Revision of the Megaselia giraudii and M. densior species complexes of Europe, including ecological notes

The Megaselia giraudii and M. densior species groups are revised to include a total of twenty-one described European species. Eight species are new to science: M. citrinella BUCK sp. n., M. glabrimargo BUCK sp. n., M. hexanopbila BUCK sp. n., M. immodensiorDISNEY sp. n. (= M. densior auctt., nec SCHMITZ), M. labellaspinata DISNEY sp. n. (= AI. septentrionalis auctt., nec SCHMITZ), Af. mimodensior'&MVri sp. n., AI. raruvesiculae DISNEY sp. n., and AI. xanthophila BUCK sp. n.. Megaseliaphoenicura (SCHMITZ), M. badia SCHMITZ and M. hayleyensis DISNEY (all described from males only) are synonymised with M. quadriseta (SCHMITZ) (previously only known in the female sex). Many specimens previously assigned to Af. badia are recognised as being M. densior SCHMITZ, and Af. coulsoni DISNEY is synonymised with this species. Lectotypes are designa­ ted for M. albicans (WOOD), Af. densior, M. offuscata SCHMITZ, M. phoenicura and M. quadriseta, and a neotype is designated for Af. comosa (SCHMITZ). The males of Af. rubricornis (SCHMITZ) and females of Af. correlata (SCHMITZ), Af. haraldlundi DISNEY and M. parnassia DISNEY are described for the first time. The putative male of AI sanguinea (SCHMITZ) and the putative female of Af. albicans are also described for the first time. The recognition of other species is clarified including four species that do not belong to the giraudii!densior complex (AI. nigrescens (WOOD), M. opacicornis SCHMITZ, M. rubida (SCHMITZ), and M. ustulata (SCHMITZ)). A revised key to males and a new key to females is provided, utilising the novel taxonomic character of the subcuticular sensilla inside the third antennal segment, and giving special attention to the morphology of the female ovipositor and vaginal sclerite. Biological data (breeding substrates of larvae, bait, colour and habitat preference of adults) are reported for most species.


Introduction
The giant genus Megaselia Rondani has been described as "the Diptera enfant terrible" (SMITH, 1984). The difficulties, however, are largely manmade, in that the pinned mate rial of the past does not allow one to observe critical fine details, with the result that type series are frequendy found to contain more than one species. In addition, these flies being sexually dimorphic, the describing of species on the basis of females alone has been a recipe for generating synonyms. Furthermore, with species that vary in size con sequential allometric effects have been assigned unwarranted taxonomic weight. The Handbook on the males of the British species (DISNEY, 1989a) represented a pionee ring attempt at a new approach to the identification of Megaselia species based upon slide mounted specimens, instead of the traditional dried specimens on micropins or card points. Subsequent revisionary studies include a closer scrutiny of the male penis com plex (D i s n e y , 1999).
The present paper deals with a relatively small group of species within this genus that comprises M. giraudii (E g g e r ) , M. densior SCHMITZ and their related species. The former complex falls in the artificial "Abteilung VI" of Megaselia (SCHMITZ & DELAGE, 1981) that is characterised by a bare mesopleuron and the presence of four scutellar bristles. The recognition of species of the M. giraudii/densior group is rendered difficult by the fact that in males the anterior pair of scutellar brisdes is often reduced to hairs. Conse quently, several species are only known in one sex, or males and females have been described under different names.
The present work was originally stimulated by an ecological investigation on small saprophagous flies (BUCK, 1997b), during which a surprisingly high number of species and specimens of the above complex was obtained in trapping and rearing experiments. The identification of this material proved to be difficult, because pre-existing keys only co ver males of the hitherto known species (DISNEY, 1989a) or rely on characters of litde taxonomic value (SCHMITZ & DELAGE, 1981). Furthermore some key voucher speci mens (including paratypes) previously examined have subsequently proved to be diffe rent from the holotypes or lectotypes. Thus, our taxonomic investigations have resulted in a partial revision of the "Abteilung VI", including the description of eight new spe cies. Apart from the taxonomic results new biological data (larval breeding substrates, bait, habitat and colour preferences of adults) are presented for many species. small or stout). Female sternum 8 with a pair of hairy lobes at rear margin. All five segments of front tarsus with longitudinal postero-dorsal hair palisade. (5) Normally some hairs below basal half of hind femur are longer than those of antero-ventral row in distal half (at least they are subequal). (6) Tibiae without rows of enlarged anterodorsal setae. (7) Spines in apical comb of posterior face of hind tibia all simple, not bifurcated. (8) Costa not considerably shortened, Cl at least 0.42. (9) Subcosta not rea ching R] (in M. densior it sometimes reaches but then its distal fifth is faint). (10) Knob of haltere whitish to dusky yellow (but tip sometimes infuscated in M. glabrimargo and o ff us cat a).
While the above diagnosis enables the recognition of females and of males with welldeveloped anterior scutellar bristles, it is not sufficient for males in which the anterior scutellars are reduced to hairs. In this case the key to British Megaselia (DISNEY, 1989a) can be used for an initial assignment of a species to this complex. In this key species of the M. giraudii I densior complex will run out at couplets 151-160 and 222-225. The revi sed key to males which is provided below replaces these couplets and refers to some figures in the key to British species.
Morphology. The basic morphological features for the identification of phorids have been outlined by DISNEY (1994). In the present study some additional characters were used that require further comments. Antennal sensilla. Various kinds of sense organs and sensilla are located on the anten nae. PFEIL et al. (1994) gave a detailed account of the types and distribution of sensilla in the antennae of Megaselia halterata (WOOD). During the present study it was discover ed that one kind of sensilla, the 'subcuticular pit sensilla' of PFEIL et a l, is of outstan ding taxonomic significance. These sensilla are present in most species examined and are recognised as subspherical vesicles inside the third antennal segment. The vesicles are in fact subcuticular pits which open through a small pore to the exterior. They are most commonly encountered on the inner face of the third antennal segment. The number and the size of these sensilla proved to be species-specific. In three species the SPS are exceptionally large (e.g., M. quadriseta, Fig. 9). In about one third of the species their number is greatly reduced (e.g., M. haraldlundi, Fig. 8) and in one (M. rubricornis) they are completely missing. The special importance of this character lies in the fact that the morphology of the SPS is basically the same in both sexes (but their number is usually somewhat reduced in females). The SPS must not be confused with the much smaller 'bulbous peg sensilla' of PFEIL et al. which have bases that look like minute, clear, round spots in the integument. Another type of sensilla, the 'domed pit sensilla' of PFEIL et al., may also be useful in some cases. A pair of these sensilla is located on the swollen base of each third aristomere. Usually they are well separated from each other but in one species (M. correlata, Fig. 6) they are closely approximated. However, aberrant specimens with closely approximated sensilla can be found in most other species, too. Labella. In the species descriptions the number of spinules below the labella is reported. The count is made for a single labellum and neither includes the spinules at the distal end of each pseudotrachea nor the small hairs near the lateral margin. The usual condi tion is only a few spinules below the labella (13-30) in both males and females. Species with densely spinose labella are sexually dimorphic with regard to the number of spinu les. In these species the male labella bear 50-200 spinules whereas females only have about half to third this number. In transitional species (males with 30-50 spinules) fe males also have fewer spinules (ca. 20-30). In order to get a clear view of the lower side of the labella it may be necessary to mount the proboscis separately. Notopleural cleft. In the past, different types of notopleural clefts were merely charac terised in terms of 'weakly' or 'strongly developed'. There are, however, clear morphologi cal differences which are highly species-specific. One important feature is the course of the cleft in its posterior half. Usually it is straight  but in one species it is distinctly curved upward and forward (Fig. 12). Another relevant feature is the presence or absence of sclerotised ridges. In most species heavily sclerotised ridges are lacking ('weak notopleural cleft'). However, in M. albicans the lower rim of the cleft is fortified posteriorly by a strongly sclerotised ridge (Fig. 12). In two species a small V-shaped ridge is lying immediately behind the cleft (Fig. 13). Hypopygium. Shape of epandrium: In many species the hind and lower margins of the epandrium form an acute or right angle on the left face just at the ventral base of the anal tube (e.g., Figs 38,42). In some species the hind margin of the epandrium continues straight below the ventral base of the anal tube to form a rounded lobe which is called here the postero-ventral lobe (Fig. 30). Because of asymmetry of the hypopygium the situation on the right side is somewhat different. On this side the hind margin of the epandrium usually continues straight below the anal tube, and normally it is more or less arched (e.g., Figs 20, 31), angled rarely ( Fig. 16), or strongly receding (Figs 18,29). On the right side the hairy lobe which forms the lowermost portion of the epandrium is called the postero-ventral lobe (e.g.,Figs 16,29,31,43). Micropubescence: A further character of importance is the micropubescence of the hypandrial lobes and the lower face of the proctiger. The microtrichia may be decumbent (i.e. they raise from the sur face at an angle of about 10-30°), semi-erect (dto., 30-80°) (Fig. 40), or erect (dto., ca. 90°). Recurved microtrichia are curved forward apically. Female ovipositor. In contrast to tergite and sternite 7 which are quite variable intraspecifically and therefore of limited use, the sternum 8 of the female ovipositor is of high est importance for identification. In the investigated group it always terminates in two posterior lobes. These lobes vary in chaetotaxy and morphology. In many species the surface of these lobes is more or less raised above the level of the hind margin to form a bulge. Sometimes the bulge projects laterally beyond the hind margin (e.g.,Figs 72,77). In all species the posterior lobes bear some more or less enlarged hairs. The distance of these hairs to each other and to the edge of the bulge or -if there is no bulge -to the hind margin of the lobe is species-specific. To avoid confusion the term hind margin is not applied to the edge of the bulge but is exclusively used for the margin of the seg ment (which is directly contiguous to the genital opening) (Figs 72,77). This margin is DOI: 10.21248/contrib.entomol.51.1.73-154 almost always glabrous (i.e. devoid of microtrichia) whereas the edge of the sternal bulge bears relatively long microtrichia in most species. The degree of sclerotisation is also characteristic. In some species the hind margin is more strongly sclerotised than the rest of the lobe, yet in other species the inner face of the lobe (i.e. a small area of the hindmost part of the wall of the genital chamber) is sclerotised . A further important character is the micropubescence of sternum 8. It varies according to species from completely glabrous to completely covered in microtrichia. Female internal genitalia. As important as the ovipositor are the various-shaped sclerites within the female vagina. They are highly species-specific and are present in most spe cies. In some species weakly sclerotised tubular or cylindrical structures can be found . The homologies of this unpaired organ are unknown. It could represent a modified spermatheca, some kind of ventral receptacle, or a modified accessory gland. Further investigation is necessary to resolve this problem. In the meantime we use the neutral term 'tubular organ'.
Keys. Two keys, for males and females, are presented below. The keys cover the Euro pean fauna but it should be noted that no extra-European Palaearctic species have been described from this group. The key to males can be used as a replacement for couplets 151-160 and 222-225 of the key to British species (DISNEY, 1989a). Figure numbers preceded by 'D -' refer to figures in the latter publication. M. labellaspinata sp. n., M. malhamensis, M. offuscata, and M. trojani are only known in the male sex and are there fore not considered in the key to females, M. septentrionalis is only known in the female sex. The female key includes seven species which cannot be named at present. Some of them may represent the females of species which are only known in the other sex, some of them are new species which will only be formally described when the male is known.

1
A notopleural cleft present in front of anterior notopleural brisde . When the cleft is inconspicuous this region of the notopleuron is bare 2 1* Notopleural cleft absent, its position occupied by hairs like those on rest of scu tum.. 3 Notopleural cleft curved upward and forward in posterior half, fortified with strong sclerotised ridges (Fig. 12). Cerci broad and each bearing more than ten hairs ( Microtrichia below proctiger recurved. Hind margin of left side of epandrium extends ventrally more than width of base of anal tube before it curves forwards ( Fig. D-400). Hind margin on right side also more vertical (Fig. 20). Left hypandrial lobe distinctly infuscated, medium brown. Cerci with 8-11 (-12) hairs parnassia DISNEY 6* Microtrichia below proctiger semi-erect, pointing rearwards. Postero-ventral bor der of left side of epandrium almost immediately curves forward from just below base of anal tube (Fig. 17). Hind margin on right side more receding below anal tube (Fig. 18). Left hypandrial lobe more or less translucent, greyish. Cerci with 7-9 ( Palps somewhat inflated, with relatively short bristles (Fig. 1). The longest hair on left side of epandrium only weakly differentiated and not quite as strong as those on cerci. Microtrichia at apex of left hypandrial lobe decumbent and relatively short ( Fig. D-393 12 Epandrium with a well developed postero-ventral lobe on left side (Fig D-403). Right hypandrial lobe longer than left one. Penis complex with a long slender process (Fig. 25). Sensilla on swollen base of third aristomere closely approxima ted (Fig. 6 Note: in M. immodensior the internal surface is sometimes weakly sclerotised, but in this species the enlarged hairs of the sternal lobes are broadly separated from hind margin (Fig. 81).

5
Lobes of sternum 8 with a postero-laterally projecting bulge, sclerotised area small (Fig. 72). Vaginal sclerite a small, weakly sclerotised ring (Fig. 101). Hairs of hypoproct not crowded towards anterior margin (cf. Fig. 93 Dufour's crop mechanism large, slightly emarginate posteriorly (Fig. 130). Ster num 8 with broad posterior lobes (Fig. 80). Vaginal sclerite present (Fig. 113).... rubricornis (SCHMITZ) 12* Third antennal segment brown, with few SPS. Width of abdominal tergites strongly decreasing posteriorly, T6 being subtriangular (Fig. 44). T7 (Fig. 50) and S7 al most linear. Dufour's crop mechanism smaller, rounded behind (Fig. 131). Ster nal lobes smaller and with fewer hairs (Fig. 79). Vaginal sclerite absent (Fig. 110) (Fig. 106). Hind margin of sternum 8 bordered by 4-8 hairs between posterior lobes (Fig 81). Legs pale brownish yellow to yellowish brown. T10 bare immodensior DISNEY sp. n. 14* Vaginal sclerite more slender, convex towards right side or just a straight transver se bar. Hind margin between sternal lobes bordered by 2-3 hairs. Legs brown. Vaginal sclerite as in Fig. 1  20 Antennae yellow to dusky yellow Posterior lobes of sternum 8 broad, with many hairs but no bulge (Fig. 86). At least one enlarged hair of lobes standing well before hind margin. Vaginal sclerite roughly semi-circular (Fig. 116). ... sanguínea (SCHMITZ) 20* Antennae brown. Posterior lobes of sternum 8 narrow, with fewer hairs, bulge present (Fig. 87). Enlarged hairs of sternal lobes close to edge of bulge. Vaginal sclerite transverse, tapered on right side and like the eye of a needle (Fig. 111) The diagnoses of the new species are, of necessity, somewhat long as the genus contains about 1,400 described species and an estimate of at least ten times as many undescribed species. To keep this work as concise as possible we refer to previous publications for general descriptions of already known species. In these cases only new information is presented.
M egaselia albicans (W OOD, 1908) (Figs 12,51,60,69,119) Phora albicans WOOD, 1908: 172 (<? 9 ). Aphiocbaeta albicans (WOOD)-BRUES, 1915: 111. Megaselia albicans (WOOD). SCHMITZ, 1928: 144. The female of M. albicans has not been adequately characterised before. In the Schmitz collection several females attributed to this species were found. They bear the same types of labels and were collected around the same dates as certain albicans males (unfor tunately the labels only indicate the collection date). Closer examination revealed that these females belong to three different species (see also below Megaselia sp. 4 and M. giraudii), but at least three specimens are probably true M. albicans. Apart from general similarity they share two significant characters with the males (reduced number of SPS in the third antennal segment and densely spinose labella). However, it cannot be ruled out that they belong to the related M. offuscata.  , 8.V.1920, d , ll.v.1920, d , 12.V.1920 (all with yellow labels), coll. Schmitz (ZFMK).

Description
For a general description see SCHMITZ & DELAGE (1981). ant closer to eye margin than to median furrow and a little closer to al than to upper sa. Third antennal segment only with 2-6, small to medium-sized (0.005-0.016 mm) SPS. Male labella with 130-150 pale spinules, about twice as many as in female (65-75). Notopleural cleft distinctly curved upward and forward in posterior half (Fig. 12). Lower rim of cleft developed as a stron gly sclerotised ridge. Abdominal venter hairy on segments 3-6. Male hypopygium figu red by DISNEY (1989a: Fig. 397). Cerci broad, each one bearing 11-14 hairs. Micropube scence of ventral edge of proctiger erect to recurved in basal half. Left hypandrial lobe longer than right one. Microtrichia below left lobe erect to recurved, those of right lobe pointing rearwards. Female abdomen: T6 trapezoid, 1.0-1.3x as broad as long. Anterior notch narrow, its width about a quarter to a sixth of maximum breadth of tergite. T7 very long and narrow (Fig. 51). T8 rectangular, with 13-21 hairs. T10 bare. S7 long and narrow (Fig. 60), with three enlarged hairs at posterior margin. Sternum 8 devoid of microtrichia except in postero-lateral parts. Posterior lobes of sternum 8 only with two long hairs (Fig. 69). Lobes otherwise completely bare or at most with some scattered microtrichia near bases of setae. Hairs of hypoproct somewhat crowded towards anteri or margin. Cerci cylindrical, about 2.5-3.5x as long as broad, with usual hairing plus a group of 4-5 short, blunt-tipped hairs at apex. Vaginal sclerite a very long plate (ca. 0.20 mm), weakly sclerotised, and widening posteriorly (Fig. 119

Type material
Previously exam ined (DISNEY, 1986). Females, having not been adequately described before, may be confused with other simi lar species. Based on a large series of reared specimens a more detailed description of M. breviseta is provided (for general characters see SCHMITZ & DELAGE, 1981).

Description
ant of frons midway between upper sa and al, or a little closer to sa. SPS numbering ca. 20-30, arranged in about three irregular rows near base of inner face of third antennal segment, lacking from outer face. SPS less numerous in female (ca. 8-20) and only in 1-2 irregular rows. Diameter of SPS small to medium (0.004-0.014 mm). Sensilla on swol len base of third aristomere well separated to closely approximated. Labella in both sexes with 16-21 spinules on lower face. Anterior scutellars well developed and as long as posterior ones in both sexes. Abdominal venter hairy on segments 3-6, in female sometimes segment 2, too, but hairs weaker and those of segments 3-4 sometimes reduced to microscopic setulae. Hypopygium see DISNEY (1989a: Fig. 402). Left side of epandrium with 10-21 hairs. Microtrichia of left hypandrial lobe erect to recurved (though usually with some semi-erect ones near outer margin). Micropubescence of right lobe decumbent. Aedeagal complex with a short slender process above right hypandrial lobe. Tip of process a little upcurved. Female abdomen: T6 trapezoid, 0.9-1.lx as broad as long. Anterior notch narrow, width only one fifth to one eighth of maximum breadth of tergite. T7 expanded towards front, shallowly emarginate anteriorly. T8 rectangular, with 10-27 hairs (mainly at or near lateral margins). T10 bare. S7 narrow, with a strongly sclerotised median strip in anterior half to two thirds (rarely missing), and with (2-)4(-5) enlarged hairs at rear (Fig. 63). Posterior third to two thirds of sternum 8 covered in microtrichia. Margins of posterior lobes narrowly sclerotised and devoid of microtri chia. Each lobe with three long hairs that stand well off margin (Fig. 88). Hairs of hypoproct equally distributed over disc. Cerci cylindrical, 3-4x as long as broad. Apart normal 5+1 long hairs with apical group of 7-8(-9) short, blunt-tipped hairs. Left porti on of vaginal sclerite convex posteriorly (Fig. 112 Longest cc 0.11-0.16 mm. (4-)5-8(-9) cal, (4-)5-6 ax, longest of the latter about as long as longest cc. Hair at base of Rs vestigial (0.012-0.030 mm), rarely absent.

Remarks
M. breviseta, M. correlates and M. rubricornis probably form a monophyletic group which is characterised by two putative synapomorphies: (1) penis complex with slender posterior process (most conspicuous in M. correlates)-, (2) female reproductive tract with peculiar tubular organ (ventral receptacle or modified spermatheca?).

Distribution
Recorded from several Central European countries as well as England, Sweden, Hunga ry (D i s n e y , 1991), and Russia (Kuril Islands) (MlCHAILOVSKAYA, 1998).

Other material examined
See below M. quadriseta (misidentified type material).

Etymology
The species is named for the yellow colour of its thorax.

Diagnosis
A bicoloured species with brown frons, yellow to yellowish brown thorax and predomi nantly brown abdominal tergites. ant about midway between upper sa and al. Third antennal segment brown, with numerous small to medium-sized SPS. Palps yellow, their longest bristles a little longer than their diameter. Labella not inflated, with few spinules below. Notopleuron with two bristles, cleft absent. Scutellum with anterior setae hair like in male, long in female. Mesopleuron bare. Abdominal venter hairy on segments 3-6. Epandrium brown, on each side with one stronger, more bristle-like hair which is about as strong as terminal hairs of proctiger. Right side of epandrium with a strongly projecting postero-ventral lobe. Anal tube yellow, shorter than dorsal surface of epandri um. Terminal hairs of proctiger stronger than hairs of cerci. Hypandrial lobes subequal, with decumbent micropubescence. Female T6 subquadrate. T7 Y-shaped. T10 bare. S7 moderately broad, expanded posteriorly. Posterior lobes of sternum of segment 8 each with three enlarged hairs, not sclerotised on internal surface. Vaginal sclerite strongly sclerotised, tapered on right side. Legs yellow, tip of hind femur darkened. Most hairs below basal half of hind femur are longer than those of antero-ventral row in distal half. Cl longer than 0.43. Costal section 1 subequal to sections 2+3 in both sexes. Sub costa fades away before reaching R . Hair at base of Rs short to moderately developed.

Description
Male. Frons about as long as broad, chestnut-brown and strongly contrasting with much paler thorax in mature specimens. Whole frons densely covered in microtrichia, and with 40-75 hairs, ant a little closer to eye margin than to median furrow, rarely midway between them, sa closer together than pre-ocellars. Area around sockets of sa yellowish brown, paler than rest of frons. Cheeks with 3-5 hairs, jowls with 2-3 brisdes. Diameter of third antennal segment at most half frontal width. SPS distributed around base of third antennal segment, but most numerous beneath inner face, there being about 4-6 irregular rows. Total number of SPS exceeding 30, diameter 0.005-0.013 mm. Arista brown, with pale basal segments. Cilia hardly longer than diameter of basal joints. Sensilla on base of third aristomere broadly separated. Palps with 5-9 brisdes (maximum length 0.048-0.070 mm) plus 8-12 hairs. Prementum with two pairs of hairs. Labrum much narrower than diameter of third antennal segment. Each labellum bears 15-20 blunt spinules and some longer hairs on lower surface. Five pairs of pseudotracheae. Ground colour of thorax normally bright yellow (only pteropleura and postnotum slighdy tinged brownish). Metafurca pale. Anterior scutellar setae about fourth to (rarely) half as long as posterior ones. Abdominal tergites brown but T1 paler and T5-6 gradually becoming paler towards rear, usually in part yellowish brown. Tergites short-haired but with some stronger setae near postero-lateral corners of T2 and a row of bristles at rear margin of T6. Abdominal venter greyish yellow, sparsely haired but hairs relatively long. Epandrium with 11-17 hairs on left side (Fig. 42). Postero-ventral lobe of right side of epandrium ( Fig. 43: pv) strongly convex, well raised above depressed hind margin (Fig. 43: hm). Each cercus with (6-)7(-8) hairs that are weaker than hairs of proctiger. Micro pubescence below the latter erect to recurved (

Distribution
Holarctic, hitherto recorded from Germany, the Netherlands, Great Britain, Canada (Ontario), and the U.S.A. (Texas).
M egaselia correlata (SCHMITZ, 1918) (Figs 6,25,44,50,79,110,131) Aphiochaeta correlata SCHMITZ, 1918: 145 (ci  This species was rescued from synonymy with M. giraudii by DISNEY (1986), after remoun ting of the holotype male. Since then a good series of males plus females has been obtained from the canopies of oak trees in England. The female is now described for the first time.

Remarks
See below M. breviseta.

Distribution
Only known from Germany, Great Britain and the Netherlands.
M egaselia densior SCHMITZ, 1927 (Figs 3, 34-35, 45, 55, 66, 68, 99 , 129) Megaselia densior SCHMITZ, 1927: 131 ( + 9  This species was recorded from Britain (DISNEY, 1985) after comparison of material from Scotland and England with a male from Limbricht (Holland) attributed to this species by Schmitz (but actually being a misidentified M. immodensior sp. n.). Since then RHLD has borrowed and remounted a male syntype of M. densior. This proves to be a different species to the previously examined voucher specimen, and has been designated the lectotype of M. densior. The species erroneously attributed to this species in Britain are described as new (see M. immodensior sp. n. and M. raruvesiculae sp. n.). The true M. densior has been previously recorded as M. badia and M. coulsoni. As previously reported (DISNEY, 1993), the unique holotype of M. coulsoni only has two, not three, notopleural bristles. Furthermore the subcosta runs to R . Consequently it runs to couplet 212 of the key to British species (DISNEY, 1989a). However, the exami nation of a greater range of material serves to indicate that the apical third of the subcosta in M. coulsoni' is variably developed. Indeed, it now seems that it typically has the subcosta free and so most specimens key out at couplet 225 (to M. badia) instead. The greatly reduced micropubescence of the hypandrial lobes in the males of this spe cies will immediately distinguish it from several related species. It is now clear that many specimens from Great Britain attributed to M. badia are in fact M. densior. Furthermore these include the holotype of M. coulsoni, which is therefore herewith synonymised with M. densior. The true M. badia is synonymised with M. quadriseta below.  -4.viii. 1993, reared from dead Arion ater (bait exposed 15.vi.-6.vii.1993); 11 3 12 9 , emerged l.-22.ix.l993, reared from buried liver baits (exposed 30.vii.-13.viii.1993), all leg. M. Buck (7 3 5 9 coll. MB, 3 3 3 9 CUMZ, 4 3 5 9 ZFMK). Ca. 750 further specimens from the same locality (see Tables  1, 2). Rhineland-Palatinate, Mainz, Lennebergwald, 3 , 26.vii. -10.viii.1988, 9 , 15.-30.viii.1989

Description
The description provided by SCHMITZ & DELAGE (1981) is to be complemented as follows: ant of frons closer to al than to upper sa. Antenna as in Fig Each lobe with only one enlarged hair standing well off posterior margin. Lobes wit hout bulge, and margin somewhat sclerotised, tinged brownish. Hairs of hypoproct somewhat crowded towards anterior margin. Cerci cylindrical, about 3x as long as broad.

Recognition
While this species is distinguished from similar species in the keys above, males with subcosta reaching Rt will be found to run to couplet 212 of the key to British Megaselia (DISNEY, 1989a), to M. largifrontalis SCHMITZ, 1939. The latter, however, has a different hypopygium (DISNEY, 1989a: Fig. 441) and much browner femora.

Distribution
The true M. densior is hitherto only known from Germany, the Netherlands and Eng land. The records from other countries need confirmation.
SANTOS A breu (1921) described a number of species from the Canary Islands, but these descriptions have frequendy proved inadequate for the recognition of species in the light of subsequent knowledge of the complexities of Megaselia taxonomy. Further more most of the type material has been lost. SANTOS ABRÉU & SCHMITZ (1934) there fore provided a critically evaluated list of the Phoridae of the Canary Islands prior to a detailed paper by SCHMITZ (1936) based upon it and much fresh material collected by R. Frey. Two of the species described by SANTOS ABRÉU were M. comosa and M. interme dia, the former on the basis of the male sex only and the latter from both sexes. The type material of both species has been lost. The existing voucher material, therefore, is that collected by Frey and then attributed to these species by SCHMITZ (1936). A series of both sexes was referred to M. comosa but with the following amendments. The females were attributed to M. intermedia but the males to M. comosa. As the male clearly differed from the description of M. intermedia, and as it was assumed that the males and females collected by Frey were conspecific, SANTOS ABRÉU & SCHMITZ (1934) synonymised the females of M. intermedia with M. comosa but not the males. They assumed that SANTOS ABRÉU had originally associated the male and female attributed to M. intermedia in error and that Frey's series of males and females, which they attributed to M. comosa, were correcdy associated. The implication was that the lost male of M. intermedia represented the only recorded specimen of this species and its female was unknown. BORGMEIER (1968) documented this but did not then list M. intermedia either as a valid species or under species dubiae. Subsequently, therefore, M. intermedia was listed under doubtful species (D is n e y , 1991).
In the Museum Koenig a series of females and a male collected by Frey are placed under M. comosa, along with a specimen labelled M. intermedia but which has lost its abdomen. The male is one of the specimens listed by SCHMITZ (1936) and has therefore been designated the neotype of M. comosa. This specimen and a female also listed by SCHMITZ (labelled "Gr. Although M. giraudii is the commonest and probably the most widely distributed species in its group it has never been adequately characterised. It is extremely similar to M. parnassia and can only be reliably distinguished from the latter by the new characters given in the refined keys above. It should be noted that Fig. 399 in DISNEY (1989a) was from a specimen that has proved to be M. parnassia, the hypopygium being tilted away from the observer. Fig. 400 of M. parnassia is with the hypopygium tilted towards the observer.

Description
As in SCHMITZ & DELAGE (1981) 40). Hypandrium as in Fig. 27. Left hypandrial lobe more or less translucent, at most slightly tinged greyish. Microtrichia of right lobe semi-erect to decumbent. Female ab domen: T6 1.20-1.45x as broad as long (Fig. 47). Anterior notch small, breadth one fifth to one seventh of breadth of tergite near anterior margin. T7 Y-shaped, shallowly not ched anteriorly (Fig. 49). T8 rectangular, with 9-32 hairs, mainly at lateral margins or close to it. T10 bare. S7 relatively broad, with (2-)3-4 enlarged hairs at apex (Fig. 64), sometimes anterior sixth to third with a stronger sclerotised median stripe. Sternum 8 micropubescent in posterior half to third, including area between posterior lobes. Ster nal lobes not sclerotised on internal surface, each one with a small, somewhat posterolaterally projecting bulge (Fig. 76). The latter with three (exceptionally four) enlarged hairs, that stand very close to each other on the apex of the bulge. Hairs of hypoproct notably crowded towards anterior margin (Fig. 94

Etymology
The species is named for the bare hind margin of the left hypandrial lobe of the male hypopygium. Female. Similar to male but differs at follows: SPS less numerous than in male, beneath inner face only in 2-3 irregular rows. Labrum stronger than in male, about as broad as third antennal segment. Labella only with 17-21 spinules below. Anterior scutellar brist les relatively short, only 0.58-0.68x as long as posterior ones. Width of abdominal tergites gradually decreasing from T2 to T6. The latter is 1.2-1.3x as broad as long, and only a little narrower posteriorly than anteriorly. Width of anterior notch about one fifth of maximum breadth of tergite. Hairs of abdominal venter scarce and short except po sterior row of segment 6. T7 broadly expanded anteriorly, similar to M. hexanophila (Fig. 58). T8 rectangular, with 7-15 hairs. S7 relatively broad, distinctly expanded and with 4(-5) enlarged hairs posteriorly, similar to M. giraudii (Fig. 64). Sternum 8 covered in microtrichia except bare hind margin of posterior lobes (Fig. 85). Each lobe with a bulge that slighdy projects laterally. Bulges with three long hairs at or very close to apex. Hairs of hypoproct a little crowded anteriorly. Cerci about 3x as long as broad, with 5+1 long hairs and an apical group of 3(-5) short blunt hairs. Vaginal sclerite posteriorly with an irregular extension (Fig. 115). Dufour's crop mechanism similar to M. quadriseta (Fig.  126). (3-)4 rectal papillae. Wing length 1.6-1.75 mm. Cl 0.46-0.47. CR 2.8-3.1 : 1.4-1.9 : 1.

Megaselia haraldlundi
This species was only known from the type series (two males). Among new material the hitherto unknown female was discovered. It is described below.

Description
The original description can be supplemented as follows: Third antennal segment in female smaller than in male, diameter (0.11-0.14 mm) a little less than half frontal width. This species is unusual in having only 3-6 SPS in the third antennal segment (Fig. 8). SPS small to large (0.004-0.020 mm), normally situated beneath medial face of third antennal segment. The upper SPS are smaller than the lower ones. Sensilla at base of third aristomere well separated, exceptionally closely approximated. Labrum in female better de veloped than in male, 0.13-0.18 mm broad, broader than third antennal segment. Num ber of spinules below labella a little higher than usual, 21-34 in male, 20-30 in female.

Etymology
The species is named after its preference for traps baited with n-hexane.

Diagnosis
A distinctly coloured species with orange-yellow thorax, chestnut-brown head and brown abdomen, ant slightly lower on frons than al, about midway between upper sa and al. Third antennal segment of normal size, with enlarged but not very numerous SPS. Palps yellow with longest bristles slightly longer than maximum diameter. Labella not inflated. Notopleuron with two bristles. Notopleural cleft of male long but weakly developed. Anterior scutellars long and brisde-like in female, variable in male (normally hair-like but sometimes enlarged). Mesopleuron bare. Abdominal venter hairy on segments 3-6. Epandrium brown, with one enlarged hair on each side. Anal tube dusky yellow, longer than dorsal face of epandrium, and pointing downwards.

Description
Male. Frons brown, usually paler anteriorly, slightly longer than wide, with 50-90 hairs, and dulled by dense microtrichia. ant a little closer to eye margin than to median furrow, sa closer together than pre-ocellars. The latter as far apart from each other as from the medio-lateral. Cheeks bearing 2-7 hairs, jowls with 1-3 bristles. Third antennal segment brown, subglobose, maximum diameter about half frontal width. With about 10-20 SPS in a single irregular row around base of segment, diameter 0.012-0.031 mm. Arista brown, basal two segments paler. Pubescence of third segment slightly longer than diameter of basal joints. Sensilla on base of third aristomere normally well separated, exceptionally closely approximated. Palps pale dusky yellow, with 6-14, short to medium long bristles of (maximum length 0.05-0.09 mm), and 6-13 hairs. Prementum with two pairs of hairs. Width of labrum distinctly less than diameter of third antennal segment. Each labellum with 5 pseudotracheae, ca. 20 pale spinules, and some longer hairs below. Ground colour of thorax orange-yellow (except medium brown postnotum), normally with at least par tially brownish pleura. Typically the anterior part of the mesopleura (below and in front of the 'mesopleura! furrow') and a spot above the upper margin of the sternopleura and the hypopleura is paler than the rest of the pleura. Rarely the whole thorax is more or less infuscated. Anterior scutellar setae always somewhat longer than hairs of scutum, normally 0.3-0.5x (rarely up to 0.85x) as long as posterior bristles. Distance between the latter more than twice the distance between the anterior and posterior ones. Abdominal venter greyish. Hairs short except those at hind margin of segment 6. Tergites also short-haired but with 1-4 longer hairs at postero-lateral comer of T2 and a row of bristle-like hairs at rear margin of T6. Hypopygium as Figs 15-16. Epandrium with 13-24 hairs on the left side. Hind margin of epandrium continues some distance below anal tube in postero-ventral direction forming a distinct angle below the proctiger. Each cercus with 7(-8) hairs. Microtrichia below proctiger semi-erect, pointing rearwards. Left hypandrial lobe slightly curved medially (Fig. 26) Female. Similar to male but differs as follows: ant tend to be a little more approximated to upper sa than in male. Third antennal segment smaller, its diameter less than half frontal width. Number of SPS slightly reduced. Bristles of palps longer, the longest one 0.085-0.11 mm long. Labrum more strongly developed, distinctly broader than diameter of third antennal segment. Anterior scutellar bristles 0.45-0.85x as long as posterior ones. Width of abdominal tergites gradually decreasing from T2 to T6. The latter is 0.85-l.lx as broad as long, its anterior notch 0.15-0.27x as broad as width of tergite. T7 as Fig. 58. T8 rectangular, with 14-33 hairs that are mostly inserted near lateral margins. S7 with (3-)4 terminal hairs (Fig. 65). Segment 8 micropubescent throughout whole surface except margins of posterior lobes of sternum (Fig. 71). On ventral face microtri chia becoming smaller and sparser anteriorly. Posterior lobes of sternum 8 only slightly bulging. The enlarged hairs only narrowly separated from apical edge of bulge. Internal sclerotisation extending up to base of lobes. Epiproct with 4 long hairs at posterior margin plus 1-5 shorter ones on disc. Hairs of hypoproct crowded towards anterior margin. Cerci cylindrical, about 3x as long as broad. Each cercus with the normal 5+1 long hairs plus a group of (4-) 5-6 short, blunt-tipped hairs at apex. Vaginal sclerite only about 0.025 mm long, oblong, with a hole in posterior half (Fig. 102). Dufour's crop mechanism (Fig. 125). Wing length 1.75-2.6 mm. Cl 0.51-0.53. CR2.7-3.8 : 2.1-2.6 :1. 4-8 cal.

Distribution
At present only known from Germany and the Netherlands.

Etymology
The name refers to the resemblance to M. densior, but it not being this species. The prefix immo-meaning "by no means".

Diagnosis
A generally brown species. The ant about midway between upper sa and al. Third anten na! segment with numerous SPS of small to medium size. Palps mainly yellow with short bristles. Male labella not inflated, somewhat spinose apicolaterally and extending onto ventral faces. Notopleuron with two bristles but no notopleural cleft. Anterior scutellars reduced to hairs in male, long in female. Mesopleuron bare. Abdominal venter hairy on segments 3-6 (in female sometimes also segment 2). Epandrium with hairs and a more bristle-like hair each side. Anal tube dusky yellow and almost as long as epandri um. Terminal hairs of proctiger at least as strong as strongest hair of epandrium. Hypandrium with a shorter right and longer left lobe, the latter of which bears long, erect to recurved microtrichia. Female T6 subquadrate. T7 Y-shaped. T10 bare. S7 moderately broad. Posterior lobes of sternum 8 with three long hairs and without bulge. Long hairs of sternal lobes inserted well before hind margin. Vaginal sclerite of roughly semi-circu lar shape, and with a small ring postero-laterally. Legs pale brownish yellow to yellowish brown. Hairs below basal half of hind femur longer than those of antero-ventral row in distal half. Cl more than 0.44. Costal section 1 longer to shorter than sections 2+3, usually subequal, cc long. Tip of subcosta fades away before reaching R . Usually with a hair at base of Rs. With (l-)2-4 ax, of which the longest is subequal to cc of section 3.
Haltere knob largely pale yellow. For differences with regard to the extremely similar M. mimodensior see below this species.

Description
As diagnosis plus following amplifications: Male. Frons brown with dense microtrichia (and hence dull), and with 30-100 hairs, ant a little closer to eye margin than to median furrow, rarely midway between them. Pre-ocellar bristles about twice as far apart as upper sa, and almost equidistant from each other as each is from a medio-lateral. All four brisdes at approximately same level. Lower sa closer together and smaller than upper pair. Typically with three strong brisdes on each jowl and three weaker ones on cheek, but sometimes only two in each situation. Third antennal segment subglobose, pale to medium brown, as broad as half width of frons. SPS distributed around base of third antennal segment but most numerous beneath inner face, there being arranged in 3-5 irregular rows. Total number of SPS at least 30, diameter 0.005-0.013 mm. Arista brown, the two basal segments a little more than twice as long as wide. Swollen basal section of third segment slighdy shorter and usually with well separated sensilla. Cilia of third aristomere as long as diameter of basal joints. Palps pale dusky yellow with (4-)5-8(-9) increasingly robust brisdes, which are relatively short (the longest one being 0.045-0.070 mm long). The palps at most only extend to level of tip of labrum. The latter pale brown and its greatest breadth distinctiy less than greatest diameter of third antennal segment. Prementum with two pairs of hairs. Labella lightly tinged brown above, with 5 pseudotracheae, 13-21 spinules and some longer hairs below. Anterior scutellars longer than hairs of thoracic scutum, ca. 0.25-0.4x as long as posterior scutellars. Abdomen with dark brown tergites and brown venter. Hairs of tergites generally short, but a patch of longer ones postero-laterally on T2, and the posterior row on T6 are longer and stronger. Posterior row of hairs of venter of segment 6 also more brisde-like. Hairs of hypopygium weaker than those at rear of segment 6 (Fig. 38). Epandrium brown, with 8-24 hairs on left face. Hind margin of left face abrupdy bent forward at ventral base of anal tube, forming an acute angle. Cerci with seven hairs that are at most as strong as strongest hair of epandrium. Micropubescence below proctiger semi-erect, pointing rearwards. Left hypandrial lobe 0.89-1.18x as long as proctiger (the length of the latter is measured at lower margin as the distance from the base to the insertion point of termi nal hairs). Microtrichia of right lobe shorter than those of left one, semi-erect to de cumbent. Internally with 2-3 rectal papillae. All five front tarsal segments with a postero-dorsal hair palisade. Dorsal hair palisade of mid tibia almost extends three quarters of length. Tip of hind femur typically darkened. With 5-9 of the hairs below basal half of hind femur longer than those of antero-ventral row of outer half. Hind tibia with 6-12 differentiated postero-dorsals and simple spines in apical comb. Wing 1.15-1.95 mm long Cl 0.45-0.48. CR 2.2-3.7 : 1.5-2.3 : 1. Section 3 cc 0.09-0.15 mm. (2-)3-4(-5) cal. Hair at base of Rs well developed (0.032-0.069 mm), sometimes missing. All veins brown, with the more anterior veins (especially the costa) more yellowish brown and membrane lightly tinged brownish grey. First thin vein originates at or beyond fork of Rs. Haltere with greyish brown stem and pale yellow knob, which may be lightly tinged pale brown in parts. Female. Similar to male but differs as follows: SPS less numerous than in male, beneath inner face of third antennal segment only in 1-3 irregular rows. Labrum as broad as third antennal segment. Longest bristle of palps 0.073-0.093 mm long. Labella with 16-21 spinules below. Anterior scutellar brisdes 0.73-0.93x as long as posterior ones. Width of abdominal tergites gradually decreasing from T2 to T6. The latter is subquadrate (1.0-1.15x as broad as long), and only a litde broader anteriorly than posteriorly. Anterior notch small, its width being 0.1-0.2x tergal breadth. Hairs of abdominal venter scarce and short, except those at hind margin of segment 6. Hairs of segments 3-4 sometimes reduced to microscopic setulae. T7 Y-shaped, moderately expanded anteriorly, similar to M. giraudii (Fig. 49). T8 rectangular, with 6-20 hairs. S7 similar to M. hexanophila (Fig. 65), with 2-4(-5) long hairs posteriorly. Segment 8 covered in microtrichia except usually basal half to two thirds of ventral median area and bare margin of sternal lobes. Lobes with (2-)3 long hairs, and margin between the lobes bordered by 4-8 smaller hairs (Fig.  81). Lobes more or less flat, not bulging ventrally. Internal face of lobes sometimes slightly sclerotised and tinged brownish. Hairs below hypoproct a little crowded towards anterior margin. Cerci 3-4x as long as broad, with the normal set of 5+1 long hairs plus 2-4 short, blunt-tipped hairs near apex. Vaginal sclerite convex towards left side (Fig.  106). It is well sclerotised except for the faint ring which is connected posteriorly to the right side. Dufour's crop mechanism with long lobes, similar to M. giraudii (Fig. 122). 3-4 rectal papillae. Wing length 1.45-2.1 mm. Cl 0.46-0.495. CR 2.6-3.5 : 1.6-2.1 : 1.

R eco g n itio n
This species is keyed out as M. densior at couplet 152 of the key to British Megaselia (DISNEY, 1989a). It is extremely similar to M. mimodensiorwith which it occurs sympatrically in Scandinavia (for differences see diagnosis of the latter).

D istrib u tio n
At present known from Britain (England and Scotland), Germany, Sweden, and the Netherlands.

(Figs 5, 22)
Megaselia septentrionalis auctt., nec (SCHMITZ, 1919). DISNEY, 1986: 120;1989a. M. septentrionalis was removed from the British List after comparing a specimen attribu ted to this species by Colyer "with a slide mount of the holotype's hind leg and a female paratype's wing" (DISNEY, 1985). However, the wing belongs to the holotype of septen trionalis (a female), and the leg is from a misidentified labellaspinata male from Finland (see below) which had been labelled as a "Prototype d " of septentrionalis (probably by Beyer). Subsequently the species was restored to the British List following procurement of a series of males from Scotland (DISNEY, 1986)

Etymology
The name refers to the spinose labella.

Diagnosis
A dark brown species with brown legs and greyish brown wings, ant a little closer to al than to upper sa but slightly lower on frons than either. The relatively large third anten nal segment dark brown and with only a few small to medium-sized SPS internally. Palps dusky pale yellow to pale brown but with the field of sensilla on the external face darker brown. The labella enlarged a little, the undersides densely spinose but with some of the outermost spinules being greatly elongated. Two notopleural bristles but no cleft. An terior scutellar bristles variable, ranging from being distinctly longer to distincdy shorter than half length of posterior ones. Mesopleuron bare. Venter greyish brown, and with hairs on segments 3-6. Epandrium almost symmetrical, with numerous hairs posterolaterally and a submarginal transverse row at rear of dorsal half. Hairs of epandrium of subequal length, without an enlarged hair near lower margin. Anal tube yellowish brown and subequal in length to epandrium. The left and right lobes of hypandrium short, subequal in length, and mainly with semi-erect micropubescence. Longest hair of antero-ventral row of hind femur subequal to or clearly shorter than longest hairs below basal half. Cl 0.46-0.52. Costal section 1 subequal to or greater than 2+3. cc long. Sub costa ends free. With a minute hair at base of vein 3 and 3-5 ax. Haltere with brown stem and dusky yellow knob.

Description
Only male known. Frons wider than long (high), with dense microtrichia and with 52-96 hairs, ant closer to eye margin than to median furrow. Lower sa not as strong as upper pair, both pairs closer together than pre-ocellar bristles. Spacing of bristles of middle row variable, with the pre-ocellars being a little closer together than either is from a medio-lateral to being clearly further apart. All four bristles in an almost straight trans verse row or medio-laterals slightly higher on frons. With 5-6 bristles on cheek and 2-3 longer ones on jowl. The dark brown third antennal segment as Fig. 5, and internally with only about a dozen SPS (diameter 0.005-0.010 mm). Arista brown, cilia of third aristomere as long as diameter of basal joints. Sensilla on swollen base of third aristomere broadly separated. Palps with (6-)7-9(-10) differentiated bristles, and as many hairs.
Longest bristle about 1.6-1.7x as long (ca. 0.12 mm) as breadth of palp. The pale sensilla of external face conspicuous against the darker surrounds. Prementum with two pairs of hairs. The maximum width of the brown labrum just over half the greatest diameter of third antennal segment. The labella brown above, somewhat enlarged and with dense fields of about 95-120 pale spinules below. Up to six of those in the outermost row on each labellum are developed as longer fine rigid bristles with curved tips. Each side of scutum with a humeral, two notopleural bristles, an intra-alar, a post-alar and a prescutellar dorsocentral bristle. Anterior scutellar setae 0.35-0.85x as long as posterior ones.
Hairs of abdominal tergites short except at rear of T6 (Fig. 22). Venter hairy below on segments 3-6, with posterior row of 6 being stronger, and with a few fine hairs behind the spiracle on 6. Much smaller hairs behind spiracles on preceding segments. Hypopygium as Fig. 22. Cerci with seven hairs of about same strength as hairs of epandrium.
Terminal hairs of proctiger a little stronger than hairs of cerci. Micropubescence below proctiger inclined rearwards. The hypandrium mainly brown with pale posterior lobes. Microtrichia of hypandrial lobes mainly semi-erect and inclined rearwards or inwards, those at hind margins erect to recurved. Legs brown but front legs paler, especially tips of coxae. All five segments of fore metatarsus with longitudinal postero-dorsal hair palisade. Ratios of lengths of front tarsal segments about 3.0 : 1.6 : 1

Description
Only male known. The description given by DISNEY (1986) can be complemented as follows: Third antennal segment with numerous (>30) small to medium-sized SPS (0.008-0.013 mm). SPS distributed around base of segment but sparse except for inner face where they are arranged in ca. three irregular rows. Sensilla on swollen base of third aristomere usually broadly separated. Palps a little inflated, 0.073 mm broad, with 8-9 conspicuously shortened brisdes (Fig. 1). The longest of these only measures 0.054-0.069 mm. Labella normal, apart some longer hairs with 18-19 short blunt spines below. Notopleural cleft simple, straight in posterior half, and without distincdy sclerotised ridges. Hypopygium see DISNEY (1989a: Fig. 393). Left face of epandrium with 18-29 hairs, none of which is brisde-like and clearly stronger than the other hairs. Proctiger (measured on lower surface) a litde longer than dorsal face of epandrium. Micropube scence below proctiger semi-erect, pointing rearwards. Left hypandrial lobe shorter than right one, both with short decumbent microtrichia. Four rectal papillae. Wing length 1. Other material examined d , same data as holotype, hypopygium deformed (coll. MB).

Etymology
The name refers to the great similarity with M. immodensior.

Diagnosis
This species is extremely similar to M. immodensior. It can be distinguished from the latter only by the following characters: On the average smaller and darker. Anal tube of male shorter than in M. immodensior. Left hypandrial lobe shorter, and with shorter micropu bescence than in M. immodensior. Vaginal sclerite of female semicircular and convex to wards right side, of less irregular shape than in M. immodensior. T10 usually bare but sometimes with 1-2 hairs. Female sternum 8 completely covered in microtrichia (except bare hind margin of posterior lobes). Sternal lobes with small ventral bulges. Lobes of Dufour's crop mechanism shorter than in M. immodensior. On the average C l lower than in M. immodensior.

Description
Male. Differs fromM. immodensior as follows: ant normally a little closer to upper sa than to al or midway between them, rarely a little closer to al. Frons with 30-75 hairs. Hypopygium as in Fig. 39. Epandrium with 7-20 hairs. Anal tube pale to medium brown. Left hypandrial lobe 0.55-0.75x as long as proctiger (the length of the latter is measured at lower margin as the distance from the base to the insertion point of terminal hairs). Microtrichia of left lobe moderately long (shorter than in M. immodensior), decumbent to semi-erect, those in outer half pointing rearwards, those beneath inner half usually cur ved medially. Legs brown, hind femur a little darkened towards tip. Longest hairs below basal half of hind femur a little longer to subequal to those of antero-ventral row in distal half.  Fig. 63, with (2-)4 long hairs posteriorly. Segment 8 completely covered in microtrichia except bare hind margin to sternal lobes (Fig. 82). Lobes similar to M. immodensior but more bulging. Each lobe with 3 enlarged hairs, and margin between the lobes only bordered by 2-3 smaller hairs. Vaginal sclerite (Fig. 107) more slender than in M. immodensior, with more regular surface, and convex towards right side. The weakly sclerotised ring on right side smaller. Dufour's crop mechanism similar to M. citrinella (Fig. 124). Wing length 1.25-2.15 mm. Cl 0.45-0.475. CR 2.6-3.5 : 1.6-1.9 : 1. Up to six cal.

Distribution
At present only known from Sweden and Norway.

Description
The original description is to be complemented as follows: ant a little closer to al than to upper sa and closer to eye margin than to median furrow. Third antennal segment of usual size, diameter a little less than half breadth of frons. Only ca. eight SPS of small to medium size (0.004-0.011 mm) beneath inner face of third segment. Palps not inflated, dusky yellow. Longest bristles of palps almost twice as long as maximum diameter of palp. Labella enlarged, each with ca. 90-100 short spinules below. Notopleural cleft straight in posterior two thirds and without strongly sclerotised ridges (Fig. 14). Abdominal venter hairy on segments 3-6. Hypopygium as in Fig. 19. Left side of epandrium with ca. 23 hairs that are weaker than those of posterior row of T6. This side acute-angled at ventral base of anal tube. Anal tube brownish. Cerci with 7-8 hairs which are almost as strong as strongest ones of epandrium. Terminal hairs of proctiger a litde stronger than other hairs of hypopygium. Microtrichia below proctiger semi-erect, inclined rearwards. Left hypandrial lobe longer than right one, and with erect to recurved micropubescence. Micropubescence of right lobe predominantly semi-erect but with some erect to recur ved microtrichia in marginal zone.

Distribution
Holarctic. The above mentioned specimens from Canada represent the first records of M. parnassia for the Nearctic Region. Other records are from Britain, Germany, Norway, and Sweden. M. parnassia seems to replace the closely related M. giraudii in boreo-alpine habitats of northern Europe.
The description of M. phoenicura-was based on males only, the females never having been reported in the literature. Unexpectedly, several females attributed to M. phoenicura were found in the Schmitz collection. While one of them is a M. hexanophila (of which the male is already known), two specimens from Austria were found to be M. quadriseta (only known in the female sex). Based on the great similarity of M. quadriseta/phoenicura it is very likely that they represent opposite sexes of the same species. The synonymy of M. phoenicura is thus being proposed. M. hajlejensis (only known in the male sex) is synonymised with M. quadriseta because of the identical hypopygium and enlarged SPS of the third antennal segment. The only significant difference between the two species is the coloration of the thorax (almost uniformly brown in M. hajlejensis, largely reddish brown in M. phoenicura). As a series of M. quadriseta females taken by MB (see above) also shows a high variability with regard to this character (but identical genitalia) it seems of little taxonomic value.

Distribution
M. quadriseta seems to be widely distributed in Europe. Revised material was collected in Austria, Britain, Denmark, Germany, Ireland, the Netherlands, Norway, Spain, and Poland. Due to confusion with M. hexanophila sp. n. and M. citrinella sp. n. older records need confirmation. Based on a misinterpreted data label M. quadriseta was erroneously repor ted from Madeira (SCHMITZ & DELAGE, 1981;and subsequently DISNEY, 1991).

Etymology
The name refers to the reduced number of SPS in the third antennal segment.

Diagnosis
Male, ant slightly lower than al or upper sa, about midway between them or a little nearer latter than former. The brown third antennal segment not enlarged, with only about a dozen small to medium-sized SPS inside. Palps yellow with short bristles. Labella densely spinose below. Thorax brown, with two bristles on notopleuron and no cleft. Anterior scutellars not quite half as long as posterior pair of bristles. Mesopleuron bare. Abdo minal tergites, venter and hypopygium (apart from left hypandrial lobe and anal tube, which are more yellowish) brown. Venter hairy on segments 3-6. Epandrium with a differentiated hair on left side. Anal tube yellow, lightly tinged brown. Right lobe of hypandrium clearly shorter than left lobe. The latter bears erect to recurved microtrichia below. Legs pale yellowish brown. The longest hairs below basal half of hind femur scarcely longer than those of antero-ventral row of outer half. Costa just under half wing length. Costal section 1 longer than sections 2+3. cc long. Hair at base of Rs about 0.05 mm long. 3-4 ax. Halteres with dusky yellow knob.

Description
Frons brown, broader than long (high), with 56-58 hairs and fine but dense microtrichia. ant about midway between eye margin than to median furrow. Lower sa at most 0.6x as long as upper pair. Pre-ocellars slighdy further apart than either is from a medio-lateral bristle, and all four in an almost straight transverse row. Two brisdes on cheek about as strong as lower sa. 2-3 stronger brisdes on jowl. Antennae litde narrower than half frontal width, greatest diameter less than 0.15 mm (Fig. 4). The SPS beneath its inner face as Fig. 7, diameter 0.004-0.014 mm. Sensilla at base of third segment of arista well separated. Palps with 6-9 brisdes, the longest one being 1.3-1.6x as long (0.07-0.09 mm) as diameter of palp. About 50-90 pale brown blunt spinules below each labellum. Grea test breadth of the pale brown labrum at most 0.1 mm, and thus less than greatest width of third antennal segment. Anterior scutellars a little longer than hairs at rear of scutum. Each side of latter, in addition to notopleurals, with a humeral, an intra-alar, a postalar and a prescutellar dorsocentral bristle. Hair of abdominal tergites 1-6 short and sparse, but stronger at rear of 6. The brisde-like hairs at rear margin of sternum 6 stronger than those at rear of T6. Epandrium with 14-21 hairs on left face, a postero-ventral lobe not being developed (Fig. 21). Cerci with seven hairs. Microtrichia below proctiger semierect, pointing rearwards. Microtrichia below right hypandrial lobe mostly semi-erect, with some at lateral and posterior margins erect to recurved. Internally with two rectal papillae. Postero-dorsal hair palisades on all five front tarsal segments. Ratios of lengths of fore-tarsal segments about 2.95 : 1.5 : 1.3 : 1.05 : 1. Near dorsal hair palisade of mid tibia extends just over three quarters of length. Hind tibia with a dozen differentiated postero-dorsal hairs, but only those in lower half are robust. Spines of apical combs of posterior face all simple. Wing 1.7-1.9 mm long. Cl 0.46-0.48. CR 2.6-3.7 : 1.4-2.4 : 1. Section 3 longest cc 0.11-0.155 mm. 3-6 cal. Hair at base of Rs is 0.051-0.062 mm long, and on one wing there is a second minute hair (only 0.02 mm long). All veins, including 7, yellowish brown. Membrane lightly tinged greyish brown. Haltere with dusky pale yellow distal two thirds of knob, a darker basal third and a brown stem. Female. Similar to male but differs as follows: Fewer SPS than in male. Greatest breadth of labrum about 0.11 mm and thus about 1.16x as broad as third antennal segment. Labella only with 28 blunt spinules. Anterior scutellars 0.84x as long as posterior ones.

Type material
Holotype: 9. Germany: Berlin, Pichelsberg, "15.viii.57" (the original description has 13.viii.1907), "Holotypus", "Aphiochaeta rubricornis Type 9 det. Schmitz" (remounted) (ZFMK). Paratype: the other type specimen mentioned by SCHMITZ is not in the ZFMK. Schmitz's personal copy of the original work, which is deposited in the ZFMK). Howe ver, a description of the male has never been published. One of the males in the Schmitz Collection is described below. It is very likely that Schmitz's determination is correct because the male shares two unique characters with the female (antennae reddish, and completely lacking the SPS).

Description
As given by SCHMITZ & DELAGE (1981) plus following amplifications: ant about mid way between al and upper sa. In the examined male there is only one pair of very un equal sa. Third antennal segment reddish and without SPS in both sexes. Diameter about half width of frons, same size in male and female. Arista longer than frons at mid-line. Sensilla at base of third aristomere not approximated (distance between them 0.008-0.012 mm). Longest bristle of palp 1.7x longer (ca. 0.12 mm) than diameter of palp. Labella normal, each bearing 32 (male), or 26-31 (female) blunt spinules. Thorax with two notopleural bristles and without cleft. Scutellum with four equal bristles in both sexes. Abdominal venter hairy on segments 3-6. Brisdes at rear margin of T6 stronger than any hair of hypopygium (Figs 23-24). Left side of epandrium with ca. 21 relatively short hairs. Hairs of epandrium weaker than stronger hairs of cerci. Left face of epandri um without postero-ventral lobe. Cerci with seven hairs. Micropubescence below proctiger predominandy semi-erect and pointing rearwards, partially erect in apical half. Left hypandrial lobe longer than right one, with erect to recurved microtrichia underneath. Micropubescence below right lobe semi-erect, pointing rearwards. Aedeagus with a short slender process that is a little longer than right hypandrial lobe. Process down-curved at tip. Abdominal venter of female hairy on segments 3-6, but hairs small and scattered except one enlarged hair at hind margin of segment 6. T6 subquadrate, a little narrower than T5. Breadth of anterior notch of T6 about one fourth to one fifth of breadth of tergite. T7 unusually short and broad (Fig. 57). T8 very small, with 11 hairs on lateral and posterior margins. T10 bare. S7 broad, anteriorly with a sclerotised median strip (Fig.  61). Posterior lobes of sternum 8 broad, with several hairs of subequal length, terminal hairs not clearly differentiated from surrounding hairs (Fig. 80). Lobes not bulging and without distinct sclerotisations. Whole of segment 8 covered in microtrichia except po sterior border of sternal lobes. Hairs of hypoproct not crowded towards anterior mar gin. Cerci 3x as long as broad, with usual 5+1 long hairs and an apical group of 5-6 short blunt hairs. Vaginal sclerite transverse, lying behind a lightly sclerotised cylindrical struc ture (ventral receptacle?) (Fig. 113). Dufour's crop mechanism large, emarginate po steriorly (Fig. 130). Fourth segment of front tarsus slightly shorter than fifth. Most hairs below basal half of hind femur longer than antero-ventral hairs in distal half. Wing length 2.75 mm (male), 3.0 mm (female Aphiochaeta sanguínea SCHMITZ, 1922: 98 ( 9 ) . Megaselia sanguínea (SCHMITZ). SCHMITZ, 1928: 145. Megaselia sanguínea (SCHMITZ). SCHMITZ & DELAGE, 1981: 707 (d1 9 ). Megaselia rubida auctt., n ec (SCHMITZ, 1918). M. sanguínea is another species which was described from females only. Six decades after its discovery DELAGE provided a description of a male found in the Schmitz collection (SCHMITZ & DELAGE, 1981). This specimen was determined by Schmitz himself as M. sanguínea but its discovery was never published. Judging from the (somewhat schema tic) figure of the hypopygium (SCHMITZ & DELAGE, 1981: Fig. 455) and some peculia rities of the description (colour of third antennal segment tending to reddish brown, left hypandrial lobe much longer than right one) it seems more probable that the inve stigated male belongs to M. hexanophila. This hypothesis is supported by the fact that a M. hexanophila female was collected at the same locality only five days later (Valkenburg, d : 20.V.1942(Valkenburg, d : 20.V. , 9: 25.V.1942. Unfortunately only an empty pin of the male was found in the ZFMK so its identity remains uncertain. Below, two specimens are descri bed which much more likely represent the male of M. sanguínea. In particular, they have the yellow third antennal segment of M. sanguínea, a character not found in any other species of the examined group.

Description
The description only includes characters which have not already been mentioned by SCHMITZ (1922) or SCHMITZ & DELAGE (1981): Position of ant approximately midway between upper sa and al. Third antennal segment clear yellow (male) to dusky yellow (female holotype), rather than 'pale red' as stated by SCHMITZ (1922). Third antennal segment of male only slightly larger than in female. Only with 1 (-2) rows of small to medium-sized SPS (0.004-0.011 mm) beneath inner face of segment, numbering about 10-20. Sensilla on base of third aristomere well separated. Labella not enlarged, with 39-42 short spines in male and ca. 22 in female. Thorax yellow, without notopleural cleft.
Anterior scutellar setae about one third (male) to two thirds (female) of length of po sterior ones. Abdominal venter with hairs on segments 3-6 (male) or 2-6 (female). Hairs short but numerous in female. Male hypopygium (Figs 36-37) very similar to M. immodensiorhut anal tube relatively shorter when compared to length of dorsal face of epandrium. Hairs of hypopygium weaker than those at rear of abdominal segment 6. Epandrium brown, with 14-15 hairs on left face. Strongest bristle on each side about as strong as hairs of proctiger, a little stronger than those of cerci. Hind margin of left side forming an acute angle at ventral edge of anal tube. Postero-ventral lobe of right side of epandrium less produced than in M. citrinella. Cerci less than half as long as dorsal face of epandrium. Microtrichia below proctiger semi-erect, pointing rearwards. Hypandrial lo bes with long and erect micropubescence, right lobe distinctly shorter than left one. Female abdomen: T6 about 1.25x as broad as long, breadth of anterior notch about one fifth of maximum width of tergite. T7 as in Fig. 52, but may be broader and distinctly notched anteriorly. T8 rectangular, with 23 hairs. T10 bare. S7 similar to M. hexanophila (cf. Fig. 65). Posterior lobes of sternum 8 broad and with many hairs (Fig. 86). The enlarged hairs of the sternal lobes inserted well before margin, relatively spaced. Micro pubescence of the lobes long, stopping shortly before the margin. Hairs of hypoproct not crowded towards anterior margin. Cerci of usual shape and chaetotaxy, 3x as long as broad, with four short blunt-tipped hairs at apex. Vaginal sclerite large (Fig. 116), roug hly semicircular and convex towards left side. Four rectal papillae in both sexes. Wing length 2.0-2.15 mm (male), 2.35 mm (female). C l 0.47-0.49 (male), 0.50 (female). CR 2.8-3.7 : 2.0-2.5 : 1 (male), not measurable in holotype female because of poor conditi on, but surely not 5 : 3 : 1 as stated by SCHMITZ (1922). Longest cc 0.15-0.175 mm. With 4-8 cal and 3-4 ax, longest of the latter slightly shorter than longest cc. Hair at base of Rs 0.065-0.100 mm.

Distribution
Hitherto only known from the Netherlands and (if the above described males in fact belong to M. sanguined) Germany.

Description
Only female known. The original description is to be complemented as follows: ant a little closer to al than to upper sa. Third antennal segment a little less than half as wide as frons, with a single irregular row of about twenty7 SPS around base. SPS medium to large size (0.009-0.030 mm). Cilia of arista about as long as diameter of basal aristomeres. Sensilla on swollen base of third aristomere well separated to closely approximated. Palps yellowish brown, longest bristle about twice as long as breadth of palp. Labrum a little broader than third antennal segment. Each labellum with 6-7 pseudotracheae and 40-51 spinules below. Notopleuron with two bristles. Abdominal venter hairy on seg ments 2-6, with those of segment 2 reduced to microscopic setulae. Hairs at hind mar gin of segment 6 enlarged. T6 trapezoid, 1.0-1.2x as broad as long. Width of anterior notch ca. one fourth to one fifth of width of tergite. T7 expanded anteriorly, at most shallowly notched, similar to M. densior (Fig. 55). T8 rectangular, with 17-25 hairs. T10 bare. S7 moderately broad, with 3-4 enlarged hairs posteriorly, similar to M. hexanophila (Fig. 65). Segment 8 micropubescent over whole surface except posterior border of sternal lobes (Fig. 91). Posterior lobes with (2-)3 enlarged hairs, at least one of which is well separated from hind margin. Lobes a litde bulging laterally. Hairs of hypoproct a litde crowded towards anterior margin. Cerci with 5+1 long hairs plus 3 (-4) short blunttipped ones near apex. Vaginal sclerite a small and lightly sclerotised ring (Fig. 104). Dufour's crop mechanism with two posterior lobes. (?3-)4 rectal papillae. Longest hairs below basal half of hind femur as long as, or longer than hairs of antero-ventral row in distal half. Wing length 2.4-2.9 mm. Cl 0.49-0.50. CR 2.7-3.7 : 1.4-2.1 : 1. Longest cc 0.15-0.17 mm. 5-9 cal, 3-6 ax. Hair at base of Rs small (0.034-0.054 mm) or absent.

Distribution
M. septentrionalis apparendy has a boreo-alpine distribution, at present being known from northern Sweden and the German Alps.

Description
Only male known with certainty. However, Megaselia sp. 1 females (see below) which were collected together with M. trojani males at two collecting sites could belong to this species. For description of the male refer to DISNEY & DURSKA (1998)

Distribution
Reported from Poland, Germany and Britain.

Etymology
The species is named for its preference for yellow water traps (see Table 3).

Diagnosis
A generally brown species with yellow palps and pale brown front and middle legs, ant a little closer to al than to upper sa, rarely midway between them or slightly closer to the latter. Third antennal segment only with a few small to medium-sized SPS. Palps nor mal. Labella of male enlarged and with dense fields of short spines on lower face. Notopleuron with two bristles and no cleft. Anterior scutellar setae usually short in male, long in female. Abdominal venter hairy on segments 3-6. Epandrium only with short hairs that are at most as strong as those of cerci. Terminal hairs of proctiger indistinctly stronger than those of cerci. Anal tube short, pale brown. Left hypandrial lobe shorter than right one. Hypandrial lobes with decumbent to semi-erect microtrichia. Female T6 subquadrate with broad anterior notch. T7 anteriorly expanded. S7 moderately broad.
Posterior lobes of sternum 8 each with a conspicuous laterally projecting bulge. Each bulge bears three enlarged hairs. Vaginal sclerite a large loop with an irregular posterior extension. Hind leg darker than middle and front leg. Longest hairs below basal half of hind femur a little longer than those of antero-ventral row in distal half, but relatively shorter and often subequal in female. Cl more than 0.44. Costal section 1 longer than sections 2+3. cc of normal length. Subcosta fades away in distal half. Normally with a hair at base of Rs. (l-)2-4 ax which are about as long as cc. Knob of haltere yellow, stem brown.

Description
Male. Frons a little broader than long, with 35-75 hairs and dense microtrichia. sa closer to eye margin than to median furrow. Third antennal segment subglobose, diameter less than half frontal width. Only 5-20 SPS which are arranged in 1 (-2) irregular rows be neath inner face of segment, absent or scarce on outer face. Diameter of SPS 0.006-0.016 mm. Arista brown, cilia a little longer than diameter of basal joints. Sensilla at base of third aristomere usually well separated. Palps with 5-8 bristles, the longest of which is about twice as long (0.09-0.13 mm) as breadth of palp. Prementum with two pairs of hairs. Labrum narrower than third antennal segment. Labella with 5 pseudotracheae, 60-120 spinules and some longer hairs below. Anterior scutellar setae longer than hairs of thoracic scutum, length about one third to (rarely) four fifths of posterior ones. Abdo minal tergites short-haired, with a group of somewhat enlarged hairs near postero-lateral corner of T2 and a row of long bristle-like hairs at rear of T6. Hairs of venter of segment 3 usually very small and those at rear of segment 6 strong. Bristles at hind margin of T6 conspicuously longer than any hair of hypopygium (Fig. 32). The latter appears somewhat laterally compressed with the surface of the left hypandrial lobe typi cally oriented in a more or less vertical plane. Epandrium with 16-31 hairs on left side. On this side hind margin of epandrium bent forward below ventral base of anal tube. Postero-ventral corner of epandrium devoid of microtrichia. Hind margin on right side of epandrium almost vertical to slightly convex below anal tube. Two rectal papillae (however, in most specimens the view on internal organs was obscured so this count might not be representative Female. Similar to male but differs as follows: Labrum strong, broader than third antennal segment. Labella hardly enlarged, only with 22-44 spinules below. Anterior scutellar bristles about two thirds as long as posterior ones. Width of abdominal tergites gradual ly decreasing from segment 2 to 6. T6 l.l-1.3 x as broad as long, width of anterior notch 0.3-0.45x breadth of tergite. Hairs of abdominal venter weak except those at rear of segment 6. Hairs beneath segments 3 and 4 often reduced to microscopic setulae. T7 moderately expanded anteriorly, similar to M. giraudii (Fig. 49). T8 rectangular, with 13-26 hairs. T10 bare. S7 somewhat variable, ranges from being as broad as in Fig. 64 to being as narrow as in Fig. 63. Posteriorly it is expanded and bears 3-4 long hairs. Posteri or half of sternum 8 covered in microtrichia, except bare hind margin of posterior lobes (Fig. 77). Bulges of sternal lobes each with three long and sometimes a weaker fourth hair on edge. Hind margin of lobes sclerotised, darker brown. Hairs of hypoproct not crowded towards anterior margin. Cerci 3-4x as long as broad, with 5+1 long hairs and a group of 3 (-6) short blunt-tipped hairs at apex. Vaginal sclerite relatively large, as in Fig. 114

Distribution
Only known from Sweden, Great Britain (Scotland) and Germany.
This could be the female of M. trojani.

Diagnosis
Generally similar to M. giraudii. Differs from this species as follows: ant midway between al and upper sa or a little closer to the latter. Labella with 15-19 blunt spinules below. Sternum 8 covered in microtrichia except sometimes small area near fore margin. Po sterior lobes indistinct, broadly separated (Fig. 75). The three enlarged bristles of these lobes further apart than in M. giraudii. Differs from all other species of the present work by possessing a third, median lobe between the two postero-lateral lobes. Vaginal sclerite absent. Posterior lobes of Dufour's crop mechanism arising from a relatively long com mon base (Fig. 127)

Diagnosis
Very similar to M. giraudii. Differs by the following details: ant a little closer to al than to upper sa. Labella with 13-17 spinules below. S7 narrow, similar to Fig. 67 but with four enlarged hairs at rear and 3-6 on disc. Posterior lobes of sternum 8 with relatively broad but not very much produced bulges (Fig. 92). Hind margin of sternal lobes translucent and delicate. Thus the bulges themselves appear to represent the hind margin of the lobes. The enlarged hairs of the bulges further apart than in M. giraudii. Hairs of hypoproct not crowded towards anterior margin. A very weakly sclerotised ring-like vaginal sclerite present (Fig. 105

Diagnosis
Very similar to M. giraudii. Differs from this species as follows: ant closer to al than to upper sa and closer to eye margin than to median furrow. S7 like in M. hexanophila (Fig.  65). Posterior lobes of sternum 8 with small and projecting bulges, internal surface sclerotised (Fig. 72). Hairs of hypoproct not crowded towards anterior margin. Vaginal sclerite developed as a weakly sclerotised ring (Fig. 101)

Diagnosis
Very similar to M. mimodensior, differs as follows: ant about midway between upper sa and al, sometimes closer to the former, sometimes closer to the latter. T10 usually bare, in one specimen with two hairs. S7 with 4(-6) long hairs posteriorly, similar to M. hexcmophila (Fig. 65). Sternum 8 covered in microtrichia except sometimes medial area of anterior third. Sternal lobes (Fig. 83) indistinguishable from M. mimodensior. Cerci with apical group of four short blunt-tipped hairs apart usual 5+1 long hairs. Vaginal sclerite narrower than in M. mimodensior, developed as a flat transverse strip, sometimes a little upcurved on right side (Fig. 108). Dufour's crop mechanism with long lobes (Fig. 128 65). Segment 8 micropubescent on whole surface, except bare hind margin to sternal lobes. The three long hairs of the latter are more separated (Fig. 90) than in M. giraudii. Cerci with apical group of four short blunt hairs apart usual 5+1 long hairs. Vaginal sclerite present, strongly sclerotised and with a hole on right side (Fig. 103). Dufour's crop mechanism similar to M. hexanophila (Fig. 125)

Diagnosis
Differs from M. gimudii as follows: Frons with ant slightly nearer to al than to upper sa. T6 l.lx as broad as long. The breadth of its notch is about one third of its maximum breadth. T7 Y-shaped (Fig. 53). T8 rectangular, with 10 hairs. S7 as in Megaselia sp. 4 (cf. Fig. 62). Sternal lobes of segment 8 only with two enlarged hairs posteriorly (Fig. 73).
Flairs of hypoproct only a little more crowded in the extreme median anterior position.
Vaginal sclerite fenestrate and strongly sclerotised (Fig. 100) Phorn nigrescens WOOD, 1910: 244 (c? 9 ). Aphiochaeta nigrescens (WOOD). BRUES, 1915: 125. Megaselia nigrescens (WOOD). SCHMITZ, 1928: 145. Specimens of several species of the M. giraudii complex were wrongly attributed to M. nigrescens prior to the designation of the lectotype (DISNEY, 1985) and the subse quent characterisation of the male (DISNEY, 1989a). Because of these confusions the records of M. nigrescens reared from fungi by I. Eisfelder (SCHMITZ, 1948) were cited as Megaselia sp. by DISNEY (1994). However, since then samples of Eisfelder's material were found in the spirit collection of the ZFMK by RHLD. Males and females of her M. nigrescens were therefore mounted on slides and the identity of the males confirmed. The recognition of the female, which was not adequately characterised by W OOD, is clarified below.

Recognition
SPS absent. The male hypopygium has been figured by DISNEY (1989a: Fig. 453). The lack of a postero-dorsal hair palisade on the fifth fore tarsal segment will distinguish both sexes from species of the M. giraudiildensior group. The simple Dufour's crop me chanism (Fig. 132), the characteristic vaginal sclerite (Fig. 120), and the short anterior scutellar setae will further distinguish the female. Lobes at rear of sternum 8 as Fig. 96. T7 as Fig. 56.

Distribution
Reported from England, Germany and the Netherlands.

Description
Differs from species of the M. giraudiil densior complex in possessing three notopleural setae and an ovipositor of the parasitic type. The original description can be comple mented as follows: Abdominal venter hairy on segments 3-6 (segment 2 only bearing some vestigial pale setulae). T6 peculiar, with extremely broad and deep anterior notch (Fig. 48). Ovipositor very long (ca. 1.3 mm), extruded almost as long as preabdomen. Segment 7 only with a few scattered setae. T7 oblong trapezoid (Fig. 59). T8 4x as long as broad, rectangular with rounded corners, and only with three hairs that are more or less directed anteriorly. T10 bare, with very long anterior apodemes (Fig. 98). Epiproct with about a dozen very fine and short setae, plus an apical pair of somewhat enlarged hairs. S7 wedge-shaped, expanded posteriorly, and with nine hairs. Its length is about 5x its width at posterior margin. Segment 8 devoid of microtrichia except anterior two fifths and margins of posterior lobes (Fig. 97). Each lobe only with 7-8 short hairs posteriorly. A row of 5-6 isolated hairs in front of these lobes extends about half-way towards anterior margin of segment. Segment 8 otherwise completely bare. Hypoproct with scattered short hairs on disc that are not crowded towards anterior margin. Apical pair of setae only slightly enlarged, a lateral pair of long hairs missing. Cerci stout, only about twice as long as broad. Dorsal surface of cerci only with four very short setae, lower side with four longer hairs plus three very short ones. Microtrichia restricted to lateral surface of cerci. Vaginal sclerite absent. Four rectal papillae. Wing length about 2.5 mm. Cl 0.52. CR 3.3-3.4 : 2.0 : 1. Longest cc at least 0.13 mm (most tips broken off). 5-6 cal and 3-4 ax, the latter slightly longer than costal cilia. Hair at base of Rs pale and vestigial (ca. 0.015 mm long).

Recognition
The female is easily recognised by its specialised ovipositor, missing vaginal sclerite and yellow third antennal segment. Most other Megaselia with an ovipositor of the parasitic type are distinguished from M. rubida by their hairy mesopleuron. Those with a bare mesopleuron differ from M. rubida in possessing very small anterior scutellar setae (M. nudipleura (BEYER, 1958)) or a differently shaped abdominal T6 plus a brown thorax (M. emarginata (WOOD, 1908)). M .funilla (SCHMITZ, 1956) (only male described) can be distinguished by its fine and short vein R^ (fork of wing vein Rs unusually small).

Distribution
Hitherto only known from Poland.

Type material
Holotype: <?. The Netherlands: Limbricht, 4.vi.l917 (pinned, right wing mounted on slide) (ZFMK). Note: the specimen is in very poor condition. Head, fore legs, one mid leg, one hind leg and left wing missing, left side of thorax eaten away. The holotype is the only specimen of this species in the Schmitz collection. A female tentatively attributed by Schmitz to M. ustulata is not that species. It differs from M. ustulata in possessing three notopleural setae.

Description
The original description is to be complemented as follows: Thorax with two notopleural bristles and without cleft. Anterior scutellar setae only about one fourth of length of posterior ones. Abdominal venter with hairs on segments 3-6. Epandrium on both sides with one hair that is a little but distinctly stronger than those of proctiger. Left side of epandrium extending far below anal tube, with at least 12 additional weaker hairs. Anal tube longer than dorsal face of epandrium. Each cercus with at least 14 hairs that are weaker than those of proctiger. Wing length 1.7 mm. Cl 0.51. CR2.6 :1.9 :1. Longest cc 0.10 mm. Subcosta fading away before reaching R . Three ax and three cal. No hair at base of Rs. Hairs below basal half of hind femur longer than those of antero-ventral row in distal half. Hind tibia with simple spines in apical comb.

Recognition
Easily distinguished from all other species mentioned in this paper by the combination of a very long anal tube and the high number of hairs on cerci. The yet unknown female will be distinguishable from M. sanguinea (which also has a yellow third antennal seg ment) by the missing hair at base of vein Rs.

Materials and Methods
Rearing experiments, baited and unbaited traps. Most specimens included in this revision were obtained by rearing and trapping experiments conducted by MB in 1992-94 (for more extensive information on the involved techniques see BUCK, 1997a, b). Experimental sites included a humid deciduous forest, a spruce stand, a meadow, and an arable field near Senden Hittistetten, Dist. Neu-Ulm, Bavaria, Germany (48°20'N 10°07'E). In rearing experiments a variety of baits (see Table 1) and some substrates collected in the field (tents containing caterpillars and pupae of the ermine moth, Yponomeuta sp. (Lep.); samples of dead wood) were used. Baits were exposed in the field and later set under laboratory emergence traps (see Fig. 1 in BUCK, 1997a) to collect the emerging fauna. Trapping involved bait traps (baits see Table 2), odour traps, emergence traps ('photo-eclectors') and pitfall traps. In odour traps three different chemicals were used: 10% 1-phenylethanol (1-PE) in n-hexane, pure n-hexane, and capronic acid. The use of 1-PE and n-hexane was inspired by the study of KAMM et al. (1987) who obser ved an attraction of certain Megaselia species to 1-PE.    Table 1. " -the second trap 20 weeks; b -the second trap 26 weeks. 400 cups filled with a 2-4 % solution of formalin were installed at ground level, at a distance of 7-20 m from each other. Half of these traps were yellow, half white, half were large, and half small (diameters see Table 3), thus counting 100 traps in each group. Traps were operated during 120 h, from 13. to 17.vii.1991. Eighty-five traps out of each group could be used for a quantitative evaluation, and their trapping results are presen ted in Table 3.

Results
Rearing experim ents. Seven species were obtained in rearing experiments conducted by MB in southern Germany (Table 1). M. giraudii, the most catholic species, was reared from nine types of substrates including different kinds of carrion (liver, kidney, snails, slugs, earthworms, insects) and rotten wheat grains. B ait, odour, em ergence and pitfall traps. In trapping sessions conducted by MB eleven species of the M. giraudii/densior complex were collected (Table 2). Except M. breviseta and M. cf. sanguínea, all species from the rearing experiments turned up in the traps as well. M. giraudii and M. densior predominated in baited traps. A total of nine species was caught in the latter, four in odour traps, and six in unbaited traps. Unbaited traps were least efficient in collecting flies of the studied group. Trapping with n-hexane and 1-PE mainly yielded M. hexanophila and M. giraudii. Unlike the study of KAMM et al. (1987), n-hexane (which should only act as a solvent for 1-PE) proved to be the effective component in attracting phorid flies. The experiments also demonstrated that the posi tion of the traps has a species-specific influence on trapping efficiency: e.g., kidneybaited traps installed 180 cm above ground yielded more M. giraudii and M. hexanophila than traps with the same bait installed at ground level or at 60 cm. In contrast, trap catches of M. densior w etz almost evenly distributed over the three levels. Furthermore, it became obvious that traps were not equally attractive for both sexes: e.g., the sex ratios of M. giraudii and M. hexanophila were significantly biased towards females in elevated kidney traps (M. . Among the two types of woods under study the deciduous forest yielded a higher number of species than the spruce stand (eleven out of twelve species were collected in the former). Judging from the whole material availa ble some species seem to colonise a wide variety of habitats. E.g., M. parnassia was collected in extreme boreo-alpine habitats as well as in different types of mesophilic forests (see records below this species). Colour Preference. The experiment conducted by von Tschirnhaus et al. gave intere sting insights in the colour preferences of different species and sexes (Table 3). On the average large traps (which had a surface about twice as large as small traps) caught about twice as many specimens as the small ones (however, in M. xanthophila this number was more than ten times as high). Yellow and white traps yielded about the same number of specimens, except for M i xanthophila which distincdy preferred yellow over white traps (As only the group totals were recorded these results could not be tested for signifi cance). In the three most common species (M. parnassia, M. mimodensior, M. xanthophila) males always dominated the trap catches. However, in M. parnassia the male bias was only significant in white traps. Only in this species the sex ratio was influenced by co lour: white traps yielded a significantly higher proportion of males than did yellow traps. Trap size did not have any significant influence on the sex ratio of the studied species.
Other observations. The crop contents of a female of Megaselia sp. 7, which was col lected in a fen habitat in eastern England, includes a mass of detrital material. This includes recognisable diatom frustules, spores and fungal bodies. It seems likely that it had ingested a moisture film from a mud surface or the surface film on a water body. Water films themselves are nutritious, apart from the microorganisms trapped in the lipo-protein film (GUTHRIE, 1989).

Taxonomy
The results of the present study give a further example of difficult taxonomic problems within Phoridae being resolved by the rigorous use of genitalic characters and other microscopic details. Once more, traditional characters such as wing indices, coloration of thorax, and others used in previous keys (e.g., SCHMITZ & DELAGE, 1981) proved to be of little taxonomic value. On the other hand, the search for new, more reliable cha racters was especially fruitful in the group under study. The antennal SPS, for instance, proved invaluable in distinguishing similar species and in correctly associating males with females in sexually dimorphic species. Furthermore, in the female sex considerable interspecific differences in vaginal sclerites and sterna 8 were discovered. It is emphasi sed that these characters are absolutely necessary for species identification in the huge genus Megaselia and we strongly encourage to routinely describe and illustrate these struc tures when describing or redescribing species. As most of the reliable characters are internal or can only be observed under a compound microscope (e.g., the inclination of microtrichia on various parts of the male hypopygium) pinned specimens could only be identified after being remounted on slides.
The present study, as with a recent study of the M. pulicaria complex (DISNEY, 1999), further underlines the confusions that have resulted from type or voucher series that contain more than one species. Thus the original remounting of a voucher specimen of M. densioritom the Schmitz Collection (DISNEY, 1985), instead of the now remounted lectotype (designated above) that proved to be a different species from the voucher specimen, set in train a set of misidentifications and the creation of unnecessary syn onyms.
As a phylogenetic approach was beyond the scope of this work we only want to briefly comment on some issues here. While the phytogeny of the Megaseliini (of DISNEY, 1989b; = Megaselia group of BROWN, 1992) is far from being understood it is currently undisputed that one monophylum within this tribe can be recognised by the presence of a notopleural cleft in the male sex (DISNEY, 1989b;BROWN, 1992). This group (the Gymnophora subgroup of BROWN, 1992) comprises part of the paraphyletic genus Mega selia, as well as the genera Gymnophora and Woodiphora. However, the present work provi des evidence that the notopleural cleft may have been lost in some lineages within the Gymnophora subgroup. An interesting example for this could be M. mimodensior and an undescribed species 1 which were discovered during the present revision. These two species are so similar that they can hardly be distinguished from each other except for the presence/absence of the notopleural cleft. The most parsimonious explanation for this finding is a supposed loss of the cleft in one of the species (viz. M. mimodensior). The occurrence of reversals in such an important character further complicates the systema tization of the Megaseliini. Future studies will have to uncover further synapomorphies in order to clarify the boundaries of the Gymnophora subgroup within Megaseliini.

Biology
R earing experiments. Very little was known about the biology of the studied species group, except for the common and widely distributed M. giraudii. This species has been reported to breed in a great variety of decaying substrates such as dead or moribund insects, dead snails, vertebrate carrion, fungi, insect frass, vole and dog dung, nests of social Hymenoptera (wasps, bumble-bees, bees) and anthropogenic substrates (e.g., im properly washed milk bottles, Drosophila rearing medium) (DISNEY, 1994). M. densior is the only other species for which a larval breeding medium is reported in the literature. According to SCHMITZ (1927) it was found developing in (already dead?) pupae of the geometrid moth Bupalus piniarius (L.). For both species the saprophagous breeding ha bits are confirmed by the present study. Furthermore, it turned out that M. citrinella, M. hexanophila, M. hreviseta, and M. sanguinea are probably also saprophages (see Table 1). As for the species developing in the Yponomeuta samples it is not entirely clear whether their larvae were living as parasites or saprophages. Given the fact that the tents contai ned a high proportion of dead and moribund specimens (due to severe food scarcity and a high rate of parasitisation by hymenopterans) the latter possibility seems more likely. In any case it appears improbable that M. hexanophila and M. hreviseta (both exclu sively reared from ermine moth tents during this study) could be specialised on this host. As ermine moth caterpillars are only available during spring the summer generati on of flies must necessarily exploit another type of substrate.
'This is one of the species that key out at lead two of couplet 7 in the key to males (see note there). TRAPPING EXPERIMENTS. The attractiveness of n-hexane to M. hexanophila and M. girau dii was quite unexpected and remains to be explained. As certain other necrophagous phorids (e.g., Chaetopleurophora erythronota, Hypocem mordellaria, Gymnophora arcuata) were also attracted to these traps it might be assumed that n-hexane is a possible side product of the microbial degradation of carcasses. Perhaps n-hexane is one of the various cue substances that leads these species to possible feeding places or breeding substrates. If this is not the case n-hexane apparently mimics one of these substances and stimulates the same receptors.
The different colour preferences among species and sexes (cf. M. xanthophila, M. parnassid) will probably also relate to some specific behavioural traits. Yet, too little is known about the biology of these species to allow speculations on the significance of this behaviour.