HEMIPENIAL MORPHOLOGY OF THE SOUTH AMERICAN XENODONTINE SNAKES, WITH A PROPOSAL FOR A MONOPHYLETIC XENODONTINAE AND A REAPPRAISAL OF COLUBROID HEMIPENES HUSSAM ZAHER BULLETIN OF THE AMERICAN MUSEUM OF NATURAL HISTORY NUMBER 240 NEW YORK: 1999 Recent issues of the Bulletin may be purchased from the Museum. Lists ofback issues of the Bulletin, Novitates, andAnthropological Papers published during the last five years are available at World Wide Web site http://nimidi.amnh.org. Or address mail orders to: American Museum of Natural History Library, Central Park West at 79th St., New York, N.Y. 10024. TEL: (212) 769- 5545. FAX: (212) 769-5009. E-MAIL: [email protected] HEMIPENIAL MORPHOLOGY OF THE SOUTH AMERICAN XENODONTINE SNAKES, WITH A PROPOSAL FOR A MONOPHYLETIC XENODONTINAE AND A REAPPRAISAL OF COLUBROID HEMIPENES HUSSAM ZAHER Traina Postdoctoral Research Fellow, Department ofHerpetology, American Museum ofNatural History Assistant Professor, Department ofZoology University ofSao Paulo, Caixa Postal 11461 Sao Paulo, SP 05422-970, Brazil BULLETIN OF THE AMERICAN MUSEUM OF NATURAL HISTORY Number 240, 168 pages, 95 figures, 1 table, 2 appendices Issued May 14, 1999 Price: $16.80 a copy Copyright C American Museum ofNatural History 1999 ISSN 0003-0090 CONTENTS Abstract .................................................................... 3 Introduction ................................................................... 3 Hemipenial Morphology in Snake Systematics, with Emphasis on the Xenodontines 5 .... Sources of Information on Hemipenial Morphology of Colubroid Snakes .... .... 7 Method of Hemipenial Preparation ........... .................................. 7 Acknowledgments ............................................................ 8 Hemipenial Terminology ...................................................... 8 Hemipenial Morphology of Colubroidea and its Bearing on Monophyly of its Constituents 12 ................................................................. Outgroups .................................................................. 13 Tropidophioidea . 13 ........................................................... Acrochordoidea . 13 ........................................................... Colubroidea 13 ................................................................. Elapidae 14 .................................................................. Viperidae 14 ................................................................. Atractaspididae 15 ............................................................ "Xenodermatinae" ... 17 ....................................................... Pareatinae ................................................................ 19 Calamariinae 20 .............................................................. "Homalopsinae".. ........................................................... 20 "Boodontinae"... 23 ........................................................... "Pseudoxyrhophiinae".. 24 ..................................................... Colubrinae 25 ................................................................ Psanunophiinae . 28 ........................................................... Pseudoxenodontinae . 28 ....................................................... Natricinae ................................................................ 32 Dipsadinae 33 ................................................................ Xenodontinae . 35 ............................................................. The West Indian Genera of Xenodontinae and Allied Taxa ..... 40 .............. The Mainland Genera of Xenodontinae .................................... 44 Comments on Some Genera ofProblematic Assignment ..... 46 ................ Comparative Morphology of Hemipenial Patterns in Colubroids: a Preliminary View ... 47 Generic Synopses of Hemipenial Features in the Xenodontinae ..... 48 ................. "Xenodontines" Incertae Sedis .............. 77 .................................... Conclusion 83 .................................................................. References 84 .................................................................... Appendix 1. List of Genera Placed in the Families Atractaspididae and "Colubridae" .... 96 Appendix 2. Colubroid Specimens Examined for Hemipenial Morphology ..... 98 ....... 2 1999 ZAHER: HEMIPENIAL MORPHOLOGY OF XENODONTINE SNAKES 3 ABSTRACT The New World xenodontine "colubrids" rep- calyces are merely enlarged capitular calyces of resent two immunologically distinct assemblag- the asulcate/medial surfaces of the lobes. In the es-the Central and South American lineages, Xenodontinae, thebodycalyces are almostalways neither ofwhich has been well diagnosed to date. separated from the calyces of the capitulum by a Ifollow this nomenclature andrecognizetheCen- more orless developed overhang (exceptin afew tral American lineage as containing 22 genera. genera). This overhang is generally retained on This clade is supported by the synapomorphy of the hemipenes where the asulcate/medial surfaces a sulcus spermaticus bifurcating within or at the of the lobes are nude (e.g., Psomophis), which base of the capitulum (Cadle, 1984; Myers and supports the view that thebody calyces were sec- Cadle, 1994). The remaining xenodontines con- ondarily lost. Body calyces are also found on the surface ofthehemipenial body inits asulcate side stitute a total of 68 presently recognized genera, (e.g., Philodryas, Pseudablabes, Xenoxybelis). of which 41 are placed in the subfamily Xeno- Because body calyces are interpreted as modified dontinae sensu stricto. The other 27 genera are "capitular calyces," which are restricted to the considered incertae sedis, pending further re- lobular region and crotch, the presence of these search. The Xenodontinae sensu stricto are hy- structures far on the hemipenial body is here pothesized as being monophyletic on the basis of viewed as a more derived state where the body the following hemipenial synapomorphies: (1) calyces extend from the lobes to the body. presence of enlarged lateral spines on the hemi- Various presumably monophyletic units are de- penial body, and (2) twodistinctly ornamentedre- fined within the Xenodontinae sensu stricto. Con- gions on the lobes, the asulcate surface bearing ophis, Heterodon, and Farancia are clearly as- enlarged spinulate or papillate calyces (= body signed to the Xenodontinae sensu stricto. calyces). Some taxa recognized as Xenodontinae The hemipenial morphology of various supra- sensu stricto lack body calyces but have a nude generic "colubrid" taxa are described and com- areainthe same topographical position (e.g., Pso- pared. The variation of some hemipenial features mophis, Tropidodryas). This pattern is viewed as within the colubroid radiation, as well as their the result ofsecondary loss. The rationale forthis bearing on the higher level phylogeny of colu- conclusion is based on the hypothesis that body broids, is investigated. INTRODUCTION No unambiguous synapomorphy is known dontines will refer solely to a South Amer- to diagnose the family "Colubridae." Addi- ican clade as redefined in the present study tionally, phylogenetic relationships within (containing 41 genera; table 1); (2) the sub- the "Colubridae" remain largely unknown family Dipsadinae or dipsadines, which (Cadle, 1994). Although some monophyletic corresponds roughly to Cadle's (1984b) and subgroups within the family have been di- Myers and Cadle's (1994) Central American agnosed successfully using morphological assemblage (containing 22 genera; table 1); features (e.g., the Thamnophiini [Rossman and (3) the New World xenodontines or and Eberle, 1977] and the Psammophiinae "xenodontines" sensu lato, which are com- [Bogert, 1940; Bourgeois, 1968]), the vast posed of the Xenodontinae, Dipsadinae, and majority of "colubrid" genera are assembled 27 genera of uncertain assignment (Cadle, on the basis ofoverall similarity, unpolarized 1984a, 1984b, 1984c, 1988, 1994; table 1). character states, and/or immunological dis- The terms South American xenodontines tances. The South American xenodontines and Central American xenodontines will represent an example ofimmunologically co- be also used herein as interchangeable with hesive assemblages (Cadle, 1984a, 1984b, Xenodontinae and Dipsadinae, respectively. 1984c, 1988) with no described synapomor- The "xenodontines" are here viewed as like- phies (Myers and Cadle, 1994). ly paraphyletic. Genera referred to the "xe- For the sake of clarity, the following nodontines" are listed in table 1; all are New names will be used in the text to refer to World snakes. three distinct "assemblages": (1) the sub- The present study reviews the hemipenial family Xenodontinae sensu stricto or xeno- variation foundin theColubroidea, providing 4 BULLETIN AMERICAN MUSEUM OF NATURAL HISTORY NO. 240 TABLE 1 List ofGenera Placed Within the Xenodontinae and Dipsadinae'l Xenodontinae Dipsadinae Alsophis Manolepis Adelphicos Antillophis Oxyrhopus Amastridium Apostolepis Phalotris Atractus Arrhyton Philodryas Chersodromus Boiruna Phimophis Coniophanes Clelia Pseudablabes Cryophis Conophis Pseudoboa Dipsas Darlingtonia Pseudoeryx Eridiphas Ditaxodon Psomophis Geophis Drepanoides Rhachidelus Hypsiglena Elapomorphus Saphenophis Imantodes Erythrolamprus Siphlophis Leptodeira Farancia Tripanurgos Ninia Helicops Tropidodryas Pliocercus Heterodon Umbrivaga Pseudoleptodeira Hydrodynastes Uromacer Rhadinaea Hydrops Uromacerina Sibon Hypsirhynchus Waglerophis Sibynomorphus laltris Xenodon Tretanorhinus Liophis Xenoxybelis Trimetopon Lystrophis Tropidodipsas Urotheca Xenodontinae and Dipsadinae Incertae Sedis Calamodontophis Echinanthera Nothopsis Tachymenis Carphophis Emmochliophis Opisthoplus Taeniophallus Cercophis Enuliophis Pseudotomodon Tantalophis Contia Enulius Ptychophis Thamnodynastes Crisantophis Gomesophis Rhadinophanes Tomodon Diadophis Hydromorphus Sordellina Xenopholis Diaphorolepis Lioheterophis Synophis aI have taken as a starting point the list of Dowling and Duellman (1978). However, Dowling and Duellman's (1978) and Jenner's (1981) tribal arrangements are notrecognized here. Various additions andcorrections tothislist have been madeinordertoinclude generaomittedbythese authors aswell astaxonomic changesproposedrecently. ThecorrectionsregardinginvalidnamesorjuniorsynonymshavebeenlargelybasedonWilliamsandWallach(1989). The additions ofnew taxa as well as the taxonomic rearrangements are as follows: Cercophis (Hoogmoed, 1982), Crisantophis (Villa, 1971), Echinanthera (sensu Myers and Cadle, 1994), Elapomorphus (sensu Ferrarezzi, 1993), Emmochliophis (Fritts and Smith, 1969; see Hillis, 1990), Enuliophis (McCranie and Villa, 1993), Hydromorphus (sensu Crother, 1989a), Phalotris (sensuFerrarezzi, 1993),Philodryas (sensuThomas andFernandes, 1996,including Platyinion), Pseudoleptodeira (Dowling and Jenner, 1987), Psomophis (Myers and Cadle, 1994), Rhadinaea (sensu Myers, 1974; Myers and Cadle, 1994; see also Di-Bernardo, 1992), Rhadinophanes (Myers and Campbell, 1981), Sibon (sensu Kofron, 1985a,b), Synophis (sensu Hillis, 1990), Tantalophis (Duellman, 1958b; see also Myers and Campbell, 1981), Tropidodipsas (Wallach, 1995), Tropidodryas (Thomas and Dixon, 1977), Urotheca (sensuMyers and Cadle, 1994; but see Savage andCrother, 1989). Genera ofuncertainphylogenetic position withinthe xenodon- tines demarcated as incertae sedis. diagnoses for the major groups of colubroid xenodontines is proposed and discussed. A snakes, with emphasis on their hemipenial detailed discussion ofthe newly defined Xe- morphology. Hemipenial evidence for the nodontinae is thenprovided, with a summary monophyly of these groups, including the of diagnosable groups within that group as two groups of "xenodontines," is evaluated, well as of taxa of problematical placement. and amonophyletic clade ofSouthAmerican This study also provides arefinedhemipenial 1999 ZAHER: HEMIPENIAL MORPHOLOGY OF XENODONTINE SNAKES 5 terminology for the South American xeno- hemipenis may be bilobed or single, and the dontines and detailed descriptions of the sulcus spermaticus is single in some. Mem- hemipenial morphology ofeach genusplaced bers of the South American lineage usually in the Xenodontinae, as well as the ones re- have a noncapitate or semicapitate hemipen- tained incertae sedis. is, with the sulcus bifurcating often near the base ofthehemipenis andusually onthebas- HEMIPENIAL MORPHOLOGY alhalfofthe organ." The author stressedthat IN SNAKE SYSTEMATICS, WITH both definitions fail to provide any synapo- EMPHASIS ON THE XENODONTINES morphy for these groups. Indeed, the defini- tion for the South American assemblage can Hemipenial morphology has beenusedex- be applied to various viperid, elapid, andAf- tensively in snake systematics, providing a rican "colubrid" snakes. However, the con- large array of phylogenetically informative dition of a sulcus spermaticus bifurcating data. After Cope's (1893, 1894, 1895, 1900) within the capitulum is unique to the Dip- innovative works on the hemipenial mor- sadinae and may represent a synapomorphy phology of snakes, Dunn (1928) and Bogert ofthis subfamily (see below). Myers andCa- (1940) attempted classifications of the New dle (1994: 27) already suggested that this World and African "colubrid" faunas, re- character, along with reduction or loss ofbi- spectively, relying on various hemipenial lobation and unicapitation, are derived fea- characters. tures characterizing the Dipsadinae. Such a Subsequently, various authors attempted definition, however, also encompasses Dia- suprageneric classifications of the "colu- dophis and Carphophis for example (except brids" in which the hemipenis was accorded for their noncapitate condition), two genera great importance but with mixed success considered by Cadle (1984a, 1984b, 1984c) (e.g., Dowling, 1975; Dowling and Duell- as only remotely related to the Central and man, 1978; Jenner, 1981; Jenner and Dowl- South American assemblages. This problem ing, 1985). A number of important works willbe treated inmore detail in the following dealt exclusively with the hemipenial mor- section. phology of particular groups of snakes (Vel- Dowling (1975) and Dowling and Duell- lard, 1928, 1946; Domergue, 1955, 1962; man (1978) assigned almost all "xenodon- Branch, 1981, 1986). tine" generato various tribes, relyingheavily Dunn (1928) was the first to recognize the on a few generalized hemipenial and osteo- "xenodontines" (his Ophiinae) as a distinct logical features. Subsequently, Jenner, in an New World "colubrid" group, characterized unpublished dissertation (1981), proposed by the presence of a bifurcated sulcus sper- for the first time to divide the xenodontine maticus in the hemipenis. However, this con- assemblage ofDunn andDowling intwo dis- dition is likely an ancestral state because, tinct groups which she called the "Northern apart from being present in almost all colu- and Southern groups." This partition is very broid lineages (Cadle, 1984c, 1987), it is un- similar to Cadle's (1984a, 1984b, 1984c) ambiguously present in the two successive subsequent proposals. However, most of the outgroups to the colubroids, Acrochordoidea tribes recognized by Jenner (1981) ontheba- and Tropidophioidea (see below). sis of exclusively hemipenial characters rep- A recent attempt to synthesize our knowl- resent paraphyletic groups (Cadle, 1984a, edge on this group was that ofCadle (1984c: 1984b, 1984c). Dowling et al. (1983) rec- 646), who suggested that the Central and ognized two subfamilies corresponding to South American xenodontine assemblages, Jenner's (1981) Northern and Southern previously recognized by him on the basis of groups, the Xenodontinae (including the albumin immunology, correspond to two dif- tribes Diaphorolepidini, Philodryadini, Pseu- ferent hemipenial types: "Members of the doboini, and Xenodontini) and the Dipsadi- Central American lineage can generally be nae (including the Alsophiini, Dipsadini, and characterized as having a capitate, calyculate Leptodeirini). Subsequently, Jenner and hemipenis with the sulcus spermaticus bifur- Dowling (1985) transferred the Alsophiini to cating (if at all) within the capitulum. The the Xenodontinae, without any discussion. 6 BULLETIN AMERICAN MUSEUM OF NATURAL HISTORY NO. 240 However, this new arrangement is supported ics. First, apart from some osteological fea- by the similarhemipenial morphology shown tures (mostly on vertebrae), the hypotheses by the Alsophiini and the Xenodontinae of relationships between the taxa belonging (Maglio, 1970; Thomas, 1976; Dowling and to the xenodontines are mostly based on a Duellman, 1978). Jenner and Dowling restricted set ofhemipenial structures and on (1985) also formalized the tribe Pseudoboini, general molecular similarity. Second, various previously recognized by Bailey (1939a, influential nomenclatural decisions have 1967). However, their inclusion ofTropidod- been made and subsequently changed with- ryas and Saphenophis in the Pseudoboini out discussions of evidence (compare Dowl- rendered the redefined tribe polyphyletic ing, 1975; Dowling and Duellman, 1978; (Dessauer et al., 1987; Zaher and Caramas- Jenner, 1981; Dowling et al., 1983). Another chi, 1992; Zaher, 1994a). Suchdisagreements important, but unrelated, nomenclaturalissue concerning the systematics oftheNewWorld is Smith's (1964) choice, as first reviser, of xenodontines is not, as one may think, due the family-group name Heterodontidae Gray, to the practical limitations inherentin theuse 1845 (a nomen oblitum) over the name Xe- ofonly one source ofcharacter datain aphy- nodontina Bonaparte, 1845. However, he logenetic analysis, but rather results from usedthe subfamily Heterodontinae to accom- oversimplified interpretations of structures, modate the genera Heterodon, Lystrophis, leading frequently to erroneous assumptions Xenodon, and Leioheterodon, thus creating a of homology (see also Myers and Cadle, highly heterogeneous group. Rossman and 1994: 26). Wilson (1964) rejected Smith's (1964) no- Although Dowling made important contri- menclatural proposition for this reason and butions to the knowledge of snake hemipen- also because the name Heterodontidae Gray, ial morphology and phylogeny, the taxonom- 1851, is widely used for a family of sharks ic arrangements proposed by him (Dowling, (type genusHeterodontus). Smith agreedand 1975, Dowling and Duellman, 1978) have withdrew his formal application to the Inter- been criticized by Cadle (1984c) and Whis- national Commission of Zoological Nomen- tler and Wright (1989). The classificatory clature for validation ofthe snake name Het- scheme proposed later by Dowling et al. erodontinae (fide Rossman and Wilson, (1983) was also rejected by Cadle (1984c) 1964). Smith et al. (1977) retained the Xe- (however, see Blackburn [1985] and Smith nodontinae Bonaparte, 1845, and elevated and Smith [1993], who recognized the tribes Dowling's (1975) tribes to familial rank. proposed by Dowling and Duellman [1978] Herein, I present evidence that supports and Jenner [1981], respectively). the monophyly of the Xenodontinae sensu McDowell (1987: 40) also proposed adef- stricto thatincludes Heterodon andXenodon, inition for the "xenodontines:" "with sulcus among others. However, for the sake of sta- spermaticus forked (except in some genera bility, only the name Xenodontinae will be with a capitate hemipenis) and centrolineal, used, even though Heterodontinae has pri- or centripetal near the crotch but becoming ority (Smith, 1964). centrolineal distally, or (the majority) cen- The tribes proposed by Dowling (1975), trifugal (i.e., its branches taking the position Dowling and Duellman (1978), Jenner farthest from the midline ofthe organ); most (1981), Dowling et al. (1983), and Jenner genera with well defined calyces; some (e.g., and Dowling (1985) are not recognized in Nothopsis, Amastridium) with natricine pos- the present study, except for the Xenodontini terior hypapophyses but most with posterior (see Dowling, 1975; Dowling and Duellman, hypapophyses reduced to keels." This defi- 1978; Dixon, 1980; Myers, 1986), and the nition, however, neither corresponds to the Pseudoboini sensu stricto (see Dessauer et two subfamilies proposed by Dowling et al. al., 1987; Zaher and Caramaschi, 1992; My- (1983), nor does it follow Cadle's (1984a, ers and Cadle, 1994; Zaher,1994a), which are 1984b, 1984c) suggestions. satisfactorily characterized. It emerges from these comments that two I do not intend to propose a comprehen- fundamental problems have contributed to sive phylogenetic hypothesis for the Xeno- the destabilization of xenodontine systemat- dontinae. Rather, I aim to reevaluate the 1999 ZAHER: HEMIPENIAL MORPHOLOGY OF XENODONTINE SNAKES 7 hemipenial variation present in the colu- and Kizirian (1995), Myers (1974, 1982), My- broids and define, based on two uniquely de- ers and Cadle (1994), Myers and Campbell rived hemipenial features, a monophyletic (1981), Myers and Donnelly (1996), Pinou and Dowling (1994), Porto and Fernandes (1996), clade-the Xenodontinae sensu stricto. Hem- Villa (1970, 1971), Walker (1945). ipenial evidence is also evaluated that sup- Other "colubrids": Amaral (1929a, 1929b, ports monophyletic subgroups as well as re- 1929c, 1929d), Blaney (1977), Bogert (1939, jects previously recognized taxa within the 1940, 1947), Branch (1976), Branch and Wade Xenodontinae. (1976), Broadley (1971), Brongersma and Wehlburg (1933), Cadle (1996a, 1996b), Clark SOURCES OF INFORMATION ON HEMIPENIAL Jr. (1964), Cliburn (1975), Cole and Hardy MORPHOLOGY OF COLUBROID SNAKES (1981), Darevsky and Orlov (1992), Dixon et al. (1993), Domergue (1955, 1962, 1963, 1972, Hemipenial materials available for each 1983, 1986, 1987, 1991, 1994), Dowling (1957, species placed in the Xenodontinae and in 1958, 1959, 1960, 1969a, 1975, 1990), Dowl- incertae sedis are listed under the genus to ing and Fries (1987), Dowling and Maxson (1990), Dowling and Price (1988), Lambiris which the species belongs, in the section ti- (1997), Lanza (1964), Lopez et al. (1993), Mao tled "Generic Synopsis of Hemipenial Fea- (1965a, 1965b), Malnate (1953), McDowell tures." (1972, 1984), Oliver (1948), Ortenburger Published descriptions and illustrations (1923), Ota and Ross (1994), Rasmussen represented an important source of informa- (1985, 1986, 1989, 1993a, 1993b, 1997), Ross- tion for determining the hemipenial pattern man and Blaney (1968), Rossman and Eberle of the remaining colubroid taxa. The two (1977), Schatti (1987, 1988), Schatti andLanza synapomorphies of the Xenodontinae pro- (1989), Schatti and McCarthy (1987), Schatti posed herein were absent (or inapplicable) in and Vanni (1986), Stull (1940), Underwood all available descriptions and figures of and Kochva (1993), Van Devender and Cole (1977), Vellard (1928, 1946), Wilson (1967, nonxenodontine colubroids. The following 1970), Ziegler et al. (1997). works were used as sources of information for such comparisons: METHOD OF General: Cope (1895), Cei (1993), Clark (1944), HEMIPENIAL PREPARATION Dowling (1975), Dowling and Duellman (1978), Dowling and Savage (1960), Dunn The present study is basedon observations (1928), McDowell (1975, 1979, 1987), Sabnis of variation in everted hemipenes of repre- and Indurkar (1977), Schatti and McCarthy sentatives of the majority of the New World (1987), Smith (1943), Underwood (1967, "xenodontine" genera and of various supra- 1979), Vellard (1928, 1946) generic colubroid groups. Almost all hemi- Elapidae: De Silva(1987), Mao (1993), Mao and penes were prepared from previously fixed Chen (1974, 1980), Mao et al. (1984), Marx and preserved specimens, using the method (1953), McDowell (1967, 1969, 1970), Ras- described by Pesantes (1994). Either the left mussen (1992), Vellard (1928, 1946), Williams ortheright organ was removed from any one and Parker (1964). specimen and submerged in a solution of3% Viperidae: Branch and Wade (1976), Campbell (1985), De Silva (1983, 1988), Domergue KOH either for 1 to 6 hours or overnight (1955, 1962), Gasc (1968, 1969), Hoge (1947, (depending on the size of the organ). When 1959), Hoge et al. (1959), Koba and Kikukawa the tissues became translucent and flexible, a (1971), Mao (1993), McCranie (1988), Murphy small incision was made at the base of the and Barker (1980), Mao et al. (1984), Malnate organ, on the asulcate side, in order to evert (1990), Pesantes (1994), Vellard (1928, 1946). manually the whole structure. This was done Dipsadines and "xenodontines" incertae sedis: by using a pair of forceps with rounded tips Bailey (1939b), Bogert (1964), Cadle (1989), which served to force the inverted organ to- Clark Jr. (1970), Crother (1989a), Di-Bernardo ward its base. The everted organ was then (1992, 1996), Duellman (1958a, 1958b), Dunn filled with- a liquid solution of agar-agar and Dowling (1957), Fernandes (1995), Grant (1943), Hoge (1952), Jenner (1981), Kofron (Manzani andAbe, 1988) orwith coloredpe- (1982, 1985a, 1985b), Lema and Deiques troleumjelly (Myers and Cadle, 1994). Ifthe (1995), McCranie and Villa (1993), Mendelson organ, particularly the lobes that are made of 8 BULLETIN AMERICAN MUSEUM OF NATURAL HISTORY NO. 240 very delicate tissue, was torn during ever- toire d'Anatomie Comparee, Museum na- sion, the hole was patched with cyanocrylate tional d'Histoire naturelle de Paris (AC); glue before the injection of agar-agar or pe- Robert Henderson, Milwaukee Public Mu- troleum jelly (excess alcohol must be re- seum (MPMH); Colin McCarthy, Natural moved before using the glue). This method History Museum, London (BMNH); Anibal of preparation worked well, even with the Melgarejo, Instituto Vital Brazil, Niteroi hemipenial material extracted from speci- (IVB); Charles W. Myers, Darrel Frost, mens that were fixed and stored at the end Charles J. Cole, and Linda Ford, American of the 19th century. Museum of Natural History, New York (AMNH); Ronald Nussbaum and Greg ACKNOWLEDGMENTS Schneider, Museum ofZoology, The Univer- sity of Michigan (UMMZ); Juan Carlos Or- I am most grateful to Darrel Frost and tiz, Universidad de Concepcion; Ivan Sazi- Charles W. Myers for their support and en- ma, Departamento de Zoologia, Universida- couragement during my postdoctoral fellow- de de Campinas (ZUEC); Gustavo Scrocchi, ship at the American Museum of Natural Fundacion Miguel Lillo, Tucumain (FML); History. I am indebted to Charles Myers, Richard Thomas, Museum of Biology, Uni- who shared with me his profound knowledge versity of Puerto Rico (RT); John E. Cadle ofthe New World snake fauna. Animportant and Jose Rosado, Museum of Comparative part ofthe work presented herein was carried Zoology, University of Harvard, Cambridge out during the elaboration of my Ph.D. dis- (MCZ); Douglas Rossman and Frank Bur- sertation in the Museum national d'Histoire brink, MuseumofNatural Science, Louisiana naturelle de Paris and under the guidance of State University, Baton Rouge (LSUMZ); Alain Dubois, to whom I owe many debts of Harold Voris and Alan Resetar, Field Muse- gratitude. um of Natural History, Chicago (FMNH, I am grateful to the following curators, in- CNHM); Hebert Ferrarezzi (private collec- dividuals, and institutions who made avail- tion). The abbreviations noted above areused able specimens from the collections under throughout the text. their care for hemipenial preparations and The manuscript benefitted from comments other dissections: Alain Dubois, Annemarie and criticisms of Darrel Frost, Charles W. Ohler, Ivan Ineich, and Roger Bour, Labor- Myers, Erika H. Zaher, Helio R. da Silva, atoire des Reptiles et Amphibiens, Museum Douglas Rossman, Samuel B. McDowell, national d'Histoire naturelle de Paris and Brian I. Crother. IthankparticularlyDar- (MNHN); Antonio J. S. Argolo, Centro de rel Frost, Charles W. Myers, and Douglas Pesquisas do Cacau, Ilheus (CEPLAC); Olin- Rossman for their insightful comments on do Bortesi and Jose M. Cei, Museo Region- the whole manuscript. ale di Scienze Naturali, Torino (MZUT, MRSN); Ulisses Caramaschi and Marcelo HEMIPENIAL TERMINOLOGY Soares, Museu Nacional do Rio de Janeiro (MNRJ); Luis Coloma, Museo de Zoologia, Although hemipenial morphology has Pontificia Universidad Catolica del Ecuador, been used extensively for taxonomic and Quito (QCAZ); Kevin De Queiroz, Roy W. phylogenetic purposes, only one work McDiarmid, and George Zug, United States (Dowling and Savage, 1960) has comprehen- National Museum, Smithsonian Institution, sively treated snake hemipenial terminology. Washington (USNM); James R. Dixon, Texas However, various other papers have contrib- A&M University, College Station (TCWC); uted to the clarification of available terms or William E. Duellman, Linda Trueb, John named new structures (Myers and Trueb, Simmons, Joseph Mendelson, and Helio R. 1967; Myers, 1974, 1986; Rossman and da Silva, Museum of Natural History, Uni- Eberle, 1977; Myers and Campbell, 1981; versity ofKansas, Lawrence (KU); Giuseppe Branch, 1986; Donnelly and Myers, 1991; Puorto, lara L. Ferreira, Hebert Ferrarezzi, Myers and Cadle, 1994). I Herein use most and Francisco L. Franco, Instituto Butantan, of the previously proposed and largely fol- Sao Paulo (IB); Jean Pierre Gasc, Labora- lowed terms (Dowling and Savage, 1960;