Trichogalumna curva ( Ewing, 1907 ) Ewing, 1907

Norton, Roy A. & Ermilov, Sergey G., 2017, Identity of the oribatid mite Oribata curva and transfer to Trichogalumna (Acari, Oribatida, Galumnidae), with discussion of nomenclatural and biogeographical issues in the ‘ curva ’ species-group, Zootaxa 4272 (4), pp. 551-564 : 552-560

Trichogalumna curva ( Ewing, 1907) comb. nov.

( Figs 1–31 )

Oribata curva Ewing 1907 , p. 113

Galumna curvum: Jacot 1933 , p. 263, emendation unjustified Pergalumna curvum: Johnston 1965 , p. 62

Pergalumna curva: Shaldybina 1975 , p. 355

Oribates tantillus Berlese 1908 , p. 7: Jacot 1933, p. 263 Galumna tantillum: Jacot 1933 , p. 263, emendation unjustified Diagnosis. Small species, length 315–400; thelytokous. Body surface not strongly textured; mostly microgranulate, with striation variably developed on parts of pteromorph and coxisternum, anal and genital plates essentially smooth. Setae ro and le of medium size, barbed, ro inserted ventrolateral to le; seta in distinctly shorter, smooth. Bothridial seta with long stalk and short, asymmetrically dilated, barbed head sharply tapered at tip. Notogastral setae short, thin; seta c present on pteromorph. Notogastral porose areas small, rounded, A1 and A2 close together, Aa equidistant between setae la and lm; cuticle above each area usually with slit-like grooves or irregular pits. Lyrifissure im closely anterior to A1. Median pore absent. Circumpedal carina directed at, but not reaching seta 3b. Epimeral setal formula: 1–0–2–2. Tarsus I with famulus inserted anterolateral to solenidion ω1; tibia IV with solenidion inserted in distal part of segment.

Redescription. Measurements (25 females, three undetermined): total length 315–400 (cotype 340); notogaster width 215–300 (cotype 250).

Integument. Body color brown ( Fig. 5 ). Cuticle surface mostly densely microgranulate, granules rounded with diameter up to 1 ( Fig. 17 ); larger granules and ridge-like striae ( Figs 28, 29, 31 ) distributed as indicated below.

Prodorsum ( Figs 1–3 ). Rostrum broadly rounded, with medial lobe hidden in dorsal view; cuticle of lobe with deep posterior excavation, into which strong medial tooth projects ( Figs 25–27 ; see Remark 2); rostrophragma relatively thick, subtriangular in outline ( Figs 26, 27 ). Lamellar (L) and sublamellar (S) lines distinct, parallel, curving backwards at ventral end; with dorsal end of S continuing into leg niche ( Fig. 3 , N). Rostral (ro) and lamellar (le) setae similar in length (32–36), barbed, well medial to L; ro inserted ventrolateral to le, not visible in dorsal view. Interlamellar seta (in, 20) setiform, smooth. Bothridial seta (bs, 61–69) with long, smooth stalk and short, asymmetrically dilated, barbed head sharply tapered at tip ( Figs 6, 7 ). Exobothridial seta and porose area Al absent. Porose area Ad small, oval (~6 × 4), transversely oriented, posterolateral to in. Dorsophragmatic apophysis represented by group of low, roughened thickenings ( Fig. 1 , hy) collectively elongated longitudinally, with variable number of components.

Notogaster ( Figs 1, 3 ). Dorsosejugal suture absent, leaving notogaster fully fused with prodorsum. With 10 pairs of thin, smooth notogastral setae, c (~16) longer than others (10–12; Fig. 8 ). With transverse band of larger tubercles (diameter up to 2) anterior to pair lp; Florida specimens often also exhibit small central region with tubercles merged into thick longitudinal striae ( Fig. 28 ). Four pairs of porose areas (Aa, A1, A2, A3) rounded, similar in diameter (6–9); A2 sometimes slightly smaller (4–6). Cuticle above each area usually with slit-like grooves, less commonly with irregular pits ( Figs. 9–12, 14, 15 , 19–24 ; see Remark 3), rarely only slightly depressed ( Figs 13, 16 ). Area Aa equidistant between setae la and lm. Median pore absent. All lyrifissures distinct: ia in posterior third of pteromorph (underside), im located closely anterior to A1 ( Figs 1 , 14 , 19 ), ip posterior to A3, ih and ips close together, anterior to p 3. Opisthonotal gland opening (gla) anterolateral to A2. Pteromorph surface texture variable within populations: usually mostly finely granulate with narrow band of striae parallel to edges of anterior lobe and wider band on posterior lobe ( Fig. 17 ); less commonly with distinct, widespread striae over most of surface ( Fig. 18 ).

Gnathosoma. Form typical for Trichogalumna (e.g. Ermilov et al. 2011; Ermilov & Tolstikov 2015). Subcapitulum size: 77–82 × 65–69. Three pairs of subcapitular setae setiform, thin, a (14–16) and m (10–12) slightly barbed, h (8–10) smooth; a thickest, h thinnest ( Fig. 2 ). Two pairs of adoral setae (8–10) slightly barbed. Palp length: 53–57. Axillary saccule distinct, slightly elongated. Chelicera length: 94–98. Cheliceral setae setiform, barbed; cha (36) longer than chb (20). Trägårdh’s organ long, elongate triangular.

Coxisternum and lateral podosomal region ( Figs 2, 3 ). Coxisternum distinctly textured: epimeres I, II microgranulate, with or without longitudinal striae in lateral region; epimeres III, IV with distinct long striae, forming band around anterior half of genital plate ( Fig. 29 ). Epimeral borders accentuated, their inner surface with pattern of short, well-spaced raised streaks ( Fig. 30 ). In ventral view, pedotectum I broadly rounded, pedotectum II rounded distally, discidium sharply triangular. Circumpedal carina (cp) thin, approaching insertion of 3b, but not reaching it. Epimeral setal formula: 1–0–2–2; setae thin, smooth, 4a (4–6) longer than 3b, 3c, 4a and 4c (8–10).

Anogenital region ( Figs 2, 3 ). Genital and anal plates without noticeable surface texture; ventral plate microgranulate, with transverse band of larger tubercles (diameter up to 2) between aggenital setae and anal plates ( Fig. 31 ). Six pairs of genital (g 1– g 3, 8; g 4– g 6, 4), one pair of aggenital (ag, 6–8), two pairs of anal (an 1, an 2, 8) and three pairs of adanal (ad 1– ad 3, 8) setae all thin, smooth. Aggenital seta inserted posterolateral to genital aperture. Adanal lyrifissure (iad) located close and parallel to anal plate. Adanal setae ad 1 and ad 2 postanal, ad 3 paraanal and lateral to iad. Distance ad 1– ad 2 shorter than ad 2– ad 3. Unpaired postanal porose area (Ap) oval (6–8 × 4). Ovipositor form typical of family (e.g. Ermilov 2010): elongated (102–110 × 28), with lobes (45–49) shorter than length of distal section (beyond middle fold; 57–61). Setae of lobes simple, smooth, ψ1 = τ1 (32) longer than ψ2 = τ a = τ b = τ c (20). Six coronal setae short (6), simple.

Legs. General form typical for Trichogalumna (e.g. Ermilov et al. 2011). Median claw distinctly thicker than laterals, all smooth on dorsal side. Porose area on femora and trochanters III, IV weakly defined. Formulas of leg setation and solenidia: I (1–4–3–4–20) [1–2–2], II (1–4–3–4–15) [1–1–2], III (1–2–1–3–15) [1–1–0], IV (1–2–2– 3–12) [0–1–0]; homology of setae and solenidia indicated in Table 1. Solenidion of tibia IV inserted in anterior part of segment. Famulus inserted anterolateral to solenidion ω1.

Remarks. 1. Generic concept. The genus Trichogalumna was proposed by Balogh (1960) with Pilogalumna (?) lunai Balogh, 1958 as type species. Generic characters were summarized by Balogh (1960), Engelbrecht (1972), Ohkubo (1984), and Villagomez & Palacios-Vargas (2013). The main morphological traits of Trichogalumna are similar to those of Pergalumna ( Ermilov et al. 2013) , but Trichogalumna differs by the presence of distinct notogastral setae, clearly emerging from the alveolus, in contrast to the non-emergent setal vestiges (sometimes referred to as ‘shaved setae’ or simply ‘alveoli’) present in Pergalumna . As noted by Aoki (1966) there are no other reliable distinguishing traits, and the generic diagnoses condense to this single character: the plesiomorphic presence or the apomorphic absence of distinct setae. It is simplistic, and analogous to various other single-trait distinctions among galumnid genera, for example that between Pergalumna and Galumna (based only on the insertion of lamellar setae relative to line L; Grandjean 1936). Generic concepts in this extremely diverse family remain to be assessed comprehensively.

2. Structure of rostral lobe. The cuticle of the rostral tectum of galumnid mites is often smooth and simple on its external (anterior) surface, but complex on its posterior surface (i.e., the surface within the camerostome), particularly in the region of the anteromedial lobe. This hidden complexity has been indicated in various literature descriptions by simple dotted lines, commonly represented as an M-shaped pattern when the rostrum is viewed anteriorly (e.g. Grandjean 1966, his Fig. 2 C), but the real meaning of these lines is usually unexplained. As shown in Fig. 27 the solid cuticle of the tectum is deeply excavated posteriorly in the region of the lobe, which therefore is mostly very thin. However, the excavation is not a simple fenestration: it leaves a conspicuous medial tooth (irt) that appears to project into the excavation from above. In anterior view the excavation and tooth are seen by transparency, and create the M-shaped outline ( Fig. 25 ). This complex excavation accommodates the tips of the paired rutella when the subcapitulum is elevated (e.g. during the defensive posture), and the tooth lies between them.

3. Structure and variation of porose areas. In most oribatid mites having an octotaxic system of porose areasthe cuticular part of dermal glands—the area is distinct only by transparency, with no external features that are visible in reflected light microscopy or scanning electron microscopy, except for a slight depression in some cases. There may be surface ‘pores’ but these are invaginations of epicuticle and do not pass through to the functional procuticular pores ( Alberti et al. 1997, their Figs. 12, 13 ). However, in T. curva the surface usually has a series of mostly parallel slits, immediately above the porose procuticle ( Figs. 9–12, 14, 15 ). A similar slitted structure occurs in the ceratozetoid mite Chamobates voigtsi (Oudemans, 1902; in Voigts 1902) and its ultrastructure is well known ( Alberti et al. 1997, their Figs. 12 f, 18b, 22a).

Leg Tr Fe Ge Ti Ta

Note: Roman letters refer to normal setae, Greek letters to solenidia (except ɛ = famulus). Single prime (’) marks setae on the anterior and double prime (”) setae on the posterior side of a given leg segment. Parentheses refer to a pseudosymmetrical pair of setae. Tr—trochanter, Fe—femur, Ge—genu, Ti—Tibia, Ta—tarsus.

In populations of T. curva , there is variation in the pattern, number and length of surface slits, and in some cases (mostly in Florida specimens) the surface pattern is more like irregular pitting ( Figs 20–22 ) than parallel slits. Some specimens from Illinois and North Carolina exhibit no surface slits or pits above any porose area of the octotaxic system; there is at most a slight overall surface concavity above these simple areas. The latter include two of Ewing’s non-type Illinois specimens—from Arcola ( Fig. 13 ) and Urbana—and several of the Jacot specimens from North Carolina ( Fig. 16 ). In each instance, all other traits of the mite match those of individuals with slitted or pitted porose areas. We found no examples of such simple porose areas in our Missouri or New York material. Whatever genetic or epigenetic mechanism controls the nature of the octotaxic system, it acts on all organs in an individual: either all eight are similarly slitted or pitted (with minor variations), or none are.

Names linked to Oribata curva . The web of uncertainty surrounding Oribata curva is difficult to untangle, but we make an attempt below, treating each of the involved names in succession. For each cited reference, we use the epithet-spelling and generic combination applied by the respective author.

Oribates tantillus Berlese, 1908 . Since Jacot’s (1933) proposal, this name has been considered a junior subjective synonym of Oribata curva Ewing, 1907 . The original Latin description was only four lines long, without illustration, and with the locality indicated only as North America (‘America boreale’). Later, Berlese (1914, p. 120, Fig. 3 ) published a slightly longer description, added a dorsal illustration, and noted the North American localities in more detail to include “Columbia, Washington , Florida ”. Norton & Kethley (1990) explained that these localities relate to Columbia , Missouri , Washington, D.C., and Lake City , Florida, which were the sources of most of Berlese’s North American oribatid mite specimens. In fact, Berlese and Ewing had the same source for specimens from Missouri, and a series of synonyms were the result. Norton & Kethley (op. cit) also noted that the original syntype series of O. tantillus included specimens from each of the three locations, and selected one of the Washington specimens as lectotype. Two other points seem important: first, Berlese (1914) noted having found the species also in Java, and second, he did not indicate from which country the illustrated specimen came. The latter point is critical, since his illustration and redescription do not match the original description in two regards. First, the bothridial seta (pseudostigmatic organ) is described and drawn as being lanceolate, whereas it was originally described as having an apical club. Second, the description and illustration note a duplicated porose area A3 on each side. No specimen we have seen from North America , including those in the Berlese collection ( Norton & Kethley 1990) have these two features. We suspect that his 1914 treatment of O. tantillus was based on the Java collection (not seen by us) and should not be relied upon without reexamination of his specimens; the mite appears to be an unnamed species of Pergalumna .

Jacot (1923) reported this mite—as Galumna tantillus Berlese, 1909 (sic)—from Tsingtao, China, in addition to two additional states of the USA, New York and Connecticut. These records are clouded by an absence of discussion concerning the two variant descriptions by Berlese. They also were odd in that the reported body lengths of all his specimens were more than twice the 320 µm indicated by Berlese (1908, 1914). Without discussion and with unjustified emendation of species epithets, Jacot (1933, p. 263) indicated that O. tantillum is a synonym of Galumna curvum . When Jacot (1934) redescribed G. curvum ( Jacot 1934) he still offered no explanation of the synonymy, nor did he mention the doubled porose area indicated in Berlese’s (1914) redescription of O. tantillus . In the same work, Jacot presented length data on specimens of G. curva from the locations in China that were mentioned in his previous work ( Jacot 1923), but with no reference to his earlier paper or to the fact that he used a different name: almost certainly, these data represent a correction, but were presented in a way that hid what probably was an initial error in micrometer calibration. After Jacot’s proposal, Berlese’s species name ( O. tantillus ) has rarely appeared in the literature except as a junior synonym of Ewing’s ( O. curva ). The only exception we know is that of Pérez-Iñigo & Baggio (1991), who noted that O. tantillus should be in Pergalumna , based on Berlese’s (1914) illustration.

Galumna ventralis Willmann, 1932 . This small species (360 µm) was described on the basis of specimens from Sumatra. [N.B. The date is often wrongly cited as 1931, which was that of the main journal volume; the supplement in which the name was proposed appeared in the following year]. Shortly after this description, Jacot (1934) considered Willmann’s mite a subspecies of Ewing’s: Galumna curvum ventralis (Willmann) . Jacot considered his Chinese specimens to represent G. c. ventralis , and his redescription and figures were based on these. He noted (1934, p. 116) that the subspecies was ‘very sparingly found’ in North America, but he gave no specific records and we have seen no specimens ourselves.

Jacot’s (1934) redescription of Galumna curvum was—for the time—relatively complete, but his brief comparative diagnosis of G. c. ventralis is problematic in a number of ways. One relates to his illustration style: he drew entire dorsal and ventral aspects (his Figs 26, 27 ), but the upper half of each purportedly represents the nominate G. curvum while the lower half represents G. c. ventralis . Indeed, the lower half of Fig. 26 shows porose area A2 as smaller than A1, a trait clearly identified in Willmann’s paper. However, according to Willmann’s Fig.

67, G. ventralis lacks emergent notogastral setae, other than seta c on the pteromorph, while Jacot’s concept of G. c. ventralis ( Fig. 26 lower half) has clear notogastral setae that are at least as long as the diameter of the larger porose areas. Similarly, Jacot’s text described the nominate G. curvum as having clear notogastral setae, yet his image ( Fig. 26 , upper half) shows only insertion points. It seems unlikely that Willmann overlooked small, fine notogastral setae, since he illustrated and described the small, fine prodorsal seta in. Another apparent discrepancy is that porose area Ad is at a level clearly lateral to seta in in Jacot’s figure, but medial to it in Willmann’s; we think this was an observational error by Willmann (see below). Jacot noted a more medial position of coxisternal seta 1a (‘gular bristle’ in his terminology) in G. c. ventralis than in the nominate G. curvum , but this character is not mentioned or drawn in Willmann’s paper. Other supposed differences were only vaguely stated by Jacot and seem impossible to decipher in either his illustrations or Willmann’s.

Decades later, Hammer (1972) added another part to the literature mystery around Galumna ventralis by recombining the name to Acrogalumna ventralis (Willmann) and redescribing it based on specimens from Tahiti. She did not highlight or explain this nomenclatural decision, nor did she mention Jacot’s (1934) opinions or his distributional records from China and the USA. To make this recombination, Hammer (op. cit) must have believed that prodorsal line L (perhaps better called ‘carina L ’) was absent, since its absence is a diagnostic trait of Acrogalumna . However, this line is clearly present in Willmann’s (1932, Fig. 67) illustration; it is the only prodorsal line of galumnid mites that shows as a carina—appearing like a point, midway along the prodorsal edge—where it curves over the contour when seen in dorsoventral aspect. In his Fig. 67, the line runs almost straight posteriorly from the ‘point’, toward the outside of the bothridium. Unfortunately, he also drew a second solid line running posteriorly from the ‘point’ more medially, close to seta in. The latter should have been a broken line, since it represents the surface contour of the lateral, concave, weakly defined pedofossa (‘leg cupboard’ in Jacot’s terminology) as seen by projection when studied under transmitted light (e.g., see Ohkubo 1984). Hammer’s (1972) figures are not as simple to interpret. She did not draw line L, but did draw its carina on the contour: a point clearly visible in dorsal (Fig. 44) and ventral (Fig. 44c) aspects; such a point does not exist without the presence of line L. She drew a solid line in Fig. 44, like the medial one of Willmann’s Fig. 67, but this clearly represents the lateral pedofossa, seen by transparency. Willmann thought seta le was absent from G. ventralis , with ro being the large seta extending beyond the rostrum in his Fig. 67. Jacot (1934) thought Willmann mistook le for ro, with the real ro being hidden ventrally. In contrast, Hammer’s (1972, Fig. 44a) interpretation was that le is minute in this species; this suggests that le, not ro was the seta missing from Willmann’s figure. Hammer’s Fig. 44a is a dorsolateral view showing both lines L and S; the two lines curve ventrad, seeming to meet at a point near the rostral margin in her figure. However, this is probably an artifact of viewing angle and projection (cf. our Fig. 3 ), and probably they remain parallel in the usual way. In our opinion, Hammer wrongly transferred G. ventralis to Acrogalumna because she misinterpreted the nature of the prodorsal lines.

If Hammer’s (op. cit.) identification of G. ventralis was correct, her figure (Fig. 44) clarifies the position of prodorsal porose area Ad: the small area has its typical position, posterolateral to seta in. We think the supposed porose area drawn by Willmann (1932, Fig. 67) in a more medial position actually represents the cluster of origins of cheliceral retractor muscles (hy, our Fig. 1 ), which Hammer correctly drew as an internal structure. Hammer’s figure also shows minute setae present in the p -row, while all other notogastral setae except c (= ta) and all ventral setae are non-emergent. Her ventral illustration (Fig. 44c) shows genital plates with the usual six pairs, rather than the two shown in Willmann’s Fig. 68: it is common for the anterior three pairs and even the posterior pair to be overlooked in galumnid mites, as they insert on the edge of the plate. Subsequently, Hammer (1973) reported Acrogalumna ventralis from Tonga Island but without additional information on morphology. Recently, Bayartogtokh and Akrami (2014) maintained this incorrect combination and did not question Hammer’s interpretations.

The next generic recombination was to Pergalumna curva ventralis (Willmann) , by Marshall et al. (1987, p. 340). This action was not explained, but was done essentially to be consistent with Johnston (1965), who first recombined the nominate Oribata curva to Pergalumna curvum (Ewing) , maintaining Jacot’s unnecessary emendation of curva . Finally, Ermilov et al. (2014) rejected the inclusion of this mite in Pergalumna , and transferred it to a fourth genus, Allogalumna . They cited Hammer’s (op. cit.) papers as evidence that line L is absent but, as reevaluated above, this is incorrect. In a recent species key to the world fauna of Allogalumna, Akrami (2015) used the combination A. (Allogalumna) curva ventralis (Willmann) ; he did not explain why only the subspecies was included.

Galumna duplicata Hammer, 1958 . This species (400 µm) was proposed based on specimens from Bolivia, and the species epithet refers to a doubling of notogastral porose area A3 ( Hammer 1958, Fig. 114); the doubling is drawn as being longitudinal (in tandem), not side-by side as in Berlese’s (1914) purported figure of Galumna tantillus . She later reported the species from Peru, and added a ventral figure ( Hammer 1961, Fig. 123) but no further discussion. Aoki (1966) then reevaluated the generic relationships of G. duplicata and proposed a Japanese subspecies, Pergalumna duplicata nipponica Aoki, 1966 . Unlike Hammer’s species, it was described and illustrated as having small, but distinct notogastral setae and a single porose area A3. According to Aoki (op. cit), Hammer reexamined the type specimens of G. duplicata at his request, and corrected two significant observations. First, she found a single porose area A3, not two, and considered the other structure (which of the two, was not indicated) to be a ‘gland’. This is odd, because other than porose organs and the opisthonotal gland, there are no glandular structures in that region of an oribatid mite notogaster. The ‘second’ structure may have been an unusually well-defined sigillum of the circumgastric muscle band, which passes across the notogaster in that region. Second, after reexamining her specimens in lateral view, Hammer corrected her original generic assessment, agreeing that G. duplicata is instead a member of Pergalumna . Although Aoki (1966) gave no specifics of this communication, her changed opinion must have resulted from seeing the presence of line L in the more favorable lateral aspect. In this reexamination, she still found no emergent notogastral or ventral setae.

Several years later, Hammer (1973, p. 54) considered G. duplicata a junior synonym of G. ventralis . Based on her descriptions and figures of both species, the synonymy seems correct. Inexplicably, in 1973 she reverted to the combination Acrogalumna ventralis , without commenting about Aoki’s (1966) discussion to the contrary.

Pergalumna duplicata nipponica Aoki, 1966 . Aoki (1966) discussed the problem of distinguishing Pergalumna from Trichogalumna , but found no unique traits that would apply to all known species; therefore, he tentatively included his Japanese mite in Pergalumna , distinguishing it from the nominate P. duplicata by the presence of small body setae. Both decisions were altered when Wen & Aoki (1983), without highlight or explanation, used the combination Trichogalumna nipponica (Aoki) . Ohkubo (1984) seems to have been unaware of this recombination and rank change in his careful treatment of Japanese Trichogalumna species, where these same changes were considered new. Ohkubo studied several Japanese Trichogalumna species that share a series of traits with Trichogalumna curva : all are small (400 µm or less), have a bothridial seta (sensillus) with a compact, asymmetrically swollen and distally tapered head, lack a dorsosejugal suture, have short and fine notogastral setae, and have small porose areas A1, A2 that are close together (within one or two diameters). He considered the type series of P. d. nipponica to include two species, distinguishable—like most species in his study—by their surface microsculpture (see below).

Figures

FIGURES 1 – 3. Trichogalumna curva (Ewing, 1907) adult (specimen from Columbia, Missouri): 1 — dorsal view (legs omitted); 2 — ventral view (legs omitted, gnathosoma represented only by slightly raised subcapitular mentum); 3 — lateral view (gnathosoma and legs omitted). Scale bar: 100 µm.

FIGURES 4 – 18. Trichogalumna curva (Ewing, 1907). Permanently slide-mounted historical adult specimens (see text). Photomicrographs (except 4) with transmitted, differential interference contrast illumination: 4 — cotype slide of Oribata curva Ewing, 1907; 5 — dorsal habitus of cotype, with egg on right side; 6 — distal part of right bothridial seta of cotype; 7 — distal part of left bothridial seta of cotype; 8 — left notogastral seta p 3 of cotype; 9 — left porose area Aa of cotype at surface focus, with insert below at deep focus; 10 — right porose area A 1 of cotype (surface focus); 11 — right porose area A 2 of cotype (surface focus); 12 — left porose area Aa of Ewing specimen from Missouri at surface (top, with base of seta lm) and deep (bottom insert) focus; 13 — porose area A 1 (surface focus) of Ewing specimen from Arcola, Illinois, showing absence of slits; 14 — porose area A 1 of Ewing specimen from Missouri at surface focus (and lyrifissure im), with insert at deep focus; 15 — right porose area Aa of Jacot specimen from North Carolina at surface (top) and deep (bottom insert) focus; 16 — same, but different specimen from North Carolina, showing absence of slits; 17 — partial right pteromorph of Jacot specimen from North Carolina (anterior to top); pteromorph of Ewing specimen from Urbana, Illinois (deteriorated glycerin medium, anterior to right), showing accentuated striae. Scale bars: 100 µm (5), 10 µm (6 – 18); 6 – 16 to same scale.

FIGURES 19 – 31. Trichogalumna curva (Ewing, 1907), dissected adult specimens. Photomicrographs with transmitted, differential interference contrast illumination: 19 — left porose area A 1 (surface focus) and surrounding region of specimen from Florida; 20 — same specimen, but right A 1; 21 — left porose area A 1 of second Florida specimen; 22 — left area A 1 of third Florida specimen; 23 — right area Aa of same specimen; 24 left area A 2 of same specimen; 25 — North Carolina specimen, anterior view of rostrum showing M-shaped inner excavation and inner rostral tooth (irt) by transparency (damage includes vertical crack * and broken edge of thin rostral lobe); 26 — same, but deeper focus to show rostrophragma (rp); 27 — same, lateral view, showing deep inner excavation of rostral lobe into which tooth irt projects; 28 — central region of notogaster, Florida specimen, showing transverse band of larger tubercles, with central region of longitudinal striae (arrow); 29 — left genital region of New York specimen showing fine granules and minute striae (arrow) on epimeres 3 and 4, and long striae near genital plate; 30 — same, but deep focus to show apodemes 3 and 4 (on either side of seta 3 b) and streaks (arrows) on inner cuticular surface in striate region; 31 — same specimen, showing transverse band of large tubercles between aggenital seta (ag) and anal plate. Scale bars: 10 µm (19 – 24 to same scale).