Main Clades of Araneomorphae:

The landmark study of Platnick et al. (1991) was extremely encouraging at the time, because it obtained quantitative phylogenetic evidence for higher groups such as Araneoclada, Austrochiloidea, Haplogynae, and Entelegynae. Subsequent analyses and new findings of morphological, behavioral, and molecular data revealed that the situation is much more complex than previously thought. The following four cases are crucial for this instability: (1) Austrochilines turned out to have presumably derived characters such as cylindrical gland spigots, both legs moving while combing cribellate silk, and median tracheae (Griswold et al., 2005; Lopardo et al., 2004; Ramírez, 2000); (2) a suite of primitive conditions were found in Filistatidae, such as M-shaped intestine, only leg IV moving while combing, and the presence of posterior book lung leaves in early juveniles (Griswold et al., 2005; Eberhard, 1988; Lopardo and Ramírez, 2007; this study); (3) several members of Palpimanoidea (Forster and Platnick, 1984) turned out to be nested inside Araneoidea (Schütt, 2000, 2002; Griswold et al., 2005; Rix et al., 2008; Blackledge et al., 2009; Lopardo et al., 2011); and, lastly and more surprisingly, (4) the leptonetid Archoleptoneta schusteri, supposedly well nested in the ecribellate Haplogynae, revealed a full-fledged cribellum and calamistrum (Ledford and Griswold, 2010).

Entelegynae and the RTA Clade:

Besides the extensive analyses made on of orb-weavers (see table 2) and the ticking time bomb of Palpimanoidea, the analyses of Griswold et al. (1999) and Griswold et al. (2005) began the cladistic exploration of the main branches of entelegynes other than orbicularians or palpimanoids. Major new hypotheses of those works are the eresoids (Oecobiidae + Eresidae), titanoecoids (Titanoecidae + Phyxelididae), and the Fused Paracribellar clade (a group of families related to Amphinectidae and Desidae). Those were followed shortly thereafter by the molecular analyses of Spagna and Gillespie (2008) and Miller et al. (2010). This last analysis did not recover eresoids, and placed the titanoecoid representatives well within the RTA clade, but obtained an Austral Cribellate clade (Spagna and Gillespie, 2008) fairly similar to the Fused Paracribellar clade but including the Stiphidiidae. Both molecular studies suggest that the RTA clade should include at least the Dictynidae.

Lycosoidea and Relatives:

Griswold (1993) made the first quantitative analysis exploring the relationships of the groups of families allied to wolf spiders, as proposed by Lehtinen (1967) and Homann (1971). The candidates for such a group usually have some striking characters: grate-shaped tapeta, oval calamistrum, tibial cracks on male legs, and a tegulum-subtegulum interlocking mechanism in the male copulatory bulb. The group endured some further analyses based on morphology (Griswold et al., 2005; Silva Davila, 2003), although the internal relationships of lycosoids and relatives (the Oval Calamistrum clade in Griswold et al., 1999, 2005; the Grate-Shaped Tapetum clade in Silva Davila, 2003) are fluctuating in the different analyses (Griswold, 1993; Silva Davila, 2003; Raven and Stumkat, 2005).

The Root of Dionycha:

Recent molecular analyses of the RTA clade (Miller et al., 2010; Spagna and Gillespie, 2008) suggested that the dionychans may be the sister group to the lycosoids, although those studies had only a limited number of representatives of those clades (three and one, respectively). Similarly, Silva Davila (2003) found Dionycha as sister to a Grate-Shaped Tapetum clade; in her analysis, the five representatives of Dionycha were joined by only one character, the precoxal sclerites (see fig. 199A).

Dionychan Relationships:

The first comprehensive study of dionychan relationships is an unpublished dissertation by Penniman (1985), including five families (Clubionidae, Anyphaenidae, Gnaphosidae, Corinnidae and Liocranidae). His analysis was based on a small number of characters, scored for groundplans of families or subfamilies, instead of representative species. Several of Penniman's characters were challenged in the last decade, and his analysis would nowadays be insufficient, but there are some coincidences with more recent studies. The clade corresponding to Dionycha in Silva Davila's (2003) tree is supported by the precoxal sclerites discovered by Penniman. Phrurolithinae was placed in Corinnidae, and Gnaphosidae appeared as the sister group of a clade consisting of Corinninae, Castianeirinae, Trachelinae, and Phrurolithinae, which has some coincidences with the results obtained here. Further cladistic analysis of dionychan spiders have focused in the relationships of Gnaphosoidea, corinnids, and liocranids. The successive studies of Gnaphosoidea (Platnick, 2000, 2002) centered on the amazing Australian diversity, consolidated the relationships and delimitation of families outlined by Platnick (1990), mostly from the morphology of spinnerets and spigots. The analysis by Bosselaers and Jocqué (2002) produced novel or unexpected groupings, of which one was considered sufficiently well supported to make the transfer of Phrurolithinae to the family Corinnidae, close to Trachelinae; the liocranids were somehow dispersed through the tree, and Castianeirinae did not appear related to corinnids. Several representatives from such analysis were further combined with gnaphosoids and additional trachelines in a recent cladistic analysis (Haddad et al., 2009), also led by Jan Bosselaers. As a result, the phrurolitines still held somehow in the vicinity of trachelines, this time near Castianeirinae as well, and the liocranids continued to be polyphyletic, and were also mixed with gnaphosoids.

“Outgroups”:

From the beginning of this study it was clear that Dionycha might not be a monophyletic group, hence the selection of outgroup taxa had to be sufficiently ample to allow for a real test of monophyly (44 out of 166 taxa, or 27%, are outgroups). The dataset includes representatives of the main lineages of the RTA clade, as well as of the main basal clades of Araneomorphae, the outgroups of the RTA clade itself. Previous analyses (Griswold et al., 1999, 2005) explicitly favored the inclusion of cribellate, presumably plesiomorphic representatives of the main RTA clades. Since dionychans are ecribellate, finding the closest relatives and internal rooting of dionychan lineages, the selection of representatives had to include ecribellate members as well. Lycosoids are especially well represented, since several dionychan groups have a grate-shaped tapetum (thomisids), and some were explicitly included among lycosoids (zorids, some miturgids). This ample taxonomic coverage allowed for a detailed review of the morphological homologies and the legacy characters relevant to dionychans. Some representatives of “zodarioids” (Homalonychus, and the zodariids Cyrioctea, Cybaeodamus, Storenomorpha, and Cryptothele) were studied in detail but not included in the final dataset, because although monophyletic, they were extremely unstable in the analyses.

“Ingroup”:

All families putatively members of Dionycha were considered in this analysis, including those that were placed within Lycosoidea. To study the relationships among families of dionychans, more than one or two members were studied for each family whenever possible, trying to sample their internal diversity. This study devoted more effort in three large dionychan families with contentious limits (Liocranidae, Corinnidae, and Miturgidae, 52 representatives in total), at the expense of a less-dense coverage of the gnaphosoid families (28 representatives), which had received focused and detailed attention in recent years (Platnick, 2000, 2002). Whenever possible, the sampling of representatives was inspired by previous phylogenetic analyses and classification in subfamilies.

Geography:

The taxa studied here are a fairly good representation of the global taxonomic breath of the families involved. A few regions were slightly favored because of the availability of good recent collections (such as those of Vietnam, thanks to the work of Diana Silva Dávila, and Argentina), but the sampling is not systematically biased toward a given biogeographical region, continent or hemisphere.

SEM Preparations:

By default, each species resulted in about eight scanning electron microscopy (SEM) preparations (female: left chelicerae, left palp, left legs I and IV, epigyne digested, spinnerets; male: left palp, abdomen). The most critical step for obtaining neat SEM preparations was selecting clean, well-preserved specimens from collections. While still in alcohol, all samples were brushed using a thin painting brush of synthetic fibers, following the direction of the setae. Very dirty samples, especially of hard structures such as the male copulatory bulb were cleaned for a few seconds in an ultrasonic cleaner. Before drying, some hairs had to be removed with fine forceps to expose structures (figs. 45B, 120B, 128F). All samples were critical-point dried after dehydration in ethanol series. Each sample was mounted on a separate aluminum stub, using adhesive carbon tabs (fig. 15A, background) or adhesive copper tape (fig. 11E, bottom right). During the mounting, some further depilation was made with fine forceps, the loose hairs or dirt then removed with brush and a jet of air blowing through a thin pipette connected to a rubber tube. To allow for a better conductivity, once positioned on the adhesive medium, the borders of the pieces were glued to the conductive substrate with colloidal graphite on isopropanol base (fig. 11E). Such conductive paint reduces the charging of the sample under the SEM, and further secures the piece. The preparation identifier was engraved on each stub with a needle, to make it visible in the SEM monitor.

KOH Digestion:

The respiratory system and spermathecae of haplogynes were examined after digestion in a hot 10%–20% KOH solution. The pieces were placed in a double boiler and heated in a hot plate or a Fuyí© heater for antimosquito tablets. After digestion, cleaning, and passing through ethanol, the samples were usually stained by quickly immersing in a saturated solution of chlorazol black in ethanol. The dissections were made as described in Platnick et al. (1999). Smaller samples (spiderlings, minute spiders) were digested inside a glass microvial with a loose cotton stopper, to prevent losing the specimens when they become transparent. In that case, the changes of fluids and staining were made through the cotton, under a stereomicroscope. Manipulation of digested pieces under the stereomicroscope was made with reflected dark field illumination, especially for the delicate pieces; this was done by placing a mirror below the Petri dish used for the samples. The digested pieces were observed in lactic acid with compound microscope on an excavated slide, using a strip cut from a glass cover to retain the sample in position.

Enzyme Digestion:

Spermathecae were cleaned of soft tissues for SEM preparation using enzymatic digestion. The dissected epigyne was placed on a small vial with water and trypsin, and incubated at 40° C overnight. Some samples were digested in pancreatine and borax solution, as in Álvarez et al. (2008).

Clove oil Clarification:

Temporary clarification of soft structures was made with clove oil. As usual, this was applied to examine spermathecae without digesting, but also to expose the path of the spermophor in the male copulatory bulb. Eye tapeta were observed after overnight clarification of the whole specimen or carapace in clove oil; the specimens used for dissection of chelicerae were convenient for clarification, since the oil can quickly penetrate into the cephalic area once a chelicerae is removed.

KOH Expansion of Male Copulatory Bulb:

The palps were dissected and brushed, sometimes sonicated, but not depilated. The expansion was made in a manner similar to Shear's (1967). The palp was placed in a 10% KOH solution for some minutes, up to half an hour, and then cycled between KOH solution and distilled water until fully expanded. Larger or harder palps needed longer cycles; in those cases the heat of a lamp in the distilled water phase was used to help expansion.

Image Medatata:

Image files were named starting with the preparation code (e.g., abbreviated “686aILeft claws female Lyssomanes.tif,” or verbose, “MJR1213e Claws I apical ff Paradiestus.tif”). Using this convention, all the specimen metadata could be easily retrieved from the preparations database. Images and their data were maintained in an IMatch database ( www.photools.com). SEM settings and other device-generated metadata were imported parsing the buddy text files (Hitachi SEM, Leica stereomicroscope) or embedded EXIF metadata (FEI SEM) using custom scripts. All images were recorded with a scale bar. For light microscopy devices that lacked scale-bar functions, the magnification was encoded in the file name between brackets (e.g., “MJR1237_01[Nx3] carapace dors ff Doliomalus.jpg”), and a calibrated scale was then overlaid on the image using a custom script in IMatch. The image metadata elements correspond to those used in Morphbank ( http://morphbank.net) plus some additional fields developed for the ATOL Spiders project (Ramírez et al., 2007), including identifiers for a web-based resource tool, the Spider Ontology (Ramírez, 2011). To allow for a quick orientation, images of spinnerets are labeled in figures with the position marked on a schematic map of the spinnerets (e.g., fig. 119, top left of each image).

Image Repository:

Since the scoring of the dataset depends heavily on high-resolution images (about 30% of the characters in this dataset can be scored only from SEM images), and it is impossible to publish all of them in this paper, the image repository is an important part of the documentation of this study. All the images produced for this study and their corresponding metadata are deposited in Morphbank collection ID 799551 ( http://www.morphbank.net/myCollection/?id=799551), openly accessible in full resolution under Attribution-Noncommercial-Share Alike 3.0 Creative Commons license.

Cells:

Scoring of dataset cells was done from actual examination of specimens or their images. Only in exceptional cases the scoring was taken from the literature, and in that case a comment was inserted indicating the source. A few characters from soft internal anatomy were scored entirely from the literature; in those cases, only primary sources were used. Polymorphisms (i.e., more than one character state assigned to a cell) were used to express variability, intermediacy, or ambiguity in the anatomical observation, and the case was explained in a comment.

Legacy Characters:

The first step for the building of the phylogenetic matrix was reviewing the characters previously used in the literature. Table 2 summarizes the previous cladistic analyses considered as sources of legacy characters relevant for the placement and internal resolution of Dionycha in Araneomorphae. Legacy characters were sorted by body region in a preliminary schema, which ended up as the skeleton of the Spider Ontology. Characters expressing equivalent evolutionary transformations were grouped together and more generally reformulated. For example, legacy characters expressing the curvature of eye rows (“straight/procurved,” “straight/recurved”) were condensed as a multistate character (“procurved/straight/recurved”). The sorting and condensing of legacy characters involved a critical review of the underlying homology hypotheses. Characters used for species-level phylogenies that are known to be very variable were not considered for this higher-level analysis. In this category fell most of the species-specific genitalic characters, such as number of coils of the male palpal embolus or the female copulatory ducts.

Character Hypotheses and Pseudocharacters:

Characters were reviewed and reformulated many times during this study, to accommodate newly found variation or revised anatomical interpretations. Special effort was placed in using the cell comments as supporting documentation for the decisions involving the reformulation or rejection of characters, and preserving finely grained observations independent of the grouping into coarsely defined character states. The following informal rules were used for the maintenance and documentation of characters, according to the capabilities of Winclada:

Invariant Characters:

Many invariant characters, which are phylogenetically uninformative within the current dataset, were kept in the matrix for documentation purposes. As shown in Ramírez et al. (2007), one of the impediments for adequate merging of phylogenetic data from multiple sources is the lack of documentation of characters considered “not informative” for a given analysis and thus not scored in the study.

Anatomical Terminology:

An effort was placed to map anatomical concepts and characters to terms in the Spider Ontology. Several new structures were discovered during this study, and those were added to the reference ontology. To avoid confusion with the usage of morphological terms and anatomical interpretations, the main character systems are preceded by a short description of the general morphology using a few fully labeled exemplar images.

Terminal Taxa and Specimens:

This analysis uses species as terminal taxa, although they are referred to by genus in the text; binomial names are used when more than one species is included, or for species whose generic placement is questioned or uncertain (e.g., Stephanopis ditissima, Odo bruchi). Occasionally some scorings and interpretations are complemented from observations from additional species, and this is noted in the comments section for each character. Two of the terminals (Ammoxenus and Zorocrates) were scored from two different species (one species for male, another for female), in the first case for availability of specimens, in the second because the taxonomy of the genus was solved when this project was already advanced. Species and voucher specimens examined for this study are detailed in appendix 1. Labels have been added to the vials (e.g., “Voucher for Dionycha study, M. Ramírez, 2000–2008”). The end dates in those labels varied as the timeline was postponed according to newly added terminals.

Implied Weighting and Sensitivity:

The dataset was analyzed under 10 different weighting regimes: equal weights, and implied weights (Goloboff, 1993) with increasing constant of concavity k  =  3, 6, 9, …, 27. The results are shown on a preferred tree corresponding to an intermediate concavity with k  =  9, which had the most groups in common with all the remaining weighting schemes (table 7). For each group, the number of weighting regimes where it is monophyletic was graphically represented as explained in figure 186. The election of weighting of characters against homoplasy rests on the experiments made in Ramírez (2003) and especially Goloboff et al. (2008b), finding a better performance when compared with analyses under equal weights; at any rate, sensitivity to changes in weighting regimes results in a lack of robustness.

Tree Searches:

For this dataset, traditional searches (e.g., 1000 replications of RAS+TBR) never hit the optimal scores, and the parsimony ratchet (Nixon, 1999; 100 replications of RAS+TBR+RAT, with 100 ratchet iterations) rarely did so. For each concavity, the dataset was analyzed using the new technologies of TNT, using sectorial searches, tree-drifting, and tree-fusion with the following commands and parameters (thanks to Pablo Goloboff for suggestions):

Subsequent searches with the preferred concavity resulted in 100 hits, at a rate of one hit every 30 seconds using a 2.8 GHz personal computer. With this rate of convergence on the same result, it is likely that the optimal tree was found.

Bremer Support:

Bremer support (BS; Bremer, 1994) values were heuristically estimated performing TBR swapping from the optimal trees, retaining suboptimal trees with increasing bounds, up to 51,000 trees. BS values are expressed in terms of fit, under concavity constant k  =  9, and are graphically displayed on branches according to the key in figure 186. To obtain greater precision, the weight of all characters was set to 100 (command “ccode/100 . ;”). Initial searches estimated a maximum value of BS  =  90. A script then ran several cycles, increasing the tree buffer to 3000 trees each cycle, increasing the suboptimal bound each time. Given the complexity of the dataset, some Bremer values may be overestimated. An additional analysis without collapsing branches (command “collapse  =  0;”) for low suboptimal values was used to correct some overestimations in the weakly supported groups.

Resampling Support Measures:

An estimation of the support upon resampling was made using 1000 pseudoreplicates of jackknifing under symmetrical resampling (Goloboff et al., 2003), each with an intermediately aggressive search (two replicates of RAS+TBR+RAT+Fusing). Absolute frequencies greater than 55% are graphically displayed along with sensitivity and Bremer support values (fig. 186).

Graphical Representation of Sensitivity and Support Measures:

The two measures of support (Bremer and jackknifing) and the stability of groups to changes in weighting regimes was summarized as a compound measure represented as branch lengths (fig. 186) (see Giribet, 2003). To recover support values for the weakly supported groups, the Bremer support was represented in a logarithmic scale.

Synapomorphies:

Synapomorphy lists were produced by taking into account only the unambiguous changes in ancestral states (e.g., 0 → 1, but not 01 → 1; 01 → 2, but not 01 → 12). Because synapomorphy lists for polytomies in consensus trees are dependent on all the optimal resolutions, all optimal dichotomous trees were first calculated, producing lists of synapomorphies that are common to all of them (Common Synapomorphies command in TNT, “apo[- ;”). A conservative estimation of the synapomorphies was made using the same procedure but considering all resolutions suboptimal by 0.01 units of fit (that is, collapsing any group of BS  =  0.01 or lower). The synapomorphies and groups that are lost after such operation are presented in parentheses (see table 11).

Evaluation of Alternative Resolutions:

Traditional groups or otherwise interesting hypotheses not found in the preferred trees were evaluated through constrained analyses, listing which characters would support or contradict such a resolution. First, a tree search was made with constraints forcing the monophyly of the group under evaluation, and the total fit difference was calculated with respect to the optimal tree. To estimate the character support against and in favor of alternative resolutions, the total fit difference is decomposed into the individual change of fit for every character. Characters that decrease their fit under the alternative resolution are in favor of such a grouping, and those that increase their fit oppose the group. The proportion of opposing and supporting characters is then calculated as C/F, where F is the sum of fit of all characters favoring a group, and C of those opposing it (Goloboff and Farris, 2001). Because constrained trees are always suboptimal, C/F is greater than 1. Values of C/F near 1 indicate that the constrained resolution is allowing almost as many characters to perform better, thus recovering a secondary signal in the data (fig. 206A). Conversely, large values of C/F indicate that such secondary signal is much smaller than the dominant one (fig. 206B, C). Because the lists of supporting and opposing characters may be extensive, only the characters adding most to the 75% of the range of F and C are reported.

Ordered Multistate Characters:

Multistate characters were considered as ordered ( =  additive), according to the following general rules (see table S2, see supplementary data:  http://dx.doi.org/10.5531/sd.sp.4):

(a) Nested states: A state definition implies the existence of a more general state: examples are “pronounced mound absent” – “present” – “mound plus dark long tooth”; “retrocoxal hymen absent” – “on leg I” – “on legs I–II” – “on legs I–III.” This case is unproblematic, as the multistate character could have been scored as two or more binary characters with well defined homologies.

(b) Continuous: The character states have an obvious continuous basis; examples are “anterior eye row procurved” – “straight” – “recurved”; “female tarsi curvature straight” – “slightly bent” – “strongly bent to coiled.” This cost regime follows all the methodological literature on continuous characters (e.g., Wiens, 2001; Goloboff et al., 2006).

(c) Intermediate: A character state is a plausible intermediate consistent with phylogenetic hypotheses or ontogenetic evidence: “two major ampullate gland spigots” – “one plus nubbin” – “one, no nubbin.” This cost regime relies on previous knowledge that may not apply to the whole tree (in the example, from Townley and Tillinghast, 2003, 2009). Not all morphological intermediates were considered as ordered. For example, the claw tuft setae (char. 163) could be construed as an ordered series “plumose” – “pseudotenent” – “tenent”; however, the results obtained here indicate that such a transformation series is not likely (fig. 198B).

(d) Gradual loss or gain: One of the states is “absent,” and another is a plausible intermediate for the loss or gain, as in the case of relictual structures: “palpal claw present” – “reduced to nubbin” – “absent”; “inferior tarsal claw large” – “small” – “absent”; “precoxal triangles, absent” – “fused to sternum” – “separate from sternum.” As in the examples above, these transformation costs rely on previous hypotheses that may be heterogeneous in the analysis. For example, the gradual palpal claw reduction is well documented in Salticidae, but not so in haplogynes. A binary scoring of this type of characters will usually score some cells as inapplicables, loosing the hypothesis of intermediacy.

(e) Counts: The character describes variation in number of discrete elements, without clear homology between each of the elements: “no tarsal trichobothria” – “single row” – “two or three rows”; “many Cy spigots” – “six” – “five” – “four” and so on. This is perhaps the most problematic case, as transformations of potentially nonhomologous structures may be conflated under the same state (e.g., the gain of basal and distal teeth, of mesal and ectal Cy spigots). Moreover, when counts are represented with several states, they may have a heavy impact on the analysis.

The five cases above are sorted from less to more contentious. In order to evaluate the impact of different treatments on the results, a few additional analyses were run with changes in the cost regimes, as in table 3.

Table 3 

Ordering of multistate characters

Experiments were made of sets of characters with ordered states (a–e), and three alternative cost schemes. U  =  unordered states. In experiment 9, * signifies that the characters are considered unordered when Lu/Lo < 0.66, where Lu is the length of characters with unordered states, and Lo is the length with ordered states, both calculated on the preferred tree (fig. 188). This index evaluates the departure from an ordered transformation series (i.e., a Lu/Lo value of 1 means there is a perfect fit to an ordered transformation series). See Phylogenetic Analysis Methodology for discussion.

Coxa:

The ventral basal corners of each coxa have fields of propriosensory setae (“hair plates”), which are deflected by the folding of the pleural membrane (fig. 44F). Most spiders have a breakage zone between coxa and trochanter, where leg autospasy occurs. For this reason, virtually all leg muscles between coxa and trochanter attach to small intermediate sclerites forming a ring (fig. 44E). Upon leg autospasy, the cleavage occurs across the sclerites, thus leaving the coxal muscles intact. The retrolateral surface of the coxa I, sometimes also II and III, may have an unsclerotized, often elevated patch, the retrocoxal hymen (fig. 45D). The coxa-trochanter joint has one prolateral condyle, and the movement is free in all directions.

Trochanter:

The trochanter may have a ventral distal indentation, the trochanteral notch (fig. 44C). The trochanter-femur joint has two condyles, one at each side, allowing movement in the vertical plane.

Femur:

The femur-patella joint has two dorsal-lateral condyles making a dorsal hinge, allowing movement in the vertical plane.

Patella:

The retrolateral distal margin of the patella is indented in an unsclerotized area, leading to one or two closely grouped lyriform organs (figs. 44D, 45E). The patella-tibia joint articulates on a single dorsal condyle, allowing horizontal movements only. The area of the retrolateral indentation is distorted by the movement, coincident with the placement of the lyriform propriosensor organs. The patella-tibia joint has a single dorsal medial condyle, allowing movements in the horizontal plane, capable of more retraction than protraction.

Tibia:

The tibia-metatarsus joint has two dorsal-lateral condyles making a dorsal hinge, allowing movement on the vertical plane.

Metatarsus:

In cribellate spiders the metatarus IV has a patch of curved thick setae, usually arranged in one or more rows along the dorsal-retrolateral margin (fig. 46C, D), but sometimes disposed on a less organized patch. The calamistrum is used for combing out silk from the cribellum. The calamistral setae usually have one or more rows of small teeth, which presumably card the cribellar fibrils (fig. 46E) (Foelix and Jung, 1978). Similarly as in most setae, calamistral setae were found to be innervated by three dentrites (Foelix and Jung, 1978). The metatarsus-tarsus joint has no condyle, and is free moving, but overflexion is limited by the dorsal metatarsal stopper. The metatarsal vibration sense organ (see Barth, 2002, and references therein) is a lyriform organ located at the dorsal end of the metatarsus, and is usually associated with a cuticular overhang, here denominated metatarsal stopper, which makes contact with a step in the tarsus (fig. 45C). The contact of the stopper triggers the response of the vibration sense organ (Barth, 2002).

Tarsus:

The tarsus usually has a dorsobasal step matching the opposing metatarsal stopper. The tarsal organ is situated approximately on the dorsal side of each leg and palpal tarsi, and has hygroreceptor and thermosensory function (see Barth, 2002). The tarsal organ is served by several sensilla with rimmed pores, where the sensory dendrites end. In entelegyne spiders the tarsal organ is protected by a capsule with a small opening (fig. 46B), while in more primitive spiders the sensilla are totally exposed (fig. 47A). The tarsus-pretarsus joint has two condyles, one at each side, allowing movement in the vertical plane (Hill, 1977). Near the condyles there is a pair of slit sensilla, one on each side, sensitive to vibrations (“claw slits,” Barth and Libera, 1970; “pretarsal slits,” Barth, 2002: 80; fig. 45A, B). There may be a ventral unsclerotized suture uniting the claw slits from both sides, the claw-slit suture, which seemingly provides room for cuticle deformation upon stress, and thus increased sensitivity on the slit sensilla (Barth, 2002: chapter VIII, fig. 8). At the tip of the tarsus there may be a group of tenent setae at each side of the claws, the claw tufts, which may arise from well delimited, articulated plates (fig. 45B).

Pretarsus:

The pretarsus is composed by a claw lever and the tarsal claws; it lacks setae. The median or inferior claw is solidly fused with the claw lever (fig. 45A), and is lost in two-clawed spiders (fig. 45B). The paired superior claws articulate with the claw lever and the tarsal tip through a movable membrane. Two tendons control the movement of the claw lever, the pretarsal levator (dorsally), and the pretarsal depressor (ventrally) (Hill, 1977). The claw lever usually has a series of longitudinal ridges at each side, the claw lever file.

Claws:

The claws are usually pectinate, that is, have a ventral line of teeth foming a comb (fig. 45A). The paired superior claws are normally larger than the median inferior claw.

97. Leg orientation: 0. Prograde (fig. 47A). 1. Laterigrade (fig. 44A). Comments: Oxyopes, Lauricius, Borboropactus, Geraesta: Intermediate (scored 01). Cycloctenus: contra Homann (1968) (scored 0).

98. Male leg I length: 0. Similar to the rest, or moderately longer (fig. 47E). 1. Much longer than the rest (fig. 47D). The males of several genera of Eutichuridae have extremely long forelegs, sometimes reminiscent of the sensory legs of Amblipygi. Comments: Hypochilus: both sexes very long (scored 0). Cheiracanthium: long, but not extremely so, male tarsus I only 1.5 of tarsus II (scored 0). Cheiramiona: femur I reaching midabdomen; leg I long in female (tarsus I twice as long as tarsus II), but not so exaggeratedly as in male (scored 1). Macerio: slightly longer than II (scored 0).

99. Leg III orientation: 0. Backward or laterally (fig. 47C). 1. Forward, here only Ariadna (fig. 47B). The third legs oriented forward is often correlated with living in tubes (see references in Izquierdo and Ramírez, 2008).

100. Tarsal cuticle texture: 0. Fingerprint (fig. 95K). 1. Smooth (figs. 95N, 96F). 2. Discrete cells adjacent (fig. 44G) or imbricate (figs. 58F, 89A, 96M). The imbricate or squamate cuticle is a classic character for araneoids (Lehtinen, 1975, 1996). A few intermediate or variable scorings were reported by Griswold et al. (2005: char. 10), and some more are reported here. Huttonia has a diffusely imbricate cuticle (fig. 94E). Galianoella has imbricate cuticle on tibiae (fig. 96J), but smooth on tarsi. Some terminals have a patchy distribution of fingerprint and smooth cuticle (fig. 94L), or with widely spaced ridges (fig. 96G). Schütt (2003: 145, char. 46) scored a few araneoid terminals as fingerprint, but from observations of the abdominal cuticle, which is differently structured. Comments: Hypochilus: from Forster et al. (1987: fig. 17) (scored 0). Thaida: from Forster et al. (1987: figs. 99, 100) (scored 0). Stegodyphus, Uloborus: from Griswold et al. (2005) (scored 1). Huttonia: diffusely imbricate (fig. 94E) (scored 2). Megadictyna: from Griswold et al. (2005), intermediate, small depressions (scored 01). Titanoeca, Dictyna: from Griswold et al. (2005) (scored 0). Neoramia, Metaltella: from Griswold et al. (2005), ambiguous (scored 01). Pimus: weak ridges (scored 01). Macrobunus: some ridged areas (scored 01). Aglaoctenus: see also Aglaoctenus sp. in Santos and Brescovit (2001) (scored 0). Cf. Medmassa THA: smooth or papillate (scored 1). Procopius: slightly papillate in zones (scored 1). Mandaneta: sometimes papillate (scored 1). Olbus: slightly papillate (Ramírez et al., 2001) (scored 1). Phrurolithus: widely spaced ridges, not fingerprint (scored 1). Camillina: intermediate (scored 01). Galianoella: imbricate on tibia, smooth on tarsi (scored 1). Legendrena: mostly smooth, but weak fingerprint on some patches (fig. 96I) (scored 01). Doliomalus: weak fingerprint in sectors, otherwise smooth (scored 01). Platyoides: there are shallow depressions delimiting cells, similar as in imbricate cuticle (scored 1). Trachycosmus, Fissarena, Cheiracanthium: weak but definite ridges (scored 0). Cithaeron: bumps and scales (see also metatarsus), also cracks (figs. 55C, 96N) (scored 2). Uliodon: tarsal tip smooth (scored 01). Polybetes: some areas about smooth (scored 0). Philodromus: rugged (scored 1). Epidius: very weak fingerprints (scored 01). Geraesta: also with bumps (scored 1). Boliscus: smooth with many papillate surfaces (scored 1). Tmarus: irregular texture (scored 1). Lyssomanes: intermediate, perhaps papillate (scored 01).

101. Tarsal cuticle discrete cells disposition: 0. Adjacent (fig. 44G). 1. Imbricate, distal cell margin elevated (figs. 89A, 96M).

102. Retrocoxal hymen: 0. Absent. 1. Present on leg I (figs. 3A, 49A–D). 2. Present on legs I–II. 3. Present on legs I–III. States are ordered. This is an unsclerotized, often elevated patch, on the retrolateral surface of the coxa, found by Robert Raven (see Bosselaers and Jocqué, 2002: 244). Bosselaers and Jocqué scored this character independently for males and females (2002: chars. 1, 2). Here I recorded separately both sexes, and concluded that males and females vary coordinately; the character may occasionally be intraspecifically variable, which may explain the discrepancies with their observations. A few scattered terminals have a hymen also on legs II (Sparianthinae VEN) and III (Xysticus). Comments: Acanthoctenus, Ctenus, Cycloctenus: wide unsclerotized patch (scored 1). Elaver: only seen in the couple INBio loc. 3339, and very faint (scored 1). Paccius: sclerotized (scored 1). Mandaneta: very large (scored 1). Neato: hard to see, the specimen is weakly sclerotized (scored 1). Miturgidae QLD: among the specimens examined, absent in several males, present in one female (scored 01). Miturga gilva: weak, weaker on male (scored 1). Syspira: Variable, more often absent. The white area is at the end of an increasing vertical series of setae (scored 01). Lauricius: some specimens with a tiny relict (scored 01). Liocranum: contra Bosselaers and Jocqué (2002) (scored 1).

103. Retrocoxal hymen size: 0. Small to medium-sized mound (fig. 49D). 1. Large unsclerotized patch. Comments: Paradiestus: larger in female (scored 0). Cf. Medmassa THA: well elevated (scored 0). Boliscus: the one on leg I much larger (scored 1).

104. Leg autospasy location: 0. Between trochanter and coxa. 1. Between patella and tibia. 2. Absent. Most of the scorings are inferred from preserved specimens, when legs are broken consistently at the same articulation. The generalized state (coxa-trochanter) was scored even when only one leg was found broken there. See character 109 for the male tibial crack.

105. Trochanter distal ventral margin: 0. Deeply notched, at least on legs I and II (fig. 47G). 1. Convex, straight, or shallowly notched (fig. 47H).

106. Trochanter IV length: 0. Less than 1.5 times the length of trochanter III. 1. 1.5 times as long as trochanter III or longer (figs. 43F, 47F, 48I).

107. Patellar indentation I–II: 0. Present (fig. 50B). 1. Absent (fig. 50G). In this dataset only Vectius lacks the patellar indentation, and in Platyoides it is reduced to a great extent. Both are extremely flat spiders with laterigrade legs, and the loss of the indentation seems related with the loss of (morphologically) lateral movements in the patella-tibia joint. Ventral to the lyriform organ there usually is one or more erect setae (fig. 50B–D) (Tharina Bird, in litt., observed in Ammoxenus). The lyriform organ is usually placed near the middle of the patella (fig. 50B, D), but a few basal representatives in this dataset have the lyriform organ of legs I and II placed near the distal margin of the patella (fig. 49E, G). In some of those terminals, there is an incision proximal to the lyriform organ (compare fig. 49F, G, with incision, and fig. 49E, without). Such an incision occurs in palpimanids, stenochilids, and Huttonia, adding support to the close relationships of those taxa. In the usual conformation with the lyriform organ placed near the middle of the patella, there may be some proximal sutures of diffuse limits. Comments: Eriauchenius: lyriform organ on anterior margin on leg I, indentation present on II–IV (scored 0). Platyoides: present on legs III and IV, on I and II distally closed, only a small narrow patch distal to a basal lyriform organ (fig. 50E, F) (scored 01). Vectius: totally closed in all legs (fig. 50G) (scored 1).

108. Patellar indentation I–II width: 0. Wide (fig. 50A). 1. Narrow (fig. 50B). This character is modified from Bosselaers and Jocqué (2002: char. 6). They described the structure as “a slit-like membranous indentation on the [retrolateral] side of the [patella]. May be very narrow or rather wide.” There are many intermediate conditions, scored (01). The indentation is frequently wider on the posterior legs. Comments: Huttonia: almost closed, lyriform organ marginal, distal to the indentation (scored 1). Eriauchenius: lyriform organ marginal on leg I, wide indentation on the rest (scored 0). Storenomorpha, Oxyopes, Apostenus, Phrurotimpus, Otacilia, Orthobula, Trachelidae ARG, Gayenna, Cheiramiona, Strotarchus, Ciniflella BRA, Tengella: intermediate (scored 01). Vectius: totally closed in all legs, lyriform organ in basal third (scored -). Platyoides: narrow on legs III and IV, but on legs I and II distally closed, only a small narrow patch distal to a basal lyriform organ (scored 1). Oedignatha: distally narrow (scored 01). Malenella: very widely open (scored 0). Cheiracanthium: not very narrow (scored 1). Lauricius, Plexippus: narrow on all legs (scored 1). Boliscus: narrow on leg I, very short on leg IV (scored 1).

109. Male tibial crack: 0. Absent. 1. Present, a suture line at the base of the leg tibiae, just distal to the basal pair of ventral spines (see Griswold, 1993: figs. 3–4; Griswold et al., 2005: char. 23). See character 104 for the patella-tibia autospasy of Filistatidae and Austrochilidae.

110. Calamistrum organization: 0. Linear, calamistrum setae in one or more lines (figs. 46C, 51A, C, E). 1. Oval, calamistrum setae in an irregular patch (fig. 52B, C). Because the presence of a calamistrum is perfectly correlated with the occurrence of a cribellum, the former was not retained as an active character in the analysis. Comments: Eresus: one line plus dorsal patch of calamistral setae (scored 01). Badumna: also dorsal patch of setae, but not of the calamistrum type (scored 0).

111. Calamistral rows: 0. Two. Only Hypochilidae (fig. 51A). 1. One (fig. 51I). 2. Three. Filistata has the calamistral setae in three staggered rows (fig. 51F, G), but the more basal calamistral setae still show the triseriate arrangement found in the basal filistatine Sahastata and in the Prithinae (fig. 51E).

112. Calamistrum rows setal arrangement: 0. Rows linear (fig. 51E, I). 1. Rows staggered on cuticular ridge. This state is unique to Filistatinae (fig. 51F) (Gray, 1995; Ramírez and Grismado, 1997).

113. Calamistrum origin: 0. Basal, 1. Median. All state 0 in this dataset. This character was used in Griswold et al. (2005: char. 28): “Calamistrum origins were classified based on the following formula: length from the metatarsus base to calamistrum origin divided by the metatarsus length. A ratio of less than 0.30 was considered basal to subbasal (figs. 143E, 145A). A ratio of greater than 0.30 was considered median origin (figs. 143D, 144A).” Comments: Stiphidion: because the calamistrum is long, but it is almost median! (scored 0).

114. Calamistrum setae teeth: 0. Absent (fig. 51H, K). 1. Present (fig. 52A, B).

115. Calamistrum setae teeth lines: 0. One line of teeth (figs. 51D, 52A). 1. Two or more lines (fig. 51B, J). Comments: Ciniflella BRA: multiple rows (scored 1). Ciniflella ARG: the more ventral setae with larger teeth (scored 1).

116. Hortipes sensor on dorsal metatarsus I and II: 0. Absent. 1. Present, an oval depression bordered by setae and two trichobothria (fig. 53F–H). A synapomorphy of Hortipes (Bosselaers and Jocqué, 2000). They report of immatures of H. contubernalis, “[w]hile running, the two pairs of front legs did not touch the substrate most of the time but were held outstretched in an antennalike fashion” (Bosselaers and Jocqué, 2000: 9). These odd structures seem like a directional air motion sensor, the lateral setae shielding signals from the sides and behind.

117. Metatarsal preening comb: 0. Brush or absent. 1. Distinct comb (fig. 53D, E). Several terminals had intermediate morphology, between a brush and a defined comb (fig. 53A–C). Comments: Ariadna: IV, retrolateral (scored 1). Donuea: comb made of very particular setae, like the chisel-shaped hairs of Jocqué (1991) (scored 1). Trachelopachys: comb plus brush (scored 1). Pseudocorinna, Jacaena, Sesieutes: intermediate (scored 01). Drassinella: brush (scored 0). Cf. Liocranidae LIB: group of thick setae, but not defined comb (scored 0). Teutamus: brush made of same kind of setae as comb in other terminals (scored 0).

118. Metatarsus ventroapical end extension: 0. Truncate or invaginated (fig. 57H). 1. Extending below tarsus (fig. 54C, F, G). Comments: Plexippus: much more evident on posterior tarsi (scored 1).

119. Metatarsal dorsodistal stopper: 0. Present (fig. 55D, G). The distal dorsal tip of the metatarsus extended beyond the lyriform organ, overhanging the tarsal base. 1. Absent (fig. 54A, B). The absence of the stopper seems to occur only in cases where the tarsus-metatarsus joint lacks movement, which in turn is apparently correlated with the metatarsus being shorter than the tarsus, as in symphytognathoids (fig. 54A, B) (see also the relictual stopper in Boliscus, fig. 54G, H). Comments: Mituliodon: membranous, more prominent on leg IV (scored 0). Xenoctenus, Uliodon: remarkably whitish, flexible, bright (scored 0). Boliscus: relictual, tarsus-metatarsus joint seemingly without much movement (fig. 54H) (scored 0).

120. Metatarsal dorsodistal stopper conformation: 0. Solid, about straight distal border (fig. 55D). 1. Membranous, trilobate (fig. 55H–J). A synapomorphy of Sparassidae, seemingly associated with their ability to overflex the tarsus-metatarsus articulation (see also char. 121). Comments: Cryptothele: perhaps not much mobility (scored 0). Zoropsis: Bosselaers (2002: 141) cites an apical, soft membranous rim on dorsum (scored 0). Pseudoctenus: not well sclerotized, cushionlike (scored 0). Hovops: that of leg IV asymmetrical (scored 0).

121. Tarsal dorsobasal step matching metatarsal stopper: 0. Present (fig. 55A, E–G). 1. Absent (fig. 54E). Lyssomanes lacks the matching step on the tarsus, which seems to be correlated with their ability to overflex the tarsus-metatarsus articulation, as sparassids do (figs. 54D, 218B). The tarsal surface in contact with the metatarsal stopper usually has regularly spaced files, which may play a role in the sensory mechanism as well. Comments: Filistata: scored from Kukulcania, a very small stopper, seemingly not much movement (scored 0). Prodidomus, Neozimiris, Lygromma: with longitudinal striations (scored 0). Austrachelas: a pair of slit sensilla just apical and at sides of step (fig. 55E) (scored 0).

122. Tarsus base sclerotization: 0. Sclerotized. 1. Weakly sclerotized ring. This character has been proposed as a synapomorphy of Archaeidae and Mecysmaucheniidae (Forster and Platnick, 1984: 104). In this matrix, it is present in Eriauchenius (Archaeidae), but also in Desis (Desidae).

123. Female tarsal cuticle continuity: 0. Entire, not cracked or pseudosegmented (fig. 58D). 1. Cracked or pseudosegmented (figs. 56A, B, 57A–D). Comments: Psechrus: flexuose, no clear cracks (scored 0). Apostenus: only tarsus IV (scored 01). Cf. Liocranidae LIB: on a median segment of tarsus IV (scored 01). Meedo: “At least tarsi III of males (often other tarsi as well, in both sexes) with cuticular cracks at about two-thirds their length” (Platnick, 2002: 23) (scored 1). Neato: cracked, irregular sclerotized fields remain (fig. 58A, B) (scored 1). Cithaeron: cracks not completing rings (fig. 57D) (scored 1). Ammoxenus: some of the cracks completing rings (fig. 57B) (scored 1).

124. Female tarsi curvature: 0. Straight (fig. 56E). 1. Slightly bent (fig. 56G). 2. Strongly bent to coiled (figs. 56A, 57A, C). States are ordered. Comments: Senoculus: straight (scored 0). Pseudoctenus: only slightly bent in female, because the tarsi are short (scored 0). Apostenus, cf. Liocranidae LIB: only IV bent (scored 0).

125. Male tarsus IV curvature and cuticle: 0. Straight or with continuous cuticle. 1. Bent, pseudosegmented. This is a modification of character 9 in Bosselaers and Jocqué (2002). Comments: Cycloctenus, Toxopsiella: curved, not pseudosegmented (scored 0). Agroeca: pseudosegmented (scored 1). Apostenus, Ammoxenus: male and female (scored 1). Cithaeron: all tarsi bent (scored 1). Cf. Liocranidae LIB: same as female (scored 1). Teutamus: a cracked area on the distal 1/5 (scored 1).

126. Tarsal organ conformation: 0. Exposed. The nerve endings are visible on the cuticle (fig. 47A). 1. Capsulate. The nerve endings are placed inside a cuticular pocket, accessible to the outside through a hole (fig. 58I, J). Comments: Hypochilus: from Forster et al. (1987: fig. 17) (scored 0). Thaida: from Forster et al. (1987: figs. 99, 100) (scored 0). Ariadna, Stegodyphus, Uloborus, Megadictyna, Titanoeca, Dictyna: from Griswold et al. (2005: figs. 151–Fig. 152.153). Corinna: see also Bonaldo (2000: fig. 67). Creugas, Micaria, Ammoxenus, Anyphops: not found (scored ?). Xenoplectus: from female palp (scored 1). Gayenna: from leg IV (scored 1). Amaurobioides: badly resolved in SEM but visible (scored 1). Liocranoides: from Platnick (1999: fig. 1) (scored 1).

127. Tarsal organ opening shape: 0. Round to oval (fig. 58L–N). 1. Teardrop or keyhole. The posterior margin of the opening is constricted (fig. 58G, K). Several terminals have limiting shapes, similar to oval (fig. 58I). 2. Long slit. The opening is very elongated in a long slit (fig. 58H). 3. Stellate. Only in Griswoldia in this dataset, this state was discovered by Griswold (1993: char. 56, state 2). Meriola is somewhat intermediate between oval and slit (fig. 58O). Comments: Uloborus, Megadictyna, Titanoeca, Dictyna: from Griswold et al. (2005: figs. 151–Fig. 152.153). Pimus: from Pimus napa (fig. 58G), but a Pimus indet. from Mendocino Co., California, has a round opening (Griswold et al., 2005: fig. 153J) (scored 1). Zoropsis: not markedly so (fig. 58I; see also Griswold, 1993: fig. 66) (scored 1). Toxopsiella: very long slit (fig. 58H) (scored 2). Clubiona: intermediate (scored 01). Elaver: most clear in leg IV (scored 1). Donuea: I teardrop, IV and palp oval (scored 01). Griswoldia: from Griswoldia acaenata, but teardrop on female palp (Griswold, personal commun.) (scored 3). Meriola: intermediate between oval and slit (scored 02). Orthobula: too dirty to see (scored ?).

128. Leg tarsal organ turret: 0. Absent. The tarsal organ is superficial (figs. 47A, 58M). 1. Present. The tarsal organ is placed at the top of a turret (fig. 68B; Griswold, 1991: fig. 29). Raven and Stumkat (2005: char. 50, state 3) report a similar structure in the zoropsid genera Megateg, Krukt, and Birrana. Comments: Griswoldia: from G. acaenata, but absent in other species (Griswold, 1991) (scored 1). Sparianthinae VEN: the tarsal organ is superficial on the female palp (scored 1). Boliscus: just slightly protruding (scored 0).

129. Tarsal sensory field depression: 0. Absent. The region bearing the trichobothria and tarsal organ at the same level as the rest of the tarsus (fig. 58C). 1. Present. The trichobothria and tarsal organ are grouped in a common apical depression (fig. 95A). Present only in Borboropactus in this dataset. This was called “tarsal pit organ” by Wunderlich (2004: 1738), one of the characters defining his “Borboropactidae” (see Thomisidae).

130. Apical ventral tarsal cuticle sclerotization: 0. Entire, sclerotized (fig. 73A). 1. Unsclerotized transverse suture below claws (figs. 45A, 59E, 60F, 61A, 62B, 68E, H, 69E, 75E, 76B). The suture seems always associated to the pair of ventral apical slit sensilla (“claw slits,” figs. 45A, 60D, 76B), thus it is called here claw-slit suture. In sparassids other than sparianthines, the claw-slit suture is served by four or more slit sensilla along the suture (fig. 68E, H). Comments: Dolomedes: only a depressed area in SEM (scored 0). Acanthoctenus: tenuous (scored 01). Senoculus: a well-defined articulation separated from claws (scored 1). Castianeira: there is a thin flexible line continued from the slit sensilla (fig. 69E) (scored 1). Ammoxenus: all pseudosegmented (scored -). Selenops: leaving a tight bunch of setae below claw tufts (scored 1). Sparianthinae VEN: wide membranous area leaving the most apical ventral setae apart (scored 1).

131. Extent of claw-slit suture: 0. Partial division, ventral, not reaching anterior superior margin (figs. 59E, 61A, F). 1. Total division, suture reaching anterior or superior margin (figs. 68E, 75E). Comments: Uloborus, Mimetus: tarsus IV, suture not a complete ring, interrupted in retrolateral dorsal sector (scored 0). Huttonia: totally divided on leg IV, almost totally on leg I (scored 01). Eriauchenius: suture reaching dorsal margin (fig. 59C) (scored 1). Stiphidion: suture reaching anterior dorsal margin (scored 1). Vulsor: suture tenuously reaching anterior margin (scored 01). Phrurolithus, Otacilia, Drassinella: suture reaching anterior margin (figs. 75E, 76B–D) (scored 1).

132. Serrate accessory claw setae: 0. Absent. 1. Present (figs. 59A, B, D, 60A). Comments: Stiphidion: very weakly serrated (Gray and Smith, 2008: fig. 3k) (scored 0). Homalonychus: Two large tenent setae. Zimiris (Prodidominae) has similar setae, without a tenent surface (Platnick and Penney, 2004: fig. 15) (scored 0). Oxyopes: one lateral short seta slightly serrate (fig. 62B) (scored 01). Senoculus: only on prolateral side (fig. 60G) (scored 1).

Fig. 15.

Structures of chelicerae, female. A. Gayenna americana (Anyphaenidae). B. Phrurotimpus alarius (Phrurolithidae) apical. C. Plexippus paykulli (Salticidae). D. Xiruana gracilipes (Anyphaenidae). E. Sicarius rupestris (Sicariidae). F. Drymusa rengan (Drymusidae).

133. Serrate accessory claw setae thickness: 0. Slender, as tactile hairs (figs. 59D, 59B, 60G). 1. Thick, as macrosetae (figs. 59A, F, 60A). Compare the thickness of macroseta and accessory claw setae in figure 59A and B.

134. Inferior tarsal claw I size: 0. Large (fig. 60D). 1. Small (fig. 60C). 2. Absent (fig. 65C). States are ordered. “Genera as Griswoldia, Phanotea (Griswold, 1991, 1994), and Janusia may have a small ITC on the anterior legs and only nubbins or none on the posterior legs” (Silva Davila, 2003: char. 111). Comments: Ctenus: Homann (1971) reports that Ctenus first and second instars have an ITC, in part with teeth (scored 12). Griswoldia: small claw on legs I–II, reduced to a nubbin in III–IV (scored 1).

135. Inferior tarsal claw teeth: 0. Toothed (fig. 60D). 1. Smooth (fig. 60E). Comments: Ariadna: one small tooth (scored 0). Huttonia: from Platnick and Forster (1984: fig. 341) (scored 0). Psechrus: three teeth in immatures, two in adult (Homann, 1971) (scored 0).

136. Inferior tarsal claw symmetry: 0. Nearly symmetric. 1. Strongly asymmetric (figs. 59A, 60B, G). Comments: Thaida: claw I curved toward prolateral side, IV toward retrolateral (scored 1). Araneus: claws rotated about 45°, I–II to prolateral, III–IV to retrolateral (scored 1). Megadictyna: Griswold et al. (2005: fig. 137B) is ambiguous for this (scored ?). Senoculus: curved to prolateral side, on all legs (scored 1).

137. Inferior tarsal claw IV curvature: 0. Dorsally convex or approximately straight (fig. 60D, E). 1. Sigmoid, dorsally concave reaching the tip (figs. 59F, 60A). The elongate, sigmoid inferior tarsal claws are characteristic of some symphytognathoids (Coddington, 1986; Griswold et al., 1998; Schütt, 2003). Comments: Thaida: claw IV, slightly sigmoid (scored 01). Dictyna: claw broken in preparation (scored ?). Mimetus: very slightly sigmoid (scored 01). Dictyna: claw broken (scored ?).

138. Superior tarsal claws teeth: 0. Toothed (fig. 60D). 1. Smooth (figs. 61B, 79D, 82B). Bosselaers and Jocqué (2000: char. 64) scored leg IV. Comments: Prodidomus: claws very long, some distal markings might be relics of teeth (fig. 82B) (scored 1). Neozimiris, Selenops, Hovops: intermediate, relictual teeth (fig. 62E) (scored 01). Phrurolithus: one blunt tooth on retroclaw I (fig. 75E) (scored 01).

139. Superior tarsal claws I teeth symmetry: 0. Both claws similarly toothed (figs. 70F, 79C, E). 1. Retroclaw with many fewer teeth than proclaw (figs. 64D, G, 65B, 66D, 67D, E). Comments: Zoropsis: male with more teeth (Homann, 1971) (scored 1). Philodromus: also noted and discussed by Homann (1975) (scored 1). Borboropactus: counting the lateral comb, which in other stephanopines is more integrated into normal teeth (scored 2). Holcolaetis: retroclaw smooth (scored 1). Falconina: two teeth each claw on I, one each on IV (scored 0). Cf. Liocranidae LIB: claws I with more, longer teeth than IV (scored 0). Vectius: I proclaw 8, retroclaw 6; IV proclaw 6, retroclaw 4 (scored 0). Anyphaena: proclaw ca. 17, retroclaw 7 (scored 1). Odo bruchi: almost identical side by side (scored 0). Stephanopoides: proclaw ca. 16 with several small basals, retroclaw 10 (fig. 64D) (scored 1). Boliscus: proclaw 10, retroclaw 12 (scored 0). Thomisus: proclaw 13, retroclaw 7 (scored 1).

140. Proclaw external comb in defined patch: 0. Absent (fig. 70F) or not well defined (fig. 64F). 1. Well defined patch of teeth forming a comb (fig. 64A). This occurs in Borboropactus in this dataset, it was also observed in other stephanopine thomisids.

141. Superior tarsal claw teeth insertion line: 0. Ectal line (figs. 64F, 67C, 79F). 1. Median line (fig. 79G). 2. Mesal line (figs. 64H, 71C, 77A, 78A, B, 79C). States are ordered. This is a modification of classic character of zodariids (Jocqué, 1991: char. 3). Several terminals have oblique (fig. 62D) or sinuous (fig. 60G) lines of teeth and were scored ambiguous. Comments: Eresus: basal ectal, apical mesal (scored 012). Nicodamus: not markedly so (scored 2). Cyrioctea: teeth insertion may be slightly mesal (scored 1). Cybaeodamus: teeth insertion slightly sinuous, basally median (scored 2). Cryptothele: teeth mesal, but reduced on leg IV (scored 2). Storenomorpha: retroclaw mesal, proclaw sinuous, basal ectal apical mesal (scored 2). Acanthoctenus, Ctenus, Vulsor, Liocranoides, Lyssomanes: oblique line, basals slightly ectal (scored 01). Elaver: claw thickness interferes a little (scored 01). Liocranum: slightly ectal (scored 01). Otacilia: reduced teeth, oblique line basals slightly ectal (scored 012). Galianoella: slightly ectal (scored 01). Ciniflella BRA: only slightly ectal (scored 01). Pseudolampona: very slightly internal, scored as median (scored 1). Borboropactus: ectal all legs, also in Onoculus and Epicadus (scored 0). Strophius: sinuous, basally slightly external (fig. 64H) (scored 2). Hispo, Plexippus: proclaw ectal (fig. 67D) (scored 0).

Tactile Hair:

The tactile hairs, usually referred simply as “hairs,” are the most frequent and widespread kind of setae. They are medium sized, have a curved shaft so that the tip is inclined toward the cuticle, and usually have barbs (fig. 84A). They are innervated by three neurons.

Macroseta:

Also known as spines, macrosetae are large articulate setae, with a thick and long shaft and robust socket (fig. 84B). Macrosetae are innervated by three neurons and become erect when the hemolymph pressure increases. Nerve impulses are generated only during the erection phase. They occur mainly on appendages, but sometimes similar setae occur on the abdomen or cephalothorax as well. The leg macrosetae occur in stereotyped patterns on the legs of spiders of the RTA clade, thus there are more or less standardized nomenclatures to describe the occurrence of macrosetae at given positions (e.g., Ramírez, 2003: 7, 51).

Scale:

Scales are setae with a small socket, bent in angle immediately after the insertion, so that they lay parallel to the cuticular surface (fig. 84C; Hill, 1979; Townsend and Felgenhauer, 1998a, 1998b, 1999; Townsend and Felgenhauer, 2001). As far as we know, scales lack innervation (Townsend and Felgenhauer, 1999). They may occur in a wide diversity of shapes.

Tenent Seta:

The claw tufts (fig. 85C) and scopulae (fig. 85D) are composed of setae specialized in adhesion to smooth surfaces. They have a defined patch of barbs with expanded tips, which make contact with the substratum and produce adherence through molecular forces. The same adhesion mechanism is used by hairs and pads of many animals, from beetles to lizards (Arzt, 2003).

Pseudotenent Setae:

These are setae with intermediate morphology between tenent and tactile hair, with acute tip and tenent barbs loosely organized on the contact side but usually not forming a pad (fig. 85E).

Trichobothria:

The trichobothria (fig. 84D) are sensory setae on the dorsal surfaces of legs and palps, specialized in detecting air movement (see Barth, 2002). The setal shaft is slender, perpendicular to the cuticle surface, usually curved backward and longer than the neighboring setae. They are disposed in longitudinal series of distally increasing length. Just above the thin articulation with the socket there is a basal expansion that may be variously sculptured. The socket forms a cup or bothrium, with an ample central cavity. The opening of the cup (alveolus) restricts the movement of the setal shaft. The bothrium is usally divided in proximal and distal plates; the proximal plate is often called trichobothrial “hood.”

Chemosensory Setae:

Sensory setae with an open tip (fig. 84E), chemosensory setae are usually innervated by 21 neurons, two of them mechanosensitive and restricted to the base, and 19 chemosensitive, with their dentrites extending into the shaft, and ending in the distal pore (see Barth, 2002). The shaft has an internal cuticular tube (fig. 84F) enclosing the dentrites, extending from the pore to about 2/3 of the shaft length (Foelix, 1970a, 1970b). The distal pore may be protected from direct contact with the substratum by an apical barb (fig. 84E). The chemosensory setae are inserted in a rather regular angle (about 70°) and have an “S” profile (Foelix, 2011). The articulation between shaft and socket is seemingly less movable; in some groups the articulation is totally exposed instead of folded, suggesting that they are not capable of much movement (fig. 85B). In several groups of small spiders and in spiderlings the chemosensory setae around the spinneret spinning fields are very similar to spigots, differing only by their lack of a base (fig. 85A) (see also Lopardo and Hormiga, 2007).

142. Tactile hair type: 0. Plumose or pseudoserrate, with many thin barbs (fig. 85A). 1. Serrate, with few, solid, triangular barbs (figs. 35D, 60A, 95F). See Lehtinen (1975) and (Griswold et al., 2005: char. 17). This is a traditional araneoid character (Lehtinen, 1967: fig. 1; 1975: 27), here also reported for some thomisids. Lehtinen (1975: fig. 7) reported “pseudoserrate” setae in the zodariid Hermippus, with similar morphology as in araneoids. Comments: Nicodamus: limiting case, close to serrate (scored 0). Psechrus: a good example for the “pseudoserrate” problem (scored 0). Geraesta, Xysticus, Stephanopis ditissima, Tmarus, Titidius: some dorsal tarsal hairs look like serrate (scored 01). Thomisus: most leg hairs are thick and smooth, except those at the sides of the claws and one median on top of the claws (scored 0). Strophius, Aphantochilus: mostly serrate (scored 1).

143. Spination legs I–II dramatically reduced: 0. With spines. 1. Virtually no spines (fig. 48G). Comments: Hypochilus: only metatarsi with spines (scored 1). Oecobius: Only some sparse spines (from Charles Griswold, in litt.) (scored 01). Huttonia: few macrosetae (scored 1). Cryptothele: somewhat reduced (scored 01). Micaria: absent on tibiae and metatarsi I and II (scored 01). Cf. Eutichuridae QLD: female spines: metatarsus I v p1-p1-p1, metatarsus II v p1 ap; male with more spines, including metatarsus I v 2-2-1 (scored 1). Cheiracanthium: tibial spines reduced (scored 0). Cheiramiona: spines fairly reduced (scored 01). Toxoniella: no spines on tibia (scored 0). Holcolaetis: present on femora (scored 01). Hispo: only femora p 1ap or p 2ap (scored 1).

144. Spination legs III–IV dramatically reduced: 0. With spines (fig. 48E, F). 1. Virtually no spines (fig. 48H). Comments: Cyrioctea: scoring taken from female C. spinifera, which has more spinose legs than other congeners. Prodidomus: only metatarsi III v 1 ap, IV v 1ap, p 1ap (scored 1). Ariadna: IV reduced, not III (scored 01). Oecobius: only some sparse spines (from Charles Griswold, in litt.) (scored 01). Neato: a few spines only on leg III (scored 1). Cheiramiona: Reduction intermediate (scored 01). Vectius: bristles in male are more like macrosetae, in thickness and in position (scored 1). Platyoides: only bristles, but not in the usual stereotyped position of macrosetae (scored 1).

145. Prolateral series of macrosetae on leg I: 0. Absent, prolateral macrosetae not organized in one row. 1. Several series of macrosetae with increasing length, as in mimetids. This character was mainly used to record the absence of a mimetid synapomorphy throughout the ingroup taxa. Araneus diadematus has some prolateral macrosetae roughly in a line, but not so well organized as in mimetids (scored 0).

146. Femoral dorsal median line of macrosetae: 0. Present, at least one macroseta (fig. 48E, F). 1. No dorsal median macroseta (fig. 48H). Comments: Ariadna: many dorsal spines in male, none on female (scored 01). Oecobius: d 1ap (scored 0). Megadictyna: d 1-0 (scored 0). Brachyphaea: male right IV d 1bas (scored 01). Pseudocorinna: I d 1, III or III and IV d 1ap (scored 0). Phrurotimpus: one basal on all legs (scored 0). Teutamus: only prolaterals, on I and II (scored 1). Jacaena: IV d 1bas (scored 0). Sesieutes: IV d 1bas (scored 0). Oedignatha: II d 1bas (scored 0). Vectius: only bristles, the basal dorsals on femora may be closer to macrosetae (scored 1). Centrothele, Legendrena: I–IV d 1bas (scored 0). Austrachelas: present on leg IV, those of leg I reduced to bristles (scored 0). Cithaeron: basal dorsal present on all legs (scored 0). Cheiracanthium: only p and r (scored 1).

Pedicel:

The pedicel is a narrow waist between cephalothorax and abdomen, and corresponds to the first abdominal segment. It has a regular pattern of two dorsal and one ventral sclerites, and there may be other small lateral sclerotizations in the area connecting the pedicel with the pleural area of cephalothorax. The dorsal anterior sclerite connects with the carapace. The posterior dorsal sclerite has a median convex area and one series of slit sensilla at each side, perpendicular to the body axis (fig. 98A; Barth, 2002). Internally, the areas bearing the slit sensilla are prolonged posteriorly into the abdomen as strong muscle apodemes (fig. 99A). The ventral sclerite (fig. 98B) is usually triangular with a forward-extending tip, but is rather variable in shape and degree of sclerotization.

Epigastrium:

The epigastric area corresponds with the second abdominal segment, between the pedicel and the epigastric furrow (fig. 98D). This region bears the anterior book lungs, the female genitalia and the male epiandrum. Some spiders, especially araneoids, have long, smooth, presumably propriosensory setae on the anterior face of abdomen around the pedicel (the “elongated pedicillate setae,” Agnarsson et al., 2008). The male epiandrum is the area just above the epigastric furrow, corresponding to the place of the female epigyne; it frequently has several epiandric spigots (fig. 98D). The uterus externus connects the female spermathecae with the ovaries, usually inside the epigastric furrow (fig. 99A). The book lung covers are usually more sclerotized and have a different sculpture than the neighboring abdominal cuticle (fig. 98C). The book lung spiracles are usually connected with the epigastric furrow (fig. 99A). The book lungs are tracheal structures organized as flat lamellae with very thin cuticle, allowing for gas exchange. The atrium and the first portion of each leaf are internally lined by a mesh of cuticular extensions (fig. 99B). On the leaf surface, there are hollow internal spacers arising from the ventral side, which prevent the collapsing of the air-filled leaf. Some Eutichuridae have tracheal tubes extending from the epigastric furrow, the epigastric median tracheae (fig. 99E); these seem to be extensions of the second entapophyses, similarly as occur with the median tracheae of the third segment.

Postepigastrium:

The third abdominal segment, here referred to as postepigastrium, extends between the epigastric furrow and the spiracles of the posterior respiratory system (fig. 98C). It contains the posterior book lungs or transformations thereof. Basal Araneomorphae without posterior book lungs still preserve book lung–like structures in their early development instars (fig. 97B). The basic tracheal pattern for Entelegynae is four simple tracheae restricted to the abdomen, opening in a single spiracle near the spinnerets (fig. 98C). The median tracheae are derived from the third entapophyses, and may retain a muscle insertion at the tip; the lateral tracheae are homologous with the atrium of the posterior book lungs (Purcell, 1909). Complex tracheal systems in anyphaenids have been shown to develop from the basic system of four tubes (Ramírez, 1995); the fully developed tracheal system appears after the dispersing molt. The main tracheal trunks are lined internally by a mesh of cuticular projections, which may form an internal layer as a spiral (fig. 99C), or reticulate plate (fig. 99D). This mesh works as an open truss for the tracheal wall, while preserving the very thin tracheal cuticle needed for gas exchange. The smaller tracheoles have a simple cuticle that breaks off as a spiral thread (fig. 99D).

198. Pedicel ventral sclerite–sternum articulation: 0. Free (fig. 41A). 1. Fused (fig. 100D, F). Comments: Castianeira: united by a thin, weakly sclerotized strip as in the precoxal or intercoxal extensions (scored 01). Cf. Moreno ARG: posterior end of sternum weakly sclerotized (scored 0). Aphantochilus: very distant, coxae IV join in between (scored 0).

199. Anterior margin of pedicel ventral sclerite: 0. Pointed (fig. 100A, G). 1. Widely truncate (fig. 100E). Comments: Ariadna: truncated, but not widely (scored 01). Eriauchenius, Trachelopachys: narrowly truncate (scored 01). Trachelas mexicanus: narrow truncation, wider in some congenerics (scored 01). Oedignatha: pedicel ventrally fused to sternum (scored -). Pronophaea: slightly truncate (scored 01). Drassinella, cf. Gnaphosoidea TEX, Neato, Tibellus: rounded (scored ?). Prodidomus: pedicel and sternum close to truncate but separate (by a pilose area!) (scored 01). Neozimiris: hourglass shaped, not very wide (scored ?). Galianoella: rounded (scored 0). Vectius: concave (scored 1). Platyoides: curved, convex anteriorly (scored 1). Cf. Eutichuridae QLD: small piece, hourglass shaped, concave anteriorly (scored ?). Eutichuridae MAD: very small piece, round in front and acute posteriorly (scored ?). Titanebo: very small sclerotization (scored ?). Hovops: the area is sclerotized, without well-defined sclerite borders (scored 01). Anyphops: ventral sclerite only faintly sclerotized (scored ?). Epidius: too pale (scored ?). Borboropactus: area of ventral sclerite partially sclerotized, but sclerite still defined (scored 0). Stephanopis ditissima: area of ventral sclerite partially sclerotized, diffuse limits (scored ?). Aphantochilus: three anterior acute ends (scored 01).

200. Pedicel ventral sclerite forming tube: 0. Absent. 1. Present, ventral sclerite embracing dorsal (fig. 100B, C). This character is here used in a more restricted sense than in Platnick (2000). The ventral sclerite is prolonged dorsally to embrace the anterior dorsal sclerite, without fusing to it. Comments: Sesieutes: the tube is abdominal (scored 0). Lampona: with lateral sclerites well separated (fig. 100H, I) (scored 0).

201. Dorsal scutum on female abdomen: 0. Absent (fig. 101G). 1. Present (fig. 101A, C, I). Comments: cf. Medmassa THA: very small, anterior (scored 1). Boliscus: entire dorsum very hard in subadult female (scored 01).

202. Extension of dorsal scutum on female abdomen: 0. Small, limited to anterior half of abdomen (fig. 101A). 1. Large, extending beyond anterior half of abdomen (fig. 101I). The small scutum found in some terminals, just above the pedicel (fig. 101C) might be scored as a separate character from the more dorsal scuta. Comments: Corinna: female has small piece above pedicel; in male same piece extends dorsally (scored 0). Castianeira: intermediate between a normal scutum and a small piece above pedicel (scored 0). Paccius: There is a small sclerotized scutum dorsal to the pedicel. The male probably has both, the small one integrated with the epigastric sclerite (scored 0). Pseudocorinna: just a small triangle above pedicel (scored 0). Jacaena: only small piece above pedicel (scored 0). Sesieutes: entire dorsum (scored 1). Oedignatha: anterior dorsal half (scored 1). Centrothele: small above pedicel (scored 0).

203. Female epigastric sclerite: 0. Absent, the epigastrium is soft, except for the epigyne and sometimes some patches around the book lung spiracles (fig. 167B). 1. Present, entire epigastrium sclerotized (fig. 178C). From Bosselaers and Jocqué (2002: char. 115). Comments: Mandaneta, Lamponella: sclerotized on pulmonary plates and behind pulmonary spiracles, but epigynum separated by soft cuticle (scored 01).

204. Abdomen epigastric sclerite in female, extension surrounding pedicel base: 0. Absent, the sclerite is limited to the ventral surface (fig. 101F). 1. Present, the sclerite forms a closed tube surrounding the pedicel (fig. 101J) (Bosselaers and Jocqué, 2002: char. 116). Comments: Otacilia: epigastrium sclerotized but not markedly so (scored 0).

205. Dorsal scutum on male abdomen: 0. Absent. 1. Present. From Bosselaers and Jocqué (2002: char. 102), see also Reiskind (1969: fig. 1). Comments: Medmassa: scutum present in M. semiaurantiaca, but absent in South East Asian species (Deeleman-Reinhold, 2001) (scored 1). Trachelas minor: faint (scored 1). Apostenus: from Ubick and Vetter (2005) (scored 0). Vectius: apparently slightly sclerotized, male not well preserved (scored 01). Legendrena: faint anterior-central sclerotization (scored 0). Cebrenninus: diffusely sclerotized (scored 1).

206. Extension of dorsal scutum on male abdomen: 0. Small, limited to anterior half of abdomen. 1. Large, extending beyond anterior half of abdomen. From Bosselaers and Jocqué (2002: char. 103). Comments: Falconina: about half the abdomen (scored 0). Meriola: diffuse (scored 1). Otacilia: anterior half (scored 01).

207. Male epigastric sclerite: 0. Absent. 1. Present. The epigastrium is sclerotized. From Bosselaers and Jocqué (2002: char. 105). Comments: Storenomorpha: slightly more sclerotized than in female (scored 1). Trachelas minor: faint (scored 1). Apostenus: from Ubick and Vetter (2005) (scored 0). Teutamus: fused with the dorsal scutum (scored 1). Pseudolampona: just a faint, homogeneous sclerotization (scored 1). Austrachelas: epiandric area sclerotized (scored 0).

208. Male epigastric sclerite surrounding pedicel base: 0. Absent. 1. Present, closed tube. From Bosselaers and Jocqué (2002: char. 106). Comments: Castianeira: fused with dorsal scutum (scored 1).

209. Ventral postepigastric scutum: 0. Absent. 1. Present in male (fig. 101D, H). I am considering here the scuta clearly posteriad of the epigatric furrow, as in Reiskind (1969). Several corinnids and lamponids have sclerotized patches just posterior to the pulmonary spiracle. Those are places of muscle insertions in many (if not all) spiders (Bosselaers and Jocqué, 2002: char. 14). This character is very variable, heterogeneous at least within Corinna and Castianeira. Comments: Pronophaea: sclerotized patches just posterior to the pulmonary spiracle (scored 0). Pseudolampona: contra Platnick (2000: 303) (scored 0).

210. Female inframammillary sclerite: 0. Absent. 1. Present. A small sclerotized patch just in front of the tracheal spiracle (not in this dataset, see fig. 101B). From Bosselaers and Jocqué (2002: char. 117). Comments: Pseudoctenus: just the spiracle protruding (scored 0). Olbus: both sexes with prolonged, slightly sclerotized anteror margin of tracheal spiracle (scored 0).

211. Male inframammillary sclerite: 0. Absent. 1. Present. From Bosselaers and Jocqué (2002: char. 107). Comments: Boliscus: ring around spinnerets (scored 1).

212. Postepigastric invaginations: 0. Absent. 1. Present. These are small depressions opposing the book lung spiracles (figs. 104B, 167B). They are most often sclerotized, although the degree of sclerotization is not easily observed in pale species. This character was first proposed by Platnick (2000: char. 11) as distinctive of Lamponidae, here reported for a wider taxonomic range. Comments: cf. Medmassa THA: present but not very deep (scored 01). Pseudocorinna: only the muscle attachment points (scored 1). Sesieutes: sclerotized plates (scored 0). Anagraphis: only small sclerotized plates (scored 0). Cf. Gnaphosoidea TEX: checked in digestion, the internal corners of book lung spiracle are dark (scored 0). Syspira: small sclerotized area from spiracle (scored 0). Boliscus: male with a continuous depressed transverse line (scored 0).

213. Abdomen anterior dorsal strong curved setae: 0. Present (fig. 102B–D). 1. Absent (fig. 102A). 2. Discrete macrosetae. From Bosselaers and Jocqué (2002: char. 101). Comments: Filistata, Eresus, Stegodyphus, Nicodamus, Cyrioctea, Cybaeodamus, Storenomorpha, Trachelopachys, Creugas, Apostenus, cf. Gnaphosoidea TEX, Malenella, Polybetes, Eusparassus, Sparianthinae VEN, Plexippus: entire dorsum with thick setae (scored 01). Araneus, Mimetus: entire abdomen with strong but not bent setae (scored 1). Badumna: not a group, entire abdomen with thick setae (scored 01). Cryptothele: pad of short, wide, modified setae (scored ?). Lampona: slightly larger setae (scored 01). Geraesta: groups of thick setae pointing medially (scored 01). Xysticus, Tmarus, Strophius: macrosetae uniformly on dorsum (scored 01).

214. Epiandrous spigots: 0. Absent (fig. 103D). 1. Present (fig. 103A, B). Comments: Filistata, Stegodyphus, Oecobius, Uloborus, Mimetus, Dictyna, Badumna, Pimus: from Griswold et al. (2005: figs. 157–Fig. 158.Fig. 159.Fig. 160.161). Senoculus: contra Silva Davila (2003) (scored 0). Aglaoctenus: absent, but patches of setae regularly distributed (scored 1). Cf. Gnaphosoidea TEX: observed in KOH digested specimen (scored 0). Centrothele, Macerio, Meedo, Austrachelas, Amaurobioides: observed with stereomicroscope (scored 0). Lamponella: just one shaft, medial (scored 01). Miturga cf. lineata: from Silva Davila (2003) and stereomicroscope (scored 0). Stephanopoides: apparently absent in stereomicroscope (scored ?). Hispo: from SEM by Junxia Zhang (scored 0).

215. Epiandrous spigots disposition: 0. Dispersed (fig. 103A). 1. Two definite bunches (fig. 103C). Comments: Filistata, Oecobius, Uloborus, Mimetus, Dictyna, Badumna, Pimus: from Griswold et al. (2005: figs. 157–Fig. 158.Fig. 159.Fig. 160.161). Medmassa: almost four groups (scored 1). Liocranum: two spigots in one side, one in the other (scored 1). Xenoplectus: two, well separated (scored 1). Eilica: only four spigots, in two pairs (scored 01). Cf. Gnaphosoidea TEX: only two medial spigots (scored 01). Legendrena: two bunches not well defined, with two spigots in the middle (scored 01). Lampona: two definite bunches in a common pit (fig. 103E). Lessertina: two areas, not very definite (scored 01). Miturga cf. lineata: under the stereomicroscope can see two dense groups plus some dispersed (scored 01). Lauricius: two groups, but not defined bunches (scored 1). Selenops: two wide groups (scored 1). Eusparassus: in several bunches (scored 0).

216. Anterior book lungs conformation: 0. Flat leaves (fig. 99F). 1. Tubular tracheae (fig. 105C). State not represented in this dataset. Comments: Zoropsis, Acanthoctenus: respiratory system from Griswold et al. (2005). Dolomedes: tracheae from Silva Davila (2003) (scored 0). Clubiona, Elaver: respiratory system from Silva Davila (2003) (scored 0). Cf. Medmassa THA: respiratory system from female maturity exuvia, decently preserved and visible (scored 0). Trachelas mexicanus: from Trachelas tranquillus (Platnick, 1974: 207) (scored 0). Prodidomus: small book lungs (scored 0). Lygromma: respiratory system from Lygromma simoni (in Ramírez, 1995) (scored 0). Ammoxenus: respiratory system from Petrunkevitch (1933) (scored 0). Cithaeron: narrow leaves (scored 0). Holcolaetis: respiratory system after H. vellerea from Wanless (1985) (scored 0). Xiruana: from Ramírez (2003) (scored 0). Stephanopoides: preparation MJR-1320 (scored 0). Portia: tracheae from Wanless (1978: fig. 1D) (scored 0). Hispo: respiratory system after H. inermis from Wanless (1981: fig. 5C) (scored 0).

217. Internal prolongations on book lung cover. 0. Absent. 1. Present. The book lung cover has internal prolongations facing the first book lung leaf. This character may be related with the report of book lung covers with pores in grate-shaped tapetum clade (L. Glatz, personal commun., in Homann, 1971: 258), “only visible in sections of old alcohol material.” Comments: Thomisus: in transverse rows (scored 1).

218. Epigastric median tracheae: 0. Absent. The epigastric furrow has internal apodemes for muscle insertion, not particularly elongated (fig. 99F). 1. Present. The muscle apodemes are remarkably elongated in thin tubes, forming a pair of tracheae (figs. 99E, 105B). A potential synapomorphy of a group of derived Eutichuridae. Comments: Cf. Gnaphosoidea TEX: only two entapophyses (scored 0). Lampona: Platnick (2000: figs. 37–Fig. 38.39) described these lobes as separating the postepigastric invaginations (scored 0). Cf. Eutichuridae QLD, Eutichuridae MAD: flat triangular extension extending in thin tube (scored 1). Cheiracanthium: in C. inclusum the apodemes are small, flat, rounded, projecting posteriorly, much larger in C. punctorium (scored 0). Cheiramiona: apodemes as flat triangles projecting posteriorly (scored 0). Eutichurus: large, round apodemes (scored 0). Macerio: round apodemes (scored 0). Miturga cf. lineata: like a wide central apodeme (scored 0). Mituliodon: apodemes small, flat, rounded, projecting posteriorly (scored 0). Zora: rectangular apodemes (scored 0). Cocalodes: dissected only below epigastric fold (scored ?).

219. Posterior book lungs or modifications: 0. Pair normal book lungs. 1. Reduced book lungs. Filistatines have reduced book lungs, with a few leaves in the hatching stage, of which only one is retained later in development (Griswold et al., 2005). Austrochilines have only one flat leaf (Ramírez, 2000). 2. Pair of tracheae (the lateral tracheae). 3. Absent (no lateral tracheae). The absence of a spiracle can be used to infer the absence of tracheae (fig. 115D). See also Ramírez (2000). Comments: Thaida: one flat lamella (scored 1). Filistata: one flat lamella (scored 1). Uloborus: tracheae from Opell (1979) (scored 2). Huttonia: after H. palpimanoides, Forster and Platnick (1984) and personal observation (scored 2). Megadictyna: tracheae from Forster (1970) (scored 2). Nicodamus: from Forster (1970) (scored 3). Titanoeca, Neoramia, Metaltella, Pimus: tracheae from Griswold et al. (2005). Desis: tracheae after D. marina from Forster (1970) (scored 2). Vulsor, Ctenus: tracheae from Silva Davila (2003) (scored 2). Cycloctenus, Toxopsiella: tracheae from Forster and Blest (1979) (scored 2). Teutamus: tracheae quickly observed from immature, preparation lost during staining (scored 23). Prodidomus: see also P. amaranthius (Lamy, 1902: figs. 26, 27) (scored 2). Legendrena: tracheae from immature partially digested in the trap liquid (scored 2). Neato: not examined, but there is a small, normal tracheal spiracle (scored ?). Trachycosmus, Desognaphosa: no spiracle (fig. 115D). Tibellus: from Lamy (1902: fig. 50) (scored 2).

220. Lateral tracheae branching: 0. Simple, linear. 1. Branched. Comments: Thaida: two “branches,” the modified book lung (scored 1). Pimus: tracheae from Griswold et al. (2005) (scored 0). Cheiramiona: very long, I have cut them during dissection of abdomen (scored ?).

221. Position of openings of posterior respiratory system (or apodemes): 0. Very close to spinnerets (fig. 114E). 1. Slightly separated from spinnerets (figs. 104C, 114C). 2. Well advanced, closer to epigastrium (fig. 104A, D, E). States are ordered. Comments: Megadictyna: only a narrow band separating from cribellum (scored 1). Doliomalus: about 1.5 times ALS length (scored 1).

222. Third entapophyses or median tracheae: 0. Present. 1. Absent, the internal cuticle is smooth, without prolongations for muscle insertion. See Ramírez (2000) and Griswold et al. (2005: char. 63).

223. Third entapophyses or median tracheae medially fused: 0. Separate. 1. Fused (fig. 105D). See also Ramírez (2000: char. 32). Comments: Huttonia: one median trunk with spicles and muscle insertion (scored 1).

224. Median tracheae: 0. Absent, or only apodemes. 1. Present.

225. Median tracheae branching: 0. Unbranched. 1. Slightly branched, from two to 10 branches. 2. Strongly branched, more than 10 branches, usually hundreds of thin tracheoles. States are ordered. Comments: Neozimiris: some branches on abdomen, main tubes pass to carapace where presumably divide (scored 2). Cheiramiona: two branches divide at middle of abdomen (scored 1). Stephanopis ditissima: flat, widened in the middle, with a short muscle insertion (scored 0).

226. Median tracheae passing to carapace: 0. Limited to abdomen. 1. Two large trunks with many ramifications passing to carapace (fig. 105A). Comments: Hortipes: median tracheae branched after pedicel (scored 1). Prodidomus: two bunches (scored 1). Thomisus: only abdomen dissected (scored ?). Plexippus: ramified in abdomen, and then passing through pedicel (scored 1).

227. Lateral tracheae dysderoidlike: 0. Absent. 1. Present. The large lateral tracheae have well-separated spiracles leading to large trunks suddenly splitting into many thin tracheoles (see Griswold et al., 2005: 39).

228. Spinnerets on abdominal tube: 0. Absent (fig. 114E). 1. Present. Just in front of the tracheal spriacle the abdominal cuticle is membranous, delimiting a short tube. This is typical of sparassids of the subfamily Sparianthinae (fig. 113A). A similar conformation has been recently reported for some Australian Zoropsidae as well (Raven and Stumkat, 2005). Comments: Cf. Gnaphosoidea TEX: not well-defined tube, but constricted before spinnerets (scored 0).

Cribellum:

The cribellum has only a short article, the cribellum base, and a wide spinning field lined with minute spigots (fig. 108D). The cribellar spigots lack a base; the shafts arise directly from the cuticle. The shafts of the cribellar spigots are different from those of spigots on spinnerets, except the paracribellars (see below). Cribellar spigots appear in the stage with most setae, still inside the eggsac, and previous to the dispersing stage (fig. 108A). In some spiders, however, the first spigots to appear on the cribellum are a pair of relatively larger spigots with shaft and base similar to those on spinnerets (fig. 108C); this has been found in Austrochilus (Austrochilidae) and Ectatosticta (Hypochilidae) (personal obs.).

Colulus:

The colulus is a relic of the cribellum, as an articulate lobe (fig. 108E), or just as a patch of setae (fig. 108F).

Anterior Lateral Spinneret:

The anterior lateral spinnerets have three articles (fig. 106C, D), of which the two distals may be reduced or lost. The basal article is the largest, nearly cylindrical. The median article is a crescent-shaped incomplete ring, covering only the ectal or anterior area. The distal article is a short ring around the spinning field, and may be interrupted in the mesal area. In Araneomorphae the spinning field has two areas more or less delimited, the ampullate field (of ampullate gland spigots) and the piriform field (of piriform gland spigots). The piriform field is often crescent shaped, covering most of the spinning field. The ampullate field is a partly sclerotized sector on the mesal area, bearing the major ampullate gland spigots and associated sensilla (fig. 106E, F). These sensilla are strain detectors similar to the slit sensilla, with a dentrite ending in a pore (Gorb and Barth, 1996).

Posterior Median Spinneret:

The posterior median spinnerets have a single article (fig. 106C, D). The spinning field is membranous, except for a slightly sclerotized area that may occur near the minor ampullate spigots. So far four gland spigot types are identified to occur in posterior median spinnerets (table 5) (fig. 107A).

Table 5

Characteristics of silk gland spigots

Summary of main characteristics of gland spigots in Araneomorphae. * Nubbins from aciniform spigots were sporadically reported for males of Argiope bruennichi (Araneidae; Müller and Westheide, 1993), some Uloboridae (Kovoor and Peters, 1988), and females of Hogna species (Lycosidae; Townley and Tillinghast, 2003). At least in Hogna the nubbins are in a very stereotyped, symmetrical location. All the detailed examination of other genera in the same families, and in many other families, shows that the nubbins are absent, thus the aciniform nubbins have very restricted occurrence (see also Townley and Tillinghast, 2009). ** Cylindrical spigots in immatures were documented in Mimetus notius, Larinioides cornutus, Neoscona theisu (Townley and Tillinghast, 2009: fig. 9, table 1; Yu and Coddington, 1989) and Meriola barrosi (this work, compare fig. 133G, H).

Table 5 

(Continued)

Posterior Lateral Spinneret:

The posterior lateral spinnerets have two articles (fig. 106C). The basal article is cylindrical, usually the longest article. The distal article is usually crescent shaped, open mesally, and contains the spinning field. So far about five gland spigot types are identified to occur in posterior median spinnerets (fig. 107B, D) (table 5). Basal araneomorphs, especially web builders, usually have a functional association of three spigots forming a triad near the distal end of the spinning field (fig. 107C, E). The identity of individual spigots forming the triad seems to vary across groups (Griswold et al., 2005: 61–62).

Spigots:

The spigots are the outlets from which the silk is extruded. They are most often inserted on spinning fields with soft, flexible cuticle. Each spigot has a base and a tapering shaft with a pore at the tip (fig. 106E), and is supposedly homologous with a seta. Chemosensory setae have a pore at the tip and reduced articulation with the socket, hence they seem the best candidates for spigot precursors. Spigots are named after the gland type they serve (Kovoor, 1977); to make the reading easier, spigot names are sometimes abbreviated, e.g., “piriform spigot” or “piriform” stands for “piriform gland spigot.” There are several silk gland types, and in general each type is served by a morphologically distinct spigot type (Coddington, 1989). The morphology of the shaft is usually more conservative than that of the base. Some spigot types occur in small number and in stereotyped positions, and it is possible to establish homology relations for individual spigots (singulars); other spigots occur in larger number, without a precise location (multiples) (Coddington, 1989). The exact location and number of multiples is slightly asymmetrical in the same individual. Reduction and specialization of gland spigot patterns resulted in several evolutionary transformations from multiples to singulars (e.g., the mesal pair of major ampullates in most Araneomorphae, the few cylindricals of araneoids and some corinnids). Table 5 summarizes the main aspects of the external morphology, ontogeny and function of the spigot types recognized thus far, updating upon the previous accounts by Coddington (1989) and Griswold et al. (2005).

Nubbins:

Nubbins are cuticular protuberances in the spinning fields, representing “a nonfunctional, only partially formed, i.e. vestigial, spigot, either morphologically singular or multiple” (as redefined in Townley and Tillinghast, 2003: 213).

Tartipores:

A tartipore is “a cuticular scar, morphologically singular or multiple, that results, after ecdysis, from a collared opening forming in the developing exoskeleton during proecdysis; the opening accommodates a silk gland duct, allowing the duct to remain attached to a spigot on the old exoskeleton during proecdysis” (as redefined in Townley and Tillinghast, 2003: 213). This mechanism allows the use of silk during the proecdysis, through gland ducts that pierce the forming cuticle to remain attached to the shedding spigots. The more generalized spigot types of araneomorphs (ampullates, piriforms, aciniforms), as well as those of mygalomorphs, have corresponding tartipores; the more specialized types lack them (cribellars and paracribellars, modified and flanking spigots in the PLS triplet, cylindricals) (table 5); the Haplogynae lack tartipores.

Anal Tubercle:

The six abdominal segments beyond the spinnerets are only superficially discernible in Mesothelae, and totally fused in Opisthothelae (Millot, 1936). The anal tubercle is the last abdominal segment (the segment 17, or abdominal segment 11), and has a tergite and a sternite, with the anus in between (fig. 109A). Spiders lack the telson (Millot, 1949).

229. Cribellum: 0. Present. 1. Absent. Comments: Stegodyphus, Megadictyna, Titanoeca, Dictyna, Neoramia, Stiphidion, Metaltella, Badumna, Psechrus: spinneret data from (Griswold et al., 2005). Desognaphosa: spinnerets from Platnick (2002).

230. Cribellum spinning field division: 0. Entire. The spigots cover the entire cribellar surface in an entire spinning field (fig. 108B, D). 1. Divided. The spigots leave a bare median band, making two separate spinning fields, one at each side (figs. 111E, 112A–C, 112E).

231. Cribellum base division: 0. Entire or slightly notched (figs. 108B, D, 111E, 112A). 1. Well divided in two lobes by a longitudinal furrow (fig. 112C, E).

232. Cribellar spigots: 0. Uniformly distributed (fig. 109C). 1. Clumped. The cribellar spigots are grouped into tightly packed groups surrounded by bare cuticle (fig. 112E, F). Comments: Zoropsis: the transverse series are here considered clumps (scored 1).

233. Clumps of cribellate spigots: 0. Entire transverse series (fig. 112C, D). 1. Transverse series of longitudinal segments (fig. 112E, F). 2. Spots. The clumps are small spots uniformly distributed (Griswold et al., 2005: fig. 97D, E). This state occurs in the zorocratid Uduba, not in this dataset.

234. Cribellum spigot morphology: 0. Strobilate. The spigots are thin, with evenly spaced annular expansions (fig. 109B). 1. Claviform. The spigots are expanded at the tip, and the annular ridges are superficial (fig. 111F). This is typical of filistatids.

234. Cribellar spigots surrounding cuticle: 0. With ridges. The cuticle between cribellar spigots has a sculpture of ridges (fig. 109C, D). 1. Smooth. The cuticle between cribellar spigots is smooth (fig. 109E). Comments: Zoropsis: too closely packed (scored -). Acanthoctenus: one circular ridge surrounding each spigot (scored 0).

236. Cribellum development, first spigots: 0. Many small spigots without base (figs. 108A, 111A, C, D), and small calamistrum. 1. Two large spigots with base (fig. 108C), without calamistrum. In filistatids and Hypochilus, the first spigots to appear in the cribellum are identical to the regular cribellar spigots, of minute size and without a base (figs. 108A, 110A–C, 111B–D), and the spiderlings of this stage have a small calamistrum with setae similar as in the adult. In Austrochilines and the hypochilid Ectatosticta, the first cribellar spigots are two large spigots with base, similar to those of spinnerets, two to three times larger than the regular cribellar spigots found in later stages (figs. 108C, 110D–H); in these cases the spiderlings lack a calamistrum. Comments: Thaida: scored from Austrochilus forsteri (scored 1).

237. Colulus: 0. Well-defined lobe (includes cribellum) (fig. 116D). 1. Hairy plate (median or paired) or two setae (fig. 108F). 2. Absent (fig. 115D). States are ordered. Cribellates are scored State 0, because the homology of the colulus with the cribellum is clear. It is often unclear whether there are one or two hairy plates, hence these conditions are not discriminated. Comments: Macrobunus: dorsal side sclerotized (scored 0). Cf. Medmassa THA: large hairy plate (scored 1). Teutamus: two setae each side (scored 1). Phrurotimpus: two setae seen in male (scored 1). Cf. Gnaphosoidea TEX: two setae (scored 1). Polybetes: a sclerotized pit! (scored 2). Eusparassus: a sclerotized depression, digestion suggests that it is a muscle apodeme (scored 2). Lessertina: large hairy plate (scored 1).

238. Spigots insertion articulation: 0. Simple, insertion of spigots continuous with cuticle or through simple fold (figs. 116E, 124C). 1. Insertion annulate, flexible (figs. 128B, 134E, 138B). Prodidomines have most of their spigots inserted on flexible, annulate articulations. Hypochilids have a superficially similar morphology, but the annulations or squamations occur all over the spigot base (fig. 116B).

239. Tartipores: 0. Present (fig. 118H). 1. Absent. Members of Haplogynae lack tartipores (fig. 117E, G). See Griswold et al. (2005: char. 70). Comments: Hypochilus: without tartipores, only small pores, commented on by Platnick et al. (1991: char. 63); these pores also occur in the hypochilid Ectatosticta (personal obs.) (scored 1). Ariadna: only some dubious scars in male PLS, and female PMS, but seemingly asymmetrical. Until new observations are available, I interpreted the male scar as an abnormality, and those on female PMS as foldings (scored 1). Cf. Gnaphosoidea TEX: a few can be seen on PLS (scored 0). Xysticus: those of PLS small (scored 0).

240. Spigot shaft surface: 0. Longitudinally ridged (fig. 123F). 1. Annulate. This is characteristic of filistatids (fig. 117F). 2. Smooth (fig. 118E). Sometimes only some spigots have sculpture while the rest are smooth; in that case I scored this character from the sculptured spigots. Griswold et al. (2005: char. 70) scored this character from ampullates only. Comments: Ariadna: Slightly irregular, mostly smooth (scored 2). Storenomorpha: Ridged only on cylindricals, the rest smooth (scored 0). Paradiestus: although major ampullates with smooth shafts (scored 0). Cf. Gnaphosoidea TEX: longitudinally ridged on PMS and PLS (scored 0). Polybetes, Eusparassus: some with weak longitudinal ridges, but mostly smooth (scored 02).

241. Anterior lateral spinnerets (ALS): 0. Present. 1. Absent. All terminals in this dataset have ALS. The loss of the ALS is characteristic of many Mygalomorphae.

242. ALS anteroposterior position: 0. Close to PMS and PLS. 1. Advanced far from PMS and PLS (fig. 115E). Prodidomids of the subfamily Molycriinae (not in this dataset) have the ALS far advanced, separated by pilose cuticle from the posterior spinnerets (Platnick and Baehr, 2006: figs. 12–Fig. 13.Fig. 14.Fig. 15.Fig. 16.17, 242, 243). These authors mention the synanthropic prodidomine genus Zimiris as the only other gnaphosoid sharing this character (see also Platnick and Penney, 2004).

Fig. 17.

Structures of chelicerae, female. A. Oxyopes heterophthalmus (Oxyopidae). B. Same, detail of basal posterior membranous mound, inset marked on A. C. Evarcha falcata (Salticidae; image by Junxia Zhang), basal posterior membranous mound. D. Eriauchenius workmani (Archaeidae), stridulatory file. E. Same, fang and peg teeth. F. Same, fang and cheliceral mound. G. Same, close-up, arrow to cheliceral mound.

243. ALS separation: 0. Contiguous to slightly separated (figs. 114D, 115B). 1. Separate about one ALS diameter or more. Gnaphosoids (fig. 115F) and cribellate taxa (fig. 106B) often have well-separated ALS. Several terminals with intermediate separations are scored as ambiguous (fig. 115C). Comments: Huttonia: slightly less than a diameter (scored 01). Cyrioctea: almost a diameter, thin ALS (scored 01). Pseudocorinna, Drassinella, Hovops, Anyphops: slightly separated (scored 0).

244. Specialized setae on ALS basal article: 0. None. 1. Mesal row of thick setae. Typical of the subfamily Filistatinae (Filistatidae) (fig. 117B, C). 2. Anterior bunch of thick setae in males. This condition occurs convergently at least in males of Zora (Miturgidae) and both sexes of Arachosia (Anyphaenidae; Ramírez, 2003) (fig. 114A, B).

245. ALS basal article crossed by diagonal membranous area: 0. Absent. Basal article entire (fig. 114D). 1. Present. Basal article crossed by a diagonal membranous area (fig. 116D). This character has been reported by Simon (1893: 310) and is seemingly a synapomorphy of Dysderoidea, as it is present at least in Trogloraptor (Trogloraptoridae) (Griswold et al., 2012), Dysdera (Dysderidae), Ariadna (Segestriidae), several genera of Oonopidae and Orsolobus (Orsolobidae) (personal obs.; Matías Izquierdo, personal commun.). Caponiidae, the reputed sister group of the Dysderoidea (see Ramírez, 2000) has an entire, two-segmented ALS without membranous area (Platnick et al., 1991: figs. 145, 150).

246. ALS intermediate article: 0. Present, incomplete lateral ring. Primitive Araneomorphae retain the intermediate ALS article as an incomplete lateral ring (figs. 117A, 118A). 1. Absent (fig. 115A). Comments: Polybetes: probably because of desclerotization of the basal article (i.e., internal side not sclerotized) (scored 0).

247. ALS distal article at ectal margin: 0. External margin entire. The distal article of the ALS is usually semilunate, mesally open in the major ampullates area. In most spiders the distal article is entire, with a continuous distal margin (figs. 121B, 126D). The sclerotization of the article is best seen with incident light in the stereomicroscope, but can also be inferred in SEM images by the setae insertions, as setal sockets occur only on sclerotized cuticle. 1. External margin interrupted. In higher gnaphosoids the distal article of the ALS is reduced, broken into relictual isolated patches with setae sockets, with at least the external margin interrupted (fig. 126B, E). The only remaining setae belonging to the distal article may be present just around the major ampullate gland spigot area (fig. 129B). In prodidomids the external margin is interpreted as being reduced to the isolated patches of setae at the base of each piriform spigot (fig. 128B). This character was scored here in a broader sense than the original use by Platnick (2002: char. 3). Here the isolated setal sockets of gnaphosoids such as Cithaeron (fig. 126E) and Platyoides (fig. 126B) are considered as interrupted ALS ectal margins. Comments: Xenoplectus: SEM defective in male, I only see the setae (scored 0).

248. ALS major ampullate gland spigots: 0. Absent. 1. Present. Present in all terminals in this dataset. Comments: Homalonychus: specimen with preparation MJR-562 did not spin a dragline, even when falling (scored 1).

249. Ampullate spigot shafts papillate: 0. Absent. The surface is smooth or slightly striated. 1. Present (fig. 131A) (Griswold et al., 2005: char. 68, state 1). Some corinnids have similarly papillate shafts in the cylindrical gland spigots figs. 133B, 136A). Comments: Corinna, Castianeira: with some papillae, not so markedly as in eresids, but mainly on cylindricals (scored 0).

250. Major ampullate and aciniform shafts shape: 0. Shafts cylindrical or tapering. 1. Shafts clavate, widened at the tip (figs. 128D, 134F).

251. Position of major ampullates relative to piriform field: 0. Only marginal cluster of major ampullate spigots. The major ampullates are placed only on the major ampullate field, a definite patch usually on the mesal margin of the ALS spinning field (fig. 116A, C), placing the ampullate spigots from both sides closer together, seemingly to help make a coherent dragline. The cuticle of the major ampullate field is smoother, more sclerotized than the piriform field, and usually has several sensilla (fig. 106E; Gorb and Barth, 1996). These sensilla assist in the identification of major ampullates in drastically modified patterns of spigots, e.g., when the piriforms are absent (fig. 126H) or highly modified, or the major ampullates are reduced. See also character 255 for a further modification of the major ampullate field. 1. Marginal cluster of major ampullates, plus some major ampullates dispersed among piriforms. Some of the relatively basal Araneomorphae (filistatids, eresids) have, in addition to a mesal marginal cluster, one or more major ampullates whithin the piriform field (fig. 117D, E, H, I). Comment: Oedignatha: there is a larger piriform close to the major ampullates (scored 0). Galianoella, Ammoxenus: piriforms absent (scored -).

252. Major ampullates, general number: 0. Three or more (fig. 116B). 1. Two or less. Most Araneomorphae have a consistent pattern of only two major ampullate spigots segregated on the mesal major ampullate field (figs. 106E, 118B). After reaching maturity, one of them may be replaced by a posterior nubbin (see char. 253 below). The ontogeny of these two spigots is well known for Araneus (Tillinghast and Townley, 1994; Townley and Tillinghast, 2003). See also Griswold et al. (2005: char. 58). Comments: Hypochilus: Two large plus several small major ampullates, similar to a piriform spigot (scored 0).

253. Major ampullates, number in female: 0. Two (generally with a major ampullate tartipore visible) (fig. 106E). 1. One plus a nubbin (generally with a major ampullate tartipore visible) (figs. 118B, 123E). 2. One, no nubbin (there may be a major ampullate tartipore) (figs. 116E, 125C). This infrequent condition is scattered in several families. In Desis, there are traces of what might be a very shallow nubbin (fig. 118E). Several Phrurolithidae in this dataset have only one major ampullate spigot without nubbin, but some closer relatives have a very small posterior major ampullate (see char. 254 below). States are ordered. This character was considered not applicable for a few terminals with more than two major ampullates (see char. 252 above). A previous character version also scored for the presence of a major ampullate tartipore; as can be seen from the comments below, the identification of such a tartipore is very often contentious. Comments: Eriauchenius: The nubbin is not smooth. Immature has two major ampullates (scored 1). Cybaeodamus: tartipore not seen but area crowded (scored 0). Medmassa: female observed with stereo, spigots visible (scored 0). Pronophaea: nubbin very small, visible on left spinneret (scored 1). Olbus: stereomicroscope, plus Ramírez et al. (2001: figs. 11–Fig. 12.13) (scored 0). Apostenus: clumped, tartipore might be hidden (scored 0). Liocranum: very small nubbin (scored 1). Toxoniella: major ampullate small, distinguishable by sensilla (scored 0). Otacilia: the anterior major ampullates very large (scored 0). Hortipes: very small nubbin (scored 1). Teutamus: tartipore not visble but shrunken cuticle (scored 0). Hortipes: very small nubbin (scored 12). Oedignatha: one is smaller, and more central (scored 0). Prodidomus: the tartipore might be between the major ampullates article and the piriform field (see Gnaphosa) (scored 0). Cf. Gnaphosoidea TEX: the tartipore might be between the major ampullates article and the piriform field (scored 2). Lampona: female tartipore on border of piriform field (scored 0). Ammoxenus: tartipore not seen, but area not well exposed (scored 0). Cithaeron: the tartipore might be in the furow between piriforms and major ampullate fields (scored 0). Gayenna: I cannot see a tartipore in male, dubious in female (scored 0). Systaria: tartipore not seen, poor preparation (scored 0). Strotarchus: I cannot see the tartipore (scored 1). Griswoldia: tartipore visible in G. urbense (Griswold, 1991) (scored 0). Ciniflella BRA: perhaps a tartipore hidden by the piriforms (scored 0). Boliscus: tartipore not seen, dirty preparation, major ampullate spigots scored from subadult female (no major ampullate reduction in adult females seen in other thomisids) (scored 0). Thomisus: dirty preparation (scored 0). Plexippus: left side two plus tartipore (scored 01).

254. Female major ampullate shaft sizes: 0. Anterior much smaller than posterior (figs. 123B, 124B, 127F). 1. Both similar size (fig. 123A). The posterior major ampullate can be slightly smaller (fig. 123F), but I distinguished only large size differences. The posterior major ampullate or its corresponding nubbin often embraces the base of the anterior major ampullate. 2. Anterior very thick, posterior thin (fig. 125A, E). States are ordered. Comments: Eriauchenius: immature similar sizes (scored ?). Homalonychus: posterior is external when invaginated (see also male nubbin) (scored 1). Phrurotimpus, Drassinella, Orthobula, Neozimiris: only one (scored ?). Neato: the bases are slightly different, the shafts less so (Platnick, 2002: figs. 47–Fig. 48.Fig. 49.Fig. 50.Fig. 51.52) (scored 1). Miturgidae QLD: poor preparation (scored ?). Xenoplectus: Anterior slightly larger (scored 1). Cf. Eutichuridae QLD: broken, but enough for scoring (scored 1). Ciniflella BRA: posterior slightly smaller (scored 1). Titanebo: posterior major ampullate slightly smaller (scored 1). Plexippus: left side has two, similar size (scored 1).

255. Female major ampullate field invagination: 0. Marginal field (fig. 123F). 1. Central invaginated field, transverse line. One of the major ampullate spigots is marginal, the other more central (figs. 119A, F, 123B). 2. Central invaginated field, longitudinal line. Both major ampullates are far from the margin of the spinning field (fig. 119B–D). States are ordered, as the transverse line has one of the major ampullates still in marginal position. This character was conceived as applicable only to the stereotyped pattern of one or two major ampullates in a definite major ampullate field (see chars. 251 and 252). The furrow or wrinkles delimiting the major ampullate field, as well as its sensilla, indicates that this is the marginal major ampullate field that has been invaginated, rather than that the major ampullate is within the piriform field, as in character 251. State 2 is a synapomorphy of Zodariidae, with a remarkable convergence in the zoropsid Uliodon (see Miller et al., 2010). Comments: Hypochilus: the two large ones might still be homologous to the marginal ones (scored -). Stegodyphus: one of the two large marginal major ampullates is out of the sensila field (scored -). Desis: Only one (scored ?). Cryptothele: adult female with nubbin and major ampullate in transverse line (fig. 119E), immature with two major ampullates in same positions (fig. 119F), both cases with marginal tartipore (scored 1). Homalonychus: posterior one is external when invaginated (see also male nubbin) (scored 1). Oedignatha: one marginal one central, also observed in several other species, including larger ones from India (scored 1). Galianoella: no piriforms (scored 02). Ammoxenus: piriforms absent, major ampullates in oblique line (scored ?).

256. Major ampullate field on anterior margin: 0. On mesal margin (fig. 116C). 1. On anterior margin (figs. 118A, D, 121C, D). This character has been proposed by Davies (Davies, 1999: char. 20). Comments: Desis: definitely anterior (scored 1). Camillina, Austrachelas, Lessertina: intermediate (scored 01). Ammoxenus: piriforms absent (scored ?).

257. Major ampullate field projection: 0. Major ampullate field on a flat (fig. 123F) or slightly domed area (figs. 126A, 127F, 128F, G, 129D–F). 1. MAmp field on a conical, well-defined, setae-bearing article (figs. 127D, 128B, E). The conical article found in some prodidomids seems an extreme of the trend found in other gnaphosoids of having the major ampullate field on a domed area. There are many intermediate conditions for a reliable scoring of such a subtly elevated area. It is not clear where the major ampullate tartipore or nubbin is placed in prodidomids with a conical article. In terminals with a domed major ampullate field, the major ampullate tartipore stays in the furrow between the major ampullate and piriform fields, but the nubbin stays on the major ampullate field (fig. 129F) and the major ampullate sensilla are still present (fig. 129E). Comments: Neozimiris: Platnick (1990: 37) referred to the fusion of the major ampullate field projection with the ALS basal article, not observed in this specimen (fig. 128B) (scored 1).

258. Major ampullates, number in male: 0. Two (generally with a major ampullate tartipore visible). 1. One plus a nubbin (generally with a major ampullate tartipore visible). 2. One, no nubbin (there may be a major ampullate tartipore). States are ordered. Same as character 253, but for the male. Comments: Uloborus: male after Waitkera from Platnick et al. (1991) (scored ?). Cybaeodamus: tartipore not seen, but poorly preserved (scored 0). Cf. Medmassa THA: male spigots seen in stereomicroscope (scored 1). Mandaneta: I cannot see the tartipore with the stereomicroscope (scored 1). Cf. Liocranidae LIB: tartipore not seen (scored 1). Hortipes: very small nubbin (scored 1). Oedignatha: the more central major ampullate is smaller, similar to a piriform spigot (scored 0). Cf. Gnaphosoidea TEX: only one, tartipore or nubbin not seen at this magnification from KOH digested specimen, needs SEM confirmation (scored 12). Meedo: one is broken (scored 0). Trachycosmus: male tartipore is marginal! (scored 0). Fissarena: tartipore not seen (scored 0). Malenella: male spinnerets not scanned (scored ?). Amaurobioides: in stereomicroscope, tartipore not seen (scored 1). Lessertina: small nubbin as in female (scored 1). Miturgidae QLD: I cannot see a nubbin, all piriforms hiding details (scored 12). Griswoldia: male spinnerets from from G. urbense (Griswold, 1991: figs: 37–40) (scored 1). Hovops: males observed with stereomicroscope (scored 1). Boliscus: nubbin or tartipore not seen, dirty preparation (scored 12). Xysticus: small nubbin (scored 1). Plexippus: dirty preparation, I cannot see the tartipore (scored 1).

259. Piriform gland spigots in adults: 0. Present (figs. 118B, 125A). 1. Absent (fig. 126C, D).

260. Piriform bases reduced: 0. Absent. Spigot base well defined from surrounding cuticle (fig. 125A). 1. Present. Spigot base not well defined, very short (figs. 116F, 118B). Comments: Huttonia: all spigot bases reduced (scored 1). Ammoxenus: piriforms absent (scored ?).

261. Piriform spigot base cuticle texture: 0. Longitudinal ridges (fig. 123F). 1. Concentric ridges (fig. 118F). 2. Smooth (fig. 119F). 3. Annulate-squamate as in Hypochilidae (fig. 116B). Comments: Oecobius: smooth (scored ?). Araneus: weak but present (scored 0). Desis: somewhat mixed (scored 1). Macrobunus: wavy, more often annulate (scored 01). Homalonychus: smooth (scored 2). Teutamus: some with very faint ridges (scored 02). Prodidomus: some spigots with annular ridges on PMS and PLS (scored 0). Lygromma: concentric waves (scored 2). Ammoxenus: piriforms absent (scored ?). Cheiracanthium: very faint (scored 0). Syspira: smooth (scored 2). Ciniflella BRA: waving (scored 01). Sparianthinae VEN: smooth (scored 2). Heteropoda: not enough magnification in SEM images (scored ?). Polybetes: very weak longitudinal (scored 02).

262. Female piriform shaft thickness relative to major ampullate shaft: 0. Piriform shaft thinner, or equal as in major ampullate (figs. 123F, 124E). 1. Piriform shaft thicker than in major ampullate (figs. 127F, 129A). In some terminals one of the major ampullates is reduced (see char. 254), in those cases this character is scored after the larger major ampullate (fig. 124E). Comments: Cyrioctea: longer (scored 0). Homalonychus: larger major ampullates in female, but male has small major ampullates (scored 0). Apostenus: piriforms rather thick, but only thicker than the smaller of the two major ampullates (scored 0). Cf. Liocranidae LIB, Pseudolampona: equal (scored 0). Gnaphosa: an equal number of piriforms and corresponding tartipores suggests that all piriforms are functional during molt (scored 1). Lygromma: very long piriform bases, but small shafts (scored 0). Lamponella: piriforms smaller than the larger major ampullate (scored 0). Doliomalus: about the same, all thick (scored 0). Cithaeron: about the same size (scored 0). Miturga gilva: the marginal piriforms larger in male (scored 0). Syspira: shaft thinner, only larger in male (scored 0). Odo bruchi: quite large (scored 0).

263. Piriform shaft-base transition: 0. Transition with a well-defined change in curvature (figs. 123C, D, 129H, J). 1. Transition on a continuous curvature, only a superficial marking (fig. 129A, G). Comments: Ariadna: slightly separate (scored 0). Storenomorpha: not sharply delimited (scored 0). Clubiona: well defined also in male (scored 0).

264. Demarcation between major ampullate and piriform fields: 0. Major ampullate field integrated with piriform field, or separated only by flat cuticle or wrinkles (figs. 116C, 123F). 1. Separated by deep furrow only in male (fig. 121A). 2. Separated by a deep furrow in male and female. The major ampullate field is well delimited from the piriform field by a deep furrow (figs. 118C, 124A, 126E, 129C). States are unordered; the distribution of the males-only demarcation does not suggest intermediacy. The furrow delimits well-separated major ampullate and piriform fields, sometimes in conjunction with a great development of the piriform field. Comments: Huttonia, Psechrus, Ciniflella BRA: superficial wrinkles (scored 0). Clubiona: superficial wrinkles in female (scored 1). Xenoplectus: not so markedly (scored 2). Phrurotimpus, Hortipes, Trachycosmus: intermediate, slightly marked separation in female (scored 012). Cf. Gnaphosoidea TEX: female from SEM, male from KOH digested specimen (scored 2). Titanebo, Polybetes, Eusparassus: major ampullate field depressed (scored 0).

265. Male separate major ampullate field with smaller piriforms: 0. Separate field only with major ampullates (fig. 125F). 1. Some small piriforms with the major ampullates (figs. 120C, D, 124G). This character was considered applicable only to terminals with major ampullate and piriform fields well delimited by a deep furrow. The piriform spigots on the major ampullate field are of similar shape as those of the female. Comments: Clubiona, Elaver, Agroeca: well defined in male (scored 1).

266. Piriform spigots size sexual dimorphism: 0. About same size in male and female (compare fig. 121E, F). 1. Male piriforms larger (compare fig. 122D, F). For those terminals with two sizes of piriforms (see chars. 265 and 267), the ectal and more numerous piriforms are used to score this character. Comments: Agroeca: I am scoring the ones in the main piriform field (there are smaller piriforms together with the major ampullates) (scored 1). Apostenus: male with only one piriform, in fact relatively smaller than those on female (scored 0). Doliomalus: male with slightly longer shafts (scored 0). Miturga gilva, Systaria: slightly so (scored 1). Miturga cf. lineata: slightly larger in male, especially ectal ones (socred 1). Syspira: very slightly larger, especially the ectal ones (scored 01). Odo bruchi: perhaps slightly larger in female than in male (scored 0). Ciniflella ARG: ectal piriforms slightly larger than in female (scored 1). Ciniflella BRA: at least the four ectal piriforms larger (scored 1).

267. Male piriforms enlargement in mesal sector: 0. All male piriforms larger than in female (fig. 122B). 1. Mesal piriforms closer to the major ampullates smaller than the rest, similar as in the female. This includes the small piriforms in the separate major ampullate field (char. 265).

268. ALS basal article cylindrical, with inflatable piriform field: 0. Absent, ALS a truncate cone (fig. 113B) or near cylindrical, but not inflated (fig. 113C). 1. Present in males. The ALS are large, cylindrical to wider distally, and the piriform field can be considerably inflated (figs. 120A, B, 122A, 124F). 2. Present in males and females (figs. 115G, 127A, 128A, 129B, C). States are ordered, because the sexually dimorphic condition is found in several taxa and seems a plausible intermediate for the morphology found in gnaphosids and prodidomids. The cylindrical ALS basal article occurs in conjunction with a modified piriform field. It is remarkable how this syndrome appears in clubionids, gnaphosoids, liocranids, and miturgines, some of which are distantly related. The inflatable piriform field seems to be a mechanism to expose and retract the enlarged piriform spigots (compare figs. 127A, 129C). This character is related with the absence of a distal segment on ALS, and enlarged piriforms (see chars. 247, 262 and 266); some terminals with inflated piriform field but not cylindrical ALS were scored absent for this character. Comments: Xenoplectus: piriform field somewhat inflated but not cylindrical (scored 0). Teutamus: intermediate, in male (scored 01). Lampona: inflated piriform field, but not cylindrical ALS (scored 0). Austrachelas: inflated piriform field, but not cylindrical ALS (scored 0). Cithaeron: female ALS not cylindrical (scored 1). Miturga gilva: basal article and piriforms slightly enlarged in male (fig. 122E, C) (scored 01).

269. Piriform spigots, number configurations by sex: 0. Male and female more than three piriforms. 1. Female several piriforms, male none to three. In phrurolithids the number of piriform spigots is consistently reduced in males (compare fig. 125A–C). This also occurs in some scattered gnaphosoids and liocranids as well (e.g., fig. 124C, D). 2. Piriforms absent in male and female (fig. 126F–H). States are ordered. Comments: Trachelas minor: female 7, male 4 (scored 0). Xenoplectus: female 5, male 3 (scored 1). Cf. Liocranidae LIB: male observed with compound, quite clear, 4 large piriforms (scored 0). Orthobula: female 7, male 3 (scored 1). Lampona: female several, male 5 (scored 0). Trachelidae ARG: both male and female 5 piriforms (scored 0). Teutamus: female many, male 3 (scored 1). Eilica: Female 3, male at least 1. Scored from Eilica bicolor female with four piriforms (Platnick, 1990: 23) (scored 0). Micaria: tentatively scored 1, the female has 1 piriform, and the male none (fig. 129H, I). Cf. Moreno ARG: male and female with 3 large piriforms (scored 0). Austrachelas: female about 33, male 11 (scored 0). Vectius: female 6, male 7 (scored 0). Cithaeron: female 4, male 1 (scored 1). Pseudolampona: male and female 1 (scored 0). Odo bruchi: female 8, male 5 (scored 0).

270. Central piriform spigot with plumose base: 0. Absent, base without barbs. 1. Present. Both sexes of Lygromma have a central piriform with barbs on its base, a morphology intermediate between spigot and seta (fig. 127E). This might be related with the occurrence of setae with a long, pore-bearing apical tube on leg and palpal tarsal tips (see char. 175), reminiscent of spigots.

271. Piriform spigots with elongate bases flanked by plumose setae: 0. Absent, the piriform bases are shorter than the shaft. 1. Present, moderately elongated bases with loosely associated setae (fig. 127F). 2. Present, extremely elongated bases closely encircled by setae. The lateral piriforms have an external arc of flanking setae (figs. 127B, C, 128B). In Neozimiris the medial piriforms have a complete circle of flanking setae (fig. 128B). The side of the flanking setae appressed against the spigot base is smooth (figs. 127C, 128C). States are ordered. The morphology of these piriform spigots was described by Platnick (1990, 2000: char. 25, 2002: char. 6) and Platnick et al. (2005) as a synapomorphy for Prodidomidae.

272. PMS minor ampullate gland spigots: 0. Absent (figs. 131D, 133C). In gnaphosoids the reduction in size of major ampullates seemingly occurs coordinately with the minor ampullates. In such cases the reduced minor ampullates are difficult to distinguish from aciniform spigots. 1. Present. Comments: Hypochilus: In living specimens (thanks to Jason Bond, same as preparation MJR-863) I see consistently two darker spigots: one anterior, one median-external. These may be homologs of the ampullate spigots not different in surface morphology from the acniforms (scored 0). Stegodyphus: interpreted according to Griswold et al. (2005) (scored 1). Cyrioctea: one tentatively identified as a minor ampullate because the aciniforms on PLS have much longer shafts (scored 1). Brachyphaea: only aciniform and cylindrical spigots (scored 0). Xenoplectus: there is at least a nubbin (scored 1). Phrurolithus: small spigot identified as a minor ampullate because of position, proximity to tartipore and comparison with male (scored 2). Trachelidae ARG: there may be a tartipore, but not all surface exposed (scored 0). Gnaphosa: at least one distinguishable (scored 1). Cf. Moreno ARG: seen in male (scored 1). Camillina: Platnick (1990: figs. 35–Fig. 36.37) illustrates a female of C. elegans, and there seems to be a second, smaller minor ampullate (scored 2). Vectius: no tartipores on PMS, subadult female perhaps with one or two tartipores (scored 2). Platyoides: anterior minor ampullate much larger, similar to modified PLS spigot (scored 0). Cf. Gnaphosoidea TEX: only one spigot, a cylindrical (scored 0). Ammoxenus: from male (scored 1). Pseudolampona: the interpretation of PMS-PLS spigots is somewhat tentative, as there is little difference between the presumed minor ampullate, modified PLS spigot, and aciniforms (scored 1). Titanebo: perhaps the tartipore in the middle, not visible (scored 0). Boliscus: tartipore not seen, dirty preparation, minor ampullate spigots scored from subadult female (no major ampullate reduction in adult females seen in other thomisids) (scored 0).

273. Female PMS minor ampullates, number: 0. Two (fig. 132D). 1. One plus nubbin (fig. 131F). 2. One, no nubbin (fig. 133E). 3. Only one nubbin (fig. 134A, B). States are ordered. There is generally a minor ampullate tartipore visible. Comments: Nicodamus, Titanoeca, Dictyna: from Griswold et al. (2005) (scored 2). Cybaeodamus: hard to distinguish (scored 01). Homalonychus: identified as minor ampullates because they are larger and male has nubbins instead (scored 0). Psechrus: tentatively identified after male nubbin (the presumable minor ampullates are similar to cylindrical gland spigots) (scored 0). Acanthoctenus: the large one anterior identified after the male (scored 0). Cycloctenus: there may be a hidden tartipore (scored 2). Trachelas minor, Pseudocorinna, Lauricius, Griswoldia, Petrichus: there may be a hidden tartipore (scored 0). Procopius: tartipore together with one of the minor ampullates in a mound (scored 0). Olbus: in same plate with anterior cylindrical gland spigot (scored 0). Liocranum: the second large spigot (from anterior) interpreted as minor ampullates because of the base not being conical (scored 0). Phrurolithus: small spigot identified as minor ampullate because of position, proximity to tartipore and comparison with male (scored 2). Orthobula: only two smaller spigots of similar size, could be interpreted as absent, or as in Phrurolithus (one minor ampullate) (scored ?). Anagraphis: not clean preparation; I can see the minor ampullates but not the tartipore (scored 0). Gnaphosa: identified tentatively as a minor ampullate and a tartipore, from G. sericata and G. taurica (scored 0). Prodidomus: There are some tartipores, perhaps of aciniform spigots. Interpreted differently from Neozimiris; here the minor ampullates are smaller than the aciniforms (scored 2). Meedo: interpreted from Platnick (2002: fig. 44) (scored 2). Trachycosmus: tartipore visible in male (scored 0). Fissarena: all very similar, not clean, at least one minor ampullate in the female, with large base (scored 02). Amaurobioides: many small tartipores, perhaps all of aciniforms (scored 2). Miturga cf. lineata: the second minor ampullate not individuated, but present in male (scored 02). Miturgidae QLD: poor preparation (scored ?). Zora: differing in the two specimens scanned (scored 01). Xenoctenus: only one seen, coded like this because any of the cylindricals might be a minor ampullate instead (scored 02). Cebrenninus: not very clear, though (scored 0). Cocalodes: perhaps the tartipore hidden close to the large minor ampullate (scored 0). Lyssomanes: the anterior minor ampullate might be a large aciniform instead (scored 02).

274. Male PMS minor ampullates, number: 0. Two. 1. One plus nubbin. 2. One, no nubbin. 3. Only a nubbin plus tartipore. 4. Only a large tartipore. Males of titanoecids have a large tartipore as the only putative remnant of minor ampullate spigots (Griswold et al., 2005). States are ordered. In states 0–2 there is generally a minor ampullate tartipore visible. Comments: Cybaeodamus: one plus nubbin? (scored 1). Homalonychus: sometimes two nubbins plus tartipore (scored 3). Acanthoctenus: the smaller minor ampullate absent on left side! (scored 0). Cycloctenus: tartipore small, of an aciniform gland spigot (scored 2). Mandaneta: tartipore visible (scored 0). Cf. Medmassa THA: tartipore not seen (scored 1). Liocranum: Many hairs, there may be a nubbin (scored 12). Cf. Liocranidae LIB: PMS with one spigot and a mound (tartipore or nubbin?), unclear homology of the spigot (scored 1234). Phrurolithus: Male with one slightly larger PMS shaft, linked to a tartipore, identified as minor ampullate. The second spigot identified as aciniform. In other Phrurolithidae these two spigots are of similar size, the one not associated with the tartipore may be missing. Identified the minor ampullate in those cases by their position besides the tartipore (scored 2). Phrurotimpus: very small (scored 2). Drassinella: minor ampullate, if present, equal to the aciniform gland spigot (scored ?). Teutamus: tartipore visible only on right side (scored 2). Anagraphis: not clean preparation (scored 1). Prodidomus: there are some tartipores, perhaps of aciniform gland spigots (scored 2). Cf. Gnaphosoidea TEX: from KOH digested male, it only shows a sclerotized scar (scored 4). Doliomalus: tartipore not seen, shrunken (scored 0). Amaurobioides: stereomicroscope, tartipore not seen (scored 2). Lessertina: bad preparation, but all the same as in female (scored 2). Miturgidae QLD: tentative interpretation (scored 1). Griswoldia: scored from image from Diana Silva Dávila (scored 1). Hovops: males observed with stereomicroscope, tartipore not seen (scored 1). Geraesta: tartipore not seen, may be hidden (scored 1). Xysticus: X. audax with a nubbin as well (E. Jantscher, personal commun.) (scored 2). Cocalodes: perhaps the tartipore hidden close to the large minor ampullate (scored 0). Holcolaetis, Portia: not good preparation, only the minor ampullate visible (scored 12). Storenomorpha: right side two minor ampullate, left side one (scored 012). Apostenus: very small minor ampullate (see sensilla at side) (scored 1). Lyssomanes: the anterior minor ampullate might be a large aciniform gland spigot instead (fig. 132C) (scored 02).

275. Minor ampullate on posterior median margin, posterior to the group of aciniforms: 0. Absent. The minor ampullates are not on the posterior median margin, or the aciniform gland spigots extend further behind the minor ampullates. 1. Present. The minor ampullates are on the posterior median margin, and the aciniforms are grouped anteriorly (figs. 130C, 131E, G, 132B). This is most evident on males, as in females some of the cylindrical spigots may extend in an external arc, behind the minor ampullates. Comments: Ariadna: only two spigots (minor ampullate and aciniform) (scored 0). Badumna: only two aciniforms more anterior to the minor ampullates (scored 0). Storenomorpha: minor ampullates in a common mound with anterior cylindrical spigot; right side of male with two minor ampullates (scored 01). Homalonychus: Only one posterior aciniform spigot (scored 0). Acanthoctenus: anterior minor ampullate much larger than posterior (scored 0). Neoanagraphis: minor ampullate and tartipore in common mound with anterior cylindrical spigot (scored 0). Cf. Liocranidae LIB: anterior group of aciniforms, median minor ampullates, and posterior group of cylindricals (scored 0). Eilica, Apodrassodes: in the middle of anterior sector (scored 0). Legendrena: only two aciniforms (scored 0). Ammoxenus: scored from male (scored 0). Anyphops: the minor ampullates are very large, much larger than the major ampullates (scored 0). Senoculus, Strophius: female with external arch of cylindrical spigots, some behind the minor ampullate (scored 1). Megadictyna: the spigots posterior to the minor ampullates might be aciniforms or cylindricals (scored 01). Galianoella: aciniforms absent, but minor ampullate clearly on anterior margin (scored 0). Macerio: one of the the minor ampullates close to the middle of the spinning field, some aciniform spigots posterior (scored 0). Eusparassus: mesal (scored 0). Borboropactus: the minor ampullates not so definitely posterior, the aciniforms describing an external arc, some of them posterior to the minor ampullates (fig. 132A) (scored 01). Stephanopis ditissima: the aciniforms also extend in an external arc, in the male reaching very slightly behind the minor ampullates (scored 1). Stephanopoides: some of the aciniforms beyond the minor ampullates (scored 1). Lyssomanes: the anterior minor ampullate might be a large aciniform spigot instead (fig. 132C) (scored 01).

276. Female PMS aciniform spigots, number: 0. Four or more. 1. Two or three. 2. One (fig. 134E). 3. None (fig. 133E). States are ordered. Comments: Mimetus: female 4 but male 3 (scored 0). Badumna: B. candida 3, B. longinqua 2 (scored 1). Neoramia, Falconina: 2 (scored 1). Zoropsis: female 5, male 3 (scored 0). Cycloctenus, Toxopsiella: many in female, 2 in male (scored 0). Castianeira: 11 on female, 6 on male (scored 0). Medmassa: female 1, male 1 or 2 (scored 2). Trachelas minor: female 6, male 4 (scored 0). Apostenus: left 2, right 3 (scored 1). Liocranum: female 3 or 4 (asymmetrical), male 4 (three posterior, one anterior) (scored 01). Drassinella: potential aciniforms are: 4 in female, 5 in male; this leaves at least 2 aciniforms in female, and 3 in male (scored 01). Prodidomus: a row of aciniforms with widened tips (scored 0). Camillina: 3 to 4 (scored 01). Neato: female 3, male 6 (scored 1). Syspira: Female several, male 2 (scored 0). Xiruana: 17–18 (scored 0). Odo bruchi: 7 (scored 0). Ciniflella ARG: Variable, 2 to 3 (scored 1). Petrichus: female 3, male 2 (scored 1). Cebrenninus: female 3, male 4 (scored 1). Boliscus: 15 (scored 0). Titidius: 3, cylindrical spigots tentatively identified by the slightly larger shaft and by not being associated with tartipores (scored 1).

277. Male PMS aciniform spigots, number: 0. Four or more. 1. Two or three. 2. One. 3. None. States are ordered. Comments: Badumna: B. candida two (scored 1). Storenomorpha: right one, left two (scored 12). Falconina, Petrichus: two (scored 1). Cf. Liocranidae LIB: only one spigot, either minor ampullate or aciniform (scored 23). Camillina: 6 (scored 0). Xiruana: 5–6 (scored 0). Odo bruchi: 8–10 (scored 0). Ciniflella ARG: variable, two to three (scored 1). Hovops: one posterior, seen with stereomicroscope (subadult none!) (scored 2). Boliscus: two aciniforms plus one aciniform with double shaft, symmetrical! (scored 01). Thomisus: 6 (scored 0). Lyssomanes: the anterior minor ampullate might be a larger aciniform instead (fig. 132C) (scored 01).

278. Aciniform spigot shafts two size classes: 0. Aciniform shafts uniform. 1. Aciniforms shafts in two size classes (figs. 132C, 137A). See Griswold et al. (2005: char. 85). Comments: Eriauchenius: only one aciniform in adult, but immature has two, same sizes (scored ?). Phrurolithus: only one or none (scored ?). Jacaena: only one (scored ?). Neozimiris: posterior male aciniform only slightly larger (scored 01). Systaria: aciniform shafts very long (scored 0). Lauricius: only one aciniform is slightly smaller in the female (scored 0). Aphantochilus: only one aciniform, basal on PLS, is slightly larger (scored 01). Strophius: PLS with smaller aciniforms in central area (scored 1).

279. Aciniform spigot shaft barbs: 0. Only with shallow sculpture, or smooth (fig. 133D). 1. With well defined barbs (figs. 132E, F, 136C, B).

280. Cylindrical gland spigots: 0. Absent. 1. Present. See Griswold et al. (2005: char. 86.) In this dataset (or elsewhere, as far as I know), when the cylindrical spigots are present, they occur in both PMS and PLS. Small cylindrical spigots in penultimate female were observed in Meriola barrosi (compare fig. 133G, H). Comments: Liocranum: the cylindricals are present in Liocranum and Mesiothelus, contra Bosselaers and Jocqué (2002: 264) (scored 1). Prodidomus: three, interpreted as cylindrical spigots because the aciniforms have expanded shafts (scored 1). Macerio: contra Ramírez et al. (1997) (scored 1). Strotarchus: PMS and PLS cylindrical spigots perhaps present in Strotarchus tropicum (SEM images by Alexandre Bonaldo, personal commun.) (scored 0). Philodromus: PMS with 12 aciniform spigots in both sexes, with the same distribution (scored 0). Epidius: cylindrical spigots tentatively identified by the longer and somewhat thicker shaft (scored 1). Stephanopis ditissima: scored uncertain because although all the spigots are similar, there is a difference in the margin of the spigot base, similar as in other thomisids (e.g., Strophius), and the female has many more spigots than the male, on PLS and PMS (scored ?). Stephanopoides: slightly but consistently larger shaft than those of the aciniforms (scored 1). Boliscus: adult female not scanned (scored ?). Xysticus: Demir et al. (2008: fig. 7) illustrate the eggsac of Xysticus pseudorectilineus, coriaceus, flat, on a stone (scored 1). Holcolaetis: Wanless (1985) reported gnaphosidlike eggsacs (scored 0).

281. Cylindrical spigot shaft rotund, incised: 0. Absent. 1. Present, shaft with longitudinal incisions (Platnick and Shadab, 1993: figs. 27, 28; Griswold et al., 2005: fig. 25C, D). State 1 was proposed as a synapomorphy of Mimetinae (Platnick and Shadab, 1993). Harms and Harvey (2009) found rotund but smooth shafts in Australomimetus (and other unspecified genera as well), and suggest that the incised shafts may be a synapomorphy of only Mimetus and Ero. No other terminal in this dataset has similar cylindrical spigots. Comments: Mimetus: cylindricals in PMS indented as well (scored 1).

282. PMS cylindrical spigot bases sunken: 0. Base raised from surrounding cuticle. 1. Base sunken in surrounding cuticle (fig. 132G; Bosselaers and Jocqué, 2000: fig. 7b). This is a synapomorphy of Hortipes. Of the pair of cylindrical spigots on PLS, the medial one is also sunken (fig. 136D).

283. PMS cylindrical gland spigots, number: 0. Many. 1. Five. 2. Four. 3. Three. 4. Two. 5. One. States are ordered. Comments: Oecobius, Dictyna: one cylindrical, interpreted as in Griswold et al. (2005) (scored 5). Senoculus: Six (scored 0). Brachyphaea: five or six (scored 01). Apostenus: Identified with difficulty with reference to PLS, then the other two large spigots identified as minor ampullates (scored 5). Liocranum: the second large spigot, counting from anterior to posterior, interpreted as a minor ampullate (scored 3). Xenoplectus: six (scored 0). Lampona: I counted six cylindricals (scored 0). Austrachelas: 10 (scored 0). Desognaphosa: variable (scored 012). Miturgidae QLD: poor preparation, identified some cylindricals, but the minor ampullates are not clear (scored 1234). Petrichus: right two, left three (scored 34).

284. PMS cylindrical spigots, clustering: 0. Mixed with aciniforms or minor ampullates (fig. 131C). 1. Isolated posterior group (figs. 133F, 134C). In this dataset, when many cylindrical spigots occur in a posterior isolated group, they are aligned in longitudinal rows; there are some cases of irregular lines, but none in a clearly disorganized pattern. Comments: Oecobius, Cyrioctea: only one (scored ?). Neoramia: few spigots, one PC close to anterior cylindrical spigot (scored 01). Pimus: only one, posterior (scored ?). Meriola, Trachelopachys, Toxoniella, Phrurolithus, Phrurotimpus, Otacilia, Drassinella, Orthobula, Jacaena, Fissarena, Ammoxenus, Meedo: isolated posterior group forming rows (scored 1). Sesieutes, Trachycosmus: rows not well formed (scored 1). Centrothele: isolated posterior group in a row (scored 1). Lamponella: only one, posterior (scored ?). Austrachelas: only a couple of aciniforms reaching anterior sector of the two cylindrical spigot rows (scored 1). Vectius: minor ampullate anterior, right side with one cylindrical anterior to minor ampullate (scored 01). Griswoldia: one of the cylindricals together with the aciniforms (scored 0).

285. PMS paracribellar spigots: 0. Absent. 1. Present (fig. 130A, E, I). Comments: Stegodyphus: note that the triad flanking spigots have longer shafts than those of the surrounding aciniforms (Griswold et al., 2005: figs. 33J, 37D) (scored 0). Zorocrates: see the shaft fused to PLS modified spigot! (Griswold et al., 2005: fig. 101D) (scored 0).

286. PMS paracribellar spigots, number: 0. Two or more (fig. 130I). 1. One (fig. 131B).

287. PMS paracribellar spigots distribution: 0. On anterior margin of spinning field (figs. 107A, 130A, H). 1. Midfield (Griswold et al., 2005: figs. 73C, 77C). 2. At posterior margin of spinning field (fig. 130E). See Griswold et al. (2005: char. 91).

288. PMS paracribellar spigots encircling anteriorly: 0. Bunched (fig. 130H). 1. In a row, encircling anteriorly (fig. 130B). See Griswold et al. (2005: char. 92.) Comments: Filistata: Pc posterior, inapplicable (scored -). Neoramia, Stiphidion: only two (scored -). Metaltella: dispersed (scored -). Badumna: midfield, inapplicable (scored -). Pimus: only one (scored -).

289. PMS paracribellar spigot base shape: 0. Cylindrical. 1. Long, narrow, flattened. See Griswold et al. (2005: char. 94.) State 1 is a synapomorphy of phyxelidids, not represented in this dataset.

290. PMS paracribellar spigot shaft: 0. Strobilate (fig. 130I). 1. Floppy (fig. 130E). The paracribellar spigots of filistatids are smooth, not strobilate, seemingly correlated with the smooth cribellar spigots (see char. 234).

291. PMS paracribellar spigot shafts grouping: 0. Every base with a single shaft. 1. Several shafts grouped on the same base. See Griswold et al. (2005: char. 93). Except for Dictyna, the fused paracribellars are monophyletic in this analysis (Clade 183). Comments: Metaltella: some paracribellars with two shafts (scored 1).

292. PMS-PLS anterior claviform setae: 0. Absent. 1. Present (figs. 130D, F, G, 138B, E). This is a classic character of filistatids (Gray, 1995: char. 18; Ramírez and Grismado, 1997: char. 7), here found in prodidomines as well. Comments: Filistata: one on PMS, contra Ramírez and Grismado (1997) (scored 1). Prodidomus: on PLS, flattened, look hyaline with stereoscope, but also on PMS in Prodidomus dalmasi (Platnick, 1990: fig. 131) (scored 1). Neozimiris: mostly on PLS, a few on PMS (scored 1).

293. PLS rows of claviform setae: 0. Absent. 1. Present. In Prodidomus and Neozimiris, the modified setae occur in one (fig. 138C) or more (fig. 138D) rows on the PLS.

294. Female PLS very short: 0. PLS long at least half the ALS. 1. Shorter than half ALS (fig. 113D). In this dataset only Cryptothele has extremely reduced PLS, with short basal segment, and distal segment just a crown of setae. The immature only bears short stubs, without spigots (fig. 113E).

295. PLS cylindrical gland spigots, number: 0. Six or more. 1. Five. 2. Four. 3. Three. 4. Two. 5. One. States are ordered. Comments: Raecius: at least four (scored 012). Petrichus: right two, left three (scored 34). Lauricius: PLS partially collapsed (scored ?).

296. PLS modified spigot: 0. Absent. 1. Present (fig. 137D). For a discussion, see (Griswold et al., 2005: char. 96). Comments: Hypochilus: see notes in Griswold et al. (2005: 30); in living specimens (thanks to Jason Bond, same as preparation MJR-863) I see consistently two spigots with thicker shafts, close together, paler (scored 1). Stegodyphus: interpreted according to Peters (1992) and Griswold et al. (2005) (scored 1). Nicodamus: apical nubbin on male PLS not symmetric (scored 0). Macrobunus: The cylindrical on female PLS is basal external, and might be confused with a modified PLS spigot. However, a subadult male has no large spigot on PLS (scored 0). Homalonychus: identified as modified PLS spigot because the male has a nubbin in that position (scored 1). Oxyopes: male with isolated apical group of longer spigots (scored 0). Meriola: small modified PLS spigot (scored 1). Teutamus: one aciniform only, distinguished from a modified PLS spigot because of the tartipore (scored 0). Gnaphosa: smaller than aciniforms, more visible in G. sericata and G. parvula (Platnick, 1990: fig. 7) (scored 1). Legendrena: identified as modified PLS spigot because it is larger than the only aciniform in female PMS (scored 1). Vectius: male without any spigot, female with a few small central nubbins (scored 0). Trachycosmus: no evident modified PLS spigot, but central spigot on male PLS with a sensilla associated (scored 0). Syspira: distinguishable only in male (scored 1). Uliodon: Imaged male with one large spigot on PLS, only on left side. Checked with more specimens: two other males lacking modified PLS spigot. The asymmetric scanned large spigot on preparation MJR-344 is interpreted as an anomaly (scored 0). Ciniflella BRA: only nubbins (scored ?). Philodromus: there is a nubbin on the male PLS, but also one on PMS, and the female does not have any larger spigot on PLS (scored 0). Lyssomanes: The apical larger spigots on PLS are accompanied by tartipores, hence they are not candidates for a modified PLS spigot. Scored as aciniform gland spigots in two size classes (scored 0).

297. PLS modified spigot conformation in adult male: 0. Nubbin (fig. 135B). 1. Spigot (fig. 136F). Comments: Eresus: reduced (scored 01). Stegodyphus: from Griswold et al. (2005) (scored 1). Psechrus: from Griswold et al. (2005); note that the modified PLS spigot occurs in a male from Papua New Guinea (their figs. 100D, 102D), but is replaced by a nubbin in a male from Thailand (fig. 54D) (scored 1). Meriola: bad preparation (scored ?).

298. PLS modified spigot position: 0. Among the aciniforms, central (fig. 137C). 1. Marginal-apical (fig. 136E) or marginal-median (fig. 135E). 2. Marginal basal, segregated from the rest of the spigots (fig. 135D). Comments: Uloborus, Araneus: marginal external median (scored 1). Homalonychus, Galianoella, Legendrena, Neozimiris: too few spigots to decide (scored 01). Trachelidae ARG: slightly marginal in female because of the very large cylindricals, central in male (scored 01).

299. PLS modified spigot accompanying spigots: 0. PLS modified spigot not particularly associated with other spigots (figs. 135G, 137B). 1. Closely associated with accompanying spigots (fig. 135A, F, H). Comments: Hypochilus: some of the two or three spigots similar to modified PLS spigots could be accompanying spigots instead (scored 01). Filistata: The two paracribellars might be homologous to the ones in the triad of other cribellate spiders. Note the presence of paracribellar spigots in male (scored 01). Titanoeca: no modified PLS spigot (scored ?). Homalonychus: male only one nubbin (scored 0).

300. PLS modified spigot and accompanying spigot shafts on common base: 0. PLS modified spigot stands alone. 1. One shaft on same base as modified spigot (fig. 135J). This character is applicable only when accompanying spigots are present. It includes character 102 of Griswold et al. (2005), but has been extended to other accompanying spigot morphologies, not only paracribellars. Comments: Neoramia: nubbins separate (scored 0). Zorocrates: present in the female, but nubbins well separated in the male (scored 01).

301. PLS spigots flanking the modified spigots, reduction in female: 0. Flanking spigots (fig. 135H). 1. Flanking nubbins (fig. 135I). Comments: Thaida: one nubbin, one paracribellar spigot (scored 01). Zorocrates: right side with nubbin, left side with spigot (scored 01).

302. Kind of spigots accompanying the PLS modified spigots: 0. Paracribellars (fig. 135A). 1. Similar to aciniforms (fig. 135C). 2. Aggregates (fig. 135F). Comments: Thaida: one paracribellar and a nubbin (scored 0). Megadictyna, Dictyna: one paracribellar (scored 0).

303. Anal tubercle size: 0. Small. 1. Very large (Griswold et al., 2005: fig. 27A). Oecobiids have a large anal tubercle fringed with setae. Comments: Austrachelas: there are two distinct holes anterior to the anal tubercle (scored 0).

Male Palpal Femur Modifications:

The ventral side of the femur may have one or more processes, of which the most frequently found are the ventral basal (fig. 140H), the ventral median (fig. 140I), and the ventral apical (fig. 140E) processes.

Male Palpal Patella Modifications:

The retrolateral side of the patella may have a process (fig. 140G).

Male Palpal Tibia Modifications:

A large clade of spiders have a retrolateral process (retrolateral tibial apophysis, RTA) on the tibia (fig. 140B), hence its name, the RTA clade. There may be other processes in addition to the RTA, such as the dorsal basal (fig. 140A), and the ventral apical (fig. 140D).

Male Palpal Cymbium and Its Modifications:

The tarsus of the male palp is modified to accommodate the copulatory bulb, and is called the cymbium. In entelegyne spiders the cymbium has a central depression, the alveolus, accommodating the spirally folded basal hematodocha (fig. 140F). The cymbium may have grooves, usually interacting with the embolus during copulation or in resting positions. Examples of such grooves are the apical cymbial groove (fig. 140F) (“cymbial conductor” of anyphaenids, Ramírez, 2003), and the retrolateral cymbial groove found in some miturgids (fig. 140B). Among the processes that usually occur on the cymbium, the most frequently found are two on the retrolateral margin of the alveolus: the retrobasal process, called the paracymbium, and the retromedial process; both processes can occur at the same time (fig. 140C). The dorsal surface of the cymbium has many chemosensory setae, which were found to be very sensitive to the pheromones released by females with their draglines (Tietjen and Rovner, 1980). The chemosensory setae may form a well-defined patch (fig. 140C).

Copulatory Bulb:

The copulatory bulb has one or more sclerotized pieces called sclerites, and an internal blind duct, the spermophore (fig. 139A–C), discharging at the tip of the intromittent structure, the embolus. The copulatory bulb is not connected to the testis, hence the male must charge his bulbs by absorbing a drop of sperm through the embolus. The copulatory bulb is attached to the tarsus by a movable membrane, the basal hematodocha (fig. 140F). In primitive spiders the movements of the bulb are controlled by a pair of muscles, but in entelegynes the movement is entirely hydraulic, with the basal hematodocha greatly developed, and producing a significant rotation of the bulb as it inflates (Huber, 2004). In most spiders the copulatory bulb is formed by several sclerites articulated to each other via flexible or inflatable hematodochae (fig. 139C), but the sclerites may be fused in many different ways. For example, in many Haplogynae all the sclerites are fused in a single piriform bulb (fig. 139A). The most conserved sclerites define the main divisions or sections of the copulatory bulb.

Basal Division of Copulatory Bulb:

The basal division spans between the basal and median hematodochae. It includes the subtegulum, and within it, the blind end of the spermophore, called the fundus.

Median Division of Copulatory Bulb:

The median division spans between the median and terminal hematodochae. It includes the tegulum, and may have a conductor and median apophysis attached to it. The spermophore runs through the tegulum.

Embolic Division of Copulatory Bulb:

The embolic division spans all structures distal to the median division, articulated to the tegulum by a terminal hematodocha. It includes the intromittent structure with the sperm outlet, called the embolus. In some groups there is an intermediate sclerite, the radix, between the embolus and tegulum, with the spermophore passing through it.

304. Endites sexually dimorphic: 0. Not dimorphic. 1. Male endites with distal ectal projection (Ramírez, 2003: fig. 96B). 2. Male endites basally globose (compare fig. 164C, D). 3. Male endites with ectal-anterior concavity (fig. 163H). Comments: Cybaeodamus: male endites more markedly depressed (scored 0). Doliomalus: just thinner in male (scored 0). Plexippus: not protruding (scored 0).

305. Male palpal femur ventral basal apophysis: 0. Absent (fig. 160G, 161E, 162E). 1. Present (fig. 140H; Opell, 1979: fig. 7). This is character 3 in Scharff and Coddington (1997). Comments: Uloborus: one large, one small, not considered homologous with the median or apical processes (scored 1).

306. Male palpal femur ventral median apophysis: 0. Absent. 1. Present (figs. 140I, 152F, 164H). Comments: Paccius: only a faint longitudinal ridge (scored 0). Jacaena: only a swelling (scored 0). Galianoella: the apical ridge-groove (scored 1).

307. Male palp femur ventral apical apophysis: 0. Absent. 1. Present (fig. 165F). The apical apophysis can cooccur with the median one (figs. 164B, 165E). Comments: Psechrus: with thicker setae (scored 0). Trachelas minor: an apical lip fitting the patellar apophysis may be homologous with the apical apophysis (fig. 140G) (scored 01). Drassinella: simple process united to the median one by a ridge (scored 1). Orthobula: hook (scored 1).

308. Male palp femur ventral longitudinal groove: 0. Absent. 1. Present (fig. 163F).

309. Male palp patella dorsal apophysis: 0. Absent. 1. Present. Comments: Anyphaena: shallow apical sclerotization (scored 01).

310. Male palp patella retrolateral apophysis: 0. Absent. 1. Present (figs. 140G, 157D, 163G). Comments: Clubiona: absent in Clubiona pallidula, but present in other species (e.g., Clubiona cf. maritima) (scored 0).

311. Male palp retrolateral tibial apophysis (RTA): 0. Absent (figs. 151C, 158D). 1. Present (figs. 161F, 166F). Here the broad homology of RTA includes the dorsal apical processes. In this study it was not possible to discern separate homology correspondences for a dorsal apical tibial apophysis, an RTA displaced to a dorsal position, or a dorsal subprocesses of a complex RTA. This resulted in the optimization of the RTA down to include titanoecids. Comments: Eresus: ambiguous (scored 01). Araneus: A ventral rounded protuberance. Because of the rotation of the palp, the homology is provisional (scored 0). Mimetus: because of the rotation of the palp, the homology is provisional (scored 0). Calacadia: only the ventral ledge plus the basal knob (scored 1). Cryptothele: male palp from Benoit (1978) (scored 1). Toxopsiella: similar to that of Cycloctenus, although in dorsal position (scored 1). Elaver: apparently moved dorsal (scored 1). Trachelopachys: should be the dorsal one (scored 1). Paccius: the dorsal branch only, the more ventral with canal is a modified seta (fig. 163C, D) (scored 1). Cf. Liocranidae LIB: thin spine as long as the cymbium, in dorsal position (scored 1). Hortipes: a trichobothria on the RTA! (fig. 162B) (scored 0). Lygromma: two processes (scored 1). Legendrena: if only one dorsal, then RTA (scored 1). Odo bruchi: a strong macroseta coming out of a retrolateral basal mound (scored 0). Hovops: palp provisionally scored from images by José Corronca (in litt.), from several species (scored 1). Epidius: in ventral-retrolateral position (scored 1). Thomisus: several setae with extremely enlarged, dark, protruding sockets (scored 1). Aphantochilus: tibia extremely short (scored 1).

312. RTA position: 0. Apical (fig. 166F). 1. Medial or basal (fig. 151A). Comments: Neoramia: complex, apical and median apophyses (scored 01). Metaltella: both (scored 01). Calacadia: both (scored 01). Macrobunus: apical and median branches (scored 01). Storenomorpha: all sides of tarsus (scored 01). Senoculus: basal (scored 1). Procopius: one apical, one basal (scored 01). Phrurolithus: all retrolateral faces (scored 01). Phrurotimpus, Otacilia: all retrolateral faces (scored 01). Orthobula: tibia very short (scored 0). Hortipes: tibia short (scored 0). Teutamus: two branches (scored 01). Jacaena: two branches (scored 01). Sesieutes: both, similar as in Jacaena (scored 01). Cf. Gnaphosoidea TEX: short tibia (scored 01). Xiruana: both (scored 01). Paravulsor: two branches, one more basal (scored 01).

313. RTA articulation: 0. RTA fixed, base sclerotized. 1. RTA articulate through membranous insertion. See comments under character 317. Comments: Titanoeca: it expands partially in KOH (scored 1).

314. RTA apical internal file: 0. Absent. 1. Present (figs. 143C, 151B, F, 152G). Lehtinen (2003: figs. 23, 25) also found files in the RTA of species of Runcinia, Misumenops, and Henricksenia. Comments: Dictyna: well spaced, irregular striations (Griswold et al., 2005 176E) (scored 01). Senoculus: thick striations (scored 1). Neato: not seen with stereomicroscope (scored 0). Corinna: a reticulate surface on one of the branches (scored 0). Brachyphaea: entire surface with longitudinal ridges, including the internal side (fig. 157B, C) (scored 1). Camillina, Apodrassodes: a few apical ridges (scored 01). Lamponella, Pseudolampona: observed in clove oil and compound microscope (scored 0). Strotarchus: irregular grains (scored 0). Ciniflella BRA: file more extended than in Thomisidae (scored 1). Polybetes: aligned cusps (scored 1). Epidius: not found with compound microscope (scored ?). Aphantochilus: internal side of RTA not examined (scored ?).

315. RTA sclerotization: 0. All sclerotized. 1. With membranous area (figs. 145A, 163A, B). Comments: Desis: canal membranous (scored 1). Macrobunus: central, dorsal, and ventral unsclerotized areas (scored 1). Brachyphaea: at the base, also surrounding all ventral processes (scored 1). Paccius: around the base of the modified seta (scored 1). Phrurolithus: internal membranous patch (scored 1). Jacaena, Teutamus: canal not sclerotized (scored 1). Miturga cf. lineata: with a ventral lobe not sclerotized (scored 1). Teminius: the canal area is membranous (scored 1). Systaria: internal side opposing cymbium, canal area membranous (scored 1). Zora: checked with stereomicroscope (scored 0). Tibellus: ventral membranous ridge extending into apophysis (scored 01). Sparianthinae VEN: between the pilose lobe and the complex apophysis (scored 1). Titanebo: membranous area extending into ventral branch of apophysis (scored 01).

316. RTA with canal: 0. Canal absent. 1. Canal present (figs. 142E, 146B, 160E, H). Comments: Toxopsiella: canal not well marked (scored 01). Brachyphaea, Mandaneta: canal on ventral hook (scored 0). Sesieutes: similar morphology as in Jacaena, but canal very shallow (fig. 160F) (scored 0). Paccius: canal on the modified seta (scored 0). Zora: canal not well defined, but very similar in general shape to the RTA of Mituliodon (scored 01). Selenops: dorsal side of major RTA branch with incomplete canal (scored 01). Anyphops: a wide canal but not as in miturgids (scored 0).

317. Male palpal tibia gland: 0. Absent. 1. Present, discharging through a dorsal or retrolateral tibial apophysis. Described in Compagnucci and Ramírez (2000: 203) for macrobunine amaurobiids, and Wanless (1979, 1984, 1987) for a few genera of spartaeine salticids. The apophysis may be fixed or movable, with an articulated base. None of the amaurobiids, salticids, or any other terminal in this analysis bears a tibial gland discharging through tibial apophyses. See also character 313. Comments: cf. Liocranidae LIB: not seen after clarification, RTA tip not perforated (scored 0).

318. Male palp tibia dorsal basal process: 0. Absent. 1. Present (figs. 140A, 144C). Here only the basal dorsal process is scored separately from the RTA. Other more apical dorsal processes are considered part of the RTA (see char. 311). The process identified by Harvey (1995) as an RTA displaced dorsally (in Nicodamus and Megadictyna) is here considered a basal dorsal process. Scoring the dorsal basal process in nicodamids as an RTA only changes the tree by making Nicodamidae paraphyletic, and tracing the RTA origin back to the common ancestor of nicodamids and the RTA clade. Comments: Storenomorpha, Galianoella: the tibia is too short to decide between apical and proximal (scored 01).

319. Male palp tibia dorsal basal process conformation: 0. Simple or absent (fig. 144C). 1. Complex (fig. 140A). In this dataset the complex condition is present only in the nicodamids Nicodamus and Megadictyna, and they look very similar to each other (scored 1).

320. Male palp tibia ventral apical apophysis: 0. Absent, simple swelling, or part of RTA complex. 1. Present, well defined (figs. 142H, 147B, 152A, C, G, 157A, B, E). The broad homology of the ventral apical apophysis is applied only to well-defined processes, separated from the RTA. Several instances of ventrally swollen tibiae were not considered as apophyses (see comments below). The hook-shaped ventral process of the thomisid Boliscus fits on a tegular ridge (fig. 152G). Huber (1995a) showed that in Misumenops tricuspidatus the ventral hook guides the rotation of the tegulum produced by hematodochal expansion. The general correspondence between the ventral apophysis and the ridges in the tegulum suggest that this function may be widespread in higher thomisids. Comments: Araneus: A ventral rounded protuberance. Because of the rotation of the palp, the homology is provisional (scored 0). Neoramia: only an inflated margin (scored 0). Stiphidion, Calacadia, Desis, Badumna, Metaltella: the straight ledge forming a canal is considered part of the RTA complex, as in Griswold et al. (2005) Oxyopes: a hook like those of thomisids (fig. 142H) (scored 1). Cf. Medmassa THA: only a round elevation close to the RTA base (scored 0). Castianeira: a longitudinal ridge (scored 0). Trachelopachys: apical retrolateral-ventral (scored 0). Pseudocorinna: prolateral-ventral simple prong (scored 1). Paccius: the more ventral process is a modified seta (fig. 163D) (scored 0). Brachyphaea: a ventral hook with a canal (fig. 157B) (scored 1). Procopius: the ventral-retrolateral process interpreted as an RTA; there is also a ventral-basal process (scored 0). Mandaneta: a flat projection and a small hook with canal and reticulate texture (scored 1). Neoanagraphis: just raised articulation margin (scored 0). Phrurolithus: extensive membranous articulation with swelling border (scored 0). Legendrena: all ventral side bulbous, similar as in Phrurolithus (scored 0). Ammoxenus: ventral and prolateral margin extended in a cup, receiving the copulatory bulb (scored 0). Cheiramiona: similar as in Eutichurus, but the articulating membrane arises just on the border (scored 01). Eutichurus: flat process, not hooked (scored 1). Systaria: swelling all around the tibia (scored 0). Raecius: interpreted as a ventral branch of the RTA (scored 0). Philodromus: P. californicus similar as in thomisids, P. aureolus broader process; perhaps with conductor function (scored 1). Geraesta: reduced but evident (scored 01). Strophius, Aphantochilus, Tmarus: retrolateral hook (scored 1). Epidius: Interpreted as an RTA. The apical ventral macrosetae are reminiscent of those found in Cebrenninus at base of RTA, but the position is different (scored 0). Boliscus: hook shaped, fitting ridge in tegulum (fig. 152G) (scored 1).

321. Male palp tarsus muscle M29: 0. Present. 1. Absent. See Huber (2004) and Griswold et al. (2005: char. 128). In this dataset M29 is expected to have the same distribution as M30 (Griswold et al., 2005: char. 129); differences were reported in Hersiliidae and Uroecobius by Huber (1994, 2004). Comments: Hypochilus: the M29 partly originates in the patella (Huber, 2004) (scored 0). Filistata: from Huber (2004), after Kukulcania hibernalis. Eresus, Oecobius, Homalonychus: from Huber (1994) (scored 1). Uloborus, Araneus, Huttonia, Dictyna: after Bernhard Huber (in litt.) in Griswold et al. (2005: char. 128). Nicodamus: some unspecified Nicodamidae examined by Huber (1994) (scored ?). Miturga cf. lineata: not observed, but absent in an unspecified Miturgidae (Huber, 1994) (scored ?).

322. Orientation of cymbium relative to bulb: 0. Dorsal. This is the general condition in the RTA clade (fig. 147A). 1. Mesal. Araneids have the copulatory bulb facing outward, with a mesal cymbium (fig. 141F) (Griswold et al., 1998: char. 2). 2. Basal. Hypochilids and filistatids have the copulatory bulb apical in the male palp (fig. 141A, E). Comments: Uloborus, Orthobula, Trachelidae ARG, Hortipes: intermediate (scored 01).

323. Cymbial tip ventral groove: 0. Absent (fig. 153C). 1. Present. The cymbium has a ventral apical smooth groove that usually holds the embolus in a resting position (figs. 156B, 158F). This character was proposed as a synapomorphy for Anyphaenidae except Malenellinae (Ramírez, 1995, 2003). In this dataset it also occurs elsewhere, including as a synapomorphy for Castianeirinae. Comments: Filistata: no cymbial tip (scored ?). Mimetus: very modified (scored 01). Huttonia: very wide, more evident in other species (Forster and Platnick, 1984: fig. 277) (scored 1). Storenomorpha: Also with canaliculate hairs, see Jocqué (1991: figs. 39, 40). He suggested that those hairs might produce the mating plug (scored 1). Zoropsis, Pseudoctenus: ambiguous, border too short (scored 01). Cf. Medmassa THA: only concave and devoid of hairs (scored 0). Castianeira: not so markedly notched (scored 1). Ammoxenus: very wide notch (scored 01). Uliodon, Liocranoides, Austrachelas: slightly notched (scored 01). Xiruana: glabrous, sclerotized groove (scored 1). Ciniflella BRA: perhaps a soft fold (scored 0). Eusparassus: groove with setae (scored 1). Hispo: perhaps slightly notched (scored 01).

324. Cymbium dorsal chemosensory patch: 0. Absent, chemosensory setae sparsely distributed (fig. 140B). 1. Present, chemosensory setae in a dense patch (figs. 142C, 149C, 155E, 162F). Often described as a “cymbial dorsal scopula,” this patch is formed by chemosensory setae (Griswold et al., 2005: char. 113). As can be reconstructed by the comments below, the chemosensory patch occurs scattered in many families, often with poorly defined limits. Comments: Acanthoctenus, Vulsor, Ctenus, Falconina, Paradiestus: borders not well defined (scored 1). Medmassa: not well defined, dorsodistal transverse bands (fig. 158B) (scored 0). Paccius: dense, not well-defined borders (scored 1). Procopius: borders not defined (scored 0). Apostenus: not dense but definite (scored 1). Cf. Liocranidae LIB: prolateral to a cymbial dorsal canal fitting the RTA (scored 1). Toxoniella, Gayenna, Macerio, Miturga cf. lineata, Systaria: not well delimited (scored 0). Phrurotimpus: larger blunt apical seta (scored 0). Malenella: the blunt setae (Ramírez, 1995) are chemosensory, similar to those below the claws (scored 1). Cheiracanthium: present but not well defined (scored 1). Mituliodon: a few sparse, thick, short setae (scored 0). Strotarchus: not well defined, but many setae present (scored 1). Zora: a bunch of thick setae (scored 0). Xenoctenus: tenent setae (scored 0). Liocranoides: small patch (scored 1). Titanebo: distal ring (scored 0).

325. Cymbial apex extension beyond alveolus: 0. Extending beyond distal margin of alveolus (fig. 148B). 1. Short, wide, not extending beyond distal margin of alveolus (figs. 142B, 143E, 144A). Comments: Hypochilus, Filistata, Ariadna: no alveolus (scored ?). Araneus: flat, wide (scored 1). Huttonia: very small alveolus (scored 01). Eriauchenius: reduced cymbium (scored 01). Apostenus, Ciniflella ARG, Boliscus, Thomisus: intermediate (scored 01). Raecius: not in R. asper (scored 1).

326. Cymbial apical ventral setae: 0. Sparse, regular setae (figs. 147E, 153C). 1. Bunch of thick setae. These setae may be ridged (fig. 165B, C), or dark and erect (figs. 147D, 148B, C, E, F). A bunch of thicker or darker setae may occur just distad of the apical margin of the cymbial alveolus. Comments: Donuea: similar as in Phrurolithidae, but thin hairs (scored 0). Trachelas mexicanus, Strotarchus, Ciniflella BRA, Polybetes: some thick setae at the tip of cymbium (scored 0). Micaria: three macrosetae (scored 0). Prodidomus: some setae slightly similar as in Phrurolithidae, with slim base, slightly expanded, flattened apically (scored 0). Neozimiris: two thick setae toward prolateral side; Zimiris has several of these setae in similar position to that in Phrurolithidae (Platnick and Penney, 2004: figs. 4, 21) (scored 1). Cheiramiona: a bunch of long setae, but not as thick as in Phrurolithidae (scored 1). Stephanopoides: a few bottle-shaped chemosensory setae (scored 0).

327. Cymbial tip apical thick setae: 0. Absent. 1. Present (figs. 150D, 165D). A few terminals have only one thick apical seta (figs. 149E, 165D). Comments: Cryptothele: not only apical (scored 1). Ctenus: some of the rakelike setae with very short thin tip (scored 01). Pronophaea: many short blunt setae (fig. 157G, H) (scored 1). Liocranum: intermediate (scored 01). Ammoxenus: there are macrosetae, but all over the cymbium (scored 01). Phrurotimpus: one, blunt (fig. 165D) (scored 1). Otacilia: at least one blunt, a second one may be broken (scored 1). Cf. Eutichuridae QLD, Eutichuridae MAD: one (fig. 149E) (scored 1). Syspira: as in most miturgids, with thick short setae, not so thick as in Zora (scored 1). Xenoctenus: mixed with the other setae (scored 1). Griswoldia: very thick (scored 1). Heteropoda: thick apical setae finely barbed, coordinately with cheliceral rake setae (scored 1). Aphantochilus: short macrosetae (scored 1).

328. Cymbial tip horn: 0. Absent, cymbial tip rounded (figs. 158F, 165D). 1. Present, cymbial tip ending in a cone (fig. 142J). In this dataset the horn is only present in Senoculus and Thaida.

329. Cymbial trichobothria: 0. Absent. 1. Present (figs. 160C, 166C).

330. Cymbial trichobothria rows: 0. One row, or single. 1. Several in two rows (fig. 160C, D). A potential synapomorphy of a group of South East Asian liocranids (Teutamus, Sesieutes, Jacaena), with multiple cymbial trichobothria in two rows. Comments: Badumna, Geraesta: two in longitudinal line (scored 0). Stiphidion, Metaltella: one row (scored 0). Calacadia: one row proximally, much widened at the end (scored 01). Medmassa: marginal lines at each side (scored 1). Teutamus: five, dispersed, perhaps two irregular rows (scored 1). Sesieutes: eight, in two rows (scored 1). Jacaena: eight, in two rows (fig. 160C) (scored 1). Cf. Gnaphosoidea TEX: several, in more than one row, the images are not clear if in definite rows (scored 1). Galianoella: at least two rows, the median one with several trichobothria, seen in stereomicroscope (scored 1).

331. Retrolateral cymbial groove: 0. Absent. 1. Present (figs. 145F, 148D, 162A, B). Comments: Pimus: small groove just at the margin, ventral (scored 0). Mandaneta: the bulb is swelling over a surface that looks like a flattened groove, with small setae on it (scored 01). Neoanagraphis: margin modified (scored 0). Oedignatha: absent in the species scored (fig. 161F), but present in O. scrobiculata (scans by R. Raven, in litt.) (scored 0). Anagraphis: widened, ventral, not on lateral pilose area (scored 0). Ammoxenus: longitudinal depression fitting large RTA (scored 01). Cf. Eutichuridae QLD, Miturga cf. lineata, Miturgidae QLD: only basal, superficial (fig. 149A) (scored 01). Systaria: fitting the RTA (scored 1).

332. Cymbial groove setae thickness: 0. Thin or absent (fig. 148D). 1. Thick setae (fig. 146B). Comments: Hortipes: only thin setae (scored 0). Oedignatha: only thin setae, more aligned in O. scrobiculata (scans by R. Raven, in litt.) (scored 0). Hortipes: only thin setae (scored 0). Miturga cf. lineata: a few macrosetae, close to the apophysis (scored 01). Teminius: thick setae (scored 1).

333. Cymbial groove posterior extension: 0. Not extending beyond articulation with tibia (figs. 146B, 148D). 1. Extending in acute conductorlike process (figs. 147E, F, 149B–D). The morphology of the extended groove suggests that it may play a role in leading the embolus during expansion. Comments: Oedignatha: present in O. scrobiculata (scans by R. Raven, in litt.) (scored ?). Cf. Eutichuridae QLD, Cheiramiona: a short cusp in that place (scored 0).

334. Cymbial retrobasal file: 0. Absent. 1. Present (Griswold et al., 2005: 12, fig. 183). See Griswold et al. (2005: 12). This is presumably a synapomorphy of Macrobuninae.

335. Cymbial retrobasal process (including the paracymbium): 0. Absent. 1. Present (figs. 140C, 142D, 159F, 165A). Here the retrobasal process arises independently in araneoids (Araneus + Mimetus), and several other groups of the RTA clade, including dionychans. For a discussion, see Griswold et al. (1998: 31). Comments: Hypochilus: very different in morphology from other terminals, and separated from the alveolus (scored 1). Megadictyna, Titanoeca, Clubiona, Castianeira: excavated area (scored 0). Nicodamus: excavated area plus projecting border (scored 01). Cyrioctea: excavated area plus median projecting lobe (scored 01). Cybaeodamus: with basal membranous area inflated on KOH expansion (scored 1). Zoropsis, Lauricius: inconspicuous projecting base (fig. 143D) (scored 01). Toxopsiella: excavated area plus shallow projection (scored 01). Creugas: retrobasal process complex, interacting with the RTA (scored 1). Falconina: retrobasal process is complex, interacting with the RTA (scored 1). Copa: retrobasal complex slightly protruding, similar to that in Corinna (scored 01). Mandaneta: small, hidden by bulb (scored 0). Apostenus: very short lobe (scored 0). Liocranum: short lobe (scored 01). Hortipes: a prolongation of the cymbial groove (scored 1). Oedignatha: Excavated, just a short projection. Present in O. scrobiculata (scans by R. Raven, in litt.) (scored 0). Micaria: protruding border (scored 0). Centrothele: a process fitting the RTA is more dorsal than the paracymbial processes of other terminals (scored 0). Austrachelas: hole matching RTA (scored 0). Neato: excavated area fitting RTA and ventral low pilose mound (scored 0). Trachycosmus: excavated plus median bump (scored 01). Fissarena: excavated plus median lobe (scored 01). Platyoides: a large glabrous area and a small hole in front of the RTA (scored 0). Ammoxenus: shallow lobe (scored 0). Cf. Eutichuridae QLD: small mounds (scored 1). Cheiracanthium: the hook where the cymbial extends (scored 1). Cheiramiona: small conic process (scored 1). Uliodon: retrobasal process similar to that in Corinna (scored 1). Zorocrates: short projecting border plus retrolateral-dorsal bump (scored 01). Selenops: depressed area with shallow projection (scored 01). Hovops: other species may have, but not this one (scored 0). Sparianthinae VEN: excavated plus median bump (scored 01). Heteropoda: small knob (fig. 154E) (scored 01).

336. Cymbial retromedian process: 0. Absent. 1. Present, without a furrow (figs. 140C, 143D, 151E, 154A). 2. Present, forming a canal, conductorlike (includes the thomisid tutaculum) (fig. 152D, E). States are ordered. A few terminals have both a retrobasal and a retromesal processes (fig. 140C). Comments: Nicodamus, Pimus: more dorsal (scored 0). Cryptothele: a simple mound, illustrated by Benoit (1978: fig. 3C) (scored 01). Paccius: there is a longitudinal furrow basal to the process, scored as a retrolateral cymbial groove (fig. 163B) (scored 1). Agroeca: very shallow mound (scored 01). Trachelopachys: shallow lobe, absent in other congeners (scored 01). Gnaphosa: intermediate, absent in other species (scored 0). Apodrassodes: a shallow mound (scored 1). Meedo, Fissarena: only a protuberance (scored 1). Miturga gilva: present, at the end of the groove (scored 1). Aphantochilus: intermediate, very thin (scored 12). Plexippus: shallow pilose lobe (scored 01).

337. Cymbium dorsobasal modifications: 0. Absent. 1. Projection (figs. 143B, 155D). 2. Transverse furrow (fig. 142I). Comments: Cyrioctea: very slightly raised, continued from retrobasal concavity (scored 01). Storenomorpha, Creugas: a transverse furrow (scored 2). Homalonychus: funny protuberance fitting the RTA (scored 1). Oxyopes: a peculiar furrow fitting the dorsal tibial apophysis (scored 2). Clubiona: Clubiona cf. maritima with dorsal cymbial projection (scored 0). Cocalodes: two short macrosetae (scored 0).

338. Cymbium-tegulum fusion: 0. Free. 1. Fused. Some prithine filistatids have the cymbium partially fused to the tegulum (Gray, 1995; Ramírez and Grismado, 1997), and many Oonopidae show several degrees of fusion (e.g., Platnick and Dupérré, 2010). The fusion does not occur in this dataset.

339. Subtegulum transverse, distally crossing a piriform bulb: 0. Other bulb conformations. 1. Subtegulum transverse, crossing, visible from both sides (figs. 158D, 159A, 164F). The apical embolar section of some phrurolithids and castianeirines is delimited by a transversely placed subtegulum, visible from both sides. Comments: Creugas: subtegulum L-shaped, with partially subdivided, more basal branch (scored 0). Copa: subtegulum mostly subdivided into two pieces (scored 1). Medmassa: intermediate (scored 01). Otacilia: intermediate between crossing and an apical cup (scored 01). Eusparassus: subtegulum mostly hidden, just visible at the center, through the loop of the embolus (scored 0). Plexippus: subtegulum a thin ring (scored 0).

340. Subtegulum-tegulum fusion: 0. Separate (figs. 139C, 141B). 1. Fused (fig. 141C, D). Comments: Hypochilus: a thick articulation (scored 0). Filistata: In Kukulcania separated on one side, fused on the other (see also Huber, 2004: fig. 10). Interpreted as fused (scored 1). Thaida: I cannot see two separate sections (see also Huber, 2004: fig. 11) (scored 1). Eresus: a thick articulation (scored 0). Huttonia: the hematodocha commented in Forster and Platnick (1984) is the basal one (scored 1). Donuea: interpretation of palp very difficult, tentative (scored 0). Hispo: it seems fused, but the tegulum is partially membranous (scored 01).

341. Subtegular locking lobe: 0. Absent (fig. 153B). 1. Present (figs. 143A, B, 156D, 164G). The subtegulum has a lobe, often fitting an opposing lobe on the embolar base. This character was originally proposed for some lycosoids by Griswold (1993), where the tegulum and subtegulum have lobes interlocking in the bulb when in resting position. The system was later reported in other families, and it was also found that the tegular lobe is the embolar base when the embolus is articulated (Griswold et al., 2005). Other locking mechanisms have been described in Theridiidae (median apophysis-cymbium; Agnarsson, 2004) and Nicodamidae (Griswold et al., 2005). In this dataset several terminals have an internal locking mechanism, exposed only after bulb expansion (e.g., in Olbus, Ramírez et al., 2001: fig. 47). In cycloctenids the subtegulum has a hole instead of a lobe, fitting the embolar lobe (fig. 142F). Comments: Oecobius: see Baum (1972: fig. 62), it looks as if it has both tegular and subtegular locking lobes (scored 0). Nicodamus: plus a tegular lobe of another kind (scored 1). Badumna: absent, but there seems to be some internal ridges (scored 0). Homalonychus: subtegular lobe more like a hole (scored 1). Vulsor: on retrolateral side, entire bulb rotated (scored 1). Ctenus: similarly as in Xenoctenus, but hidden by the tegulum (scored 1). Oxyopes: at the base of the embolus, embolar lobe (scored 0). Aglaoctenus: the tegular notch harbors most of the subtegulum (scored 0). Cycloctenus: very similar to that of Toxopsiella, including hole in subtegulum (scored 0). Toxopsiella: impressive embolar lobe fitting in subtegular hole (fig. 142F) (scored 0). Paccius: very well defined, retrolateral (scored 1). Mandaneta: subtegular lobe fitting in a hole just below the median apophysis (scored 1). Olbus: internal locking, Ramírez et al. (2001: fig. 47) (scored 0). Agroeca: lobes tightly coupled (scored 1). Phrurolithus: embolar base fits lobe in subtegulum (fig. 164G) (scored 1). Lampona: not even internal locking (scored 0). Centrothele: there may be an internal locking (scored 0). Meedo: locking lobes visible in retrolateral view (scored 1). Trachycosmus, Fissarena: with internal locking mechanism with lobes on both tegulum and subtegulum, not close to the base of embolus (scored 0). Desognaphosa: a small locking lobe near embolus base (scored 1). Teminius: opposing embolar base (scored 1). Syspira: retrolateral-basal (scored 1). Xenoctenus: basal (scored 1). Uliodon: matching embolar lobe (scored 1). Ciniflella BRA: internal locking (scored 0). Anyphops: Subtegular lobe plus hole. There are additional tegular-subtegular locking lobes on the other side, as in Nicodamus (scored 1). Tmarus: as in Nicodamus, both locking lobes not related with embolar base (scored 0). Holcolaetis: subtegular lobe just shallow depression (scored 01). Portia: base of embolus fitting shallow depression and lobe in subtegulum (scored 01). Galianoella: plus a dorsal lobe locking on the cymbium! (scored 1).

342. Tegular (embolar base) locking lobe: 0. Absent (fig. 153B). 1. Present. The embolar base has a lobe, often fitting an opposing lobe on subtegulum (fig. 143A, B). See Griswold et al. (2005: char. 116). Comments: Nicodamus: with a lobe arising close to the beginning of sperm duct, not at the base of embolus (scored 0). Metaltella, Calacadia: the anterior sclerotized hook, unmatched (scored 01). Macrobunus: the tegular lobe is ventral to the embolus, considered as an embolar process (scored 0). Homalonychus: lobe at base of embolus. Zoropsis: embolar lobe (scored 0). Pseudoctenus: embolar base lobe conspicuously protruding (scored 0). Trachelas mexicanus: a projecting triangle before embolar base (scored 0). Paccius: very well defined, retrolateral (scored 1). Agroeca: lobes tightly coupled (scored 1). Phrurolithus: embolar base fits lobe in subtegulum (scored 1). Xenoplectus: The tegular lobe is on a partially separated part where the embolus arises. It might be part of the embolar base (scored 1). Cf. Moreno ARG: embolar base (scored 1). Vectius: there is a locking lobe, but not clear if part of the embolus (scored ?). Fissarena: a tegular lobe, but not so close to the embolus insertion (scored 01). Miturgidae QLD: the base of the embolus is protruding, although not locking anything (scored 01). Miturga gilva: there is a lobe there, but not matching the subtegular one (scored 0). Syspira: embolus free, an irregular separate lobe from tegulum (not embolar base) in front of that of subtegulum (scored 0). Uliodon, Raecius: the base of the embolus (scored 1). Xenoctenus: basal, tegular lobe at the base of long embolus (scored 1). Liocranoides: large tegular lobe at embolus base (scored 1). Sparianthinae VEN: a not well-defined tegular lobe at base of embolus (scored 01). Lauricius: embolar lobe, because embolus is free (scored 1). Holcolaetis: basal lobe (scored 1).

343. Tegular distal division at embolar base: 0. Absent. 1. Present. In Xenoctenus and related genera the tegular region where the embolus arises is delimited from the rest of the tegulum by a membranous area, and extends forward. This division may be small (fig. 150A) or very large, with a furrow leading the embolus (fig. 150B, C, E). This structure was proposed by Silva Davila (2003: fig. 18, char. 38, “Odo-like” sclerotized tegular process) as a synapomorphy joining some Xenoctenus and Odo species. Because it is not clear whether this region belongs to the tegulum or is instead a basal embolar process, character 351 was scored as uncertain for the terminals having this structure (the Xenoctenus group: Xenoctenus, Odo bruchi, Paravulsor). In Lyssomanes the tegulum has a deep and narrow transverse furrow arising close to the embolus base (fig. 155A), thus delimiting a thin strip of tegulum somehow similar to the configuration found in the Xenoctenus group (scored 01) (see char. 344).

Fig. 18.

Chelicerae of female. A. Hypochilus pococki (Hypochilidae), cephalothorax and chelicerae. B. Hypochilus pococki (Hypochilidae), left, posterior view, arrow to median cheliceral concavity. C. Mimetus hesperus (Mimetidae), right, anterior view of promargin. D. Kukulcania hibernalis (Filistatidae) posterior view. E. Filistata insidiatrix (Filistatidae), left, anterior view of promargin. F. Same, posterior view of right fang and cheliceral lamina. G. Ariadna boesenbergi (Segestriidae) female, left chelicera, anterior view of promargin. H. Same, chelicerae ectal view. I. Same, posterior-mesal view.

344. Tegular notches: 0. None. 1. Amaurobioidinae-like. The tegulum has a deep basal indentation, occupied by the median hematodocha (fig. 156C). 2. Lycosidae-like. The tegulum has a basal indentation, occupied by the subtegulum (fig. 142G). 3. Basal transverse furrow. The tegulum has deep and narrow transverse furrow arising close to the embolus base (fig. 155A). Comments: Thaida: tegulum just a sclerotized band (scored ?). Oxyopes: most of the tegulum is membranous (scored 0). Strotarchus: the median hematodocha is well exposed at the base of the tegulum, in ventral view (scored ?). Liocranoides: large central membranous area extending posteriorly, mesal to trajectory of sperm duct (scored 0). Plexippus: unsclerotized area projecting basally (scored 0).

345. Proximal sperm duct constriction: 0. Absent. The proximal section of the sperm duct not constricted. 1. Present, near fundus. The proximal section of the sperm has a narrow constriction (Huber, 1994: fig. 1). According to Huber (1994) this is a potential synapomorphy of Oecobius and Uroctea. Comments: Eresus: from Huber (1994) (scored 0). Oecobius: from Huber (1994) (scored 1). Storenomorpha: too dark (scored ?). Ctenus: internal, not clarified (scored ?). Dolomedes: from Sierwald (1990) (scored 0).

346. Sperm duct distal thickness: 0. Gradually tapering, or thinned before embolus (fig. 159D). 1. Thick sclerotized apical bulb, described as “sclerotized area on distal reservoir” by Bonaldo (2000: figs. 90–Fig. 91.Fig. 92.Fig. 93.Fig. 94.Fig. 95.Fig. 96.Fig. 97.Fig. 98.Fig. 99.Fig. 100.Fig. 101.Fig. 102.Fig. 103.Fig. 104.105) (see also fig. 158E). Bonaldo (2000) proposed the distal thick sclerotization of the sperm duct as a synapomorphy of Corinninae, although his group has recently found that the structure occurs in most Castianeirinae as well (A. Bonaldo and D. Candiani, personal commun.); in this dataset it is found in Castianeira trilineata (fig. 158E). Comments: Corinna: the AER from C. ducke (Bonaldo, 2000) (scored 1). Falconina: from Bonaldo (2000: fig. 101) (scored 1). Paradiestus: from Bonaldo (2000) (scored 1). Phrurotimpus, Phrurolithus: sperm duct with thick walls, thin before embolus, similarly illustrated in P. difficilis by Wiehle (1967: fig. 76) (scored 0). Sesieutes: with apical widened section, but not thickly sclerotized (fig. 161B) (scored 01). Titanebo: sperm duct markedly sinuous (scored 0).

347. Sperm duct sclerotization: 0. Sclerotized, thick wall, at least in basal section (fig. 153D). 1. Membranous, thin wall. In this dataset the only terminal with a membranous sperm duct is Thaida (Austrochilidae, personal obs.). The sperm duct has lost its sclerotized walls also in derived oonopids (Platnick et al., 2012).

348. Sperm duct spiral meander in ventral tegulum: 0. Absent (figs. 139C, 158E). 1. Present (fig. 159E). Corinnines have a meander of the sperm duct describing a regular spiral on the ventral side of tegulum (Platnick and Baptista, 1995; Bonaldo, 2000). Comments: Titanoeca: complex, internal sperm duct (scored 0). Clubiona: similar as in corinnines, other species different, C. maritima fairly contorted (fig. 144B) (scored 01). Orthobula: there is a gland discharging close to embolus (scored 0). Uliodon: parts of sperm duct not touching tegulum walls, with voluminous tegular gland (scored 0). Hispo: sperm duct crosses internally through the tegulum (scored 0).

349. Embolus origin internal: 0. Exposed. 1. Internal to complex conductor (fig. 139D). See Griswold et al. (2005: char. 122). Comments: Titanoeca: the embolus arises between tegulum and cymbium, and has a very complex associated conductor, but the base is exposed (scored 0). Eilica: the embolus has a long, sclerotized base arising internally in a tegular-embolar section (scored 0).

350. Separate embolic division: 0. Absent, embolus in one sclerotized piece (fig. 153A). 1. Present, a sclerotized basal section as a cylinder containing the sperm duct, separated from the embolus by a membranous articulation (fig. 155B). This embolic division includes the araneoid radix and the “distal sclerotized tube of apical division,” which may be fully articulated as in Dolomedes (Sierwald, 1990). Comments: Xenoplectus: I interpreted the articulated piece where the embolus arises as part of the tegulum, but at the base of embolus there is a further partial division (fig. 162D) (scored 01). Camillina: there is a separate basal division, but not containing the sperm duct (scored 0). Eilica: the embolar base is telescoped inside this section, not a complete cylinder (scored 0). Gayenna: the embolus base is partially divided by an incomplete membranous strip (fig. 156D, F) (scored 01).

351. Embolus attachment: 0. Fixed (figs. 144E, 153E, F). 1. Flexibly attached (figs. 139B, 145G, 148A, 153A). Comments: Thaida: the subtegulum-tegulum is a continuous sclerotized band (scored 0). Eriauchenius: embolus base not exposed; Wood et al. (2012: fig. 9a) show dramatic distal expansion of the palpal bulb of Eriauchenius, exposing the embolus and other distal sclerites (scored ?). Titanoeca: complex, unclear (scored 1). Metaltella: not exposed, but arising from a membranous area (scored 1). Calacadia, Desis: internal, not exposed (scored ?). Oxyopes: fused to one of the two sclerotized remains of the tegulum (scored 0). Trachelas mexicanus: base of embolus with sclerotized rings (scored 01). Phrurotimpus: partially membranous areas surrounding embolar base (scored 01). Eilica: very large embolar base (scored 1). Cf. Moreno ARG: incomplete division (scored 01). Lampona: embolus arising from slightly unesclerotized area (scored 0). Austrachelas: anterior basal part is continuously sclerotized with tegulum (scored 01). Tengella: intermediate, weakly sclerotized articulation (scored 01). Xenoctenus: scored ambiguous, because the tegular distal division at embolar base might be an embolar projection instead (scored 01). Philodromus: with wide partial suture, although not movable (Huber, 1994) (scored 01). Titanebo: well-defined suture (scored 1). Petrichus: no suture at all (scored 0). Heteropoda: a meander in the sperm duct may indicate the place of fusion (scored 0). Plexippus: unsclerotized notch at embolus base (scored 01).

352. Embolar basal process: 0. Absent (figs. 140D, 145E). 1. Present. The base of embolus has a sclerotized process, continuous with the embolus (figs. 144A, 156E, 162D, 166D). Comments: cf. Liocranidae LIB: a lamellar complex process (scored 1). Camillina: flat wide piece (scored 0). Austrachelas: ectal embolar process (!) (scored 1). Platyoides: a small process at the tip (scored 0). Liocranoides: complex embolus, some of the folds could be interpreted as processes (scored 01). Raecius: the sclerotized tegular process 2 (STP2) in Griswold (2002: fig. 51) (scored 1). Odo bruchi: a very long and thin process (scored 1).

353. Embolus prolateral furrow: 0. Absent. 1. Present. Comments: Filistata: there is a short furrow (scored 01). Thaida, Agroeca, Lauricius: embolus very complex (scored 01). Donuea: thin furrow on complex embolus, unclear homology (scored 01). Trachelas minor, Castianeira, Copa: modified, screw shaped (scored 01). Jacaena: just a line (scored 0). Camillina: too modified (scored ?). Eutichurus: only apically (scored 01). Liocranoides: quite complex, but without furrow (scored 0). Ciniflella BRA: ventral-retrolateral furrow (scored 1).

354. Embolus screw shaped: 0. Absent, other shapes (fig. 158A). 1. Present. The embolus forms a tapering screw (figs. 158C, 159A, B, 163E), which usually corresponds to a complementary screw in the female copulatory openings (figs. 158G, 159C, 167D). Comments: Filistata: very slightly so (scored 01). Trachelas minor: very slightly screw shaped, female copulatory opening also very slightly screwed (scored 01). Trachelas mexicanus: screw in the opposite direction compared to castianeirines, provisionally interpreted as homologous (scored 1). Tibellus: only the tip of the embolus forms a well-defined screw (fig. 36G) (scored 0).

355. Embolus tip wide, truncate, opening on thin transverse tube: 0. Absent. 1. Present (fig. 144D, F). Comments: Eriauchenius: conformation too different (scored -). Apostenus: similarly as in Liocranum, a thin, short transverse ending of the sperm duct (scored 1). Cf. Liocranidae LIB: extensive, waving ribbon, but not tube (scored 0).

356. Median apophysis: 0. Present. The median apophysis typically is a hook-shaped, articulated sclerite, arising from the retrolateral side of the copulatory bulb (figs. 156A, G, 166B). 1. Absent (figs. 159A, 164A, 166A). Comments: Hypochilus: the structure labeled “MA?” in Coddington (1990: fig. 9) is the base of the embolus, and continues spiraling on the other side (scored 1). Thaida: the thin piece previously identified as embolus in Forster et al. (1987) here interpreted as in Griswold et al. (2005) (scored 0). Eresus, Megadictyna, Macrobunus: interpreted as in Griswold et al. (2005) (scored 1). Oecobius: the articulated sclerite (scored 0). Uloborus: interpreted as in Coddington (1990) and Griswold et al. (2005) (scored 0). Eriauchenius: interpreted as in Griswold et al. (2005) (scored 0). Nicodamus: interpreted as in Harvey (1995), apically membranous, may be a conductor instead, but there are other apophyses related to the embolus (scored 0). Psechrus: present in Fecenia (scored 1). Cyrioctea: median apophysis conductor-shaped, with small basal branch (scored 0). Pseudoctenus: median apophysis as a sclerotized hook plus posteriorly directed fleshy lobe (scored 0). Corinna: the Corinna tegular process (PTC) of Bonaldo (2000) may be interpreted as a median apophysis, but it is prolateral to the conductor (scored 01). Hortipes: interpreted as in Bosselaers and Jocqué (2000) (scored 0). Cf. Liocranidae LIB: two sclerotized articulate sclerites, the one closer to embolus base interpreted as a conductor (scored 0). Jacaena, Sesieutes: tentatively identified as the small whitish prolongation besides the conductor tip (scored 0). Oedignatha: tegular sclerite identified as median apophysis, but might be identified as a conductor instead (fig. 161G, H) (scored 0). Trachycosmus: two sclerites, one identified as the median apophysis, the other as the conductor (scored 1). Fissarena: both conductor and median apophysis guide the embolus (scored 0). Desognaphosa: Just a fleshy lobe. Homology different from Platnick (2002); the large sclerite leading the embolus arises just from the base of embolus, has granulose basal texture, and has a large canal to receive the embolus, it is identified as conductor (scored 1). Cithaeron: interpretation as in Platnick (1991), the median apophysis looks like a hyaline conductor (scored 0). Macerio: median apophysis stick-shaped (scored 0). Strotarchus: there is only a central sclerite, part of the embolic division (scored 1). Uliodon: present but vestigial (scored 0). Philodromus: with “no apparent function during copulation” (Huber, 1995a: 156) (scored 0). Petrichus: no vestige of median apophysis found with SEM (scored 1). Stephanopis ditissima: just a vestige (scored 0). Hispo: not a subtegular apophysis (see Szüts and Scharff, 2009) (scored 1).

357. Median apophysis articulation: 0. Flexibly attached. The median apophysis connects with the tegulum via an area of soft cuticle (fig. 156A). 1. Fixed insertion. In this dataset, a fixed median apophysis occurs only in outgroups and in cases where the fusion is obviously a sclerotized articulation (fig. 147C). Comments: Mimetus: continuous with conductor (scored 0). Aglaoctenus: the median apophysis arises from a partially sclerotized area (scored 01). Miturga cf. lineata: two sclerites arising at the base of the embolus, the median apophysis identified as the more conservative, usually hook shaped, as in Miturgidae QLD, Zora, Teminius, and other Miturgidae (see next character) (scored 0). Lauricius: partially fused (scored 01). Philodromus: small, more prominent in C. californicus (scored 0). Hovops: at embolus base, on top of a turret! (scored 0). Anyphops: placed on a long membranous tube (fig. 154B) (scored 0).

358. Median apophysis continuous with embolus base, directed forward: 0. Absent, other conformations (figs. 148A, E, 150A). 1. At embolic base, directed forward (miturgid conformation). The median apophysis is continuous with the base of the embolus (figs. 145G, 146A, D). This character state represents a palpal conformation characteristic of miturgids. Comments: Donuea, Liocranoides, Cocalodes: unclear but compatible (scored 01). Miturga gilva: there may be a further synapomorphy with Miturga cf. lineata in the pointed anterior projection of the median apophysis (fig. 146C) (scored 1).

359. Conductor: 0. Present (figs. 148A, 154C, 159G, 162A, C). The conductor is typically a semimembranous sclerite, which may be partially or totally sclerotized or hyaline, often with a canal or depression fitting part of the embolus (hence its name). 1. Absent (figs. 163E, 166E). The homology of the conductor is always contentious in spiders. In this study the conductor was identified as a tegular sclerite or membranous outgrowth, somehow associated with the embolus (e.g., with a matching canal, holding the embolus tip), and in the prolateral-ventral-apical region of the copulatory bulb; the median apophysis is much more conservative, usually in the retrolateral region. One of the main issues is the simultaneous occurrence of two structures, each of which might arguably be scored as a conductor (e.g., a “membranous tegular process” and a “hyaline conductor,” as in Griswold, 1993: chars. 8 and 22). Not surprisingly, this character is among the most homoplasious in the dataset, and the implied weighting makes it uninfluential in the analysis (inactivating chars. 359–361 produces the same trees, except for the placement of Lyssomanes sister to Plexippus + Hispo). Comments: Megadictyna: interpreted as in Griswold et al. (2005) (scored 0). Titanoeca: long tegular groove, but also a closed loop protecting the embolus (scored 0). Metaltella, Calacadia, Desis: the embolar cover, interpreted as in Griswold et al. (2005) (scored 0). Aglaoctenus: I identified the hook leading the embolus as a part of the conductor, as in Santos and Brescovit (2001) and Santos et al. (2003: LAC, “lateral apophysis of conductor”), but see Sierwald (2000), who reports many apophyses in Sossipus, not evident which one may be a conductor (scored 0). Clubiona: the fleshy apical lobe (scored 0). Trachelopachys: a membranous structure close to the embolus base, similar to the membranous tegular process in other terminals (e.g., Zoropsis) (scored 0). Brachyphaea: Perhaps fused to tegulum, because there is a suture at embolus base. Coded as a tegular furrow as in Paccius (but in Paccius the furrow is prolonged in a membraneous structure) (scored 1). Phrurolithus, Phrurotimpus: a membranous lobe at embolus base may be a relict of conductor, may be a synapomorphy (figs. 164E, 165C) (scored 1). Teutamus: fused to the tegulum, can be tentatively identified because of the similarity with Sesieutes (scored 0). Jacaena: tentatively identified as the sclerite leading the embolus, fused to the tegulum (see comments on median apophysis), it can be identified because of the similarity with Sesieutes; there may be an additional character to unite Sesieutes, Jacaena and Teutamus in this conformation (figs. 160A, B, 161A–D) (scored 0). Oedignatha: here interpreted as reduced (the apical projecting border of the tegulum, more projecting in O. scrobiculata) (scored 1). Prodidomus: there are membranous folds at embulus base (scored 1). Doliomalus: only a membranous window at the side of an extension of the tegulum (scored 01). Gayenna: the “C2” of Ramírez (2003), reinterpreted according to Ramírez (2007) (scored 0). Systaria: labeled as median apophysis by Deeleman-Reinhold (2001) (scored 0). Philodromus: Perhaps sclerotized but translucent, membranous according to Huber (1995b: 156): “Rotation [of the copulatory bulb] is apparently stopped by the membranous “conductor” that finally becomes arrested by contact with the ventral tibial apophysis. It is unclear whether this “conductor” also assists the introduction of the embolus into the insemination duct”) (scored 1). Petrichus: only a membranous apical area with a furrow where the embolus fits (scored 01). Titanebo: only a membranous apical area with a furrow (scored 01). Eusparassus: just a membranous apical part of the tegulum (scored 0). Aphantochilus: a sclerotized apical lamella totally fused with the tegulum (scored 0).

360. Conductor sclerotization: 0. Sclerotized (fig. 147D). 1. Hyaline (fig. 142B) or membranous (figs. 145C, E, F, 154F). Comments: Homalonychus: completely sclerotized but with the same shape (apical fan) as in many hyaline conductors (scored 0). Vulsor: basal portion membranous, median hyaline, apical sclerotized (scored 01). Pronophaea: sclerotized process with basal hyaline conductor (scored 01). Olbus, Pseudocorinna: heterogeneous, complex (scored 01). Xenoplectus: very evident tube in an expanded bulb, hidden in SEM images (scored 1). Cf. Gnaphosoidea TEX: apparently two conductors, one sclerotized, the other membranous (scored ?). Lampona, Lamponella: membranous (scored 1). Anyphaena: both membranous and sclerotized parts (scored 01). Zora: hyaline, but base sclerotized, articulated (scored 1). Zorocrates: the tegular apophysis (Griswold et al., 2005: fig. 186B) looks like an extra sclerotized conductor (may be part of the embolar division instead) (scored 1). Philodromus: sclerotized but translucent (scored 0). Tibellus: intermediate (scored 01). Anyphops: sclerotized tegular furrow ending in translucent conductor (scored 01). Heteropoda: intermediate (scored 01).

361. Sclerotized conductor articulation: 0. Fixed insertion (fig. 152B). 1. Articulate (figs. 151D, 155C). This character was considered not applicable when the conductor is totally membranous. Comments: Metaltella: extensively fused to tegulum (scored 0). Meedo: all central tegular area membranous (scored 01). Anyphops: unclear, A. barbertonensis seemingly separated by unsclerotized areas (scored 01). Tibellus: intermediate (scored 01).

362. Other tegular articulate sclerites (in addition to the conductor and the median apophysis): 0. None. 1. At base of conductor. 2. At base of embolus (figs. 154D, 157F). 3. At base of median apophysis. Some miturgines have a spine-shaped sclerite parallel to the median apophysis (fig. 145B; Raven and Stumkat, 2003: fig. 5). The sclerite was also found in the diaprograptine Diaprograpta (Raven, 2009, as “accessory spine”). Comments: Titanoeca: a piece at the base of embolus is interpreted as an embolar process (scored 0). Metaltella: the anterior hook is the embolar base (scored 0). Calacadia: there is an apical sclerotized hook, not clear whether it is a separate sclerite; it seems deeply connected with the tegular base (scored 012). Desis: the fleshy finger, not clear at base of what (scored 012). Macrobunus: a small sclerite, tegular apophysis (TA) in Griswold et al. (2005: fig. 193B) (scored 3). Cyrioctea: a large central swelling in C. spinifera, smaller in C. calderoni, absent in C. aschaensis (scored 02). Zoropsis, Uliodon: a hyaline flap, not a sclerite, that would have been coded as conductor if the apical one was absent (fig. 142B; membranous tegular process (MTP) in Boesselaers, 2002: fig. 1C) (scored 0). Oxyopes: sclerotized relics of the tegulum fused to the conductor medially and dorsally (scored 0). Aglaoctenus: there is a projection at the base of the conductor interpreted as part of the conductor (lateral apophysis of conductor, Santos and Brescovit, 2001) (scored 0). Corinna: the Corinna tegular process (PTC) of Bonaldo (2000) may be interpreted as a median apophysis or as a further sclerite (scored 01). Pronophaea: a flat tegular process (scored 2). Procopius: two apophyses at sides of conductor (scored 0). Phrurolithus: retrolateral apical tegular projection also found in Otacilia (scored 0). Camillina: the embolus comes from a more or less demarcated tegular section (scored 2). Cf. Gnaphosoidea TEX: it depends on the interpretation of the conductors (scored ?). Lamponella: two sclerotized cusps may come from one or two separate sclerites, at base of embolus and conductor, all close together (scored 12). Platyoides: a small projection of the tegulum at the base of embolus (scored 0). Anyphaena: secondary conductor in Ramírez (2007) (scored 2). Gayenna, Amaurobioides: the paramedian apophysis (Ramírez, 2003, 2007) (scored 1). Miturga gilva: very small (scored 3). Syspira: there may be a translucent extension parallel to the median apophysis, apparently variable within the same species, not scored as a sclerite (scored 0). Xenoctenus: the central sclerite found in Odo patricius is more fused, but still some weakly sclerotized separations remain; this reinforces the distinction of this piece with the miturgid median apophysis (scored 02). Zorocrates: the tegular apophysis (Griswold et al., 2005: fig. 186B). Heteropoda: There is an apical dorsal tegular projection similar to a conductor, hidden in the unexpanded bulb. This might instead be the conductor, and the large articulated conductor might be the median apophysis (scored 0). Sparianthinae VEN: bulbous area, similar to the conductor in Amaurobioidinae (scored 2).