Updated catalogue and taxonomic notes on the Old-World scorpion genus Buthus Leach, 1815 (Scorpiones, Buthidae)

Abstract Since the publication of the ground-breaking “Catalogue of the scorpions of the world (1758–1998)” (Fet et al. 2000) the number of species in the scorpion genus Buthus Leach, 1815 has increased 10-fold, and this genus is now the fourth largest within the Buthidae, with 52 valid named species. Here we revise and update the available information regarding Buthus. A new combination is proposed: Buthus halius (C. L. Koch, 1839), comb. n. from Portugal and Spain. B. halius is removed from junior synonymy with Buthus occitanus (Amoreux, 1789), and proposed as a senior synonym of B. ibericus Lourenço & Vachon, 2004, syn. n. Moreover, following I.C.Z.N. article 23.9.2 we propose to maintain as valid B. ibericus (nomen protectum) and to consider the disued B. halius as a nomen oblitum. Buthus europaeus tridentatus Franganillo, 1918 is proposed as a junior synonym of B. occitanus (Amoreux, 1789), syn. n. Buthus sabulicola Touloun, 2012 is proposed as a junior synonym of Buthus bonito Lourenço & Geniez, 2005, syn. n. Buthus occitanus tunetanus neeli Gysin, 1969 is proposed as an informal senior synonym of Buthus tassili Lourenço, 2002, informal syn. n. Two taxa are rised to species rank, Buthus nigrovesiculosus Hirst, 1925, stat. n. and Buthus parroti Vachon, 1949, stat. n.. We further confirm the restricted distribution of B. occitanus that is confined to southeastern France and northwestern Iberian Peninsula and does not occur in North Africa. Additionally, Androctonus barbouri (Werner, 1932), comb. n. from the Agadir region of Morocco, is hereby transferred to the genus Androctonus. We summarize and provide a critical appraisal of the diagnostic characters currently in use for the genus. The catalogue section considers the names for species, subspecies and varieties that have been used for Buthus scorpions. Information about types, including collection numbers and localities are included when available. Finally, an annotated listing of synonymies and an updated bibliography are given.


Introduction
Members of the genus Buthus Leach, 1815 are medium-sized scorpions, usually yellowish in colour, with a robust metasoma that ends in a telson with a globular vesicle and a curved aculeus (Fig. 1). Buthus rest during the day in burrows under stones or shrubs and are active from dusk till dawn, although their activity typically peaks at the beginning of the night (Cloudsley-Thompson 1956). They are successful scorpions that, when present, tend to be the most abundant scorpion in their habitat, as for example in the Iberian Peninsula and Morocco. Buthus are usually described as sit-andwait predators, although they can also actively search for prey (Skutelsky 1995, Piñero et al. 2013. Buthus is among the most venomous of all scorpion genera (Chippaux and Goyffon 2008). Buthus venom toxicity is considered much lower in Europe than in North Africa, which can be empirically corroborated by the few severe cases of scorpionism reported for Western Europe when compared to the North African countries (Chippaux and Goyffon 2008).
Buthus exhibits a wide distribution range, spanning over two biogeographic realms, the Palearctic (Western) and the Afrotropical (Udvardy 1975, Olson et al. 2001. Interestingly, none of the chorotypes proposed by Vigna Taglianti et al. (1999) satisfactorily describes Buthus distribution. The genus extends from the temperate Mediterranean areas of south-western Europe to the tropical and sub-tropical grasslands south of the Sahel and into the Horn of Africa, including the semi-arid and arid regions of North Africa and the Middle East (Fig. 2). Although first considered of European origin (Vachon 1952a), current data support the hypothesis that the centre of origin of the genus is North Africa (Lourenço 2002). North Africa harbours a disproportionate number of species (Fig. 2) as well as four of the five main genetic clades found in Buthus, as defined by mitochondrial DNA sequence variation (Sousa et al. 2012, Pedroso et al. 2013 (Fig. 7).
Buthus species are known from 17 countries in Africa: Algeria, Cameroon, Central African Republic, Chad, Egypt, Eritrea, Ethiopia, Guinea, Libya, Mauritania, Morocco, Niger, Senegal, Somalia, South Sudan, Sudan, and Tunisia; five countries in Asia: Cyprus, Egypt (Sinai), Israel, Jordan, and Yemen; and four European countries: France, Italy (Sicily), Portugal, and Spain (Fig. 2). Unidentified Buthus species have also been reported from Burkina Faso, Djibouti, Gambia, Ghana, Guinea-Bissau, Ivory Coast, Nigeria, Iraq, and Lebanon (Fig. 2). No records exist for Saudi Arabia or Syria, however the first might have been confused in the past with citations for the Arabian Peninsula (e.g. Vachon 1952a), although the existence of Buthus in either of these countries cannot be excluded. The frequently cited occurrence of Buthus in Iraq is based on a single specimen, deposited in the Czech National Museum of Natural History (Táborský 1934, Kovařík 1992. As such the actual distribution of the genus remains poorly delimited. Old records from mainland Greece and Turkey are highly doubtful, as these are well-studied areas with no recent Buthus collections (Ersen Yağmur pers. comm. for Turkey) (Fig. 2). The former records most likely refer to the genus Mesobuthus Vachon, 1949. As for the record for Malta, it was considered dubious by Fet and Lowe (2000), although other Buthus reported on islands that were previously regarded as doubtful have turned out to be correct, namely B. kunti Yağmur, Koc & Lourenço, 2011, from Cyprus, described from freshly collected material and B. trinacrius Lourenço & Rossi, 2013, from Sicily, based on 130-year-old material. Other island records include B. occitanus in the Columbrete islands (Castilla and Pons 2007) and B. tunetanus (Herbst, 1800) in the Tunisian islands of Djerba, Kerkena, and Zembra (Vachon 1952a).  Table 2) where known or the best possible approximation. Actual distribution within each country can be much smaller, but detailed distribution information is unknown for the majority of species.
Buthus is the type genus of the Buthidae C. L. Koch, 1837(Koch 1837, 1850, the most diverse family within Scorpiones, with almost half of all known extant scorpion species (1101 of the 2311 known species) (Rein 2016). The Buthidae also includes most of the species venomous to humans (Chippaux and Goyffon 2008). The genus Buthus is the second oldest valid genus of the order Scorpiones C. L. Koch, 1837, only surpassed by the single genus created by Linnaeus in 1758, Scorpio, to accommodate all the scorpion species he described.
Buthus was first proposed by Leach (1815), with Scorpio occitanus Amoreux, 1789 as its type species (Fig. 3). The only diagnostic character for the genus proposed by Leach was the presence of eight eyes "Oculi octo." (Leach 1815, page 391), apparently referring to the pair of central eyes and three pairs of lateral eyes. This character was mentioned as diagnostic for different scorpions early during Scorpiones taxonomy (De Geer 1778, Fabricius 1781. The same diagnostic character was used by Ehrenberg 1828, 1829) for defining several genera and subgenera of scorpions with a varying number of eyes, ranging from six to 12. A more detailed explanation on the usage of the number of eyes in the classification of scorpions is given in Thorell (1876).
The poor description of Leach (1815) led to a rapid increase in the number of species included in the genus, which lacked any internal coherence. This taxonomic conundrum arose through the misidentification of the number of lateral eyes of B. occitanus (Amoreux, 1789), originally stated by Amoreux as three pairs. Several taxonomists of that century realised that there were actually four pairs of lateral eyes (e.g. Gervais 1844b; Simon 1879), but this information was not appreciated by some later authors. It has recently been shown that most Buthidae species (including Buthus) have five pairs of lateral eyes, although in many species two pairs of lateral eyes are much smaller in size and require extreme care and the help of UV light to be recognised (Yang et al. 2013;Loria and Prendini 2014). Ehrenberg (in Hemprich and Ehrenberg 1828) modified the original meaning of the genus to include the species that are now part of Heterometrus Ehrenberg in Ehrenberg 1828 (Family Scorpionidae Latreille, 1802), all with five pairs of lateral eyes. Ehrenberg described several other genera that were soon synonymized with Buthus, at least by some taxonomists, which further exacerbated the taxonomic confusion within Buthus.  (Amoreux 1789a). The work was retrieved from the Biodiversity Heritage Library and images were rearranged for compactness without re-scaling.  (Hjelle 1990); B -Tip of pedipalp movable finger highlighting the three distal granules (distal denticle not included) (Lourenço 2002). Because of the poor definition of the genus, many members (≈100 species) of the Buthidae family with no close relationship to the type species, were included in the genus Buthus up to the mid-20 th century (Vachon 1952a, Levy and Amitai 1980, Lourenço 2002. Unfortunately, this obsolete taxonomy is still in use, for example in many toxicology and venom related papers on scorpions (e.g. Gopalakrishnakone et al. 2015). From 1948 to 1951, Vachon conducted a major taxonomic revision of the genus (compiled in Vachon 1952a), providing a more informative and exclusive definition, retaining only the species that were morphologically similar to the type species and hence restricting also the distribution range of the genus. He proposed two main morphological characters that in combination separate Buthus from all other known Buthidae genera: the central-lateral and posterior-median prosomal keels fused in a lyra shape (character shared with Cicileiurus Teruel, 2007, Leiurus Ehrenberg in Hemprich and Ehrenberg 1828, Mesobuthus Vachon, 1950, and Odontobuthus Vachon, 1950, and the presence of only three granules on the tip of the movable finger (or tarsus) of the pedipalp chela (character shared with Androctonus Ehrenberg in Hemprich & Ehrenberg, 1828, Fig. 4B).
While studying specimens from northwest Africa, Vachon recognized that the genus included a large amount of undescribed diversity. Vachon took a very conservative approach to Buthus taxonomy, recognizing only four species, further split in 12 subspecies, ten of which under B. occitanus, and naming four different varieties, along with other forms with no formal rank, all within B. occitanus. This was partly justified by Vachon's view that Buthus species exhibited a large morphological plasticity, at least in the characters he used to diagnose the different taxa (Vachon 1952a). Stahnke (1972), in his key to Buthidae genera, recognized 21 species and subspecies in Buthus, without further explanation, although this is probably an error resulting from an outdated interpretation of the genus taxonomy. In accordance with the ICZN article 45, none of Vachon's infra-subspecific varieties were included in the Catalogue of the Scorpions of the World . The Buthus Catalogue recognized as good five species and 12 subspecies, although the authors recognized that some taxa were probably not taxonomically good  Subsequently, Rossi (2015) transferred Buthus insolitus Borelli, 1925 to the recently erected genus Gint Kovařík et al., 2013. Lourenço (2003 marked a renewed interest in the taxonomy and diversity of the genus, describing six new species, some of which corresponding to Vachon's infra-subspecific varieties. During the last 15 years, the rate of description of new Buthus species has increased exponentially ( Fig. 5). At present, the genus is composed of 52 species, three of which were described in 2016, making it the fourth most diverse genus of Buthidae, only surpassed by the megadiverse scorpion genera Tityus C. L. Koch, 1836, Centruroides Marx, 1890and Ananteris Thorell, 1891(Rein 2016. Thirty authors have been involved in the description of recent Buthus species, and most species (21) have been described in collaborative studies. Wilson Lourenço is by far the most prolific author, having authored or co-authored 29 Buthus species, 55% of the total.

Materials and methods
Nomenclature and measurements follow Stahnke (1970), except for trichobothriotaxy (Vachon 1974, Fet et al. 2005. All diagnostic morphological characters mentioned in the text refer to adults (or large sub adults) of both sexes, unless otherwise noted.
Most references prior to 1998 cited by Fet and Lowe (2000) are confirmed, but not all original literature could be obtained, and we made some corrections following comparisons with additional sources (Vachon 1952a, Lamoral 1979, Polis 1990, Hendrixson 2006, Dupré 2013. The criteria applied by Fet and Lowe (2000) citing both taxonomic and faunistic works are broadly followed. To the best of our knowledge we cited all works that follow these criteria up to November 2016. Fet and Lowe (2000) cited approximately 180 articles pertaining to the genus Buthus, we added approximately 80 new articles, ten of which were published before 1998. Whenever possible, coordinates for the type localities are provided, using information available in articles or, if not available, finding approximate coordinates with the help of Google Maps (maps.google.com) and the GEOnet Names Server (geonames. nga.mil/gns/html). All coordinates are in WGS 1984 datum, in Latitude/Longitude format, in decimal degrees.
Collections abbreviation codes are listed below. Abbreviation codes follow Sabaj (2016), except for those marked with an asterisk that are not presented there. 1876, now a junior synonym of B. occitanus (Amoreux, 1789) (for further details see Braunwalder 1997, Fet et al. 2002. Finally, although Buthus is considered the nominal genus of Buthidae, Koch (1837) used for the typification of the family, the species Buthus spinnifer Ehrenberg, 1828, which is currently the type species of the genus Heterometrus Ehrenberg, 1828 (Scorpionidae Latreille, 1802), and as such according to ICZN Article 65.2.1. ["type genus was misidentified (that is, interpreted in a sense other than that defined by its type species)] when the family-group name was established") we will submit to the ICZN a petition to fix the type species of the Buthidae. This was not done by  as the authors probably interpret it as ICZN Article 65.2.3. ("type genus was, when established, based on a type species then misidentified"), which does not necessarily require a ruling by the Commission.

Buthus taxonomy
The revised classification of Sharma et al. (2015) is followed, based on the first phylogenomics study on extant scorpions, which resolved most relationships between scorpion families.
Class Arachnida Lamarck, 1801 Order Scorpiones C. L. Koch, 1850Suborder Neoscorpiones Thorell & Lindström, 1885Infraorder Orthosterni Pocock, 1911Parvoder Buthida Soleglad & Fet, 2003 Superfamily Buthoidea C. L. Koch, 1837 Family Buthidae C. L. Koch, 1837 There are no subfamilies in use within the Buthidae, although many have been proposed and rejected (Fet et al. , 2005. Fet et al. (2005) defined six groups within the Buthidae, and placed Buthus in the Buthus group along with 38 additional genera. The phylogenomic study of Sharma et al. (2015) provided strong support for most groups (although generic level sampling was limited), including the sister group relationship of the Buthus group with the remaining members of the familiy. Vachon (1952a) considered Androctonus to be the sister taxa to Buthus. However, the only molecular phylogenetic study addressing the relationships between Buthidae genera that includes both genera , recovered Buthus as the sister taxa to a clade formed by Androctonus and Leiurus, albeit with low support.
There are no taxonomically distinct groups within the genus Buthus, although two "species complexes" are generally recognised. Vachon (1952a, p. 251) suggested that "la «lignée» atlantis se sépare avec facilité de l'ensemble des autres Buthus par la forme élancée des appendices de la queue, de la vésicule et divers autres caractères que nos tableaux de détermination préciseront." This distinction was retained by Lourenço (2002Lourenço ( , 2003 who referred to B. occitanus as a "«complexe de forms»", and subsequently also adopted it in most subsequent taxonomic works describing new Buthus species. According to Lourenço and Geniez (2005), the two complexes are distinguished by the level of keel development, weaker in the atlantis group. However, Lourenço (2005a) subsequently wrote that B. occitanus from Europe has a weak keel development in contradiction with the previous morphological definition. This statement has been used by subsequent authors. For example, Rossi (2012) described B. elongatus Rossi, 2012 as belonging to the occitanus complex, but if the author had applied Vachon's definition it should have included it in the atlantis complex because of the slender metasoma (at least its fifth segment) when compared to B. occitanus. None of the published molecular phylogenies of Buthus supports the existence of the atlantis complex. The results of Gantenbein and Largiadèr (2003) grouped B. atlantis within the species included in the occitanus complex. Although Lourenço and Vachon (2004) acknowledged the work of Gantenbein and Largiadèr (2003), they rejected the phylogenetic position of B. atlantis presented in this study.
Recently, based on the information provided by a cox1 mtDNA tree, Sousa et al. (2012) and Pedroso et al. (2013) have defined a series of groups, (see Table 1). These groups have been renamed to facilitate communication and have been expanded to include all available molecular data with reliable species identifications.
In Fig. 7 the current distribution of these five groups in the Maghreb, the Iberian Peninsula, and southern France is presented, based exclusively on specimens with available molecular data. The group assignment does not necessarily correspond to the species assignment in the original publications.

Diagnostic characters used in Buthus taxonomy
Several morphological traits have been used by recent authors as diagnostic characters (in the sense of Winston 1999) for Buthus species.
Colour is of limited utility in Buthus taxonomy, as the underlying colour varies in tones of yellow, orange, reddish or light brown within and between species. Only one species has a fully dark body, Buthus maroccanus Birula, 1903, in some cases even black. Other species also have the mesosoma of a darker colour than the rest of the body. Of greater taxonomic use are colour patterns, such as darker marks, over a lighter background colour, that can be present on the carapace, the mesosoma or the metasoma; the latter being the more informative.
Adult size may also be diagnostic (Fig. 8). Buthus adult body sizes range from 38 to 90 mm (telson included) (the maximum size of 110 mm reported by Vachon (1952a) is presumably a mistake). Most species have a maximum size between 60 and 70 mm in females, and 55 to 70 mm for males (Fig. 8). On the 5% percentiles we have the smaller species of Buthus (less than 45 mm long), and the larger species of Buthus (more than 85 mm for females and 80 mm long for males).
Two additional meristic traits have been used as diagnostic characters, namely the number of rows of granules on the cutting edge of the movable finger of the pedipalp chela, and the number of pectinal teeth, a sexual dimorphic trait. Variation in the number of rows of granules is not very informative because species show an incremental overlap in the numbers of rows, which range from 8 to 14 (Fig. 9). Pectinal teeth number, although carrying a potentially greater amount of information as they have a wider range to vary from, is actually of limited usefulness because of the interspecific overlap (Fig. 10); female counts range from 18-34 and males from 24-37. Pectinal teeth number is also of limited use due to the lack of any information for several species and the unknown range of variability for many other Buthus species (Fig. 10). It should be noted, however, that Buthus elizabethae Lourenço, 2005 is unique in having male pectines that do not overlap in their proximal portion (Lourenço 2005a).
Trichobothria number and position are not useful for Buthus species diagnosis, as their location shows little variation and have as much intraspecific as interspecific variability (P. Sousa pers. obs.). Conversely, body chaetotaxy (other than trichobothria) is very useful for taxonomy. Vachon (1952a) defined three, albeit diffuse, states in Buthus body chaetotaxy: low ("oligotriche"), high ("polytriche") and medium ("mésotriche"), and used the number of setae on the fifth segment of the metasoma as example of the ranges: low has three or fewer setae, high more than 5-6 setae and medium four setae, although this latter category was fluid. Confusion can further arise from the fact that these categories apply to the metasoma and the pedipalp, and in the same species these two body parts can have different ranges of chaetotaxy. Nevertheless, this is a useful trait, and one that needs to be explicitly stated in species descriptions to avoid misinterpretations. The chaetotaxy of the leg tarsi and mesosoma terguites is also useful.
Most other diagnostic traits in use for Buthus species are found in the metasoma and the pedipalp chela.
The length/width ratio of the first metasomal segment, which is typically square in most species but can be elongated or sturdy in certain species, is informative. This ratio is also applied to the fifth metasomal segment, and Vachon (1952a) further compared the ratios of the first and second segment, and sometimes even the third segment. The number of keel rows in the metasoma segments is also useful, with special attention paid to the presence, and in some cases the relative length (Vachon 1952a), of the median lateral keel in the second, third and fourth segments. Both the degree of development of the inferior median keels of the five segments (except perhaps the forth), and the existence of larger granules may also be used for taxonomic purposes. The number of lateral lobes in the anal arch, either two or three (the latter only in B. atlantis Pocock, 1889 andB. lourencoi Rossi, Tropea &Yagmur, 2013) may be misleading because in some species, or even specimens (Vachon 1952a), a third smaller lobe may be present between the two larger lobes, which has been interpreted as a third state (e.g. Sadine et al. 2015). For instance, Lourenço and Qi (2006) state that in Buthus mariefranceae Lourenço, 2003 the anal arch may sometimes have 3 lobes but this is mentioned neither in the original description nor in Vachon's descriptions. The relationship between the length of the aculeus and the length of the vesicle that form the Telson are also used in Buthus taxonomy. In most species, the aculeus is shorter than the vesicle, or as long as the vesicle at most. For a few species the aculeus is clearly shorter than the vesicle and for another handful of species, the aculeus is clearly longer. The states are defined here using a 10% difference threshold, but other authors have used a 5% difference. This ratio is correlated with the shape of the aculeus, also in use, which can be more or less curved.  The shape of the pedipalp chelae in Buthus taxonomy has gained increased usage in recent years. The shape can be approximated by using the length to width ratio of the chela, which reflects its specific robustness or slenderness. However, in many Buthus species the chela shape is sexually dimorphic, a trait that was first used in a species key by Kovařík (2006), although its use goes back at least to Vachon (1952a). As a measure of sexual dimorphism the pedipalp chelae has three possible states: 1) no sexual dimorphism (male = female); 2) slender chela in male (male > female); 3) chela of male more robust (male < female). There is data available for 29 species (56% of the known species), and from these we can estimate that 38% do not have sexual dimorphism (+/-10% threshold as a cut-off point), 52% of species have males with slender chelae and only 3 species (10%) show males with more robust chelae (Fig. 11). Interestingly, although chelae play a role in mating, defense, and as a sensory organ (van der Meijden et al. 2012), their prime importance in prey capture and handling (Polis 1990) may limit the slendering of female chelae, as these may be more prone to breakage (van der  data was contradictory, and they are further discussed below. Another useful pedipalp trait is the interrupted dorso-median keel of the patella in Buthus rochati Lourenço, 2003(Vachon 1952a. Several partial keys have been published over the years to assist Buthus species identification. However, due to the high rate of new species description (Fig. 5), they have become incomplete and even outdated in their taxonomy (Birula 1903, Vachon 1952a, Lourenço 2003, Lourenço and Vachon 2004, Kovařík 2006 Pedipalp chela length to width ratio Females Males Figure 11. Graphical representation of the relation between female and male pedipalp chela aspect ratio in Buthus species. Only those species with available data from both sexes were plotted. The grey area represents species without perceived sexual dimorphism. Species with males above that area have slender pedipalp chela than females, while species with males plotted below have more robust pedipalp chela than females.  Rossi et al. (2013, 13 species), and for the Iberian Peninsula by Lourenço and Vachon (2004), Rossi (2012), and Teruel and Melic (2015). Unfortunately, the identification of the majority of Buthus species remains difficult, in part because of the limited number of diagnostic characters and the incomplete knowledge regarding their intraspecific variation. Sexual dimorphism of pedipalp chelae is a promising trait, but for many species the male or female is still undescribed, which limits its applicability. Authors are urged to mention the variation on all the traits mentioned here (see  for a nice example) in future Buthus species descriptions.
Confirming the fast pace of new Buthus species descriptions, a new species, Buthus danyii Rossi, 2017 was published from Ghana (Rossi 2017) while this study was under revision. As such we were unable to include this species in the present update.
It is hoped that the present catalogue will facilitate a more precise, informative and comparative description of future species. Buthus are an important component of the scorpions' fauna of North Africa and Western Europe, but it is only now becoming apparent that they are also diverse in the southern Sahara Desert, an area that should be prioritized in future surveys of Buthus scorpions. Etymology. Leach did not provide an explanation for his selection of the genus name. A search on the original usage of the word may shed some light on the intended meaning. Buthus is the Latin form of the Greek name βοῦθος (Bouthos), an unusual name of a winning athlete of the ancient Pythian Games, mentioned by Hesychius and Aristotle (Müller 1848, Christesen 2007). The name was more familiar in antiquity when used in the adage "Βοῦθος περιφοιτᾷ", translated to the Latin as "Buthus obambulat", which translates into "Buthus who wanders", which apparently was applied to stupid and simple people (Müller 1848, Christesen 2007). In Hofmann et al. (1698) the entry for Buthus reads "athleta nobilis, qui bovem integrum unô die devorare solebat; unde natum proverbium in edaces, Buthus obambulat" which roughly translates to "a noble athlete, who used to devour a great ox in a day, and who gave rise to the proverb, Buthus obambulat". Noël (1824) entry for Buthus also refers to an athlete that devoured an ox in a single day, and that this voracity was the origin of the proverb "Buthus obambulat", which according to the author refers to gluttony. Interestingly Noël also states that Buthus, in combination with "βῦς, θúιεν (thuein)", also refers to sacrifice. This opinion shares roots with the meaning of two other words with similar etymology, būthysĭa (used by Nero) that translates to "sacrifice of an ox" and būthytēs (used by Pliny the Elder), that translates to a "sacrificed ox", according to the Gaffiot Latin-French dictionary (Various 2016). Recently Dupré (2016) reached a similar conclusion, although he states that Buthus originates from the composition of the Greek word "Gr. bous, ox; -thouéin [Greek suffix?], killer". Potentially therefore, Buthus refers to a stupid or voracious animal, an ox killer or to a sacrifice of an ox, from the latter two we can interpret it as a powerful and dangerous animal. In our opinion the later makes more sense and agrees well with what was known at the time about the potent venom of Buthus scorpions. As such, it is our opinion that Buthus is a singular masculine Latin word (of Greek origin), which Leach intended as homage to an ancient hero (a trend at that time), and that refers to an animal so venomous that it could kill an ox.  Figure 10 offers an additional zoom to the most diverse region of Buthus species diversity, the Maghreb.
Remarks. There are several old records of Buthus, marked with a question mark above, which have never been found again (independently of the material in which they were based being lost or not). As such, many have not been checked since the genus was reduced in scope by Vachon (1949), or those localities remain doubtful because no Buthus has been collected there since. This is of special significance in countries like Greece and Turkey that have been in recent years reasonably well prospected. Type specimens for several Buthus species described early on were not designated or have since become lost, but this does not necessarily represent a taxonomic problem. For example B. occitanus has no type specimen , but its type locality is well established and no other Buthus species occurs nearby. In this case the designation of a neotype is not justified under the ICZN (article 75.2). However, this is not the case for other species that have neither type specimens nor localities, and that we will further discuss below.
Remarks. It is the largest known Buthus species.

Buthus brignolii
Remarks. it is one of the four known "inland island" species of Buthus that have been found in the Mountainous regions in the heart of the Sahara Desert.

Buthus chambiensis Kovařík, 2006
Buthus chambiensis: Remarks. Based on the colour pattern and pigmentation, the original authors suggested that B. confluens was the closest phylogenetic relative in Morocco to B. ibericus, from the Iberian Peninsula. However, because the presence of three dark bands on the metasoma is shared among several Buthus species, this claim should be further confirmed with additional data.
Type material. 1 F holotype (ZMH N° A20/12), Siwa (approx. 29.17°, 25.46°), Egypt. Distribution. known only from the type locality. Remarks. One of the four known "inland island" species of Buthus that have been found within the Sahara Desert, although in this case from an Oasis. It is also one of the largest known Buthus species.

Buthus elhennawyi Lourenço, 2005
https://science.mnhn.fr/institution/mnhn/collection/rs/item/rs8637 Distribution. this species is known from Niger and Senegal, from a single locality in each country, which are almost 2,000 Km apart.
Remarks. We used the location of Fété-Olé given in Vincke et al. (2010), a locality that has been part of long term ecological studies, to map this locality, instead of the original spelling "Félé-Olé".

Buthus elizabethae Lourenço, 2005
https://science.mnhn.fr/institution/mnhn/collection/rs/item/rs8638 Distribution. this species is known from Guinea and Senegal, Remarks. Given the geographical proximity, it is possible that the Buthus material reported to have been found in Guinea-Bissau might very well correspond to this species. Unfortunately the Guinea-Bissau material was lost in a fire, and hence only newly collected material could confirm this possibility.
Remarks. According to ICZN article 45.5, Vachon's (1949) infrasubspecific name is unavailable. Although the name was published before 1961, it was only used as infrasubspecific by all subsequent authors. Distribution. this species is known from the southern Iberian Mediterranean coast, close to Marbella.

Buthus elongatus
Remarks. The second locality given by Rossi as Alicante (sic), had a typographic error, as the coordinates given by the author, together with their map in Fig. 7, provide sufficient evidence for the correct mapping of this locality. Both localities are under severe anthropomorphic pressure.
Remarks. Another of the four known "inland island" species of Buthus that have been found in a Mountainous region in the heart of the Sahara Desert. The type locality was pinpointed following the map provided by the authors (fig. 90). Distribution. this species seems to have a wide distribution range in the western part of the Iberian Peninsula, although the limits of its distribution remain poorly defined.

Buthus ibericus
Remarks. B. ibericus was first described from Spain and subsequently reported for Portugal (e.g. Sousa et al. 2010, Rossi 2012). Simon (1879) synonymized Androctonus halius with B. occitanus based on its type locality, which was wrongly stated to be Spain (page 98), because Simon considered B. occitanus (as B. europaeus) to be the only good species in Spain. Koch's original description is not by itself enough to synonymize both species. However, Koch's fig. 383 illustration of Androctonus halius includes a basal lobe in the movable finger ( Fig. 13a and B), which is the key diagnostic character for B. ibericus Vachon 2004, Rossi 2012), together with its type locality (Portugal), supports this synonymy. Nonetheless, according to the I.C.Z.N. article 23.9, the junior synonym can remain valid to maintain taxonomic stability. To our knowledge the name B. halius has not been used since 1879 (article 23.9.1.1), and more than 25 works have been published in the past 12 years by more than 10 authors using the name B. ibericus (article 23.9.1.2). Not all works are cited here because they are neither taxonomic nor faunistic. As such we propose to maintain as valid the junior synonym B. ibericus (nomen protectum) according to prevailing usage (article 23.9.1), and to consider the senior synonym B. halius a nomen oblitum. Rossi, 2012 (page 278), for B. ibericus, states erroneously "Sexual dimorphism is not noticeable in the chela manus", which is in contrast to the original descriptions of both Koch (1839) and Lourenço and Vachon (2004) and to the complementary description in Teruel and Pérez-Bote (2005). In both works it can be observed that the male manus is more bulbous than the female's, which results in the males having a smaller length to width ratio. The type locality of B. ibericus was precisely located following the indications of Iñigo Sánchez, the original collector. Distribution. Known only from the type locality. Remarks. Fet and Lowe (2000) considered the locality as probably wrong since at that time no other Buthus had been collected again in Yemen. However, Lourenço's (2008) Buthus yemenensis revalidated Ehrenberg species' by providing concrete proof for the existence of Buthus species in this country. Doubst about B. intermedius type locality were the only evidence given by Kovařík (2006) for its synonimization with Buthus intumescens (Ehrenberg in Hemprich & Ehrenberg, 1829).  : Kovařík 2006 (part): 10-11, 15, fig. 20;Kaltsas et al. 2008 (part): 215;Rossi 2013: 191-192;Rossi, Tropea and Yağmur 2013: 3, 6-8.

Buthus lienhardi Lourenço, 2003
Buthus Distribution. the species is known from a wide range across the High-Atlas Mountains.
Remarks. Vachon (1949) infrasubspecific name is not available as explained previously.  Remarks. The type locality is now part of the large city of Tripoli. It is fairly unlikely that the species still occur within the boundaries of the city given the level of urban development. However, Mellaha, which was originally a military airport, is now the Mitiga International Airport, where large patches of unconstructed ground that may be suitable fot the species still exist.
Remarks. Vachon (1949d: 358) corrected what he considered Simon's incorrect original spelling of "mardoche" to "mardochei", since the form was named as a patronym after its collector, Rabbi Mardoché. Nevertheless, it is our understanding that this was an unjustified emendation, because the ICZN article 31.1 admit the use of a noun in apposition as was the case with "mardoche", however the ICZN article 33.2.3.1 admits the prevalence of this emendation as it continues to be attributed to "the original author and date" and is "in prevailing usage" and as such we refrain from any change to the name. Vachon (1949Vachon ( , 1952 also established that the species occurs roughly between Essaouira and Agadir, but not near the coast where it is replaced by B. atlantis. Remarks. Vachon (1949) infrasubspecific name is not available as explained above.
Remarks. it remains the only known Buthus species with a uniformly darkened body.
Remarks. All material collected outside of the range here proposed should be considered as Buthus sp. Only the re-examination of those specimens could reveal their appropriate identity. Vachon (1952a) included animals from the west (Atlantic) coast of Morocco, from Kenitra to El Jadida, within his definition of B. o. occitanus. As explained above this material is now considered not to be part of B. occitanus and thus remains unnamed. We opted to leave Androctonus eurilochus in synonymy with B. occitanus, despite the uncertain about the species provenance, to help to stabilise the genus' taxonomy. To further bring stability to Buthus taxonomy we propose Franganillo's subspecies, B. o. tridentatus, as a junior synonym of B. occitanus. Although the type specimen is not known (if it ever existed as such), its type locality is clearly stated as Javier, in Navarre, Spain. Extensive sampling on the left bank of the Ebro River (Sousa 2017) indicates that only B. occitanus occurs in this part of Spain and hence we here propose this new synonymy.
Type material. Holotype lost according to Fet and Lowe (2000), Algeria. Vachon (1949cVachon ( , 1952a wrote that the types came from Alger without further explanation. Distribution. the species is currently distributed across Algeria, Morocco and Tunisia. Remarks. Because of the description of new species from Algeria, and the lack of both type specimen and locality (beyond the country), a neotype for B. paris is necessary to stabilize the taxonomy of Algerian Buthus, which may challenge the status of some newly described species. This is further complicated by recent diagnoses of B. paris (at least in part: Lourenço 2003, Kovařík 2006, Lourenço and Sadine 2016) that differ from those offered by Vachon (1952a). The differences between the different diagnoses include the number of rows in the movable finger, the aspect ratio of the first metasomal segment, the body chaetotaxie, the aculeus to vesicle length and the type of sexual dimorphism of the pedipalp chelae. Vachon (1952a) studied a large number of specimens from the entire Maghreb region, unrivalled by any subsequent study, which leads us to consider Vachon's description as the "gold standard". Vachon himself stated that most of the specimens used in his redescriptions were stored at the MNHN. If this material is ever located, it should have priority in the designation of a neotype. Although Vachon (1952a) did not formally describe any varieties within B. paris, he split the specimens that compose the species into three regions: 1) The typical region (from Algiers to northern Tunisia); 2) Specimens from the Oujda region (Morocco); 3) Those from northern Morocco and the Middle Atlas flanks. Interestingly the split of B. paris into these three regions corresponds well with the distribution of the genetic variability in the cox1 gene (Gantenbein and Largiadèr 2003, Sousa et al. 2012, Pedroso et al. 2013, placing B. paris in two different groups: in occitanus which include all B. paris specimens from regions two and three, and tunetanus, which include B. paris specimens from the typical region one, along the split of the two cox1 groups in the middle of Algeria (Fig. 7). If confirmed, this will mean that B. paris does not occur in Morocco. It is unclear if the variety from the third region above might correspond to B. confluens Lourenço, Touloun & Boumezzough, 2012, although these authors (page 22) refrained from suggesting this possibility because they could not find any of the material used by Vachon, purportedly to be in the MNHN, to describe this variety.
Remarks. B. parroti was first described as a subspecies of B. atlantis. The two species occupy different habitats in Western Morocco; B. parroti is a forest species and B. atlantis is a sand dune dweller (Vachon 1952a). Furthermore, B. atlantis is clearly larger than B. parroti. Additionaly, the first metasomal segment of B. parroti is wider than long, the aculeus is shorter than the vesicle and the anal arch has only two lobes (Vachon 1952a), while in B. atlantis this segment is longer than wide, the aculeus is as long as or longer than the vesicle and the anal arch has three lobes (Vachon 1952a). Three additional Buthus species occur in the same area of Morocco as B. parroti, namely B. elmoutaouakili Qi, 2006, B. mardochei Simon, 1878 andB. mariefranceae. B. parroti can be distinguished from all three species by the presence of macrosetae in the terguites (Vachon 1952a Distribution. known only from the type locality. Remarks. type locality toponym derived from the coordinates given in the original description, as the type locality given was only Region of Sanguéré-Djoi, Cameroon.

Buthus pusillus Lourenço, 2013
Buthus pusillus: Lourenço 2013: 64-67, fig. 1 Remarks. The locality we present here corresponds to the coordinates given in the paper, although these do not agree with the elevation also reported in the paper: 2150 m a.s.l. This is very close to the maximum altitude of the highest peaks of the Djurdjura Mountains, and much higher than the 935 m a.s.l. of Tizi Oumalou.
Remarks. The Tin Tazarif coordinates given here, standing at 880 m a.s.l., do not match the altitude given for the point by Lourenço (2002), of 1.800 m a.s.l., but correspond well to the map location given by the author in Figure 6 of the same article. Nevertheless there are several locations in the Tassili N'Ajjer Mountains at or above 1.800 m a.s.l., located closer to Jebel Azao, its highest peak. The species ranges from the Hoggar Mountains to the Tassili N'Ajjer Mountains, including Ghat in Libya. Gysin's name, B. o. t. neeli, is not available under the ICZN article 10.2 and 45.5, as already stated by Fet and Lowe (2000). Therefore, it cannot enter formal synonymy. Nevertheless, because Gysin's description (1969) brings relevant taxonomic information (figures and new localities) for B. tassili, we have decided to clearly state this new informal synonymy. The specimens studied of both species come from the Algerian Hoggar Mountains and share a typical darkened fifth segment of the metasoma (see B. nigrovesiculosus above).

Buthus trinacrius Lourenço & Rossi, 2013
Buthus trinacrius: Lourenço and Rossi 2013: 10-12, fig. 1 Lourenço and Rossi (2013) report that the collector is not mentioned in Simon's notes, Kraepelin (1901), in his list of all the scorpion material present in the MNHN Paris, writes that Letourneur collected the Buthus material from Sicily and Corfu (Greece), which causes doubts regarding the correct collection locality of the specimens used to describe this species. Simon (1879) had doubts about the actual existence of Buthus in Sicily (when examining the specimens that would eventually be designated as type material for B. trinacrius): "il habite probablement aussi le midi de l'Italie et la Sicile", and Buthus has never been found in mainland Italy. Furthermore, the authors also claimed that Simon (1910) "referred to the almost impossibility to distinguish Buthus populations from North of Africa with those from Spain and Sicily", which is only partially correct. Simon (1910) solely refered to the distinction of Buthus populations of Algeria and Spain from those of Egypt, although in a subsequent paper, Simon does state that Buthus exist in Sicily without any further comments.
Type material. Types lost according to Fet and Lowe (2000), Tunisia.
Distribution. The species is currently distributed across Algeria, Libya, Morocco and Tunisia, and doubtfully in the island of Malta.
Remarks. Because of the description of new species from Tunisia, and the lack of both type specimen and locality (beyond the country), a neotype for B. tunetanus is necessary to stabilize the taxonomy of Tunisian Buthus. As explained for B. paris, this is further complicated by recent diagnoses of B. tunetanus that differ from those offered by Vachon (1952a), and as such if Vachon's B. tuntetanus material is found in the MNHN it should be given priority in the future designation of a neotype. Vachon (1952a) did not formally described any variety of B. tunetanus, but he again split the specimens that compose the species into four regions: 1) the typical region, cor-responding to north and central Tunisia; 2) the southern montane region of Algeria, specimens from which have subsequently been described as B. tassili; 3) the Algerian Saharan Atlas and the southern region of the High Plateau; and 4) the disjunct desert regions of southern Tunisia, western central Algeria and eastern central Morocco. It is unclear whether region 3 or 4 might either correspond to B. dunlopi or B. saharicus. As explained in Fet and Lowe (2000), the name A. (Leiurus) t. genuinus refers to the nominotypical form of the species and as such the adjective "genuinus" is not an available subspecific name.
Distribution. Known only from the type locality.
Remarks. The species B. barbouri was described by Werner from Agadir, Morocco. However, this species cannot be linked to any of the Buthus species known from the reported type locality. This problem was already recognized by Vachon (1952a) that considered this species to have been "imperfectly described", but then failed to make any taxonomic or nomenclatural act to fix it. Buthus barbouri and B. marrocanus are the only fully dark Buthus species known from Morocco. B. marrocanus was described from the Rabat region, 500 km distant from Agadir, and is distinguished from the former species by the shape of the inferior lateral keels of the fifth segment of the metasoma and by the number of pectinal teeth (Vachon, 1952a). Because of the chaetotaxy of the pedipalps Vachon (1952a) considered B. barbouri to be closer to Androctonus mauritanicus (Pocock, 1902). In agreement with these diagnostic characters, and Vachon's (1952) opinion, we transfer B. barbouri to the genus Androctonus, with the new combination Androctonus barbouri (Werner, 1932). The relationship between A. barbouri and A. mauritanicus bourdoni Vachon, 1948, a subspecies of A. mauritanicus (Pocock, 1902), another black Androctonus species that also occurs in the Sous River valley, should be investigated.

Acknowledgments
Pedro Sousa is grateful to Arie van der Meijden for the opportunity he was given to sample for Buthus and other scorpions in Morocco, Tunisia, and Egypt, and also for the financial support to visit the MNHN, as part of an unrelated research project. Pedro Sousa would also like to thank Mark Judson, Elise-Anne Leguin and Wilson Lourenço who welcomed him at the MNHN. José Templado Gonzalez, Javier Sanchez Almazan and Carolina Noreña Janssen also welcomed Pedro Sousa during is stay at the MNCN.
We would like to thank our friend Sérgio Henriques who provided us with photographs of B. nigrovesiculosus with the kind permission of Janet Beccaloni from the NHMUK; and to extend our thank to Anja Friederichs and Jason Dunlop from the ZMB who provided us with photographs of B. intermedius and B. intumescens. We would also express our thanks to Oulaid Toulon who provided clear information about the location of the type material of B. draa and to Iñigo Sánchez who gave us the coodinates of the type locality of B. ibericus.
For their help in providing us with otherwise difficult-to-obtain bibliography, we thank Wilson Lourenço and Andrea Rossi, as well as Alison Corley, Elisa Mora, Gerard Dupré, Petr Dolejš and Petr Kment (Czech National Museum of Natural History, Prague). Isabel Sá from the Library of the Faculty of Sciences of Porto´s University helped us to obtain publications located in the Library of Coimbra University. Porto's Public Library helped us to obtain some difficult-to-locate Portuguese articles. Several users of the webpage etimologias.dechile.net/ provided useful comments regarding Buthus etymology. We also express our gratitude to Miguel Angel Zarazaga for his expert opinion in the interpretation of several articles of the ICZN regarding nomenclature problems in Buthus, although all conclusions about such matters in the present manuscript are our responsibility alone. We extend our thanks to Jan Ove Rein for keeping an updated and freely available list of valid scorpion species names (www.ntnu.no/ub/ scorpion-files/). Wilson Lourenço as Editor, Lionel Monod, and an additional anonymous reviewer provided helpful comments on an earlier draft of the manuscript.