): male with mite parasite, right lateral view ( x 8.2). 131, H. ischiaca Loew: puparia, left one with braconid Hymenoptera parasite, right one with pharate adult fly, dorsal view ( x 9.5). 132, H. cruralis Coquillett: puparium, ventral view, and emerged adult hymcnopterous parasite, dorsal view ( X 6.2; photograph from slide borrowed from C. O. Berg). NUMBER 6 8 145 OR- M3+Cu, 135 FIGURES 133-135.?133, Hydrellia griseola (Fallen): female internal genitalia, left view of parasagittal section ( x 17). 134, H. griseola (Fallen), male: proboscis, uncleared, left lateral view ( x 54). 135, H. griseola (Fallen), female: left wing, dorsal view ( x 18.7). 146 SMITHSONIAN CONTRIBUTIONS TO ZOOLOGY 138 FIGURES 136-138.?136, Hydrellia griseola (Fallen), male: abdomen, ventral view ( x 34). 137, H. griseola (Fallen), male: abdomen, left lateral view (X 34). 138, H. bilobifera Cresson, male: abdomen, left lateral view ( x 36.5), phallus partly depressed. NUMBER 6 8 147 FIGURES 139-142.?139, Hydrellia griseola (Fal len) , male: head, front view ( x 32 ) . 140, H. griseola (Fal len) , male: cibarium of probosis, front view ( X 32) . 141, H. griseola (Fal len) , male: cibarium, posterior view ( X 80) . 142, H. griseola (Fal len) , male: cibarium, right lateral view ( X 80) . Publication in Smithsonian Contributions to Zoology Manuscripts for serial publications are accepted by the Smithsonian Institution Press, subject to substantive review, only through departments of the various Smithsonian museums. Non- Smithsonian authors should address inquiries to the appropriate department. If submission is invited, the following format requirements of the Press will govern the preparation of copy. (An instruction sheet for the preparation of illustrations is available from the Press on request.) Copy must be typewritten, double-spaced, on one side of standard white bond paper, with V/2" top and left margins, submitted in ribbon copy with a carbon or duplicate, and accompanied by the original artwork. Duplicate copies of all material, including illustrations, should be retained by the author. There may be several paragraphs to a page, but each page should begin with a new paragraph. Number consecutively all pages, including title page, abstract, text, literature cited, legends, and tables. The minimum length is 30 pages of typescript and illustrations. The title should be complete and clear for easy indexing by abstracting services. Taxonomic titles will carry a final line indicating the higher categories to which the taxon is referable: "(Hymenoptera: Sphecidae)." Include an abstract as an introductory part of the text. Identify the author on the first page of text with an unnumbered footnote that includes his professional mailing address. A table of contents is optional. An index, if required, may be supplied by the author when he returns page proof. Two headings are used: (1) text heads (boldface in print) for major sections and chapters and (2) paragraph sideheads (caps and small caps in print) for subdivisions. Further headings may be worked out with the editor. In taxonomic keys, number only the first item of each couplet; if there is only one couplet, omit the number. For easy reference, number also the taxa and their corresponding headings throughout the text; do not incorporate page references in the key. In synonymy, use the short form (taxon, author, date, page) with a full reference at the end of the paper under "Literature Cited." Begin each taxon at the left margin with subsequent lines indented about three spaces. Within a taxon, use a period-dash (.?) to separate each reference. Enclose with square brackets any annotation in or at the end of the taxon. For references within the text, use the author-date system: "(Jones, 1910)" or "Jones (1910)." If the reference is expanded, abbreviate the data: "Jones (1910, p. 122, pi. 20: fig. 1)." Simple tabulations in the text (e.g., columns of data) may carry headings or not, but they should not contain rules. Formal tables must be submitted as pages separate from the text, and each table, no matter how large, should be pasted up as a single sheet of copy. For measurements and weights, use the metric system instead of (or in addition to) the English system. Illustrations (line drawings, maps, photographs, shaded drawings) can be intermixed throughout the printed text. They will be termed Figures and should be numbered consecutively; however, if a group of figures is treated as a single figure, the individual components should be indicated by lowercase italic letters on the illustration, in the legend, and in text references: "Figure 9b." If illustrations (usually tone photographs) are printed separately from the text as full pages on a different stock of paper, they will be termed Plates, and individual components should be lettered (Plate 9b) but may be numbered (Plate 9: figure 2). Never combine the numbering system of text illustrations with that of plate illustrations. Submit all legends on pages separate from the text and not attached to the artwork. In the bibliography (usually called "Literature Cited"), spell out book, journal, and article titles, using initial caps with all words except minor terms such as "and, of, the." (For capitaliza- tion of titles in foreign languages, follow the national practice of each language.) Underscore (for italics) book and journal titles. Use the colon-parentheses system for volume, number, and page citations: "10(2) :5-9." Spell out such words as "figures" and "plates" (or "pages" when used alone). For free copies of his own paper, a Smithsonian author should indicate his requirements on "Form 36" (submitted to the Press with the manuscript). A non-Smithsonian author will receive 50 free copies; order forms for quantities above this amount with instructions for payment will be supplied when page proof is forwarded.