Reclassification of the Sack-bearer Moths (Lepidoptera, Mimallonoidea, Mimallonidae)

Abstract A backbone molecular phylogeny of Mimallonidae, based on 47 species and 515 loci, was recently published. That study resolved some of the major relationships in the family, but taxon sampling was limited and a classification of the family was not formally presented for all species. Here morphological phylogenetic analyses in parsimony and maximum likelihood (ML) frameworks were conducted that included 192 species and 55 morphological characters. A phylogenetic analysis was also conducted on the morphological dataset with a topological constraint based on the 515 locus tree from the previous study. Results show that nearly all species can be confidently placed in a genus using morphological phylogenetics. The presence of a frenulum, a character that was historically used to distinguish major groups of Mimallonidae, varies within and among genera. Based on our phylogenetic results, the classification of Mimallonidae is revised, which now includes 291 species in 41 genera. Descriptions of three new genera are included: Fatellallagen. n., Citrallagen. n., and Lepismallagen. n. The following taxonomic changes were made in the present article: 43 new/revived combinations (in Aceclostria Vuillot, Arcinnus Herbin, Bedosia Schaus, Bedosiallo St Laurent & Kawahara, Cicinnus Blanchard, Citralla, Druentica Strand, Fatellalla, Lacosoma Grote, Lepismalla, Mimallo Hübner, Procinnus Herbin, Psychocampa Grote, Roelmana Schaus, and Thaelia Herbin), two new species-level synonyms (melini Bryk is synonymized with viemanda Schaus, jaruga Jones is synonymized with hamata Walker), one revived synonymy (roscida Dognin is resynonymized with externa Moore), seven new statuses (in Druentica, Macessoga Schaus, and Trogoptera Herrich-Schäffer), six revived statuses (in Aceclostria, Cicinnus, Druentica, Psychocampa, and Zaphanta Dyar), and one new designation of nomen nudum. In order to alleviate nomenclatural problems, twelve lectotypes are designated (for Tolypidaamaryllis (Schaus), Trogopteraalthora Schaus, Adalgisacroesa Schaus, Alheitapulloides (Dognin), LacosomabriasiaSchaus, Lacosomadiederica Schaus, Lacosomaraydela Schaus, Psychocampalacuna (Schaus), Cicinnuscorallina Dognin, Cicinnuslatris Schaus, Cicinnussolvens Schaus, Cicinnustuisana Schaus) as well as a neotype for Mimallodespecta Walker (= Cicinnusdespecta). This paper also provides apomorphies for each genus and a morphological key to genera. Annotations are given to aid researchers in understanding all changes made herein, and images of male and female and their genitalia are present for nearly all type species.


Introduction
The sack-bearer moths (Mimallonidae) have been poorly studied in terms of their evolution, systematics, and natural history. The family is the only representative in the Mimallonoidea, and nearly all species are Neotropical (Lemaire and Minet 1998). Prior research on the family has focused largely on taxonomy of individual genera (e.g., Pearson 1951, 1984, St Laurent and Dombroskie 2016 and the only systematic treatment of the entire family was the study of Schaus (1928). Schaus described most of the taxa that are valid today, and since his work, only a handful of authors have studied the taxonomy and systematics of Mimallonidae (e.g., Pearson 1951, 1984, Franclemont 1973, Herbin 2012, St Laurent and Dombroskie 2015. Life histories of few mimallonids have been published (e.g., Dyar 1900, Wagner 2005, and adult behavior has been poorly studied except for one diurnal North American species (St Laurent and Carvalho 2017a).
Mimallonidae have extraordinary morphological diversity. Schaus (1928) discovered the variation in the presence of the frenula across mimallonid genera and divided the family into two subfamilies based on this trait. Pearson (1951) noticed that this trait is inoconsistent within genera, and Lemaire and Minet (1998) postulated that the presence of the frenulum is a symplesiomorphy, but neither study conducted a formal phylogenetic analysis. Since Schaus' (1928) mimallonid classification, two family-level checklists were published (Gaede 1931, Becker 1996, and in both cases, Schaus' classification was retained with only minor taxonomic changes. Because of the significant morphological variation in Mimallonidae, studies on mimallonid taxonomy utilized features of the male genitalia to define genera (e.g., Dombroskie 2016, St Laurent andHerbin 2017). However, whether these characters serve as apomorphies for genera has never been formally tested because a morphological phylogenetic framework for the family has been lacking.
The only formal phylogenetic study of Mimallonidae is the molecular analysis of St Laurent et al. (2018a), which sampled 47 species. Their phylogeny was well-resolved and most clades were robustly supported. Six subfamilies and eight tribes were established or received a revived status. These authors revised the classification of several genera, proposing 19 new combinations, three new genera, and one generic synonym. However, many genus-and species-level taxonomic changes were not made because taxon sampling was limited.
The current study aims to incorporate morphological data to assess apomorphies and revise the classification of the family. This article has four main goals: 1 Conduct morphological phylogenetic analyses of Mimallonidae. 2 Present a key to genera of Mimallonidae, based on male genitalia morphology. 3 Provide a generic classification of Mimallonidae with apomorphies based on new phylogenetic results. 4 Present a complete, annotated checklist of Mimallonidae that includes all species in the family.

Phylogenetic methods
In order build a comprehensive phylogeny of Mimallonidae (Goal 1), 192 species of Mimallonidae (66% of the family considering that 291 species are treated as valid here) were coded for 55 morphological adult characters (see Suppl. material 1: Table 1), including characters pertaining to male genitalia (25 characters), female genitalia (six characters), and external morphology (24 characters). All characters were coded as unordered, with 25 of them binary and 30 multi-state. Explanations of characters and their states can be found in Suppl. material 2, with illustrations of these characters and states in Suppl. material 3. We only included species for which we were able to dissect at least one male specimen, because male genitalia are generally the most reliable suite of characters for genus and species identification in the family (St Laurent pers. obs.). All Mimallonidae species sequenced for anchored hybrid enrichment (AHE) by St Laurent et al. (2018a) were coded for morphological characters so that a topological constraint could be applied to our morphological phylogenetic analysis. We chose this approach as morphological characters alone were not able to conclusively provide information on relationships among higher mimallonid groups (see Suppl. materials 5, 7). Two outgroups from the AHE study were also coded for morphology, the saturniid Citheronia splendens (Druce), and the pyralid Galleria mellonella (Linnaeus). Morphological phylogenetic analyses were carried out in a framework of maximum likelihood (ML) and parsimony. ML analyses were conducted in IQ-TREE v. 1.6.1 (Nguyen et al. 2015) using a Jukes-Cantor type model for morphological data + equal states frequency (option "-m MK+FQ"). We conducted two ML phylogenetic analyses that included morphology: 1 ML analysis of morphology that implements the higher-level topological constraint of St Laurent et al. (2018a) and 2 ML analysis of morphology that does not implement the higher-level topological constraint of St Laurent et al. (2018a). The -g option was used to perform the constrained tree search, using the dataset2_PART2 tree file from St Laurent et al. (2018a). The constraint tree topology is provided as a supplementary file (Suppl. material 4: constraint.tre). Nodal support in all IQ-TREE analyses was tested with 1000 ultrafast bootstraps (option "-bb 1000") (Hoang et al. 2018) and SH-aLRT (option "-alrt 1000") support.
For the parsimony analyses, we used TNT v. 1.5 (Goloboff and Catalano 2016) and conducted a New Technology Search, with default settings for sectorial, ratchet, drift, and tree fusing. Random addition sequences were set to 100 with the random seed set to 1 and all other parameters set to default. All characters were unordered and weighted equally. Nodal support was inferred using 100 jackknife replicates (Farris et al. 1996) using Traditional Search, and all other parameters set to default.

Taxonomic key, classification, and checklist
Our taxonomic key (Goal 2) is based on male genitalia morphology, developed from the male genitalia characters in our morphological matrix (characters 1-25). We created this key so that researchers can readily identify genera of Mimallonidae, and place new or unrecognized taxa in the according genus. Our revised classification (Goal 3) follows the results from the topologically constrained, morphological phylogenetic analysis (Fig. 1). The section is arranged so that subfamilies are listed in phylogenetic order from the root of Mimallonidae to the crown. Tribes are listed in a consistent order below their respective subfamilies, throughout the text. Genera and species are listed alphabetically for convenience. We figure habitus of a male specimen and male genitalia of species from all genera, including those of type species when possible. Female genitalia, while often bearing genus-specific features, could not be examined extensively due to the relative rarity of female mimallonid specimens in collections. We provide images of females and their genitalia for most genera, but because we lack data for many of them, we abstain from discussing generic apomorphies based on these features.
For the checklist (Goal 4), species that were not explicitly mentioned or sampled in St Laurent et al. (2018a) have been examined morphologically in order to determine their likely placement. In addition to taxonomic information for each species in the checklist, we also include the type locality given as the country and state/department (when known), as well as the natural history collection abbreviation where primary type(s) is/are located, when known. All names considered valid in the present checklist are listed in boldface italicized type, whereas synonyms, unavailable names, and nomina nuda are given in regular italicized type. Unavailable names are denoted by the " ‡" symbol, similar to the method of Fletcher and Nye (1982). Type species of each genus are signified by an asterisk.
Lectotypes are designated when syntypes from multiple localities exist for a given species, or when lectotypification is necessary to determine the identity of a species. Lectotypes for species described from syntypes from the same location, which we deem to be conspecific, will not be designated here, but rather later in more in-depth genuslevel revisions where interspecific relationships can be more robustly determined. A neotype is designated for Mimallo despecta since the original type(s) is/are almost certainly lost (see Becker 2001).  Laurent et al. (2018a). Branch lengths are removed so that relationships can be easily visualized. Asterisks denote nodes that were topologically constrained. For branch lengths and all support values, see Suppl. material 6.
Morphology terminology follows Kristensen (2003), Lemaire and Minet (1998), and St Laurent et al. (2018a). The term "valva apodeme" is derived from Zwick (2009). Dissections were performed as in Lafontaine (2004). Genitalia are slide mounted or stored in microcentrifuge/ microvials with glycerol. Genitalia dissection numbers are given in the legends for the respective figures, "diss." is used as an abbreviation for "dissection". Authorship of "St Laurent," although occasionally published "St. Laurent," is here standardized as "St Laurent" without the period as the author originally intended. All taxon authorships using this name should omit the period, though original article citations may include the period for standardized purposes. Furthermore, due the existence of multiple sets of different coauthors, with the first author St Laurent published in the same year, "St Laurent et al." references are further denoted by a letter following the year, even if the "et al." authors are not the same.

Phylogeny of Mimallonidae
Our morphological analysis which utilized a molecular topological constraint provides the most up-to-date picture on the evolutionary relationships of Mimallonidae (Fig. 1). This phylogeny presents many new relationships, especially within and between genera. All recognized mimallonid genera can now be placed within subfamilies and tribes (where applicable) using our phylogenetic results from this study, building on the results of St Laurent et al. (2018a). For example, all genera that were not sampled by St Laurent et al. (2018a)  Relationships between species were also well-supported, with essentially all species placed in the genera to which they are assigned in the present study as per the taxonomic sections of this article (Sections 3 and 4). Although the majority of genera were recovered as monophyletic in both the constrained and unconstrained ML trees (Fig. 1,Suppl. materials 5,6), some genera were recovered as paraphyletic/polyphyletic in these analyses. Such incongruence can be summarized as follows: in both the unconstrained and constrained ML analyses Alheita Schaus and Cicinnus Blanchard were paraphyletic; Lacosoma Grote, Druentica Strand, Procinnus Herbin, Psychocampa Grote and Robinson, and Roelmana Schaus were only rendered paraphyletic in the unconstrained ML analysis; with Bedosiallo and Isoscella recovered as paraphyletic only in the constrained ML analysis. Despite these issues, ongoing molecular phylogenetic work continues to support the monophyly of all Mimallonidae genera as they are treated in the present article (St Laurent et al. in prep.).
Generic clade recovery in the unconstrained ML analysis (Suppl. material 5) and parsimony analysis (Suppl. material 7) were not significantly conflicting with results from our ML constrained analysis, with the fundamental differences lying in the backbone topology that was constrained by molecular data. There were no major genusdefining nodes, that is, clades which include all representatives of a given genus, that were recovered with UFBS ≥ 95 and SH-aLRT ≥ 80 in the unconstrained ML analysis that were not also recovered in the constrained ML analysis.
Our results also demonstrate that morphological traits, such as the degree of development of the frenulum (Schaus 1928), is not a diagnostic character to distinguish subfamilies due to the variability of this character within genera (particularly in Mimallo and Druentica). Other seemingly robust external characteristics that would appear diagnostic of genera, such as the presence or absence of hyaline patches, antennae formation (filiform, dentate, bipectinate), also vary within genera (e.g., antennae structure within Roelofa Schaus). We note that one of the autapomorphies of Mimallonidae given by Lemaire and Minet (1998: 323): "distal section of male antenna provided with short rami (distinctly shorter than those of the proximal section) or even wholly devoid of rami" should be revalidated because this degree of rami shortening along male antennae is quite variable across Mimallonidae.

1
Uncus deeply bifurcated, bifurcation deeper than uncus width (Fig. 4a)  Genitalia structurally asymmetrical, such that phallus and single long tusk are oriented to one side of genitalia (Fig. 33a)  Sternite VIII with claw-like spines (Fig. 26a)  Valvae apodemes lobe-like, covered in setae (Fig. 27b)  Uncus apex and gnathos both significantly more heavily sclerotized than surrounding tegumen and situated very near each other, appearing pincer-like from the lateral aspect (Fig. 2a)  Gnathos arms with minute teeth ventrally, or with short protuberance near apex (Fig. 29c)  Mesal base of valvae with upturned projection, this projection usually spined (Fig. 6a)  Gnathos synscleritous with juxtal complex and/or connected to diaphragm by patches of setae (Fig. 20b), variable in configuration with variable extensions, subuncal projections also present (Fig. 20a)

Generic classification of Mimallonidae
For the generic classification of Mimallonidae, we follow the phylogeny in Fig. 1. Subfamilies are in phylogenetic order from the root of Mimallonidae to the crown in this tree, with genera listed alphabetically. For each genus, a brief diagnosis is given, allowing recognition of species belonging to the genus based on external morphology and maculation. Apomorphies based on male genitalia are provided to formally identify the genus and allow additional new species to be placed accurately in the genera as we have defined them. Three new monotypic genera, Fatellalla gen. n., Citralla gen. n., and Lepismalla gen. n., are newly described and therefore also include formal descriptions.

Cunicumara St Laurent, 2016 Figs 5, 77
Type species. Cunicumara anae  Diagnosis. Hoary appearance caused by interspersion of gray, pale-khaki, and dark-brown scales layered upon salmon to orange-brown, sandy tan ground color. Extremely long bipectinate antennae extending more than half length of the stout forewings, with distinctly long pectinations. Apomorphies.
Remarks. St Laurent et al. (2018a) did not include this genus in their phylogeny and instead treated this genus as incertae sedis based on morphology, due to similarities with the other incertae sedis genera. The morphological phylogenetic results of the present article support a close relationship between Cunicumara and other incertae sedis genera, particularly with Menevia and Roelofa (Fig. 1, see also Suppl. materials 5-7).
The female of C. anae is unknown, and thus we are unable to figure it or its genitalia.

Apomorphy.
(1) Paired, elongated, thin, tusk-like extensions reaching outwards from modified transtilla/juxtal complex which itself extends inward into the body cavity from attachments on either side of the inner costal apodemes of the valvae (Fig. 3a). Schaus, 1928 Figs 6, 43, 81, 82;Suppl. material 3: Plate 3 Type species. Perophora olivia Schaus, 1896. Diagnosis. Combination of the following two characters: forewing with black apical streak which forms a connection with a dark colored, straight or concave (or convex in some females) postmedial line and with pair of elongate darkly colored terminal abdominal tufts. Such tufts are seen in other genera but never in combination with the presence of a darkly colored forewing apical streak.

Roelofa
Apomorphy. Valvae with basal mesal flap-like extension which may be finger-like or more triangular and spined; valva flap extends into center of genitalia and curves outwards (Fig. 6a).
Apomorphy. Deeply bifid uncus such that bifurcation deeper than uncus width (Fig. 4a). Diagnosis. Orange-brown ground color with diffuse pink coloration on all regions of wings, with a tornal notch on the forewing only, anterior margin of the hindwing smooth. Similarly colored brown and pink Mimallonidae have a notch on the anterior margin of the hindwing (though this may be weak) as well as a usually deeper notch on the forewing tornus, and/or the pink coloration is not suffused across the whole wing in these species, but rather clearly delimited by the postmedial line. For example, see Fatellalla gen. n. below, a similarly colored genus with more clearly distinct medial and submarginal pink coloration.
Remarks. St Laurent et al. (2018a) did not include this genus in their phylogeny and instead placed this genus in Mimalloninae based on morphology, due to clear similarities to the genus Macessoga, a genus that was included. Our ongoing molecular work, which includes broadened taxon coverage of Mimallonidae, including Eadmuna, fully supports this placement (St Laurent et al. in prep.). Additionally, all of our morphological phylogenetic results consistently place Eadmuna as sister to Macessoga (Fig. 1, see also Suppl. materials 5-7), therefore confirming the placement of this genus in Mimalloninae. Schaus, 1928 Figs 9, 46, 87, 88;Suppl. material 3: Plate 2 Type species. Perophora fabia Druce, 1887. Diagnosis. Similar to Eadmuna, but ground coloration more yellow to yellowbrown, with continuous dorsal postmedial line, which is mostly straight (except for a sharp angle toward costa), not incomplete and crenulate as in Eadmuna.
Remarks. Tostallo was not included in St Laurent et al. (2018a), but was placed in Mimalloninae based on morphology due to the presence of all apomorphies of the subfamily. Furthermore, the shape of the uncus in Tostallo is nearly identical those of some Mimallo and Eadmuna, and the transtilla/valva apodeme configuration is similar to this structure these other genera as well. Our morphological phylogenetic analyses, except the unconstrained ML analysis, robustly support the inclusion of Tostallo in Mimalloninae, and we consider this evidence to confirm the subfamily placement of this unique genus. Ongoing molecular phylogenetics of Mimallonidae, which include Tostallo, support the placement of this genus in Mimalloninae (St Laurent in prep.).
Diagnosis. Similar in overall size and shape to Reinmara, but wings broader, margins more squared, ground coloration more earthen in tone, most species are light khaki brown, some darker. Males and females with bipectinate antennae.
Remarks. Although the morphology of Reinmara and Trogoptera are similar, we note that in ongoing molecular phylogenetics of the family, which now includes denser sampling of both genera than were available in St Laurent et al. (2018a), the two genera are quite divergent genetically. Eventually it will be worthwhile to include R. ignea which displays characters found in both genera (Reinmara genitalia but Trogoptera antennae structure) in order to clarify the phylogeny of Trogopterini.
(1) Phallus very long (longer than length from base of vinculum to tip of valva) and sharp, covered in fine spine-like setae (Fig. 12a), Suppl. material 3: Plate 6, 24: 4. Alheita Schaus, 1928 Figs 13, 50, 101, 162 Type species. Cicinnus anoca Schaus, 1905. Diagnosis. Small mimallonids with consistently short, triangular wings, always with a white postmedial lunule on the forewing, though this may be faint in species with accented veins in the broad, darker postmedial region (a character which itself is rather diagnostic of Alheita). Ventrally largely unmarked except for discal spot and significantly darker brown region of forewing delimited by the outline of the dorsal postmedial lunule. The similar genus Tarema is not so cleanly marked ventrally and has less falcate forewings. Apomorphies.
Remarks. We were unable to include a female A. anoca in our figures. Female Alheita are generally not largely distinct from males, with only the usual degree of sexual dimorphism (larger size, broader wings) which is usually observed in related genera.
Arianula Herbin, 2012 Figs 17, 95, 96 Type species. Arianula haxairei Herbin, 2012. Diagnosis. Arianula is the only other mimallonid genus besides Adalgisa which has irregularly scattered hyaline patches on all wings. However, Arianula is easily recognized by the alternate configuration of these hyaline patches, such that there are fewer patches on the forewing than in Adalgisa, with the most apparent patch being between Rs4 and M1 (conversely in Adalgisa, the most apparent patch is between Rs2 and Rs3); and a unique trio of rectangular patches on the hindwing between M3, CuA1, and CuA2, not observed in this arrangement in Adalgisa. Apomorphies.
(1) Bizarre claw-like valvae each similar in size to the tegumen + uncus, giving the overall genitalia a triangular appearance (Fig. 17a); (2) Phallus weakly sclerotized, tube-like ( Fig. 17b), with elongated narrow juxtal process dorsal to phallus, which itself terminates in a multi-pronged, spined upward curved claw. Type species. Cicinnus fatella Schaus, 1905: 326, by present designation. Etymology. The name for this new genus is derived from the type, and only known species belonging to Fatellalla: Fatellalla fatella comb. n. The name is feminine.
Diagnosis. This new genus can be recognized by the following combination of characters in the male (the only known sex): triangular forewings with pink antemedial and medial areas, which contrast with the orange-brown submarginal area. The submarginal area is clearly delimited by a pair of thin, straight, preapical, blackish brown postmedial lines. The rounded hindwings display the same patterning as the forewings. The coloration of the body is that of the ground color of the wings: orange-brown, with a distinct darker brown lateral line along either side of the abdomen, as well as a distinct dark brown tuft of elongated scales extending from the terminus of the abdomen. The genitalia are distinct due to the truncated, flattened uncus and the knob-like saccular process at the base of each valva, characters together not observed in any other known Mimallonidae. The gnathos plate is typical of alheitine Lacosominae; and the short, rounded valvae, narrow pair of valva apodeme arms, and thick bunch of elongate setae originating from the diaphragm are similar to those seen in Tarema and Alheita. In Fatellalla, the arms cross over each other mesally and are sharply tipped unlike in these other genera. Apomorphies.
Description. Male. Head: Tan-brown ground color, frons dark brown, eyes very large, occupying more than two-thirds area of head, bordered posteriorly by dark scales; antenna coloration dark tan, antenna almost entirely bipectinate, distalmost 9-10 antennomers dentate; labial palpus three segmented. Thorax: Coloration pinkish orange with red hue (fading to light brown in old specimens), scales covering prothorax grayer, contrasting against lighter remainder of thorax. Legs: Coloration as for thorax, vestiture thick, long. Tibial spurs elongate, covered in scales except for tip, roughly half length of first tarsal segment. Forewing dorsum: Forewing length: 15.5-17.0 mm, avg.: 16.3 mm, wingspan: 29.0-4.5 mm, n = 2. Triangular, margin mostly straight except slightly convex below apex and slight indentation at tornus. Ground color orange brown, but antemedial and medial areas evenly tinted with pink, faint gray suffusion present submarginally. Antemedial line absent; preapical postmedial line formed by two thin, parallel dark brown lines with medial area pink coloration between them. Postmedial lines slightly wavy basally. Pink coloration of medial area does not extend beyond postmedial lines. Discal mark present as pink ovoid region slightly lighter than surrounding medial area. Fringe poorly preserved in examined specimens, but darker brown than submarginal area. Forewing ventrum: Darker brown than forewing dorsum, medial pink suffusion bleeds into submarginal area, postmedial line convex and indented at intersections with veins. Hindwing dorsum: Coloration, patterning as for forewing dorsum, but discal mark absent, submarginal area comparatively wider than medial area. Hindwing ventrum: Following same pattern as forewing ventrum, but lighter overall. Frenulum present as single bristle. Venation: Typical of Mimallonidae. Abdomen: Dorsal coloration as for thorax, laterally very dark brown, nearly black, dark coloration continues on to elongated paintbrush-like tuft of dark-brown scales extending from terminus of abdomen. Vestiture thick, long. Genitalia: (Fig. 106) n= 4. Vinculum ovoid, ventrally rounded. Paired sclerotized arms extend outward from sclerotized base of costal valva apodemes. Uncus simple, truncated distally, rectangular. Gnathos well-sclerotized, forming mesally lobed plate below uncus, gnathos spanning tegumen. Valvae small relative to tegumen + uncus, valvae rounded, hardly extending beyond vinculum-tegumen juncture. Saccular edge of valvae with heavily sclerotized knobs near valval base. Diaphragm with dense setae extending outward above phallus. Juxta fused to phallus, encircling it, lateral margins of juxta extend one quarter length of phallus as triangle directed toward apex of phallus, ventral lip of juxta connects to vinculum. Base of phallus weakly sclerotized with undefined coecum phallus. Phallus cylindrical, downwardly angled distally, vesica spiculate. Female. Unknown Remarks. We describe Fatellalla for F. fatella, an Amazonian species widely distributed in Ecuador, the Brazilian Amazon, and French Guiana. This species is exceptionally rare in collections, with less than ten specimens known to us in global collections. Although St Laurent et al. (2018a) did not sample F. fatella, this species clearly belongs to Lacosominae: Alheitini due to the presence of the fused gnathos "plate" which is also observed in alheitine genera Thaelia, Tarema, and Alheita. The genitalia of Fatellalla are most similar to those of Tarema and Alheita (see ), the presence of elongated valvae apodeme extensions basally not unlike those of Tarema, support a close relationship between these genera. Our morphological phylogenetic analyses also support the placement of Fatellalla within Alheitini due to close placement with several other alheitine genera, namely Alheita ( Fig. 1, see also Suppl. materials 5-7). Although Fatellalla is nested within a poorly supported Alheita clade, we maintain this genus as valid pending molecular phylogenetic analyses due to the extremely divergent uncus shape and external appearance which is inconsistent with any observed Alheita. We note also a particularly long branch length (relative to other Alheita) in our ML analyses, emphasizing the divergent morphological characters of this odd genus.
Schaus ' (1928) Schaus, 1905. Diagnosis. Antemedial/ medial areas strongly contrast against the darker, chestnut brown postmedial/submarginal areas, these two distinct regions of color are divided by a pure white, sinuate postmedial line. Trogoptera mana Schaus displays a similar coloration scheme but the forewing has a distinct tornal notch and an anterior hindwing notch typical of Trogopterini, which is absent in Herbinalla. The particularly acute hindwing anal angle, which is accentuated by darker scales at the tips, is largely unique to this genus. Apomorphies.
(1) Seemingly floating sclerotized plate apparently homologous with plate-like gnathos of Alheitini, covered in fine setae, situated centrally within the diaphragm, plate not connected to remainder of the genitalia by any sclerotization (Fig. 15b); (2) The only alheitine genus with narrow and downwardly curved uncus ( Fig. 15a) (this itself is not an apomorphy due to elongated/narrow uncus in other mimallonid genera, however its presence coupled with other typical alheitine traits such as valva shape, are unique to this genus).  Schaus, 1896. Diagnosis. Hoary in appearance due to generous amounts of light gray scales present over the entirety of the wings and body. Tarema are similar to Alheita, particularly by the small size of the moths of both genera, and the postmedial lunule. However, the ventral surface of Tarema is not as plainly maculated due to the presence of postmedial lines and the light gray scales. Apomorphies.
Remarks. St Laurent et al. (2018a) did not include this genus in their phylogeny and instead placed this genus in Lacosominae: Alheitini based on morphology, due to strong similarities with the genus Alheita that was included. Our ongoing molecular work, which includes broadened taxon coverage, and Tarema, fully supports this placement (St Laurent et al. in prep.). Furthermore, the morphological analyses carried out in the present study place Tarema nested within Alheitini, either sister to Herbinalla (constrained ML analysis, Fig. 1) or sister to the remainder of Alheitini (unconstrained ML and parsimony analyses, Suppl. materials 5, 7 respectively).
Apomorphies. See diagnosis and remarks below.
Remarks. The definition of Thaelia here follows St Laurent et al. (2018a), based on phylogenetic placement of T. linamariae Herbin (type species of Thaelia), T. subrubiginosa (Dognin), and T. anysia (Schaus) within a robustly supported clade ( fig. 1 in St Laurent et al. (2018a)). All Thaelia species, sensu this work, were examined morphologically, and included in our morphological phylogenetic analyses, and all form a well-supported clade in all analyses (see Fig. 1, but also Suppl. materials 5 and 7). There are three distinct male genitalia patterns within this genus: (1) gnathos present, valva apodeme extensions absent, and cornuti present (T. linamariae, T. beniensis Herbin); (2) gnathos present, valva apodeme extensions present, cornuti absent (T. anysia); 3) gnathos absent, valva apodeme extensions present, cornuti absent (T. subrubiginosa, T. inornata (Druce)). We do not consider it warranted to further break Thaelia into separate genera at this time considering the consistent external morphology among these species. Ongoing molecular phylogenetics which include denser sampling of Thaelia further support a single genus concept for the species that we place in Thaelia. Etymology. The name for this new genus is derived from citrus (Latin) referring to the lemon-yellow coloration of the type species of rumina, the only recognized species in the genus. The name is feminine.

Lacosomini
Diagnosis. This new genus can be recognized by the following combination of characters: bright yellow coloration with gray and pink shading on the tornal region of the forewing and anal angle of the hindwing. The postmedial line of the forewing is faint, crenulate, and incomplete, existing only apically, as a small splotch halfway across the wing, and along the tornal shading. The hindwing displays similar maculation. Ventrally, the antemedial area of the forewing is shaded gray and pink, making Citralla and Zaphanta the only Mimallonidae genera with completely shaded antemedial regions of the ventral surface of the forewings. Citralla, however, lacks the ventral antemedial line present in Zaphanta. The prothoracic tibia has a prominent tuft of pink scales that is seen nowhere else in Mimallonidae. The male genitalia are simple, but unique in the absence of gnathos and transtilla projections, and by the simple triangular uncus and narrow valvae. The phallus is nondescript and largely similar to that of Lacosoma. The female genitalia are most similar to the related Vanenga, but display narrower papillae anales, ostium bursae, and ductus bursae. In Vanenga the confluence of the ostium bursae and ductus bursae is almost as wide as segment VIII, but in Citralla this part of the ductus bursae is only about one quarter the width of VIII, compare Figs 57, 119.
Apomorphies. Combination of the following characters: (1) Gnathos reduced to narrow bars below uncus which lack both mesal extensions and subuncus projections typical of related genera (Lacosoma and Vanenga); (2) Simple, smooth, triangular uncus and narrow valvae (relative to sharply triangular or extremely narrow uncus of Lacosoma and Vanenga).
Description. Male. Head: Gray-brown, eyes very large, occupying more than twothirds area of head, bordered posteriorly by dark scales; antenna coloration light tan, antenna bipectinate to tip, distalmost 10-12 pectinations significantly shorter; labial palpus three segmented, but segments difficult to discern due to compact scaling. Thorax: Coloration light yellow with scales along posterior prothoracic margin and junc-tion with abdomen very faint pink and gray, ventrum pale gray. Legs: Coloration variable, prothoracic leg: femur light purple-gray, tibia yellow with light gray scales before juncture with tarsus, prominent tuft of pink scales present on inner margin of tibia apex, tarsus yellow. Mesothoracic leg: femur and tibia light gray, tarsus yellow with some gray scaling apically. Metathoracic leg: all segments predominantly yellow with some gray scaling at terminus of tibia and apex of tarsus. Tibial spurs elongate, narrow, dorsally covered in scales, ventral surface and tip naked, length roughly half length of first tarsal segment. Forewing dorsum: Forewing length: 9-14 mm, avg: 11 mm, wingspan: 19-27 mm, n = 16. Triangular, margin nearly straight. Ground color light yellow. Antemedial faint, pink, irregular, antemedial area may be slightly suffused with pink; preapical postmedial line irregular, incomplete, existing only near tornus, apex, and halfway across length of wing as single splotch. Postmedial line outwardly shaded with gray and pink, particularly along tornus where pink suffusion reaches wing margin. Discal mark present as light gray ovoid splotch. Fringe checkered off-white and orange-brown, slightly crenulate. Forewing ventrum: Nearly identical to forewing dorsum, but antemedial line absent and antemedial area completely shaded by pink and gray, discal spot more pronounced. Hindwing dorsum: Coloration, patterning as for forewing dorsum, but discal mark faint or absent, antemedial line absent. Hindwing ventrum: Following same pattern as forewing ventrum. Frenulum present as single bristle. Venation: Typical of Mimallonidae. Abdomen: Dorsal coloration as for thorax, but slightly darker. Genitalia: (Fig. 118) n = 3. Vinculum ovoid, ventrally inwardly notched. Uncus simple, triangular, ventrally membranous. Gnathos and transtilla absent, but sclerotized bars extend downward from uncus/tegumen junction. Valvae narrow, triangular, simple; base of valvae extend centrally above vinculum base such that valvae cannot be fully spread. Juxta fused to phallus, encircling it, extending dorsally above phallus as flattened process. Phallus cylindrical, basally truncated. Female. Head: As for male, but antennae smaller overall. Thorax, Legs: As for male. Forewing dorsum: Forewing length: 11.5-15.0 mm, avg: 13.8, wingspan: 26-30 mm, n = 7. As for male, but slightly broader overall. Forewing ventrum: As for male, but slightly broader overall. Hindwing dorsum: As for male, but slightly broader overall. Hindwing ventrum: Following same pattern as forewing ventrum. Frenulum as multiple bristles. Abdomen: As for male, but more robust. Genitalia: (Fig. 119) n = 1. Tergite VIII forms smooth, thickened posteriorly directed arch, mesally with cup-like indentation at dorsal base of papillae anales. VIII weakly sclerotized laterally. Apophyses anteriores thick, truncated distally, roughly half length of apophyses posteriores which are outwardly bent halfway along length. Lamella ante-and postvaginalis poorly preserved, but weakly sclerotized without distinguishing features. Ductus bursae long, narrow, about three times the length of VIII-X, ductus widest at convergence with ostium bursae, but remaining very narrow along remainder of length. Corpus bursae small in length in comparison to elongated ductus bursae and large papillae anales, shape balloon-like. Papillae anales narrow, elongated, ventrally angled such that apical ridges of papillae anales and opening between them situated ventrally in nearly same plane as ostium bursae.
Remarks. Citralla is here described for the unique species C. rumina comb. n., which is distributed from Guatemala to southeastern Brazil. Further taxonomic investigations into the various populations of C. rumina will undoubtedly reveal several cryptic species, as we have observed slight external morphological distinctions in populations in Southeastern Brazil and the Amazon, in comparison with topotypical material from Panama and nearby Costa Rica.
Ongoing molecular phylogenetic work which includes Citralla consistently places this genus sister to Vanenga, which together form a clade sister to Lacosoma (St Laurent in prep.). Morphological analyses are less consistent in placement, with our morphological analyses recovering Citralla sister to (unconstrained ML analysis) or nested within Lacosoma (constrained ML and MP analyses). Regardless, tribal placement of the new genus is confidently in Lacosomini. The substantial reduction of the gnathos and juxtal configuration are most similar to those of both Lacosoma and Vanenga than to any other know Mimallonidae genus. But the uncus shape and pink scale tufts on the forelegs are unique to C. rumina. The larvae of C. rumina ( Fig. 120) have been reared and photographed by Janzen and Hallwachs (2017), revealing morphology remarkably similar to various species of Lacosoma, two species of which have been reared by the first author (St Carvalho 2017b, St Laurent et al. 2017). In particular, the striated appearance of the head and prothoracic shield, as well as the less rugose anal plate are largely consistent with Lacosoma as opposed to the more uniform coloration and generally more robust anal shield of other groups such as Mimalloninae and Cicinninae. We therefore, consider these characters as additional information supporting our decision to place Citralla within the Lacosominae: Lacosomini.
According to Janzen and Hallwachs (2017), the host plant of C. rumina in Costa Rica is Eugenia salamensis Donn. Smith (Myrtaceae). Myrtaceae is a host plant family that is frequently a larval resource for Mimallonidae (St Laurent et al. 2018).
Lacosoma Grote, 1864 Figs 20, 56, 111, 112, 163-165; Suppl. material 3: Plate 1 Type species. Lacosoma chiridota Grote, 1864. Diagnosis. Lacosoma is one of the most speciose genera of Mimallonidae, and displays a great deal of variation. They can easily be recognized by their genitalia (see Apomorphies below), but are externally more variable. Externally, Lacosoma are some of the smallest Mimallonidae in overall size, most species have crenulated wing margins and pink, salmon, and gray coloration. The general shape and size of Lacosoma species is more diagnostic than any one other external character. The combination of the general characters of small size, falcate forewings, and crenulated margins allow for the recognition of most species in the genus. Few Central American species display straight wing margins (such as L. elassa (Franclemont) and L. morgani Herbin), but are of the usual size and coloration for Lacosoma in general. The more uniquely colored species display the typical falcate wing shape and crenulate wing margins. Examples of uniquely colored species include those that are all pink: L. maldera Schaus; nearly all black: L. syrinx (Druce) and L. briasia Schaus (Fig. 163); and pink and yellow: L. valera Schaus and L. valeroides Herbin and C. Mielke. Apomorphies.
(1) Uncus as long (roughly length of valva), narrow spine-like in shape, not obviously triangular as in essentially all other mimallonid genera (Fig. 21a); (2) Subuncus projections similar in shape to those of Lacosoma, but originating from downward lateral projections of tegumen extending well below costal margin of valva ( Fig. 21b), as opposed to extending directly outward from ventral margin of tegumen above or very near costal valva margin as in Lacosoma (se Fig. 20a).

Druenticinae St Laurent & Kawahara, 2018 Druenticini St Laurent & Kawahara, 2018
Druentica Strand, 1932 Figs 22, 58, 121, 122 Type species. Cicinnus partha Schaus, 1905. Diagnosis. Most Druentica are silvery gray medium sized mimallonids with a distinct, usually straight postmedial line, and a preapical black dot along the costa where the postmedial line angles towards the costa, meeting it. Most species have straight forewing edges, but a few (such as D. rotundula (Dognin), D. muta (Dognin), D. mutara (Schaus), and D. brosica (Schaus)) have crenulated wing margins not unlike some Lacosoma and Mimallo, hence erroneous placement of several species of Druentica in these unrelated genera. Druentica is another of the larger mimallonid genera, and thus there is more interspecific variation in this genus than in other genera previously treated above. However, male genitalia of Druentica are immediately recognizable as such.
Apomorphies. Combination of the following characters: (1) Slightly curved tusklike arms extending from valva base (Fig. 22a); (2) Paired, mesally separated, elongated gnathos arms with distal flattening extend from immediately below uncus, gnathos arms roughly equal to length of uncus (Fig. 22b). Etymology. The name for this new genus is derived from the small, silvery coloration of the sole Lepismalla species, reminiscent of silvery Zygentoma in the genus Lepisma Linneaus. The ending -alla/-allo has been commonly applied in Mimallonidae. The genus name is feminine.

Lepismalla
Diagnosis. The single species of Lepismalla, L. montagnaniae comb. n., is recognizable by the almost complete lack of markings (except for faint irregular postmedial lines and heavy black discal marking) on the dorsal surface of the wings, which combined with the small size and falcate wing shape is unique in the family Mimallonidae. Genitalia are robust structures, with prominent gnathos arms that extend distally below the saccular margin of the valvae. In this way the gnathos arms are somewhat similar to those of related genera Procinnus Herbin and Micrallo St Laurent and C. Mielke, but extend much farther ventrally relative to the valvae. The basal valva arms typical of Druenticini are present, but highly reduced, and are flanked by setae covered regions of the sacculus. This genus lacks the claw-like sternite VIII extensions in the male genitalia which are typical of the related genus Procinnus, and has much shorter valvae tusks than either Procinnus or Micrallo. Apomorphies.
Description. Male. Head: Light brown, eyes very large, occupying more than two-thirds area of head; antenna coloration light tan, antenna entirely bipectinate, distalmost quarter length of antennae with pectinations markedly shorter; labial palpus apparently two segmented, though a third segment may be present but small. Thorax: Light brown, lightly speckled with darker petiolate scales. Legs: Coloration as for thorax, vestiture thick, long. Tibial spurs elongate. Forewing dorsum: Forewing length: 14.5-16.0 mm, avg.: 14.6 mm, wingspan: 28-35 mm, n = 6. Triangular, margin concave mesally forming falcate apex. Ground color silvery gray and light brown, overall lightly flecked with dark brown petiolate scales. Antemedial line absent; preapical postmedial line faint, outwardly convex, consisting of numerous barely distinguishable individual dark brown petiolate scales. Discal mark present as irregular black splotch heavily contrasting against light ground color and otherwise largely unmarked surface. Fringe light brown, lightening to cream near tornus. Forewing ventrum: As for fore-wing dorsum, but more brown than gray, anal region light tan, apex lighter gray than surrounding brown area; postmedial line slightly more well-defined; discal mark fainter than for dorsum. Hindwing dorsum: Coloration, patterning as for forewing dorsum, but discal mark much fainter to nearly absent. Fringe nearly white along most of wing margin. Hindwing ventrum: Following same pattern as forewing ventrum, but lighter overall, more of a continuation of light tan of anal area of forewing ventrum. Frenulum as single bristle, though difficult to see. Venation: Typical of Mimallonidae. Abdomen: Dorsal coloration as for thorax, ventrally lighter gray. Vestiture appearing thinner in comparison with thorax. Sternite of VIII with narrow pair of sclerotizations. Genitalia: (Fig. 131) n = 2. Vinculum ovoid, ventrally slightly pointed. Uncus simple, broad, triangular, distally narrowed to thin point. Gnathos massive, the most distinct aspect of the genitalia, gnathos heavily sclerotized with paired mesal arms which extend beyond and below saccular edge of valvae, gnathos arms distally and inwardly membranous, small tooth present at apex of each arm, gnathos arms fused mesally at base by narrow sclerotized band. Valvae small relative to gnathos, somewhat rectangular in shape, terminally with slight saccular lobe. Pair of small (length less than one third length of valvae), tusk-like curled sclerotized arms extend outward from saccular edge of valvae, tusk-like arms inwardly flanked by protruding setae covered sacculus. Juxta as thin sclerotization ventral to phallus forming connection with vinculum. Phallus curved when viewed laterally, viewed dorsally/ventrally phallus flattened distally with membranous opening such that phallus appears spade-like with sclerotized ring forming margins of spade shape, basal half of phallus as membranous sack. Female. Unknown Remarks. Lepismalla is described for the unique Amazonian taxon L. montagnaniae, which is so far outwardly and by male genitalia, unlike any other Mimallonidae, though is most similar to related genera Procinnus and Micrallo. In the original description, it was suggested that "Cicinnus" montagnaniae might be more properly placed in a new genus (Herbin 2012). After examining genitalia of all genera in prior to and in preparation of the present work, it is clear the genitalia of L. montagnaniae do not conform to any other generic concept of Mimallonidae, and certainly this species is unique externally as well. However, the subfamily and tribal apomorphies provided by St Laurent et al. (2018a) for Druenticini are consistent with the general male genitalia characters seen in L. montagnaniae. Our morphological and ongoing molecular phylogenetic analyses also consistently place this taxon as sister to the druenticine genus Procinnus, within a robustly supported Druenticini clade (St Laurent et al. in prep.).
The shape of the phallus of Lepismalla is very similar to that of other druenticine genera Micrallo and Procinnus. The phallus in all three genera is flattened with a unsclerotized central region in the distal half, which is clearly visible from the dorsal or ventral aspect. These genera also display a membranous sack-like region along the basal dorsal half of the phallus, though the size of this structure differs between the genera. Paired sclerotizations of the VIII sternite typical of Druenticinae, including Micrallo and Procinnus, are present in Lepismalla, but are reduced to narrow sclerotized strips in the otherwise membranous intersegmental region. The silvery gray coloration of Lepismalla is a coloration scheme seen in almost all Druenticini genera (except the darkly colored Ulaluma and some Procinnus), and the gnathos configuration is not unlike that of Procinnus and Micrallo, but simply more robust and ventrally farther reaching, such that in Lepismalla the gnathos arms actually reach below the saccular edge of the valvae. The basal valvae arms typical of Druenticini are present in Lepismalla, albeit greatly reduced.
Lepismalla montagnaniae is rare in collections, with only a handful of specimens known to us in global collections. Almost all examined material comes from the Amazon Rainforest, although one specimen, unfortunately destroyed in the fire at the Museu Nacional, Rio de Janeiro, Brazil, was from Mato Grosso do Sul, Brazil, in quite a different habitat on the border of Cerrado and Pantanal. An additional specimen in CPAC is from Distrito Federal in the Brazilian Cerrado. Future, finer scale, examinations of all known Lepismalla specimens may eventually reveal that this is not a monotypic genus, though it is certainly not diverse. Diagnosis. Tiny to small (forewing length: 11.5-15 mm) silver gray mimallonids, with elongate triangular forewings, and well-defined, straight forewing postmedial lines that are perpendicularly angled toward the costa, immediately separates this genus from others in the family.
Remarks. St Laurent et al. (2018a) did not include this genus in their phylogeny and instead placed this genus in Druenticinae: Druenticini based on morphology, due to strong genitalia similarities with the sampled genus Procinnus, particularly the flattened, spade-like phallus, which we now also observe in the newly described Lepismalla. All morphological phylogenetic analyses carried out here also place Micrallo within Druenticini, either sister to Ulaluma (parsimony analysis) or to Procinnus + Lepismalla (both ML analyses), thus confirming the tribal placement of St Laurent et al. (2018a).
An undescribed species that likely belongs to this genus is larger and browner than M. minutus, but otherwise bears the same arrangements of wing markings and genital configuration.
Pamea Walker, 1855 Figs 25, 62, 125, 126 Type species. Pamea albistriga Walker, 1855. Diagnosis. Quite similar to some of the smaller Andean Druentica species, but instead consistently have white outlined postmedial lines and white suffusions sub-marginally which contrast against a pale gray ground color finely stippled with darker petiolate scales. Pamea lack the black costal dot of Druentica.
Remarks. St Laurent et al. (2018a) did not include this genus in their phylogeny and instead placed this genus in Druenticinae: Druenticini based on morphology, due to clear similarities to the genus Druentica (particularly the gnathos shape), Druentica was included by these authors. As was the case with the two previously treated Druenticini genera not formally included in any molecular phylogenetic analyses, the male genitalia morphology of Pamea allow confident placement of this genus within Druenticini. Our morphological phylogenetic analyses confirm this placement, consistently placing Pamea sister to the remainder of Druenticini (both ML analyses). Our ongoing molecular work which includes Pamea, also support this placement (St Laurent et al. in prep.).
(1) Pair of elongated, sharp projections extend from either side of VIII sternite, which together with the basal valva arms typical of Druenticini and lowset gnathos arms, give the lower half of the genitalia an overall appearance of possessing a massive multi-fingered claw (Fig. 26a).
Remarks. We note that one undescribed species is known to the first author to belong to this genus based on clear male genitalia characters, but has significantly stouter wings than other Procinnus and a preapical postmedial line. The coloration and discal markings of this undescribed taxon are however, typical of this genus. Also in this and another undescribed Procinnus species known to the first author, there are fewer sternite VIII arms, but those that are present are the same shape and size as in other described taxa. This combined with the valvae and gnathos shapes are not unlike those of more typical Procinnus.
Apomorphy. The combined absence of the gnathos and the presence of valvae tusks (Fig. 24).
Apomorphies. Combination of the following characters: (1) Rectangular valvae, distally concave along margin (Fig. 27a), (2) Parallel setae-covered lobes projecting from the valva apodeme ( Type species. Cicinnus rotunda Dognin, 1916. Diagnosis. Unique, blueish (even somewhat iridescent in fresh specimens) black mimallonids usually with crenulated wing margins, broad wings; comparatively large antennae with some of the longest pectinations in Mimallonidae (this trait is similar to Cunicumara). The toothed postmedial lines are inwardly lined with pale brown which contrasts against the dark ground color.

Cicinninae Schaus, 1912 Psychocampini St Laurent & Kawahara, 2018
Biterolfa Schaus, 1928 Figs 29, 65, 138, 139 Type species. Cicinnus althea Schaus, 1905. Diagnosis. All three species of Biterolfa are similar, primarily differentiated by genitalia characters. Externally Biterolfa are brown to red-brown with a dark graybrown postmedial line outwardly margined with a thin gray submarginal region which narrows from the tornus to the falcate apex. Distinctive, parallel, white outlined dark streaks span across the middle of the forewing towards the costa.
Remarks. See remarks of Psychocampa below. Diagnosis. Externally Psychocampa display typical cicinnine shape and large size, and always lack hyaline patches, however it is difficult to generalize about this relatively large genus beyond that. Despite external variability across the genus, Psychocampa genitalia are quite homogenous, and therefore species belonging to this genus are easily recognizable by their genitalia, which are dramatically distinct and simple in structure in comparison with all other Cicinninae genera (except for the sister genus Biterolfa, see below).
Apomorphies. Largely as for Biterolfa but gnathos arms more gradually narrowed, never with teeth or dorsal apical protrusions (in addition to the tapering tip).
Remarks. This is one of the largest Mimallonidae genera in terms of number of species, particularly after the present checklist and the work of St Laurent et al. (2018a) which cumulatively transferred many species from Cicinnus to Psychocampa based on the exceptionally homogenous and readily diagnostic genitalia of the latter genus. As stated above, it is difficult to diagnose Psychocampa based on external morphology because there are several clear species-groups within the genus which may eventually prove to be more appropriately placed in new genera that can more succinctly be unified by external morphology (though genitalia morphology is so consistent across all of these groups that additional atomization of Psychocampa into several genera may prove to be unwarranted). Genitalia morphology is largely homogenous across all Psychocampini, not only in Psychocampa, and this is clearly evident in comparing the genitalia of sister genera Psychocampa and Biterolfa, see Figs 29 and 30. However, external maculation and convincing phylogenetic results of St Laurent et al. (2018a) placing Biterolfa as sister to the larger clade of Psychocampa, which was intentionally sampled to include most of the clear species-groups of the genus, supports their valid separation as two distinct genera. Further ongoing molecular phylogenetics which continues to sample Psychocampini more densely also continues to support the separation of these two genera. Our morphological phylogenetic results here fully support the transfer of various species from Cicinnus to Psychocampa, as well as the close relationship between Psychocampa and Biterolfa, though with our increased taxon sampling, the validity of Biterolfa is called into question in a morphology-only context because it is nested within Psychocampa in unconstrained ML and parsimony morphological phylogenetic analyses. Despite this, we maintain Biterolfa and Psychocampa as valid separate genera considering the molecular results of St Laurent et al. (2018a) and the previously mentioned ongoing molecular work.

Bedosiini St Laurent & Kawahara, 2018
Bedosia Schaus, 1928 Figs 31, 67, 142, 143 Type species. Cicinnus fraterna Schaus, 1905. Diagnosis. Bedosia is a rather diverse genus in coloration, shape, and size, but species belonging to this genus always have a B-shaped forewing hyaline patch (as is common in Cicinninae) and well defined postmedial lines on all wings. Most species have brown accented veins and triangular wings with straight margins or those that are slightly convex. Male genitalia, however, clearly unify this genus and distinguish Bedosia from the externally very similar Bedosiallo.
Apomorphies. Combination of the following characters: (1) Valvae situated on dorsal half of vinculum (Fig. 31a) immediately below uncus such that ventral portion of vinculum extends downward as a well-defined projection (Fig. 31d); (2) Heavily concentrated paired set of setae emanate from diaphragm (Fig. 31b); (3) Gnathos arms massive and extremely robust, variable in shape but always arm-like and with speciesspecific modified tips, extended outward more so than upward (Fig. 31c).

Bedosiallo St Laurent & Kawahara, 2018
Figs 32, 68, 144, 145 Type species. Cicinnus forbesi Schaus, 1928. Diagnosis. Much like Bedosia in external shape and key features, but usually with much narrower wings; the genitalia, however, are very different from those of Bedosia, and are diagnostic of the genus (see apomorphies below).
Remarks. Bedosiallo was found to be a strongly supported clade in St Laurent et al. (2018a), where two species of this genus that apparently represented two distinct species-groups within Bedosiallo were sampled and found sister to one another, which together were sister to Bedosia fraterna, type specie Bedosia. Although sampling of Bedosia was relatively poor in St Laurent et al. (2018a), we have since examined the genitalia of nearly all Bedosia species (including those erroneously placed in Cicinnus by Schaus (1928)), finding consistent complex male genitalia, which are readily divergent from the much simpler genitalia of Bedosiallo. In the description of Bedosiallo, all described species were dissected and their genitalia figured, displaying the consistent genitalia in that genus, as well as the consistent difference from all known Bedosia species (St Laurent et al. 2018a). Below in the checklist we transfer two additional species to Bedosiallo from Cicinnus: B. minimalis Herbin and C. Mielke, comb. n. and B. gentilis (Schaus), comb. n. These two species display genitalia much more in line with those of the described Bedosiallo species than to any Bedosia species, but do differ in some respects, particularly in the more closely fused gnathos, triangular valvae shape, juxtal configuration, and presence of cornuti (see Annotations in Section 4). Bedosiallo minimalis was included in our morphological phylogenetic analyses, and was placed within Bedosiallo in all analyses (Suppl. materials 5, 7) except the constrained ML analysis (Fig. 1). We attribute the somewhat unique genitalia of B. minimalis to result in this conflicting placement, and while B. gentilis and B. minimalis may eventually prove to belong to an entirely new genus considering such unique traits, we are confident that these small moths are more closely related to Bedosiallo than Bedosia in comparing the genitalia of all species in Bedosiini, and certainly do not belong in Cicinnus as they were originally placed.
Diagnosis. Silvery-gray to gray brown ground color, sinuate, diffuse postmedial lines, falcate forewings, and broad darker gray to dark gray brown discal marks that vary from fully scaled to fully hyaline within species (often from the same series); maculation overall is dark and diffuse giving the wings a soiled appearance.
Apomorphy. Asymmetric, complex genitalia with phallus situated on left side (when viewed ventrally) of vinculum ( Fig. 33a) with single vinculum tusk reaching out above phallus.
Remarks. St Laurent et al. (2018a) did not include this genus in their phylogeny and instead placed this genus in Cicinninae: Cicinnini based on morphology, namely the presence of all apomorphies of Cicinninae and Cicinnini in what was at the time the single Aceclostria species, A. mus. Our ongoing molecular work, which includes broadened taxon coverage, and Aceclostria, fully supports this placement (St Laurent et al. in prep.). The morphological phylogenetic analyses carried out here support this placement as well, with Aceclostria always recovered within Cicinnini, sister to (unconstrained ML and parsimony analysis, Suppl. materials 5 and 7 respectively) or nested within (constrained ML analysis, Fig. 1) the broader, so far poorly resolved Cicinnus sensu lato clade.
See annotations in Section 4 for information pertaining to the novel inclusion of A. cordubensis comb. n. and A. nigrescens comb. n. et stat. rev. in Aceclostria. Aceclostria was long considered a monotypic genus, but morphological examinations, including the genitalia of A. cordubensis, conclusively support these additional poorly known taxa in this genus. This genus is apparently closely associated with Anacardiaceae (see Annotations).
The Cicinnus s.l. clade is discussed in more detail below as it is the most poorly resolved set of taxa in the family.
Aleyda Schaus, 1928 Figs 34, 70, 148, 149 Type species. Cicinnus accipiter Dognin, 1916. Diagnosis. Wings are exceptionally narrow and elongate with thin hyaline patch streaks on all wings (hyaline streak mesally kinked on hindwing, lunate on forewing). The antemedial line forms a convex semicircle, whereas the postmedial line is incomplete and mostly limited to the proximal margin of the forewing. The ante-and postmedial lines are curved toward each other or are tangent below discal cell.

Arcinnus Herbin, 2016
Figs 37, 150, 151 Type species. Arcinnus hoedli Herbin, 2016. Diagnosis. Externally not particularly distinct from several other species of Cicinnus. Arcinnus are generally smaller than most Cicinnus and always lack hyaline patches on the hindwings (similarly patterned Cicinnus have hyaline patches on both fore and hindwings). Genitalia were the primary means by which this genus was originally described, and we also consider the male genitalia to have the most robust diagnostic characters defining this genus. Apomorphies.
Remarks. St Laurent et al. (2018a) recovered Arcinnus as a valid genus sister to Euphaneta + Aleyda. In the present study, Arcinnus is recovered nested within the poorly resolved Cicinnus sensu lato clade. We hesitate to synonymize this genus at this time pending broader molecular sampling of the Cicinnus s.l. clade, which is ongoing and consistently supports the validity of Arcinnus.
We were unable to figure female genitalia of Arcinnus, but direct the reader to figs 54 and 55 in Herbin (2016), where the female genitalia of the type species, A. hoedli are presented. Diagnosis. See below in the checklist for specific information pertaining to different species-groups within Cicinnus, but in general Cicinnus can be recognized by their falcate forewings, moderate to relatively large size compared to most other Mimallonidae genera. Small, B-shaped hyaline patches, when present, are found on all four wings or forewings only. Species without hyaline patches are usually heavily marked with irregular blotches and speckles.
Remarks. Annotations following the checklist are provided for each group of Cicinnus as we currently define them: Cicinnus s. s. (Cicinnus Group 1) and Cicinnus s. l. (which includes Cicinnus s. l. Groups 2, 3, and some species of uncertain placement). The annotations include diagnostic characters which can be treated as apomorphic of these preliminary groupings of similar species, and may eventually be used to define formal genera (e.g., the available name Gonogramma Boisduval, see Annotation 98, may eventually be resurrected to include the species treated here as Cicinnus s. l.). The morphological phylogenetic analyses do not fully resolve the relationships of species within Cicinnus, with relatively low support values for the various groups within the genus. However, Cicinnus s. l. is consistently recovered within a broader Cicinnini clade, thus tribal placement of the included species is not in question. Additional molecular sampling of Cicinnus and related genera in order to better define the clades within Cicinnus s. l. will be a focus of our future molecular work.
Euphaneta Schaus, 1928 Figs 35, 72, 152 Type species. Phaneta divisa Walker, 1855. Diagnosis. This genus is similar to Aleyda, but Euphaneta are larger, have broader, more ovoid wings, lunate hyaline patches on all wings, and lack clearly distinct antemedial and postmedial lines, but rather have a dark brown antemedial area which merges with a brown costal region that extends along the length of the wing. Euphaneta are cryptically colored, apparently camouflaged resembling a woody stem.
Apomorphy. Hood-like juxtal process (Fig. 35a) dorsal to weakly sclerotized phallus which is shorter than the juxtal hood above it; this juxtal process is a large, substantial structure comprising almost entire diaphragmal area.

Remarks.
We were unable to include the female of E. divisa, however sexual dimorphism in Euphaneta is not overly significant, with females only being larger and with broader wings than the males. Diagnosis. Isoscella can be recognized by the following combination of characters: narrow, triangular wings with single ovoid (or more circular in some species) discal hyaline patch. The similar genus Roelmana always has broader, less elongated wings, and may lack hyaline patches entirely.
Remarks. In our morphological phylogenetic analyses Isoscella and the related Roelmana below, were poorly differentiated phylogenetically due to strong morphological similarity between these two genera. However, Roelmana, which displays more interspecific variation than within Isoscella (see below) was strongly supported as a monophyletic group in St Laurent et al. (2018a), as well as in ongoing molecular phylogenetics that continues to sample more Roelmana species. These molecular phylogenetic studies include the several morphologically distinct species-groups of Roelmana, which formed a clade sister to Isoscella. The most morphologically divergent Isoscella species, I. andina St Laurent and Carvalho, was not sequenced, but was included in our morphological analyses. This unique species is nested within Isoscella in all analyses except the constrained ML analysis (Fig. 1), rendering Isoscella paraphyletic in that analysis. We consider the placement of this taxon as sister to all Cicinnini in this analysis anomalous considering the presence of all generic Isoscella apomorphies in I. andina. The observed placement in Fig. 1 may be due to certain aspects of patterning (e.g., complete, straight ventral forewing postmedial lines in I. andina) which are unique to this species, whereas genitalia are typical of Isoscella. Therefore, based on the unconstrained analysis, we consider Isoscella to represent a monophyletic lineage, whose relationship to species of Roelmana will likely be more clearly elucidated with denser taxon sampling in our future molecular work. Schaus, 1928 Figs 39, 74, 155, 156;Suppl. material 3: Plate 6 Type species. Cicinnus maloba Schaus, 1905. Diagnosis. Wing shape of Roelmana is typical of Cicinnini, but much less variable overall within this genus than in other cicinnine genera. Forewings are sharply falcate, with the anterior margin of the forewing concave, particularly along the tornus. Two distinct wing morphological groups are found in this genus: one that has a hyaline patch on each wing (such as type species R. maloba (Schaus) and R. pluridiscata (Schaus)), whereas the other group entirely lacks hyaline patches and are instead diffusely colored with faint patterning (such as R. laguerrei (Herbin) and R. beneluzi (Herbin), comb. n.). However, genitalia are mostly homogenous in this genus and provide clear generic placement for these two otherwise outwardly distinct groups.

1
The type locality of Z. infantilis has remained unclear since its original description. Dyar (1910: 85) mentioned four male specimens from two locations (Rockstone, Guyana and St. Jean, French Guiana), but did not say which locality corresponded to the type that he did explicitly reference ("Type: No, 13065, U.S. National Museum"). We were able to locate this type from Rockstone, Guyana and consider it to be the holotype. Two additional specimens are curated with the holotype in the type collection of the USNM matching the other locality (St. Jean), and are therefore considered paratypes.

2
As part of an ongoing revision of Zaphanta, which includes dissections of all Zaphanta species and populations, it is apparent that the Central American Z. fraterna stat. rev. is markedly distinct from the Amazonian Z. infantilis (St Laurent and Giusti in prep.). We therefore revalidate Z. fraterna in the present checklist. 3 The monotypic Cunicumara was considered "incertae sedis" along with Menevia, Roelofa, and Tolypida by St Laurent et al. (2018a), due to the fact that these genera did not form a monophyletic group, despite displaying similar genitalia structures in each genus. Our morphological phylogenetic results also support the close association between Cunicumara and incertae sedis genera, as stated above in the remarks of that genus treatment. 4 Unfortunately, the type locality of M. parostia is unknown. It has been suggested that M. pallida may be a synonym of M. parostia (St Laurent and Dombroskie 2016) due to their similar appearance (small size, maculation). However, the genitalia of the type of M. parostia have not been compared to those of female M. pallida, thus the validity of M. pallida remains uncertain, especially considering that the type locality of M. parostia will likely never be known. 5 The The original primary type of M. plagiata is lost, along with many types described by Walker originating from the Fry collection (Becker 2001). St Laurent and Dombroskie (2016) designated a neotype for this species, thereby stabilizing the uncertainty of the name. 7 Becker (1996) incorrectly listed Veracruz, Mexico as the type locality for R. hegewischi, when in fact the primary type in ZSM only has written on data labels "Mexico". In Druce's (1887) original description, he stated that a specific locality within Mexico was not known "but most likely came from the southern part of that country" (Druce 1887: 227). 8 Confusion has surrounded the type locality and original combination of Tolypida amaryllis, the type species of Tolypida. Motada amaryllis Schaus, 1896a: 142 and Hydrias amaryllis Schaus, 1896a: 143, both bearing the specific name amaryllis, were described by Schaus in the same publication, and both had the type locality of Castro, Paraná, Brazil (Schaus 1896a). Motada amaryllis is a synonym of the Arctiinae Paracles brunnea (Hübner), whereas Hydrias amaryllis is a mimallonid and was transferred to Tolypida by Schaus (1928). Problems with the incorrect original combination can be traced back to Fletcher and Nye (1982) who first listed Motada as the original combination for T. amaryllis when listing the type species of the mimallonid genus Tolypida. This incorrect original combination was repeated by Becker (1996). Furthermore, an apparently incorrect type locality (Castro, Paraná, Brazil) was given by Schaus (1896a) in the original description of Hydrias amaryllis, which was repeated by Becker (1996). The name-bearing type specimen in the USNM is labeled as being from São Paulo. The inconsistency in the reported type locality between Schaus' original description and the apparent type specimen in the USNM was first noted by Pearson (1984) in his revision of Tolypida. Pearson (1984) mentioned that the type locality given by Schaus (1896a) did not match what was written on the labels of the type specimen, and also pointed out that wing measurements reported by Schaus do not match original description. Therefore, we designate a lectotype for the specimen in the USNM from São Paulo (Fig. 159)  Eadmuna paloa was treated by Becker (1996) as a synonym of E. esperans without justification. St Laurent and Dombroskie (2015) showed that these two sympatric species display distinct morphological differences warranting their treatment as separate species. The type locality listed in Becker (1996)

(Paraguay), is incorrect. 10
Eadmuna pulverula was transferred from Cicinnus to Eadmuna by St Laurent and Dombroskie (2015) based on external and genital morphology. This species is known only from a single female specimen, though a second specimen has since been discovered in the NHMUK. However, this other specimen is lacking an abdomen so it will be impossible to know for certain its identity since the diagnostic characters of this rare species are in the genitalia and a ventral abdominal stripe. 11 Macessoga hoppia stat. n. and M. hyginia stat. n. were both originally described as "forms" of M. aelfrida by Schaus (1928). However, an ongoing revision of Macessoga (St Laurent and Herbin, in prep.) has determined that these "forms" will likely prove to be valid species. Article 45.6.4 permits such a status change, from a "form" to a species (ICZN 1999 Rubio (2004), in which the collecting data for "Psychocampa nocturna ‡" is given as: "Ecuador, Napo, San Rafael, 1550 m" leading us to believe that the specimen figured in Piñas Rubio (2004) is not likely R. occidentalis, a species that seems to be restricted to low elevations of the western slopes of the Andes. Reinmara occidentalis also lacks a forewing tornal notch, which is present but reduced in the specimen figured in Piñas (2004). As mentioned in St Laurent et al. (2017c), all names proposed by Piñas Rubio in "Mariposas del Ecuador" are unavailable nomina nuda as they do not satisfy Article 13.1.1 of the ICZN (no written description given) (ICZN 1999). 17 Trogoptera althora was described from an indeterminate number of specimens from Guatemala and Costa Rica. Considering that this may lead to some confusion in identifying "true" T. althora, we designate a lectotype for the specimen labeled as the "type" in the USNM (Fig. 160). This specimen bears the following labels: Cayuga Guat/ May/ Schaus and Barnes coll./ Type No. 33569 U.S.N.M/ USNM-Mimal: 1048/ Trogoptera althora Type Schaus/. A red handwritten label reading "LECTOTYPE♂ Trogoptera althora Schaus, des. by St Laurent" has been added to the specimen. 18 The identity of Trogoptera erosa has not been clear due to the lack of any type material, besides the figure in Herrich-Schäffer, [1856]. Trogoptera erosa is darker, with more sharply angled forewing apices than in the other common southeastern Brazilian species Trogoptera notata/excavata which may or may not prove to be synonymous names since the type of T. notata is male and that of T. excavata is female. 19 St Laurent and Cock (2017) designated a lectotype for T. guianaca from French Guiana because Schaus (1928) described this species from syntypes from different localities, likely including multiple taxa. Trogoptera guianaca was originally described as a subspecies of T. notata but was later treated as a valid species by Becker (1996) without any indication of its new status. We maintain T. guianaca as a valid species. 20 Trogoptera jonica stat. n. was originally treated as a subspecies of T. althora by Schaus (1928). However, externally T. jonica is unlike T. althora, particularly in terms of wing shape and maculation. Trogoptera jonica is more similar to species from Brazil than to Central American species such as T. salvita. Schaus (1928) drew this comparison in the description of T. salvita, despite naming T. jonica as a subspecies of Central American T. althora in the same work. Trogoptera jonica is mostly known from Paraguay, Bolivia and Brazil, and is very similar to the Brazilian T. salvita which displays straighter wing margins in the female and darker maculation in both sexes. A substantial series of specimens from Paraguay, Bolivia, and Mato Grosso, Brazil (T. jonica) and Goiás, Brazil (T. salvita) in the NHMUK, and comparison to the types of both species revealed that the identities of these two species are often confused in collections. To further complicate the matter, the type of T. jonica is male, whereas that of T. salvita is female. The aforementioned series in the NHMUK, which contains both sexes of both species, allowed us to determine their distinction as two separate species. 21 Trogoptera mahlaha stat. n. was originally treated as a "form" of T. althora by Schaus (1928). However, T. mahlaha is undoubtedly a distinct species based on external morphology and distribution (T. mahlaha is from Ecuador, whereas T. althora was described from Guatemala). 22 Trogoptera maroniensis (Dyar) is a senior homonym of T. maroniensis (Dognin). Schaus (1928), as the first reviser, treated Dyar's name as the valid name to apply to this Amazonian Trogoptera species. Slight maculation differences (the straightness of the forewing postmedial line) between the respective types of T. maroniensis (Dyar) and T. maroniensis (Dognin) can likely be attributed to individual variation as examination of several other specimens from French Guiana likewise revealed variation in the same character. In order to clarify their synonymy, genitalia should be examined of both types in a formal revision of Trogoptera. 23 Trogoptera micalha stat. n. was originally treated as a subspecies of T. althora by Schaus (1928). The wing shape of T. micalha is broader, and more rounded than either T. althora or T. noaha (also described from Veracruz, Mexico). This taxon likely should be treated as a valid species, and we propose this classification pending a comprehensive revision of Trogoptera. 24 Trogoptera noaha stat. n. was originally treated as a "form" of T. althora by Schaus (1928). Externally, T. noaha is similar in appearance to T. althora, but displays reduced maculation on all wings. Rather than maintain the name noaha as a "form", we elevate this taxon to species pending a comprehensive revision of Trogoptera. 25 Trogoptera notata is a problematic species because its holotype specimen lacks specific locality data within Brazil, though which was probably Rio de Janeiro since most of the Walker-era Mimallonidae were collected in that Brazilian state. Schaus (1896c) originally thought that T. notata and T. excavata represented different sexes of the same species, though in his revision of Mimallonidae, he treated them as distinct species (Schaus 1928). We believe that T. notata and T. excavata may be conspecific, but pending a more thorough revision of Trogoptera, we treat them as distinct as did Schaus (1928). 26 Trogoptera salvita has been often confused with T. jonica. However, the two species appear to be distinct, with the wing margins of female T. salvita being more convex.
Trogoptera salvita is found in the Brazilian Cerrado, whereas T. jonica is so far known from Paraguay and neighboring Argentina, Brazil, and Bolivia. The type locality given by Schaus (1928) "Santa Catarina" is the incorrect Brazilian state; the type locality of T. salvita, which is Casa Branca, Brazil, is an old location in western/northern São Paulo state (C. Mielke pers. comm.). Becker (1996) was seemingly uncertain about this locality, listing it as "SC/SP". This distinction is important because the type locality of T. salvita at Casa Branca is within the Cerrado biome, to which this species is endemic. This information allowed direct comparisons to several specimens in the NHMUK, which were also collected in the Cerrado (Goiás, Brazil), thereby allowing the recognition of this species, and its distinction from T. jonica. 27 Adalgisa croesa was described from an indeterminate number of specimens from Santa Catarina, Brazil, and French Guiana. Considering the likelihood that Adalgisa from these two localities may eventually prove to be distinct species, and in order to stabilize the identity of A. croesa, we designate a lectotype for the specimen labeled as the "type" in the USNM. This specimen (Fig. 161)  Alheita pulloides was described from four male specimens from Guatemala, Colombia, French Guiana, and Peru. Although Dognin (1921) listed the type being from Medina, Colombia, the presence of possible syntypes from distant localities encouraged us to designate a lectotype for the specimen labeled as the "type" from Medina, Colombia in the USNM. The lectotype specimen (Fig. 162) Becker (1996) incorrectly listed the type locality of A. subnotata as "Brazil (AM)". The correct locality is Pebas, Peru. 30 We describe a new genus, Fatellalla gen. n., for F. fatella comb. n. Although this genus was not part of the phylogenetic analyses of St Laurent et al. (2018a), Fatellalla is unambiguously part of Lacosominae: Alheitini based on male genitalia characteristics and our morphological phylogenetic analyses. See genus description above for diagnosis, apomorphies, description, and figures for this genus. 31 The monotypic genus Herbinalla was described by St Laurent and Kawahara in St Laurent et al. (2018a) to include the enigmatic species H. caudina based on both phylogenomic and morphological analyses, this species was incorrectly placed in Alheita by Schaus (1928). 32 Tarema macarina was treated as synonym of T. rivara by St , who noted that the type of T. macarina was the sexually dimorphic female of T. rivara. 33 Thaelia subrubiginosa and T. anysia were erroneously placed in Alheita and Cicinnus (respectively) by Schaus. Both species were transferred to Thaelia by St Laurent et al.
(2018a) based on their phylogenetic results. External morphology also supports the new combinations of these two species. However, genitalia of T. subrubiginosa and T. anysia are rather distinct from the type species of Thaelia figured in Herbin (2016). 34 Thaelia inornata comb. n. was omitted from Becker (1996). Although T. inornata was not included in St Laurent et al. (2018a), examination of genitalia and external morphology of T. inornata suggests it is morphologically very similar to T. subrubiginosa, and we therefore transfer it to Thaelia. The morphological phylogenetic analyses consistently recover T. inornata nested within Thaelia. 35 We describe a new genus, Citralla gen. n., for C. rumina comb. n. Based on male genitalia, larval characteristics, and our morphological phylogenetic analyses, we can unambiguously place this new genus in Lacosominae: Lacosomini and ongoing molecular phylogenetics supports the novelty of this new genus (St Laurent in prep.). See genus description above for diagnosis, apomorphies, description, and figures for this genus. The type locality for "Trogoptera" rumina in Becker (1996) was given as Costa Rica, but this was erroneous as the type specimen and original description clearly indicate "Volcan de Chiriqui," Panama, as the type locality. 36 Naniteta was described by Franclemont (1973) to include Lacosoma elassa, described in the same work. The new genus was compared only to the unrelated Cicinnus by Franclemont. St Laurent et al. (2018a) showed that another species, L. morgani, which displays genitalia morphology that is essentially identical to those of L. elassa, and similar external morphology, was nested within Lacosoma. Naniteta was thereby synonymized by St Laurent et al. (2018a). We can confirm this synonymy here as L. elassa was firmly nested within Lacosoma in all analyses. It remains uncertain whether L. elassa and the similar L. morgani are synonymous, as the genitalia and external morphology of the two species are quite similar. Lacosoma perplexa and L. miradorensis are also similar to these two species, but they are slightly larger, and an examination of the genitalia of a topotypical specimen of L. perplexa (St Laurent diss.: 4-16-18:1, USNM) revealed characters distinct from types of L. miradorensis, L. morgani, and L. elassa. 37 Lacosoma briasia was described from an indeterminate number of specimens from far reaching localities in Brazil, Colombia, and French Guiana. Considering that this may lead to some confusion in identifying "true" L. briasia, we designate a lectotype for the specimen labeled as the "type" in the USNM (Fig. 163) Lacosoma diederica was described from an indeterminate number of specimens from Brazil, Bolivia, and Paraguay. Considering that this may lead to some confusion in identifying "true" L. diederica, we designate a lectotype for the specimen labeled as the "type" in the USNM (Fig. 164) There is a specimen labeled "Type" in CMNH of "Lacosoma streckeri ‡" from San Esteban Valley (Carabobo), Venezuela. The specimen appears to be L. ludolpha and perhaps for this reason was never formally described, hence the name appears to be an unavailable manuscript name. 40 Lacosoma morgani may be a synonym of L. elassa; both species are nearly identical in male genitalia and external morphology, but the two species are described from distant localities (Guatemala for L. morgani, and southern Texas, USA for L. elassa). More material of L. elassa, especially from farther south, would be needed to compare directly with L. morgani. 41 Lacosoma raydela was described from an indeterminate number of specimens from Guatemala and Costa Rica. Considering that this may lead to some confusion in identifying "true" L. raydela, we designate a lectotype for the specimen labeled as the "type" in the USNM (Fig. 165) Lacosoma schausi was proposed as a replacement name by Dognin (1923) for L. rosea ‡ Schaus, 1905 (preoccupied). This is not to be confused with L. rosea (Dognin, 1905: 120), a separate species, and the source of the preoccupation of the name "rosea". 43 Lacosoma subrufescens was transferred to Lacosoma from Trogoptera by St Laurent and Mielke (2018) based on genitalia. This species is closely related to L. asea and L. ursmara. All three of these Lacosoma species are Andean, inhabiting moderate to high elevations, and together form a unique morphological group within the genus. 44 Lacosoma ursmara comb. n. is transferred from Druentica based on the discovery of female L. asea, which is remarkably similar to the female type of L. ursmara. The two species are from far separated type localities (L. asea in Colombia, L. ursmara in Bolivia), so we do not treat them as synonyms pending an inclusive revision of Lacosoma. Both species are Andean endemics and are relatively rare in collections. The morphological phylogenetic analyses consistently recover L. ursmara nested within Lacosoma. 45 Lacosoma violacea is a taxon of uncertain identity. No specimens have been found attributed to this name, and the only information pertaining to this species are the original figures and text in Sepp [1818]. Externally, the specimen figured in Sepp is unlike any other known Mimallonidae, particularly the blue coloration and arrangement of the postmedial line on the ventrum of the forewings. In fact, the color blue is not known from any Mimallonidae. It is possible that Sepp's illustration is a misrepresentation of a more characteristically colored mimallonid species, and thus we consider Sepp's illustration unreliable for accurately assigning the name violacea. Forbes (1942) believed that the specimen figured in Sepp is a member of Thyrididae, and Schaus (1928) was uncertain that L. violacea should be treated as a Lacosoma. However, the larva and feeding behavior illustrated and described in Sepp [1818] unquestionably belongs to Mimallonidae, though it is impossible to ascertain whether the larva and "blue" adult are conspecific. 46 Vanenga mediorosea was described for the Vanenga populations inhabiting the Brazilian Atlantic Forest, the Pampa, Paraguay, and Argentina. This species has had at least three manuscript names variously applied to it in the CUIC, NHMUK, and USNM, but was never formally described until St Laurent and Herbin (2017 Druentica garretti comb. n. displays external and genital morphology (MWM genitalia preparation 35.544) similar to D. brosica, and the former is unlike any described Mimallo. Pearson (1951), in his revision of Mimallo, recognized that D. garretti and D. brosica did not belong in Mimallo, but did not transfer these species out of Mimallo, or suggest a better genus in which to place them. All morphological phylogenetic analyses consistently recover D. garretti nested within Druentica, thus supporting this new combination. 51 Druentica inscitoides stat. rev. was treated as a synonym of D. scissa by Schaus (1928). Upon the first author's examination of both types, we determine that D. inscitoides should be maintained as a valid species pending a revision of Druentica. The identity of the three species of Pamea is uncertain. We have not been able to locate the type(s) of P. nana, but the figures in Herrich-Schäffer (1856) display a male specimen not unlike the male types of P. albistriga and P. dotta. Pending a revision of Pamea, the three species are currently treated as valid. Kirby (1892) considered P. nana a synonym of P. albistriga, though this synonymy was overlooked by later authors (e.g., Schaus 1928, Becker 1996. Examination of the Pamea specimens at the NHMUK reveals several specimens labeled by earlier workers as either nana or albistriga, supporting the idea that the names are likely synonymous. Other cases do exist of Herrich-Schäffer describing the same species previously (by a matter of a year or two) described by Walker, for example Procinnus plana (Walker) and its synonym P. diagonalis (Herrich-Schäffer). Pamea was not included in St Laurent et al. (2018a), but our morphological phylogenetics and ongoing molecular phylogenomics that both do include Pamea, strongly support the placement of this genus in Druenticini. 57 We transfer Procinnus producta comb. n. from Cicinnus to Procinnus based on external and genitalia morphology. The genitalia of a male were examined by brushing, and are unambiguous, matching very closely the described Procinnus species. Due to the rarity of this species we did not fully dissect P. producta at this time, and thus did not code it for morphology. The genitalia were readily visible due to brushing, however, and thus we are confident of its generic placement. In the original description, Dognin (1901) recognized the similarity to P. plana (as its synonym P. diagonalis) The date of publication of the name was incorrectly listed as 1905 in Schaus (1928), Gaede (1931), and Becker (1996). 58 As explained in St , the genitalia preparation of the type, and only known specimen of L. penia, is missing. However, this species is externally distinct from L. quindiuna and thus both species were treated as valid, which is followed here. 59 St  determined that the abdomen attached to the type of L. quindiuna originated from an entirely different Lepidoptera specimen. Unfortunately, the lack of genitalia from either type specime of Lurama results in some uncertainty as to the identity of the two species in this genus. However, examination of 12 male genitalia preparations from the entire distribution of L. quindiuna (with the exception of the type locality), revealed uniform genitalia characteristics despite minor external differences across populations . Attempts should be made to rediscover L. penia in nature or in natural history collections. Furthermore, examination of the genitalia, and a study of additional topotypical L. quindiuna would be desirable as well. 60 Subsequent to the description of U. azurula, an additional specimen from Peru: Junín: 38 km SW of Mazamari, 2540 m (deposited in the MGCL) has been examined. This specimen was collected roughly halfway between the type localities of U. azurula and U. dombroskiei, and did not show any intermediate characters between these two species. This specimen displays external and male genitalia characteristics typical of U. azurula, supporting the validity of this rarely collected species. 61 As Psychocampa alcuna stat. rev. was previously treated as synonym of P. hamata (Dognin 1923, Schaus 1928, Gaede 1931, Becker 1996. Genitalia and external morphology support the revalidation of P. alcuna, but we recommend a formal revision of this species complex. Psychocampa hamata was described from Rio de Janeiro, Brazil whereas P. alcuna was described from French Guiana. Therefore, previous discussions of "Psychocampa hamata" in St Laurent et al. (2018a) (2016)). Thus, the transfer made in St Laurent et al. (2018a) remains valid and allows the transfer of alcuna and hamata to Psychocampa. Furthermore, since the previous status of the name alcuna was as a synonym of "Cicinnus hamata" in Becker (1996), and hamata was transferred to Psychocampa by St Laurent et al. (2018a) along with its associated synonyms by default, the present combination of P. alcuna is not a new combination, but only a revived status. Psychocampa alcuna was consistently recovered within a broader Psychocampa + Biterolfa clade in all morphological phylogenetic analyses. 64 Psychocampa bactriana comb. n. belongs to a relatively large group of similar species including P. marona, a species included in the phylogenetic study of St Laurent et al. (2018a). Psychocampa marona was nested within Psychocampa in St Laurent et al. (2018a), and genitalia are nearly identical to those of P. concolor, the type species of Psychocampa. We therefore transfer P. bactriana (and the remaining similar species) to Psychocampa from Cicinnus based on this molecular and morphological evidence. Genitalia information is based on males that we have examined that correspond to the female holotype of P. bactriana. The type of P. bactriana was historically dissected, but the preparation has since been lost. 65 Psychocampa batesii comb. n. is very similar to P. bactriana. Upon direct comparison of the holotype of P. bactriana (NHMUK) and several syntypes of P. batesii (NHMUK, HEC), P. batesii is a decidedly smaller and lighter colored with reduced maculation. The forewing postmedial line in P. batesii runs closer to the wing margin than in any other similar species. Psychocampa batesii also resembles P. callipius, but in the latter species the forewing postmedial line approaches the discal spot and is more distal from the wing margin. An in-depth revision of this complex of species is warranted. 66 Although outwardly similar to Cicinnus s. s., a dissection of a male P. bilinea comb. n. surprisingly revealed genitalia diagnostic of Psychocampa, and therefore we move this species to Psychocampa from Cicinnus. All of our morphological phylogenetic analyses support this transfer as well. 67 Psychocampa callipius comb. n., P. candacus comb. n., P. gaujoni comb. n., and P. narseres comb. n. are hereby transferred to Psychocampa from Cicinnus based on phylogenetic and morphological data which places P. marona within the Psychocampa clade sensu St Laurent et al. (2018a). These species were not included in the morphological phylogenetic analyses due their close similarity to P. marona, which was sampled for molecular data in St Laurent et al. (2018a). Genitalia of all of these species have since been examined subsequent to the morphological phylogenetic analyses in this article, and support their placement in Psychocampa.

68
Since the original description of "Cicinnus" camarinus in Schaus (1928), there has been uncertainty about the identity of this species. Recently, the first author examined the female type specimen in the MNHU. Surprisingly, P. camarinus comb. n. is externally extremely similar to P. alcuna from French Guiana and P. hamata from Rio de Janeiro, Brazil (see Annotations 63 and 70 regarding the separation of P. alcuna and P. hamata). Psychocampa camarinus was described from Iquitos, Loreto, Peru. Many Amazonian mimallonids described from French Guiana are known from Peru (e.g., Dombroskie 2016, St Laurent et al. 2017d), therefore it is possible that P. camarinus may actually prove to be a synonym of P. alcuna. However, upon examining several specimens of P. alcuna from the eastern Amazon and French Guiana (the latter is the type locality of P. alcuna), and comparing them to NHMUK specimens from Peru, it is clear that Peruvian material is distinctly larger and displays slightly more elongated wings. Genitalia are also distinct between the two species, in P. camarinus (Peru, St Laurent dissection: NHMUK010402320, NHMUK) the phallus is broader and relatively shorter than the more narrowed phallus of P. alcuna (French Guiana, St Laurent dissection 17-3-18:9, MGCL). We therefore maintain P. camarinus as valid, and merely transfer it to Psychocampa based on its close similarity to P. alcuna and P. hamata. All of our morphological phylogenetic analyses support this new combination as P. camarinus was consistently recovered nested within a broader Psychocampa + Biterolfa clade. 69 Psychocampa funebris comb. n. is hereby transferred to Psychocampa based on examination of genitalia which revealed characters completely in line with those of Psychocampa sensu St Laurent et al. (2018a). This species is likely quite closely related to the previously listed species such as P. callipius, P. candacus, P. gaujoni, and P. narseres, but displays much darker coloration such that most patterning is obscured. This uniquely colored species is one of the only described Psychocampa to display diurnal behavior (St Laurent and Carvalho 2017a, St Laurent and Carvalho, in prep.). All of our morphological phylogenetic analyses support this new combination as P. funebris was consistently recovered nested within a broader Psychocampa + Biterolfa clade. 70 As previously discussed in Annotation 63, Psychocampa hamata was transferred to Psychocampa based on morphology and the molecular results of St Laurent et al. (2018a). However, the specimen used in their study was from French Guiana, and thus actually represented P. alcuna (reinstated from synonymy under P. hamata). Psychocampa alcuna and P. hamata are likely to be sister species considering external and genitalia similarities, thus the transfer made in St Laurent et al. (2018a) was correct. The name P. hamata is restricted to the species present in Southeastern Brazil and adjacent regions, genitalia are largely consistent between these populations, but the phallus is narrower and flatter in P. hamata than P. alcuna. 71 Yet another taxonomic issue concerning the Psychocampa hamata/alcuna complex arises from the name jaruga, proposed by Jones (1912). Jones described Perophora jaruga syn. n. from an indeterminate number of specimens from Guarujá, Santos, São Paulo, Brazil. A male specimen labeled as the type is present in the NHMUK and its examination by the first author reveals that it is a close morphological match for the other southeastern Brazilian species: P. hamata. In his revision of Mimallonidae, Schaus (1928: 649) stated the follow-ing in regards to "Cicinnus jaruga": "The species is unknown to me and the description faulty. Owing to the peculiar black tipped white scales, it is evidently close to hamata and camarinus." Therefore, as early as Schaus (1928), there was some concern regarding the identity of this species, already attributing the name jaruga to at least being a relative of P. hamata/ alcuna/ camarinus. Therefore, it is uncertain as to why Becker (1996) synonymized the name jaruga with the quite distinct "Cicinnus" joanna (=Psychocampa joanna sensu St Laurent et al. (2018a) and this paper). We therefore newly synonymize jaruga with P. hamata, as its synonymy with P. joanna was apparently erroneous. 72 Psychocampa joanna and P. unalca were transferred to Psychocampa from Cicinnus by St Laurent et al. (2018a) on molecular and morphological grounds. The genitalia of these, and related species (e.g., P. lacuna, P. undiscata, and P. viemanda) are typical of Psychocampa and are transferred to that genus here. 73 Psychocampa lacuna stat. rev. was treated as synonym by Schaus (1928), and this synonymy was followed by subsequent authors (Gaede 1931, Becker 1996. St Laurent and Cock (2017) hypothesized that the synonyms of P. joanna would likely prove to be valid species. We revalidate P. lacuna based on external and genitalia differences between topotypical P. joanna (French Guiana, St Laurent diss.: LEP24627 in MGCL) and near topotypocal P. lacuna (Costa Rica, Sixaola River, St Laurent diss.: NHMUK010402321 in NHMUK). In the darker colored P. joanna, the gnathos arms are much thinner in comparison with the thick and robust gnathos arms of P. lacuna. The phallus of P. joanna is also thinner from the lateral aspect (wider when viewed dorsally, flatter when viewed anteriorly) than in P. lacuna. Furthermore, the two species have clearly disjunct distributions. The type locality of P. lacuna is Costa Rica, but the fist author has examined material from as far north as Guatemala, suggesting that this species is likely widely distributed in Central America. Psychocampa joanna, on the other hand, is an Amazonian species. Because the previous status of the name lacuna was as a synonym of "Cicinnus joanna" in Becker (1996), and joanna was transferred to Psychocampa by St Laurent et al. (2018a) along with its associated synonyms by default, the present combination of P. lacuna is not a new combination, but only a revived status. Schaus described P. lacuna from an indeterminate number of syntypes from Costa Rica, though two specimens (a male and a female) in the USNM are labeled "type", therefore we designate the male specimen (Fig.  166)  Psychocampa madenus comb. n. possesses the genitalia typical of Psychocampa (sensu St Laurent et al. 2018a; based on dissection of genitalia NHMUK010402309 in NHMUK), and is hereby transferred to Psychocampa from Cicinnus. Psychocampa madenus is most similar in external morphology to P. manalca, which we also transfer to Psychocampa on morphological grounds. All of our morpho-logical phylogenetic analyses support this new combination as P. madenus was consistently recovered nested within a broader Psychocampa + Biterolfa clade. 75 Psychocampa magnapuncta comb. n. belongs to a complex of similar species in this genus. As per St Laurent and Cock (2017), only females from Trinidad have so far been collected or photographed, and thus there is some uncertainty as to the identity of the male and validity of this species. St Laurent and Cock (2017) designated and figured the lectotype of P. magnapuncta. This species was not included in our morphological phylogenetic analyses due to its rarity, and the fact that males remain unknown (and thus could not be coded for male genitalia). External morphology is sufficiently close to other Psychocampa species coded for morphology and sequenced in St Laurent et al. (2018a)  Psychocampa mulatro comb. n. was first noted to be related to P. undiscata by Dognin (1923). Although P. mulatro is rare in collections, the first author recently located several males belonging to this species (other known specimens are female) among the reared material originating from D. Janzen and W.
Hallwach's rearing efforts in Guanacaste, Costa Rica, deposited in the USNM. Based on external morphology and genitalia (examined from brushing), P. mulatro undoubtedly belongs in Psychocampa and we transfer this species here. 79 Psychocampa unalca was transferred to Psychocampa from Cicinnus by St Laurent et al. (2018a) on molecular and morphological grounds. 80 Psychocampa undiscata stat. rev. was erroneously synonymized with P. joanna in Becker (1996), who incorrectly listed the type locality of P. undiscata as "Fr. Guiana" which gave the appearance of a valid synonymy considering the type locality of P. joanna is French Guiana. However, the correct type locality of P. undiscata is São Paulo, Brazil, and the morphology of this southeastern Brazilian and Argentinian species is noticeably different from the Amazonian P. joanna and Central American P. lacuna. Namely in P. undiscata (Franclemont dissections 1431(Franclemont dissections , 1432 in CUIC) the gnathos arms are thicker and more robust than in P. joanna, and in this way more closely resemble P. lacuna. Furthermore, the phallus of P. undiscata is substantially flattened, wide, and spoon-like, immediately distinguishing it from related species. Dark, chocolate brown coloration and distribution in southeast Brazil and Argentina also help distinguish P. un-discata from the comparatively lighter colored Amazonian P. joanna and even lighter Central American P. lacuna. We therefore revalidate this taxon. Because the previous status of the name undiscata was as a synonym of "Cicinnus joanna" in Becker (1996), and because joanna was transferred to Psychocampa along with its associated synonyms by St Laurent et al. (2018a), the present combination of Psychocampa undiscata is not a new combination, but a revived status. 81 Psychocampa viemanda comb. n. is externally difficult to distinguish from P. unalca, but P. viemanda is larger with darker maculation and has wider submarginal areas.
Considering the close relationship of P. unalca and P. viemanda, and the incontrovertible evidence placing P. unalca in Psychocampa based on external morphology and molecular data in St Laurent et al. (2018a), we hereby transfer P. viemanda to Psychocampa. The genitalia of P. viemanda are also typical of Psychocampa. All of our morphological phylogenetic analyses support this new combination as P. viemanda was consistently recovered nested within a broader Psychocampa + Biterolfa clade. 82 The name melini syn. n. was proposed as a subspecies of P. joanna by Bryk (1953), and the name was omitted by Becker (1996) Bedosia olasis comb. n. is similar to B. dulcis, as originally mentioned by Schaus (1928), and is therefore transferred to Bedosia. This is a relatively rare species in collections, and unfortunately the type locality, "Central America", is not specific, though the first author is aware of a potentially conspecific species that is commonly collected in Guatemala (several specimens in MGCL and CRAS). This species appears to be closely related to B. dulcis and B. euthymius which are found in Central America and western Colombia respectively, suggesting a unique Central American/western Andean lineage of Bedosia lacking morphologically similar species in other regions of South America. Examination of the genitalia of the type of B. olasis (by brushing) and full dissections of the Guatemalan population, displays genitalia typical of Bedosia. 87 Authorship of B. strigifera (and other Mimallonidae described by C. and R. Felder (1874)) was determined with the aid of Nässig and Speidel (2007), hence it is in disagreement with the authorship listed in Becker (1996). The type locality of B. strigifera is given only as "Brazil" but the type specimen in the NHMUK is a good match for a common southeastern Brazilian species, therefore the original collecting locality of B. strigifera was likely one of the heavily collected regions from the 1800s such as Rio de Janeiro or São Paulo. 88 In the original description, Schaus (1928) stated that the type deposition for B. forbesi was the CUIC. However, we confirm that the type is not currently in the CUIC and is instead in the USNM. 89 The species Bedosiallo gentilis comb. n. and B. minimalis comb. n., previously both placed in Cicinnus, display genitalia characteristics very close to Bedosiallo. At the time of the Bedosiallo description, St Laurent et al. (2018a) did not realize this close similarity, and thus did not place these two species in this genus. We hereby transfer these two species to Bedosiallo. Upon examining the male genitalia of B. minimalis, and the genitalia of both sexes of two undescribed species near B. gentilis (B. gentilis is known to the first author only by the female type), it is clear that these species share a close relation to those placed in Bedosiallo, namely in the external appearance in which the hyaline discal spot is lined with dark scales (a trait shared with most Bedosiini) and the wing veins which are accentuated by light brown scales. The male genitalia are quite similar to those in the B. forbesi species-group, both have a large, concave juxtal complex attached dorsally to the phallus. The simple triangular valvae and robust gnathos are also similar to those described from Bedosiallo, although in B. cf gentilis and B. minimalis the gnathos arms appear distally fused. However, upon close examination, they are slightly separated as in all other Bedosiallo, only substantially thicker and shorter. In an undescribed species of Bedosiallo near B. gentilis from Southeast Brazil, the gnathos arms are fully fused. The phallus of these newly included species is thicker than in all others in the genus, and contain cornuti. Bedosiallo minimalis was consistently placed within Bedosiini in all morphological phylogenetic analyses, and within Bedosiallo in all but the constrained ML analysis (in which B. minimalis was placed sister to Bedosia + Bedosiallo). Regardless, Bedosiallo is the most appropriate genus based on morphology for these two species.
Because of the noted unique features of B. gentilis and B. minimalis, we place them within a third species group within Bedosiallo. 90 Aceclostria cordubensis comb. n. displays the complex asymmetrical genitalia typical of A. mus, type species of Aceclostria. In A. cordubensis the complicated juxtal arrangement has more robust and more distinctly curved dorsal processes, a shorter spine that protrudes over the phallus, and the gnathos arms are smaller than in A. mus; but otherwise these features are extremely similar to all examined Aceclostria dissections examined thus far. Externally A. cordubensis is earthen in coloration, with indistinct suffuse maculation, and postmedial lines arranged in much the same way as in A. mus. Furthermore, in A. cordubensis and A. mus, the discal cell varies from fully scaled to completely hyaline in specimens collected from the same series. Such pronounced intraspecific variation in hyaline patches appears to be so far unique to Aceclostria. Distribution in dryer regions of central South America and usage of Schinus Linnaeus (Anacardiaceae) as host plants also supports the inclusion of A. cordubensis in Aceclostria (Schreiter 1943). Although the type of A. cordubensis has not been located in collections in Argentina or Europe (F. Penco pers. comm., St Laurent pers. obs.), the type of its synonym, schulzii, was found in the MNHU. Despite the missing type of A. cordubensis, the identity of this species is not in question due to the fact that both cordubensis and schulzii come from the same type locality (Cordobá, Argentina) and figures of A. cordubensis in Schreiter (1943) and the type specimen of schulzii appear to be conspecific. Aceclostria cordubensis is one of the few Mimallonidae found far south in Argentina (Córdoba), and additional specimens examined by the first author in the ZSM from localities even farther south in Argentina and near the Chilean border, are a close match to A. cordubensis. 91 A specimen of Aceclostria mus labeled as "Cicinnus callinica ‡ Schaus" is present in the USNM. This specimen apparently represents an unpublished manuscript name because no description exists in the literature using this name and this specimen, and therefore is an unavailable name. The specimen is from the same type locality as A. villaricensis, which is a synonym of A. mus. It is worth noting that "Cicinnus callinica ‡" should not be confused with the valid and obviously distinct Trogoptera callinica Schaus. 92 Upon dissecting A. mus specimens from the type localities of both mus (São Paulo, Brazil, genitalia preparation NHMUK010402316 in NHMUK) and villaricensis (Villarica, Paraguay genitalia preparation NHMUK010402317 in NHMUK), we determined that there are no external nor male genitalia differences between these populations and maintain the synonymy of Becker (1996).
Other populations from the Cerrado of Central Brazil, however, do differ externally and in genitalia, and likely represent an undescribed species.  (2016)), it is evident that the blunt vincular arms typical of Arcinnus are much more robust in the type of A. xingua than in any of the A. hoedli dissections. The valvae of A. xingua also terminate much more abruptly than in A. hoedli, forming a sharp distal edge. All of our morphological phylogenetic analyses support this transfer. 96 Cicinnus is the most species-rich mimallonid genus. A great deal of taxonomic confusion has been associated with Cicinnus, and the genus has largely existed as a wastebasket taxon since its inception. Much of the confusion of the identification of "true" Cicinnus can be traced back to the type species, C. orthane. This species, described from Chile, is one of the only mimallonids recorded from that country, and therefore has been speculated to be erroneously labeled (St Laurent et al. 2018c). Comparisons of the putative female syntype of C. orthane, which was subsequently designated as the lectotype, in the NHMUK, revealed external and genitalia morphology consistent with a common southeastern Brazilian species similar to C. incerta, C. solvens, and related species. Examination of males from the Brazilian population helped to solidify our understanding of the morphology of "true" members of this genus. St Laurent et al. (2018a)  Cicinnus: Cicinnus s. l. and Cicinnus s. s., the latter which included C. incerta, a close relative of type species C. orthane. This dichotomy of Cicinnus suggests at least two genera must be recognized to eliminate paraphyly in Cicinnus s. l. In order to reorganize Cicinnus for the present study, we break the genus into three morphologically distinct groups, and a fourth uncertain grouping lacking defining morphological characteristics. We do not formally assign names to these groups at this time, but separate them for organizational purposes. The group numbers are unrelated to Schaus' (1928) Cicinnus numbering system. Group 1 is Cicinnus s. s. and includes the "true" Cicinnus as per the phylogenetic results of St Laurent et al. (2018a). The remaining Cicinnus s. l. may eventually be transferred to a revalidated Gonogramma, see Annotation 97 below. 97 Gonogramma was named by Boisduval (1872) in association with Psychocampa, and has since been considered, apparently only inexplicitly, a synonym of Cicinnus. Gonogramma was proposed without any associated species, so Fletcher and Nye (1982) gave C. despecta as the type species of the genus but suggested that C. despecta was congeneric with the type of Cicinnus, and hence inferred that Gonogramma and Cicinnus were synonymous. The generic combination was apparently overlooked in the following years as Becker (1996) maintained C. despecta in Cicinnus and did not provide Gonogramma in the list of synonymies of Cicinnus. St Laurent et al. (2018a) raised the possibility that Gonogramma could be revalidated to include the "Cicinnus sensu lato" species that have hyaline patches on the wings since this feature, as well as the genitalia of these taxa, are distinct from those of Cicinnus s. s. Ongoing molecular research, which includes many more of the Cicinnus s. l. species, as well as the type species of Gonogramma, C. despecta, supports the use of Gonogramma and the eventual transfer of all Cicinnus s. l. (Groups 2-4) species to this genus, including at least one that does not have hyaline patches on the wings (St Laurent et al. in prep.). 98 Group 1, which we treat here as Cicinnus sensu stricto, includes species lacking hyaline patches on both the fore and hindwings, diffuse coloration with heavy speckling of dark brown to black petiolate scales, and short mesal arms of the gnathos. The gnathos originates from near the dorsal juncture between the vinculum and the base of the uncus. Vincular arms may be present or absent (the only Cicinnini group to have this character vary). St Laurent et al. (2018a) sampled C. incerta and C. melsheimeri (the former being a proxy for Cicinnus type species C. orthane) and found them sister to each other in a robustly supported clade. 99 The year of description for C. incerta was incorrectly given as 1877 by Becker (1996), which was apparently when the first volume of Möschler's "Beiträge zur Schmetterlings-Fauna von Surinam" was published, when in fact C. incerta was described in the second volume from 1878. 100 Cicinnus latris was described from an indeterminate number of specimens from two localities in Costa Rica. Considering that syntypes and different localities are involved in the type series representing C. latris, we designate a lectotype for the male specimen labeled as the "type" in the USNM (Fig. 168). This speci-men bears the following labels: Tuis CR/ Sept./ Collection W. Schaus/ Type No. 18958 [or 16958, the (Harris 1841a). Hopefully our clarification here will prevent others from making this mistake. Neither Perophora nor Saccophora are valid names as both are preoccupied. Grote (1896) proposed Ptochopsyche as a replacement name for Perophora, but this name is a synonym of the senior name Cicinnus Blanchard, 1852 since both C. melsheimeri and type species C. orthane are congeneric. 102 St Laurent and McCabe (2017) provided discussion regarding the names egenaria and primolus, confirming the synonymy of egenaria and primolus with C. melsheimeri. 103 As previously mentioned in Annotation 96, the type locality of C. orthane may be erroneous. Mimallonidae are exceedingly rarely collected in Chile, with only one other record of the family reported by Ureta (1957). There are also no mimallonid specimens in the major collection in Santiago, Chile (C. Mielke and A. Ugarte pers. comm.). The first author has not seen any mimallonid specimens from Chile despite examining specimens in over 50 institutional and private collections in Europe, Latin America and North America. A specimen labeled "Cicinnus orthane Chili" in the NHMUK has been designated as a lectotype by St Laurent et al. (2018c). External and genital morphology of the lectotype matches that of a common southeastern Brazilian species that has gone unnamed, and thus it is possible that the original specimen of C. orthane was of Brazilian origin and was mislabeled. Assuming that the lectotype is the same specimen on which Blanchard based his illustration and description, the concept of Cicinnus s. s. is solidified. 104 Cicinnus solvens was described from three syntypes, two male and one female, from two different localities in Panama. Considering that syntypes and different localities are involved in the type series representing C. solvens, we designate a lectotype for the male specimen labeled as the "type" in the USNM (Fig. 169). This specimen bears the following labels: CabimaPan May 20. .11 August Busck/ Type No. 16102 U.S.N.M/ USNM-Mimal: 1038/. A red handwritten label reading "LECTOTYPE♂ Cicinnus solvens Dyar, designated by St Laurent" has been added to the specimen. 105 Cicinnus tuisana was described from an indeterminate number of specimens from two localities in Costa Rica. Considering that syntypes and different local-ities are involved in the type series representing C. tuisana, we designate a lectotype for the male specimen labeled as the "type" in the USNM (Fig. 170). This specimen bears the following labels: JuanVinas CR/ Jan 09/ Cicinnus tuisana Type Schaus/Type No. 16956 U.S.N.M/ USNM-Mimal: 1138/. A red handwritten label reading "LECTOTYPE♂ Cicinnus tuisana Schaus, designated by St Laurent" has been added to the specimen. 106 Group 2 (Cicinnus s. l., in part) contains species recognizable by hyaline patches on the forewings or on both fore and hindwings, as well as by the absence gnathos arms and the presence of vincular arms in the male genitalia. These taxa will likely eventually be placed in a revalidated Gonogramma once molecular phylogenetics can more robustly determine the relationships within Cicinnini. 107 The first author dissected a topotypical male specimen of C. belaria comb. n. (St Laurent diss.: 3-15-16:1 in USNM) and hereby transferred this species Cicinnus from Psychocampa due to genitalia morphology, which are largely indistinguishable from C. despecta. Upon close examination of external maculation, it is apparent that C. belaria also resembles C. despecta externally, differing primarily in more obscure coloration and in size and narrowness of the forewings. All of our morphological phylogenetic analyses support this transfer, consistently placing C. belaria within a broader Cicinnus s. l. clade. 108 Cicinnus corallina was described from three syntypes from Colombia and Panama. Because syntypes and specimens from different localities are part of the type series of C. corallina, we designate a lectotype for the male specimen labeled as the "type" in the USNM (Fig. 167). This specimen bears the following labels:  St Laurent" has been added to the specimen. 109 As for Menevia plagiata (Annotation 5), the original primary type of C. despecta is lost, along with many types described by Walker originating from the Fry collection (Becker 2001). However, Walker's (1855) original description clearly refers to subsequent authors' and our concept of C. despecta, namely by the presence of a "broad red dorsal stripe" of the abdomen and the "red band" accompanying the outer edge of postmedial lines, characters that do not fit any other mimallonid species known from Rio de Janeiro. Considering the broad distribution of C. despecta s. l., and the various names treated as synonymous with C. despecta (but from distant type localities) we hereby designate a neotype for a specimen from Rio de Janeiro in NHMUK that most accurately reflects Walker's description of this species, and thus establishes nomenclatural stability to the name despecta (Fig. 171) Note that the name "sanguinolenta" is on one of these labels, but this name was commonly applied to specimens of C. despecta in the NHMUK by historical workers. The specimen that we designate as the neotype is not a syntype of C. sanguinolenta. The true syntype of C. sanguinolenta is a specimen of known location (deposited in the NHMUK), and bears very different labels of Felder's time. The three synonyms (sensu Becker (1996)) of C. despecta come from three localities: curtisea from Argentina, sanguinolenta from Brazil, and sachinius from Minas Gerais, Brazil. After comparing the type of sanguinolenta to the neotype of despecta, it is clear that sanguinolenta should remain a synonym of C. despecta. However, the type of curtisea is not known to us, and the type of sachinius appears quite distinct from the typical C. despecta. Therefore, we are unable to clearly rectify the synonymy of these names until a thorough revision of Cicinnus s. l. is completed. 110 Cicinnus curtisea potentially represents a valid species considering the geographic separation of the type localities of C. curtisea and C. despecta (see also  Dognin (1924), but this synonymy was overlooked by Schaus (1928), Gaede (1931), and Becker (1996). Cicinnus roscida was treated as valid as recently as Becker (1996), and we hereby revive this synonymy upon examining numerous specimens of this species and the type of C. roscida. 115 Although C. manicora comb. n. displays valva and phallus shapes rather unique to this species, the presence of vincular arms, absence of gnathos arms, and presence of hyaline patches on the forewings all are clear indicators of Cicinnus s. l. Group 2, and we therefore transfer this species accordingly from Psychocampa, where it was placed apparently due to overly narrowed wings and sphingiform shape, characters that are superficially similar to some Psychocampa. All of our morphological phylogenetic analyses, as well as ongoing molecular phylogenomics that include this species (St Laurent et al. in prep.), support this transfer, consistently placing C. manicora within a broader Cicinnus s. l. clade. 116 Schaus (1928) suggested that C. thermesia may be a synonym of C. corcovada, and that the types are merely the male and female (respectively) of the same species. Both species are similar, but upon examining several specimens in the NHMUK and USNM and comparing these specimens directly to type material, we believe these two names belong to separate species. Cicinnus corcovada displays slightly more rounded forewings, generally with a more convex postmedial line, and redder coloration, whereas C. thermesia has more angulate wings, often straighter postmedial lines, and a grayer coloration. A comparison is more easily made between the females of the two species. Unfortunately, whoever dissected the holotype of C. thermesia did not properly mount the entire genitalia (only abdominal segments are mounted on the slide); therefore, it will be impossible to accurately identify C. thermesia with a proper examination of type genitalia. 117 Group 3 (Cicinnus s. l., in part) contains species recognizable by hyaline patches on both the forewings and hindwings, as well as by the presence of gnathos and vincular arms in the male genitalia. Most species in this group were either already described in Cicinnus, or were transferred to Cicinnus from Psychocampa by St Laurent and Cock (2017) based on male genitalia morphology. These taxa will likely eventually be placed in a revalidated Gonogramma once molecular phylogenetics can more robustly determine the relationships within Cicinnini. 118 St Laurent and Cock (2017) transferred C. beta, along with several other species treated here in Cicinnus s. l. Group 3, to Cicinnus from Psychocampa based on genitalia morphology. However, these authors incorrectly listed this taxonomic change as "comb. n." when in fact this was a revived combination since Schaus (1910) had originally described this species in Cicinnus. 119 Cicinnus gaia displays bizarre, atrophied valvae, seemingly making this species appear not to be part of Cicinnus s. l. Group 3. However, all characters unique to Group 3 are present in C. gaia, including vincular and gnathos arms, and hyaline patches on the fore and hindwings. Tornal diffusions that are present in the above species are also present in C. gaia, but are less discernable due to the lack of contrast in this more darkly colored moth. 120 Cicinnus bibula comb. n. is a very rare species in natural history collections. We are aware of only seven specimens in collections, and all appear to be female, although one in the MNHN (only examined via photograph) may be male. Considering the rarity of this species, and our inability to examine male genitalia, we preliminarily transfer C. bibula to Cicinnus s. l. from Psychocampa due to the presence of hyaline patches on the forewings, a character absent from all Psychocampa. Examination of female genitalia of C. bibula, reveals characters quite similar to those of female Cicinnus s. l., Isoscella, and Roelmana, namely the longitudinal wrinkles of the lamella antevaginalis and setae covered lobes on either side of the lamella antevaginalis. These same characters are characteristic of at least three Cicinnini genera, therefore although our generic placement of C. bibula is preliminary, the tribal assignment is most certainly correct.  Schaus (1928), Gaede (1931), and Becker (1996). St Laurent and Carvalho (2017b) recognized the distinctness of this species, as well as three additional similar new species, and therefore described a new genus in which to place them. St Laurent et al. (2018a) included I. ecuadoriana in their phylogenetic analysis and found Isoscella to belong to a distinct lineage including several species also postulated by St Laurent and Carvalho (2017b) Laurent and Carvalho (2017b). Roelmana contains two distinct groups of species, which may formally be determined to be species-groups or genera in a revision of Roelmana or a more densely sampled phylogeny. The typical group includes species with hyaline patches on the fore and hindwings (as in the sister genus Isoscella), whereas the other group contains species lacking hyaline patches. 126 Of all species previously and here transferred to Roelmana, R. beneluzi comb. n.
displays the most divergent male genitalia, and may eventually prove to be more appropriately placed in a separate new genus. We hesitate to do so here pending ongoing molecular phylogenetic work that will likely be able to include R. beneluzi. The gnathos shape however, is consistent with other taxa placed in Roelmana here and in St Laurent et al. (2018a), and externally this species displays similar diffuse, weakly defined patterning as in R. laguerrei and R. prominens. Therefore, Roelmana is certainly a more systemically relevant generic assignment than Cicinnus. Furthermore, R. beneluzi was recovered sister to R. laguerrei in all morphological phylogenetic analyses, thus supporting this preliminary transfer. 127 Externally, R. brasiliensis comb. n. is nearly identical to the rare R. fenestrata comb. n., and in the original description of R. brasiliensis this similarity was unfortunately not addressed. Roelmana fenestrata is only known from the female type (erroneously reported and labeled as a male), collected in Castro, Paraná, Brazil, a locality that sits between Cerrado and Mata Atlântica. Roelmana brasiliensis is a Cerrado endemic, therefore it is possible that these two names are synonymous. However, the hyaline discal marking on the hindwing is closer to the postmedial line in R. fenestrata than in females of R. brasiliensis. Pending the availability of male specimens from near the type locality of R. fenestrata, and the loss of the type genitalia of R. fenestrata, it is not possible to reliably synonymize these names at this time.