Varadia, a new helicarionoidean semi-slug genus from India’s Western Ghats (Stylommatophora: Helicarionoidea)

We here describe a new Indian helicarionoidean genus, Varadia Bhosale & Raheem gen. nov., containing the single species Varadia amboliensis Bhosale, Thackeray, Muley & Raheem gen. et sp. nov. This new semi-slug is endemic to the northern a nd central Western Ghats and is primarily a forestBHOSALE A.R. et al., Varadia, a new semi-slug genus from India’s Western Ghats 51 living species. We describe and fi gure the shell, reproductive system, radula, spermatophore and external morphology of thi s new species, and detail its known distribution. We explore its relationships to other helicarionoideans using phylogenetic analyses of DNA sequence data for part of the ribosomal RNA gene cluster and discuss the morphology of the new genus in relation to other, primarily South Indian, helicarionoidean taxa. Based on characters of the reproductive system, particularly the male genitalia and the gametolytic sac, we provisionally place Varadia gen. nov. in the Macrochlamydinae (Ariophantidae). This is consistent with the results of our molecular phylogenetic analyses. The combination of large size, broad, densely tuberculated shell lobes and a shell with ca 4 whorls and a disproportionately large body whorl makes V. amboliensis gen. et sp. nov. unique among the helicarionoidean taxa of the Western Ghats. The new semi-slug is also highly distinctive in the morphology of its male genitalia.

Our current understanding of this taxonomically and morphologically diverse fauna has many gaps and is largely based on publications and collections from the British colonial period (see monographs by Blanford & Godwin-Austen 1908;Raheem et al. 2014). Many taxa were poorly described / defi ned and / or were not illustrated in the original literature, hence making their identifi cation problematic. The genus-level placement of some taxa is uncertain and questionable (e.g., at least some of the western and southern Indian species assigned to the genus Macrochlamys may belong to Eurychlamys; Blanford & Godwin-Austen 1908;Raheem et al. 2014). Data on the reproductive anatomy, external morphology and radula are scarce or lacking for most species. Knowledge of species' distributions is poor; while many species are known only from single localities and have not been recorded for over a century, the snail fauna of substantial parts of the Western Ghats remains to be explored (Raheem et al. 2014). To date, although South Asian taxa have been included in molecular phylogenetic analyses of deep-level pulmonate relationships (e.g., Wade et al. 2001Wade et al. , 2006Herbert & Mitchell 2009;Ramirez et al. 2012), studies focussing on the molecular systematics and diversifi cation of South Asian helicarionoideans have yet to be carried out. With the exception of the substantial work done on Australian helicarionids (e.g., Hyman et al. 2007Hyman et al. , 2017Hyman & Köhler 2019), this refl ects the general scarcity of such studies globally (but see Schilthuizen et al. 2019;Pholyotha et al. 2020a).
The taxonomic study of the WGSL land-snail fauna has seen a renewed interest in recent years (e.g., Raheem & Naggs 2006;Raheem et al. 2014;Aravind & Páll-Gergely 2018;Bhosale et al. 2019, in press;Páll-Gergely et al. 2020). This includes the fi rst modern surveys of the land-snail fauna of the northern Western Ghats by one of us (Bhosale et al. 2016Bhosale 2018). Begun as PhD fi eldwork, these surveys focussed initially on Kolhapur District, Maharashtra, a 7685-km 2 area in the extreme south of the northern Western Ghats, which had not been previously explored by malacologists. It was during these surveys that a large, forest-living semi-slug with a Macrochlamys-like shell and broad, conspicuously tuberculate shell lobes was encountered. Based on comparative morphological study and molecular phylogenetic analyses of part of the ribosomal RNA (rRNA) gene cluster, we here describe this semi-slug as a new Indian helicarionoidean genus, Varadia Bhosale & Raheem gen. nov., containing the single species V. amboliensis Bhosale, Thackeray, Muley & Raheem gen. et sp. nov. We discuss the morphology of this new genus in relation to other, primarily South Indian, taxa and explore its phylogenetic relationships to other helicarionoideans, provisionally placing it in the subfamily Macrochlamydinae (Ariophantidae).

Material and methods
Living snails and shells of Varadia amboliensis gen. et sp. nov. were collected and / or observed by A. Bhosale at 5 localities in the state of Maharashtra, India in September 2017, November 2019 (during the autumn) and in September 2020, towards the end of the monsoon rains ( Fig. 1). After sampling of foot tissue for DNA analysis, snails were euthanized following the guidelines of the American Veterinary Medical Association (2020) and preserved in 80% ethanol for anatomical study. The whole radulae were prepared following Bhosale et al. (2019) and SEM images were taken on a TESCAN VEGA3 SEM. All radula counts were carried out in the area of greatest width. Morphological studies of collected material were carried out at the Department of Zoology, Shivaji University, Kolhapur, India, with type and other material being deposited in the collections of the Bombay Natural History Society (BNHS), Mumbai and Zoological Survey of India (ZSI), Western Regional Centre (WRC), Pune, India. Unless stated otherwise, we follow the morphological terminology of Raheem & Naggs (2006) and Bhosale et al. (2019) for the mantle and reproductive system (the terms 'proximate' and 'distal' are used with reference to the genital orifi ce); Hyman & Köhler (2019) for the spermatophore; and Cox (1960) for the shell (including shell sculpture). Shell whorl counts (to the nearest quarter whorl) are based on the approach of Kerney & Cameron (1979). Shell height and width were measured as shown in Fig. 2. Unless stated otherwise, all photographic images are by A. Bhosale.
For the molecular phylogenetic study, our ingroup consisted of 30 taxa belonging to the order Limacoidea Batsch, 1789 (Table 1). The outgroup comprised three species of Arionoidea Gray, 1840; a recent study has shown that the orders Arionoidea and Limacoidea are sister groups (Saadi & Wade 2019). Unless stated otherwise, we have followed the higher-level classifi cation of Bouchet et al. (2017). Phylogenetic analyses were based on the region of the rRNA gene cluster described by Wade & Mordan (2000). DNA sequence data for eight of the 33 taxa were generated during the course of the present study; data for the remainder were generated in previous studies by Wade et al. (2001Wade et al. ( , 2006. For the newly sequenced samples, DNA was extracted from a 1-2-mm 3 piece of foot tissue using a CTAB DNA extraction protocol (Goodacre & Wade 2001). PCR amplifi cations were done by adding 2.5 μl of DNA extract to a reaction mix containing 1× PCR reaction buff er (including 1.5 mm MgCl 2 ), 0.2 mm each dNTP (Sigma-Aldrich, USA), 0.2 μM each primer and 1 unit Taq DNA polymerase (Sigma-Aldrich, USA) in a fi nal volume of 25 μl. The PCR conditions were as follows: 96°C for 1 min; 30 cycles of 94°C for 30 s, 50°C for 30 s and 72°C for 1 min; and a fi nal extension step of 72°C for 5 min. Amplifi ed products were purifi ed using the HiPurA™ Quick Gel Purifi cation Kit. Sequencing was carried out using a BigDye Terminator ver. 3.1 Cycle Sequencing Kit (Applied Biosystems, USA) with both DNA strands sequenced on a 3500 Genetic Analyzer (Applied Biosystems, USA); sequencing was done at the Department of Biochemistry, Shivaji University, Kolhapur. DNA sequences were assembled using STADEN ver. 1.5.3 (Staden et al. 2000). Subsequently, the sequences were manually aligned using the Genetic Data Environment ver. 2.2 package (Smith et al. 1994).
Phylogenetic analyses were based on a dataset of 874 unambiguously aligned nucleotide sites for 33 taxa. We used maximum likelihood (ML; Felsenstein 1981), neighbour joining (NJ; Saitou & Nei 1987) and Bayesian (Larget & Simon 1999) approaches, with the nucleotide substitution model being GTR + Γ (Lanave et al. 1984;Gu et al. 1995). ML analyses were carried out using RAxMLHPC2 (Stamatakis 2014). Each RAxML analysis involved a single programme run and consisted of a rapid bootstrap analysis with the extended majority rule bootstopping criterion, followed by a search for the best-scoring ML tree (Stamatakis 2014); analyses were run on the CIPRES Science Gateway ver. 3.3 (http://www.phylo.org/index.php, Miller et al. 2010) and each run was repeated at least once. NJ analysis was performed using PAUP* ver. 4.0b10 (Swoff ord 2002) with model parameters estimated following an iteration process; for each tree the parameters were estimated and used to build the next tree until there was no further improvement of the likelihood score. Bootstrap resampling (Felsenstein 1985) with 1000 replicates was undertaken for the NJ trees. Bayesian analyses were done using MrBayes ver. 3.1.2 (Ronquist & Huelsenbeck 2003). Each MrBayes analysis involved two independent MCMC runs (with four chains per run) for 5 million generations, sampling every 100 generations; each analysis was repeated at least once. To ensure adequate chain swapping, a range of heating parameters were tested with the optimal parameter used to construct the fi nal trees. Only after the Bayesian MCMC searches had reached a stationary phase (indicating convergence of the chains onto the target distribution) was the run ended. A consensus tree was built using the last 75% of trees (burnin = 5001). Trees were rooted on a composite outgroup comprising three members of the Arionoidea, Arion hortensis Férussac, 1819, Meghimatium bilineatum (Benson, 1842) and Philomycus carolinianus (Bosc, 1802). Branches with bootstrap support (BS) values ≥ 70% and posterior probabilities (PP) ≥ 0.95 were considered to be well / strongly supported (Hillis & Bull 1993;Alfaro & Holder 2006). The frequency of all bipartitions in the bootstrap trees used for computing the NJ and ML consensus (50% majority-rule) trees were estimated using PAUP*.

Diagnosis
Only known large semi-slug species from the Western Ghats having broad shell lobes (covering nearly all of the dorsal shell surface when fully extended) and a depressed, discoid shell with a large body whorl. Shell thin, glossy golden brown, with 4-4½ rapidly increasing whorls and barely raised spire. Extensive, largely free penial sheath enclosing substantial part of male genitalia. When sheath is in situ and penis in relaxed state, male genitalia constitute a single elongate, irregularly-shaped mass, with all of penis and epiphallic caecum, most of epiphallus and part of fl agellum enclosed by sheath; only vas deferens and most of fl agellum is visible outside intact sheath. Penial sheath divisible into thick proximal part and thin distal part, with transition between these two parts occurring between distal penis and epiphallic caecum. With penial sheath removed, following evident: region extending from penis to epiphallic caecum long and cylindrical, with distinctive S-shaped bend in penis and associated band of muscle running along length of penis; proximal ¾ of epiphallus held in long, conspicuous loop; and penial retractor muscle with two branches, one inserting subterminally on epiphallic caecum and one inserting on apex of loop of epiphallus. Inner wall of penis divisible into three regions: proximal penis (one major and several minor longitudinal pilasters), mid-penis (several minor longitudinal pilasters) and distal penis (a few minor pilasters with associated regular transverse ridges in interspaces). Inner wall of epiphallic caecum shows one long major longitudinal pilaster, a large mass of reticulate ridges proximally and several minor longitudinal pilasters distally. Gametolytic gland elongated and long; gametolytic sac 3-3.5 times as long as gametolytic duct. Amatorial organ absent. Spermatophore consists of elongated, soft capsule and long tail-pipe, with U-shaped bend at capsule-tail-pipe junction and funnel-like opening at tip of tail-pipe; surface smooth apart from four ribs running obliquely along length of tail-pipe and short spines near end of tail-pipe.

Etymology
The new genus is named in honour of the herpetologist Dr Varad Giri, who has made a major contribution to the modern study and conservation of the Indian herpetofauna; masculine.

S
. Adult shell thin, depressed, glossy and appearing non-umbilic ate, with ca 4-4.5 rapidly increasing whorls and colour ranging from golden-brown to reddish yellow (Fig. 3). Shell measurements (n = 35): width 20.2-26.3 mm; height 10.2-15.0 mm. Spire only slightly raised with fl at apex and suture only slightly impressed. Body whorl disproportionately large, rounded at periphery, gently convex beneath. Aperture large, crescent-shaped, with width greater than height (Fig. 3A, E). Apertural margin simple, thin and delicate; in lateral view angled forward, with upper apertural margin noticeably anterior to the lower margin. When shell is viewed from below, basal margin curved (not straight) and expanded columellar margin refl ected over, covering umbilical region. Shell surface smooth and glossy to naked eye, with irregular, faint collabral striae; under SEM, seen to be fi nely and closely sculptured with well-defi ned spiral lines on protoconch (fi rst 1.5-2 whorls) and indistinct and irregular oblique lines on teleoconch (Fig. 4).

B
. Total adult body length, excluding extended tentacles, ranges from 4.8 to 6.9 cm (n = 5). Living snail glossy grey or greyish white with irregular dark mottling; head and tail dark grey or blackish with tentacles tending to be paler at their tips (Fig. 5-6). Surface of mantle densely and conspicuously covered by small, irregular tubercles that appear lighter on top. Sole tripartite with welldefi ned sole furrows dividing it into three distinct longitudinal tracts; central tract paler than lateral ones. Tail with large slit-like caudal pit (sensu Hausdorf 1998: 51); caudal horn prominent when extended (Fig. 6A) but when retracted gives tail truncated and blunt appearance (Fig. 6B). Mantle consists of two broad shell lobes (right and left) and two dorsal lobes (right and left) (Fig. 7). Shell lobes may cover nearly all of dorsal surface of shell (Fig. 5), but individuals have also been observed with shell lobes largely retracted and much of shell exposed (Fig. 6). Left dorsal lobe extends as far as base of tentacles when snail is resting (i.e., body not fully extended) and tentacles are retracted.

R
. Male genitalia consist of proximally penis and distally epiphallic caecum (= epiphallic retractor caecum of Hyman & Ponder 2010: 139) and epiphallus. These three regions are  held together by a penial sheath (= penial tunica of Hausdorf 1998), which is largely independent of wall of penis, epiphallic caecum and epiphallus (Figs 8A, 9A-C). Penial sheath encloses all of penis, all of epiphallic caecum, much of epiphallus and part of fl agellum; it holds proximal three quarters of epiphallus in a loop against epiphallic caecum, with remaining part of epiphallus (i.e., part closest to vas deferens and fl agellum) lying outside sheath, along with a substantial part of fl agellum (Fig. 9A). Proximal part of penial sheath is thick and covers penis; distal half of this sheath is thin and covers epiphallus and epiphallic caecum. Thick penial sheath attached to proximal end of penis, close to genital atrium. Thin penial sheath attached to distal end of epiphallic caecum and is open where penial retractor muscle inserts on epiphallic caecum (Fig. 9A-C); an extension of the thin penial sheath also encloses a sizeable section of fl agellum (this section located about halfway along length of fl agellum). Epiphallus passes through and is attached to penial sheath in region where thick penial sheath transitions into thin penial sheath.
With penial sheath dissected open, penis seen to have noticeable S-shaped bend midway; this bend is associated with a band of muscle that extends for some distance along penis, on either side of bend ( Fig. 9D-E). Distally, penis branches into wider-lumened epiphallic caecum and narrower-lumened epiphallus. Epiphallus passes into much narrower-lumened vas deferens; junction between these two regions marked by long, bluntly pointed fl agellum, which is similar in length to epiphallus. Penial retractor muscle, which originates on inner lung wall, inserts in two places (Fig. 9E): subterminally on epiphallic caecum, and on apex of loop of epiphallus (i.e., about three quarters of distance from vas deferens to penis). Junction between two branches of penial retractor muscle located near most distal part of epiphallic caecum. Irregular small holes / pores visible on inner surface of thin part of penial sheath (i.e., with sheath cut open and pinned out).
On the basis of the morphology of its inner wall, penis divisible into three morphologically distinct regions, proximal penis, mid-penis and distal penis, with S-shaped bend of penis including all of mid-and distal penial regions (Fig. 10). Proximal penis shows one major and several minor longitudinal pilasters; close study at low magnifi cation (4 ×) of holotype and one paratype (BNHS GAS 114) showed that pilasters are interspersed by fi ne, obliquely longitudinal ridges that are close and irregular. Mid-penis ornamented by several thin longitudinal pilasters. Distal penis also with thin longitudinal pilasters, but here they are fewer in number and are contiguous with uniform, widely spaced transverse ridges that extend outwards on either side of each pilaster. Opening of epiphallus into most proximal part of epiphallic caecum clearly visible (Fig. 10). Inner wall of epiphallic caecum (Fig. 10) has one major longitudinal pilaster (surface marked by irregular, fi ne longitudinal and / or transverse ridges; not shown in Fig. 10) running along its length; a large, reticulate mass of ridges proximally; and several short longitudinal pilasters distally. The short pilasters tend to be crenulated proximally and are smoother distally. Lumen of vas deferens widens with increasing distance from epiphallus, with part of vas deferens nearest to epiphallus being noticeably narrower-lumened than remaining two thirds (Fig. 8A). Right eye retractor Abbreviations: ag = albumen gland; cg = capsular gland; ec = epiphallic caecum; ep = epiphallus; esh = extension of penial sheath around epiphallus; fl = fl agellum; ga = genital atrium; gd = gametolytic duct; gs = gametolytic sac; hd = hermaphrodite duct; ir = irregularly-marked papillate ridges; lw = lung wall; ot = ovatestes; ov = oviduct; pe = penis; pg = prostate gland; pr = penial retractor muscle; thsh = thick penial sheath; tsh = thin penial sheath; so = spermoviduct; va = vagina; vd = vas deferens. muscle passes between male and female genitalia. Amatorial organ absent. Genital atrium cylindrical, well defi ned but short, with junction between male genitalia and vagina located at a short distance from genital orifi ce. Vagina cylindrical and shorter in length than genital atrium (Fig. 8A). Proximal part of oviduct, near junction with gametolytic gland, consists of pale yellowish, indistinctly-defi ned region, which is most likely the capsular gland (see Dasen 1933); inner wall of this gland irregularly marked by papillate ridges and papillae (Fig. 8). Gametolytic gland (Fig. 8A) comprises narrow gametolytic duct and long, voluminous sac that is ca 3-3.5 times length of duct; duct noticeably constricted at its junction with sac and has 1-3 longitudinal ridges on its inner wall.   A. Fragments of two spermatophores (part of the tail-pipe is missing in both), as found in situ in the gametolytic sac of the paratype BNHS GAS 125. B-D. Morphology of the fully-intact spermatophore found in the gametolytic sac of the paratype ZSI Moll/1820. B. The upper arrow indicates the U-shaped bend at the junction of the capsule and tail-pipe, and the lower arrow shows the funnel-like opening at the tip of the tail-pipe. Also shown is the cross section of the spermatophore (indicated by the dotted line) close to the tip of the tail-pipe; note the hollow central area. C. Detail of a small section of the tailpipe (indicated by rectangle in B), showing one of the four longitudinal ribs and the hair-like spines on the spermatophore surface. D. Detail of tip of tail-pipe, showing the funnel-like opening that leads into the hollow central area (shown by arrow). Note the hair-like spines. Abbreviations: cap = capsule; gd = gametolytic duct; gs = gametolytic sac; o = opening at tip of tail-pipe; sp = spermatophore; tp = tail-pipe.
One or two spermatophores (i.e., only one wholly intact; the rest damaged / partially digested) present in gametolytic gland of each of six specimens (Fig. 11A). Intact spermatophore consists of elongated, soft capsule with long tail-pipe. Sharply-angled, U-shaped bend at junction of capsule and tail-pipe; apex of bend noticeably hooked towards tail-pipe. Capsule wider-lumened than tail-pipe and twisted spirally (Fig. 11B). Tail-pipe fl exible, internally hollow and externally sculptured obliquely along its length with four fi ne ribs. Tail-pipe in vicinity of tip hollow centrally and this passes into funnel-like opening (perforation) (Fig. 11D); surface of spermatophore near tip of tail-pipe has short, hair-like spines that point towards capsule (Fig. 11C).

R
. Central tooth tricuspid, with large mesocone, which is shorter than tooth base, and smaller, more basal ectocones (Fig. 12A-B). Inner laterals 17-21, uniformly tricuspid ( Fig. 12A-B); mesocone large, equal in size to those of central tooth and shorter than tooth base, endocone barely defi ned and ectocone prominent but more basal than other cusps. Outer 2 lateral teeth grade into marginal teeth. Marginal teeth 45-53, uniformly bicuspid (endocone absent), with shorter, narrower and more basal ectocone (Fig. 12C-D). Formulae for the 8 specimens examined are as follows (the plus sign indicates that the outermost marginal teeth could not be counted): Jaw oxygnath (smooth), having a concave cutting edge with well-defi ned or barely evident median projection (Fig. 12E). Varadia amboliensis gen. et sp. nov. is primarily a ground-living snail. It can be encountered at night in leaf litter or on rocks and the bases of trees; in rainy weather it can be seen on the exterior walls of buildings close to the forest edge (e.g., it was observed at the entrance of Amboli Forest Park in September 2017). The species can be seen throughout the monsoon (June to October) and as late as the end of November.

Distribution and ecology
A few individuals have been seen in late February (late winter) on the banks of fast-fl owing streams at Amboli.
This species appears to be omnivorous. It has been observed feeding on decaying plant matter (leaf litter, discarded banana peel) and on the remains of at least two diff erent invertebrate taxa (a cricket and an earthworm) (A. Bhosale, personal observation) (Supp. fi le 1). Data on its predators are scarce,

DNA analysis
The three phylogenetic analyses (NJ, ML and Bayesian) yielded broadly similar results (Fig. 14), with disagreements occurring only for internal branches lacking strong support in any of the analyses. Most deeper relationships within the Limacoidea were not strongly supported, with the optimal ML and NJ trees having fewer strongly supported branches than the Bayesian tree. All three trees included a sister-  Table 1). The tree is rooted with an outgroup consisting of three members of the Arionoidea (Arion hortensis Férussac, 1819, Meghimatium bilineatum (Benson, 1842) and Philomycus carolinianus (Bosc, 1802)). Clades A and B are indicated by arrows. Branch support values are shown in the following sequence: Bayesian posterior probabilities, ML bootstraps and NJ bootstraps. Symbols: '*' = maximal support (Bayesian = 1; NJ, ML = 100%); '-' = diff erences in branching pattern between Bayesian and NJ trees. Scale bar indicates substitutions per site.

Discussion
Two of the three phylogenetic analyses (ML and Bayesian) showed strong support for clade A (consisting of clade B, Rhysotina hepatzion, the Macrochlamys clade, Varadia gen. nov. and Mariaella dussumieri). We note, however, that while this clade is strongly supported (BS = 71%) in the ML tree it is only very weakly supported in the NJ tree (BS = 30%), so its robustness requires substantial further investigation (e.g., by expanded sampling of taxa and / or gene regions). The fi ve members of the Ariophantinae sensu stricto (Blanford & Godwin-Austen 1908; an almost exclusively Peninsular Indian-Sri Lankan group) lie outside clade A. Although this suggests that Varadia gen. nov. is not a member of the Ariophantinae sensu stricto, the position of Varadia gen. nov. within clade A was not compellingly resolved, and that includes uncertainty about the sister group of Varadia gen. nov.
On the basis of characters of the reproductive system, particularly the male genitalia and the gametolytic sac, we provisionally place Varadia gen. nov. in the Macrochlamydinae. This is consistent with the results of our DNA analyses and there are two key points here. First, although the sister-group relationship of Varadia gen. nov. to Macrochlamys was not strongly supported (Bayesian: PP = 0.62; ML: BS = 60%; NJ: BS = 44%), it was consistently recovered by all analyses, and is much better supported than the best supported incompatible bipartition in both the ML (BS = 12%) and NJ (BS = 14%) analyses, though, taxon sampling could be an issue here, and further investigations are required to address this. Second, in two of the three analyses, Varadia gen. nov. was strongly supported (ML: BS = 71; Bayesian: PP = 0.98) as a member of a clade uniting the helicarionids with the Macrochlamys clade and the urocyclid Rhysotina hepatzion. In V. amboliensis gen. et sp. nov., as in many Macrochlamydinae, the penial retractor muscle inserts not on the tip or distal-most end of the epiphallic caecum but on the side. Hausdorf (1998) considered the lateral insertion of the penial retractor muscle on the epiphallic caecum to be the only known autapomorphy of the Macrochlamydinae; we note, however, that in the Indian species Also, on the basis of available data, the insertion of the penial retractor muscle at two separate points (near the tip of epiphallic caecum and on the epiphallus) and the presence of a thin but noticeable band of muscle on the outer wall of the penis are unique to Varadia gen. nov.
Published data on the morphology of the penial sheath and inner wall of the penis are scarce for South Asian helicarionoideans (but see Hyman & Ponder: M. petrosa;Raheem & Naggs 2006: Ratnadvipia Godwin-Austen, 1899Bhosale et al. 2019: Eurychlamys). While detailed anatomical studies of Thai and Myanma Macrochlamydinae have been published recently (e.g., Pholyotha et al. 2018Pholyotha et al. , 2020aPholyotha et al. , 2020b, these do not include details on the structure of the penial sheath. We compared M. cf. indica from Kolkata (formerly Calcutta), West Bengal, India (4 specimens dissected; reg no. BNHS GAS 142 to 145) with V. amboliensis gen. et sp. nov., and this revealed some clear diff erences between the two. In the Kolkata examples of M. cf. indica, the penial sheath extends from the most proximal part of the penis to the area where the penis gives way to the coiled epiphallic caecum; it is attached to the penial wall at these two ends but is otherwise free. In V. amboliensis gen. et sp. nov., in contrast, the penial sheath extends over the whole of the epiphallic caecum, part of the epiphallus and a small part of the fl agellum. Interestingly, our dissections of M. cf. indica indicate that like in V. amboliensis gen. et sp. nov., the penial sheath is divisible into a proximal section that is thick and opaque, and a distal section that is thin and transparent (to our knowledge, this has not been reported before). These dissections also suggest that the section of the epiphallus nearest to the epiphallic caecum and penis is fi rmly attached to the penial sheath; again, this is similar to V. amboliensis gen. et sp. nov. Further studies of other species of Macrochlamys as well of other taxa in the subfamily Macrochlamydinae are required to investigate the generality of these observations. Internally, the penis of M. cf. indica is divisible into proximal and distal parts, with a well-developed penial verge (= penial papilla) demarcating the two; a similar arrangement has also been reported for Thai and Myanma Macrochlamys (Pholyotha et al. 2018(Pholyotha et al. , 2020b. In contrast, in V. amboliensis gen. et sp. nov., a penial verge is absent and the penis is divisible into three parts on the basis of the morphology of the inner surface. The whole spermatophore and spermatophore fragments recovered from V. amboliensis gen. et sp. nov. were found in a loosely coiled state in the gametolytic sac. These spermatophores are broadly similar with those described and fi gured for M. pedina and M. fl emingi (L. Pfeiff er, 1856) by Godwin-Austen (1899: 133, pl. 83: fi gs 5, 5a-c; pl. 87: fi gs 2, 2a-d), the capsule being relatively long and elongated in all three species. In M. pedina and M. fl emingi, however, the entire length of the tail-pipe is covered with relatively large, branching spines that point towards the capsule; in V. amboliensis gen. et sp. nov., the spines are minute and confi ned to the free, terminal part of the tail-pipe (cf. the Australian helicarionid genus Fastosarion; Hyman & Köhler 2019). Godwin-Austen (1899: 133, 134; pl. 94: 1, 2, 2a, 4) also described and fi gured the spermatophores of M. udus Godwin-Austen, 1899 and M. lecythis (Benson, 1852). In these species, the capsule is short relative to the tail-pipe and spines (of varied complexity) are present only at the two ends of the tail-pipe. Spermatophores of broadly similar morphology have also been reported for some Thai Macrochlamys species (e.g., M. aurantia Pholyotha & Panha, 2018 and M. coleus Pholyotha & Panha, 2018;Pholyotha et al. 2018). The spermatophore of V. amboliensis gen. et sp. nov., unlike those described to date for Macrochlamys, has a pronounced U-shaped bend at the junction of the capsule and tail-pipe. We note that similarly bent spermatophores are characteristic of some urocyclid taxa (Van Goethem 1977: 34-35); these urocyclids, like V. amboliensis gen. et sp. nov., also have a perforated tip to the tail-pipe.
In external morphology (the combination of large size, broad, densely tuberculated shell lobes and a shell with ca 4 whorls and a disproportionately large body whorl), V. amboliensis gen. et sp. nov. is unique among the helicarionoidean taxa of the Western Ghats. Eurychlamys platychlamys, M. indica and M. pedina are superfi cially similar, but the fi rst two are smaller (particularly E. platychlamys) and all have tightly-wound shells composed of ca 5 whorls (cf. Raheem et al. 2014: fi g. 67a-b; Bhosale et al.