Subsequent literature

According to Mielke (2005), only seven publications refer to busiris following its description: Donovan (1800–1804), Latreille ([1824]), Westwood (1852), Butler ([1870]), Kirby (1871), Mabille & Boullet (1919), and Zimsen (1964). To these, citations by Westwood (1842) and Scudder (1874–1875), and that of Mielke (2004, 2005) himself, should be added. None of these sheds much additional light on the identity of the nominal taxon. Donovan (1800–1804) illustrated the dorsal aspect of the species (Fig. 1F), but the figure appears less detailed than Jones’s originals and looks like an amalgamation of Jones’s dorsal and ventral views, i.e. it shows pale-lined veins in the distal band of the hindwing. The greyish tinge along the forewing apex is also broader and the apex of the antennae is less reflexed. It is unclear whether this illustration was made from a specimen or is a copy of Jones’s illustration, but it could have been based on either, for Donovan (1800–1804) acknowledged Drury in the introduction for allowing inspection of his collection, whereas he referred to Jones as a ‘worthy friend’ (Donovan, 1805) and is known to have described species directly from his paintings (Vane-Wright, 2010). It should be noted that the identical position of the wings in Donovan’s illustration and Jones’s painting strongly supports the interpretation that the former was based on either the original specimen(s) or its ‘iconotype’, and not on any additional subsequent material. Inconsistencies in the antennal apices and patterns between the two illustrations are minor and can be ascribed to the poor reliability by Donovan, who reputedly introduced many errors into his artwork (R.I. Vane-Wright, personal communication, 2018).

Donovan’s illustration was subsequently reissued in the revised edition of his work by Westwood (1842), who added an indication of the size of the species. This was evidently taken from the plate itself, as contrary to other species for which he had supplementary materials, Westwood simply recorded the same basic information previously given by Fabricius and, subsequently, Latreille ([1824]).

Interestingly, all the authors treated the taxon within the Hesperiidae and most of them assigned it to one of the gaudy-patterned genera of Neotropical skippers presently placed in the subfamily Eudaminae (Butler, [1870]; Kirby, 1871; Mielke, 2004, 2005). Westwood(1852) and Butler ([1870]) used Phareas busiris, and Kirby (1871), followed by Mabille & Boullet (1919, misspelled as busirus), used Entheus busiris. Butler ([1870]) likened busiris to the species currently known as Cabirus procas (Cramer), misspelling it as ‘phorcas’. Species of Entheus Hübner (formerly placed in Phareas Westwood) and Cabirus Hübner (formerly placed in Entheus, and before that in Phareas) are characterized by white, yellow and orange spotting on black wings, with hindwings being mostly orange with black margins in several species, matching Jones’s illustrations well (Warren et al., 2018). However, the forewing pattern of Hesperia busiris is significantly different from that of any known Entheus or other Hesperiidae (cf. Warren et al., 2018).

Strangely, Mabille & Boullet (1919) admitted to not having seen any specimens, but stressed that the species was known only on the basis of the male sex, evidently by analogy with other Entheus, in which only the males have yellow bands on forewings. They also mentioned the elongated wings, unusual for the Hesperiidae, and which are, following a suggestion by Latreille ([1824]), more typical of Heliconius Kluk (Nymphalidae: Heliconiinae).

On the early fate of the type of Hesperia busiris

As clearly stated by Jones ([1785–1787]) and Fabricius (1793), the type(s) of Hesperia busiris belonged to Dru Drury (1725–1803, or –1804), a London silversmith and jeweller who established a collection of over 11 000 insect specimens ‘from all the various parts of the World where we have any intercourse as Englishmen’ (Drury, 1778), mostly in short series and in 1788 including exactly 2148 Lepidoptera species (Smith, 1842; Chalmers-Hunt, 1976). The accession date of the busiris specimen(s) into Drury’s collection should not necessarily be assumed to be after 1782 solely on grounds that otherwise he would have described such a gaily patterned species himself (Drury, 1770–1782). In fact, a review of the catalogue of his collection of exotic Lepidoptera, dated 1784 on the frontispiece but updated to 1789 (Drury, 1784[–1789]a,b), reveals that he described only a minority of the species he owned, many of which remained unidentified. On the other hand, Hesperia busiris should have been illustrated by 1787, the date inferred for Fabricius’s visit to Jones, a fond admirer of Drury (Salmon, 2000), to see the plates that had made been from his material (Vane-Wright, 2010). In fact, in that year the Dane visited England once again with all his family, whereas up to 1782 he had not mentioned Jones in his correspondence despite thoroughly reporting on all the people and places he had seen during his numerous stays in London (Fabricius, 1784; Hope, 1845; Armitage, 1958).

There is no indication that the type(s) of Hesperia busiris may have been retained by Fabricius following its description. In fact, Zimsen (1964: 565) specifies for Hesperia busiris: ‘»in Indiis Dom. Drury« – ‘ with no mention of specimen(s) location after the dash. In contrast, for ‘Hesperia Pisistratus’, the species described next by Fabricius (1793), Zimsen lists ‘Dom Drury«, – Kiel 2 specimens.’ The Fabrician collection of Lepidoptera belonging to ZMUK has long been united with that of ZMUC, where indeed there are presently two specimens presumed to be H. pisistratus types, but nothing referable to Hesperia busiris is present in either of the two holdings (cf. ZMUC, 2015). In fact, it is known that rather than borrowing specimens from Drury, Fabricius spent long visits with him to work with his collection, as recalled by Drury himself (‘He is now in London, and very busy in making descriptions from Mr. Banks’ and my Collections, where he will have employment for some months’; Smith, 1842). Fabricius himself admits he simply examined and described Drury’s material (Fabricius, 1784: ‘Die mehrsten derselben habe ich bei den Banks und bei den Drury gesehen, beschrieben’).

Considering the extreme rarity of a species that, despite its outstanding pattern, remained unrecognized and was never subsequently collected following its description (see later for a discussion on its distribution), it is very unlikely that, by the end of 1787, Drury owned more than one specimen, by which time busiris had already been portrayed by Jones. Although Drury’s collection was considered by a contemporary evaluator as ‘la plus parfaite Collection au Monde’ (Anonymous, 1780), it was not that large according to modern standards, if we consider that in 1788 it contained 2462 Lepidoptera specimens belonging to 2148 species (Smith, 1842), thus averaging 1.15 specimens per species. Although the possibility that the original series consisted of multiple syntypes cannot be ruled out with certainty, we will nevertheless, for convenience and brevity, refer subsequently to only a single type specimen of Hesperia busiris. Furthermore, the available information is fully consistent with this specimen being retained by Drury, and eventually auctioned with the sale of Drury’s collection (Anonymous, 1805).

Heraclia contigua is a perfect match for Hesperia busiris

Stripes and spots of metallic-blue scales between yellow bands and spots on forewings are not present in any known Hesperiidae but are characteristic of many Agaristinae (Noctuidae). The conspicuously ringed abdomen also closely corresponds with that seen in many agaristine taxa. Wide diagonal yellow bands and spots on the forewings are typical of African agaristine species, in particular in the genus Heraclia Hübner, which also has clubbed antennae with a reflexed tip (Jordan, 1913–1934; Kiriakoff, 1977a). However, the wing and body patterns of most common and widespread species of Heraclia do not correspond to those of Hesperia busiris. Thus, we searched the Agaristinae holdings at NHMUK for specimens that agree closely with Jones’s illustrations. We found a single specimen, the holotype of Eusemia contigua Walker (now in Heraclia), which is a perfect match (Fig. 1G–J). The agreement in pattern is so close that it is far more parsimonious to conclude that busiris is an agaristine moth than to envisage the existence of a mimetic hesperiid that has yet to be rediscovered since Fabricius’s time. According to provisions of ICZN (1999), the valid name for this species thus becomes Heraclia busiris (Fabricius, 1793) comb.n. (= Eusemia contigua Walker, 1854, syn.n.). In fact, conditions for reversal of precedence between names and considering busiris as an unused senior synonym are not met (ICZN 1999: art. 23.9.1), because Entheus busiris has been used as a valid name after 1899. Furthermore, the junior name has not been used frequently enough to be considered in ‘prevailing usage’ and therefore it does not warrant referral of the case to the International Commission on Zoological Nomenclature for suppression of precedence.

The reasons for the long-standing failure to recognize the close similarity between Hesperia busiris and E. contigua probably reside in various circumstances, but the main one possibly arises because Jones’s work was unpublished. Despite being in the OUMNH since at least 1933 (Vane-Wright, 2010), the work was made freely available only recently (Jones’ Icones Online, 2015). In addition, as noted earlier, the only subsequent illustration of busiris was quite different from the original (Donovan, 1800–1804; Westwood, 1842). Walker’s contigua is itself, in turn, a poorly known taxon. Only the male holotype is so far known (Mabille, 1890; Kirby, 1891; Jordan, 1913–1934; Kiriakoff, 1977a) and was illustrated only once by Butler (1877) (Fig. 1G,H). Although Jordan (1913–1934: 8) claimed to have included a new illustration of contigua on plate 5 of his extensive work on the group, he eventually failed to do so.

Systematic position of Heraclia busiris within Agaristinae

Regarding its systematic position and affinities, Kirby (1891) tentatively placed the species (as contigua) in the endemic African genus Xanthospilopteryx Wallengren, 1858 (currently a junior synonym of Heraclia), stating that it was an unmistakable species probably allied to Polacanthopoda tigrina (Druce, 1882) (cited under Xanthospilopteryx). Later, he assigned it fully to Xanthospilopteryx (Kirby, 1892). Karsch (1895), reporting an observation by Hermann Dewitz, suggested an association with his Mitrophrys fabricata Karsch, 1895, which is still, however, a name relating to Druce’s (1882) tigrina, either as a race or a form (Kiriakoff, 1977a; Poole, 1989). By contrast, Hampson (1901) placed contigua in Rothia Westwood, 1877, another African agaristine genus, his concept of which included species that are currently placed in other genera, such as Heraclia. Strangely enough, Hampson dismissed any similarity between fabricata and tigrina, considering the former as a synonym of contigua and the latter as the type species of his new monotypic genus, Polacanthopoda Hampson, 1901. Strand (1912) echoed Hampson (1901) by placing contigua in Rothia and considering fabricata to be an aberration of contigua.

Jordan (1913–1934) and Kiriakoff (1977a) conclusively placed contigua in Heraclia (= Xanthospilopteryx) and fabricata as infraspecific within P. tigrina. The two species do not show any particularly stronger resemblance than they do with a number of other agaristine taxa (cf. Jordan, 1913–1934; Kiriakoff, 1977a), and exhibit clear differences, including the orientation of inner yellow band(s) of the forewing, the abdominal pattern (Figs 1I,J, 2A–D), the configuration of labial palps, the tibial spining and features of the genitalia (Fig. 3A,B).

Open in a separate window

Fig. 2.

Polacanthopoda tigrina. (A) Original illustration of Mitrophrys fabricata (a junior subjective synonym of Polacanthopoda tigrina) by Karsch (1895: pl. 2, f. 4). (B) Original illustration of P. tigrina by Druce (1882: pl. 60, fig. 4, under Hespagarista). (C, D) Dorsal and ventral aspects, respectively, of a male with wing pattern intermediate between f. fabricata and typical tigrina [Congo, Katanga, Kafakumba, specimen no. BMNH(E) 1377147, in NHMUK]. Images (C) and (D) are copyright of the Trustees of the Natural History Museum, London (used with permission). [Colour figure can be viewed atwileyonlinelibrary.com].

Open in a separate window

Fig. 3.

Male genitalia of Noctuidae Agaristinae. (A) Same as Fig. 1 (I, J), BMNH Noct. slide 8601. (B) Polacanthopoda tigrina, holotype, Nigeria, Akwa Akpa (‘Old Calabar’), BMNH Noct. slide 8562. (C) Heraclia geryon, type species of genus Heraclia (as Phalaena euphemia), Sierra Leone, BMNH Noct. slide 8597. Aedeagi are shown on the right. All specimens are in the NHMUK collection and their photographs are copyright of the Trustees of the Natural History Museum, London (used with permission).

The systematic position of contigua within Heraclia has never seriously been questioned. In this context, we should draw attention both to its placement in another agaristine genus, namely Aegocera Latreille, 1809, and to the synonymy of Polacanthopoda with this genus, as seen in some web sources. Neither of these taxonomic acts has been substantiated in any publications, so that the taxonomic arrangement recorded by Poole (1989) remains in force. Nevertheless, it should be noted that many genera of Agaristinae, including Heraclia, are only imperfectly characterized, essentially on features of the wing venation alone, and other structural characters, such as those of the genitalia, have been stated to be quite variable within species, even in formal diagnoses (cf. Kiriakoff, 1977a: 13). However, even though the holotype of contigua matches the general wing venation of Heraclia, the fine details of the forewing radial system place this species somewhat aside from the main assemblage of the genus. So Jordan (1913–1934: 7–8) associated contigua with a small group of deviant species characterized by vein Rs₁ (‘2. Subcostalast’ in his unusual terminology, cf. Miller, 1970) arising from the areole, whereas the condition of Rs₁ stalked with Rs₂ and arising from beyond the areole is considered to be diagnostic for Heraclia s.s. (Jordan, 1913–1934: 1; Kiriakoff, 1977a: 3, 12). Other features, such as mouthparts, appear in agreement with Heraclia, but the male genitalia (Fig. 3A) again show several discrepancies with respect to most congeners. These, including the type species, Phalaena euphemia Stoll, 1781 [currently a junior synonym of Heraclia geryon (Fabricius, 1781)], generally show a very tall tegumen and broadly axe-shaped valvae (Fig. 3C), whereas, at least in the general shape of the valvae and saccus, the male genitalia of Heraclia busiris (= contigua) resemble those of P. tigrina (Fig. 3B).

The distribution of Heraclia busiris:everywhere and nowhere

Although the identity of Fabricius’ busiris has now been clarified, the actual distribution of the species remains a mystery. The original type locality of Hesperia busiris was stated to be ‘Indiis’ (Indies), a notation that, during Fabricius’s time, was inconsistently used in reference to both the Oriental Indies (East Indies), with particular reference to east side of the Indian subcontinent, and the West Indies.

Based on data provided in ZMUC (2015), a comparison of citations of ‘Indian’ localities for Lepidoptera in Fabricius (1793) yielded the following results: (in) Indiis (81), India (35), India orientali (21), Tranquebariae (7), India occidentalis (1), Bengalia (1), Coromandel (1), Madras (1), Malabaria (1), and nova Hollandia [= W. Australia, cf. Scott, 1914] (1). Thus, being fully aware of the distinctions (or lack thereof) among India, East Indies, West Indies and Australia, one can really only assume a general region stretching from Indochina to Malesia. Of course, Fabricius could only rely on the information associated with the specimens or supplied by the various owners of the material that he was describing. In addition, the actual provenance of specimens described as being from ‘Indiis’ may have been other areas, due to imperfect recording of localities in early lepidopterology. For example, several Fabrician species from ‘Indiis’ were later found to originate from other regions; for example, Hesperia nerva Fabricius (Hesperiidae, presently in the genus Kedestes Watson) is South African, and Hesperia aemulius Fabricius (Riodinidae, presently in the genus Catocyclotis Stichel) is South American.

Donovan (1800–1804) included busiris in his work on Lepidoptera from India and islands of the Indian seas. Latreille ([1824]) maintained ‘Indies’, whereas Westwood (1842) narrowed down its provenance specifically to India, but, as noted earlier, most authors associated the species with Neotropical eudamine hesperiids. In particular, Mabille & Boullet (1919) explicitly gave ‘Amérique méridionale’ as the type locality. Since then, busiris has been included in lists of Neotropical Hesperiidae (Mielke, 2004, 2005).

The original locality of E. contigua was clearly unknown to Walker (1854), who simply described it as coming from G. [George] Milne’s (aka Mylne) collection, without stating the provenance of the type specimen. Butler (1877), Kirby (1891, 1892), Hampson (1901) and all subsequent authors did likewise. Nonetheless, the taxon was very early considered to be African (e.g. Butler, 1877: ‘this and the following are doubtless African species’) or associated with African genera of the Agaristinae (e.g. Kirby, 1891, 1892). It was eventually included in synopses of African species of this subfamily (e.g. Jordan, 1913–1934; Kiriakoff, 1977a).

Hampson (1901) was the only author to mention a possible second specimen of contigua, and one of known origin, by listing a ‘record’ from West Africa, Lower Guinea, in addition to the holotype. However, Hampson did not state the sex of this ‘specimen’, as he did for all the other species that he treated in his account, and there is no other specimen of this species from Lower Guinea in the NHMUK collections under contigua, fabricata or tigrina. Therefore, we surmise that there was no second specimen and that he simply appended information taken from Karsch’s (1895) description of Mitrophrys fabricata, which Hampson wrongly listed as a synonym of contigua. The type locality of fabricata is indeed ‘Nieder-Guinea’. Two decades later, Jordan (p. 7, dated 1913, in 1913–1934), who had access to the then BMNH collections, clearly reconfirmed that contigua was known from ‘Nur 1 Exemplar ohne Vaterland’. Strand (1912) then took the next step, explicitly stating ‘Westafrica (Nieder-guinea)’ as the locality for contigua, evidently referring to the locality offabricata, a taxon that he incorrectly considered to be an aberration of contigua.

The type of contigua has been repinned, reset and possibly partly damaged since it was illustrated by Butler (1877). Its genitalia have been dissected and mounted with the abdomen on slide ‘BMNH Noct. 8601’ by M.L. (Maureen Lane) in 1973 (Fig. 3A). The moth bears a label with the following data: ‘Africa | 39.6.19 | 899.’ (Fig. 1I,J), where the numbers refer to an entry in the museum’s Accessions Register. Examination of the Accessions Register in the NHMUK (BM, 1837–1839) showed that this code associated exactly with specimen no. 899, registered on 19 June 1839, in a large lot of insects bought at the sale of the late Mr Milne’s collection, and that it was entered as a species of Agarista.

We are now left with an intriguing situation: a species seemingly with only two known individuals, viz. those of Drury and Milne (only the latter extant), which had variously been recorded or mentioned as being of American (Neotropical), African (Afrotropical) or Oriental (Indoaustralian) origin. The absence of subsequent records under either the hesperiid or the agaristid name has so far prevented resolution of the species’ distribution.

Even though there is an extant specimen, which led the species to be reassigned subsequently to the exclusively African genus Heraclia, it seems that rediagnosis of this and other related agaristine genera, including reassessment of the taxonomic value of the characters so far employed, should be undertaken within a broader phylogenetic context. As shown earlier, Heraclia busiris is a fairly aberrant member of the genus, sharing some characters with P. tigrina (Fig. 2A–D), so that its African origin is likely, although corroboration would require new material. However, a search of other museums with large holdings of African Agaristinae, namely MfNB and CMNH, yielded no further specimens of Hesperia busiris.

From his correspondence (Drury, 1768a,b,c; Cockerell, 1922; Sherborn, 1937), it is evident that to obtain specimens for his entomological collection, Drury liaised with so many ships’ crews travelling around the globe that unambiguously tracking down the provenance of busiris from his writings is unfeasible. However, he did consider that ‘The Brazills & Africa have yielded a very considerable number more than half the Collection’ (Drury, 1778).

Finally, it is difficult to ascertain when the label stating ‘Africa’ was added to the holotype of E. contigua. The specimen was not considered to bear any indication of locality by Hampson (1901) and, as late as 1977, Kiriakoff (1977a), explicitly stating: ‘The single specimen in British Museum does not bear any indication.’ On the other hand, the label appears to be quite old and yellowed. Moreover, pinned near to the specimen, is a black-and-white version of Butler’s (1877) illustration onto which someone has annotated in pencil ‘W. Africa’ (Fig. 1H). Finally, Mrs Maureen Lane, who prepared dissections for Kiriakoff’s revision (Kiriakoff, 1977b: 2), recorded the specimen as being from Africa, both on the genitalia slide label and in the BMNH slide register. Whether she did this by implication because the taxon was being treated in an exclusively African genus, or after directly reading that particular specimen label, we are unable to determine. If the latter is true, she appears to have failed to inform Kiriakoff that an indication of a locality was present. Nevertheless, this apparently puzzling situation could be easily explained if we consider that the concept of ‘locality’ can be interpreted in a broader or narrower sense by different persons. The handwritten label bearing the indication ‘Africa’ matches those that were added during recuration of BMNH specimens in the second half of 19th century, and was probably added following Butler’s (1877) suggestion that E. contigua had to be African. As all subsequent authors treating contigua took it for granted that this taxon was African, their comments that no locality was present may simply relate to the fact that a more restricted locality or region was unknown.

We are grateful to Kathleen Diston (University Museum of Natural History, Oxford, U.K.) for granting permission to reproduce the illustration of Hesperia busiris by Jones; Robert Blackburn (Macleay Museum, The University of Sydney, Sydney, Australia) for discussions of the Macleay collection and searching for possible Hesperia busiris type(s); John E. Rawlins (Carnegie Museum of Natural History, Pittsburgh) and Wolfram Mey (Museum fur Naturkunde, Berlin) for searching in the agaristine holdings under their charge; John V. Calhoun (Palm Harbor, FL, U.S.A.) for information on historical facts and literature; Ruth Benny, Laura Brown, Lisa Cardy, Paul Martyn Cooper, Hellen Pethers, Lorraine Portch, Roberto Portela-Miguez and Sarah Sworder (Natural History Museum, London, U.K.) for help with literature and archival documents; Gerardo Lamas (Museo de Historia Natural, Universidad Nacional Mayor de San Marcos, Lima, Peru), James S. Miller (American Museum of Natural History, New York, NY, U.S.A.), and Bernard Hermier (Cayenne, French Guiana) for many enlightening discussions; and to John V. Calhoun, Bernard Hermier and Richard I. Vane-Wright for critical reviews of the manuscript and their numerous suggestions and corrections. Last but not least, a grateful thank you goes also to Ian J. Kitching (Natural History Museum, London, U.K.) for thoroughly reviewing the English style.

The authors declare that there are no conflict interests in the manuscript. This work was supported in part by NIH grant GM127390 to NVG.