Uropeltis caudomaculata, Gower & Das & Deepak & Gerard & Narayanan, 2024

Gower, David J., Das, Sandeep, Deepak, V., Gerard, Jason D. & Narayanan, Surya, 2024, A new species of Uropeltis Cuvier, 1829 (Serpentes: Uropeltidae) from the eastern escarpment of the southern Western Ghats, India, Zootaxa 5477 (3), pp. 333-357 : 336-351

Uropeltis caudomaculata sp. nov.

urn:lsid:zoobank.org:act:B404F4AC-CEF5-4225-9595-4CD47313E204

Figs. 1–6 , Tables 1–4 , Appendix 2

Chresonymy

Uropeltis dindigalensis : Ganesh et al. (2008)

Uropeltis cf. dindigalensis : Chandramouli & Ganesh (2010), Ganesh et al. (2014)

Uropeltis pulneyensis : Hutton & David (2008), Chaitanya et al. (2018), Deepak et al. (2022, in part), Ganesh et al. (2023, in part)

Uropeltis cf. pulneyensis : Ganesh et al. (2022)

Holotype ( Fig. 2 ). BMNH 1955.1.2.69 , female. Collected from “High Wavys” (= High Wavy Mountains = Meghamalai), Tamil Nadu, India ( Fig. 3 ), 3,000–5,000 feet elevation (914–1,524 m) by Angus F. Hutton. No collection date recorded in catalogue, but specimen accessioned into the London collection on 2 January, 1955, and was probably collected between 1946 and 1948 (see Hutton & David 2008) .

Paratypes (n = 7). BMNH 1955.1.2.68 (female), 1955.1.2.70 (male), 1955.1.2.71 (male), and 1955.1.2.72 (female), same data as for holotype. BNHS 3765 ( Figs. 4 , 5 ), female, collected from Manalar, Meghmalai Tiger Reserve, Tamil Nadu (9.621255° N, 77.340202° E, 1,541m asl) by J.D. Gerard on 24 th July 2021. BNHS 3763 and BNHS 3764 ( Fig. 4 ), female, collected from Yellapetty, Munnar, Kerala (10.119513° N, 77.206169° E, 1,952 m asl) by S. Das on 7 th February 2022 and 15 th March 2023, respectively.

Diagnosis. Uropeltis caudomaculata sp. nov. differs from all other congeners except U. pulneyensis and some individuals of U. bhupathyi and U. grandis (see Jins et al. 2018) in having a rostral scale that extends posteriorly to prevent midline contact between the opposite nasal scales. The new species differs from U. pulneyensis (and its current junior synonym Silybura guentheri Beddome, 1878 ) in having more ventral scales (among examined specimens, 185–187 and 187–197 versus 156–175 and 166–182 in males and females, respectively: Tables 3 and 4 ). Uropeltis caudomaculata sp. nov. differs from U. grandis in having 17 versus 19 dorsal scale rows at midbody, and in having fewer versus more than 190 ventrals. The new species differs from U. bhupathyi in having fewer than 190 versus more than 200 ventrals, and in having a much less distinct tail shield, corresponding to Smith’s (1943) shield Type III versus Type I.

Description of holotype ( Fig. 2 ). See Table 3 for morphometric and meristic data. Female with large (ca. 19 mm long), yolky ova in oviducts. Generally good condition; longitudinal incision into coelom on right of midventer, ca. 22 mm long, extending anteriorly from 12.5 mm anterior to vent. The specimen smells of wintergreen oil, sometimes added as a bitterant to surgical spirit or rubbing alcohol.

Body subcylindrical, venter slightly flattened; tapering over c. 10 mm immediately behind head; very gently tapering posteriorly up to vent, more abruptly on tail. Head strongly tapering in dorsal view, sides straight to slightly convex, pointed; tapering also in lateral view, gently with straight edges to anterior margin of prefrontal then much more abruptly to tip of rostral.

Rostral rounded anteriorly; longer than wide in dorsal view, about as long as deep in lateral view; at widest slightly in front of and below nostril. Rostral approximately twice as long (in dorsal view) than rostral-frontal gap. Frontal six-sided; lateralmost edges in contact with oculars strongly diverging anteriorly. Frontal much longer, wider than rostral. Paired nasals not in contact behind rostral. Left prefrontal overlaps right along midline. External naris small, subcircular, slightly countersunk within small depression, located in anteroventral corner of undivided nasal shield. Nasal shield subequal in length and height, contacts first and second supralabials (SLs). Prefrontal about as wide as long, much shorter than frontal, contacts SL2 and SL3. Supralabials four, SL1 smallest, making least contribution to margin of mouth; SL2 larger than SL1, smaller than SL3; SL4 much the largest and longest. Ocular contacts SL3 and SL4. Eye small but distinct, diameter approximately 0.4 times length of ocular shield, located near anteroventral corner of ocular; bulges very slightly from ocular surface, pupil appears subcircular; eyes directed antero- (and very slightly dorso-) laterally. Seven maxillary and eight mandibular teeth on each side. Teeth simple, pointed, distinctly retrorse, straight to slightly posteriorly curved distally; evenly spaced, smallest posteriorly.

Paired parietals slightly shorter than frontal, posteriorly broadly rounded. Opposite parietals in brief midline contact, left overlapping right. Parietals much longer than wide, wider than frontal and rostral. Each parietal contacts four scales other than head shields. Infralabials (ILs) 3,3, IL2 slightly longer than IL1, more notably longer than IL3. Left and right first infralabials briefly in midline contact, separating small, slightly prominent mental from first midline ventral scale. First and second ventrals (in contact) and third longer than wide, fourth about as long as wide, fifth and subsequent ventrals wider than long.

Head and body scales macroscopically smooth, lacking keels, except on far posterior of tail. Body scales generally evenly sized on dorsum and along body except for those involved in dorsal scale row reductions. Midline ventral scales between mental and anal of even size though anterior- and posteriormost ones gradually narrow, posteriormost ventral with V-shaped posterior margin (gently rounded in preceding ventrals). Ventrals 187, though could be interpreted as 188 if the small posteriormost element on the left side is counted. Ventrals at midbody approximately 1.6 times as broad as exposed part of adjacent first dorsal scale row. Dorsal scale rows 19 immediately behind head, reducing to 17 by level with 31 st ventral, maintained thereafter along most of body with minor fluctuation on right between 50 th and 60 th ventral:

4+5 (31) +4 (50), 4+5 (53), +4 (57), -4 (58) 19 --------------- 17 ---------------------------------------------- 17 4+5 (28) [b3, s4: 53]

Dorsal scale rows 13 at base of tail. Paired anal scales (right overlying left) considerably larger than posteriormost ventrals and subcaudals. Distal margin of each anal overlaps three other scales in addition to anteriormost subcaudals. Subcaudals 8,7. Posteriormost tail scales (most notably on dorsum) bear few, very short and low, subparallel ridgelike tubercles.

Tail shield not apparent; terminus most closely matching Type III among Smith’s (1943) states for Uropeltis tails (also Pyron et al. ’s [2016] Type III). Terminal scute longer and taller than wide, surrounded by nine scales including last subcaudals; ventrally convex, smooth; dorsally concave shortly anterior to terminus; laterally with few, small, low tubercles. Posterior edge of terminal scute ends in pair of narrowly separated (0.7 mm), paramedian, short, inconspicuous points; these not disposed on thin horizontal plate-like ridge.

In alcohol, venter dark (blackish) brown with mostly asymmetrical and somewhat irregularly spaced pale yellow crossbars, approximately one (to two) scales long, mostly extending from midline across first two to three dorsal scale rows; anteriorly these cross bars do not encroach onto ventrals or first dorsal scale rows, instead on dorsal scale rows 4–5, 3–5, 2–4 or 2–3 (from anterior to posterior). On body, 27 and 25 pale cross bars on left and right, respectively. Posteriormost cross bar level with and encroaching slightly onto lateralmost edges of anal shields. Some ventral pale bars, pale flecking on ventrals and on dorsal scale rows 1–3 but overall bicoloured; mostly dark below, pale above. Many scales with pale yellow to off-white posterolateral edges. Pale dorsum in the form of an irregularly edged but large (more than 50% of area of each scale) mottled pale tan spot on every scale of (9–)11 dorsalmost scale rows, except for anteriormost 60 mm and posteriormost 30 mm. Tail dark brown except for yellow on posteriormost two thirds of terminal scute. Scales not notably iridescent. Underside of head brown except for pale yellow flecks on first infralabials and anterior of mental shield. Dorsally, head dark olive-brown; diffuse pale tip to rostral; large pale yellow spots covering posterior half of each posteriormost (fourth) supralabial.

Variation in paratypes. Some morphological data and images presented in Fig. 4 , Table 3 , and Appendix 2. Among the four BMNH specimens, 1955.1.2.68 and more notably 1955.1.2.71 and 72 are somewhat dehydrated. There are longitudinal incisions in the tails of 1955.1.2.70 and 72, and a long ventral incision into the coelom of 1955.1.2.68, which also has some skin on the left side of the head reflected by dissection. BMNH 1955.1.2.70 was preserved close to ecdysis. BMNH 1955.1.2.70 and 71 both have small subrectangular supernumerary shields on the temporal region of the head region—between the fourth supralabial and parietal, immediately behind the ocular. In association with this scalation pattern, the parietal shields in these two specimens contact 4,5 and 5,6 (rather than 4,4) other head shields, respectively. In BMNH 1955.1.2.68 the anal shields are ‘fused’, though at the distal edge the right appears to slightly overlap the left (as in all other specimens). In BMNH 1955.1.2.71 the right prefrontal overlaps the left rather than vice versa. In 1955.1.2.68 three of the subcaudals are unpaired (‘fused’). As with the holotype, the female BMNH paratypes have some low and inconspicuous, longitudinal ridges on some of the tail scales, especially more dorsally and posteriorly (not on the subcaudals). These ridges are slightly more prominent in the male BMNH paratypes, more so in 1955.1.2.71 than in 1955.1.2.70, and are more widespread, occurring also on the anal shields and a few of the posteriormost ventrals. The pale mottled spots on the dorsalmost 9–11 scales are more or less conspicuous; as conspicuous as in hlotype in 1955.1.2.71 and less conspicuous in1955.1.2.68 and 72.

Among the BNHS paratypes, BNHS 3765 is slightly dehydrated, with one ventral incision into the coelom at midbody (extending across six ventral shields) and one close to the anus (extending across four ventrals). BNHS 3763 is well preserved, with a coelomic incision at midbody covering six ventrals. BNHS 3764 is slightly dehydrated, with one coelomic incision anterior to midbody (15 ventrals) and one posterior to midbody (8 ventrals). The right parietal in 3763 contacts only three head shields. The parietals are posteriorly narrow in 3765 and 3763. The third ventral is much shorter and smaller than the first two in 3763. The last (posteriormost) ventral shield is ‘divided’ in 3763, with one paramedian scale on the right and two small scales on the left. The tips of the points on the terminal scute are blunt in BNHS 3763 and 3764. BNHS 3765 resembles the holotype in its colour pattern, but BNHS 3763 and 3764 differ in having a lateral yellow stripe (instead of yellow spots) from the head to anterior of the body and on the base of the tail.

Colour in life. See Fig. 5 . The darker background is generally dark brown. The pale spots and mottling on the uppermost 9–11 dorsal scale rows are olive-green to pale yellow. The larger pale spots (or stripes) on and behind the lateral surface of the head, along the ventrolateral and lateral aspect of the body, and on the lateral surface of (at least the base of) the tail are bright yellow, more vivid than the more diffuse dorsal pale markings.

Sexual dimorphism. See Fig. 6 . The available sample size is too small to confirm probable sexual dimorphism in numbers of ventral shields, but the data do show that males have relatively longer tails with more subcaudals than females.

Etymology. The name of the new species is in reference to the lateral yellow spot on each side of the base of the tail (in specimens from Meghamalai), rather than a stripe as occurs in many congeners, including, typically, its closest known relative, U. pulneyensis . For nomenclatural purposes, the species epithet is a noun in apposition.

Suggested vernacular names. Tailspot uropeltis or Tailspot shieldtail (English).

Distribution, habitat and conservation. The new species is known only from three localities, Meghamalai Tiger Reserve in Tamil Nadu state, Periyar Tiger Reserve, and Yellapetty, Munnar in Kerala state ( Fig. 2 ). The holotype and four BMNH paratypes were probably collected between 1946 and 1948 (see Hutton & David 2008), according to the BMNH catalogue from 3,000 –5,000 feet (914–1,524 m) elevation (but according to Hutton & David [2008], 914–1,066 m). It is not possible to be certain that the snakes from Meghamalai reported as U. pulneyensis by Hutton (1949) are actually U. caudomaculata sp. nov. because no scalation or colour pattern data were presented. However, Hutton (1949) reported the species identified by him as U. pulneyensis to be relatively common, to be encountered on roads, in tea nurseries and under or in decaying wood and heaps of stones, and to be viviparous, giving birth to 8 to 17 young of approximately 60 mm total length.

Paratype BNHS 3765 was a roadkill specimen collected from the Manalar region of Meghamalai ( Fig. 7 ), bordered by tea plantations, with sparsely distributed windbreak trees and the Manalar catchment on the other side. Three other (uncollected) roadkill individuals were sighted by JDG during the day between tea plantation and forest or reservoir in June and July, 2021 near Meghamalai village (9.67789° N, 77.38296° E; 9.65795° N, 77.36677° E; 9.61945° N, 77.34194° E, all at ca. 1,500 m elevation). Four uncollected individuals were found by SD and colleagues while rolling logs within Periyar Tiger Reserve (9.587788° N, 77.320336° E, 1721m) on 22–23 July 2017. One of these individuals is illustrated in Fig. 5 . Ventrals and other scales of this individual were not counted in the field but from photographs it is a female with approximately 7,7 subcaudal scales and has 200 middorsal scales (compared with 198 and ca. 186 in the female holotypes of U. caudomaculata sp. nov. and U. pulneyensis , respectively). Paratypes BNHS 3763 & 3764 were collected from secondary forest in the Yellapetty area at an elevation of ca. 1,950 m. Both individuals were collected during the day under logs within the forest area. This site is surrounded by tea plantations on one side and hill slope on the other. The general habitat ( Fig. 7 ) of this area is shola-forest mosaic.

Uropeltis caudomaculata sp. nov. is known from multiple localities up to ca. 60 km apart, and though threatened at least locally by road traffic it appears able to tolerate at least some human-mediated environmental disturbance. However, there is a notable intervening gap between two clusters of locality records ( Fig. 3 ), only a single point locality is known at the northernmost edge of its recorded range, and very little is known about its ecological tolerances or preferences. Thus, we suggest that for the present, the species is likely to qualify for Data Deficient status under IUCN Red List criteria ( IUCN Species Survival Commission 2012).

Remarks. Beddome (1863) did not specify the number of specimens available to him in his original description of Uropeltis pulneyensis (as Plectrurus pulneyensis ). It can be inferred that there were likely at least two (probably at least three) type specimens, given that Beddome reported the number of subcaudals as “male, about 12, female 6–8”. Subsequent workers have suggested up to nine syntypes, currently stored in Kolkata (ZSI 4381 & 6948 from “Palni Hills”; 6950 & 6972–75 from “Madras Presidency”: but see Gans [1966], who reported that there were no ZSI types of this species), London ( BMNH 1946.1.17.6, previously BMNH 64.3.9.1, “Anamallays” according to accession register but Pulney Hills according to Boulenger 1863: Fig. 8 ), and Paris (MNHN 1895.78, Pulney Hills) ( Das et al. 1998; McDiarmid et al. 1999; Pyron et al. 2016; Ganesh et al. 2023). We did not examine ZSI specimens, but MNHN 1895.78 has 12,11 subcaudals and BMNH 1946.1.17.6 has 6,7 subcaudals so both could be from Beddome’s type series. The accession date of the BMNH specimen (March 1864) is also consistent with this. Given that we have examined it in detail, that it is in good condition (unlike the ZSI specimens: Ganesh et al. 2023), and colour photographs are provided here ( Fig. 8 ), we designate BMNH 1946.1.17.6 (previously BMNH 64.3.9.1) as the lectotype of Plectrurus pulneyensis Beddome, 1863 . At least MNHN 1895.78 is a paralectotype, other paralectotypes are likely present in the ZSI collection (e.g., Ganesh et al. 2023). Beddome (1886) thought there was no sexual dimorphism in numbers of ventrals in U. pulneyensis , but his sample size might not have been large, and examination of more specimens shows that females, on average, have more ventrals as well as nonoverlapping numbers of subcaudals ( Fig. 6 ; see also Boettger 1892, noting that it is not clear how the anteriormost ventral shield was identified in that study).

As far as we know, U. pulneyensis occurs only in the Palani Hills, the westernmost specimen records that we know of being Kurduvadi (Kadavari?; ca. 10.216° N, 77.286° E, ca. 2,200 m elevation according to Google Earth), Kukkal (Kookal; ca. 10.295° N, 77.365° E, ca. 1,800 m elevation according to Google Earth) and Marian Shola (ca. 10.155° N, 77.77.348° E, ca. 2,200 m elevation according to Google Earth). The imprecise locality records of “near Madras” (MCZ 3870, 47042–4703; AMNH 77607) and “Nilgherries” (ZMB 5538) are here considered questionable. Ganesh et al. (2023) additionally listed localities for U. pulneyensis in Meghamalai (High Wavy Mountains) and in Nilambur. We have not examined the reported voucher specimens, and Ganesh et al. (2023) presented ranges of scalation data rather than disaggregated for individual specimens. However, photographs of Meghamalai specimens previously published appear to be of U. caudomaculata sp. nov. based on colour and number of midline scales (see Ganesh et al. 2022; specimens identified as U. cf. pulneyensis ). Nilambur (11.276944° N, 76.225833° E) is highly disjunct from verified localities for both U. pulneyensis and U. caudomaculata sp. nov., lying at only ca. 400 m elevation, more than 150 km north of Munnar on the other side of the lowland Palghat (Palakkad) Gap, so we consider that this locality is doubtful for either species. Four MNHN specimens are reportedly from Kurduvadi (Appendix 1), though it seems extremely unlikely that this is the town in southern Maharashtra (18.097° N, 75.425° E), almost 900 km to the north of the Palani Hills at below 600 m elevation in the Deccan Plateau rather than Western Ghats.

Rajendran (1985) included a population from Sevenmalai, near Munnar under U. pulneyensis (see also Ganesh et al. 2023), but we have seen no specimens of U. pulneyensis from in or very close to Munnar, and all of the Uropeltis specimens we have seen from that locality that have a tail morphology generally like that of U. pulneyensis were subsequently identified as U. tricuspida Gower, 2023 ( Gower 2023). Rajendran (1985) also reported one specimen of U. pulneyensis from a 250 m elevation site in the Alargarkoyil hills (ca. 10.12° N, 78.23° E), a range below 900 m elevation near Madurai, disjunct from the main ranges of the Western Ghats. We do not know of any museum specimens or published images of uropeltids from the Alargarkoyil Hills and are so unable to comment on the identity of that population. The known distributions of U. pulneyensis and U. caudomaculata sp. nov. do not overlap (assuming that Silybura guentheri is not a synonym of U. pulneyensis : see below), but the westernmost known localities of U. pulneyensis (Kadavari, Kukkal and Marian Shola) , and northernmost-known locality of U. caudomaculata sp. nov. (Yellapetty, Munnar) are only approximately 14, 16 and 25 km apart, respectively, so it is possible that the two species are sympatric somewhere in this region. In Deepak et al. ’s (2022: 55) map of the distribution of U. pulneyensis , the westernmost dot (Munnar) is Rajendran’s (1985) record from Sevenmalai that possibly pertains instead to U. tricuspida (see Gower 2023) and the dot shortly to the east of that (Yellapetty) and in Meghamalai pertain instead to U. caudomaculata sp. nov.

Hutton & David (2008) identified the BMNH type specimens of U. caudomaculata sp. nov. as U. pulneyensis . Hutton & David (2008: 302) mentioned seven BMNH specimens of U. pulneyensis from Meghamalai, but we assume that this is an inadvertent error because they reported only the same five BMNH specimens that we report here. The sex, size and scalation data presented by Hutton & David (2008) for these five specimens are greatly at odds with our observations (see Table 3 ). We have no explanation for these discrepancies, even allowing for some potential mixing of data and/or specimen tags among the five specimens subsequent to Hutton & David’s observations.

All of the Uropeltis caudomaculata sp. nov. specimens we have examined from Meghamalai differ from U. pulneyensis in having yellow spots or blotches rather than stripes on the lateral surface of the base of the tail and on the back of the head and anterior end of the body. The two collected specimens of the new species from Yellapetty near Munnar resemble U. pulneyensis in this feature, but are identified here as U. caudomaculata sp. nov. because they much more closely match the Meghamalai material in terms of numbers of ventral shields, and in mitochondrial DNA sequences.

Beddome (1878) decribed Silybura guentheri on the basis of a single specimen ( BMNH 1946.1.16.74, previously BMNH 83.1.12.34: Fig. 9 ) from High Wavy Mountains (Meghamalai). This species was listed as a junior subjective synonym of Silybura pulneyensis by Boulenger (1893), which seems to have been followed by subsequent taxonomic works (e.g., Smith 1943; Gans 1966; McDiarmid et al. 1999; Pyron et al. 2016). The holotype of S. guentheri does, indeed, resemble U. pulneyensis in having the nasal shields separated by the rostral, 17 dorsal scale rows at midbody, a tail matching Smith’s (1943) Group III, and in being a male with 167 ventral shields and 12,12 subcaudals ( Fig. 6 ). However, as well as originating from Meghamalai versus the Palani Hills, the specimen differs from U. pulneyensis notably in colour, as also noted by Beddome (1878, 1886) and Boulenger (1893) —having an almost entirely pale venter ( Fig. 9 ), having a stripe (versus spots) on the head and tail (in which it differs from at least the Meghamalai specimens of U. caudomaculata sp. nov.) and with pale spots on the uppermost dorsal scale rows being less dense (more mottled). It is not only colour and number of ventral scales in which the S. guentheri holotype differs from U. caudomaculata sp. nov. —the former has relatively larger eyes (compared to, for example, the length of the ocular shield: Table 3 ), and also lacks midline contact between the opposite first infralabial shields (which make contact in all but one of the known specimens of U. caudomaculata sp. nov.; rare outliers for this feature occur in U. pulneyensis : present in only 2 out of 25 of the specimens examined here for that feature). Pending further data, we consider Uropeltis guentheri comb. nov. to be a valid species, most similar to U. pulneyensis in external morphology but (if historical locality data are correct) occurring in Meghamalai rather than the Palani Hills. Other than finding additional specimens that resemble the only known specimen of U. guentheri , additional data might come in the form of internal anatomy (including osteology) and/or DNA sequences for the holotype.

There are, to the best of our knowledge, no other voucher specimens of the new species other than the eight specimens reported here, though we suspect that several observations of U. caudomaculata sp. nov. have been reported previously under various identifications. Probable observations of U. caudomaculata sp. nov. from Meghamalai were reported by Ganesh et al. (2008) and Chandramouli & Ganesh (2010; also Ganesh et al. 2014) as U. dindigalensis and U. cf. dindigalensis , respectively, and as U. pulneyensis by Chaitanya et al. (2018), and as U. cf. pulneyensis by Ganesh et al. (2022). Ganesh et al. (2022) reported that one uncollected specimen had approximately 190 ventral shields, 17 dorsal scale rows at midbody, and a tail matching Smith’s (1943) Group III—data that, along with the colour pattern seen in published photographs ( Ganesh et al. 2014: fig. 6a; Chandramouli & Ganesh 2010: fig. 16; Chaitanya et al. 2018: fig. 3a), allow us to reasonably confidently identify this as U. caudomaculata sp. nov. The photograph of U. pulneyensis provided by Deepak et al. (2022: 55) and reproduced here in Fig. 5A is of one of the paratypes of U. caudomaculata sp. nov. (BNHS 3765).

Figures

FIGURE 1. Maximum Likelihood (ML) phylogeny for selected samples of Uropeltis. Numbers on internal branches are Bayesian posterior probability/ML bootstrap support values greater than 0.9/60. See Table 1 for data sources.

FIGURE 2. Holotype of Uropeltis caudomaculata sp. nov. (BMNH 1955.1.2.69). Scale bar increments in millimetres. Photographs by Kevin Webb.

FIGURE 3. Map indicating localities for Uropeltis pulneyensis and U. caudomaculata sp. nov. Locality symbols with a black dot inside are associated with voucher specimens; those symbols without a black dot indicate where specimens have been observed but not collected. The imprecise type locality for (the only known specimen of) U. guentheri comb. nov. is ca. 1,500 m, possibly in or close to Meghamalai (“Cumbum Hills”).

FIGURE 4. The three most-recently collected paratypes of Uropeltis caudomaculata sp. nov. in approximately dorsal (left column) and ventral (right column) views. A & B, BNHS 3765, C & D, BNHS 3763, and E & F, BNHS 3764.

FIGURE 5. Uropeltis caudomaculata sp. nov. in life. A. Paratype BNHS 3765 from Manalar, Meghamalai, Tamil Nadu; photograph by JDG. B. Uncollected live individual from Periyar Tiger Reserve; photograph by SD. C. Paratype BNHS 3764 from Yellapetty; photograph by Sidharth S.

FIGURE 6. Variation in numbers of ventral and subcaudal shields in females (circles) and males (triangles) of Uropeltis pulneyensis (25 females, 19 males), U. caudomaculata sp. nov. (6 females, 2 males) and the holotype of U. guentheri comb. nov. (one male). See Tables 3 and 4 for sources of data.

FIGURE 7. Habitats of recent localities for Uropeltis caudomaculata sp. nov. A. Locality where paratype BNHS 3765 was collected, at Manalar, Meghamalai, Tamil Nadu. B. General habitat close to where paratypes BNHS 3763 and 3764 were collected; Yellapetty, Munnar, Kerala. Photographs by JDG (A) and SD (B).

FIGURE 8. Lectotype (by present designation) of Uropeltis pulneyensis (Boulenger, 1863) (BMNH 1946.1.17.6, previously BMNH 64.3.9.1). Scale bar increments in millimetres. Photographs by Kevin Webb.

FIGURE 9. Holotype of Uropeltis guentheri (Beddome, 1878) (BMNH 1946.1.16.74, previously BMNH 83.1.12.34). Scale bar increments in millimetres. Photographs by Kevin Webb.