Revision of the pseudo-orbweavers of the genus Fecenia Simon, 1887 (Araneae, Psechridae), with emphasis on their pre-epigyne

Abstract The present paper provides a taxonomic revision of the genus Fecenia with emphasis on the characteristics of the pre-epigynes which are integrated for the first time into an identification key. As a result, one species is revalidated, Fecenia protensa Thorell, 1891, stat. n., and two new junior synonyms for Fecenia protensa are recognised: Fecenia sumatrana Kulczyński, 1908, syn. n. and Fecenia nicobarensis (Tikader, 1977), syn. n. New records are reported: Fecenia ochracea (Doleschall, 1859)from Malaysian Borneo, Fecenia macilenta (Simon, 1885) from Sumatra, Indonesia, Fecenia protensa from Thailand and Malaysia, Fecenia travancoria Pocock, 1899 from Sri Lanka and Thailand, and Fecenia cylindrata Thorell, 1895 from Thailand and Laos. Additional information on the biology of Fecenia is provided and the validity of characters for identifying Fecenia species is discussed.


Introduction
Representatives of the spider genus Fecenia are distributed from southern India to the Solomon Islands. They are not known beyond the latitudes of 25°N and 15°S. To date (Platnick 2011) this genus comprises five valid species. Fecenia species possess relatively long and prograde legs. The first two pairs are directed anteriorly whereas the third and fourth leg pairs are directed posteriorly. Fecenia species have a flat carapace and a slender body shape (Thorell 1881). Their chelicerae are short and strong and bear a distinct condyle latero-proximally (Levi 1982). Adults build a vertical web, which is similar to the typical orbwebs of (most) Araneidae and related families like Tetragnathidae, respectively. Thus representatives of Fecenia are here called 'pseudo-orbweavers'. Simon (1892) described their web as more irregular than the webs of Araneidae. Furthermore he stated that it contains an enrolled leaf as a retreat in the centre. Despite this somewhat similar web style, Fecenia is not closely related to the Araneidae and does not belong to the Orbiculariae either (Coddington 1990). Together with Psechrus Thorell, 1878, this genus belongs to the Psechridae Simon, 1890 (Simon 1892;Lehtinen 1967;Levi 1982;Griswold 1993;Griswold et al. 2005;Platnick 2011). Previously the pseudo-orbweavers were revised twice. Levi (1982) provided a worldwide revision and Wang and Yin (2001) covered Chinese representatives. In the study of Lehtinen (1967) several Fecenia species were synonymised. Levi (1982) matched a female of a different species with the male of F. macilenta (Simon, 1885). Murphy (1986) recognised this mistake and described the female of F. macilenta for the first time. At present, further taxonomic ambiguities still persist. Some of these were caused by descriptions of new species using subadult females (which only possess pre-epigynes) as type specimens.
Pre-epigynes do not occur in all entelegyne spiders, but seem to be common within the families supposed to be related to Psechridae (Griswold 1993;Griswold et al. 1999Griswold et al. , 2005, e.g. Pisauridae, Lycosidae, Stiphidiidae, Zoropsidae and Ctenidae. Up to now pre-epigynes were mostly disregarded in arachnological studies. There are some first descriptions where pre-epigynes had been erroneously regarded as epigynes (e.g. Psechrus mimus Chamberlin, 1924, Heteropoda shillongensis Sethi and Tikader, 1988, Psechrus ghecuanus Thorell, 1897). A study on Agelena labyrinthica (Clerck, 1757) noted the presence of a primordial copulatory organ in females (Strand 1906). Jäger and Ono (2000), Jäger (2008) as well as Jäger and Bayer (2009) illustrated pre-epigynes of a few particular species of Olios Walckenaer, 1837 and Heteropoda Latreille, 1804 respectively. Several differently developed primordial copulatory organs in different stages of immature females of Cupiennius salei (Keyserling, 1877) were documented in Lachmuth et al. (1985). In Psechridae pre-epigynes were illustrated for the first time by Levi (1982). However, he studied only a few species in this regard. Moreover, in the case of Psechrus himalayanus Simon, 1906, he regarded a pre-epigyne as an adult epigyne. This led to misunderstandings in species determination and characterisation. As an ongoing revision shows (Bayer unpubl. data), Griswold (1993) examined a subadult female of P. marsyandi Levi, 1982, identified as P. himalayanus, as the female representative of the genus Psechrus in his study on the phylogeny of Lycosoidea. Wang and Yin (2001) showed the pre-epigyne of one species, Psechrus rani Wang and Yin, 2001, and compared it with features of the conspecific adult female. A fairly complete investigation on the pisaurid genus Thalassius Simon, 1885 was carried out by Sierwald (1987) where most species concerned were characterised by their pre-epigyne II (penultimate instars) and some even by their pre-epigyne I (antepenultimate instars). An even more detailed study on American Pisauridae described changes in the development of pre-epigynes of different stages via very detailed illustrations (Sierwald 1989). Nevertheless, no study to date has examined variation within pre-epigynes of penultimate instar females within a species, nor has there been any attempt to integrate the pre-epigyne and pre-vulva-features into an identification key. In this context, the intention of this paper is to provide a thorough taxonomic revision of Fecenia including some remarks on their biology and above all the character states of pre-epigynes.

Material and methods
Part of the spider material was collected by hand during an expedition in Thailand and Laos from October-December 2009. Further material was obtained from colleagues, who collected specimens in different regions of SE Asia. Most of the material examined in the present study was borrowed from several natural history museums, which are listed below. Examinations and illustrations were made using a Leica MZ 165 C stereomicroscope with a drawing mirror. Photos of living spiders were taken with a Canon EOS 500D (equipped with a Sigma 105 macro lens and a Canon ringlite). Photos of preserved spiders and copulatory organs were taken with a Sony DSC W70 compact camera via the ocular of the stereomicroscope. The material was preserved in 70% denatured ethanol. Female copulatory organs were cleared from surrounding hairs and dissected. The opaque tissue surrounding the vulva was removed. Vulvae were cleared in 96% DL-lactic acid (C 3 H 6 O 3 ). As the cuticle surrounding the epigyne may curl and structures may get shifted in the course of applying lactic acid, this method could not be applied to every specimen. In males, hairs along the margin of the cymbium were removed to give a clear view of the bulb structures.
All measurements are in millimetres (mm). Leg formula (from longest to shortest leg) and leg spination pattern follow those in Bayer and Jäger (2010). In leg/palp spination the femur, patella, tibia and metatarsus (tarsus in palp) are listed in exactly this sequence. First, all spines on the prolateral surface of the respective limb article are counted and listed, then the ones on the dorsal, then retrolateral and finally the ventral surface. Thus the resulting number is generally one of 4-digits. Some limb articles, e.g. the femur and patella, always lack ventral spines, so here the number is of 3-digits. If a spination pattern from a certain limb article differs between the left and right sides, the pattern for the right article is listed in parenthesis behind, without a blank. Palp and leg lengths are listed as: total (femur, patella, tibia, metatarsus, tarsus).
Terminology of structures belonging to the copulatory organs is given as follows: The female epigyne consists of two slits, which separate the lateral lobes (LL) from the median septum. The latter is folded transversely, resulting in a transverse edge or ridge (TR) (Fig. 79). Consequently, an anterior part of the septum (AS) and a posterior part (PS) can be distinguished (Fig. 79). Anteriorly, each of the LL exhibits a more or less sclerotised margin (anterior margin of lateral lobe, AML). The entire epigyne is surrounded by an epigynal field (EF), which is a sclerotised area. It is not as intensively sclerotised as the median septum or the LL and is distinguished from the adjacent areas of the ventral opisthosoma by a darker colour. The following structures certainly do not belong to the epigyne, but they may be of additional taxonomic information, so they are illustrated and described here, too. Namely the two muscle sigilla (epigynal muscle sigilla, EM) in front of EF (sometimes they are integrated into the epigynal field) and the slit sense organs (SO) near the epigyne (Fig. 79). The vulva consists of an internal duct system (more precisely a folded slit system, cf. Sierwald 1987). It is divided into an initial, rather transparent section (TSI), a strongly sclerotised section (SSI) and the fertilisation duct (FD) (Fig. 83). The border line (BL) between TSI and SSI is clearly visible (Fig. 83). The initial section of SSI features a wide area with pores leading to associated glands. As this area is presumably homologous to the spermathecal head in Psechrus (for location of the spermathecal head see Wang andYin 2001 or Bayer andJäger 2010) the term spermathecal head (SH) is used here for Fecenia too, despite its different shape (Fig. 83). Griswold (1993: p. 21) even denominated the entire SSI as "head of spermatheca", which is not followed here. In Fecenia it is very difficult to locate the receptaculum. It is not clear where the functional copulatory duct actually ends. Moreover, nobody has ever observed how far a Fecenia embolus penetrates within the internal duct system or where the sperm are finally stored.
Apart from structures of a male palp that are well known in arachnology, e.g. conductor, sperm duct or RTA, the Fecenia palp shows a retrolateral patellar apophysis (RPA), a ventral patellar apophysis (VPA) and a membranous process (MP) close to the embolus base (Fig. 8). In one species, F. macilenta, an additional large apophysis arises dorso-retrolaterally from the tibia (dorso-retrolateral tibial apophysis, DRTA, Fig. 53). Presently it cannot be clarified whether this apophysis is just the dorsal branch of an extended RTA or an additional apophysis. In either case, the DRTA can be regarded as an autapomorphy of this species.
Symbols/styles used in the illustrations: Regular solid lines indicate edges/margins/ rims of structures as recognised in the respective view; Weak solid lines indicate edges of fine structures, e.g. membranous structures, or wrinkles in the area of the epigyne; Dashed lines indicate inner walls of ducts and/or slits; Dotted lines (wide) indicate structures shining through the cuticule (e.g. parts of vulva shining through epigynal cuticula). Dotted lines (fine) indicate clear colour differences (e.g. border of epigynal field). In schematic illustrations showing the course of the internal duct system the spermathecal head area is marked with several "T" marks, the copulatory opening with a circle and the end of the fertilisation duct in the direction of the uterus externus with an arrow (see e.g. Fig. 3). When a copulatory opening comprises an elongated slit/area, the circle is put at the central position of that slit/area. Arising points and/or directions of tegular appendages in males are described as clock-positions of the left palp in ventral view. This refers also to directions of some structures of the female vulva. As a convention in this latter case: In every species only the right vulva half is considered.
Museum In the species descriptions the spider material is listed as follows: localities are listed from North to South, then from West to East; countries, provinces and towns/villages are listed as far as possible by their presently valid names.

Characteristics of pre-epigynes
Distinction of pre-epigyne from adult epigyne. Pre-epigynes are considerably smaller than epigynes. If there is no adult female available to compare the size of the epigyne with that of the pre-epigyne of a subadult female the slit sense organs (SO) and epigynal muscle sigillae (EM) in front of the pre-epigyne can help. The distance between the SO from left to right side is about twice as long as the width of a pre-epigyne, but only slightly longer than the width of an adult epigyne. Furthermore, the pre-epigyne is only slightly longer than one EM. The adult epigyne, in contrast, is at least twice as long as EM. Moreover, the pre-epigyne exhibits either no epigynal field or the latter does not reach SO and/or EM.
Ontogeny of the epigyne. Pre-epigynes from four pseudo-orbweaver species were examined and found to exhibit apparently species-specific characteristics. Basic structures of adult epigynes can be recognised as primordial structures in the pre-epigynes. The following general ontogenetic process apparently leads from the primordial to the adult female copulatory organ: The anterior part of median septum (AS) and the anterior margins of lateral lobes (AML) extend strongly anterio-laterally.
In the subadult female of F. protensa Thorell, 1891 the transverse ridge/edge of the median septum (TR) is clearly recognisable as a broad "W"-shaped edge (Fig.  69). In addition to the changes that happen from the subadult to adult stage described above, the median section of TR becomes strongly notched, together with a distinct median folding of AS. The result is the characteristic adult epigyne (Figs 55,64,108).
In F. cylindrata Thorell, 1895, AML run at more or less a right angle anteriorly and face each other. This can be recognised overall in pre-subadult, subadult and adult females . A clearly developed TR is only present in subadult females and adults. In pre-subadult females the TR is at best only slightly indicated (dotted line in Fig. 82).
In F. ochracea (Doleschall, 1859), it is easy to identify corresponding structures of subadult females and adults, because the pre-epigyne (Figs 20-21) already strongly resembles the adult one (Fig. 19). TR is present in subadult females. As on both sides TR is strongly curved anteriorly the characteristic broad-"nose-like" AS, like in adults ( Fig. 19), is already recognisable. By contrast, in pre-subadult females TR is at best very weakly developed (Fig. 22).
In F. travancoria Pocock, 1899, the situation is very similar to that in F. protensa, although its pre-epigyne ( Fig. 74) slightly differs from that of F. protensa (Figs 58,69) (see respective species descriptions). Different developmental stages of pre-epigynes. Epigynes of adult females within the same species are similarly shaped (this is the reason why they can serve as an identification tool). In general this applies to the pre-epigynes, too. Yet, in one out of fifteen subadult females of F. cylindrata the pre-epigyne was larger and somewhat differently shaped (Fig. 80) than generally (Figs 81,94). It gives the impression that it may be further developed than the others. This phenomenon of a differing character state of the pre-epigyne does not mean that identification via the pre-epigyne is not possible. Because if the respective pre-epigyne is interpreted accurately, it is noticeable that it tends to fall along a developmental continuum together with the "regularly" shaped pre-epigynes, the pre-pre-epigynes of p.s.a. ♀♀ and the adult epigynes (Figs 79-82). The s.a. ♀ of F. cylindrata illustrated in Fig. 80 is already more similar to the adult (Fig. 79). Its pre-vulva already exhibits a clear division into a transparent section of internal duct system (TSI) and a strongly sclerotised one (SSI) (Fig. 86). Hence, it is clearly recognised as F. cylindrata.
In summary, pre-epigynes are easily distinguished from adult epigynes and apparently exhibit species-specific characters (note that one species pair F. protensa/F. travancoria is difficult to distinguish, but this is not surprising as it applies to the adults too; see respective species descriptions). In rare cases, in F. cylindrata pre-epigynes of particular subadult females may differ from the general type. But by the means of an accurate interpretation of those pre-epigynes the respective subadult females can be recognised as F. cylindrata, anyway. So, in Fecenia the pre-epigynes can be used as an identification tool. Here they are integrated in an identification key for the first time.

Psechridae Simon, 1890
In combination, the following characters are diagnostic for Psechridae: cribellum and calamistrum present; claw tufts distally on the 3-clawed tarsi; rectangular calamistrum comprising at least 3 rows of setae; indirect eyes with grate shaped tapetum (Simon 1892;Homann 1950;Lehtinen 1967;Levi 1982;Griswold et al. 2005). Diagnosis. Fecenia species differ from Psechrus in the following characters: AME larger than all other eyes (in Psechrus, AME smaller or at most as large as other eyes); ventral side of opisthosoma centrally with pair of two white or beige patches, never with light median line like in Psechrus; clypeus flatter than in Psechrus, not or just slightly higher than diameter of AME, hence cephalic part of carapace rather flat; leg IV always shorter than leg II (in Psechrus, leg IV slightly longer or as long as leg II); in contrast to Psechrus, males with RTA, RPA, VPA and MA; females with clearly divided median septum of epigyne, vulva always lacking spherical spermathecal heads (in Psechrus females, median septum simple and spherical spermathecal heads generally present). Description. Medium sized to large Psechridae, body length in males: 7.2-13.2 mm; females: 7.7-20.2 mm. Cephalic part of carapace not distinctly narrower than broadest (thoracic) section. Anterior eye row recurved, posterior row straight (or at least almost straight). Chelicerae strong, shorter than in Psechrus, basal article at most 2.5 times longer than broad. Cheliceral furrow with three promarginal and four retromarginal teeth. Basal article of chelicerae ventrally with long field of short, transverse striae. Ventral surface of former distally with semicircular lobe with long, curved hairs ( Fig. 6). Labium slightly longer than broad (Fig. 5). Gnathocoxae ca. twice as long as broad, distal section slightly broader than basal one (Fig. 5). Serrula with ca. 130-170 (size-dependant) very small, dark, apically blunt teeth, very densely arranged. Sternum slightly longer than broad, with pointed posterior ending and broad-angled (160°) anterior ending (Fig. 4). Pedipalp in females with single claw (Fig. 51) containing 8-12 teeth. Legs extremely long in males (metatarsus I ca. three times longer than carapace (Fig. 117), relatively long in females (metatarsus I ca. 1.5-2 times longer than carapace, Fig. 119). Leg formula 1243. Coxae of legs I, II broader than III, IV. Calamistrum dorso-retrolaterally on metatarsus IV consisting of 3-4 rows of setae (inner rows irregular). Spination of palp and legs: Highly variable within each species. Therefore, no species-specific and no common genus-specific spination pattern could be found. Consequently the spination will only be listed for the primary type specimen in the species descriptions. At the following positions spines are always absent: All patellae, dorsal surface of all tibiae and all metatarsi. Palpal femur spination varies from 000, 010, 110, 120, 130, which are the most common ones, to 141. Palpal patella, tibia and tarsus mostly without spines, if present, then very small, the most common patterns in this case are: patella 110, tibia 0100, tarsus 1004. Femora of legs I and II with even more variable spination, e.g. 100, 110, 210, 300, 310, 312, 320, 401, 412, 501, or 613. The most common one is 310. The same for those of legs III and IV, but here the number of spines is lower on average, most common is 010. The tibial spination pattern in Fecenia includes a characteristic aspect: Legs I and II: retrolateral spines absent; legs III and IV: prolateral ones absent. At each opposite side the number of spines varies from 0 to 4, with legs I and II mostly having one to two spines more than III and IV. Ventrally at tibiae I and II there are mostly 6, at tibiae III and IV mostly 4 spines (paired spines at all tibiae). The spination of metatarsi is more conservative: I-II 2015, III 1025 or 1015, IV 1015 (ventrally the four proximal spines are paired). But there are exceptions, too. Colouration: Chelicerae, carapace and sternum yellowish brown to dark brown. In rare cases specimens exhibiting a darker carapace margin and a median longitudinal band. Sternum unicoloured. Legs from yellowish brown or light brown to brown, may be annulated. Tibiae I and II in some cases darker than other limbs/legs. Femora at distal third often with dark, annulated patches. Opisthosoma dorsally greyish-brown with yellowish patches. Heart region with darker lanceolate patch with light centre (Fig. 119). Distal half of opisthosoma dorsally with two converging rows of dark brown spots. Lateral surface of opisthosoma is covered with 3-4 larger yellowish patches running diagonally. Opisthosoma ventrally dark brown to black, centrally with a pair of white to beige patches (Figs 116,118), which differ intraspecifically in size and shape. In some cases those patches are fused, in extremely rare cases absent. Additionally, with white to beige transverse patch in front of spinnerets/cribellum (Fig.  116). The whole body is covered with grey hairs (Fig. 116). Spinnerets are relatively short and conical, except for median ones, which are distinctly smaller, slender and cylindrical. Bipartite character clearly visible in posterior spinnerets. Copulatory organs: Male palp with almost round tegulum (T). Embolus (E) filiform, arising in prolateral half of tegulum (T) and at least twice as long as conductor (C). The latter membranous, mostly arising centrally on upper half of T (Fig. 8) and mostly shorter than median apophysis (MA). T next to E-base ( Fig. 8) with membranous process (MP). MA relatively large with general retrolateral direction (e.g. Fig. 89). Cymbium distinctly broader than palpal tibia and patella (e.g. Fig. 62). RTA differently shaped among the particular species, DRTA only present in F. macilenta (Simon, 1885) (Figs 53-54). VPA often slightly bent anteriorly (e.g. Fig. 87). RPA mostly small and inconspicuous. Palpal femur modifications, e.g. ventral bulge as present in some Psechrus species, absent in all Fecenia species. Scopula dorsally on cymbium present in the same form in all Fecenia species (Figs 99-101), but less distinct than in most Psechrus species. Female epigyne generally broader than long, with folded median septum (e.g. Fig. 55). Anterior part of median septum (AS) larger than posterior part (PS). Anterior margins of lateral lobes (AML) iin some species strongly sclerotised (Fig. 108). Vulva simple, with internal duct system divided in three sections: Transversal section (TSI), strongly sclerotised section (SSI) and fertilisation duct (FD) (Fig. 83). Borderline (BL) between TSI and SSI clearly recognisable and often of taxonomic importance.

Key to genera
Biological notes. The pseudo-orbweavers are found in shrubs and trees, and also in the canopy (Deeleman, pers. comm.). Fecenia suspends its vertical pseudo-orbweb ( Fig. 120) in the vegetation, mostly between twigs. The web possesses an enrolled leaf at the hub serving as a retreat. This is true for adults and later instar juveniles of all Fecenia species. Earlier instars build an elongate cone-shaped tube as a retreat, which is disguised with small prey remains and soil-and leaf-particles. The very early instars do not even build a pseudo-orbweb, but a rather conical or dome-shaped web with the retreat at the top of the cone. This web can be found in the herb layer too (Robinson and Lubin 1979). The pseudo-orbweb ( Fig. 120) is more irregular than the webs of araneids and related families building orb-webs. In Fecenia there is no regular spiral of capturing thread(s) as in araneids etc. In Fecenia, one cannot speak of a real spiral as the distance between the threads and their orientation differs. The irregularity applies to the radii too. In many cases they are not continuous.
Predatory behaviour was observed in the lab using several F. cylindrata and F. protensa specimens. In each case the spider was transferred to a large cylindrical glass (30 cm high, diameter 20 cm) with a leaf, already partly enrolled, placed at the bottom. The next day the leaf was suspended by threads in the middle of the glass, a day later it was already fixed at the top. The pseudo-orbweb was completed another day later. After placing a house fly into the lower area of the web it took a few seconds until the spider stretched its two forelegs out of the retreat, and after ca. 1 minute it came out. The fly was grabbed with the chelicerae and immediately dragged into the retreat. A few centimetres before the leaf entrance the spider turned and crawled backwards into the retreat. In the case of larger prey items like crickets, the spider was extremely shy and careful. It took two or three attempts of coming out of the leaf and escaping back into it, sometimes interrupted by 5-15 minutes within the retreat. During the last attempt the cricket was bitten for about 7 minutes at the capturing site of the web before it was dragged to the retreat. Binding behaviour, as described in Robinson and Lubin (1979), was observed after providing an even larger cricket. But in addition to their observations I could recognise that Fecenia took threads out of the web, too, in order to bind its prey. An attack-wrapping behaviour like in Araneidae does not exist. Robinson and Lubin (1979) observed the mating behaviour of a male F. ochracea (however in their publication identified as Fecenia sp.) approaching the female retreat, which I corroborate observing (raised) F. cylindrata from Champasak Province, Laos (males SB 509, 510 and females SB 486-487, 511, 514, see list in description of F. cylindrata, additional material examined). These were maintained in cylindrical glasses (see above) and fed on house flies and crickets. A few days after one male's final moult its web was reduced to a few frame threads. In two corners of that thread-framework sperm webs were found (Fig. 123). The bulb filling procedure was not observed. A female was transferred into a terrarium (30 cm high, diameter 20 cm) and offered a small "cone" of transparent film as retreat, which was accepted and later on integrated in the new web. In the respective trial the male was placed into the female's terrarium. After a while it approached the retreat from the top of the terrarium by roping down onto it. There it tapped and stroked the retreat carefully (Fig. 121). Later on it moved to the margin of the opening of the leaf retreat and repeated this behaviour. After some more repeats it stayed there motionless. Unfortunately, neither the moment of entering the retreat nor the copulation itself could be observed. The next day the male was sitting within the leaf retreat, together with the female (Fig. 122). In another trial a raised F. protensa male from Flores (SB 196, see description of F. protensa, list of additional material examined) was transferred to a terrarium with an already adult conspecific female from Bali. The behaviour was the same as described above, but in this case the next day saw the male half digested lying at the bottom underneath the retreat, which means it had been attacked and killed by the female. In one further trial a F. protensa male was put into the terrarium of a F. cylindrata female. The approaching behaviour upon the leaf was executed up to the point when the male reached the leaf opening. Here he turned and disappeared to an upper corner of the terrarium and stayed there motionless for more than one day.

Key to species:
1 Male ( AS flat (at least anteriorly), without longitudinal folding (Fig. 79), TR without notch, pre-epigyne with continuous TR (Fig. 81), the latter slightly curved, pre-receptaculum with lateral extension (Fig. 85) ............ cylindrata 8 In vulva BL running almost longitudinal (Fig. 77), lateral prongs of the "crown" in pre-epigyne narrow (Fig. 74)  Note on the holotype of Tegenaria ochracea. The first description of Doleschall (1859) lacks any remarks concerning deposition of the type specimen. Generally, material recorded by naturalists of the "Natuurkundige Vereeniging in Nederlandsch Indie" has been deposited either in RMNH or in ZMA. Lehtinen (1967) stated that the type deposition was unknown (to him). Levi (1982) (Doleschall, 1857). Both Lehtinen (1967) and Levi (1982) believed that the syntypes of this species had been lost. However, for this latter species found on Ambon, too, and recorded and described by the same author just two years before, it is evident that at least a part of the original syntype series was once deposited in RMNH (Van Hasselt 1877). After consulting Jürgen Gruber and Verena Stagl (both NHMW) I learned that Doleschall sent only a part of his spider-and insect material collected on Ambon to the museum in Leiden; a large part of the material was sent to the museum in Vienna (Stagl 1999). In the spider collection of NHMW I found a Fecenia female (SB 94), which was labelled "Fecenia -Insel Ambon" (oldest label). According to Gruber (pers. comm.) the handwriting is that of E. Reimoser, the curator of NHMW from 1923-1940. It is well known that Reimoser often discarded old labels and substituted them with new ones (Gruber pers. comm.). It is most likely that in this case the same had happened. Assuming that the handwriting on the original label from Doleschall was unclear, it is likely that Reimoser discarded that label, determined the female as Fecenia and just added the locality on the new label. Anyway, it is evident that before 1950 nobody other than Doleschall sent spider material from the island Ambon to the natural history museum in Vienna (Gruber pers. comm.). Hence, the female SB 94 (see synonymy list above) can be considered the holotype of Tegenaria ochracea.  Diagnosis. Distinguished from other Fecenia species by the epigyne with diverging anterior margins of lateral lobes (AML) (Fig. 1). Males differ from all other Fecenia species by RTA at least as long as width of palpal tibia, MA large and massive, at least as long as width of tegulum (T) (Fig. 8).
Remarks. The vulvae of the holotype of F. cinerea (SB 404) (Fig. 40) and the specimens recorded from Mindiptana, Eastern Papua Province of Indonesia (SB 96-98, 442) (SB 98 illustrated in Fig. 44) differ from all other females examined. The duct of SSI is somewhat longer, especially the second curve (Figs 40, 44). Consequently, the course of the internal duct system of these specimens (Figs 41, 45) differs from the remaining F. ochracea females (Figs 3, 35, 37, 39, 43, 47). However, the vulvae of the holotype of F. cinerea (SB 404) and female SB 98 do not correspond completely. In SB 404 the second curve of SSI protrudes more strongly in a lateral direction. In one specimen (SB 97, not illustrated) from Mindiptana the second curve of SSI is a bit shorter than in the others from this locality. The epigynes of SB 96-98, 404 and 442 differ in shape (SB 404: Fig.  28; SB 98: Fig. 32; others not illustrated). According to the differences in the shape of the vulvae (see above) it may be justified to revalidate F. cinerea Hogg, 1914. However, the difference is little (second curve of SSI slightly longer than in F. ochracea) and thus does not provide evidence for a clear species delimitation; especially considering that in one specimen from Mindiptana the second curve is again slightly shorter. In addition, if the females from Mindiptana should be regarded as F. cinerea, then the male (SB 95, Figs 9, 13, 17), which was recorded from exactly the same locality, should be placed here, too. However, as discussed above, the palp structures of this male only slightly differ from the ones of other F. ochracea males (though these differences are worth mentioning as intraspecific variation). Moreover, no males have been recorded from the type locality of F. cinerea so far. Consequently, I refrain from changing the taxonomic status of F. cinerea. More material from the type locality of F. cinerea, especially males may enlighten this "problematic case".
Pre-vulva may be slightly more structured (Fig. 67). Remarks. The reasons for revalidation of F. protensa and the synonymy of F. sumatrana with the former are as follows: In Thailand, Malaysia, Singapore and on Bali at several localities subadult Fecenia females were collected together with adult females respectively, which showed the characteristic rounded-"W"-shaped epigyne. The preepigyne of the subadult female holotype of F. protensa (SB 620) matches the ones of the subadult females mentioned above. In 1908 Kulczyński described F. sumatrana. The (adult) female holotype of this species exhibits the characteristic rounded-"W"-shaped epigyne. The adult females mentioned above match the holotype of F. sumatrana. F. protensa is the oldest name available and hence the valid name for this taxonomical species. It is distinguished from F. travancoria by the BL of the vulva running almost transversal. Consequently, F. sumatrana is not a junior synonym of F. travancoria as postulated in Murphy (1986), but a junior synonym of F. protensa. Both, F. protensa and F. travancoria are regarded as valid species (see also remarks sub species description of F. travancoria).
Reason for synonymy of F. nicobarensis with F. protensa: Although the types of Psechrus nicobarensis were not available on request it became obvious that Tikader (1977) dealt with F. protensa. The drawing of the female epigyne in Tikader (1977: 208, fig. 27B) is not very informative, however, the rounded-"W"-shaped character of the epigyne is very clear. His fig. 27E of the right male palp is more detailed. However, the proportions probably do not reflect the real situation. Additionally, this illustration does not represent an exact ventral view of the palp. If the left palps of the males examined herein (SB 128,136,137,218,219,512) were arranged in the same way/ position, they would match the (mirrored) drawing in Tikader (1977 Diagnosis. Females distinguished from other Fecenia species except F. protensa by having anterior margins of lateral lobes (AML) anteriorly more or less converging and surrounding epigynal pit partly and the anterior part of median septum (AS) comprising a longitudinal, anteriorly pointed folding (Fig. 76); moreover, by having a notched transversal edge (TR) of median septum. Females are distinguished from F. protensa by the almost longitudinal borderline (BL) between strongly sclerotised section (SSI) and the transparent section of internal duct system (TSI) in vulva (Fig. 77).
Remarks. This species is very similar to F. protensa. There are only fine differences in characters of the vulva (see diagnosis). Up to now, no intermediate forms concerning the shape of vulva have been found. Though it cannot be fully excluded, it seems rather unlikely that F. travancoria is a junior synonym of F. protensa. Generally, in Fecenia species the vulva shows less intraspecific variation than the epigyne. By now I consider F. travancoria as valid species. But with more material from the southern Provinces of India, especially males, it may be possible to clarify this 'difficult taxonomic case'.

Fecenia cylindrata
Remarks. Thorell (1895) described this species based on juvenile types. Two years later Thorell himself redescribed this species based on ♂♂ and ♀♀ recorded just ca. 70 km away from type locality Tharrawaddy (Thorell 1897). This material is deposited in NRS, ZMH and MCSN and was examined (see material list above). Moreover, to date no other Fecenia species than the one described above had been found in Myanmar. For that reason there are no doubts about the identiy of Fecenia cylindrata.
Fecenia hainanensis Wang, 1990 was synonymised with F. cylindrata by Wang and Yin (2001). The female holotype from Tonqian, Hainan Province, China was not available on request. According to the illustrations in Wang (1990), which are not very detailed, it is more likely that his F. hainanensis was in fact conspecific with F. cylindrata. The specimens from Hainan checked in the present study are considered belonging to F. cylindrata, though there are slight differences (see variation of copulatory organs in the description of F. cylindrata). More material from Hainan and also from regions of South East China and Northern Vietnam is necessary to assess the consistency of those slight differences among the different specimens. At the moment F. hainanensis is regarded as junior synonym of F. cylindrata.

Characteristics of the pre-epigyne
The pre-pre-epigyne (antepenultimate instar), although hardly useful for species determination, may bear important information. In some Fecenia species both presubadult and subadult females were available for examination. A continuous developmental trend from pre-pre-epigyne (p.s.a. ♀♀) to the epigyne of adults can be traced (e.g. Figs 19-22 for F. ochracea, Figs 79-82 for F. cylindrata). Sierwald (1989) showed that in most of the American Pisauridae even more primordial epigyne stages exist. In Pisaurina mira (Walckenaer, 1837) up to five stages with differently developed primordial copulatory organs (which Sierwald denominated as "anlagen") occur. Gradually from earlier to later stages the anlagen resemble more and more the adult. The changes from penultimate instar to adult constitute the largest developmental step as the shapes of pre-epigynes and adult epigynes differ the most. The number of primordial stages in Pisaurina mira varies between three and five (Sierwald 1989). Interestingly, in specimens with only three primordial stages, the anlagen of the antepenultimate and penultimate instars were less developed and differentiated. Anyway, these specimens moult following their third anlage to "normal" mature females (Sierwald 1989). The total number of juvenile stages varies in Pisauridae. For example in Dolomedes triton (Walckenaer, 1837) the number ranges from 10 to 15 in males and 9 to 15 in females (Zimmermann and Spence 1998).
The present study reveals the occurrence of a different developmental stage of the pre-epigyne (penultimate instar) in the pseudo-orbweaver F. cylindrata (Fig. 80). The following preliminary considerations may explain this phenomenon: In insects a juvenile hormone (JH) regulates the development of the larva throughout the several moults up to the imago. Following Wigglesworth (1952), a controlled hormone balance between JH and prothoracotrope hormone is essential for regular development of the bug Rhodnius prolixus Stål, 1859. From 1 st to 4 th stage larva the concentration of JH decreases more or less continuously, but from the 4 th to 5 th stage the decrease is much stronger and from 5 th stage to imago JH is completely absent (Wigglesworth 1952). It is likely that JH exists in spiders, too (Webber 2005). Prothoracotrope hormone does not exist in spiders, but instead of this it is possible, that another, equivalent hormone exists.
On the other hand it is known from spiders that the number of moults, and thus the number of instars, to reach maturity may differ, for example in Pisauridae (see above). In Latrodectus mactans (Fabricius, 1775) the number of instars varies from 7 to 9 depending on food supply (Deevey 1949). From particular species of Stegodyphus Simon, 1873 it is also known that maturity is reached after different numbers of moults in different specimens examined, irrespective of their sex (Kullmann et al. 1972, Kraus andKraus 1988). Furthermore, Kraus and Kraus (1988) state that the enormous size variation in species of Stegodyphus seems to be caused mainly by this flexibility. At least in the species F. ochracea and F. cylindrata, the size variation is high. This becomes obvious by their carapace-length size ranges (see respective descriptions). It is possible that in Fecenia the number of moults required to reach maturity differs intraspecifically, too. Considering that the number of stages of immature females with differently developed primordial copulatory organs varies in Pisauridae (see above), a family also belonging to the Lycosoidea (Griswold et al. 2005), it is not unlikely that this applies to the pseudo-orbweavers too. A preepigyne of a s.a. ♀ of the 6 th instar would then most likely differ from the one of an 8 th instar.
In Fecenia it seems to be rare, that the pre-epigyne of a particular subadult female differs from the ones of the others belonging to the same species. But, anyway, as the example of the subadult female of F. cylindrata (Fig. 80) shows, this phenomenon may appear. In such a case additional consideration concerning the identification of subadults is necessary. Does the respective subadult female fit into a conceivable developmental continuum for the species in question? This is, of course, much easier if several "regularly" developed s.a. ♀♀ and/or p.s.a. ♀♀ are available. As the pre-epigyne of a "further developed" s.a. ♀ most likely resembles more an adult epigyne than a "regularly developed" one does, it should not be too difficult to identify it. Thus, in Fecenia the pre-epigynal characters apparently are species-specific (pre-epigynes, take notice; this must not inevitably mean that this applies also to the pre-pre-epigynes or other primordial epigynes of instars below subadult females!). Following the studies of Sierwald (1987Sierwald ( , 1989) the pre-epigynes of the Pisauridae species examined seem to be specific, too. Hence, it is justified to use the preepigyne as a tool for identification.

Validity of characters in Fecenia.
Somatic characters are not useful for species determination in Fecenia. Colouration and spination, for example, are highly variable intraspecifically. Figures 69 and 82 in Levi (1982) suggest that species discrimination between F. ochracea and F. cylindrata via colouration of the ventral surface of the opisthosoma is possible. According to the present study, this cannot be confirmed. Species identification is only possible by checking the copulatory organs.

Remarks on spination
In the description of the genus Fecenia above a characteristic aspect of the spination pattern on the tibiae is mentioned. This may be explained by the life style of the pseudo-orbweavers. Fecenia is the only spider genus in which all representatives spend at least 95% of their lifetime in a very narrow enrolled-leaf retreat or cone retreat in early juveniles. In Araneidae there are several genera including species, that have similar lifestyles, e.g. Acusilas, Cyclosa, Neoscona, Araneus, Cyrtophora, also in Theridiidae, e.g. Parasteatoda simulans (Thorell, 1875). In any case, there is no genus in which all representatives use enrolled leaves as a retreat. Furthermore, in representatives of the families mentioned above the leaf-retreat is never as narrow (in relation to body size) as in Fecenia. A pseudo-orbweaver enters its retreat always with its opisthosoma first. The patellae and tibiae have the most intensive contact with the inner walls of the leaf retreat. As the legs are prograde with leg pairs I-II held anteriorly and III-IV posteriorly it becomes obvious that in the first two leg pairs the retrolateral and in the last two leg pairs the prolateral spines on the tibiae would be an impediment while moving inside the retreat. Perhaps in the course of the evolution of this genus, specimens with shorter spines or even no more spines at these respective positions were preferred? Like in Psechrus the patellae completely lack spines (Lehtinen 1967). This characteristic aspect of the tibial spination pattern in Fecenia may be an adaptation to this special life style. It would be interesting to check if the tibial spination pattern of species from the Araneidae and Theridiidae genera listed above using enrolled leaves, differ from the ones with different lifestyles. But in contrast to Psechridae in Araneidae and Theridiidae the spines are in any case not so prominent in comparison to leg diameter.