A phylogenetic and taxonomic review of baviine jumping spiders (Araneae, Salticidae, Baviini)

Abstract The systematics and taxonomy of the tropical Asian jumping spiders of the tribe Baviini is reviewed, with a molecular phylogenetic study (UCE sequence capture, traditional Sanger sequencing) guiding a reclassification of the group’s genera. The well-studied members of the group are placed into six genera: Bavia Simon, 1877, Indopadilla Caleb & Sankaran, 2019, Padillothorax Simon, 1901, Piranthus Thorell, 1895, Stagetillus Simon, 1885, and one new genus, Maripanthus Maddison, gen. nov. The identity of Padillothorax is clarified, and Bavirecta Kanesharatnam & Benjamin, 2018 synonymized with it. Hyctiota Strand, 1911 is synonymized with Stagetillus. The molecular phylogeny divides the baviines into three clades, the Piranthus clade with a long embolus (Piranthus, Maripanthus), the genus Padillothorax with a flat body and short embolus, and the Bavia clade with a higher body and (usually) short embolus (remaining genera). In general, morphological synapomorphies support or extend the molecularly delimited groups. Eighteen new species are described: Bavia nessagyna, Indopadilla bamilin, I. kodagura, I. nesinor, I. redunca, I. redynis, I. sabivia, I. vimedaba, Maripanthus draconis (type species of Maripanthus), M. jubatus, M. reinholdae, Padillothorax badut, P. mulu, Piranthus api, P. bakau, P. kohi, P. mandai, and Stagetillus irri, all sp. nov., with taxonomic authority W. Maddison. The distinctions between baviines and the astioid Nungia Żabka, 1985 are reviewed, leading to four species being moved into Nungia from Bavia and other genera. Fifteen new combinations are established: Bavia maurerae (Freudenschuss & Seiter, 2016), Indopadilla annamita (Simon, 1903), I. kahariana (Prószyński & Deeleman-Reinhold, 2013), I. sonsorol (Berry, Beatty & Prószyński, 1997), I. suhartoi (Prószyński & Deeleman-Reinhold, 2013), Maripanthus menghaiensis (Cao & Li, 2016), M. smedleyi (Reimoser, 1929), Nungia hatamensis (Thorell, 1881), N. modesta (Keyserling, 1883), N. papakula (Strand, 1911), N. xiaolonghaensis (Cao & Li, 2016), Padillothorax casteti (Simon, 1900), P. exilis (Cao & Li, 2016), P. flavopunctus (Kanesharatnam & Benjamin, 2018), Stagetillus banda (Strand, 1911), all comb. nov. One combination is restored, Bavia capistrata (C. L. Koch, 1846). Five of these new or restored combinations correct previous errors of placing species in genera that have superficially similar palps but extremely different body forms, in fact belonging in distantly related tribes, emphasizing that the general shape of male palps should be used with caution in determining relationships. A little-studied genus, Padillothorus Prószyński, 2018, is tentatively assigned to the Baviini. Ligdus Thorell, 1895 is assigned to the Ballini.


Introduction
Baviines are tropical Asian jumping spiders with elongate medium to large bodies, living above ground on leaves or branches of vegetation or in suspended litter (Maddison 2015a). The Baviini is by far the smallest of the four major groups that compose the clade Salticoida (Maddison, 2015a), with far fewer described species (31 to date) than the other three groups (Astioida, ~ 600 species; Marpissoida, ~ 850; Saltafresia, ~ 3500). Baviines have been little studied, but recent works have begun to add to our knowledge quickly (Caleb and Sanap 2017;Kanesharatnam and Benjamin 2018;Malamel et al. 2019;Caleb et al. 2019;Nafin et al. 2020), resulting so far in at least six genera currently placed in the Baviini (Bavia Simon, 1877, Bavirecta Kanesharatnam & Benjamin, 2018, Indopadilla Caleb & Sankaran, 2019, Padillothorax Simon, 1901, Piranthus Thorell, 1895, and Stagetillus Simon, 1885Maddison 2015a;Kanesharatnam and Benjamin 2018;Malamel et al. 2015Malamel et al. , 2019Caleb et al. 2019). The frequency of new species in recent collecting suggests that the group is much more diverse than it appears; e.g., Piranthus includes only two South Asian species described, but there are at least three new species in Singapore alone. The purpose of this paper is to describe some of the recently discovered species, and to present the first phylogeny for the group. The phylogenetic work leads to a revised classification of species into genera, one of which is new.

Material examined
Spider specimens examined for this study are stored in the University of British Columbia Spencer Entomological Collection, Canada (UBCZ), the Lee Kong Chian Natural History Museum, Singapore (LKCNHM, https://lkcnhm.nus.edu.sg), the Research Collections at National Centre for Biological Sciences, Bengaluru, Karnataka, India (NCBS, http://collections.ncbs.res.in), the Florida State Collection of Arthropods (FSCA), and the Senckenberg Museum, Frankfurt, Germany (SMF). Work on Indian material was done at the National Centre for Biological Sciences, Bengaluru. Work on Sri Lankan material was done at the National Institute of Fundamental Studies, Kandy. Data for specimens collected during fieldwork by WPM and colleagues includes a code beginning "WPM#" representing not a specimen identification, but a collecting event of location and time.

Morphology
Preserved specimens were examined under both dissecting microscopes and a compound microscope with reflected light. Drawings were made with a drawing tube on a Nikon ME600L compound microscope. Most photographs of living specimens were made with either a Pentax Optio 33WR digital camera with a small lens glued to it for macro capability (2016 and earlier) or an Olympus OM-D E-M10 II camera with 60 mm macro lens (2017 and later). Microscope photographs were made either on a Nikon ME600L compound microscope or an Olympus SZX12 stereoscope and focus stacked using Helicon Focus 4.2.7.
All measurements are given in millimeters. Descriptions of color pattern are based on the alcohol-preserved specimen. Carapace length was measured from the base of the anterior median eyes not including the lenses to the rear margin of the carapace medially; abdomen length to the end of the anal tubercle. The following abbreviations are used: ALE anterior lateral eyes; ECP epigynal coupling pocket (also known as a hood or notch); PLE posterior lateral eyes; PME posterior median eyes (the "small eyes"); RTA retrolateral tibial apophysis; TmA terminal apophysis.

Molecular data
Molecular data was gathered for three gene regions by traditional Sanger PCR methods, and for many genes by Ultra-Conserved Element (UCE) target enrichment sequencing methods (Faircloth 2017), combined with data from the literature, to assemble a dataset of nine outgroup species and 22 baviine species. Table 1 lists the specimens used in the molecular study; Table 2 lists the data gathered for each. Table 3 lists specimens used for a small auxiliary study of Nungia Żabka, 1985 andCapeyorkia Richardson, 2016. Table 2. Molecular data used for phylogenetics of baviines. UCE probeset was either arachnid (A) or spider (S). SRA indicates sequence read archive accession number. "Reads Pass QC" indicates number of reads retained after quality control and adapter removal via Illumiprocessor. "UCE loci" counts number of loci both as obtained by mixed spider-arachnid file, and after filtering by occupancy and branch length criteria. "Filtered length" is number of non-missing sites, i.e. total base pairs or sequence length. For 28S, mtDNA, 16SND1, and COI, numbers indicate sequence length as obtained by Sanger sequencing (σ) or bycatch from UCE sequencing (β). New Sanger sequences show Genbank accession numbers of the form MW0818##. mtDNA indicates taxa for which entire mitochondrial genome recovered as bycatch. Published sequences indicated by citation (MH03: Maddison and Hedin 2003   , to assist as outgroups. Raw demultiplexed reads were processed with Phyluce version 1.6 (Faircloth 2016), quality control and adapter removal were conducted with the Illumiprocessor wrapper (Faircloth 2013), and assemblies were created with SPAdes version 3.14.1 (Nurk et al. 2013), using the meta option, at default settings. From among the contigs thus assembled, those matching particular UCE probes were pulled out using the Phyluce pipeline at default settings. Because some taxa were captured using the arachnid probeset (outgroups Attulus, Breda, Colonus, Salticus), and others using the spider probeset (remaining outgroups, and all baviines), a blended probeset file was needed to best pull out UCE contigs, because each of the arachnid and spider probesets includes loci not included by the other. Kulkarni et al.'s (2019) spider probeset includes (i) some of Starrett et al.'s (2017) arachnid probes directly, (ii) others for the same loci but modified to target spiders better, and (iii) others for new loci. Because Kulkarni et al. do not identify probes of the second category as such, we sought to identify whether spider probes are orthologous to arachnid probes. We then deleted from the probeset file those arachnid probes matching spider probes, as including duplicate homologs reduces data recovery (contigs matching two probes are removed by Phyluce for being problematical). To determine homology, contigs from 18 diverse species in the Salticinae, 12 captured with arachnid probes, 6 with spider probes, were each matched against both arachnid and spider probesets. Any instance of a contig matching both a spider probe and arachnid probe, as assessed using a script examining .lastz files, was taken as indicating homology between the probes. Arachnid probes that showed no such hint of homology to spider probes were then added to Kulkarni et al.'s (2019) spider probeset to generate the blended probeset (see Suppl. material 1). The spider probeset includes 15015 probes and probe parts; the arachnid probeset, 14799; blended, 25689. The efficacy of the blended probeset can be seen in the numbers of loci recovered in the baviine dataset reported here: the arachnid probset pulled out on average 134 loci from spider-enriched taxa and 411 from arachnidenriched taxa; the spider probeset pulled out on average 1118 and 113 respectively; the blended probeset pulled out on average 1123 and 415. Nonetheless, many of the UCE loci recovered from the arachnid-enriched taxa were only among those taxa; this explains why many were subsequently deleted when a filter for occupancy among ingroups (see below) was applied.
Recovered UCE loci were aligned with MAFFT (Katoh and Standley 2013) and trimmed with Gblocks (Castresana 2000;Talavera and Castresana 2007), using -b1 0.5, -b2 0.5, -b3 10, -b4 4 settings in the Phyluce pipeline. Among the loci recovered, those with fewer than 6 taxa total or fewer than 3 ingroups were deleted. As in the analysis of , loci were also deleted over concerns about paralogy if their gene tree showed a very long branch, at least 5× longer than the second longest branch.

Data for 28S and mitochondrial genes
For the new Sanger-sequenced data, specimens were preserved, their DNA extracted, and sequences obtained for the nuclear gene 28S and the mitochondrial gene regions 16SNDI and COI following the protocols using the protocols of Zhang and Maddison (2013) and Maddison et al. (2014). Alignments were done by MAFFT with the L-INS-i option.
The same three gene regions were also present among the sequence capture genomic contigs as untargeted bycatch. We recovered them by constructing a local BLAST database of the contigs of each taxon, and querying it with 28S, 16SND1 and COI sequences from eight to nine different salticid species (Bavia, Indopadilla, Bathippus, Harmochirus, Idastrandia, Langerra, Phintella, Platycryptus, Salticus, Attulus, Lyssomanes), retaining any contigs matching with an e-value of less than 10 -10 and length greater than 200. In a few cases, multiple contigs from a taxon were recovered as matching a single locus, but after alignment against others these could be interpreted as different parts of the target gene, and were thus stitched together to a single sequence.
3. "restricted mtDNA+28S" -A restricted version of 28S and mtDNA with the long bycatch sequences trimmed to put the taxa with legacy Sanger data on almost equal footing (i.e., approximately as much data) with those with bycatch. 28S sites at start and end were trimmed until at least 3 of the shorter legacy Sanger sequences were represented. The same rule for trimming was used for the mtDNA before 16SND1, between 16SND1 and CO1 genes, and after COI. Analyzed with 2 partitions, 28S and mtDNA.
For the partitioned analyses, the options -m MFP -spp were used (extended model selection followed by tree inference, edge-linked partition model, with partition-specific rates); for the unpartitioned analyses, -m MFP (extended model selection followed by tree inference, edge-linked partition model, no partition-specific rates).
Several species of elongate brown Asian and Australasian salticids were initially identified in the field as baviines, but were excluded from the Baviini by preliminary molecular analyses and closer morphological study, which showed them to be near Nungia epigynalis Żabka, 1985. To document this and clarify the limits of the Baviini, we did a small analysis based on 28S and 16SND1, using previously published sequences of baviines, astioids, and other groups from Maddison and Hedin (2003), Bodner and Maddison (2012), and Maddison et al. (2014) combined with new sequences of the species in question (Table 3). Details are given in the Taxonomy section under Viciriini: Nungia.
Raw sequence reads from UCE capture are deposited in the Sequence Read Archive (BioProject PRJNA667925, https://www.ncbi.nlm.nih.gov/sra/PRJNA667925) with accession numbers shown in Table 2. Genbank accession numbers of Sanger sequenced genes are shown in Tables 2 and 3. Alignments and trees are deposited in the Dryad data repository (https://doi.org/10.5061/dryad.4f4qrfj9j).

Taxonomic authority
The taxonomic authority for all nomenclatural acts (synonymies, new combinations, new species) is W. Maddison.

Molecular data obtained
UCE data obtained are outlined in Table 2. The Phyluce pipeline recovered aligned matrices of 1837 loci using the blended spider-arachnid probeset. In these initially recovered loci, the 16 taxa originally sequence-captured with the spider probeset had on average 1123 loci and 797,222 base pairs of sequence; the four originally sequencecaptured with the arachnid set had on average 415 loci and 158,475 base pairs. From the original 1837 loci, 511 were deleted because they had fewer than six taxa total or fewer than three ingroups. Thirteen loci were deleted for having the longest branch more than 5 × longer than the second longest. This resulted in resulting in a final set of 1313 loci, in which the 16 spider-probeset taxa had on average 1076 loci and 775,244 base pairs of sequence; the four arachnid-probeset taxa had on average 104 loci and 53,936 base pairs. The strong decline in arachnid-probeset taxa from 415 loci to 104 after occupancy filtering suggests that Kulkarni et al.'s (2019) incorporation of arachnid-unique probes into the spider probeset included only a portion. The 1313 loci were concatenated into a single alignment with 1,050,217 sites.
The bycatch 28S sequences were 435-6894 bp long (average 5675.9) and aligned well with the Sanger 28S sequences using MAFFT at default settings. Bycatch 28S were obtained for three specimens for which previous Sanger sequences were available and identical (I. kahariana SWK12-1163, 1021 base pairs; M. reinholdae SWK-01991, 1037 bp; S. cf. opaciceps MRB079, 1067 bp) except for a one base difference at the start of the I. kahariana sequence. An initial alignment of 28S including UCE bycatch and legacy Sanger data was 18646 bp long, but the first 1593 bp and the last 9793 bp of the alignment were poorly aligned and represented by only a few bycatch sequences, and so were trimmed. After addition of a few other taxa and realignment, the final 28S alignment was 7281 bp long.
Among the bycatch contigs for 10 taxa were long sequences containing both the 16SND1 and COI regions, and whose size (12568-15601 bp) suggests they may be the whole or nearly whole mitochondrial genome (marked in the column "mtDNA" in Table 2). For other taxa, the mitochondrial contigs recovered as matching 16SND1 and COI were separate and fairly short (16SND1: 362, 1136, 1835 bp; COI: 467, 596, 1083 bp). Bycatch 16SND1 were obtained for two species for which previous Sanger sequences were available and identical. For S. cf. opaciceps (MRB079) the 362 bp bycatch sequence was identical to the 908 bp Sanger sequence in the 303 bp of overlap; the sequence used was their concatenation. For Helpis, the 956 bp Sanger sequence published by Maddison and Hedin (2003) matches exactly the longer bycatch sequence here obtained from a different specimen, NZ19-9152, also from New Zealand. The COI obtained from those same two specimens (Sanger and bycatch likewise, respectively) is identical except for one nucleotide, but the 721 bp sequence reported by Maddison et al. (2014) for a different specimen of Helpis from Papua New Guinea differs at 35 sites, and is thus likely a different species. The bycatch sequences were used in their entirety, except for a terminal 183 bp of a bycatch 16SND1 from M. reinhol-dae SWK12-1991, which was deleted because it showed no clear alignment with that portion of other taxa; we suspect that portion may have mistakenly assembled with 16SND1 because of a shared poly-AT repeat.
The whole mtDNA sequences aligned well against each other using MAFFT at default settings. The shorter bycatch sequences and Sanger 16SND1 and COI aligned well against the whole mtDNA. The whole mtDNA sequences initially differed in their (arbitrary) starting point on the circular mitochondrial genome, but after a preliminary alignment they were adjusted to all begin at a conserved region in 16S. After adjustments by hand to align ND1 and COI without gaps, the other regions (before ND1, between ND1 and COI, and between COI and the end) were re-aligned using MAFFT.
Trimming of the 28S and mtDNA alignments as explained for dataset #3 ("restricted mtDNA+28S") yielded a 28S alignment of 1181 base pairs, and a mitochondrial alignment of 2211 bp.

Phylogenetic results
The primary phylogenetic results are shown in Figs 1-3. The different datasets (UCE, Sanger, and combined) gave substantially concordant results, with just a few points of disagreement. All confirm the monophyly of the baviines, and divide the group into three major clades (marked in Figs 1-3), one with a relatively long embolus that appears freely moveable (the Piranthus clade), one with a short erect embolus and flat body (the genus Padillothorax), and one with a relatively short embolus largely fixed to the tegulum and a higher body (the Bavia clade). The relationship among these three groups varies by dataset: Padillothorax is sister to the Piranthus clade by the mitochondrial and 28S data (Fig. 3), but by the much larger UCE and combined datasets it is placed reasonably securely as a deep-branching sister to the Bavia clade.
Relationships within each of the three major clades is reasonably stable across datasets. Within the Piranthus clade, the morphologically distinctive Piranthus is monophyletic, as is Maripanthus. The morphologically similar Padillothorax badut and P. mulu are sisters, as are P. flavopunctus and P. cf. flavopunctus. Within the Bavia clade, molecular results more or less match morphological groups: Bavia with relatively large bodies, Indopadilla with ridged chelicerae, thoracic bulges, and exposed clypeal arthrodial membrane, and the elongate yellow-orange Stagetillus. Accordingly, the concepts of genera here come from both morphological and molecular evidence.
Nungia are generally smaller than baviines, with a more flat-topped carapace; they are quite distinct by molecular data , and see below under Viciriini), but the palps are similar. Nungia is discussed further under Viciriini, below.
Padillothorus (see Prószyński 1984Prószyński , 2018 is included tentatively among the baviines based primarily on the similarity of its body form with baviines, and its geographic distribution. Padillothorus has five retromarginal cheliceral teeth (Reimoser, 1927), which is typical for baviines. Although Ligdus Thorell, 1895 might appear as a candidate to belong in the Baviini, with narrow body and raptorial front legs, the juvenile type specimen (in NHM London, examined) is such a close match to Copocrossa Simon, 1901 as to be possibly a senior synonym thereof. Ligdus shares with Copocrossa and Mantisatta the peculiar feature of a first metatarsus bent ventrally almost 90 degrees near its base. Because of its apparent relationship with these ballines, Ligdus is therefore moved to the Ballini.
Figures 4-35 illustrate character variation in baviines, some of which is used to diagnose genera. Among baviines, the genera vary in these characters:  Shape of the carapace -flatter (height < 36% of length) in most Piranthus and Padillothorax, higher in others. Indopadilla has distinctive bulges on the thorax sides (Figs 5,14).
Lateral margin of male endites -Indopadilla and Piranthus males have the endite with a simple rounded margin, but the other genera have varying projections (Figs 6-12). Maripanthus bears a sharp distal-retrolateral corner. Bavia and Padillothorax have a lobe, in some thumb-like (Figs 6, 10), in others broad and triangular (Fig. 9). Stagetillus opaciceps has a large broad lobe projecting retrolaterally (Fig. 8), though S. irri lacks this lobe, and instead has only a mild corner at the retrolateral distal tip, not too different from Indopadilla.  malala 2007], and some marpissines [Edwards 2006]) they are in a more dorsal or distal position; in Padillothorax they are at least one quarter of the femur's length from the distal tip, approx. midway between dorsal and ventral.
In several baviine genera, not each others' closest relatives, there is a characteristic series of markings consisting of small patches of pale scales on the thorax: one patch medially between the PLE, a short longitudinal stripe at the top of the thoracic slope, and one behind each PLE (e.g., Bavia, Fig The taxonomic account below presents in sequence the Bavia clade, Padillothorax, and then the Piranthus clade.  Larger-bodied than most other baviines. Carapace relatively broad and having hexagonal shape, widest at or just behind the PLEs (Fig. 13). Chelicerae lack the sharp lateral ridge (Fig. 4) of Indopadilla. Embolus shorter than length of tegulum, arising in all known species on bulb's distal prolateral corner. ECP on a prominent medial bulge. Male endite with small thumb-like lobe laterally (Fig. 6), as in the Padillothorax badut group.
Illustrations are given here of some of the well-known species of Bavia, includ- The one figured here as B. cf. capistrata could be different from that figured by Cao, Li, and Żabka. The former shows a slightly wider embolus and more delicate ECP, and possibly more contrasting markings. Regardless, C. L. Koch's species is removed from synonymy with Evarcha flavocincta and returned to Bavia.
In addition to the species below, we have seen an undescribed species near B. nessagyna from Mulu National Park (single female) and a species near B. intermedia (single male, here represented as specimen d079 in the Sanger data). Etymology. From the Greek nessa, duck, and gyne, female, referring to the resemblance of the epigyne to a duck's bill. Other names: In WPM's lab notebooks the informal code for this species was "BVDUC-S".

Bavia nessagyna
Diagnosis. One of the more delicate Bavia, along with B. capistrata, having legs II-IV very much paler than I, and thus resembling Indopadilla. Its distinction from B. fedor is slight: the embolus of both appears as a curved and narrowing blade with a series of retrolateral teeth. The teeth are short, triangular, and closely spaced in B. nessagyna, but larger in B. fedor, appearing as broad pillars whose bases are well separate (photographs of holotype kindly supplied by J. Boone, Bishop Museum). In B. nessagyna the teeth are not on the embolus proper but on a TmA that parallels the embolus (Fig. 64). In B. fedor it is unclear whether the teeth are on a TmA or on the embolus itself; Prószyński's illustration (Berry, Beatty and Prószyński 1997) shows no division into two processes, but it may be that they are closely adpressed. Even still, the overall shapes of the embolic division differ: broad at the base in B. nessagyna but abruptly narrowing; narrower at the base in B. fedor and narrowing more gradually. Epigyne with arcing ridges lateral to Description. Male (based on holotype, specimen IDWM.20005). Carapace length 3.6; abdomen length 4.8. Carapace (Fig. 71): Broad, seeming even broader because of small tuft of setae at widest point just posterior to PLEs. Integument of ocular area black, thorax dark brown, except in alcohol pale orange yellow area just behind the ocular area. Bands of cream coloured scales lie just posterior to and lateral to the ocular area; thorax with a few small spots of cream scales, the two posterior ones of which are prominent and well separated, as in B. capistrata. Clypeus dark and glabrous, but a disorderly fringe of long cream setae overhangs the chelicerae (Fig. 70). Chelicerae simple and mostly vertical, dark, with cream setae that, with those of the clypeus, give the appearance of an unkempt moustache. At least two teeth on retromargin (paratype with 6). Palp as in Fig. 64; embolus thin and accompanied by toothed TmA. Femur dark basally, but terminally is pale, as are all more distal segments. Endites with thumblike lobe laterally (Fig. 6). Legs II-IV notably paler than legs I, which bear annulae of white scales on patella and distally on tibia. Metatarsus I dark; tarsus pale. Ventral fringe of black setae beneath tibia and metatarsus I. Abdomen with dorsal markings primarily transverse, with four transverse pale bands separated by three dark bands.
Female (based on paratype IDWM.20003). Carapace length 3.3; abdomen length 5.9. Carapace as in male, but lacking tuft at widest point. Clypeus and Chelicerae as in male. Five retromarginal teeth. Legs similar to those of male, but with first leg not quite so dark, and without the ventral fringe so distinctly developed. Abdomen with only a hint of transverse bands, instead dominated by longitudinal stripes: a narrow central pale band flanked by wide dark bands. Epigyne with ECP beneath a posteriorlyprojecting mound (Figs 66-67).
Natural history. All specimens from Lambir Hills were collected on big-leaved foliage (e.g., palms) except for IDWM.20004 whose collecting record says simply "foliage".
Additional material examined.

Species included.
Indopadilla annamita (Simon, 1903) (Berry, Beatty & Prószyński, 1997), comb. nov., transferred from Bavia Indopadilla suhartoi (Prószyński & Deeleman-Reinhold, 2013), comb. nov., transferred from Bavia Indopadilla thorelli (Simon, 1901) Indopadilla vimedaba Maddison, sp. nov. Diagnosis. Front face of chelicerae concave or flat in both males and females, bordered laterally by a sharp ridge (Fig. 5). The ridge projects laterally in some males (e.g., Fig. 98, triangle, or see the male of I. vimedaba figured by Prószyński 1987, p. 105, as "Stagetilus semiostrinus"). (Males of Bavia may have chelicerae with a flat front face, and males of Padillothorax may have the front face concave medially, but none have such a sharp or extended lateral ridge.) Clypeus very narrow at centre, exposing a broad expanse of arthrodial membrane that in most species is white (Figs 96, 115, 128;Żabka 1988 Spermatheca anterior to copulatory openings (posterior in other baviines). Retromarginal teeth of male not clustered onto a raised hump; rather, the teeth occur on a long sharp ridge that extends medially from base of fang (e.g., Prószyński 1987, p. 105). Male endite rounded, lacking lobe or sharp corner (Fig. 7). Thorax with a bulge just behind the PLE that gives the carapace a "muscular" appearance (Figs 5, 14; see triangle on Fig. 14). (This area contains the muscle attachments for the palps, first legs, and possibly the second legs.) Abdominal markings varied but include a pale longitudinal stripe along each side that is broken approx. half way toward the back, followed by a more posterior pale spot that extends slightly dorsally (Figs 79,99,103). When the embolus is long (e.g., Fig. 93), it arises simply as an extension of the tegulum, not being clearly divided from the tegulum as in Piranthus and Maripanthus.
This distinctive group may have many dozens or even hundreds of species, judging from the rate of discovery of new species among the few specimens being collected. We show in Figs 129-142 some of the apparently-new species that we are not naming. Matching males and females with such sparse collecting can be difficult. We have co-collected both males and females of I. kahariana, and they fall together on the molecular phylogeny (Fig. 3). Among the new species, only two are represented by both sexes, I. redunca and I. vimedaba. We interpret the male I. bamilin, I. kodagura, and I. sabivia as not being matches for the female I. nesinor because they are distinct on the molecular phylogeny or have mismatching genitalia (length of embolus vs. copulatory ducts; elaboration of RTA vs. ECP).
Embolus length varies through Indopadilla, short in many species, in other species (e.g., I. kodagura, I. suhartoi) as long as in some Piranthus and Marapathus. While we might be tempted to split the group into two -Indopadilla with a long embolus in South Asia and southeast Asia, and a new genus with a short embolus in southeast Asia -the most complete molecular data nests the long-embolus I. kodagura among short embolus species (Fig. 1). Also, even if the long and short embolus species formed mutually monophyletic groups, dividing them would break one very distinctive and easily recognized clade (cheliceral ridge, thoracic bulges, wide arthrodial membrane on face, etc.) into two much more subtle groups. We therefore maintain the whole group as a single distinctive genus.
The peculiar carapace bulge and exposed arthrodial membrane on the clypeus hint to the possibility that Indopadilla may use unusual biomechanics. The third leg claw tufts appear noticeably larger than the others. Indopadilla are excellent jumpers, difficult to collect even on a beating sheet, from which they can escape in a single decisive bound. They are usually collected from foliage. A video of living males of I. kodagura and the undescribed species "BVBTN" (Figs 134-136) is available in Maddison (2020). Etymology. An arbitrary combination of letters, ungendered. Other names: In Maddison (2015b) and WPM's lab notebooks the informal code for this species was "BVBML".

Indopadilla bamilin
Diagnosis. Both embolus and RTA short and simple (Fig. 76). Other species with such a short embolus have a distinct TmA (I. annamita, I. kahariana, I. redunca, I. sabivia), which seems lacking in I. bamilin. First legs and body dark. None of the known females is an obvious match for this male in terms of markings or molecular phylogenetic placement, but also the unmatched females have longer copulatory ducts than would be expected for such a short embolus, or have a more dramatic ECP than expected for the simple RTA.
Description. Male (based on holotype, specimen SWK12-1618). Carapace length 2.3; abdomen length 3.0. Carapace dark brown, slightly paler around fovea, with a few patches of pale scales on thorax (Figs 78,79). Clypeus dark, narrow; status of arthrodial membrane unclear, as chelicerae are somewhat sunken into the prosoma. Chelicerae vertical, brown, concave in front and with lateral ridge. Retromarginal teeth at least three, on long ridge. Palp femur dark; more distal segments pale. Embo- Etymology. In the Kodava language, kodagura means from Kodagu. Other names: In WPM's field or lab notebooks the informal code for this species was "BVHVW".
Diagnosis. Very similar to I. insularis, contrastingly marked in dark brown and yellow, with the face appearing white because the clypeus is withdrawn toward the eyes to expose a bright white arthrodial membrane. Like I. insularis, I. darjeeling, I. sonsorol, I. suhartoi, and I. thorelli in having a long thin embolus, but even longer than in those species, arising from the retrolateral basal corner of the bulb.
Description. Male (based on holotype, specimen NCBS-BN351). Carapace length 2.6; abdomen length 4.0. Carapace integument dark brown to black except to either side of fovea and narrow medial stripe on thorax, and with small patches of yellow scales in pattern typical for Indopadilla. Clypeus dark, extremely narrow, exposing white arthrodial membrane. Chelicerae dark, with lateral ridge bearing a tooth (Fig.  98, triangle). At least five retromarginal teeth, on long ridge. Palp yellow except for black cymbium. Embolus simple, long (Fig. 93); RTA a flat blade. Legs strongly contrasting between dark brown to black first legs, and yellowish posterior legs. Darkness of first legs relieved by pale tarsus, and honey coloured path on patella. Abdomen black except for two prominent yellow-white patches along each side.
Natural history. A video of the living holotype is available in Maddison (2020).  Etymology. An arbitrary combination of letters, ungendered. Other names: In WPM's lab notebooks the informal code for this species was "BVNES".
Diagnosis. As in I. vimedaba, the face appears white because of the exposed arthrodial membrane, whose boundary with the very narrow clypeus is indistinct. Differs from I. vimedaba in having chevroned abdominal markings visible in alcohol and epigynal openings small and copulatory ducts densely tangled and fused (Fig. 101).
Description. Female (based on holotype, specimen MRB076). Carapace length 2.5; abdomen length 3.6. Carapace integument in alcohol dark brown, paler in ocular area and along midline of thorax, with patches of yellow scales (Fig. 102). Clypeus extremely narrow, but broad arthrodial membrane is exposed and bright white. Chelicerae brown, with concave front and lateral ridge. At least four retromarginal teeth. Legs pale except first, whose femur is brown darkening to black ventrally, and whose tibia and metatarsus are red-brown in the middle. Abdomen dark except for large patches of yellow scales along each side. Epigyne with shallow ECP and small openings (Fig. 100). Etymology. Latin, meaning bent backward, referring to both the RTA and the ridge in front of the epigynal openings. Other names: In Maddison (2015b) and WPM's field or lab notebooks the informal code for this species was "BVMTT". In J. Koh's notebooks it was referred to as "Bent RTA".

Indopadilla redunca
Notes. The living paratype male 12.01.22.0024 is shown as figure E on p. 203 of Koh and Bay 2019.
Diagnosis. Male palp distinctive for bent RTA and thick curled embolus accompanied by TmA (Fig. 104). Epigyne similar to that of I. redynis, with cavernous ECP, but with each opening framed by an anterior curved ridge. Posterior third of abdomen dark except for prominent pale transverse band just in front of anal tubercle. First femur pale basally, lacking dark spot of I. redynis.
Description. Male (based on paratype, specimen SWK12-M0009). Carapace length 2.9; abdomen length 3.6. Carapace in alcohol dark brown except medium brown near fovea, and a small medium brown stripe medially on thoracic slope. Clypeus dark, glabrous, very narrow in middle, exposing a white arthrodial membrane beneath. Chelicerae dark, concave and with lateral ridge. At least four retromarginal teeth, on long sharp ridge. Palp pale except dark gray retrolateral face of femur. Embolus thick and curved, accompanied by similarly curved TmA (Figs 104, 106). RTA projects out away from axis of palp, curved toward the proximal (Figs 105 106). Legs strongly contrasting between the first legs (dark, except for patella, tarsus, and much of femur). Abdomen as described under Male-female matching.
Female (based on holotype, specimen IDWM.20011). Carapace length 3.2; abdomen length 4.1. Carapace dark brown, paler around fovea and in a narrow medial band on thoracic slope. Clypeus narrow, dark, glabrous, exposing white arthrodial membrane. Chelicerae vertical, concave in front. Five retromarginal teeth. Legs pale except for first leg, whose tibia and metatarsus are dark brown in the middle, and the femur which grades to dark brown terminally. Abdomen marked as described under Male-female matching. Epigyne (Figs 107, 108) with large ECP medially. Openings behind arcing ridges.
Male-female matching. The male and female, both collected at Mulu National Park, are matched primarily on the basis of markings and expected genitalic correlations. They share abdominal markings (Figs 110-112): just in front of the anal tubercle is a prominent pale transverse mark, more prominent than in other species, and in front of the pale mark is a dark more or less unmarked area that extends to cover the posterior third of the abdomen. Just in front of that, in the middle third, are two uneven narrow broken longitudinal bands, besides which are similar uneven pale bands on the side. While other Indopadilla species have uneven longitudinal bands, no others known have the bands so restricted by a dark posterior third, nor have the pale pre-anal band so prominent. In genitalia, the large ECP is expected to correspond to a dramatic RTA, which the matched male has. A female was chosen as holotype as it is in best condition, and the molecular data are from a matching female. Etymology. An arbitrary combination of letters, ungendered. Other names: In WPM's lab notebooks the informal code for this species was "BVMT2"; it was also sometimes confused with I. redunca and labelled "BVMTT".

Indopadilla redynis
Diagnosis. Similar in colour and epigyne to I. redunca. The epigyne differs in having the anterior part of the cavernous ECP with a more sharply defined boundary, the openings not behind a curved ridge, and the edge of the opening clearly sinuate (Fig. 113). The abdominal dorsum has fine reticulate markings throughout, not showing the unmarked posterior third. First leg markings match those of I. redunca except that anterior base of first femur with dark patch. Pale area on thoracic slope much broader than in I. redunca.
Description. Female (based on holotype, specimen IDWM.20012). Carapace length 3.6; abdomen length 5.7. Carapace integument medium to dark red-brown, except upper part of thorax between and beside eyes, continuing as a broad medial pale area to pedicel. Clypeus dark, extremely narrow, exposing broad white arthrodial membrane. Chelicerae light brown, concave in front. Five teeth on retromargin. Legs pale except dark brown on middle of first tibia and metatarsus, and darker patch at front base of first femur. Abdomen red-brown, with delicate reticulate pale markings. Epigyne (Fig. 113)  Etymology. An arbitrary combination of letters, ungendered. Other names: In WPM's lab notebooks the informal code for this species was "BVSAB".
Diagnosis. Palp similar to I. kahariana with a short sharp embolus with a TmA behind it, but the TmA is broader and more retrolaterally placed than in I. kahariana. RTA with sharp point, unlike the broad flat RTA of I. kahariana. These two species are sister groups on the molecular phylogeny (Fig. 1).
Description. Male (based on holotype, specimen d107). Carapace length 2.8; abdomen length 3.8. Overall appearance similar to that of I. kahariana, paler coloured than many male Indopadilla, honey to medium brown. Carapace medium brown to honey coloured, darker on sides, with ocular area and larger portion of dorsal thorax quite pale. Clypeus narrow, exposing white arthrodial membrane. Chelicerae brown, with lateral ridge. Six retromarginal teeth, on long ridge. Palp pale except basal part of femur. Embolus short and sharp, with broad TmA behind (Fig. 117). Legs mostly honey coloured, first somewhat darker and with ventral fringe on patella through metatarsus. Abdomen in alcohol similar to that of female I. kahariana, with series of darker paired spots dorsally and elongated pale spots along the side. Etymology. An arbitrary combination of letters, ungendered. Other names: In Maddison (2015b) and WPM's field or lab notebooks the informal code for this species was "BVMDB".

Indopadilla vimedaba
Notes. As noted under Padillothorax, this species was mistaken for Padillothorax semiostrinus, and illustrated by Prószyński (1987) under that name. He illustrates a male and female in the same vial 15151 in the MNHN Paris from the "Malayana" peninsula.
Diagnosis. Palp with embolus longer that half the length of the bulb, straight and tapering (Fig. 121). Other Indopadilla have the embolus either longer and curved, or shorter. Epigyne with small but distinct ECP on back margin medially; openings simple and posterior (Fig. 123). Body dark brown except for highly contrasting but narrow pale cream markings, reticulate on abdomen.
Description. Male (based on holotype, specimen JK 13.09.03.0011). Carapace length 2.4; abdomen length 3.5. Carapace in alcohol dark brown to black, slightly paler around fovea and in narrow medial band along thoracic slope, which also has some white scales; other patches of white scales along borders of ocular area. Clypeus extremely narrow at centre, and beneath it is a broad expanse of white arthrodial membrane; precise boundary between the clypeus and arthrodial membrane indistinct. Chelicerae with strong lateral ridge that extends into a flange near the fang, as drawn by Prószyński (1987: 105). Three promarginal teeth and at least five retromarginal. Palp honey coloured except darker base of femur. Embolus a narrow long triangle (Fig. 121). Legs honey coloured except much larger first leg, which bears a ventral fringe. Abdomen dark above and below, with two pair of lateral pale spots, the anterior of which connects with a few thin pale lines.
Female (based on specimen SWK12-3620). Carapace length 3.3; abdomen length 5.4. Carapace integument in alcohol medium red-brown except for narrow medial pale band along thoracic slope. Clypeus extremely narrow, exposing broad white arthrodial membrane (Fig. 128). Chelicerae concave in front with lateral ridge. At least four retromarginal teeth. Legs pale except first leg, dark except tarsus. Abdomen brown with distinct reticulation of pale scales. Epigyne (Fig. 123) with simple openings, from which broad copulatory ducts proceed anteriorly before narrowing considerably and proceeding to the posterior, then back anteriorly to the spermathecae.
Male-female matching. The male and female described above were not co-collected, but they match well the male and female (MNHN 15151, photographs examined) described by Prószyński, which were in the same vial, and thus likely co-collected. These males and females match in markings: the first leg is all dark except the tarsus; the abdominal dorsum is dark with some narrow pale lines, the venter dark; the pale medial band of the integument of the thoracic slope is narrow; there is a narrow band of pale scales on the midline low on the thoracic slope; the triangle of scales near the fovea is narrow. The male has fewer lines in its abdominal markings than the female, but those it has are precise matches to the female. Other Indopadilla differ; e.g., another candidate female, I. nesinor, has the first leg considerably more strongly banded. suggesting her male should have a first patella paler than seen in Fig. 125, more like that of I redunca (Koh and Bay 2019:. 203, fig. E), whose female's leg is banded similarly to I. nesinor. Also, I. nesinor differs in having a pale underside of the abdomen, and different thoracic markings. The relatively long but wide embolus of the male is a good fit to the openings and first broad part of the duct of the matched female, being ~ 0.07 mm wide at its base and ~ 0.3 mm long. In contrast, the openings of I. nesinor are only ~ 0.02 mm wide, enough to accommodate only a small portion of the embolus.

Stagetillus opaciceps Simon, 1885
Stagetillus irri Maddison, sp. nov. Strand, 1911 is based on a juvenile (SMF 2511 in SMF, examined) and not easily placed. Its being a baviine is suggested by its elongate body (most like that of Indopadilla or Stagetillus), plurident chelicerae, and Asian origin. Its macrosetae sockets resemble those of baviines (Figs 21-23). It lacks the thoracic bulge and ridged chelicerae of Indopadilla, and the flat carapace of Padillothorax. Like Stagetillus, the fovea is displaced slightly to the posterior. The two dark bands running through the PLE shown in Strand's (1911) original figure (no longer visible in specimen) resemble those of male S. opaciceps. The carapace is widest posteriorly, though not so clearly as in Figs 15 and 16. Although its synonymy with Stagetillus is not certain, by placing it with Stagetillus, a salticid genus otherwise incertae sedis is provisionally settled.

Notes. Hyctiota banda
Diagnosis. Carapace distinctive in shape, widest point toward the posterior, approx. half way between the back eyes and the pedicel, and in colour, orange or yellow with the white digestive diverticulum showing beneath the transparent ocular quadrangle. Palp much like that of Bavia, with short blade-like embolus. Simon, 1885 Figs 8, 15, 16, 23, 31, 143-150 Stagetillus opaciceps Simon, 1885: 32. Note. Prószyński's (1987) illustrations of Simon's type specimen characterize the male well. Two additional males illustrated here (Figs 143-147) are conspecific or at least very close (the male from Brunei, Figs 146, 147, has a slightly narrower embolus). Females of the species have not been reported, although we now tentatively identify, as the female of S. opaciceps, the specimen MRB079 from Ulu Gombak used by Maddison et al. (2014) for molecular phylogeny (Figs 148-150). That the female is a Stagetillus is suggested by its similarities to the male in size and body form, carapace shape, yellow coloration, and visibility of the ocular digestive diverticulum. The modest ECP and epigynal openings are what might be expected from the small simple RTA and embolus of the known male of S. opaciceps. Given that Ulu Gombak is only 70 km from a known male of S. opaciceps (Figs 143, 144), the match of the Ulu Gombak female to S. opaciceps is credible. We therefore label the female, used here for molecular data, as S. cf. opaciceps (Figs  148-150). In WPM's lab notebooks the informal code for the female was "STULG". Etymology. From the acronym for the type locality. Other names: This species was referred to by Maddison and Hedin 2003 as "unident. (Phil.)" In WPM's lab notebooks the informal code for this species was "STPHL".

Stagetillus opaciceps
Notes. The FSCA holds more material of this species, including from Mt. Makiling, near the type locality but at higher elevation (G. B. Edwards, pers. comm.).
Diagnosis. Similar in overall appearance to S. opaciceps, from which it differs in palp (narrower and longer embolus) and epigyne (shorter openings and having a central mound, presumably bearing the ECP). Male lacking the dense fringe beneath the first leg (Figs 155 vs. 146), but with a spur on the paturon (Fig. 155, arrow) that is lacking in S. opaciceps (Fig. 146). Male endite lacks the lateral bulge seen in S. opaciceps.
Description. Male (based on holotype, specimen IDWM.20023). Carapace length 3.3; abdomen length 4.6. Carapace yellow-orange with two darker stripes passing along PME, PLE, and to posterior margin, with transparent ocular areas showing bright white digestive diverticular beneath. Shape as in S. opaciceps (Fig 150). Clypeus narrow, more or less glabrous. Chelicerae orange, projecting slightly, with a spur on each paturon just above the fang base (arrow, Fig. 155). Palp pale except darker cymbium. Embolus arising at ~ 9:00 on the bulb, narrowing to a fine tip, with transparent flange retrolaterally (Fig. 151). Long lobe of tegulum overhangs the tibia. RTA broad and obtuse. Legs pale yellow to dark orange, first pair darkest, with femora of first two pair darker below. First leg long, with only a weak fringe beneath. Abdomen with dark central band with three wider areas, flanked by bright white.
Female (based on paratype, specimen IDWM.20014). Carapace length 2.9; abdomen length 5.2. Entirely light in colour, from white to medium orange, except for the black of eyes, and small black patch at the front distal tip of the first leg femur. Carapace pale orange, with transparent ocular area showing bright white digestive diverticular beneath. Shape as in S. opaciceps (Fig 157). Chelicerae orange, with five retromarginal teeth. Legs pale to dark honey-orange, except for black on prolateral tip of femur I. Abdominal integument transparent in alcohol, showing white digestive glands to either side of beige heart and mottled stercoral pocket. Epigyne (Fig. 153) similar to those of Bavia, with sclerotized openings on either side of a central mound.

Genus Padillothorax
Padillothorax Simon, 1901Padillothorax Simon, 1901. Type species Padillothorax semiostrinus Simon, 1901. Bavirecta Kanesharatnam & Benjamin, 2018 Simon, 1901 Padillothorax taprobanicus Simon, 1902 Notes. The synonymy of Bavirecta with Padillothorax can be established now that the identity of the type species of the latter has been clarified (see below). The synapomorphies uniting them include the position of macrosetae on the first femur, flattened carapace, placement of fovea, pale thoracic "window" (Kanesharatnam and Benjamin 2018), and palp form. The deep genetic divide seen in Fig. 3 between P. flavopunctus and the P. badut species group (P. badut and P. mulu) might permit us to divide this into more than one genus, but this is unnecessary and not yet justified. The known Padillothorax species together make a recognizable genus, holding together well through their synapomorphies. In addition, there is no known data guiding us as to how to divide them and maintain monophyly, as we see no strong evidence resolving the trichotomy (P. semiostrinus, P. flavopunctus, P. badut + P.mulu).
Diagnosis. Distinctive for the macroseta(e) in the middle of the front surface of the first leg femur, the palp with narrow distally-pointing embolus, the pale trapezoidal "window" dorsally on the thorax (Kanesharatnam and Benjamin 2018), and the flat carapace with fovea well posterior to the eyes. The first femur's prolateral macrosetae, either one or two, are present in both males and females (Figs 32, 33). While salticids usually have macrosetae dorsally or on the front surface distally (Figs 29-31, 34, 35), the more central location (more ventral and more basal, Figs 32, 33) of Padillothorax is unusual, known from only a few other genera (Epocilla, Padilla, some marpissines). The thoracic window is also different from the pale area on the top of the thorax seen in other baviines (e.g.,

Bavia nessagyna in alcohol)
, where it is neither as pale nor does it extend so far posteriorly as in Padillothorax. Stagetillus spp., and Maripanthus draconis have a pale thorax, but it is not framed by dark, as most of the thorax is pale. Palp is also distinctively simple: embolus narrow, pointing distally (i.e., appears erect in standard ventral view), fused to the tegulum, with no tegular fold covering its base as in Bavia and many Indopadilla. Epigynal openings anterior with simple copulatory ducts proceeding posteriorly. Abdomen with distinct transverse bands. First legs distinctly the longest. Male endite with lobe or sharp corner, unlike Indopadilla (whose endite is rounded, without lobes or corners). Simon, 1901 Figs 9, 17, 24, 32, 158-167 Padillothorax semiostrinus Simon, 1901: 71. Notes. There has been confusion about the identity of P. semiostrinus. Prószyński (1987 p. 105) illustrated as P. semiostrinus a male and female from Simon's collection that clearly belong to the group here called Indopadilla, as Prószyński's illustrations show the group's typical diagnostic features. If his illustrations had shown P. semiostrinus, then Indopadilla would have been junior synonym of Padillothorax. However, Prószyński's illustrations are misidentified, representing a species quite distinct from P. semiostrinus, and which we describe above as Indopadilla vimedaba. Simon's (1901a, b) descriptions are sufficient to show the distinctions. He notes the thorax of P. semiostrinus almost twice as long as the ocular quadrangle (approx. equal in Prószyński's drawings and in Indopadilla generally), the carapace flatter than Bavia (as high or higher in Indopadilla), the fovea well back of the posterior eyes (immediately behind in Indopadilla), the retromarginal cheliceral teeth on a conical elevation (Indopadilla without elevation), and the male endite with a corner (well rounded in male Indopadilla). The last two distinctions are clear also in Simon's illustration (1901b, p. 461), which shows the teeth on a mound and the lateral margin of the chelicera simple, in contrast to Prószyński's illustration which shows the teeth not on a mound but spread along a ridge, and the lateral margin of the chelicera with an extended ridge (as in Indopadilla). Even Simon's name, Padillothorax, emphasizes the similarity to Padilla, whose distinctive carapace (flat, long, with short ocular quadrangle) is quite unlike the more standard carapace shown in Prószyński's drawings and Indopadilla in general. Thus, Prószyński's drawings are misidentified. He was likely misled by the vial's label, which appears much like those of Simon's other types.

Padillothorax semiostrinus
As to what is Padillothorax semiostrinus, we have not been able to examine the type specimens, as they have not yet been located in the MNHN Paris. However, specimens found recently in Singapore and Taiwan match well Simon's (1901a) description, which we translate here to English (with the assistance of Charmaine Gorrie and Anna Bazzicalupo): "♂. Length 7.5 mm. Cephalothorax low, long and oval, red-brown, darker towards the border; texture very wormy-coriaceous except for the middle of the thoracic part which is smoother. Cephalic area in front and at both sides, [and?] near the eyes, decorated with white-silver hairs. Two wide medial thoracic bands, nearly contiguous; a thin marginal line decorated with white-silver hairs. Few white hairs around the eyes. Clypeus very narrow, bald. Abdomen narrow and very long, decorated above with dark violet, a medial band that is wide, entire, and yellow-brick red, bordered in front with lines and behind with a series of spots covered with silver-white hairs, marked on each side with a straight line in front and two white oblique [or crosswise?] lines behind. Venter reddish-yellow. Spinnerets dark. Chelicerae shiny black, short and diverging, convex outside, inside somewhat ribbed, inferior margin having a sunken furrow, then very raised and armed with a series of contiguous teeth, the middle larger. Mouth area black. Laminae truncate at the tip, convex, but with a compressed corner that is slightly extended. Sternum yellow. First pair of legs much longer and thicker than the others, femur clavate, tibia long and ovate, dark brown, coxa and femur black, tarsus yellow, tibia and metatarsus with fringe of reasonably long but not very dense black hairs. Remaining legs pale yellow, armed by a few small spines, as in Bavia. Palps reasonably small, pale yellow, thick with white hairs; tibia and patella rather short, equipped at the outside with a [long?] apophysis with a straight front and a black and acute tip." There are two apparent or possible conflicts between this description and the specimen seen in Figs 158-164. The most serious conflict is that the RTA in Fig. 158 is short, but Simon's description suggests the RTA is long. However, we were unable to understand the apparent use of genitive "apophysi" in "tibia patella breviore, extus ad apicem apophysi sat longa, antice directa, apice nigra et acuta, instructa". What exactly is long: the tip of the apophysis, the apophysis, or the outer edge of the tibia to the tip of the apophysis? Second, his description refers to marks at the back of the abdomen that are "obliquus", which would differ from Figs 164, 166 if it were translated as "oblique", but match if translated as "crosswise". Otherwise Simon's description and his mouthparts illustration are an excellent match to the Singapore specimen, including the distinctive pair of nearly-touching wide thoracic bands, the pattern of the abdomen, the thorax rugosity except in the middle, the shape of the carapace and position of the fovea, the shininess and shape of the chelicerae, the mound bearing the retromarginal cheliceral teeth, the shape of the endite, the colours of the legs, and the yellow sternum.
We therefore provisionally identify the specimen of Figs 9, 17, 24, 32, 158-164 as P. semiostrinus. Although we might have added "cf." to its label to indicate our uncertainty ("P. cf. semiostrinus"), we avoid this so as to propose a stable concept of the species that could endure if Simon's types are never found.
A juvenile found in Singapore (Fig. 165) suggests the likely appearance of females. Males and females of this or a very similar species have been found in Taiwan, not yet examined, but with photographs posted in the website Facebook (Figs 166, 167).
Natural history. The male in Singapore was found in the open on a simpoh air leaf (Dillenia suffruticosa). A video of the living juvenile (specimen AS19.2448) is available in Maddison (2020).
Notes. This species shares the diagnostic features of the genus, including macrosetae on the middle of the front face of the first femur. P. flavopunctus has two such macrosetae on each femur, as in P. badut and P. mulu, but larger and placed even more proximally. We have concerns that the paratype female (IFS_SAL_679) is not conspecific with the male holotype, but if not conspecific, it is a closely related species, both by its strong morphological similarity (body form, markings) to the male and by its molecular proximity to a juvenile (IFS_SAL_1017) collected alongside the male at the type locality. Etymology. From the Malay word badut, meaning clown. In the field we called these (and P. mulu) the "banded clowns". Other names: In Maddison (2015b) and WPM's field or lab notebooks this species was grouped with P. mulu under the informal name "BVBND", until it was distinguished from that species as "BVBND-L".

Padillothorax badut
Diagnosis. Very similar to P. mulu, differing most notably in details of genitalia. The embolus of P. badut lacks the prolateral basal ridges and has a longer terminal part; the epigyne has the openings hidden beneath a fold.
Notes. This and the other new Malaysian species (P. mulu) are very similar in appearance, thin and banded; together we consider them as the P. badut species group. When the first legs are held forward in life (e.g., Fig. 174), the spider appears as a series of transverse white bands approximately evenly spaced from anterior to posterior: the first leg annuli, the two transverse bands on the carapace, and the four transverse bands on the abdomen. Their bodies are more delicate and parallel-sided than other Padillothorax, and they have more complex emboli.  (holotype, specimen IDWM.20007). Carapace length 2.3; abdomen length 3.1. Carapace (Fig. 172): Black to dark brown except for band just back of the anterior eye row, and the pale trapezoidal window on thorax. Unusually flat (carapace height 0.72), with ocular area and most of the thorax on a plane, declining sudden-ly only near back of thorax. Fovea well behind the PLEs (by more than their diameter). Clypeus very narrow, dark. Chelicerae vertical, glabrous, dark. Teeth not examined in holotype, but another male from Lambir Hills has five retromarginal teeth, together in a mound as in P. semiostrinus. Palp pale except for base of femur. Base of embolus with two peaked projections on the retrolateral side (Fig. 168). Endite with lateral lobe (Fig. 10). Legs: First legs very distinctly longest and darkest, red-brown to black except for strong annulae with white scales. Legs II-IV pale yellow with just a smudge of dark laterally on a few segments of the fourth leg. Abdomen: narrow, with four white transverse bands.

Description. Male
Female (paratype, specimen IDWM.20008). Carapace length 2.1; abdomen length 3.1. Colour and structure matches that of male in nearly all aspects, with the most distinct difference being the slightly shorter first legs. Cheliceral teeth not examined in this specimen, but another female from Lambir Hills has four retromarginal teeth, together in a mound. Epigyne (Fig. 170) with two anterior openings, hidden under folds, and simple copulatory ducts visible without clearing. The copulatory duct has a diverticulum for the accessory gland extending laterally just before entering the simple spermatheca.
Natural history. On large-leaved understory plants such as palms. In life, they often hold the front legs out or to the front.
Additional material examined. All from Malaysia: Sarawak: Lambir Hills Nat. Pk.,  Etymology. From the name of the type locality. Other names: In Maddison (2015b) and WPM's field or lab notebooks this species was grouped with P. badut under the informal name "BVBND", until it was distinguished from that species as "BVBND-M".
Diagnosis. Very similar to P. badut, differing most notably in details of genitalia. The embolus of P. mulu is shorter, and has a toothed ridge on the prolateral base; the epigyne has the openings exposed on a more or less flat surface.
Description. Male (holotype, specimen IDWM.20009). Carapace length 2.5; abdomen length 3.5. Colour and structure matches that of P. badut as described above. Cheliceral teeth not examined. Palp pale except for base of femur. Base of embolus with various fine teeth; embolus shorter than in P. badut (Fig. 176).
Female (paratype, specimen SWK12-EP0105). Carapace length 2.1; abdomen length 3.3. Colour and structure matches that of P. badut as described above. Cheliceral teeth not examined in this specimen, but another female from Mulu has five retromarginal teeth. Epigyne (Fig. 178) similar to that of P. badut, except for the openings being exposed, not under a fold.
Natural history. As for P. badut, on large-leaved understory plants. They often hold the front legs out or to the front.
Additional material examined. All from Malaysia: Sarawak: Mulu Nat. Pk., in UBCZ. One female (specimen SWK12-EP0107) from Base  (Reimoser, 1929), comb. nov., transferred from Bavia Etymology. An arbitrary combination of letters, reminiscent of Marpissa (as the females resemble) and Piranthus (to which it is closely related). To be treated grammatically as masculine.
Diagnosis. Epigynal atria long and gaping, anteriorly placed. Embolus long and beginning on the basal side of the tegulum, apparently freely articulated from the tegulum (as in the related Piranthus). Retromarginal cheliceral teeth close together, forming a single short ridge. Male endite with sharp corner (Fig. 11). Abdomen with central longitudinal pale band flanked by dark and stuttered into a chevron. Body unremark-able for the Salticinae, in contrast to its sister genus Piranthus which are unusually flat and robust. The small M. menghaiensis and M. reinholdae in some respects resemble Indopadilla or small Bavia, but they lack the ridged chelicera l paturon of the former and the short embolus of the latter. Etymology.

Maripanthus draconis
Greek, δράκων, referring to the fiery colours of the male. Other names: In Maddison (2015b) and WPM's field or lab notebooks the informal code for this species was "CFMAR". A specimen that is likely a closely related but distinct species is shown in Koh and Bay 2019 as "Bavia" sp. B Black-collared long-bellied jumping spider.
Diagnosis. Most similar to M. smedleyi, of which only the female is known; M. draconis differs in having longer epigynal atria (greater than half of epigynal length) and less distinct atrial cliff ("ac",Fig. 191). Similar to M. jubatus in being large bodied, but M. draconis differs in having a honey-coloured translucent thorax, which in the female is speckled with black, and which in the male contrasts strongly against the black ocular area. Male front is impressively red-orange, with red-orange first femora, palps, and rings around eyes, and a golden face. Embolus slightly shorter than the other three species for which males are known. Ventral bump on palp femur.
Description. Male (based on holotype, specimen AS19.2232). Carapace length 4.0; abdomen length 5.0. Carapace (Figs 196,197) black in ocular area and in a thin line along ventral and posterior margin; thorax strongly contrasting honey coloured, glabrous except for a few vertical black setae. Clypeus covered with yellow-cream setae. Chelicerae vertical, a soft orange-red, covered with yellow-cream setae. Three promarginal and five retromarginal teeth. Palp femur pale, darkening to red-brown distally. Bulb elongate (Figs 188, 189); embolus arising on the prolateral basal corner and proceeding distally. Femur with small but distinct distal ventral bump. Endite with small corner projection. Legs with pale femora and darker markings more distally. First femur striking orange-red, patella to tarsus dark brown to black. First patella through metatarsus with sparse ventral fringe of setae that are black at base, white at tip. Abdomen with brown longitudinal medial band dorsally, flanked by scattered red scales, with some scattered white scales basally.
Female (paratype, specimen AS19.2250). Carapace length 4.2; abdomen length 6.1. Carapace (Fig. 199) integument black in ocular area, except for a pale spot on each side just medial to the PME (in other females, these pale spots form a transverse band that intersects with a longitudinal medial pale band, forming a pale cross in the ocular area). Thorax honey-coloured with distinct black speckles. Clypeus with scattered white setae. Chelicerae brown with darker patches, with scattered white setae. Etymology. Latin, meaning maned or crested, referring to the field of short black setae on the male ocular area. Other names: In WPM's field or lab notebooks the informal code for this species was "CFMA2". Description. Male (based on holotype, NCBS-BN352). Carapace length 3.9; abdomen length 4.8. Carapace  black, with scattered white scales on the top of the thorax. Front half of ocular area covered with vertical black setae, giving the appearance of a mane from in front (Fig. 210). Clypeus black, with long black setae. Chelicerae vertical, black. At least five retromarginal teeth. Palp black, but cymbium with white setae. Embolus long and swooping , beginning with a broad base over the tibia, proceeding prolaterally, then dorsally, then distally. RTA long and more or less straight. Femur lacks the ventral bump seen in M. draconis. Endite with small corner projection. Legs similar to those of M. draconis, with some segments of first legs dark, but differs in having the first femur pale honey darkening terminally to brown, and tarsus pale. Abdomen light brown mid-dorsally, darker laterally and with scattered cream coloured scales.
Female (based on paratype, NCBS-BN353). Carapace length 4.2; abdomen length 5.1. Carapace black except medium brown areas (yellow in alcohol) around fovea and along margin, covered with yellow cream scales in band along ventral margin, and dorsally on ocular area and anterior part of thorax . Clypeus covered with long cream-coloured setae. Chelicerae dark, covered with long cream setae. Three promarginal and five retromarginal teeth. Legs pale honey-coloured with various darker patches. Abdomen like that of male, but paler (Fig. 213). Epigyne (Fig. 205)  Etymology. Named in honour of Christa Deeleman-Reinhold, whose extensive work on southeast Asian spiders has greatly increased our knowledge of the area's fauna. She has discovered and described over 350 new species, including 54 new salticids. Other names: In Maddison (2015b) and WPM's field or lab notebooks the informal code for this species was "BVBGB". This species is shown in Koh and Bay 2019 as "Bavia" sp. C Two-lined long-bellied jumping spider; the upper photograph on p. 208 is of the holotype.
Diagnosis. Very similar to M. menghaiensis, and like it smaller and more Indopadilla-like in body form than M. draconis and M. jubatus (shorter first legs, more elongate abdomen). Differs from M. menghaiensis in the longer epigynal openings, and in details of the palp's bulb. In view from the retrolateral, the embolus is first directed to the distal then quickly turns dorsally (M. menghaiensis, embolus begins toward the dorsal). When the embolus comes out from behind the tegulum it is directed slightly proximally (slightly distally in M. menghaiensis). The embolus is thinner near the tip than in M. menghaiensis.
Description. Male (based on holotype, specimen JK.11.12.24.0006; living holotype shown on p. 208 of Koh and Bay 2019). Carapace length 2.9; abdomen length 4.0. Carapace (Fig. 222) black except for brown around fovea. Cream scales form a band crossing the thorax behind the PLEs, a small streak on the midline of the thoracic slope, and a narrow band along the lateral margins. Clypeus dark but centrally with a cluster of erect white scales that overhang the chelicerae. Chelicerae simple, vertical, dark. Plurident. Palp pale except for dark femur. Embolus long, arising retrolaterally before curling under the bulb prolaterally then proceeding distally (Figs 215-217). A short spur diverges from the embolus before it turns distally (Figs 215,216). RTA vertical except for a bend and curl distally. Endite with small sharp corner, similar but smaller to that in the larger species of Maripanthus. Legs pale honey coloured except for first, which is black in all segments except the tarsus. Third and fourth legs darker near more distal joints. Abdomen thin, dark, with a pale middorsal band, just lateral to which are indistinct longitudinal streaks of white scales.
Female (based on paratype, specimen SWK12-1934). Carapace length 3.7; abdomen length 5.0. Carapace as in male, but slightly paler in integument. Chelicerae with three promarginal and five retromarginal teeth. Legs as in male but with first legs only slightly darker than the others. Abdomen brown with central pale chevroned band, and thin white streaks as in male. Epigyne (Fig. 218)  Diagnosis. Carapace surface rugose, with a coarse reticulate sculpturing throughout. Carapace flat (height well less than half the length), with ocular area and front part of thorax on a plane, and fovea well back of PLE, 1.3-1.5 × further from front of carapace than is the back of the PLE. Legs robust, especially the first pair. Embolus begins at prolateral basal corner of bulb; epigyne with central septum. Tip of abdomen black.
Two of the species (P. bakau and P. kohi) are distinctive for their black-and-white banding, three others (P. decorus, P. mandai, and P. planolancis) are more simply marked with brown and black, while the last (P. api) is a red-orange-black ember.
The four new species described here extend the range of Piranthus eastward as far as Borneo. The two previously described species, P. decorus (Thorell 1895; Caleb and Sanap 2017) and P. planolancis (Malamel et al. 2019;Nafin et al. 2020), are from Myanmar and India. A video of a living female of P. planolancis (specimen AS19.5940) is available in Maddison (2020). Etymology. From the Malay word, "api", meaning "fire", referring to the colour. Other names: In WPM's field or lab notebooks the informal code for this species was "PIORG".

Piranthus api
Diagnosis. A distinctively narrow species with bright red-orange legs. Description. Female (holotype, specimen AS19.3205). Carapace length 3.0; abdomen length 3.5. Carapace narrow and low, orange-red-brown with central black area covering ocular area and medial part of thorax (Figs 229, 233). Clypeus black. Chelicerae short and vertical, dark. Legs red-orange except for black patella and tibia of first legs. Abdomen dark brown above, with median longitudinal paler band that in alcohol appears as to uneven parallel cream bands, extending backward until the black end of the abdomen. Epigyne (Fig. 235) with medial septum and a small medial ECP. Natural history. The two specimens were both found deep within large grass tussocks overhanging a moist ditch. A video of the living holotype is available in Maddison (2020).
Additional material examined. A second female (specimen AS19.3217, in UBCZ), raised in captivity, same data as the holotype. Etymology. Referring to the type locality and to the holotype's habitat, mangroves (Malay, bakau = mangrove). Other names: In Maddison (2015b) and WPM's lab notebooks the informal code for this species was "BKOMG". This species is shown in Koh and Bay 2019 as the male of "Bavia" sp. D Strong-armed flat jumping spider.  Diagnosis. This and the closely similar P. kohi differ from other known Piranthus in having white transverse banding on the body and legs, and the posterior legs striped with black and translucent white. P. bakau (Figs 237-254) differs from P. kohi  in having:

Piranthus bakau
• Second white transverse band on the dorsum of the abdomen (i.e., the first behind the basal band) complete or broken by only a slight space; • Sides of thorax lacking the three distinct narrow vertical lines seen in P. kohi (at most only a hint of two); • Carapace lateral to the PLE with a bare patch, lacking golden scales, extending from PME back to behind PLE, and lateral to the bare patch is a stripe of denser golden scales (Figs 245, 249); • First and second tibia bicoloured (black with white tip, arrow in Fig. 248); • Second femur bicoloured (white basally, black terminally, Fig. 249); • Black dorsal band on the fourth tibia incomplete, beginning mid-segment and reaching to the end (Figs 246, 247).
• Embolus (Fig. 237) notably longer than in P. kohi (Fig. 255), closely resembling that of P. planolancis (Nafin et al. 2020); • Epigyne with cavernous atria framed by a curved medial ridge, and relatively long copulatory ducts leading to a posterior tangle of tubes and spermathecae.
Juveniles have markings consistent with those of adults, and thus can be distinguished by the non-genitalic features above.
Description. Male (based on holotype, SWK12-0561). Carapace length 2.7; abdomen length 3.0. Carapace with rugose surface, black on ocular area, dark brown otherwise, covered thinly with narrow golden to white scales except bare patch lateral to PLE, and on posterior slope. Clypeus narrow, dark, with a few white setae. Chelicerae small and vertical, dark, with a few pale setae. Palp (Figs 237-239) black except band of white scales terminally on femur, and cymbium, which is dark basally but then fades to white distally. Embolus long, arising on prolateral basal corner of bulb, looping first dorsally, then proximally, then distally. RTA a long blade. Endite margin rounded laterally. Legs relatively larger (compared to the body) than in many baviines, with first pair especially robust. First and second legs mostly dark and with terminal white annuli on the segments. First patella and tibia with ventral fringe of black setae. Third and fourth leg segments from patella to tarsus translucent white with black bands and stripes. Fourth tibia and patella with a pattern of short black and white bands. On the tibia, a black dorsal band begins not basally but half way to the end, and extends distally to the tip; basally, it is interrupted by white ( Fig. 246; compare with Fig. 266). Abdomen with three narrow transverse white bands, a basal one, a second one behind that, and a third one behind that. Posterior to the third band the abdomen is a shiny black, but anterior it is dusted with golden setae.
Female (based on specimen AS19.2895). Carapace length 3.0; abdomen length 4.1. The specimen was collected and photographed as a small juvenile (Figs 251-254); as an adult its basic appearance is similar (Figs 249, 250). Structure and markings of carapace and legs as in male. Abdomen with second transverse band oblique. Epigyne (Fig. 240) with medial septum dividing deep broad atria. What serves as the ECP is not obvious; there may be two, folds posterior to the atrium on each side, midway between the epigastric furrow and the posteriormost part of the medial septum. If so, it may be intermediate between P. decorus, with a medial ECP (Caleb and Sanap 2017), and P. planolancis (Malamel et al. 2019;Nafin et al. 2020), with a pair of distantly separated lateral ECPs (Nafin et al. 2020: fig. 17).
Male-female matching. P. bakau and P. kohi are similar in general appearance, have overlapping geographical ranges, and have been collected to date with only adult males or adult females at a locality, not both. This leads to a question of which male matches which female. Unless there are additional closely similar species in the same areas, the inferred matching is well supported by the differences in markings, carapace shape, and lengths of embolus/copulatory ducts. The male of P. bakau and the female from Tengkorak inferred to match it share the diagnostic traits mentioned above. Doubt might arise because of one difference in their markings: the female has the second transverse abdominal band more oblique, with its two sides meeting at a central peak, while in the holotype the band is straight across. However, the second male, from Brunei, shows a peak (Fig. 242). A juvenile co-collected with the male holotype has the band peaked, and looks very much like the female when immature. The female is not designated as a paratype, however, because of the possibility it is a different but very closely related species, given the geographical distance between it and the holotype and their different habitats.
Natural history. Holotype male collected from mangroves; female from Tengkorak collected by shaking vines and understory trees near waterfall.
Additional material examined. Two juveniles with same data as holotype. Also, one male (specimen JK.11.04.17.0040), in LKCNHM, from Brunei: Tutong, Tasek  similar to those of P. planolancis; darkest in the femur, patella, and tibia of the first leg, and the femora of the other legs. Abdomen brown above, lacking the transverse bands of P. bakau and P. kohi, but with the dark end of the abdomen (Fig. 274).

Tribe Viciriini Simon, 1901
Nungia Żabka, 1985Nungia Zabka, 1985. Type species Nungia epigynalis Żabka, 1985. Note. At least 16 species of elongate dull brown salticids collected in field work in Asia and New Guinea were initially thought to be baviines (example photographs, Figs 275-282), but an analysis of two genes (Fig. 283) in nine of the species show that they form a clade with the similar Nungia epigynalis, nested within the Viciriini. Photographs of them were released by Maddison (2015b) under the name Nungia. Their distinction from baviines is also morphological. They have unident or fissident chelicerae and a carapace that is dark, matte (not shiny), distinctly flat-topped, and relatively square. The male first legs are fringed below; the embolus is usually short and fixed to the tegulum. The group is unreported but diverse in New Guinea (based partly on collecting by WPM, Maddison and Zhang 2011; Nungia there mistakenly reported under "Bavia and related genera"), as might be expected for a member of the Astioida. Capeyorkia falls within this group (Fig. 282, 283; NPNGE and NPNGF are listed as Capeyorkia because of similarity of palps), and likely Pungalina as well. Both Pungalina and Capeyorkia may merit synonymy with Nungia. At least some species of Muziris Simon, 1901 belong here as well. Also within the group are two species misplaced in Bavia, one in Diplocanthopoda, and one in Cosmophasis. While we might have delayed their consideration until a future revision of Nungia, we address them here as part of organizing the baviines. We therefore make the following new combinations: Nungia hatamensis (Thorell, 1881), comb. nov., transferred from Diplocanthopoda Abraham, 1925. See Fig. 280. Nungia modesta (Keyserling, 1883), comb. nov., transferred from Bavia. Based on Keyserling's (1883) figures, body form and colouration typical of the New Guinea Nungia. Nungia papakula (Strand, 1911), comb. nov., transferred from Bavia. Male and female syntypes in SMF examined; appears closely related to N. hatamensis and a likely senior synonym of Muziris wiehlei Berland, 1938. See Fig. 281. Nungia xiaolonghaensis (Cao & Li, 2016), comb. nov., transferred from Cosmophasis Simon, 1901. See Fig. 279.
Some of these new combinations, and others made above in baviines, correct mistaken placements that may have resulted from convergence in the general form of the male palp. The palps of Nungia hatamensis, Nungia xiaolonghaensis, Maripanthus menghaiensis, Bavia capistrata, and Bavia maurerae do in fact resemble those of the genera in which they had been placed, respectively, Diplocanthopoda (a hasariine), Cosmophasis (a chrysilline), Nannenus (a nannenine), Evarcha (a plexippine), and Epidelaxia (a nannenine). Those genera are all in the Saltafresia, a group phylogenetically distinct from the astioids and baviines. The resemblance is largely restricted to the palps, as the remainder of the body is quite distinct in each case. Convergent evolution of the basic form of the palp is widespread in salticids, insofar as the palps are simple and variation in a single dimension (embolus length) can generate palps that look superficially quite similar. The general shape of the palp should be used with caution in determining relationships.

Species incertae sedis
The following species are too poorly known to assign to a known baviine genus, or for that matter to confirm their placement in the Baviini: Of these, B. sinoamerica is almost certainly not a baviine by body shape. Based on the form of the palp, and its being unident with a compact body, it may be a hasariine. B. albolineata has a palp that is credibly baviine, but its location in Madagascar and diverging chelicerae suggest it is not baviine.