The status of the nomen Hyperolius guttatus Peters, 1875 (Amphibia: Anura) and allied nomina

Transcription

1 RESEARCH ARTICLE 2014 VOLUME 30 PAGES The status of the nomen Hyperolius guttatus Peters, 1875 (Amphibia: Anura) and allied nomina Thierry Frétey 1, Alain Dubois 2, Annemarie Ohler 2 * 1. Association RACINE, 5, allée des cygnes, Saint Maugan, France 2. Reptiles & Amphibiens, UMR 7205 OSEB, Département de Systématique et Evolution, Muséum National d Histoire Naturelle, 25 rue Cuvier, CP 30, Paris, France The nomenclatural status of the nomen Hyperolius guttatus Peters, 1875 was investigated. The specimens on which the original description was based were collected at Bootry in Ghana and in Cameroon, but their precise collection localities are unknown. We established the chronology of the successive taxonomic allocations of the specimens from Ghana and Cameroon. The study of old documents allowed to identify, without ambiguity, the type locality of Hyperolius guttatus in Cameroon as the vicinity of Douala. This also applies to several other species of anurans and snakes described in the same publication, which are listed in the Appendix. An analysis of the colour pattern and of morphometrical characters allowed us to compare the specimen of Hyperolius guttatus from Cameroon with the other species of Hyperolius living in the same region, and to ascertain the status of this nomen. Hyperolius guttatus is proposed as the nomen of a subspecies of Hyperolius concolor from Cameroon. The syntypes from Boutry (Ghana) are confirmed to be Hyperolius fusciventris burtonii. We redescribe the type specimens of Hyperolius guttatus Peter, 1875 and Hyperolius pulcher Ahl, 1931 to present evidence for the synonymy. INTRODUCTION In 1875, W. C. H. Peters presented to the Academy of Sciences of Berlin several species of Amphibians and Reptiles collected by R. Buchholz in West and Central Africa. Among them, he described Hyperolius guttatus on several specimens (syntypes or symphoronts; Dubois, 2005: 403) coming from Cameroon and Ghana (Peters, 1875: 207). These specimens are stored in the Museum für Naturkunde, Leibniz Institute for Research on Evolution and Biodiversity (ZMB; formerly Zoologisches Museum at Berlin) and the Netherlands Centre for Biodiversity Naturalis at Leiden (RMNH). They have the following collection numbers: ZMB 8378 for the specimen of Cameroon (Bauer et al., 1995: 44) and ZMB 4489 and RMNH RENA.1788 for those from Ghana (de Ruiter, pers. comm.). Ahl (1931) only considered the specimens from the Museum für Naturkunde Berlin and ever since these two specimens are considered to be two distinct species. The sample from Ghana does not require further investigations and has been attributed to Hyperolius fusciventris burtonii (Boulenger, 1883) (Schiøtz, 1967), yet the one from Cameroon seems more problematic because the taxonomic conclusions are not consensual (e. g., Perret, 1975, 1976; Lötters et al., 2001) and the description of the type locality (onymotope; Dubois, 2005: 404), Cameruns seems too general to indicate the onymotopic population (sensu Dubois & Ohler, 1997; Dubois, 2005: 381). MATERIALS AND METHODS This published work and the nomenclatural acts it contains have been registered in ZooBank. The ZooBank LSIDs (Life Science Identifiers) can be resolved and the associated information viewed through any standard web browser. The LSID for this publication is: urn:lsid:zoobank.org:pub:a23cef58-b23c-4762-bc ec09e3be Received 18 May 2012 *Corresponding author Accepted 18 February 2013 Published Online 31 January 2014 ISSCA and authors 2014 annemarie.ohler@mnhn.fr

2 THIERRY FRÉTEY et al. List of abbreviations BMNH: Natural History Museum, London, United Kingdom. JLA: Jean-Louis Amiet Collection, Nyons, France. MNHN: Muséum national d Histoire naturelle, Paris, France. RMNH: Netherlands Centre for Biodiversity Naturalis, Leiden, Netherlands. ZMB: Museum für Naturkunde, Berlin, Germany. Nomenclatural terms used Holophoront, holotype (Dubois, 2005). Symphoront, syntype (Dubois, 2005). Lectophoront, lectotype (Dubois, 2005). Onymotope, type locality (Dubois, 2005). Specimens studied See Appendix 1. Measurements used in the descriptions and the morphometrical analyses The following measurements were taken on 111 specimens with a slide caliper or the ocular micrometer of a binocular microscope by the same observer (AO); the measurements in bold were used in multivariate analyses: SVL: Snout-vent length. Head: HW, head width (largest width of the head, at level of jaw articulations); HL, head length (distance from back of mandible to tip of snout); MN, mandible to nostril (distance from back of mandible to posterior border of nostril); MFE, mandible to front of eye (distance from back of mandible to anterior margin of eye); MBE, mandible to back of eye (distance from back of mandible to posterior margin of eye); IFE, inter front of eyes (distance between anterior margins of eyes); IBE, inter back of eyes (distance between posterior margins of eyes); IN, internaral space (distance between internal margins of nostrils); EN, eye to nostril (distance between posterior margin of nostrils and anterior margin of eye); EL, eye length; NS, snout to nostril (distance between tip of snout and anterior margin of nostril); SL, snout length (distance between tip of snout and anterior margin of eye); TYD, greatest tympanum diameter; TYE, tympanum to eye (distance between anterior margin of tympanum and posterior margin of eye); IUE, inter upper eyelid (minimum distance between inner margins of eyelid); UEW, upper eyelid width) (maximum width of upper eyelid). Forelimbs: HAL, hand length (length of hand from basis of external palmar tubercle to tip of finger III; FLL, forelimb length (length of forelimb between elbow and base of external palmar tubercle); TFL, third finger length (length of finger III from basis of proximal subarticular tubercle); PA1 to PA4, pad width fingers I to IV; WA1 to WA4, width of fingers I to IV; AW, anterior web (distance from basis of external palmar tubercle to maximum concave curvature of web between fingers III and IV). Hindlimbs: TL, tibia length; TW, tibia width (maximal width of shank); FOL, foot length (length of foot between proximal margin of metatarsal tubercle and tip of toe IV); TFOL, tarsus-foot length (length of tarsus and foot combined); FTL, fourth toe length (length of toe IV from base of proximal subarticular tubercle); FL, femur length (length of thigh between vent and tip of knee); PP1 to PP5, pad width toes I to V; WP1 to WP5, width of toe I to V; IMT, inner metatarsal tubercle length; ITL, inner toe length. Webbing: MTTF, distance from distal edge of metatarsal tubercle to maximum concave curvature of web between third and fourth toe; TFTF, distance from maximum concave curvature of web between third and fourth toe to tip of fourth toe; MTFF, distance from distal edge of metatarsal tubercle to maximum concave curvature of web between fourth and fifth toe; FFTF, distance from maximum concave curvature of web between fourth and fifth toe to tip of fourth toe. Statistical analyses The measurements were transformed into ratios of SVL to take growth allometry into account. A principal 12

3 ALYTES component analysis was performed using SPSS program. A scatterplot was drawn with the factors 1 and 2 obtained from the principal component analysis to visualize the distribution of specimens along the first two axes. Citation of references The three authors of this paper consider the Code as an anonymous work, following the definition of anonymous work given in the Glossary of the Code itself (ICZN 1999: 100), followed in all issues of Alytes published so far over 30 years, but the citation as ICZN was suggested by the International Commission on Zoological Nomenclature and followed by the current editor of the journal. Additionally, the suppression of the paginations of books in the bibliographic references was determined by the editor. RESULTS The onomatophores of Hyperolius guttatus Peters, 1875 In 1875 Wilhelm Carl Hartwig Peters cited specimens under the nomen Hyperolius guttatus, which he credited to Schlegel. Because this last author never published this nomen, Peters (1875), who is the author of the description, is the author of the nomen according to the International Code of Zoological Nomenclature (further named Code ; ICZN, 1999). The description of Peters was based on a specimen collected by Reinhold Buchholz at Cameruns which the author compared with a specimen originating from Boutry. Peters considered this specimen as one of the original specimens in the possession of Schlegel, for which he had received a sample in exchange: welches mit einem Originalexemplar aus Boutry übereinstimmt [ which corresponds to one of the original samples from Boutry ]. This sentence clearly indicates that Peters considered that several other specimens were available to Schlegel when creating this nomen. If he had thought that a single specimen came from Boutry, he would have written mit dem Originalexemplar [ which corresponds to the original sample from Boutry ]. In conclusion, the nomen Hyperolius guttatus Peters, 1875 was proposed for several syntypes: the specimen from Cameruns ZMB 8378, primary syntype (sensu Dubois & Ohler, 1997), the specimen from Boutry ZMB 4489 which he had obtained in exchange from Leiden, primary syntype, and one or several other specimens, secondary syntypes. The collection of Leiden currently holds four other specimens from Boutry under the collection number RMNH RENA.1788; they had been collected by Hendrik Severinus Pel and entered in the collection with the nomen Hyperolius guttatus. On the jar of these specimens, an old label indicates Hyla guttata and what appears to be the year 1848 (de Ruiter, pers. comm.), the period during which Pel was holding a position at Boutry (Holthuis, 1968). These specimens would then be secondary syntypes of Hyperolius guttatus Peters, Origin and status of the specimens from Ghana Beside the sample from Cameroon, Peters (1875: 207) cited a sample originating from Boutry, now Bootry (04 49 N, W) in Ghana, stating that this original sample [ Originalexemplar ] coincides [ übereinstimmt ] with the specimen from Cameroun and is part of a series of specimens studied by Schlegel. As we have seen, there Figure 1. Drawing of Ahl (1931) (A) and photography of the paralectotype (B) (one of the original primary symphoronts) of Hyperolius guttatus Peters, 1875, ZMB 4489, young female, SVL 21.3 mm, from Boutry, Ghana. 13

4 THIERRY FRÉTEY et al. are four other specimens, RMNH RENA.1788 from this locality, which are secondary syntypes. Ahl (1931: 354) kept the nomen Hyperolius guttatus for the sample of Boutry in the Berlin Museum, of which he gave a drawing (figure 229) and a detailed description; he indicated the size of the specimen as 21 mm. Article 74.5 of the Code (ICZN, 1999) allows a designation of lectotype by non-ambiguous selection of a particular syntype. Did Ahl designate a lectotype by restricting the species Hyperolius guttatus to the specimen from Boutry? Mertens (1938: 28) and Laurent (1950: 107) considered this designation as valid. But later Laurent (1961) went back on his decision and designated the specimen from Cameruns as lectotype. Given that the species was based on several syntypes originating from Boutry, the nomenclatural action of Ahl (1931) cannot be considered unambiguous because he did not give any indication allowing individualisation of the chosen specimen. The drawing that he presented (Ahl, 1931: 354, figure 229) is too schematic to allow identification of the specimen. In conclusion, Ahl (1931) did not designate a lectotype for Hyperolius guttatus Peters, 1875 (see fig. 1). Laurent (1958: 287) attributed this specimen to Hyperolius rosaceus described by Ahl (1931: 379) from Klein Popo (Togo). He wrote: L exemplaire de Boutry, bien que qualifié d original (sans doute parce que Schlegel lui avait donné un nom qui n a d ailleurs jamais été publié par lui) est simplement comparé à celui du Cameroun. Il a donc tout au plus la valeur d un paratype et c est bien à tort que Ahl (1931) lui attribue le sens restreint de H. guttatus Peters [ the sample from Boutry, although qualified as original (without doubt because Schlegel gave it a name which however was never published by him), is merely compared with that from Cameroon. It has thus at most the value of a paratype and it is by error that Ahl (1931) attributes it the restricted meaning of H. guttatus Peters ] (Laurent, 1958: 288). Later, Laurent (1961: 70, 71) changed his opinion and considered this specimen as a probable member of Hyperolius wermuthi which he described. He also corrected his error by citing in the synonymy of this new species: Hyperolius rosaceus (non Ahl) Laurent He compared his material to the Paratype de H. guttatus, for which he indicated some ratios of measurements. However he added: Il s agira donc de voir, lorsqu on pourra étudier des séries du Ghana, s il n y existe pas une forme correspondant mieux au spécimen de Boutry. [ It should be explored, when series of specimens from Ghana are available for study, if there is a form there corresponding better to the Boutry specimen ] (Laurent, 1961: 72). Finally, Schiøtz (1967: 149, 159), attributed this sample to the taxon Hyperolius fusciventris burtoni [sic] described by Boulenger (1883: 163) under the nomen Rappia Burtonii. Schiøtz (1967: 159) detailed that the specimen was: H. f. burtoni phase J. Hyperolius wermuthi is not present in Ghana, its distribution being limited to Ivory Coast, Liberia and Guinea (Schiøtz, 1967: 155) but this species can sometimes resemble Hyperolius fusciventris, particularly with regards to its coloration and size. Measurements and colour characters of the species allied to Hyperolius fusciventris are given by Rödel (1999). Table 1. Measurements of specimens collected by Hendrik Severinus Pel at Bootry, Ghana, considered as Hyla guttata by Schlegel (unpublished) and symphoronts of Hyperolius guttatus Peters (1875). 14 ZMB4489 young female RMNH 1788.A adult male RMNH 1788.B adult male RMNH 1788.C adult male RMNH 1788.D adult male AW EL EN FL HL HW IFE IN IUE MTTF SL SVL TFL TL UEW

5 ALYTES Figure 2. Drawing of Peters (1875) (A) and of Ahl (1931) (B) and photograph of the specimen ZMB 8378 (C), juvenile female, SVL 23.0 mm, lectophoront of Hyperolius guttatus Peters, 1875 and holophoront of Hyperolius hildebrandti Ahl, 1931, from Douala, Cameroon. The four syntypes from the Leiden Musem are adult males of Hyperolius fusciventris burtonii showing phase J (phase A of Schiøtz, 1963: plate I, figure 11): the back is bluish brown with round spots, a dark canthal line and white, fine laterodorsal lines; the belly shows rare brown fleckings. The four specimens are very similar morphologically and in their colour pattern (tab. 1, fig. 2). History of the nomina attributed to the specimen from Cameroon In his description of Hyperolius guttatus, Peters (1875: 207) cited a specimen originating from Cameruns, provided an illustration (Peters, 1875: plate 2, figure 3) and indicating its size as 21 mm. Ahl (1931: 334) described Hyperolius hildebrandti based on this sample of Hyperolius guttatus from Cameroon. He reproduced the figure of Peters (1875: plate 2, figure 3) to represent his new species (figure 207), giving a size of 22 mm. Laurent (1950: 107) referred this specimen to a subspecies Hyperolius ocellatus hildebrandti. He specified that: La dénomination hildebrandti Ahl, [est] basée visiblement sur une jeune femelle [ The denomination hildebrandi Ahl, [is] obviously based on a young female] (Laurent, 1950: 109). Laurent (1958: 288), in his discussion relating to Hyperolius rosaceus, noted that: La description de H. guttatus Peters (1875) a été basée sur un spécimen du Cameroun qui a été figuré (plate II, figure 3) et dont les mensurations sont précisées (p. 207) [ The description of H. guttatus Peters (1875) was based on a specimen from Cameroon that has been drawn (plate II, figure 3) and for which the measurements have been indicated (p. 207) ]. He added just below that Hyperolius hildebrandti, described by Ahl (1931), is a synonym of Hyperolius guttatus Peters. Finally, he stipulated in a footnote: Mes références à H. ocellatus hildebrandti Ahl [ ], dues au fait que l erreur de Ahl m avait échappé, doivent donc se lire H. ocellatus guttatus Peters [ My references to H. ocellatus hildebrandi Ahl [ ], given the fact that the mistake of Ahl had escaped me, must read H. ocellatus guttatus Peters ]. Laurent (1961: 73) considered this specimen under the nomen Hyperolius ocellatus guttatus. He cited: Type 1 (8378), Kamerun (aussi type de H. hildebrandti) [ Type 1 (8378), Kamerun (also type of H. hildebrandti) ]. Laurent thus designated this specimen as lectotype according to Article 74.5 of the Code. In addition he cited in the same article the sample from Boutry as paratype (Laurent, 1961: 72). Perret (1966: 401) and Schiøtz (1967: 243) followed the conclusion of Laurent and attributed this specimen to Hyperolius ocellatus guttatus. 15

6 THIERRY FRÉTEY et al. Perret (1975: 190) summarized the history of the two specimens of Peters, before adding to the one from Cameroon: Aujourd hui, ayant réexaminé le type, je l identifie sans hésitation à Hyperolius steindachneri Bocage. Dans mes collections, j ai bien des femelles subadultes qui présentent le même pattern et la morphologie céphalique concorde parfaitement alors qu elle diffère de celle d ocellatus [ Today, having examined the type, I identify it without hesitation as Hyperolius steindachneri Bocage. In my collections, I definitely have subadult females that present the same pattern and the head morphology agrees perfectly though it differs from that of ocellatus ]. He added: Le nom de guttatus, relégué dans la synonymie de steindachneri, disparaît donc [ The name guttatus, relegated to the synonymy of steindachneri thus disappears ]. He finally stated: Etat de conservation du type: Complètement décoloré [ Condition of preservation of the type: completely faded ], which suggests that, when he referred to the pattern, he considered the figure presented by Peters and not the specimen that he had observed. The following year, Perret (1976: 25-26), pushed his conclusions further. Under Hyperolius steindachneri, he explained: Une sous-espèce de steindachneri a été décrite au Cameroun par Laurent (1947): Hyperolius steindachneri pardalis. J ai abondamment récolté cette dernière (Perret, 1966) et je me rends compte maintenant que pardalis est certainement spécifiquement différent de steindachneri [ A subspecies of steindachneri was described from Cameroon by Laurent (1947): Hyperolius steindachneri pardalis. I have collected a large sample of it (Perret, 1966) and I now realize that pardalis is certainly a species distinct from steindachneri ]. He added: La comparaison du type de steindachneri avec mon riche matériel de pardalis révèle des différences de coloration importantes comme d autres morphologiques plus subtiles. Les taches ventrales sont jaunes d or allongées en bandes, sur les côtés du ventre et sous la gorge, chez steindachneri, tandis qu elles sont blanches, plus petites et arrondies chez pardalis. La coloration dorsale qui peut être uniforme, il est vrai, chez les deux espèces, est décrite: d un beau vert violacé par Bocage (cf. 1895, plate 19, figure 3), alors qu elle est jaune ou brunâtre chez pardalis. Enfin, il faut relever que steindachneri est une forme de plateau élevé angolais largement séparée de pardalis, planitiaire de brousse et de forêt secondaire, camerouno-gabonaise. Cette précision écologique qui semble discriminante dans le cas (on n a jamais trouvé pardalis hors de la forêt) corrobore à mon sens la reconnaissance des deux espèces qui pour moi ne fait aucun doute. [ The comparison of the type of steindachneri with my rich material of pardalis reveals important differences of coloration as well as other more subtle morphological differences. The ventral markings are golden yellow elongated in bands on the sides of the belly and under the throat, in steindachneri, but they are white, smaller and rounded in pardalis. The dorsal coloration that can be uniform in both species is described: of a beautiful violet green by Bocage (cf. 1895, plate 19, figure 3), but it is yellow or brownish in pardalis. Finally, it should be mentioned that steindachneri is a form from the high Angolan plateau largely separated from pardalis, living in Cameroon-Gabonese savannah and secondary forest. This ecological precision that seems discriminating in the case (pardalis has never been found outside forest) corroborates in my opinion the recognition of two species that for me is beyond doubt. ]. For Perret, but without explicitly writing it, the sample of Hyperolius guttatus from Cameroon belonged in fact in Hyperolius pardalis. Giving that the specimen ZMB 8378 has been designated as lectotype of Hyperolius guttatus Peters, 1875 by Laurent (1961: 73) and that following Perret (1975, 1976) this specimen is similar to Hyperolius pardalis Laurent, 1948, Hyperolius guttatus Peters, 1875 should be a senior synonym of Hyperolius pardalis Laurent, 1948 and have priority over it. The situation seems clear, but before going further in our analysis, it is necessary to consider a recent work where Hyperolius guttatus is involved. Lötters et al. (2001: 29) examined the lectotype of Hyperolius guttatus (which they stated is a largely faded female of 21.3 mm) and reached the conclusion that this specimen is conspecific with the material studied from Gabon (2 males) and from Cameroon (without indicating the number and sexes of individuals), but not with H. steindachneri. Not wanting to discuss the validity of a sub-species purpurescens, they attribute the lectotype of Hyperolius guttatus to Hyperolius ocellatus. Lötters et al. (2001) thus do not follow the opinion of Perret. But, although they correctly cite the 1975 work of Perret in their article, they do not at all mention his 1976 work that provides the final explanation! Thus, actually, their material from Gabon and from Cameroon cannot be conspecific with Hyperolius steindachneri as this species is present neither in Gabon nor in Cameroon. Additionally, it should be noted that, in the genus Hyperolius, males and females might present extremely different forms and particularly patterns and that it is therefore not recommended to compare individuals of different sexes. It is even more difficult to compare a completely faded female, like the lectotype (ZMB 8378), with males. Having reviewed these various opinions, there is a need to examine the origin of the specimens of Hyperolius guttatus in more detail. 16

7 ALYTES Geographical origin of the specimen from Cameroon The geographical origin of this specimen is cited as Cameruns in the description of Peters (1875: 207) and thus does not seem very precise. This specimen has always been considered as coming from Cameroon, without a precise indication of locality. Nevertheless this is incorrect, as will be demonstrated below! To have a better understanding, we need to examine a short historical perspective. In 1472, the sailors of the Portuguese Fernando Póo entered in what is now the estuary of the Wouri, that they named Río dos Camarões (River of shrimps) because of the abundance of a certain kind of shrimp (Froelich & Gautron, 1996). In 1868, the German Woerman founded a commercial settling in the modern-day Douala (Roche, 2011). In 1884, the Germans established a protectorate in this sector with Kamerunstadt (now Douala) as capital. In 1888, Yaoundé was founded. Only in 1907 Kamerunstadt was changed to Douala (Room, 1994) and Kamerun became the name of the whole country (Foondé, 2011). The collections of Buchholz and the description of Peters thus predate the establishment of the German protectorate and only few places close to the coast had been visited by the Europeans by the time. In 1880, Heinersdorff depicted the route of Reinhold Wilhelm Buchholz, collector of the specimen described by Peters. Heinersdorff (1880: ) tells us that between 1872 and 1874 Buchholz visited the surroundings of Camaroons, the locality indicated as Cameruns on the map that he presents at the end of his work (fig. 3). Cameruns is thus a place that grouped together a series of villages on the left shore of the estuary of the Camerun Fluss (Wouri river) and corresponds to the modern-day Douala (04 03 N, E). The onymotope of Hyperolius guttatus Peters, 1875 is therefore Douala. It is nevertheless possible that the frogs came from other places that are mentioned as collection sites of Amphibians in the book of Heinersdorff (1880), such as Victoria or Abo. Buchholz never stayed for a long period at Camaroons and he used his stays there to prepare the specimens and packages for shipping to Germany. He spent more time at Victoria ( , ), where he had collected several new remarkable species of Amphibians [ mehrere neue merkwürdige Arten ]. Heinersdorff also mentioned the collection of Amphibians at Abo ( , ). His stay at Balong, the furthest inland site did not yield any results, neither did a stay in Jenssoki (south of Douala). The locality Camaroons (or Cameruns ) might refer to the place of shipping of the specimens. Nevertheless the explorations of Buchholz did not reach beyond a 40 km radius around modern-day Douala. He generally used boats on the rivers to access to the interior of the country, but he often must have walked in very difficult conditions because marshes were abundant in that region. His explorations were also restricted by his health because from November 1872 onwards he suffered from fever and bad headaches. The onymotope of Hyperolius guttatus Peters, 1875 is therefore Douala (04 03 N, E). We indicate in Appendix 1 the other species of reptiles and amphibians that have been described from this onymotope. Figure 3. Part of the map published in Heinersdorff (1880), the arrow pointing to Cameruns, modernday Douala, Cameroon. 17

8 THIERRY FRÉTEY et al. Comparison with other Hyperolius Starting from the origin (now rather precise) of the specimen from Cameroon, we have retained all species of the genus Hyperolius from Cameroon having a known distribution within 40 km around Douala (which corresponds to at least two days of walk at the time of Buchholz). We included Hyperolius pardalis, even though it has only been found more in the interior of the country, because Perret thought that this species was identical to Hyperolius guttatus. The following species are present in the region of Douala (Amiet, pers. comm.): Hyperolius acutirostris Buchholz & Peters, 1875, H. bolifambae Mertens, 1938, H. bopeleti Amiet, 1980, H. concolor (Hallowell, 1844), H. fusciventris Peters, 1876, H. guttulatus Günther, 1858, H. kuligae Mertens, 1940, H. ocellatus Günther, 1858 and H. sylvaticus Schiøtz, Comparison of patterns. Based on the description of Hyperolius guttatus by Peters (1875: 207) and on the illustration of that species (plate 2, figure 3), two distinctive characters might be retained: the presence of a black canthal line and a dorsal pattern showing dark spots. The specimen, a young female, measured 21 mm. Among the species retained, Hyperolius acutirostris, H. bopeleti, H. kuligae and H. sylvaticus do not show a coloration morph (so-called phase ) with dark spots on the back and dark canthal lines. The young females of Hyperolius pardalis have neither canthal line nor spots on the back. Only the adult females may present spots on the back that then are distinct, but they never show a dark canthal line. This species can thus also be removed from the list. Additionally, it is not present in the region of Douala as the limit of its distribution range is more than 100 km to the East (Amiet, pers. comm.). In the other species, females or juveniles may have patterns close to that observed in the lectotype of Hyperolius guttatus. They thus must be considered candidates for the identity of the species of Peters. In the species Hyperolius bolifambae, H. concolor, H. fusciventris, H. guttulatus and H. ocellatus, in young females and sometimes in the males, the presence of a canthal line and dark spots as seen in the type specimens of Hyperolius guttatus from Douala can be observed. Nevertheless, the presence of dark spots on the back is not known in H. bolifambae phase J, which excludes this species. In H. ocellatus, the spots present on the back in phase J-2 are very regularly round and distributed on the back, and the light laterodorsal lines join to from a clear triangle on the snout, which is very different from the pattern observed on the drawing of Peters (1875). On the contrary, in the three species left there are coloration patterns that resemble the type specimen illustrated and other data are needed to identify the specimen. Figure 4. Several dorsal patterns in the male specimens of Hyperolius concolor of Douala, Cameroon: A. ZFMK 19819; B. MHNG ; C. MHNG ; D. MNHG

9 ALYTES In Hyperolius concolor, three different coloration phases have been described (fig. 4 ; see Schiøtz, 1963: plate I, figure 1-3), none of which correspond to the coloration pattern of the type specimen of Hyperolius guttatus. But Joger (1982) published under the nomen Hyperolius balfouri a description of a series of frogs from Douala with a dark canthal line, dark laterodorsal bands and spots on the back. After studying these specimens (ZMFK ) we confirm that they belong to Hyperolius concolor, thus adding another coloration phase to the very variable phase J of H. concolor belonging to the still undescribed subspecies (Schiøtz 1967: 185; 1999: 104). The holotype of Hyperolus pulcher Ahl, 1931 (see Appendix 3) is an example of this variation as it presents an extremely clear pattern of the phase J-1. Morphometrical analysis. Principal component analysis of eight measurements taken on 42 specimens allowed a discrimination on axis 1 of a group including the various specimens of Hyperolius concolor from Cameroon and of H. pardalis from a second group that includes the specimens of Hyperolius concolor concolor (from Mount Nimba) and H. ocellatus purpurescens (tab. 2; fig. 5). This factor mainly includes measurements of the head (length of snout, distance of eyes) and of the hand (web of hand). The type specimen of Hyperolius guttatus is among the specimens of Hyperolius concolor from Cameroon. A paratype of H. concolor ibadanensis is close to this group. The important difference between the two subspecies H. c. concolor and H. c. ssp. indet. should be stressed. Figure 5. Principal component analysis including 8 measurements transformed in ratios of SVL of 42 specimens of Hyperolius. In conclusion, the type specimen of Hyperolius guttatus should be attributed to H. concolor, more precisely to the unnamed subspecies of Schiøtz (1967). The identity of H. c. ibadanensis must be studied including more specimens and using genetic methods. The big differentiation between H. concolor concolor and the other subspecies might indicate that these two groups are more probably distinct species than subspecies, but further bioacoustic and genetic studies should confirm this hypothesis. 19

10 THIERRY FRÉTEY et al. Status of the specimen from Cameroon In conclusion, we identify the specimen ZMB 8378, lectotype of Hyperolius guttatus, as belonging to the unnamed subspecies of Hyperolius concolor of Schiøtz (1967: 185), which should therefore bear the name Hyperolius concolor guttatus. Hyperolius concolor guttatus Peter, 1875 comb. nov. Hyperolius guttatus Peters, 1875: 207. Onomatophore: Lectophoront by subsequent designation of Laurent (1961: 73), ZMB 8378, young female. Onymotope: Douala ( Cameruns ), Cameroon. Rappia sordida Fischer, 1889: 10. Onomatophores: Symphoronts, including BMNH (Parker, 1936). Onymotope: Kamerun, Cameroon. Synonymy with Hyperolius concolor by Loveridge, 1938: 50, 67; synonymy with Hyperolius concolor ssp. indet. by Schiøtz, 1967: 185. New synonymy with Hyperoliius guttatus. Remark. The other syntypes studied by Fischer (1889) could not be found, neither in the Zoologisches Museum Hamburg (Germany), the institution where Fischer worked (Hallermann, 2007; Hallermann, pers. comm.), nor in the Museum für Natur und Umwelt of Lübeck (Germany) where the specimens came from according to Fischer (1889) (Füting, pers. comm.). The collections of the two museums have been largely destroyed by bombings during Second World War. Thus only the two syntypes of the Natural History Museum of London still exist. Hyperolius hildebrandti Ahl, 1931: 207. Onomatophore: Holophoront by monotypy, ZMB 8378, young female. Onymotope: Douala ( Cameruns ), Cameroon. Junior objective synonym of Hyperolius guttatus Peters, Hyperolius maximus Ahl, 1931: 366. Onomatophore: Holophoront by monotypy, ZMB 36113, adult female. Onymotope: Osidenge, Mamfe (05 45 N, E), Cameroon. Synonymy with Hyperolius concolor by Loveridge, 1938: 50, 67; synonymy with Hyperolius concolor ssp. indet. by Schiøtz 1967: 185. New synonymy with Hyperoliius guttatus. Hyperolius pulcher Ahl, 1931: 308. Onomatophore: Holophoront by monotypy, ZMB 36088, adult female. Onymotope: Japoma ( N, E), suburbs of Douala, Cameroon. Synonymy with Hyperolius concolor concolor by Laurent, 1961: 77; synonymy with Hyperolius concolor ssp. indet. by Schiøtz 1967: 185. New synonymy with Hyperoliius guttatus. Table 2. Results of principal component analysis including 8 measurements of 42 specimens of Hyperolius: A. indicating the variance explained; B. the factor matrix. A Factors rsl riue ruew rtfl rtl raw rfl rmttf B Factor Eigenvalue % of variance Cumulated variance 1 5,788 72,356 72, ,279 15,987 88, ,360 4,494 92,836 Redescription of the lectophoront (lectotype) of Hyperolius guttatus Peters, 1975 ZMB 8378, young female (fig. 2): Small size (23.0 mm), body rather elongate. Head moderately large, wider (8.4 mm) than long (8.1 mm), flat. Snout truncated, projecting slightly over mouth, its length (3.42 mm) slightly shorter than eye diameter (3.68 mm). Canthus rostralis rounded; loreal region concave, angle with surface of head slightly obtuse. Interorbital region flat, twice as large (2.98 mm) as width of upper eyelid (1.49 mm) and slightly larger than distance between nostrils (2.59 mm). Nostril of oval outline without dermal tubercle, closer 20

11 ALYTES to tip of snout (1.23 mm) than to eye (2.33 mm). Pupil rounded. Tympanum poorly distinct, rounded, its diameter (0.84 mm) corresponding to a fourth of diameter of eye, distance from tympanum to eye (0.45 mm) corresponding to half its diameter. Pineal organ absent. Crests and vomerine teeth absent. Tongue of moderate size, cordate, emarginated. Arm short, thin; forearm (4.7 mm) shorter than hand (6.42 mm), not enlarged. Fingers thin, short (length of finger III 3.24 mm). Relative length of fingers from shortest to longest: I < II < IV < III. Tips of fingers rounded, bearing discs with terminal grooves, rather enlarged as compared to finger width (PA mm, WA mm; PA mm, WA mm; PA mm, WA mm; PA mm, WA mm). Fingers linked by a narrow web followed by dermal fringes to pads: I 2 2 II 1 3 III 2 2 IV. Subarticular tubercles scarcely prominent, rounded, all present, the proximal ones of fingers III and IV smaller. Prepollex oval, scarcely distinct; palmar tubercles absent; supernumerary tubercles absent. Shank almost four times longer (11.8 mm) than wide (2.8 mm), longer than thigh (9.8 mm) and foot (9.8 mm). Toes short and thin, the fourth (5.19 mm) about one third of distance from tarsus to tip of toe IV (15.7 mm). Relative length of toes from shortest to longest: I < II < V < III < IV. Tips of toes rounded, bearing pads with terminal grooves, rather enlarged compared to toe width (PP mm, WP mm; PP mm, WP mm; PP mm, WP mm; PP mm, WP mm; PP mm, WP mm). Palmar formula: I 1 ½ 2 II 1 2 III 1 2 IV 2 1 V (MTTF 5.38 mm, MTFF 5.83 mm, FTTF 3.63 mm, FFTF 3.31 mm). Dermal ridge along toe V present, until level of proximal subarticular tubercle. Subarticular tubercles slightly prominent, rounded, all present, the proximal ones smaller than the terminal. Internal metatarsal tubercle slightly elevated, rather long, its length (0.96 mm) 2.3 times the length of toe I (2.24 mm). Tarsal fold absent. External metatarsal tubercle, supernumerary tubercles and tarsal tubercle absents. Dorsal and lateral parts of specimen smooth. Supratympanic fold scarcely distinct, from eye to above arm. Forearm, thigh, shank and tarsus smooth. Ventral parts smooth. Macroglands absent. Colouration in alcohol. Specimen entirely clear beige, only eyeballs still showing pigments. Female sexual characters. Observation of an oviduct through an aperture in the posterior belly. Oviduct feebly enlarged, folded, not in reproductive condition. Remarks. Peters (1875) indicated 21 mm as size for the specimen. Lötters et al. (2001) gave 21.3 mm for this measurement. The specimen is in very poor state, not only faded, but also very soft. We measured snout-vent length several times and obtained values around 23.0 mm every time. This growth is surely due to the relaxation of the tissues. The absence of dermal structures such as subarticular or supernumerary tubercles on the hands and feet might also be the consequence of the very poor preservation condition of the specimen. Such a phenomenon has been observed on the symphoronts of Rana bibroni Hallowell, 1845 which, even though they show the characteristic nuptial pads of the species on fingers and on the base of the hand, do not show the supernumerary tubercles at the base of toes present in the well preserved specimens of this species (Lamotte & Ohler, 1997). The presence or absence of tubercles on the skin of the type specimen of Hyperolius guttatus therefore cannot be confirmed. Sex and ontogenetic stage of the specimen were confirmed. The development of the oviduct confirms that it is a young female because the oviduct is not completely developed. It shows folds that are distinctly separated. In adult anurans the development of the tissue of the oviduct leads to foldings that are not individualised but stuck together. We could not observe the ovaries because this would have required another incision on the flank in this specimen that is already in a very poor state. RÉSUMÉ Le statut nomenclatural du nomen Hyperolius guttatus Peters, 1875 a été réétudié. Les spécimens qui ont servi à sa description proviennent de Bootry au Ghana et du Cameroun, a priori sans localité précise. Nous avons établi la chronologie des différentes attributions taxonomiques des spécimens du Ghana et du Cameroun. L étude de documents anciens nous a permis de préciser, sans ambiguïté, la localité type (ou onymotope) d Hyperolius guttatus au Cameroun dans la région de Douala, ainsi que celle de plusieurs autres espèces d Anoures et de Serpents décrits dans la même publication et donnés en Annexe. Une analyse de la livrée et de la morphométrie nous a permis de comparer le spécimen d Hyperolius guttatus du Cameroun avec les autres espèces d Hyperolius présentes dans la même région, nous permettant ainsi de statuer sur ce nomen. Hyperolius guttatus est proposé comme nomen pour une sous-espèce de Hyperolius concolor du Cameroun encore non nommée. L identité des syntypes de Boutry (Ghana) est confirmée être Hyperolius fusciventris burtonii. Les spécimens-types d Hyperolius guttatus Peters, 1875 et d Hyperolius pulcher Ahl, 1931 ont été redécrits pour documenter leur identité. 21

12 THIERRY FRÉTEY et al. ACKNOWLEDGEMENTS We wish to cordially acknowledge Jean-Louis Amiet, for several reasons. First because he is the initiator of this publication, and because he made us aware that at the time of the collections of Buchholz, Hyperolius pardalis was inaccessible due to its distribution. This was the starting point for studying the history of the early batrachological exploration of Cameroon. We also thank him for his inestimable help regarding the coastal distribution of the species and the access to his collections. We finally are grateful to him for his comments on a first version of our manuscript. We also acknowledge Mark-Oliver Rödel and Frank Tillack, for sending specimens from Berlin and Ronald de Ruiter for sending the specimens from Leiden. Jakob Hallermann (Hamburg) and Susanne Füting (Lübeck) are thanked for searching lost specimens and giving information of their collections. Finally thanks to Victoire Koyamba (Paris) for the copy of the map from Heinersdorff (1880). Uwe Jung from Goethe-Institut (Yaounde, Cameroon) gave valuable access to documents on the history of Cameroon. Arne Schiøtz and Mark-Oliver Rödel are acknowledged for their useful comments on the manuscript. 22 LITERATURE CITED Ahl, E. (1931). Amphibia: Anura: III Polypedatidae. Das Tierreich, 55: i-xvi Bauer, A.M., Günther, R., Klipfel, M. (1995). The herpetological contribution of Wilhelm C. H. Peters ( ), with an introduction, annotated bibliography, and synopsis of taxa. Society for the Study of Amphibians and Reptiles, St. Louis. Boulenger, G.A. 1883). Description of new species of reptiles and batrachians in the British Museum. Annals and Magazine of natural History, (5), 12: Broadley, D.G., Wallach, V. (2007). A review of East and Central African species of Letheobia Cope, revived from the synonymy of Rhinotyphlops Fitzinger, with descriptions of five new species (Serpentes: Typhlopidae). Zootaxa, 1515: Dubois, A. (2005). Proposed rules for the incorporation of nomina of higher-ranked zoological taxa in the International Code of Zoological Nomenclature. 1. Some general questions, concepts and terms of biological nomenclature. Zoosystema, 27(2): Dubois, A., Ohler, A. (1997). Early scientific names of Amphibia Anura. I. Introduction. Bulletin du Muséum National d Histoire Naturelle, Paris, (4), 18: Fischer, J.G. (1889). Herpetologische Mitteilungen. I. Über zwei neue Schlangen und einen neuen Laubfrosch von Kamerun. Jahrbuch der Hamburgischen Wissenschaftlichen Anstalten, 1888, 5: Foondé, J.P.M. (2011). Douala: Toponymie, histoire et cultures. Ifrikiya, Yaoundé. Froelich, J.C., Gautron, J.C., (1996). Cameroun. Encyclopædia Universalis, France, 4: Hallermann, J. (2007). Zur Geschichte der herpetologischen Sammlung des Zoologischen Museums Hamburgs, mit besonderer Berücksichtigung von Dr. Johann Gustav Fischer ( ). Sekretär, 7: Heinersdorff, C. (1880). Reinhold Buchholz Reisen in West-Afrika nach seinen hinterlassenen Tagebu chern und Briefen, nebst einem Lebensabriss des Verstorbenen. F. A. Brockhaus, Leipzig. Holthuis, L.B. (1968). Biografische notities betreffende verzamelaars voor het Rijksmuseum van Natuurlijke Historie te Leiden. I. Hendrik Severinus Pel ( ). Zoologische Bijdragen, 10: Hughes, B. (1977). Chlorophis carinatus Andersson, 1901, proposed nomenclatural precedence over Philothamnus nigrofasciatus Buchholz and Peters, 1875, its senior subjective synonym (Reptilia; Colubridae). Z.N.(S) Bulletin of Zoological Nomenclature, 33: International Commission on Zoological Nomenclature (ICZN) (1999). International Code on Zoological Nomenclature, 4 th edition. International Trust for Zoological Nomenclature, London. Joger, U. (1982). Zur Herpetofaunistik Kameruns (II). Bonner Zoologische Beiträge, 33(2-4): Lamotte, M., Ohler, A. (1997). Redécouverte de syntypes de Rana bibroni Hallowell, 1845, désignation d un lectotype et description d une espèce nouvelle de Ptychadena. Zoosystema, 19: Laurent, R.F. (1948). Two new forms of the genus Hyperolius. Annals and Magazine of Natural History, 1947, (11), 14: Laurent, R.F. (1950). Genres Afrixalus et Hyperolius. Exploration du Parc national Albert. Mission G.F. de Witte ( ), 64. Institut des Parcs nationaux du Congo Belge: Laurent, R.F. (1958). La réserve naturelle intégrale du mont Nimba, XIII. Les rainettes du genre Hyperolius. Mé-

13 ALYTES moires de l Institut Français d Afrique Noire, 53: Laurent, R.F. (1961). Note sur les Hyperolius et quelques Afrixalus (Salientia) du Musée de Berlin. Revue de Zoologie et de Botanique Africaines, 64(1-2): Lötters, S., Gossmann, V., Obame, F., Böhme, W. (2001). Zur Herpetofauna Gabuns. Teil I: Eineitung, Untersuchungsgebiet und Methodik, kommentierte Artenliste der gefundenen Froschlurche. Herpetofauna, 23(133): Loveridge, A. (1938). On a collection of reptiles and amphibians from Liberia. Proceedings of the New England Zoölogical Club, 17: Mertens, R. (1938). Herpetologische Ergebnisse einer Reise nach Kamerun. Abhandlungen der Senckenbergischen Naturforschenden Gesellschaft, 442: Pauwels, O.S.G., Branch, W.R., Burger, M. (2004). Reptiles of Loango National Park, Ogooué-Maritime Province, south-western Gabon. Hamadryad, 29: Perret, J.-L. (1958). Observations sur des rainettes africaines du genre Leptopelis Günther. Revue Suisse de Zoologie 65: Perret, J.-L. (1962). Révision des types de Leptopelis et note sur quelques Hyperolius (Amphibia Salientia) de la région camerounaise, conserves au Muséum de Berlin. Revue de Zoologie et de Botanique Africaines, 65: Perret, J.-L. (1966). Les Amphibiens du Cameroun. Zoologische Jahrbu cher, Abteilung fu r Systematik Ökologie und Geographie der Tiere, 8: Perret, J.-L. (1975). Révision critique de quelques types de Reptiles et Batraciens africains. Revue Suisse de Zoologie, 82(1): Perret, J.-L. (1976). Révision des Amphibiens africains et principalement des types, conservés au Musée Bocage de Lisbonne. Arquivos do Museu Bocage, (2), 6: Peters, W. (1875). Über die von Herrn Professor Dr. R. Buchholz in Westafrika gesammelten Amphibien. Monatsberichte der Preussischen Akademie der Wissenschaften zu Berlin, 1875: Roche, C. (2011). Afrique noire et la France au XIXe siècle. Karthala, Paris. Rödel, M.-O. (1998). A new Hyperolius species from Tai National Park, Ivory Coast (Anura: Hyperoliidae: Hyperoliinae). Revue française d Aquariologie et d Herpétologie, 25: Room, A. (1994). African placenames. McFarlan, Jefferson, NC. Schiøtz, A. (1963). The Amphibians of Nigeria. Videnskabelige Meddelelser fra Dansk naturhistorik Forening i Københaun, 125: Schiøtz, A. (1967). The treefrogs (Rhacophoridae) of West Africa. Spolia Zoologica Musei Hauniensis, 25: Specimens studied APPENDIX 1 The collection numbers of specimens included in the morphometrical analysis are in bold: Hyperolius acutirostris, JLA , , adult females, Kala, Cameroon; JLA , , adult females, Saxenhof, Cameroon. Hyperolius balfouri, MNHN , adult male, Managouba, Cameroon. Hyperolius bolifambae, JLA , adult female, Bdjouka, Cameroon; JLA , adult female, Bomnyebel, Cameroon; JLA , adult female, Ehom, Cameroon; JLA , adult female, Fineschang, Cameroon; JLA , adult female, Bolifamba, Cameroon. Hyperolius concolor concolor, MNHN , , , , , , , 5 males and 5 adult females, Mont Nimba, Guinea. Hyperolius concolor ibadanensis, MNHN (ex. RO 07727), paratype, adult male, 4 km north of University of Ibadan, Nigeria. 23

14 THIERRY FRÉTEY et al. Hyperolius concolor ssp. indet., JLA , , , adult females, Edéa, Cameroon; JLA , , adult females, Debundscha, Cameroon; MHNG , adult males, Douala, Cameroon; ZMFK , adult males, Douala, Cameroon. Hyperolius fusciventris, JLA , adult female, Nyomwe, Cameroon; JLA , adult females, Debundscha, Cameroon; JLA , , adult females, Nlohé, Cameroon; MNHN , , juvenils, Assakra, Ivory Coast; MNHN , , male and adult females, Lamto, Ivory Coast; MNHN , , , , , 1 juvenile and 4 adult males, Orombo Boka, Ivory Coast. Hyperolius fusciventris burtonii, ZMB 4489, paralectotype of Hyperolius guttatus Peters, 1875 (Boutry [Bootry (4 49 N, 1 55 O)], Ghana; RMNH.RENA.1788, 4 adult males, paralectotypes of Hyperolius guttatus Peters, 1875, same locality. Hyperolius guttatus, ZMB 8378, young female, lectotype of Hyperolius guttatus Peters, 1875 and holotype of Hyperolius hildebrandti Ahl, 1931, Cameruns [Douala], Cameroon. Hyperolius guttulatus, JLA , adult females, Bonendale, Cameroon; JLA , adult female, Nyonwe, Cameroon; JLA , adult females, Apouh, Cameroon; MNHN , , adult males, Mount Nimba, Guinea. Hyperolius kuligae, JLA , adult female, Santchou, Cameroon; JLA , , adult females, Elomzok, Cameroon; JLA , adult female, Mintom, Cameroon; JLA , adult female, Mbet, Cameroon. Hyperolius ocellatus purpurescens, MNHN , adult female, San Benito, river, Gabon; MNHN , adult female, Gabon; MNHN , Makak, Cameroon. Hyperolius pardalis, JLA , adult female, Binguela, Cameroon; JLA , adult female, Ototomo, Cameroon; JLA , , adult females, Matomb, Cameroon; JLA , adult female, Essazok, Cameroon. Hyperolius pulcher, ZMB 36088, holotype, young female, Japoma, Cameroon. Hyperolius sylvaticus ivorensis, MNHN (ex RO ), adult males, 35 km south of Abidjan, Ivory Coast; MNHN (ex RO 76251, 76341, ), adult males, just south of Ndenou, Ivory Coast; MNHN , adult male, Ivory Coast. Hyperolius sylvaticus nigeriensis, JLA , adult female, JLA , , adult males, Mbio, Cameroon; JLA , , adult males, Bowonda, Cameroon; LA , adult male, Yoké, Cameroon; MNHN (ex RO 71673), paratype, adult male, Iperin, Nigeria; MNHN (ex RO 71804), paratype, adult male, 4 km orth of University of Ibadan, Nigeria. Hyperolius sylvaticus sylvaticus, MNHN (ex RO 74796), paratype, adult male, Bobiri, Ghana. 24 APPENDIX 2 Species described in Peters (1875) from the type locality Cameruns (Douala) Squamata Philothamnus nigrofasciatus Buchholz & Peters in Peters, 1875: The classification of this taxon is not clear: according to the authors it is either a synonym of P. carinatus (Hughes, 1977; Pauwels et al., 2004) or of P. nitidus (Chirio & LeBreton, 2007). Synonym of Philothamnus carinatus (Andersson, 1901) [in this case nigrofasciatus should have priority, but Opinion 1214 (Melville, 1982) reversed the precedence] or of Philothamnus nitidus (Günther, 1863). Thrasops pustulatus Buchholz & Peters in Peters, 1875: 199. Synonym of Thrasops flavigularis Hallowell, 1852: Pauwels et al., Typhlops (Ophthalmidion) decorosus Buchholz & Peters in Peters, 1875: Current valid nomen Letheobia decorosa): Broadley & Wallach, Amphibia Hylambates notatus Buchholz & Peters in Peters, 1875: [holophoront lost according to Perret, 1962]. Current valid nomen Leptopelis notatus: Perret, 1958: 259. Hyperolius acutirostris Buchholz & Peters in Peters, 1875: Hyperolius guttatus Peters, 1875: 207.

15 ALYTES Hyperolius spinosus Buchholz & Peters in Peters, 1875: Current valid nomen Acanthixalus spinosus: Laurent, Nectophryne afra Buchholz & Peters in Peters, 1875: APPENDIX 3 Redescription of the holotype of Hyperolius pulcher Ahl, 1931 ZMB 36088, young female (fig. 6): Size rather small (25.8 mm), body rather elongate. Head moderately wide, longer (10.2 mm) than wide (9.2 mm), flat. Snout pointed, rounded, slightly projecting beyond mouth, its length (4.5 mm) longer than diameter of eye (3.6 mm). Canthus rostralis rounded, loreal region flat, at right angle to upper head. Interorbital space flat, less than twice larger (3.5 mm) than width of upper eyelid (2.1 mm) and distinctly larger than distance between nostrils (2.6 mm). Nostril of oval outline without dermal tubercles, closer to tip of snout (1.69 mm) than to eye (2.66 mm). Pupilla oval, horizontal. Tympanum scarcely distinct, rounded, its diameter (1.23 mm) one third of eye diameter, distance from tympanum to eye (0.32 mm) corresponding to one fourth its diameter. Pineal ocellus absent. Vomerine crests and teeth absent. Tongue of moderate size, cordate, emarginated. Arm short, thin; forearm (5.8 mm) shorter than hand (7.2 mm), not enlarged. Fingers thin, short (length of finger III 4.02 mm). Relative length of fingers from shortest to longest: I < II < IV < III. Tips of fingers rounded, bearing pads with terminal grooves, rather enlarged compared to fingers (PA mm, WA mm; PA mm, WA mm; PA mm, WA mm; PA mm, WA mm). Fingers linked by a narrow web prolonged by dermal fringes up to pads: I 2 2 1/3 II 2 3 III 2 1/3 2 IV. Subarticular tubercles scarcely prominent, rounded, all present, the proximal subarticular tubercles of fingers III and IV smaller. Prepollex oval, scarcely distinct; numerous tubercles flat tubercles on palm. Shank five times longer (12.8 mm) than wide (2.5 mm), longer than thigh (12.4 mm) and foot (11.0 mm). Toes short and thin, the fourth (5.8 mm) about one third the distance from tarsus to tip of toe IV (17.3 mm). Relative length of toes from shortest to longest: I < II < V < III < IV. Tip of toes rounded, bearing pads with terminal grooves, rather enlarged compared to toes (PP mm, WP mm; PP mm, WP mm; PP mm, WP mm; PP mm, WP mm; PP mm, WP mm). Palmar formula: I 1 2 II 1 2 III 1 2 ¼ IV 2 1 V (MTTF 6.05 mm, MTFF 7.24 mm, FTTF 3.95, FFTF 3.68 mm). Dermal ridge along toe V present, until level of proximal subarticular tubercle. Subarticular tubercles scarcely prominent, rounded, all present, the proximal ones smaller than the terminal ones. Inner metatarsal tubercle poorly prominent, rather long, its length (1.10 mm) included 2.4 times in the length of toe I (2.66 mm). Tarsal fold absent. External metatarsal tubercle, supernumerary tubercles and tarsal tubercle absents. Dorsal and lateral parts of specimen smooth. Distinct tubercles posterior to mouth slit. Supratympanic fold feeble, from eye to above arm. Fore-arm, thigh, shank and tarsus smooth. Throat and chest smooth, belly and ventral part of thigh with rounded densely set tubercles ( treefrog belly-skin ). Macroglands absent. Coloration in alcohol. Dorsal parts of head and back brown very clear with poorly distinct darker spots, Figure 6. Drawing of Ahl (1931) (A) and photograph of the holophoront of Hyperolius pulcher Ahl, 1931 (B), ZMB 36088, young female, SVL 25.8 mm, Japoma, suburbs of Douala, Cameroon. 25