f^lO o HARVARD UNIVERSITY Library of the Museum of Comparative Zoology (US ISSN 0027-4100) auUetin of the Museum of Comparative Zoology ^« / 2 1995 Types of Shelled Indo-PaGffj^'iVibtlusks Described by W. H. Pease RICHARD i. JOHNSON HARVARD UNIVERSITY CAMBRIDGE, MASSACHUSETTS, U.S.A. VOLUME 154, NUMBER 1 /DECEMBER 1994 (US ISSN 0027-4100) PUBLICATIONS ISSUED OR DISTRIBUTED BY THE MUSEUM OF COMPARATIVE ZOOLOGY HARVARD UNIVERSITY Breviora 1952- BULLETIN 1863- Memoirs 1865-1938 JOHNSONIA, Department of Mollusks, 1941-1974 Occasional Papers on Mollusks, 1945- SPECIAL PUBLICATIONS. 1. Whittington, H. B., and W. D. I. Rolfe (eds.), 1963. Phylogeny and Evolution of Crustacea. 192 pp. 2. Turner, R. D., 1966. A Survey and Illustrated Catalogue of the Tere- dinidae (Mollusca: Bivalvia). 265 pp. 3. Sprinkle, J., 1973. Morphology and Evolution of Blastozoan Echino- derms. 284 pp. 4. Eaton, R. J., 1974. A Flora of Concord from Thoreau's Time to the Present Day. 236 pp. 5. Rhodin, A. G. J., and K. Miyata (eds.), 1983. Advances in Herpetology and Evolutionary Biology: Essays in Honor of Ernest E. Williams. 725 pp. 6. Angelo, R., 1990. Concord Area Trees and Shrubs. 118 pp. Other Publications. Bigelow, H. B., and W. C. Schroeder, 1953. Fishes of the Gulf of Maine. "Reprinted 1964. Brues, C. T, A. L. Melander, and F. M. Carpenter, 1954. Classification of Insects. {Bulletin of the M.C.Z., Vol. 108.) Reprinted 1971. Creighton, W. S., 1950. The Ants of North America. Reprinted 1966. Lyman, C. P., and A. R. Dawe (eds.), 1960. Proceedings of the First Inter- national Symposium on Natural Mammalian Hibernation. {Bulletin of the M.C.Z.,Wol 124.) Ornithological Gazetteers of the Neotropics (1975-). Peters' Check-list of Birds of the World, vols. 1-16. Proceedings of the New England Zoological Club 1899-1947. (Complete sets only.) Proceedings of the Boston Society of Natural History. Price list and catalog of MCZ publications may be obtained from Publications Office, Museum of Comparative Zoology, Har\'ard Universitv, Cambridge, Massachusetts 02138, U.S.A. This piil)licati<)n lias been printed on acid-free pernianent paper stock. © The President and Fellows of Harvard College 1 994. I MCZ LIBRARY JAN 1 '^ 1995 HARVARD UNIVERSITY TYPES OF SHELLED INDO-PACIFIC MOLLUSKS DESCRIBED BY WILLIAM HARPER PEASE (1824-71) RICHARD I. JOHNSON^ TABLE OF CONTENTS Abstract 1 Introduction 1 Remarks 4 Acknowledgments 5 Abbreviations 5 List of the Taxa of Shelled Mollusks Described b\ Pease 5 Additional or Corrected Locality Data 29 Literature Cited 29 Index 34 Plates 42 Abstract. Pease described some 500 species of mollusks from the Indo-Pacific region. Of these, some 380 were shell-bearing. Type material was located for all but 42, consisting of syntypes, some previously figured or measured, which have hitherto been re- garded as holotypes or selected as lectotypes. Many lectotypes are chosen here and illustrated for the first time. All of the holotypes and lectotypes in the Mu- seum of Comparative Zoology are figured. INTRODUCTION The original William Harper Pease col- lection is in the Museum of Comparative Zoology, Harvard University. After Pease died in 1871, his collection was sent to Boston to be sold "entire." It was offered to a Mrs. Witthaus of New York for $3,000. In a letter to John Gould Anthony, Curator of Mollusks at the Museum of Compara- tive Zoology, she said that she had seen the collection and found it to be damaged by careless packing. She arranged to pur- chase only the specimens she wanted, the larger pretty shells (R. D. Turner, personal ' Museum of Comparative Zoology, Harvard Uni- versity, Cambridge, Massachusetts 02L38. communication). The remaining collec- tion was sold to Louis Agassiz, director of the Museum of Comparative Zoology. Most of the types and smaller specimens were carefully glued on pieces of glass or slate for exhibition by curator Anthony. This process, as well as the formula for the glue, is included in Turner's (1946) definitive biography of Anthony. In this century, the shells were removed from the plaques by William J. Clench, cataloged, and num- bered. Each lot is usually accompanied by labels in Anthony's distinctive and accu- rate calligraphy as well as one penciled on a scrap of paper by Pease. There is a type- written list of the collection in the De- partment of Mollusks compiled by Wil- liam F. Clapp (1923, Dept. Library, no. 5912), "copied from Pease catalogue writ- ten on loose sheets of stationery." Aside from the type material, and that collected by Andrew Garrett, most of the lots have locality data that are so general as to ren- der the specimens useless for study. The Witthaus material became the property of Vassar College, Poughkeepsie, New York, and was presented to the Museum of Com- parative Zoology in 1944. Solem (1976: 9) took such a dim view of the Pease collection that he virtually ignored it when monographing the En- dodontoid Land Snails from Pacific Is- lands, preferring to select lectotypes from any other collection that had Pease ma- terial, such as the National Museum of Natural History, Washington, D.C., or the Bernice Pauahi Bishop Museum, Honolu- Bull. Mus. Comp. Zool., 154(1): 1-61, December, 1994 1 Bulletin Museum of Comparative Zoology, Vol. 154, No. 1 lu, even when the material had been orig- inally received from the Museum of Com- parative Zoology as duplicates. He claimed that when the collection was shipped "from England [sic] to Harvard University, ap- parently there was extensive mixing of sets when cabinets were tilted and handled. In subsequent years (obviously referring to duplicates sent out from the Museum of Comparative Zoology rather than by Pease), these shells have been traded wide- ly to other museums and amateur collec- tors. Virtually all traded Pease material that I examined contained more than one species, living on different islands" (ibid.). Solem further declared that "Pease him- self was careless in handling his collection since in a letter to a correspondent [An- drew Garrett] wrote that his small daugh- ter delighted in playing with shells in the cabinets" (Alison Kay, personal commu- nication). Kay (1975: 18) had actually quoted Pease as having written, "Our little girl is growing and troubles me in my col- lections, by assorting my duplicates and arranging them in boxes." Pease began describing shells in the Pro- ceedings of the Zoological Society of Lon- don, often remarking that examples were in the collection of Hugh Cuming, who had died in 1865. The Cuming collection was acquired by the British Museum (Nat- ural History). Kay (1965) located all of the available types of the marine species, re- described them, selected lectotypes when necessary, and figured them, most for the first time. Many of those species not found by Kay were found in the Pease collection and are figured here. Syntypes of all but 14 of the land and freshwater species de- scribed in the Proceedings were located in the Museum of Comparative Zoology. None of those missing could be located in the British Museum (Natural History) (P. B. Mordan, in litt.). Clench (1975) listed all of the Pease taxa but made no effort to supply any information as to the location of any of the type specimens, neither those in the British Museum (Natural History) nor those described in the Journal de Con- chyliologie, which were sent to H. Crosse, who translated the papers into French. Fi- scher-Piette (1950) listed all of the types that had been in the collection of the Jour- nal, and later presented them to the Mu- seum National d'Histoire Naturelle, Paris, figuring those not done so previously. Clench also did not notice that Baker (1963, 1964), who listed all of the land and fresh- water types in the Academy of Natural Sciences of Philadelphia, included those by Pease described in the American Jour- nal of Conchology . After the death of Cuming, Pease chose P. P. Carpenter to sponsor his publications in the Proceedings of the Zoological So- ciety, but Carpenter (in Pease, 1865a: 675- 676) included a list of synonyms based on specimens Pease sent to Cuming deter- mined by Cuming and H. Adams. Later (in Pease, 1865d: 516-517), Carpenter published a list of synonyms based on his own observations. This ended their rela- tionship. It was not until 1872 that another paper by Pease appeared in the Proceed- ings. In it were described 14 new species of Triphoris, none of which were sent to the British Museum (Natural History). Twelve of them were found in Pease's col- lection and are figured here for the first time. While the opisthobranch mollusks that are shell-less are not included in the following list, those that were studied by Pruvot-Fol (1947) have been noted in the accompanying index of all of the Pease taxa. William Harper Pease was born in Brooklyn, New York, in January 1824. He joined the Lyceum of Natural History of New York in 1841; traveled with General Winfield Scott to Mexico in 1848, his as- signment a matter of conjecture; arrived in Honolulu on December 11, 1849; and in 1850 applied for citizenship in the Ha- waiian Kingdom, purchased land, and be- came Local Agent of the Government on the island of Kauai. Little else is known about him to this time. Returning to Ho- nolulu about 1856, Pease became Com- missioner of Water Rights and Rights of MoLLUSKS Described by W. H. Pease • Johnson Way as well as Assessor of the City of Ho- nolulu, though Solem (1976-77) thought he was a merchant. His main interest was natural history, especially the shells of the Pacific. In a letter of 1865, quoted by Kay (1975: 6), he wrote, "That is all I think or care about." Prominent among his friends was Levi Haalelea, Chamberlain to the Court. He kept his shell collection, library, and an office in Haalelea's home opposite the Palace. When Haalelea died in 1865, Pease moved the library to a small building adjoining his own house at 25 Liliha Street. Most of the material that Pease de- scribed after 1860 was collected by An- drew Garrett (1823-87), whom Solem (1976: 7) mistakenly thought was a mis- sionary. Garrett was actually a self-trained naturalist, who had spent some of his life as a sailor before becoming a professional collector of shells, fishes, and other natural history objects of the South Pacific Islands. He arrived in Honolulu in 1852 and later moved to the island of Hawaii, where he remained, except for collecting trips to other Pacific Islands, until 1863 when he settled permanently in the Society Islands. While in the Hawaiian Islands he became a close friend and employee of Pease, and the two carried on an active correspon- dence until the death of the latter in 1871. Garrett collected for Louis Agassiz of the Museum of Comparative Zoology, for the Museum Godeffroy, Hamburg, and for Pease, who sponsored his trips in part and acted as his agent in Honolulu. He sent Pease descriptions of shells and nudi- branchs, which the latter used in his pub- lications. The relations of these two men was friendly, although Pease took almost a paternalistic attitude toward Garrett. There was a difficult side for Gar- rett. It is true that, in a sense, he was an employee of the other man, but in some ways Pease seems to have taken advantage of Garrett's abilities and ef- forts. The latter worked hard and assiduously. He collected amid the most trying and dangerous con- ditions. He scrupulously described and drew pic- tures of many of the specimens gathered, partic- ularly the fishes. In the case of fishes and shells he suggested the scientific names. But despite all this, he lived in the shadow of Pease, while the more articulate writer and more assertive man earned the glory. (Thomas, 1979: 20) After the death of Pease, beginning in 1872 until his death in 1887, Garrett began publishing his own descriptions of new species and monographs of various groups of mollusks, which included much more accurate locality data than those given by Pease. Few details of the life of Pease are re- peated here. His biographer, Karl W. Greene (1960), based his account on the letters Pease wrote to Garrett, which, along with the latter's shell collection, are now in the Bernice Pauahi Bishop Museum, Honolulu. While details of the purchase of the Garrett collection are lacking, it was in the museum by 1899. Kay (1975) also wrote a biography of Pease. Although Greene (1960 (8): 12) mentioned Kay's ar- rival in Honolulu, and while she must have been aware of his articles in Hawaiian Shell News, Kay did not allude to Greene's work. The life and travels of Andrew Garrett are treated definitively by W. Stephen Thomas (1979), who devoted many years to the study. Thomas' interest in Garrett was inspired by his obtaining from some of his cousins over 250 drawings of fishes and shells that Garrett had made for Cap- tain John W. Leonard of the whaleship Lydia. Clench (1979) listed all of the mol- lusks described by Garrett but again did not make any effort to locate his type spec- imens, many of which are in the Academy of Natural Sciences of Philadelphia, the Museum of Comparative Zoology, and, of course, the Bishop Museum. Kay (1975: 18) mentioned that neither portrait nor photograph of Pease had ever been found, but that Lady Franklin, the widow of Sir John Franklin, the Arctic ex- plorer, met Pease at the plantation of Rob- ert G. Wyllie, a Scotsman who was Ka- mehameha IV's minister of foreign affairs. Lady Franklin wrote, I find it so difficult to make out what he says that much is lost to me — this proceeds partly perhaps from a want of some teeth in front of his mouth, Bulletin Museum of Comparative Zoology, Vol. 154, No. 1 but chiefly from his holding and chewing tobacco which not only thickens his speech, but causes him to be constantly spitting. (Korn, 1958) Perhaps, it is just as well that no portrait was found. In spite of his appearance, Pease seems to have moved in the best circle of Ha- waiian society of the time. As already men- tioned, he was an intimate of Levi Haa- lelea, an acquaintance of Robert G. Wyllie, Minister of Foreign Affairs, and James Walker Austin, sometime Associate Justice of the Supreme Court of the Kingdom of Hawaii. Austin, a relative of my children, was executor of all of the preceding estates including that of Pease (Austin, 1921: 64). REMARKS For convenience the following list of taxa is arranged alphabetically by species name. It contains references only to those that are shelled: a complete list of molluscan names introduced by Pease was compiled by Clench (1975) and may be found here in the Index alphabetically by genus with supplemental data. In addition to the orig- inal reference, subsequent ones are also cited if they include the first figure of a type or other relevant information. More detailed locality information is included from subsequent references by both Pease and Garrett, the latter having collected most of the material. Pease put a question mark before some genera when he was not sure where the species belonged. A generic name in pa- rentheses is a second citation indicating that he later changed the generic place- ment. Data in brackets have been found on original labels, are additions or correc- tions from recent maps, or are comments by this author. No attempt has been made to discuss the present status of any of the taxa. This is a task for individual revisers. In the past, if Kay (1965), Fischer-Piette (1950), Baker (1963, 1964), Houbrick (1992), or the present author was able to locate the single figured or measured syn- type, it was usually regarded as the holo- type. Most of these designations, included in this paper, were made long before the promulgation of the third edition of the International Code of Zoological Nomen- clature (1985), which invalidates this prac- tice under Article 74 (b): ". . . should an- other syntype or syntypes be discovered [the first subsequent author is to be re- garded] to have chosen a lectotype." If this change is to be made it is left here, in these instances, to the first subsequent reviser. Baker (1963, 1964) listed all of the types of land snails in the Academy of Natural Sciences of Philadelphia and used his own system of abbreviations for type designa- tions, which have been interpreted here as follows: TOD type by original designation [ho- lotype] TOM type because only one example was included in the original de- scription, or was indicated by only one set of dimensions . . . [holo- type] TSD type by subsequent selection, fol- lowed by "now" if apparently first designated in these lists . . . [lec- totype] At the request of Dr. Gary Rosenberg, of the Academy of Natural Sciences of Philadelphia, all of the types, either fig- ured for the first time or refigured, are regarded as lectotypes in compliance of ICZN 74 (b). Unless previously selected as lectotypes by someone else, the types of the species of genus Parttila from the Society Islands have been included only as syntypes, be- cause so many were regarded as mere color forms by Crampton (1916, 1932) and be- cause most of these are now extinct (Tudge, 1992). Pease introduced a number of names in the literature that are believed to be nude. When these were represented by speci- mens, the number of the lot is included. These specimens are of no significance un- less, peradventure, the taxa were subse- quently validated by another author and overlooked by the present one. Several of MoLLUSKS Described by W. H. Pease • Johnson the land shells from the Society Islands are known only from the specimens figured by Garrett (1884) and are regarded as lec- totypes. Examples of some of the described taxa that are missing may occur in the Garrett collection in the Bernice Pauahi Bishop Museum under the names of Pease and, while not exactly strictly speaking types, might serve as neotypes. ACKNOWLEDGMENTS It gives me great pleasure to thank those people whose cooperation eased the task of gathering information. Peter Mordan, of the Mollusca Section, British Museum (Natural History), was able to inform me that certain of the 14 non-marine mollusks not in the Pease collection were repre- sented in that collection and Ms. Kathie Way supplied other data. Drs. George M. Davis and Gary Rosenberg of the Acade- my of Natural Sciences of Philadelphia an- swered every request for material and data, and there were many; Ms. Doree Bardes also helped. Dr. Alison Kay left a number of helpful notes in many lots of the marine types in the Museum of Comparative Zo- ology in conjunction with her work (1975). Mr. Tan Koh Siang identified the types of Sistrum triangulatum, which otherwise would have been overlooked. With the ex- ception of two photographs by Dr. James H. McLean, Los Angeles County Museum of Natural History, and two others by Dr. Kenneth J. Boss, all of the photographs and measurements were made by Dr. Silvard P. Kool. Dr. Kenneth J. Boss made many helpful suggestions. Finally, thanks are ex- tended to Mrs. Marion D. Britz, who cop- ied and recopied as many corrected copies of this manuscript as did Sonia Tolstoy of Anna Karenina. Drs. Alan R. Kabat and Gary Rosenberg kindly reviewed the manuscript and made a number of frank criticisms that saved this author from some really egregious errors. I would also like to thank Drs. C. C. Wang, William U. Shi- pley, Merrill Goldstein, Mandel E. Cohen, and Blair Ardman. Publication costs of this study were covered in part by the Wet- more Colles Fund. ABBREVIATIONS In general, this author disdains refer- ences that merely give a date, forcing the use of the bibliography over and over. To partially alleviate this problem, the prin- cipal references are cited in the catalog by the following abbreviations. ACM Annals of the Carnegie Museum AJC American Journal of Conchology ANSP Academy of Natural Sciences of Philadelphia, Philadelphia, Penn- sylvania BMCZ Bulletin of the Museum of Com- parative Zoology, Cambridge, Massachusetts BMNH British Museum (Natural History) BPBM Bernice Pauahi Bishop Museum, Honolulu, Hawaii CM Carnegie Museum, Pittsburgh, Pennsylvania JANSP Journal of the Academy of Nat- ural Sciences of Philadelphia JdeC Journal de Conchyliologie MCZ Museum of Comparative Zoology, Cambridge, Massachusetts MNHN Museum National d'Histoire Na- turelle, Paris, France MofC Manual of Conchology PZS Proceedings of the Zoological So- ciety of London USNM National Museum of Natural His- tory, Washington, D.C. Titles of other journals cited in the catalog are usually given in full. LIST OF THE TAXA OF SHELLED MOLLUSKS DESCRIBED BY PEASE abbreviata, Partula 1865a, PZS for 1864: 675 [>2omen nudum]. MCZ 2505. abbreviata, Planaxis Plate 9, Figure 2 1865d, PZS for 1865: 515 (Islands of the Central Pacific); 1868g, AJC 4: 101, pi. 12, fig. 16 (Tahiti). Lectotype, here selected, ANSP 18261; paralecto- types MCZ 187833 (Plate 8, Figure 13) and 187834; not located in BMNH by Kay (1965: 86). Bulletin Museum of Comparative Zoology, Vol. 154, No. 1 abbreviata, Realia 1865a, PZS for 1864: 674 (Islands of the Central Pacific); 1869a, JdeC 17: 155, pi. 7, fig. 5, as Om- phalotropis (Tahiti). Holotype in MNHN, teste Fi- scher-Piette (1950: 72); paratypes MCZ 74936 and 187842. acetabulum, Helix 1861d, PZS for 1861: 242 (Tahiti). Holotype ANSP 47844a, teste Baker (1963: 231) selected as lectotype by Solem (1976: 357, figs. 156d-f); paralectotypes MCZ 17248 and 176567. aciculata, Eulima 1861b, PZS 28: 438 (Sandwich Islands). Lectotype BMNH 1962839 selected by Kay (1965: 66, pl. 9, fig. 2); paralectotype MCZ 31705 figured by Pilsbry (1917: 222, figs. 11a, b); paralectotype MCZ 187747. aculeata, Scutellina Plate 1, Figure 9 1868g, AJC 4: 100 (Hawaii [Island]). Lectotype, here selected, MCZ 83720 labeled as holotype. affine, Sistrum 1863b, PZS for 1862: 244 (Kingsmill Islands); 1868d, AJC 3: 277, pl. 23, fig. 13. Lectotype BMNH 1964304 selected by Kay (1965: 79, pl. 13, figs. 13, 14). affinis, Partula 1865a, PZS for 1864: 675 [nometi nudum]; 1868b, AJC 3: 224 (Tahiti). Holotype ANSP 59560a, teste Baker (1963: 204); paratypes MCZ 25313. affinis, Realia 1865e, AJC 1: 288 (Polynesia); Tryon, 1866a, AJC 2: 82, pl. 5, fig. 4 (Hervey Isles); Pease, 1869a, JdeC 17: 152 as Omphalotropis (Aitutake, Hervey Isles). Holotype ANSP 13346a, teste Baker (1964: 178); paratypes MCZ 187840 and 187841. affinis, Triphoris 1861b, PZS 28: 434 ([Kauai] Sandwich Islands). Ho- lotype BMNH 1962808 figured by Kay (1965: 51, pi. 10, fig. 1); paratypes MCZ 50074. alba, Thala Plate 9, Figure 20 1868a, AJC 3: 215, pl. 15, fig. 8 (Paumotus). Lecto- type ANSP 28755 selected by Cernohorsky (1976: 501, pl. 450, fig. 6); three paralectotypes ANSP 325385; paralectotypes MCZ 260612. albinea, Helicina maugeriane 1871d, PZS for 1871: 466 (Raiatea [restricted to a single valley on the east side of Tahaa, teste Garrett, 1884: 101, pl. 3, fig. 64]). Lectotype, here selected, ANSP 14538 is the specimen figured by Garrett; five paralectotypes MCZ 202599 from Garrett; not located in BMNH, teste Mordan, personal com- munication. albocincta, Engina 1860b, PZS 28: 142 (Sandwich Islands). Lectotype BMNH 1961454 selected by Kay (1965: 16, pl. 2, figs. 9, 10); paralectotypes MCZ 297948. alta. Helix 1868), AJC 4: 153, pl. 12, fig. 1 (Ponape). Holotype ANSP 49016, teste Baker (1963: 233); paratypes MCZ 94770. alternata 'Pease' H. H. Smith, Partula 1865a, PZS for 1864: 675 [nomen nudum]; 1902, ACM 1: 447 (Moorea). Syntypes MCZ 25069. alternata, Triphoris 1861b, PZS 28: 434 ([Kauai] Sandwich Islands); 1868h, AJC 4: 127 non Adams 1852. Changed to: Triphoris btcolor. Lectotype BMNH 1962815 selected by Kay (1965: 54, pl. 10, fig. 4); paralectotypes MCZ 50057 and 73735. ambigua, Limnaea 1870b, AJC 6: 6, pl. 3, fig. 5 (Hawaiian Islands). Holotype ANSP 21240a, teste Baker (1964: 152); paratypes MCZ 73469. anibusta, Auriculella Plate 2, Figure 10 1868f, JdeC 16: 345 ([Waianae Mountains, Oahu]). Lectotype, here selected, MCZ 45152; paralecto- types MCZ 298488. analogica, Pithys 1870c, JdeC 18: 396 (Marquesas [Islands]); 1871b, PZS for 1871: 454. Lectotype BPBM 115307 from MCZ 17260 selected by Solem (1976: 328, figs. 143a- d); paralectotypes MCZ 17260. angasii, Rissoina 1871e, AJC 7: 20. New name for turricula Angas 1867 non Pease 1860. angicostata, Hindsia 1860b, PZS 28: 142 (Sandwich Islands). Lectotype BMNH 1961159 selected by Kay (1965: 16, pl. 1, figs. 15, 16). angiostoma, Cythara 1868h, AJC 4: 105. New name for Pleurotoma triticea Reeve 1843 non Kiener 1839. angiostoma, Thala Plate 9, Figure 22 1868a, AJC 3: 216, pl. 15, fig. 9 (Paumotus). Lecto- type, here selected, ANSP 28754; paralectotypes MCZ 256500. angulata, Carelia adusta \870c, JdeC 18: 403 (Kauai). Lectotype ANSP 23434 selected by Baker (1963: 197) figured by Hyatt and Pilsbry (1911: 116, pl. 20, fig. 16); paralectotypes MCZ 45167. angulatus, Euchelus Plate 7, Figure 13 1868d, AJC 3: 283, pl. 23, fig. 27 (Annaa Island). Lectotype, here selected, ANSP 40671 is the figured type; two paralectotypes ANSP 391031; paralecto- types MCZ 89796. annectens, Bulimus [Partula] 1865a, PZS for 1864: 671 (Islands of the Central Pacific [two valleys on the west coast of Huaheine, teste Garrett, 1884: 66]). Syntypes MCZ 26352. antiqua, Leptachatma 1870a, JdeC 18: 94 (Kauai); Crosse, 1876, JdeC 24: 97, pl. 3, fig. 6. Holotype in MNHN, teste Fischer- Piette (1950: 149). aperta, Haminea Plate 4, Figure 17 I868e, AJC 4: 72 (Tahiti). Lectotype, here selected, ANSP 57575. aperta, Tornatellina Plate 2, Figure 18 1865b, PZS for 1864: 673 (Islands of the Central Pacific); 1871d, PZS for 1871: 473 (Tahiti). Lecto- type, here selected, MCZ 175722; paralectotypes MCZ 298457. approximala, Nassa Plate 8, Figure 4 1868d, AJC 3: 272, pl. 23, fig. 3 ([Ascension Island]). MoLLUSKS Described by W. H. Pease • Johnson Lectotype MCZ 151800 selected by Cernohorsky (1984: 128, pi. 23, fig. 13); paralectotype MCZ 303191. approximata 'Pease' Garrett, Partula 1884, JANSP (2) 9: 75 (several small valleys on the south west part of Raiatea). Lectotype ANSP 59451a selected by Baker (1964; 204) figured by Pilsbry (1909: 243, pi. 17, fig. 15); paralectotypes MCZ 50851, 24927, and 25197. approximata, Turricula 1860b, PZS 28: 146 (Sandwich Islands). Lectotype BMNH 1961192 selected by Kay (1965: 27, pi. 3, figs. 1, 2). argutus, Bulimus [Partula] 'l865b, PZS for 1864: 670 (Islands of the Central Pacific [Huaheine, teste Garrett, 1884, JANSP (2) 9: 62, pi. 3, fig. 57]). Syntypes MCZ 26353. armata, Vertigo 1871d, PZS for 1871: 461 (Bolabola [Borabora Island, Society Islands]). Holotype MCZ 48315 figured by Pilsbry and Cooke (1920: 327, pi. 30, figs. 12, 13); paratypes MCZ 31398. arwatus, "7" Acanthochites Plate 1, Figure 1 1872a, AJC 7: 195 ([Pauloa] Oahu). Holotype MCZ 73452, only specimen. asperum, Cerithium 1861b, PZS 28: 433 (Sandwich Islands). Holotype BMNH 1961203 figured by Kay (1965: 47, pi. 5, figs. 11, 12); paratype MCZ 239650. assimilis, Mitra Plate 7, Figure 20 1868a, AJC 3: 211, pi. 15, fig. 1 ([Oahu] Polynesia). Lectotype ANSP 28718 selected by Cernohorsky (1976: 487, pi. 436, fig. 1) and the type locality re- stricted to: Huahine [sic] Island, Society Islands; paralectotypes MCZ 260601 and 260603; paralec- totype ANSP 391048. assimilis, Partula 1868b, AJC 3: 230, pi. 15, figs. 28, 29 [both as 28] (Roratonga); 1870c, JdeC 18: 401. Syntypes MCZ 24906 and 25137; not located in ANSP by Baker (1963: 204). assimilis, Terebra 1869d, AJC 5: 67 (Oahu); 1871e, AJC 7: 20 is Terebra contigua Pease 1871, new name for T. assimilis non Angus 1867. Not located in ANSP. aliensis, Pithys Plate 2, Figure 8 1870c, JdeC 18: 394 (Atiu [Island, Cook Islands]); 1871d, PZS for 1871: 453. Lectotype, here selected, MCZ 17335; paralectotypes MCZ 17336. atra, Planaxis Plate 9, Figure 1 1869d, AJC 5: 72, pi. 6, fig. 4 (Marquesas [Islands]). Lectotype, here selected, ANSP 18282; three para- lectotypes ANSP 391038; paralectotypes MCZ 187836 and 187837. attenuala, Cirsotrema. 1861a, PZS 28: 400 (Sandwich Islands). Lectotype BMNH 1962796 selected by Kay (1965: 43, pi. 10, figs. 9, 10). attenuata, Partula 1865a, PZS for 1864: 672 (Islands of the Central Pacific [Tahiti]). Syntypes MCZ 24901 and 25047. aurantmm, Operculatum 1868d, AJC 3: 287 (Hawaii [Island]). Not located in ANSP. australis, Tritonidea 1871e, AJC 7: 21 (Australia). New name for Triton- idea assimilis Angas 1867 non Buccitium assimile Reeve 1846. bacca. Vertigo 187 Id, PZS for 1871: 462 (Kalapana [Puna District] Hawaii Island); specimens lost, teste Pease. baetica 'Pease' Paetel, Lampania 1887, Catalog der Conchylten-Sammlung (Berlin) 1: 351 [nomen tiudum]. balteata, Clathurella 1860b, PZS 28: 143 (Sandwich Islands). Lectotype BMNH 1962786 selected by Kay (1965: 34, pi. 5, figs. 3, 4); paralectotypes MCZ 50004. balteata, Leptachatina 1870a, JdeC 18: 91 ([Waimea] Kauai); Crosse, 1876, JdeC 24: 97, pi. 4, fig. 4. Holotype in MNHN, teste Fischer-Piette (1950: 149); paratypes MCZ 142986. balteata, Nassa 1869d, AJC 5: 71, pi. 8, fig. 4 (Ebon Island). Lecto- type, here selected, ANSP 16366 is the figured type. balteata, Rissoina 1869d, AJC 5: 72 (Hawaii [Island]). Not located in ANSP. bell a, Daphnella 1860b, PZS 28: 147 (Sandwich Islands). Lectotype BMNH 1962786 selected by Kay (1965: 34, pi. 5, figs. 3, 4); paralectotypes MCZ 50013 and 50014. bella, Helicitia 1865a, PZS for 1864: 676 [numen nudum]. MCZ 297930. bella 'Pease' Hartman, Partula 1871a, PZS for 1871: 473 [nomen nudum]; 1881, BMCZ 9: 193 ([Raiatea]). Syntypes MCZ 3639. bella, Turricula 1860b, PZS 28: 145 (Sandwich Islands). Lectotype BMNH 1961204 selected by Kay (1965: 26, pi. 3, figs. 12, 14); paralectotypes MCZ 50017. biangulatum, Cyclostoma Plate 2, Figure 13 1865a, PZS for 1864: 674 (Islands of the Central Pacific [Aitutaki, Cook Islands, teste Garrett (1881: 404) who collected the specimens]). Lectotype, here selected, MCZ 141031; paralectotypes MCZ 141032. bicarinata, Amathina 1861a, PZS 28: 399 (Sandwich Islands). Lectotype BMNH 1962792 selected by Kay (1965: 40, pi. 6, figs. 3, 4). bicarinata, Clathurella Plate 6, Figure 17 1863c, PZS for 1862: 243 (KingsmiU Islands); 1868a, AJC 3: 222, pi. 15, fig. 23. Lectotype, here selected, ANSP 15806; not located in BMNH by Kay (1965: 86). bicolor, Partula 1871f, AJC 7: 26, pi. 9, fig. 4 (Guam). Syntype MCZ 25345; not located in ANSP by Baker (1963: 204). bicolor, Triphoris 1868h, AJC 4: 127. New name for Triphoris alternata 8 Bulletin Museum of Comparative Zoology, Vol. 154, No. 1 Pease 1860 non C. B. Adams 1852. See under: alter- tiata, Triphoris. bifasciata, Borsonia 1860b, PZS 28: 143 (Sandwich Islands). Holotype BMNH 1962758, teste Kay (1965; 18, pi. 8, figs. 4, 5). bilineata, Partula 1865b, PZS for 1864: 675 [tiomen nudum]; 1866a, AJC 2: 201 (Tahaa [confined to Faa-apa Valley on the east coast, teste Garrett, 1884: 62]); 1867a, AJC 3: 81, pi. 1, fig. 10; 1871b, PZS for 1871: 473 (Tahiti). Lec- totype ANSP 59438a selected by Baker (1963: 204) is the figured type; paralectotypes MCZ 24903. hipes, Syphonota 1860a, PZS 28: 23 (Sandwich Islands). Syntype MCZ 297867; not located in BMNH by Kay (1965: 84). boeticum, Centhmm 1861b, PZS 28: 433 (Sandwich Islands). Lectotype BMNH 1962802 selected by Kay (1965: 48, pl. 10, fig. 8). brazieri, Helicina Plate 5, Figure 11 1870c, JdeC 18: 397 (Niue [Island]). Lectotype, here selected, MCZ 74947; paralectotypes MCZ 298500. brazieri, Partula 18711, AJC 7: 27, pl. 9, fig. 5 (Tutuila). Holotype ANSP 59846, teste Baker (1963: 204). brevicula, Leptachatina 1869c, JdeC 17: 169 (Kauai). Measured holotype and paratype in MNHN, teste Fischer-Piette (1950: 72); paratype ANSP 57802 figured by Cooke (1910: 24, pl. 8, fig. 54); paratype MCZ 45195. brevicula 'Pease' Garrett, Partula 1884, JANSP (2) 9: 48 (Faahuaite Valley, Tahiti). Lectotype ANSP 59558a selected by Baker (1963: 204) figured by Pilsbry (1909: 191, pl. 26, fig. 12); paralectotypes MCZ 24997. brevis, Cithara [Cythara] Plate 6, Figure 10 1868a, AJC 3: 217, pl. 15, fig. 11 ([Anaa Island] Pau- motus). Lectotype, here selected, ANSP 15652 is the figured type; paralectotype ANSP 391038; para- types MCZ 49988 and 49989. brunnea, Avicula Plate 1, Figure 3 1863c, PZS for 1862: 244 ([Molokai] Sandwich Is- lands). Lectotype, here selected, MCZ 298465; para- lectotypes MCZ 297881; not located in BMNH by Kay (1965: 86). brunnea, Clatliurella 1860b, PZS 28: 143 (Sandwich Islands). Lectotype BMNH 1962763 selected by Kay (1965: 19, pl. 8, figs. 1-3); paralectotypes MCZ 50006 and 50007. brunnea, Strigatella Plate 9, Figure 10 1868a, AJC 3: 215, pl. 15, fig. 7 ([Paumotus] Polyne- sia). Lectotype, here selected, ANSP 29722 is the figured type; five paralectotypes ANSP 391049; paralectotype MCZ 260610. brunneus, Triphoris Plate 10, Figure 6 1871a, PZS for 1870: 777 (Apaiang [Abaiang Is- land]). Lectotype, here selected, MCZ 73922. buccmoides, Clatliurella 1860d, PZS 28: 144 (Sandwich Islands). Holotype BMNH 1961156, teste Kay (1965: 23, pl. 2, figs. 1, 2). calliostoma, Helicina Plate 5, Figure 5 1871d, PZS for 1871: 466 (Marquesas [Islands]). Lec- totype, here selected, MCZ 74950; paralectotypes MCZ 298501. canaliculata, Clatliurella Plate 6, Figure 16 1868a, AJC 3: 219, pl. 15, fig. 17 (Paumotus). Lec- totype, here selected, MCZ 231978; paralectotypes MCZ 303197; not located in ANSP. cancellata, Coralliobia 1861a, PZS 28: 399 (Sandwich Islands), "only a sin- gle dead specimen found." Not located in BMNH by Kay (1965: 86). cancellata, Scutellina 1861b, PZS 28: 437 (Sandwich Islands). Lectotype BMNH 1962833 selected by Kay (1965: 64, pl. 11, figs. 8, 9). cancellatus, Strombus 1861a, PZS 28: 398 (Sandwich Islands). Lectotype BMNH 1961180 selected by Kay (1965: 40, pl. 4, figs. 8, 9). Candida, "?" Collonia 1861b, PZS 28: 436 (Sandwich Islands). Lectotype BMNH 1962826 selected by Kay (1965: 61, pl. 9, fig. 8); paralectotypes MCZ 245260. Candida, Cypraea 1865d, PZS for 1865: 515 (Islands of the Central Pacific); 1868g, AJC 4: 95, pl. 11, figs. 12, 13 (Apaian Island). Lectotype BMNH 1964318 selected by Kay (1965: 83, pl. 14, figs. 3, 4); paralectotypes MCZ 297949. Candida, Volvatella 1868e, AJC 4: 73, pl. 7, fig. 6 ([Tahiti] Polynesia). Syntypes MCZ 297940, mostly fragments. capillata. Helix Plate 2, Figure 1 1866b, AJC 2: 292 (Sandwich Islands); 1871b, PZS for 1871: 474 as Pitys (Kauai). Lectotype ANSP 1975a (now 1975) selected by Solem (1976: 368 [not fig- ured]) is the measured type, teste Baker (1963: 232); paralectotypes MCZ 17585. castanea 'Pease' Garrett, Partula 1884, JANSP (2) 9: 75 (Faaloa, on the east coast [of Raiatea]). Lectotype ANSP 59977a selected by Baker (1963: 204) figured by Pilsbry (1909: 244, pl. 17, fig. 2); paralectotypes MCZ 297856. celsa, "?" Pitliys 1870c, JdeC 18: 396 (Raiatea); 1871b, PZS for 1871: 445 as Endodonta. Lectotype BPBM 3484 selected by Solem (1976: 358, figs. 158a-c); paralectotypes MCZ 17243 and 17249. cerithiopsis, Rissoina 1862b, JdeC 10: 382 [nomen nudum]. cmcta, Helicina 1865a, PZS for 1864: 676 [nomen nudum]. cmctus, Melampus 1865a, PZS for 1864; 676 [nomen nudum]. cinerea, Littorina 1869d, AJC 5: 78, pl. 8, fig. 14 (Marquesas [Islands]). Lectotype, here selected, ANSP 18811 is the figured type. cmgulifera, Triphoris i861b, PZS 28: 434 ([Haena Point, Kauai] Sandwich Islands). Lectotype BMNH 1962812 selected by Kay MoLLUSKs Described by W. H. Pease • Johnson (1965: 52, pi. 6, figs. 9, 10); paralectotypes MCZ 50056, 50076, and 73737. citrina, Partula 1865b, PZS for 1864: 675 [nomcn nudum] 1866a, AJC 2: 195 (Tahitian Archipelago [restricted to a single valley called Uparu on the west coast of Raiatea, teste Garrett, 1884: 64]). Syntypes MCZ 24885 and 25027. clara, Partula 1865b, PZS for 1864: 671 (Islands of the Central Pacific); 1871d, PZS for 1871: 473 (Tahiti). Syntypes MCZ 24931 and 25141. clathrata, Emarginula 1863b, PZS 'for 1862: 241 (Pacific Islands); 1868g, AJC 4: 99, pi. 1 1, fig. 24 (Rowland Island). Lectotype BMNH 1964290 selected by Kay (1965: 74, pi. 12, fig. 11); paralectotypes MCZ 179161. clavata, Triphons 1861b, PZS 28: 434 (Sandwich Islands). Lectotype BMNH 1962814 selected by Kay (1965: 53, pi. 10, fig. 2); paralectotype MCZ 50060. cognata 'Pease' Garrett, Partula 1865a, PZS for 1864: 675 [nomen nudum]; 1884, JANSP (2) 9: 67, 68 (Faahiti Valley, Huaheine). Specimens figured by Pilsbry (1909: 254, pi. 22, figs. 12, 17; pi. 33, fig. 3). Lectotype ANSP 59991a selected by Baker (1963: 204); paralectotypes MCZ 24817 and 25117. colorata, Helicina 1868), AJC 4: 156, pi. 12, fig. 9 (Annaa [Island]). Holotype ANSP 145541a, teste Baker (1964: 101); paratypes MCZ 297929. compacta, Labiella Plate 4, Figure 5 1869c, JdeC 17: 172 ([Palauea] Maui; Coll. [of] Pease). Lectotype, here selected, MCZ 45196; paralectotype MCZ 298489. compacta, Limnaea 1870b, AJC 6: 6, pi. 3, fig. 4 (Oahu). Syntypes MCZ 302381; not located in ANSP by Baker (1964: 151). Is Physa mexicana Philippi [introduced], teste D. W. Taylor, personal communication. compacta, Partula 1865b, PZS for 1864: 675 [nomen nudum]; 1866a, AJC 2: 200 (Raiatea [Hamoa Valley on the east coast of Raiatea, teste Garrett, 1884: 55]); 1867a, AJC 3: 81, pi. 1, fig. 9. Lectotype ANSP 59983 selected by Baker (1963: 204) is the figured type; paralectotypes MCZ 24993. complanalum, Registoma Plate 4, Figure 3 1861c, PZS 28: 440 (Ebon Island). Lectotype, here selected, MCZ 141018; paralectotypes MCZ 141019. compressa, Scutellina Plate 1, Figure 10 1868g, AJC 4: 99, pi. 11, figs. 25, 27 (Tahiti). Lec- totype, here selected, MCZ 302552. compta, Cypraea 1860c, PZS 28: 189 (Jarvis Island). Holotype BMNH 1964276, teste Kay (1965: 36, pi. 12, figs. 1, 2). compta, Partulma 1869c, JdeC 17: 175 (Molokai). Holotype in MNHN figured by Fischer-Piette (1950: 73, fig. 54); para- types MCZ 25826 and 25828. concinna, Partula 1872a, AJC 7: 196 (New Hebrides). Though not a type Carnegie Museum 4244, "Tanna Island, New Hebrides [Vanuatu] (Cox)" was figured by Hartman (1886: 35, pi. 2, fig. 16) as this species and by Pilsbry (1909: 288, pi. 36, figs. 9, 12); not located in ANSP by Baker (1963: 204). concinna, Truncatella Plate 10, Figure 18 1871d, PZS for 1871: 468 (Apaiang [Abaiang Island, Kingsmill Islands]). Lectotype, here selected, MCZ 178650; paralectotypes MCZ 298459. congrua. Helix Plate 2, Figure 4 1868), AJC 4: 154, pi. 12, figs. 3, 4 (Ponape); 1871b, PZS for 1871: 457 as Trochomorpha contigua to replace H. congrua 1868 non Pfeiffer 1858. Lectotype, here selected, MCZ 12161; paralectotypes MCZ 12159 and 297861; not located in ANSP by Baker (1963). conica, Laimodonta 1863b, PZS for 1862: 242 (Pacific Islands); 1868g, AJC 4: 101, pi. 12, fig. 15 (Paumotus); 1871b, JdeC 19: 94 (Annaa). Lectotype ANSP 22610a selected by Baker (1964: 151) is the figured type; a lectotype BMNH 1964292 was also subsequently selected by Kay (1965: 75, pi. 13, figs. 7, 8). conica, Tormia 1865d, PZS for 1865: 514 (Islands of the Central Pacific). Not located in BMNH by Kay (1965: 86). conoidalis, Tectura Plate 1, Figure 6 1868g, AJC 4: 98, pi. 11, fig. 22 (Roratonga [Island]). Lectotype, here selected, MCZ 302558; paralecto- types MCZ 298470; specimens identified by Pease from a different locality [idiotypes] ANSP 40993. conoidalis, Trochus Plate 9, Figure 26 1868d, AJC 3: 287, pi. 24, fig. 8 (Paumotus). Lecto- type, here selected, ANSP 18868 is the figured type; three paralectotypes ANSP 267209; paralectotypes MCZ 104618, 150597, and 298236. consimilis. Helix 1865b, PZS for 1864: 675 [nomen nudum]; 1868b, AJC 3: 227 (Tahiti [error for Raiatea]). Lectotype BMNH 71 .1.5.28 selected by Solem (1976: 174 [not figured]); paralectotypes MCZ 17262 and 297952. conspersa. Bulla 1869d, AJC 5: 72, pi. 8, fig. 9 (Marquesas [Islands]). Lectotype, here selected, ANSP 57505 is the figured type; paralectotypes MCZ 297880 and 298464. contigua, Melania Plate 4, Figure 19 1870b, AJC 6: 7 (Kauai). Lectotype, here selected, MCZ 74887; paralectotypes MCZ 298908; not lo- cated in ANSP by Baker (1964: 181). contigua, Terebra 1871e, AJC 7: 20. New name for Terebra assimilis Pease 1869 non Angas 1867. See under: assimilis, Terebra. contigua, Trochomorpha 1871d, PZS for 1871: 457. New name for Helix con- grua Pease 1868 non Pfeiffer 1858. See under: con- grua. Helix. conula. Helix Plate 3, Figure 6 1861d, PZS for 1861: 243 (Tahiti). Lectotype, here selected, MCZ 297945; paralectotypes MCZ 298469. coreensis. Turcica Plate 10, Figure 23 1860c, PZS 28: 189, pi. 51, fig. 2 (Corea [Korea] Sea). 10 Bulletin Museum of Comparative Zoology, Vol. 154, No. 1 Lectotype, here selected, MCZ 104609 is the figured type; not located in BMNH by Kay (1965: 85). corrugata 'Pease' Tryon, Borsonia 1884, MofC (1)6: 228 [nomen nudum]. corrugata, Helkina Plate 5, Figure 4 1865a, PZS for 1864: 673 (Islands of the Central Pacific); 1871d, PZS for 1871: 476 (Raiatea). Lecto- type, here selected, MCZ 297925; paralectotypes MCZ 298904. corrugata, Trivia Plate 10, Figure 25 1868g, AJC 4: 95, pi. 11, figs. 14, 15 (Paumotus). Lectotype, here selected, MCZ 303451; paralecto- types MCZ 303197; three specimens identified by Pease [idiotypes] ANSP 39703 labeled as from the Sandwich Islands. corrugatus, Euchelus 1861b, PZS 28: 435 (Sandwich Islands). Lectotype BMNH 1962821 selected by Kay (1965: 58, pi. 8, figs. 12, 13); paralectotypes MCZ 89897. costata, Engina 1860b, PZS 28: 142 (Sandwich Islands). Lectotype BMNH 1961163 selected by Kay (1965: 14, pi. 1, figs. 17, 18). costata, Hydrocaena 1865b, PZS for 1864: 676 [nomen nudum]. costata, Leptothyra Plate 7, Figure 19 1869d, AJC 5: 70 (Maui). Lectotype, here selected, MCZ 245261; paralectotypes MCZ 298461. costata, Realia 1868b, AJC 3: 225 (Tahaa); 1869a, JdeC 17: 158, pi. 7, fig. 2 as Scalinella. Holotype and paratype in MNHN, teste Fischer-Piette (1950: 72); Baker (1964: 178) claimed the holotype is ANSP 13359a; para- types MCZ 74939, 74951, and 187864. costata. Vertigo Plate 2, Figure 20 1871d, PZS for 1871: 461 ([Kona] Hawaii [Island]). Listed as an undetermined species by Pilsbry and Cooke (1920: 272). Holotype MCZ 45238, teste Cooke on label; not MCZ 45327 which is a Goniobasis as mentioned by Pilsbry and Cooke (1926: 224). cost at us, Triphoris Plate 10, Figure 1 1871a, PZS for 1870: 775 (Annaa [Island]). Lecto- type, here selected, MCZ 273206; paralectotypes MCZ 298481. costellifera, Anachis 1863d, PZS for 1862: 279 (Pacific Islands). Not lo- cated in BMNH by Kay (1965: 86). costellifera, Terebra Plate 9, Figure 19 1869d, AJC 5: 66 ([Honolulu] Oahu). Lectotype MCZ 248800 selected by Bratcher and Cernohorsky (1987: 207, pi. 64, fig. 252b). costellifera. Truncal ella 1871d, PZS for 1871: 468 (Vavau Island [Tonga Is- lands]). Not located in BMNH, teste Mordan (per- sonal communication). "Mr. Pease's T. costellifera, which Mr. Brazier obtained at Vavau, Tonga Islands is undoubtedly the same as T. rustica (Mousson)." (Garrett, 1887: 300). costulata, Rissoina Plate 9, Figure 6 1868d, AJC 3: 295, pi. 24, fig. 28 (Paumotus). Lec- totype, here selected, ANSP 19241 is the figured type; three paralectotypes ANSP 391036; paralec- totypes MCZ 178856. costulosa, Atys Plate 6, Figure 5 1869d, AJC 5: 73 ([Waimalu] Oahu). Holotype MCZ 31714, only specimen, figured by Pilsbry (1917: 218, fig. 6). costulosa, Leptachatina 1870a, JdeC 18: 90 (no locality); Crosse, 1876, JdeC 24: 97, pi. 3, fig. 4 (Kauai). Holotype in MNHN, teste Fischer-Piette (1950: 149); paratype MCZ 45191. costulosa, Succinea Plate 4, Figure 15 1865b, PZS for 1864: 677 (Tahitian Archipelago); 1871d, PZS for 1871: 472 (Tahiti). Lectotype, here selected, MCZ 31406; paralectotypes MCZ 298485. costulosa. Vertigo Plate 2, Figure 22 1871d, PZS for 1871: 462 (Hawaii). Lectotype, here selected, MCZ 45244; paralectotypes MCZ 32294; not located in BMNH, teste Mordon (personal com- munication). coxi, Bulimus ("?" Borus) Plate 2, Figure 12 1872b, AJC 7: 197 (Solomon Islands). Holotype MCZ 86495, [New Hebrides Islands], teste Clench (1932: 69). crassa 'Pease' Garrett, Partula 1865a, PZS for 1864: 675 [nomen nudum]; 1884, JANSP (2) 9: 49 (Faahuaite [Valley], Huaheine). Lectotype ANSP 59921a selected by Baker (1963: 204); para- lectotypes MCZ 25078. crassicostata, Borsonia 1860b, PZS 28: 143 (Sandwich Islands). Lectotype BMNH 1962847 selected by Kay (1965: 17, pi. 2, figs. 5, 6). crassilabris, Partula 1865a, PZS for 1864: 675 [nomen nudum]; 1866a, AJC 2: 199, 201 ([Hapai Valley] Raiatea); 1867a, AJC 3: 81, pi. 1, fig. 6, as crassilabrum. Holotype ANSP 59477a, teste Pilsbry (1909: 226, pi. 21, fig. 10) se- lected as lectotype by Baker (1963: 205); paralec- totypes MCZ 25131, 25139, and 25292. crenulata, Dapyhnella Plate 7, Figure 3 1868c, AJC 3: 221, pi. 15, fig. 20 ([Howland Island] Polynesia). Lectotype, here selected, ANSP 15694; paralectotype MCZ 221177. crenulata, Scalaria 1868d, AJC 3: 290, pi. 24, fig. 13 (Tahiti). Holotype ANSP 19553, only specimen; refigured by DuShane (1990: 3, fig. 6). crispata, Scalaria Plate 9, Figure 13 1868d, AJC 3: 289, pi. 24, fig. 12 (Paumotus). Lec- totype ANSP 19575 selected by DuShane (1988 (5): 9, fig. [not numbered]) and refigured (1990: 8, fig. 35); two paralectotypes ANSP 352472; paralecto- types MCZ 187528. crocata, Haminca 1861b, PZS 28: 19, 432 (Sandwich Islands). Lecto- type BMNH 1961191 selected by Kay (1965: 7, pi. 1, figs. 9, 10); paralectotypes MCZ 88127, 207387, and 297883. curta, Daphnella Plate 7, Figure 2 1868a, AJC 3: 221, pi. 15, fig. 22 (Paumotus). Lec- totype, here selected, ANSP 16956; paralectotype MCZ 221178. MoLLUSKS Described by W. H. Pease • Johnson 11 cylmdracca Tease' Nevill, Triiucatclla 1878, HanJ List of Mollusca in the Indian Museum 1: 253 [nonien nudum]. MCZ 161490. cyUndraceum, Cerithium ' 1869d, AJC 5: 77 (Paumotus). Holotype ANSP 17703, teste, Houbrick (1992: 49, fig. 31j); paratypes MCZ 297939 (Plate 6, Figure 6) and 298468. cylmdrata, Leptachatina 1869c, JdeC 17: 168 (Kauai). Lectotype, here select- ed, ANSP 57806 is the syntype figured by Cooke (1 91 0: 18, pi. 8, figs. 63, 64); paralectotypes in MNHN, teste Fischer-Piette (1950: 72). cylindrica, Clathurella ' 1860b, PZS 28: 143 (Sandwich Islands). Lectotype BMNH 1962765 selected by Kay (1965: 20, pi. 8, fig. 8). cylindrica, Marginella ' 1863c, PZS for 1862: 244 (Kingsmill Islands); 1868d, AJC 3: 280, pi. 23, fig. 19 [Tarawa Island] as Mar- ginella polita to replace M. cylindracea [sic] non Sow- erby 1846. Lectotype BMNH 1964300 selected by Kay (1965: 77, pi. 13, figs. 5, 6) is the syntype figured by Reeve (1865, Conchologia Iconica 15: Marginella, pi. 21, fig. 108) as Marginella peasei replacing polita Pease non Carpenter 1857; paralectotypes MCZ 24965. cylindrica, Truncatella ' 1865a, PZS for 1864: 676 [nomen nudum]. MCZ 181051. cylindricus, Triphorus ' 1871a, PZS for 1870: 776 (Apaiang Island). Not lo- cated in BMNH, teste Way (personal communica- tion). cylindricus, Triton Plate 10, Figure 17 ' 1868g, AJC 4: 94, pi. 11, fig. 9 (Tahiti). Lectotype, here selected, MCZ 239749; paralectotypes MCZ 288003; not located in ANSP. cytharoides 'Pease' Pace, Columbella 1 902, Proceedings of the Malacological Society of London 5: 74 [nomen nudum]. daedalea, Cithara [Cythara] Plate 6, Figure 1 1 1868a, AJC 3: 218, pi. 15, fig. 13 (Paumotus); 1868h, AJC 4: 105 as Cythara debilis Pease to replace C. dae- dalea 1868 non Reeve 1846. Lectotype, here selected, ANSP 15663; paralectotype ANSP 391052; paralec- totypes MCZ 231920 and 231921. debilis, Atys 1860a, PZS 28: 20 (Sandwich Islands). Lectotype BMNH 1961197 selected by Kay (1965: 11, pi. 1, figs. 5, 6); probable paralectotypes MCZ 31713 though labeled as from Tahiti. See: Martens and Langkavel (1871, Donum Bismarckianum, p. 53), who also received specimens from Pease thus labeled. debilis, Cythara 1868h, AJC 4: 105. New name for Cithara daedalea Pease 1861 non Reeve 1846. See under: daedalea, Cithara [Cythara]. debilis, Marginella 1871e, AJC 7: 22. New name for Marginella oryza Pease 1860 non Lamarck 1822. debilis, Odostomia Plate 8, Figure 15 1868d, AJC 3: 292, pi. 24, fig. 21 (Howland [Island]). Lectotype, here selected, MCZ 297933; not located in ANSP. decussata, Cithara [Cythara] Plate 6, Figure 7 1868a, AJC 3: 217", pi. 15, fig. 10 ([Anaa Island] Pau- motus). Lectotype, here selected, ANSP 15651; para- lectotype ANSP 391051; paralectotypes MCZ 231922 and 231923. decussata, Scalaria Plate 9, Figure 14 1868d, AJC 3: 289, pi. 24, fig. 10 (Hawaii [Island]). Lectotype ANSP 19585 selected by DuShane (1988 (7): 4, fig. [not numbered]) and refigured (1990: 7, fig. 30); paralectotype ANSP 391064; paralectotypes MCZ 187529 and 187530. decussata, Turbonilla 1861b, PZS 28: 438 (Sandwich Islands). Lectotype BMNH 1962842 selected by Kay (1965: 68, pi. 5, figs. 15, 16). decussatula. Helix 1866b, AJC 2: 291 (Sandwich Islands); 1871d, PZS for 1871: 474 as Pitys (Molokai). Specimens iden- tified by Pease [idiotypes] from Waimea or Wahimi, Kauai MCZ 17273 and 17274; not located in ANSP by Baker (1963: 232) or elsewhere by Solem (1976: 377). deformis. Sty lifer Plate 9, Figure 15 1868d, AJC 3: 293, pi. 24, fig. 23 (Paumotus). Lec- totype, here selected, ANSP 19836 is the figured type; three paralectotypes ANSP 391040; paralec- totypes MCZ 248839 and 248840. delicata, Narica Plate 8, Figure 3 1868d, AJC 3: 282, pi. 23, fig. 25 (Paumotus). Lec- totype, here selected, ANSP 40201; paralectotypes MCZ 231415. delicatus, Pleurobranchus 1861d, PZS for 1861: 245 (no locality); 1868e, AJC 4: 79, pi. 9, fig. 1 (Huaheine). Possible syntype MCZ 297873 though labeled as from Tahiti; not located in BMNH by Kay (1965: 86). dentata, Tornatellina Plate 4, Figure 9 1871d, PZS for 1871: 460 (Hawaii [Island]). Lecto- type [so labeled], here selected, MCZ 28918; para- lectotypes MCZ 175730. dentifera, Vertigo 1871d, PZS for 1871: 462 (Roratonga [error for Ai- tutaki. Cook Islands, teste Garrett (1881: 401), who collected the specimens]). Lectotype, here selected, MCZ 258352 figured by Pilsbry and Cooke (1920: 329, pi. 30, fig. 14); paralectotype MCZ 48314. depressa, Siphonaria Plate 1, Figure 5 1863d, PZS for 1862: 279 (Pacific Islands); 1868g, AJC 4: 98, pi. 1 1, fig. 23 (Apaiang Island). Lectotype ANSP 22199 selected by Baker (1964: 159); not lo- cated in BMNH by Kay (1965: 86). depressiformis. Helix Plate 3, Figure 1 1865b, PZS for 1864: 670 (Islands of the Central Pacific); 1871d, PZS for 1871: 475 (Tahiti). Lecto- type, here selected, MCZ 17342; paralectotypes MCZ 297962. "Is Trochomorpha swainsoni (Pfeiffer), a Raia- tean (Society Island) species," teste Pilsbry and Cooke (1922: 17). depressiformis, Vitrina 1865b, PZS for 1864: 675 [nomen nudum]. 12 Bulletin Museum of Comparative Zoology, Vol. 154, No. 1 deshayesii, Neritina 1868h, AJC 4: 130. New name for Neritma sand- wichensis Deshayes 1838 non Philippi 1843. discoidea, Helicina Plate 5, Figure 6 1865a, PZS for 1864: 676 [nomen nudum]; 1868b, AJC 3: 226 (Tahaa). Holotype ANSP 14398a (now 14398) is the measured type, teste Baker (1964: 161); prob- able paratypes MCZ 297927 though labeled as from Raiatea. discoidea, Torinia Plate 9, Figure 24 1868g, AJC 4: 102, pi. 12, fig. 18 (Paumotus). Lec- totype, here selected, ANSP 38804; paralectotype ANSP 391039; paralectotype MCZ 297941. dispar, Neritina Plate 8, Figure 8 1868d, AJC 3: 285, pi. 24, fig. 3 (Roratonga). Lec- totype ANSP 37714 selected by Baker (1964: 160); nine paralectotypes ANSP 358497; paralectotypes MCZ 73472. distans. Helix 1866b, AJC 2: 290 (Sandwich Islands). Not located in ANSP. distorta, Leiostraca 1861b, PZS 28: 438 (Sandwich Islands). Holotype BMNH 1962841 figured by Kay (1965: 67, pi. 11, figs. 12, 13); paratype MCZ 31706 figured by Pilsbry (1917: 229, figs. 13c, d). dubia, Partula affinis 1865b, PZS for 1864: 675 [nomen nudum]. dubia Tease' Garrett, Partula otaheitana 1865b, PZS for 1864: 675 [jwmen nudum]; 1884, J ANSP (2) 9: 49 (valley several miles from Papinoo [Tahiti]). Syntypes ANSP from Pease figured by Pilsbry (1909: 188, pi. 25, figs. 10, 11) [not in Baker, 1963]; syntypes MCZ 25315. elegans, Clathurella 1860b, PZS 28: 144 (Sandwich Islands). Lectotype BMNH 1961166 selected by Kay (1965: 21, pi. 2, figs. 21, 22; pi. 8, fig. 6); paralectotype MCZ 50009. elongata, Hydrocaena 1865b, PZS for 1864: 676 [nomen nudum]. elongata, Partula 1865b, PZS for 1864: 676 [nometi nudum]; 1866a, AJC 2: 196 ([Vianae Ravine] Moorea); 1867a, AJC 3: 81, pi. 1, fig. 2. Lectotype ANSP 59477a selected by Baker (1963: 205) is the figured type; paralectotypes MCZ 24899, 24921, 25127, and 25294. elongata, Realia 1868b, AJC 3: 225 (Raiatea); 1869a, JdeC 17: 152, pi. 7, fig. 4 as Omphalotropis. Holotype in MNHN, teste Fischer-Piette (1950: 72); Baker (1964: 178), appar- ently unaware of this, selected as lectotype ANSP 13350a; paratypes MCZ 187861 and 187862. elongata, Succtnea Plate 4, Figure 14 1870a, JdeC 18: 96 (Kauai). Lectotype, here selected, MCZ 161665, probable measured type. elongata, Syphonota 1860a, PZS 28: 24 (Sandwich Islands). Syntypes MCZ 31442 and 298486; not located in BMNH by Kay (1965: 84). elongata, Turbonilla Plate 10, Figure 22 1868d, AJC 3: 293, pi. 24, fig. 22 (Paumotus). Lec- totype, here selected, MCZ 10537; paralectotypes MCZ 10539; not located in ANSP. elongata, Volutella 1868d, AJC 3: 281, pi. 23, fig. 23 (Fanning Island). Not located in ANSP. erecta, Laminella 1869c, JdeC 17: 174 (Maui). Holotype in MNHN figured by Fischer-Piette (1950: 73, fig. 53); para- types MCZ 23338. ericea, Mitra 1860b, PZS 28: 146 (Sandwich Islands); 1869f, AJC 5: 85 is Mttra turgida Reeve 1845. Lectotype BMNH 1961161 selected by Kay (1965: 28, pi. 3, figs. 3, 4); paralectotypes MCZ 260605. exaratus, Rhizochilus 1861a, PZS 28: 399 (Sandwich Islands); 1868h, AJC 4: 115 is Coralliophda deformis Lamarck 1816. Lec- totype BMNH 1961177 selected by Kay (1965: 41, pi. 4, figs. 10, 11); paralectotypes MCZ 297936. exilis, Clathurella 1860b, PZS 28: 144 (Sandwich Islands). Lectotype BMNH 1961166 selected by Kay (1965: 21, pi. 2, figs. 21, 22; pi. 8, fig. 6); paralectotype MCZ 50009. exilis, DrUlia 1868a, AJC 3: 220, pi. 15, fig. 19 (Tahiti). Lectotype, here selected, ANSP 15690 is the figured type; para- lectotype ANSP 316068. exilis, Eulima Plate 7, Figure 11 1863b, PZS for 1862: 242 (Pacific Islands); 1868d, AJC 3: 294, pi. 24, fig. 25 (Paumotus). Lectotype, here selected, MCZ 187831; paralectotype MCZ 187832; not located in BMNH by Kay (1965: 86). exilis, Trochus Plate 10, Figure 26 1868d, AJC 3: 286, pi. 24, fig. 7 (Paumotus). Lecto- type, here selected, MCZ 104617; paralectotype MCZ 303193; not located in ANSP. expansa, Auriculella 1868f, JdeC 16: 343, pi. 14, fig. 8 (Hawaiian Islands). Holotype in MNHN, teste Fischer-Piette (1950: 71); paratypes MCZ 45155. expansa, Partula 1871f, AJC 7: 26, pi. 9, fig. 3 (Tutuila). Holotype ANSP 59453, teste Baker (1963: 205). extensa, Leptachatina 1870a, JdeC 18: 92 (Kauai). Not mentioned by Fi- scher-Piette (1950: 74). faba, Helicina 1865a, PZS for 1864: 676 [nomen nudum]; 1868b, AJC 3: 226 [nomen nudum] (neighboring island near Ta- haa). faba 'Pease' Garrett, Helicina 1884, JANSP (2) 9: 105, pi. 3, figs. 61, 61a, b (Raiatea and Moorea). Holotype ANSP 14546a, teste Baker (1964: 161). fabrefacta, Helix Plate 3, Figure 2 1865a, PZS for 1864: 669 ([Raiatea]). Lectotype, here selected, MCZ 17238; paralectotypes MCZ 176568. The neotype USNM 42427 selected by Solem (1976: MoLLUSKS Described by W. H. Pease • Johnson 13 363) but not figured is considered here to be invalid since authentic types are available. fasciata, Partula 1865a, PZS for 1864: 675 [nomcu nudum]; 1866a, AJC 2: 202 (Marquesas Islands) is Partula ganymedes Pfeif- fer, teste Pease (1866, AJC 2: 293). Syntypes MCZ 25113 and 25324. fascuita, Planaxis 1868g, AJC 4: 102, pi. 12, fig. 17 (Paumotus). Ho- lotype ANSP 18286, only specimen; specimens sub- sequently identified by Pease [idiotypes] MCZ 187838. ficta, Helix 1865a, PZS for 1864: 669 (no locality, Tahaa [teste Garrett 1884: 38]). Lectotype USNM 24213 selected by Solem (1976: 362 [not figured]); paralectotypes MCZ 17240. filocostata, Pitys 1871d, PZS for 1871: 454 (Kauai). Not located in BMNH, teste Mordan personal communication). fimbriatus, Euchelus 1861c, PZS 28: 435 (Sandwich Islands). Holotype BMNH 1962823 figured by Kay (1965: 58, pi. 8, figs. 14, 15); paratype MCZ 89896. fisheri Tease' Paetel, Dolahrifera 1888, Catalog der Conchylien-Sammlung (Berlin) 1: 635 [nomen nudum]. flammea, Rissoa Plate 9, Figure 7 1868d, AJC 3: 297, pi. 24 [not 14], fig. 33 (Caroline [Islands]). Lectotype MCZ [not ANSP] 178868 [nor MCZ 178863 as on plate caption] selected by Ponder and de Keyzer (1992; 1058, fig. 3C) labeled as ho- lotype; six paralectotypes MCZ 178867 [not 178861] borrov/ed from the MCZ in 1990 but, "not exam- ined"; not located in ANSP. flammulata, Mitra 1868a, AJC 3: 212 (Sandwich [Islands] and Pau- motus). Not located in ANSP. flammulata, Tnphoris 1861b, PZS 28: 434 ([Haena, Kauai] Sandwich Is- lands). Lectotype BMNH 1961175 selected by Kay (1965: 52, pi. 6, figs. 15, 16); paralectotypes MCZ 50065 and 50066. flai'escens, Helicwa 1868b, AJC 3: 228, pi. 15, fig. 25 (Mangaia [Cook Islands]); 1871d, PZS for 1871: 476; here Pease ad- mits that Helicina flavescens is a redescription of his H. paafica. Holotype ANSP 14401a, teste Baker (1964: 161); paratypes MCZ 297928. formosa, Cylmdra Plate 7, Figure 1 1868d, AJC 3: 271, pi. 23, fig. 1 (Ascension [Island]). Holotype MCZ 260605 figured by Cernohorsky (1991: 132, pi. 130, fig. 2); not located in ANSP. formosa 'Pease' Garrett, Partula 1884, JANSP (2) 9: 60, pi. 3, fig. 49 (Fatimu on the southwest part of Raiatea). Holotype ANSP 59453 refigured by Pilsbry (1909: 218, pi. 20, fig. 5); para- types MCZ 25193. fortiplicata, Tumcula (Costellana) Plate 10, Figure 21 1868a, AJC 3: 213, pi. 15, fig. 3 (Paumotus). Lecto- type, here selected, ANSP 28844; paralectotype ANSP 391043; paralectotypes MCZ 260609. foveolatus, Murex 1869e, AJC 5: 83, pi. 8, fig. 3 (Gulf of [Golfo de] California, La Paz [Baja California Sur, Mexico]). Tryon, 1880, MofC (1)2: 129 as Murex peasei to re- place M. foveolatus non Hinds 1844. Lectotype ANSP 36144 selected by Myers and D'Attilio (1989: 155, figs. 1, 2); two paralectotypes MCZ 304068. fragilts, Hydrocena 1861c, PZS 28: 439 (Ebon Island); 1869a, JdeC 17: 145, pi. 7, fig. 6 as Omphalotropis. Holotype in MNHN, teste Fischer-Piette (1950: 72); paratypes MCZ 297957. fragilis, Volvatella 1860a, PZS 28: 20 (Sandwich Islands); 1868e, AJC 4: 73, pi. 7, fig. 4. Holotype BMNH 1962753, teste Kay (1965: 12 [not figured]). fratercula, Helix 1867b, AJC 3: 104. New name for Helix sculptilis Pease 1865 non Bland 1858. See under: sculptilis, Helix. fratercula, Pitys 1871d, PZS for 1871: 452 [nomen nudum] (Hervey Islands). MCZ 176558 and 297860. frwola. Helix 1866b, AJC 2: 290, pi. 21, fig. 3 (Polynesia); 1871d, PZS for 1871: 475 as Helicopsis (Oualan [Island]). Holotype ANSP 49296a, teste Baker (1963: 334); paratypes MCZ 135675. fucata, Scalaria 1861a, PZS 28: 400 (Sandwich Islands). Lectotype BMNH 1961168 selected by Kay (1965: 43, pi. 6, figs. 11, 12); paralectotypes MCZ 187526. fucata, Triphoris 1861b, PZS 28: 433 ([Kauai] Sandwich Islands). Lec- totype BMNH 1961171 selected by Kay (1965: 51, pl. 6, figs. 13, 14); paralectotypes MCZ 50067 and 73736. fucatum, Cerithium 1861b, PZS 28: 432 (Sandwich Islands). Lectotype BMNH 1962800 selected by Kay (1965: 46, pi. 5, figs. 7, 8). fusca, Dolabifera 1868e, AJC 4: 76, pl. 8, fig. 4 ([Tahiti] Polynesia). Syntype MCZ 297870. fusca, Partula 1865b, PZS for 1864: 675 [nomen nudum]; 1866a, AJC 2: 193 (Raiatea). Syntypes MCZ 25363; not located in ANSP by Baker (1965: 205). fusca, Vitrina 1868), AJC 4: 155, pl. 12, fig. 6 (Marquesas [Islands]). Holotype ANSP 49270a, teste Baker (1963: 234). fuscata. Helix 1865a, PZS for 1864: 675 [nomen nudum]. fuscata 'Pease' Pilsbry, Trochomorpha 1896, Nautilus 9: 120 (Ponape, Caroline Islands). Ho- lotype ANSP 1934, teste Baker (1963: 238). fuscescens, Strigatella 1860b, PZS 28: 146 (Sandwich Islands). Lectotype BMNH 1961184 selected by Kay (1965: 29, pl. 3, figs. 7, 8). 14 Bulletin Museum of Comparative Zoology, Vol. 154, No. 1 fuscolineata, Neptunea 1860c, PZS 28: 189 (Corea [Korea] Sea). Holotype BMNH 1964277, teste Kay (1965: 35, pi. 12, figs. 14, 15). fuscomaculata, Clathurella 1860b, PZS 28: 144 (Sandwich Islands). Lectotype BMNH 1961153 selected by Kay (1965: 23, pi. 2, figs. 7, 8). fuscomaculata, Cypraea 1865d, PZS for 1865: 515 (Islands of the Central Pacific); 1868g, AJC 4: 95, pi. 1 1, figs. 10, 1 1 (Apaian Island [Garrett, 1879: 113, stated that he collected the types from the outer reefs of the Kingsmill Islands]). Lectotype BMNH 1964316 selected by Kay (1965: 83, pi. 14, figs. 1, 2). fusconigra, Vexilla 1860b, PZS 28: 141 (Sandwich Islands); 1868h, AJC 4: 115. Lectotype BMNH 1961164 selected by Kay (1965: 13, pi. 4, figs. 3, 4). fuscus Thilippi' Pease, Melampus 1865a, PZS for 1864: 676 [rwmen nudum]. fusiformis, Columbella 1868h, AJC 4: 122. New name for Columbella pusilla Pease 1862 non Sowerby 1844. See under: pusilla, Columbella. fusiformis, Conus 1861a, PZS 28: 398 (Sandwich Islands); 1868h, AJC 4: 126 non Fischer 1807. Changed to: Conus parvus. Lectotype BMNH 1962788 selected by Kay (1965: 38, pi. 10, fig. 12); paralectotypes MCZ 197346 and 197347. fusiformis, Engina 1865d, PZS for 1865: 513 (Islands of the Central Pacific); 1868d, AJC 3: 273, pi. 23, fig. 5 (Howland Island). Lectotype BMNH 1964309 selected by Kay (1965: 81, pi. 13, figs. 15, 16); paralectotypes MCZ 297954. fusiformis, Mitropsis 1868a, AJC 3: 212, pi. 15, fig. 2 (Paumotus). Lecto- type, here selected, ANSP 28756 is the figured type. galba, Haminea 1861b, PZS 28: 432 (Sandwich Islands). Lectotype BMNH 1961 194 selected by Kay (1965: 8, pi. 1, figs. 11, 12). garrettii, Cythara 1860b, PZS 28: 147 (Sandwich Islands); 1868h, AJC 4: 105. Lectotype BMNH 1962780 selected by Kay (1965: 32, pi. 10, fig. 11); paralectotypes MCZ 49999 and 298242. garrettii, Fossar 1868h, AJC 4: 128. New name for Adeorbis costata Garrett 1857 non Philippi 1844. garrettii, Murex 1868h, AJC 4: 103. New name for Murex exiguus Garrett 1857 non Broderip 1832. garrettii, Partula 1865a, PZS for 1864: 672 (Islands of the Central Pacific [Vaioara on the west coast of Raiatea, teste Garrett, 1884: 56]). Syntype ANSP figured by Pils- bry (1909: 228, pi. 21, fig. 15), but not located by Baker (1963); syntypes MCZ 24894 and 25306. gibbus, Latirus Plate 7, Figure 16 1865c, PZS for 1865: 54 (Pacific Islands); 1868d, AJC 3: 279, pi. 23, fig. 17 (Howland Island). Lectotype, here selected, MCZ 261182; paralectotype MCZ 303192; not located in BMNH by Kay (1965: 86) or in ANSP. glabra, Mitra 1868d, AJC 3: 272, pi. 23, fig. 2 (Ascension [Island]); 1869f, AJC 5: 85 is Mitra lubrica Pease 1869, new name for M. glabra Pease 1868 non Swainson 1821 non Risso 1826. Lectotype, here selected, ANSP 28854 is the figured type. globosa 'Pease' Pilsbry, Partula 1865a, PZS for 1864: 675 [rwmen nudum]; 1909, MofC (2)20: 224. Manuscript name listed under the syn- onymy of Partula heba (Pfeiffer) (south end of Ra- iatea, Garrett ms); MCZ 25275. gracile, Cerithium 1861b, PZS 28: 432 (Sandwich Islands). Lectotype BMNH 1961173 selected by Kay (1965: 45, pi. 5, figs. 5, 6). gracilior 'Pease' Hartman, Partula 1881, BMCZ 9: 183 ([Moorea]). Syntype MCZ 25059. gracilis, Blauneria 1860b, PZS 28: 145 (Sandwich Islands). Lectotype BMNH 1962770 selected by Kay (1965: 26, pi. 9, fig. 3); paralectotypes MCZ 297790. gracilis, Citharopsis Plate 6, Figure 13 1868g, AJC 4: 97, pi. 11, fig. 20 (Paumotus). Lecto- type, here selected, ANSP 16921. gracilis, Mucronalia Plate 7, Figure 24 1868d, AJC 3: 295, pi. 24, fig. 27 (Tahiti). Lectotype MCZ 288506 selected by Waren (1980: 294, fig. 100); paralectotypes MCZ 248841 and ANSP 391065. gracilis, Nassa Plate 8, Figure 5 1868d, AJC 3: 273, pi. 23, fig. 4 (Ascension [Island]). Lectotype, here selected, MCZ 228822; "Appears to be lost," Cernohorsky (1984: 66). gracilis, Odostomia 1868d, AJC 3: 292, pi. 24, fig. 20 (Hawaii [Island]). Lectotype ANSP 19967 selected by Pilsbry (1917: 321, fig. 18); paralectotypes MCZ 10489. gracilis, Partula 1865b, PZS for 1864: 675 [nomen nudum]; 1866a, AJC 2: 197 (Tahitian Archipelago [upper portions of all the central valleys on both the east and west sides of Raiatea. It is more abundant in Tolo and Hapai Valleys than elsewhere. It occurs more rarely at Tahiti, teste Garrett, 1884: 70]); 1867a AJC 3: 81, pi. 1, fig. 3. Not located in ANSP by Baker (1963: 205); syntypes MCZ 24905 and 25115 both from Raiatea. gracilis, Rissoa Plate 9, Figure 3 1861b, PZS 28: 438 ([Kauai] Sandwich Islands); 1862b, JdeC 10: 382 as Rissoina. Lectotype, here selected, MCZ 178853; paralectotype MCZ 298460; not lo- cated in BMNH by Kay (1965: 86) or in MNHN by Fischer-Piette (1950: 20). MoLLUSKS Described by W. H. Pease • Johnson 15 gracilis, Tornalcllina Plate 4, Figure 10 1871d, PZS for 1871: 460 (Kauai). Lectotype, here selected, MCZ 302554. gracilis, Triphoris Plate 10, Figure 10 1871a, PZS for 1870: 774 (Kauai). Lectotype, here selected, MCZ 50058; paralectotype MCZ 298493. gracilis, Triphoris 1871a, PZS for 1870: 777 (Kauai). The description of this and the preceding Triphoris gracilis differ but slightly and though it appears that Pease described the same shells under the same name, Tryon (1872, AJC 7: 206) renamed the second gracilis, T. peasei. grandis, Syphonota 1860a, PZS 28: 23 (Sandwich Islands). Syntype MCZ 297866 [of questionable diagnostic value]; not lo- cated in BMNH by Kay (1965: 84). granifera, Fissurella 1861d, PZS for 1861: 244 (Sandwich Islands). Lec- totype BMNH 1964278 selected by Kay (1965: 70, pi. 12, figs. 3, 4); paralectotypes MCZ 150701. granifera, Narica Plate 8, Figure 2 1869d, AJC 5: 78, pi. 8, fig. 13 (Jarvis [Island]). Lec- totype, here selected, ANSP 37301; two paralecto- types ANSP 391041; paralectotypes MCZ 231414. grantferus, Vertagus 1861b, PZS 28: 433 (Sandwich Islands). Holotype BMNH 1961208, teste Kay (1965: 49, pi. 5, figs. 9, 10). granocostata, Scutellina Plate 5, Figure 1 1868g, AJC 4: 100 (Hawaii [Island]). Lectotype, here selected, MCZ 83723 labeled as holotype. grariosus, Triphoris Plate 10, Figure 9 1871a, PZS for 1870: 776 (Tahiti). Lectotype, here selected, MCZ 273207; paralectotypes MCZ 288954. granulata, Cypraea 1863d, PZS for 1862: 278 (Pacific Islands). Lectotype BMNH 1964306 selected by Kay (1965: 79, pi. 14, figs. 17, 18); paralectotypes MCZ 297955 from An- naa Island, Tuamotu Archipelago. granulosa, Collonia Plate 6, Figure 20 1868g, AJC 4: 92, pi. 11, fig. 4 (Ponape). Lectotype, here selected, MCZ 245264; paralectotypes MCZ 297961; not located in ANSP. granulosa, Rissoina 1862b, JdeC 10: 382, pi. 13, fig. 10 (Sandwich [Is- lands]). Holotype and paratype in MNHN, teste Fi- scher-Piette (1950: 20); paratype MCZ 178862. granulosus, Latirus Plate 7, Figure 17 1868d, AJC 3: 279, pi. 23, fig. 10 [fig. 18 in error] (Paumotus). Lectotype MCZ 261181 selected by Cernohorsky (1987: 97, fig. 5) is the figured type; paralectotypes MCZ 297961. guppyi, Helicina 1871d, PZS for 1871: 467. New name for Helicina humilis Guppy 1868 non Rousseau 1854. harpa, Clathurella 1860b, PZS 28: 144 (Sandwich Islands). Lectotype BMNH 1961206 selected by Kay (1965: 22, pi. 2, figs. 23, 24). hawaiiensis, Perna 1871f, AJC 7: 25. Invalid new name for Perna cali- f arnica Conrad 1834. Changed because calif arnica was a misnomer. hutchinsonii, Helicter Plate 4, Figure 2 1862a, PZS for 1862: 7 (Maui). Lectotype, here se- lected, MCZ 45254; paralectotype MCZ 141500. ignominiosus 'Pease' Paetel, Achatinella 1873, Catalog der Conchylien-Sammlung (Berlin), p. 105 [nomem nudum]. imbricata, Vanikara 1861b, PZS 28: 435 (Sandwich Islands). Holotype BMNH 1962820, teste Kay (1965: 57, pi. 8, fig. 11). imperforata 'Pease' Garrett, Partula 1884d, JANSP (2) 9: 54, pi. 3, fig. 53 (Toloa and Hapai Valleys, west coast of Raiatea [Island], Society Islands). Holotype ANSP 59502a, teste Baker (1963: 205); holotype and two paratypes figured by Pilsbry (1909: 221, pi. 20, figs. 13-15); paratypes MCZ 25262 from Hapai Valley. imperforata, Pithys 1870c, JdeC 18: 394 (Aitutake [Island, Cook Is- lands]); 1871, PZS for 1871: 453. Lectotype BPBM 2322 selected by Solem (1976: 170, figs. 76e, f); para- lectotypes MCZ 17279. impressa 'Pease' Paetel, Achatinella 1873, Catalog der Conchylien-Sammlung (Berlin), p. 105 [nomen nudum]. incisa, Triphoris 1861b, PZS 28: 434 ([Haena Point, Kauai] Sandwich Islands). Lectotype BMNH 1961151 selected by Kay (1965: 54, pi. 6, figs. 19, 20); paralectotypes MCZ 50061 and 73738. "?" inconspicua, Helicina 1865b, PZS for 1864: 676 [namen nudum]. inflexa, Eulima Plate 7, Figure 9 1868d, AJC 3: 294, pi. 24, fig. 26 (Paumotus). Lec- totype, here selected, ANSP 59334 is the figured type; five paralectotypes ANSP 391056; paralecto- types MCZ 31703 and 31704. inflexa, Terebra szvainsonii 1869d, AJC 5: 64. Based on specimen figured by Reeve, 1860, Conchologia Iconica 12, Terebra, pi. 22, fig. 118. Holotype BMNH 1979113 is the figured type. interlirata, Neritopsis Plate 8, Figure 9 1868d, AJC 3: 282, pi. 23, fig. 26 (Annaa Island). Lectotype, here selected, MCZ 73479; not located in ANSP. intermedius, Tritan Plate 10, Figure 16 1869d, AJC 5: 74 (Oahu). Lectotype MCZ 191331 selected by Clench and Turner (1957: 217, pi. 122, fig. 2); paralectotypes MCZ 191330. interrupta, Daphnella 1860b, PZS 28: 147 (Sandwich Islands). Lectotype BMNH 1962849 selected by Kay (1965: 34, pi. 5, fig. 13); paralectotypes MCZ 50011. kauaiensis, Melania 1870b, AJC 6: 7, pi. 3, fig. 6 (Kauai). Holotype ANSP 26510, teste Baker (1964: 190); paratypes MCZ 74929. 16 Bulletin Museum of Comparative Zoology, Vol. 154, No. 1 labiata Tease' Pilsbry, Partula 1909, MofC (2) 20: 217, pi. 20, fig. 9 ([Vairahi Valley, Raiatea]). Holotype ANSP 59460, teste Baker (1963: 205); paratype MCZ 24882. labiata Tease' Schmeltz, Partula 1865a, PZS for 1864: 675 [nomen nudum]; 1874, Mu- seum Godeffroy (Hamburg). Catalog 5: 207 [nomen nudum]. labiata, Succinea Plate 4, Figure 16 1865a, PZS for 1864: 675 [nomen nudum]; 1868b, AJC 3: 227 (Raiatea). Lectotype, here selected, MCZ 298909; paralectotypes MCZ 31409; not located in ANSP by Baker (1963: 215). laevis, Gena Plate 6, Figure 2 1868d, AJC 3: 283, pi. 23, figs. 7, 28, 29 (Tahiti). Lectotype, here selected, ANSP 40754; paralecto- types MCZ 297919. laevis, Lamellina Plate 2, Figure 16 1865a, PZS for 1864: 672 (Islands of the Central Pacific); 1871, PZS for 1871: 473 (Tahiti). Lectotype, here selected, MCZ 154942 labeled "Hervey Is- lands"; syntype ANSP figured by Pilsbry (1915: 165, pi. 33, fig. 6), but not located in ANSP by Baker (1963). laevis, Leptachatina 1870a, JdeC 18: 91 (Kauai); Crosse, 1876, JdeC 24: 97, pi. 4, fig. 6. Holotype in MNHN, teste Fischer- Piette (1950: 149); paratypes MCZ 45173. laevis, Realia 1865e, AJC 1: 289 (Polynesia); Tryon, 1866a, AJC 2: 82, pi. 5, fig. 5 (Oualan Island); Pease, 1869a, JdeC 17: 145 as Omphalotropis (Ponape [or] Ascension [Is- land]). Holotype ANSP 12225a, teste Baker (1964: 178); paratypes MCZ 187921, 187922, and 187923. laminata. Helix Plate 3, Figure 4 1866b, AJC 2: 292 (Sandwich Islands); 1871d, PZS for 1871: 474 as Endodonta (Kauai). Lectotype, here selected, MCZ 17233; paralectotypes MCZ 298477 and BPBM, teste MCZ label; not located in ANSP by Baker (1963: 232) or elsewhere by Solem (1976: 377). lateritia 'A. Adams' Pease, Assiminea 1869b, JdeC 17: 164 [nomen nudum]. Error ioT Assimi- nea latericea H. and A. Adams 1864. lauta, Bullina 1860a, PZS 28: 19 (Sandwich Islands). Holotype BMNH 1961201, teste Kay (1965: 6, pi. 1, figs. 3, 4). lauta, Drillia Plate 7, Figure 8 1868a, AJC 3: 220, pi. 15, fig. 18 ([Anaa Island] Pau- motus). Lectotype, here selected, ANSP 15692 is the figured type; paralectotypes MCZ 49981. lauta, Terebra Plate 9, Figure 21 1869d, AJC 5: 66 (Oahu). Holotype ANSP 33589 figured by Tryon (1885: 33, pi. 10, fig. 91), teste Cernohorsky and Bratcher (1976: 139). lent a. Helix 1865a, PZS for 1864: 675 [nomen nudum]. lenticulina, Helicina 1865a, PZS for 1864: 676 [nomen nudum]. liberatus, Capulus 1868d, AJC 3: 285, pi. 24, fig. 2 (Paumotus). Not located in ANSP. lignaria, Partula '^ 1865b, PZS for 1864: 671 (Islands of the Central Pacific [valley about two miles west of Fautaua [Ta- hiti], teste Garrett, 1884: 48]). Syntypes MCZ 25136, 25142, and 25058. lineata, Alcyna Plate 6, Figure 4 1869d, AJC 5: 69 ([Puuloa] Oahu). Holotype MCZ 31724 figured by Pilsbry (1917: 212, pi. 15, fig. 4). Imeata, Columbella 1861a, PZS 28: 399 (Sandwich Islands). Not located in BMNH by Kay (1965: 85). lineolata, Partula 1865a, PZS for 1864: 675 [nomen nudum]; 1868b, AJC 3: 224 (Tahiti). Holotype ANSP 59933a, teste Baker (1963: 205); paratypes MCZ 24933. lirata, Pleurotoma 1869d, AJC 5: 68 (Oahu). Not located in ANSP. liratus, Latirus Plate 7, Figure 15 1868i, AJC 4: 152, figured by Reeve, 1847, Conchol- ogia Iconica 4: Turbinella, pi. 12, figs. 61a, b (Mar- quesas [Islands]). Lectotype, here selected, MCZ 302628; paralectotypes MCZ 297938; possible para- lectotypes BMNH 1979031; location of figured type unknown. lubrica, Mitra 18691, AJC 5: 85. New name for Mitra glabra Pease 1868 non Swainson 1821 non Risso 1826. See under: glabra, Mitra. lucida, Assiminea 18691, JdeC 17: 166, pi. 7, fig. 10 (Annaa Island). Holotype in MNHN, teste Fischer-Piette (1950: 72); paratypes MCZ 74952. lucida, Leptachatina 1869g, PZS for 1869: 650 [nomen nudum]; 1870a, JdeC 18: 93 (Kauai). Not mentioned by Fischer-Piette (1950: 74). lucidus, Melampus Plate 4, Figure 22 1 869d, AJC 5: 75 ([Honolulu] Oahu). Lectotype ANSP 22284 selected by Baker (1964: 151); paralectotype ANSP 391058; paralectotypes MCZ 297789. lugubris, Partula "var.?" 1865a, PZS for 1864: 672 (Islands of the Central Pacific [Hapai Valley, Raiatea, teste Garrett, 1884: 77]); 1871d, PZS for 1871: 473. Syntypes MCZ 3641 and 25288. lutea, Borsonia 1860b, PZS 28: 143 (Sandwich Islands). Lectotype BMNH 1962756 selected by Kay (1965: 17, pi. 2, figs. 15, 16); paralectotypes MCZ 49978. luteostoma, Ranella 1861a, PZS 28: 397 (Sandwich Islands). Not located in BMNH by Kay (1965: 85). maculata, Engina lineata 1869d, AJC 5: 76, pi. 8, fig. 12 (Apaian). Lectotype, here selected, ANSP 34538 is the figured type. maculata, Nerita Plate 8, Figure 7 1868d, AJC 3: 286, pi. 24, fig. 6 (Tahiti). Lectotype, here selected, ANSP 37490; nine paralectotypes ANSP 391034. MoLLUSKs Described by W. H. Pease • Johnson 17 niticulatuf. Tnphons Plate 10, Figure 13 1871a, PZS for 1870: 777 (Kauai). Lectotype, here selected, MCZ 50069; paralectotype MCZ 298494. maculosa, Clathurella Plate 6, Figure 14 1863b, PZS for 1862: 242 (Pacific Islands); 1868a, AJC 3: 219, pi. 15, fig. 16 (Paumotus). Lectotype, here selected, ANSP 48693; not located in BMNH by Kay (1965: 86). maculosa, Columbella 1871e, AJC 7: 22. New name for Columbella dermes- toides Angas 1867 non Sowerby 1844. maculosa, Daphnella 1860b, PZS 28: 148 (Sandwich Islands). Not located in BMNH by Kay (1965: 85). maculosus Euchclus 1863c, PZS for 1862: 243 (Pacific Islands); 1868g, AJC 4: 91, pi. 11, fig. 1 as Collotiia maculosa ([Anaa Island] Paumotus). Lectotype BMNH 1964298 se- lected by Kay (1965: 77, pi. 13, figs. 1, 2); figured paralectotype ANSP 38414; paralectotypes MCZ 89828. mammillata, Succinea Plate 4, Figure 12 1871b, PZS for 1871: 459 (Nukahiva). Lectotype, here selected, MCZ 155145; paralectotype MCZ 298906; not located in BMNH, teste Mordan (per- sonal communication). marginatus, Pleurobranchus 1860a, PZS 28: 25 (Sandwich Islands). Not located in BMNH by Kay (1965: 84). marmorata. Purpura Plate 8, Figure 19 1865d, PZS for 1865: 515 (Islands of the Central Pacific); 1868g, AJC 4: 92, pi. 11, fig. 5 (Apaian Is- land). Lectotype, here selected, MCZ 177824; para- lectotypes MCZ 302627; not located in BMNH by Kay (1965: 86). marmorata Tease' Martens and Langkavel, Triforis [sic] Plate 10, Figure 5 1871, Donum Bismarckianum, p. 38, pi. 2, fig. 7 ([Kau- ai] Sandwich Islands). Lectotype, here selected, MCZ 50055; paralectotypes MCZ 298492. marmorea. Bulla 1861b, PZS 28: 431 (Sandwich Islands). Holotype BMNH 1961209 figured by Kay (1965: 44, pi. 1, figs. 13, 14); paratypes MCZ 297878. marmorea 'Pease' Paetel, Dolabnfera 1888, Catalog der Conchylietr-Sammlung (Berlin) 1: 635 [nomen nudum]. marmorea, Margarita 1861b, PZS 28: 435 (Sandwich Islands). Lectotype BMNH 1962824 selected by Kay (1965: 59, pi. 9, fig. 5). marmoreus, Trochus Plate 9, Figure 25 1868b, AJC 3: 287, pi. 24, fig. 9 (Paumotus). Lecto- type, here selected, ANSP 40614; paralectotype ANSP 391032; paralectotype MCZ 89894. marquesana, Helix Plate 3, Figure 5 1868), AJC 4: 153, pi. 12, fig. 2 (Marquesas [Islands]). Lectotype, here selected, MCZ 302557; paralecto- types MCZ 297937; not located in ANSP by Baker (1963). megastoma 'Pease' Pilsbry, Partula 1909, MofC (2) 20: 214, pi. 20, fig. 5 ([Raiatea: re- stricted to the higher portion of Haamoa Valley]). Holotype ANSP 5945a, teste Baker (1963: 205); para- type MCZ 25019. megastoma 'Pease' Schmeltz, Partula 1865b, PZS for 1864: 675 [twmen nudum], 1874, Mu- seum Godeffroy (Hamburg). Catalog 5: 92 [nomen nu- dum]. micans 'Pease' Tryon, Columbella Plate 6, Figure 19 1883, MofC (1) 5: 124, pi. 48, fig. 85 (Paumotus; Viti Isles). Lectotype, here selected, MCZ 304063 from the Paumotus. microstoma, Nassa 1860b, PZS 28: 145 (Sandwich Islands). Holotype BMNH 1961458 figured by Kay (1965: 24, pi. 3, figs. 5, 6); paratype MCZ 25341. microstoma 'Pease' Hartman, Partula 1881, BMCZ 9: 184 ([Vairahi Valley] Raiatea). Syn- type ANSP figured by Pilsbry (1909: 233, pi. 27, fig. 14); not located in ANSP by Baker (1963: 205); syn- types MCZ 25341. millecostata, Scalaria 1861a, PZS 28: 400 (Sandwich Islands). Holotype BMNH 1961170, teste Kay (1965: 42, pi. 6, figs. 5, 6). minimus, Tnphons Plate 10, Figure 7 1871a, PZS for 1870: 774 (Howland Island; [Haena] Kauai). Lectotype, here selected, MCZ 50071; para- lectotypes MCZ 50070 and 298495 all from Kauai. modest a, Turricula Plate 10, Figure 20 1868a, AJC 3: 212, pi. 15, fig. 6 ([Ponape] Polynesia). Lectotype, here selected, ANSP 28780; paralecto- types MCZ 260607. monilifera, Engina 1860b, PZS 28: 142 (Sandwich Islands). Lectotype BMNH 1961460 selected by Kay (1965: 15, pi. 2, figs. 13, 14). monilifera, Pleurotoma 1869d, AJC 5: 68 (Oahu). Not located in ANSP. monilifera. Turns 1861a, PZS 28: 398 (Sandwich Islands). Holotype BMNH 1961196, teste Kay (1965: 39, pi. 5, figs. 17, 18; pi. 8, fig. 7). moussoni, Omphalotropis 1869a, JdeC 17: 147. New name for Omphalotropis ovata Mousson 1865 non Pease 1861. "?" multicolor, Helicina 1865b, PZS for 1864: 676 [nomen nudum]. multicostatus, Fossar 1861a, PZS 28: 398 (Sandwich Islands). Lectotype BMNH 1962790 selected by Kay (1965: 38, pi. 6, figs. 1, 2). multiplicata, Mitroidea 1865d, PZS for 1865: 514 (Islands of the Central Pacific). Not located in BMNH by Kay (1965: 86) or elsewhere by Cernohorsky (1976: 471). multistriata 'Pease' Sowerby, Collonia 1886, Thesaurus Conchyliorum 5: 212 [nomen nudum]. multistriatus 'Pease' Paetel Turbo 1888, Catalog der Conchylien-Sammlung (Berlin) 1: 537 [nomen nudum]. 18 Bulletin Museum of Comparative Zoology, Vol. 154, No. 1 nebulosa, Borsonia 1860b, PZS 28: 143 (Sandwich Islands). Not located in BMNH by Kay (1965: 85). nebulosa, Omphalotropis Plate 4, Figure 4 1872b, AJC 7: 197 ([Makela = San Cristobal Island] Solomon Islands). Lectotype, here selected, MCZ 72347 labeled as holotype; paralectotype MCZ 72348. neglecta Nentina 1861b, PZS 28: 435 (Sandwich Islands). Lectotype BMNH 1961186 selected by Kay (1965: 56, pi. 4, figs. 5, 6). neglectus, Cotius 1861a, PZS 28: 398 (Sandwich Islands); 1871b, JdeC 19: 99 is a variety of Conus flaindus Lamarck; not located in BMNH by Kay (1965: 85). neivcombii Tease' Brot, Melauia 1872, Materiaux . . . des Melaniens III: 43 [nomen nu- dum]. Listed as a synonym of M. kauaiensis Pease, non neu'combii Lea 1866. newcombii, Mitra 1869d, AJC 5: 69 (Oahu). Not located in ANSP. nigra, Philinopsis "l860a, PZS 28: 22 (Sandwich Islands). Not located in BMNH by Kay (1965: 84). nigricans, Mitra 1865d, PZS for 1865: 514 (Islands of the Central Pacific [Marquesas Islands]). Holotype BMNH 1964312 figured by Kay (1965: 82, pi. 14, figs. 11, 12); paratypes MCZ 260604. nigritella. Helix 1865b, PZS for 1864: 675 [nomen nudum]. ntgropunctata, Hamwea 'l868e, AJC 4: 71, pi. 7, fig. 1 (Raiatea). Probable syntypes MCZ 297876 though labeled from Tahiti. nitens. Vertigo Plate 2, Figure 21 1861b, PZS 28: 439 (Ebon Island). Lectotype, here selected, MCZ 151650; paralectotypes MCZ 298456. nitida, Hydrocena 1865b," PZS for 1864: 674 (Islands of the Central Pacific); 1869b, JdeC 17: 165, pi. 7, fig. 11 as/lss/m/wea (Huaheine). Holotype in MNHN, teste Fischer-Piette (1950: 72); paratypes MCZ 139120. nitida, Tornatellina Plate 4, Figure 7 1861c, PZS 28: 439 (Ebon Island). Lectotype, here selected, MCZ 28921; paralectotypes MCZ 28922 and 302556. nitidula, Mucronalia 1861b, PZS 28: 437 (Sandwich Islands). Lectotype BMNH 1962835 selected by Kay (1965: 65, pi. 11, fig. 3); paralectotype MCZ 31711 figured by Pilsbry (1917: 226, fig. 12c); paralectotype MCZ 248842. nodicostata, Engina Plate 7, Figure 25 1868d, AJC 3: 274, pi. 23, fig. 8 (Paumotus). Lecto- type, here selected, MCZ 260614; paralectotype MCZ 260617; "type" ANSP 34543 with the note, "match- es the description but not the figure." nodifera, Drillia 1860b, PZS 28: 145 ([Haena, Hawaii] Sandwich Is- lands). Lectotype BMNH 1961210 selected by Kay (1965: 24, pi. 5, figs. 1, 2); paralectotype MCZ 49982. nodulosa, Drillia 1863d, PZS for 1862: 279 ([Sandwich Islands]). Ho- lotype BMNH 1964308 selected by Kay (1965: 80, pi. 14, figs. 9, 10); paratypes MCZ 49987. nodulosa, Engina 1869d, AJC 5: 71, pi. 8, fig. 11 (Ebon Island). Lec- totype, here selected, ANSP 34513 is the figured type. nodulosa, Turricula (Pusia) Plate 10, Figure 24 1868a, AJC 3: 214, pi. 15, fig. 5 (Paumotus). Lecto- type, here selected, ANSP 28713; four paralecto- types ANSP 391045; paralectotypes MCZ 260611. normalis. Helix Plate 3, Figure 9 1865a, PZS for 1864: 669 (Islands of the Central Pacific); 1871d, PZS for 1871: 475 (Moorea). Lecto- type, here selected, MCZ 11543; paralectotypes MCZ 11544. nucea, Nassa 1869d, AJC 7: 70, pi. 8, fig. 7 (Caroline Islands). Not located in ANSP. nucleola, Doris 1860a, PZS 28: 29 (Sandwich Islands). Redescription and taxonomic reappraisal by Brodie and VVillan (1993: 124-133). nucleola 'Pease' Garrett, Partula 1884, JANSP (2) 9: 72 ([Moorea]). Lectotype ANSP 59531a selected by Baker (1963: 205); paralectotypes MCZ 24897, 24937, and 50854. oahuensis 'Pease' Brot, Melania 1872, Materiaux . . . des Melaniens IIL- 43, pi. 3, fig. 2 (Oahu). Type should be in Musee d'Histoire Na- turelle de Geneve [not confirmed]. obconica. Helix 1865a, PZS for 1864: 669 (Islands of the Central Pacific [Raiatea, teste Garrett, 1884: 22, pi. 3, figs. 37, 37a, b]); figured specimen was not located in ANSP by Baker (1964). Two syntypes ANSP 49337 from Pease; not located in BMNH, teste Mordan, personal communication. ohesa, Partula 1868b, AJC 3: 223, pi. 15, fig. 12 (Polynesia). Ho- lotype, only specimen, not located in ANSP by Ba- ker (1963: 205). obliqua, Nassa Plate 8, Figure 6 1865d, PZS for 1865: 513 (Islands of the Central Pacific). Lectotype, here selected, MCZ 228823 la- beled as probable measured holotype by Kay; not located in BMNH by Kay (1965: 86) or elsewhere by Cernohorsky (1984: 76). oblonga, Tornatellina Plate 4, Figure 6 1865b, PZS for 1864: 673 (Islands of the Central Pacific); 1871d, PZS for 1871: 473 (Tahiti). Lecto- type, here selected, MCZ 154941; paralectotypes MCZ 297947. oblonga, Tugalia 1861b, PZS 28: 437 (Sandwich Islands). Lectotype BMNH 1962831 selected by Kay (1965: 64, pi. 11, figs. 4, 5); paralectotypes MCZ 298471. oceanica, Helicina Plate 5, Figure 9 1868b, AJC 3: 226 (Kingsmill [Islands]). Lectotype, here selected, MCZ 176561; paralectotypes MCZ 298458; not located in ANSP by Baker (1964: 162). MoLLUSKS Described by W. H. Pease • Johnson 19 ochrostoma, Realia 1865e, AJC 1: 287 (Polynesia); Tryon, 1866a, AJC 2: 82, pi. 5, fig. 1 (Hervey Isles); Pease, 1869a, JdeC 17: 147 as Omphalotropns. Holotype ANSP 13322a, teste Baker (1964: 178); paratypes MCZ 176573, 187848, and 187849. olwacea, Carclia Plate 2, Figure 14 1866b, AJC 2: 293 (Sandwich Islands); 1871d, PZS for 1871: 473 (Kauai). Holotype MCZ 57114, only specimen, teste Pease (1870c: 18: 402); two speci- mens subsequently identified by Pease [idiotypes] MCZ 23343 figured by Pilsbry and Cooke (1914: 16, pi. 9, figs. 11, 15). olivacea, Dolahrifera 1860a, PZS 28: 22 (Sandwich Islands). Holotype BMNH 1964376, teste Kay (1965: 13 [not figured]). opprcssa, Trochomorpha mgritella Plate 2, Figure 6 1870c, JdeC 18: 400 (Ponape); 1871d, PZS for 1871: 457. Lectotype, here selected, MCZ 12187; paralec- totypes MCZ 12181 and 298478. ordmata 'Pease' Paetel, Mitra 1887, Cat alogderConchiflien-Sammlung{BeThn) 1: 184 [nomen nudum]. ornata, Citharopsis Plate 6, Figure 12 1868g, AJC 4: 97, pi. 11, fig. 19 (Tahiti). Lectotype, here selected, ANSP 16919; paralectotype ANSP 391046; paralectotypes MCZ 49994. oryza, Marginella "l860b, PZS 28: 147 (Sandwich Islands); 1871e, AJC 7: 22 non Lamarck 1822. Changed to: Marginella debilis; not located in BMNH by Kay (1965: 85). oryza, Triphoris Plate 10, Figure 11 "l871a, PZS for 1870: 776 (Kauai). Lectotype, here selected, MCZ 50072. oualanensis. Helix Plate 3, Figure 8 1866b, AJC 2: 289, pi. 21, fig. 1 (Oualan [Island]). Lectotype, here selected, ANSP 47763; four para- lectotypes ANSP 391059. oualanensis 'Pease' Tryon, Melania Plate 4, Figure 20 1866b, AJC 2: 299, pi. 20, fig. 4 (Oualan [Island]). Lectotype ANSP 26274 selected by Baker (1964: 191); paralectotypes MCZ 89614. ovalis, Hammea Plate 4, Figure 23 1868e, AJC 4: 71, pi. 7, fig. 2 (Tahiti). Lectotype, here selected, MCZ 297877; paralectotypes MCZ 303195. ovalis, Partula 1865a, PZS for 1864: 675 [nomen nudum]; 1866a, AJC 2: 194 (Raiatea). Holotype ANSP 59448a, teste Baker (1963: 205) figured by Pilsbry (1909: 172, pi. 19, fig. 7); paratypes MCZ 25007. ovata, Engina Plate 7, Figure 4 1865d, PZS for 1865: 513 (Islands of the Central Pacific); 1868d, AJC 3: 274, pi. 23, fig. 6 (Howland Island). Lectotype, here selected, ANSP 34536 is the figured type; paralectotypes MCZ 260598 and 260599; not located in BMNH by Kay (1965: 86). ovata, Hydrocena 1865a; PZS for 1864: 674 (Mangiers [error for Man- gaia. Cook Islands]); 1869a, JdeC 17: 148 as Omphalo- tropis. Not located in BMNH, teste Mordan (personal communication) or mentioned as in MNHN by Fi- scher-Piette (1950: 72). ovata, Mucronalia 1861b, PZS 28: 437 (Sandwich Islands). Lectotype BMNH 1962810 selected by Kay (1965: 65, pi. 11, fig. 2). ovata, Succinea 1865a, PZS for 1864: 675 [nomen nudum]; 1868b, AJC 3: 227 "is Succinea papillata Pfeiffer." pachystoma, Labiella 1869b, JdeC 17: 171 (Kauai). Measured holotype in MNHN, teste Fischer-Piette (1950: 73, fig. 52); para- types MCZ 45181. pacifica, Helicina Plate 2, Figure 3 1865e, AJC 1: 291 (Polynesia); Tryon, 1866a, AJC 2: 82, pi. 5, fig. 7 (Oualan Island). Lectotype ANSP 14443 selected by Baker (1964: 162); paralectotype ANSP 358505. Redescribed by Pease. See under: flavescens, Helicina. pacifica, Marginella Plate 7, Figure 22 1868d, AJC 3: 280, pi. 23, fig. 20 (Paumotus). Lec- totype, here selected, ANSP 29415; paralectotype ANSP 391023. pacifica, Pedicularia 1865d, PZS for 1865: 516 (Islands of the Central Pacific); 1868g, AJC 4: 96, pi. 11, figs. 17, 18 (Apaian Island). Lectotype BMNH 1964312 selected by Kay (1965: 84, pi. 14, figs. 13, 14); paralectotypes MCZ 297958. pacifica. Truncal ella Plate 10, Figure 19 1868b, AJC 3: 230, pi. 15, fig. 27 (Oualan [Island]). Lectotype, here selected, MCZ 59799; two paralec- totypes ANSP 161791 ex MCZ. pallens, Trochomorpha trochiformis Plate 2, Figure 5 1870c, JdeC 18: 399 (Tahiti; Moorea); 1871d, PZS for 1871: 457 (Tahaa). Lectotype, here selected, MCZ 12188; paralectotypes MCZ 12188; both from Mo- orea. pallida 'Pease' Pace, Columbella 1902, Proceedings of the Malacological Society of London 5: 118 [nomen nudum]. pallida, Mitra 1860b, PZS 28: 146 (Sandwich Islands). Not located in BMNH by Kay (1965: 85); type lost, teste Cer- nohorsky (1976: 292). pallida, Taheitea [sic] Plate 9, Figure 17 1868b, AJC 3: 229 (Tahiti; Huaheine). Lectotype, here selected, ANSP 12659 is the measured type specimen [as 13559a], teste Baker (1964: 175), no locality on label; four paralectotypes ANSP 391035; paralectotypes MCZ 178656 from Tahiti and 178660 from Huaheine. pallidus, Triphoris Plate 10, Figure 8 1871a, PZS for 1870: 774 (Kauai). Lectotype, here selected, MCZ 50073; paralectotypes MCZ 288955. parva, Engina 1868d,'AJC 3: 276, pi. 23, fig. 11 (Paumotus). Lec- totype, here selected, ANSP 34542 is the syntype 20 Bulletin Museum of Comparative Zoology, Vol. 154, No. 1 figured by Cernohorsky (1987: 100, figs. 17, 18); paralectotypes MCZ 49995. pari'a, Pterocyclos Plate 5, Figure 10 1865e, AJC 1: 290 (Polynesia); Tryon, 1866a, AJC 2: 82, pi. 5, fig. 8 (Hervey Isles); Pease, 1868j, AJC 4: 158 (Aitutake, Hervey Isles). Lectotype ANSP 13406 selected by Baker (1964: 178); two paralectotypes ANSP 372687; paralectotypes MCZ 141027, 141028, and 141029. parvidens, Helix 1861d, PZS for 1861: 243 (Tahiti). Lectotype BPBM 170888 selected by Solem (1976: 171, figs. 77a, b); paralectotypes MCZ 17267. parvula, Helicina 1868), AJC 4: 156, pi. 12, fig. 10 (Atiu [Island, Cook Islands]). Holotype ANSP 14444a, teste Baker (1964: 162); paratypes MCZ 297920. parvum, Cyclostoma Plate 2, Figure 15 1865a, PZS for 1864: 674 (Islands of the Central Pacific). Lectotype, here selected, MCZ 74949; para- lectotypes MCZ 187914; both from Tahiti. parvus, Conus 1868h, AJC 4: 126. New name for Conus fusiformis Pease 1861 non Fischer 1801 non Lamarck 1810. See under: fusiformis, Conus. paucicostata, Cithara [Ci/thara] Plate 6, Figure 8 1868a, AJC 3: 217 (Tahiti). Lectotype, here selected, MCZ 231925. paucicostata, Clathurella 1860b, PZS 28: 144 (Sandwich Islands). Holotype BMNH 1961158, teste Kay (1965: 22, pi. 2, figs. 11, 12). paucicostata, Pithys Plate 2, Figure 7 1870c, JdeC 18: 395 (Kauai). Lectotype, here select- ed, MCZ 17271; paralectotypes MCZ 298476; not mentioned by Fischer-Piette (1950: 76). paumotensis, Marginella Plate 7, Figure 23 1868d, AJC 3: 281, pi. 23, fig. 22 (Paumotus). Lec- totype, here selected, ANSP 29497; paralectotype ANSP 391050; paralectotypes MCZ 297943. paumotensis, Scalaria Plate 9, Figure 16 1868d, AJC 3: 289, pi. 24, fig. 11 (Paumotus). Lec- totype ANSP 19581 selected by DuShane (1988 (7): 5, fig. [not numbered]) and refigured (1990: 9, fig. 34); two paralectotypes ANSP 352473; paralecto- types MCZ 187394. paxillum, Cerithium 1861b, PZS 28: 433 (Sandwich Islands). Lectotype BMNH 1964804 selected by Kay (1965: 47, pi. 10, fig. 6). peasei Martens and Langkavel, Columbella {Seminella) 1871, Donum Bismarckianum, p. 23. New name for Cythara varia Pease 1860 non Columbella varia Sow- erby 1832. peasei Reeve, Marginella 1865, Conchologia Iconica 15: Marginella, pi. 21, fig. 108. New name for Marginella cylindrica non Sow- erby 1846 and M. polita Pease non Carpenter 1857. See under: cylindrica, Marginella. peasei Tryon, Triphoris 1872d, AJC 7: 206. New name for T. gracilis Pease, 1871: 777 non Pease, 1871: 774. See under: gracilis, Triphoris. pellucida, Columbella 1861a, PZS 28: 399 (Sandwich Islands). Lectotype BMNH 1962794 selected by Kay (1965: 41, pi. 10, fig. 7). Is Columbella rorida Reeve, teste Pease (1868h: AJC 4: 122). pellucida, Partula 1871d, PZS for 1871: 457 (Guadalcanal, Solomon Islands). Though not a type, MCZ 94837 from "Ysa- bel [Island] Solomon Is. (coll. Cox)" was figured by Hartman (1886: 35, pi. 2, fig. 17) as this species and by Pilsbry (1909: 297, pi. 36, fig. 6). The original lot, in the Hartman collection at the Carnegie Mu- seum 6246, consisted of one adult and two imma- ture examples. The former and one of the latter were transferred to the MCZ in 1935; not located in BMNH, teste Mordan (personal communication). pellucidus, Pleurobranchus 1860b, PZS 28: 145 (Sandwich Islands). Not located in BMNH by Kay (1965: 84). perfectus, Triphoris Plate 10, Figure 15 1871a, PZS for 1870: 775 (Kauai). Lectotype, here selected, MCZ 302553; paralectotype MCZ 50075. perlonga. Vertigo 1871d, PZS for 1871: 462 ([Nuuanu] Oahu). Holo- type MCZ 48063 figured by Pilsbry and Cooke (1920: 258-259, pi. 23, figs. 1, 2). perplexa 'Pease' H. H. Smith, Partula 1902, ACM 1: 463 (Huaheine). Syntypes MCZ 25358. perplexa, Scalaria Plate 9, Figure 1 1 1868d, AJC 3: 288 (Hawaii [Island]). Lectotype [so labeled], here selected, MCZ 181978; paralectotypes MCZ 181979 and 181980. perversa 'Pease' H. H. Smith, Partula 1902, ACM 1: 442 (Tahiti). Syntypes MCZ 25317. picea, Strigatella 1860b, VZS 28: 146 (Sandwich Islands). Holotype BMNH 1962772 figured by Kay (1965: 28, pi. 9, fig. 9). picta, Collonia Plate 6, Figure 21 1868g, AJC 4: 91, pi. 11, figs. 2, 3 ([Anaa Island] Paumotus). Lectotype, here selected, MCZ 245268. Labeled, "Chosen [lectotype] by P. A. Maxwell, ms in prep. 1963" but apparently never published; paralectotypes MCZ 288001 and 288002; not located in ANSP. picta, Helicina 1865b, PZS for 1864: 676 [nomen nudum]. MCZ 74428. plana, Libratula Plate 5, Figure 2 1865d, PZS for 1865: 512 (Islands of the Central Pacific). Lectotype, here selected, MCZ 119162. planilabrum, Partula 1865a, PZS for 1864: 671 (Islands of the Central Pacific [Haamene Valley, east coast of Tahaa, teste Garrett, 1884: 63]). Syntypes MCZ 25004 and 25298. plicatula, Turricula (Costellaria) Plate 10, Figure 27 1868a, AJC 3: 213, pi. 15, fig. 4 (Paumotus). Lecto- type, here selected, ANSP 28845; four paralecto- types ANSP 391044; paralectotypes MCZ 260602. MoLLUSKS Described by W. H. Pease • Johnson 21 plumbea, Plaiiaxis 1861d, PZS for 1861: 244 (Sandwich Islands). Lec- totype BMNH 1964280 selected by Kay (1965: 71, pi. 13, figs. 11, 12); paralectotypes MCZ 39062, 187835, and 207329. polita, Marginella 1868d, AJC 3: 280, pi. 23, fig. 19 (Tarawa Island). New name for Margmella cylindracea [sic] Pease 1863 non Sowerby 1846 non Carpenter 1857. See under: cylindrka, Marginella. polita, Odostowta Plate 8, Figure 10 1868d, AJC 3: 291, pi. 24, fig. 17 (Tahiti). Lectotype, here selected, MCZ 10562; paralectotype MCZ 298903; not located in ANSP. poryhyrostoma, A mastra 1869c, JdeC 17: 172 (Oahu; Coll. [of] Pease). Two syntypes MCZ 45256, one of which was figured by Hyatt and Pilsbry (1911: 226, pi. 37, fig. 13). producta, Clathurella 1860b, PZS 28: 143 (Sandwich Islands). Lectotype BMNH 1962761 selected by Kay (1965: 19, pi. 2, figs. 3, 4); paralectotypes MCZ 49974. producta, Omphalotropis 1869a, JdeC 17: 151, pi. 7, fig. 8 (Tahaa); 1871d, PZS for 1871: 471 as Atropis (Raiatea; Tahaa). Holotype in MNHN, teste Fischer-Piette (1950: 72); paratypes MCZ 74932, 74933, and 187855 all from Tahaa. producta, Partula 1865a, PZS for 1864: 671 (Islands of the Central Pacific); 1871d, PZS for 1871: 473 (Tahiti). Syntypes MCZ 25016. producta, Ranella 1861a, PZS 28: 397 (Sandwich Islands). Holotype BMNH 1961157, teste Kay (1965: 37, pi. 6, figs. 17, 18). producta, Realia 1865b, PZS for 1864: 673 (Islands of the Central Pacific); 1869a, JdeC 17: 151, pi. 7, fig. 8 as Om- phalotropis (Tahaa); 1871d, PZS for 1871: 476 (Ra- iatea) as Atropis. Holotype in MNHN, teste Fischer- Piette (1950: 72); paratypes MCZ 74943, 187855, and 187856. propinqua 'Pease' Hartman, Partula 1881, BMCZ 9: 185 (Tahaa). Syntypes MCZ 25124 and 25308. propinqua, Terebra 1869d, AJC 5: 66 (Hawaii [Island]). Not located in ANSP. prostrata. Helix 1865a, PZS for 1864: 670 (Islands of the Central Pacific); 1871d, PZS for 1871: 475 ("?" Lanai). Syn- types MCZ 17343 "are Planorbis opercularis Gould from the West Indies," teste Pilsbry and Cooke (1922: 117). protea 'Pease' Schmeltz, Partula 1865a, PZS for 1864: 675 [nomeyi nudum]; 1874, Mu- seum Godeffroy (Hamburg). Catalog 5: 92 (Raiatea) [nomen nudum]. MCZ 24996 and 25349. proximus, Helicter Plate 4, Figure 1 1862a, PZS for 1862: 6 (Molokai). Lectotype, here selected, MCZ 25823. pudica, Mitra 1860b, PZS 28: 146 (Sandwich Islands). Lectotype BMNH 1961190 selected by Kay (1965: 27, pi. 3, figs. 11, 12). pulchella, Clathurella 1860b, PZS 28: 144 (Sandwich Islands). Lectotype BMNH 1962768 selected by Kay (1965: 22, pi. 2, figs. 19, 20); paralectotypes MCZ 50010. pulchra, Auriculella 1868f, JdeC 16: 346, pi. 14, fig. 6 ([Oahu]). Holotype and four paratypes in MNHN, teste Fischer-Piette (1950: 71); paratypes MCZ 161609. pulchra, Partula varia 1871d, PZS for 1871: 473 (Huaheine); Schmeltz, 1874, Museum Godeffroy (Hamburg). Catalog 5: 92; [both nom- ma nuda]. MCZ 24878 and 24920. punctata, Syphonota 1868e, AJC 4: 77, pi. 9, fig. 2 (Huaheine). Probable syntype MCZ 297868 though labeled as from Tahiti. punctatus, Triphoris 1871a, PZS for 1870: 775 (Annaa Island). Not located in BMNH, teste Way (personal communication). purus, Conus Plate 6, Figure 23 1863d, PZS for 1862: 279 (Pacific Islands); 1871c, JdeC 19: 98 (Niihau Island). Holotype MCZ 72331, only specimen; not located in BMNH by Kay (1965: 89). pusilla, Columbella 1863b, PZS for 1862: 244 (Kingsmill Islands); 1868h, AJC 4: 1 22 as Columbella fusiformis to replace C. pusilla non Sowerby 1844. Lectotype BMNH 1964302 se- lected by Kay (1965: 78, pi. 13, figs. 9, 10); paralec- totypes MCZ 136617. pusilla, Cythara 1860b,' PZS 28: 147 (Sandwich Islands). Lectotype BMNH 1962784 selected by Kay (1965: 33, pi. 10, fig. 14); paralectotypes MCZ 50003. pusilla, Distorsio 1861a, PZS 28: 397. Holotype BMNH 1961155, teste Kay (1965: 37, pi. 3, figs. 15, 16). pusilla, Haminea 1860a, PZS 28: 20 (Sandwich Islands). Lectotype BMNH 1962754 selected by Kay (1965: 9, pi. 9, fig. 1); paralectotypes MCZ 297879. pusilla, Triton 1861b, PZS 28: 434 (Sandwich Islands). Lectotype BMNH 1962818 selected by Kay (1965: 55, pi. 5, figs. 19, 20); paralectotypes MCZ 297942. pustulosus, Triphoris Plate 10, Figure 2 1871a, PZS for 1870: 776 (Kauai). Lectotype, here selected, MCZ 50077. putillus, Turricula 1865d, PZS for 1865: 514 (Central Pacific); 1868a, AJC 3: 214, pi. 15, fig. 24. Holotype BMNH 1964311, teste Kay (1965: 81, pi. 14, figs. 7, 8); paratypes MCZ 260605. pyriformis, Marginella Plate 7, Figure 21 ' 1868d, AJC 3: 280, pi. 23, fig. 21 (Paumotus). Lec- totype, here selected, ANSP 29541; four paralec- totypes ANSP 391061; paralectotype MCZ 24968. 22 Bulletin Museum of Comparative Zoology, Vol. 154, No. 1 pi/riformis, Voh^atella 1868e, AJC 4: 73, pi. 7, fig. 5 (Huaheine). Syntypes MCZ 297944, mostly fragments. radiata, Aviciila Plate 1, Figure 4 1863c, PZS for 1862: 244 (Kingsmill Islands). Lec- totype, here selected, MCZ 298466; paralectotypes MCZ 297882; not located in BMNH by Kay (1965: 86). radiata Tease' Garrett, Partula 1884, JANSP (2) 9: 74, pi. 3, fig. 45 (Hamoa Valley, east coast of Raiatea Island, Society Islands). Lec- totype ANSP 59409 selected by Baker (1963: 205) figured by Pilsbry (1909: 232, pi. 18, fig. 5); para- lectotypes MCZ 24957. radiata, Tectura 1861b, PZS 28: 437 (Sandwich Islands). Lectotype BMNH 1962837 selected by Kay (1965: 66, pi. 11, figs. 6, 7). recta, Partula 1865a, PZS for 1864: 675 [nomen nudum]; 1868j, AJC 4: 155, pi. 12, figs. 8, 9 ([Nukahiva] Marquesas [Is- lands]). Holotype ANSP 59789a fig. 8, teste Baker (1963: 205); paratypes MCZ 25338 and 25343. retkulatus, Pleurobranchus 1860a, PZS 28: 25 (Sandwich Islands); 1864, PZS for 1863: 510 is Pleurobranchus violaceus Pease 1864, new name for P. reticulatus Pease 1860 non Rang 1832 non Kelaart 1859; not located in BMNH by Kay (1965: 84). retunsa. Helix 1865b, PZS for 1864: 670 (Pacific Islands); 1871d, PZS for 1871: 475 (Tahiti). Lectotype BPBM 170913 selected by Solem (1976: 412, figs. 178d-f); paralec- totypes MCZ 17224. robusta, Hi/drocaena 1865b, PZS for 1864: 676 [twmen nudum]. robusta, Omphalotropis 1869a, JdeC 17: 148, pi. 7, fig. 3 ([Raiatea]). Holotype in MNHN, teste Fischer-Piette (1950: 72); paratypes MCZ 74933 and 187853. robusta, 'Pease' H. H. Smith, Partula 1865b, PZS for 1864: 675 [nomen nudum]; 1902, ACM 1: 436 (Raiatea). Lectotype ANSP 59444a selected by Baker (1963: 205) figured by Pilsbry (1909: 248, pi. 18, fig. 14); paralectotypes MCZ 3651. robustus, Stylifer 1861b, PZS 28: 437 (Sandwich Islands). Not located in BMNH by Kay (1965: 85). robustus, Triphoris Plate 10, Figure 12 1871a, PZS for 1870: 775 (Makaimo[Makemo Island, Taumotu Archipelago]). Lectotype, here selected, MCZ 73923; paralectotypes MCZ 298499. roratongensis, Pithys 1870c, JdeC 18: 395 (Roratonga); 1871d, PZS for 1871: 457. Lectotype BPBM 170885 selected by So- lem (1976: 174, figs. 77e, i) who changed it to rara- tongensis; paralectotypes MCZ 17345. rosacea, Gena Plate 6, Figure 1 1868d, AJC 3: 284, pi. 24, fig. 1 (Paumotus). Lecto- type, here selected, ANSP 40756; two paralecto- types ANSP 391033; paralectotypes MCZ 297918. rosacea, Odostomia Plate 8, Figure 11 1868d, AJC 3: 292, pi. 24, fig. 19 (Paumotus). Lec- totype, here selected, ANSP 19959; paralectotypes MCZ 391054. rosacea, Terebra 1869d, AJC 5: 65 (Oahu). Not located in ANSP. rosea, Mucronalia 1861b, PZS 28: 437 (Sandwich Islands). Holotype BMNH 1962809, teste Kay (1965: 64, pi. 11, fig. 1). rotella, Diadema 1868), AJC 4: 158, pi. 12, fig. 13 (Atiu [Island, Cook Islands]). Holotype ANSP 13409a, teste Baker (1964: 178). rotellina, Pithys 1870c, JdeC 18: 393 (Aitutake [Island, Cook Is- lands]); 1871d, PZS for 1871: 453. Lectotype BPBM 2312 selected by Solem (1976: 139, figs. 62e, f); para- lectotypes MCZ 176559. rubella, Succinea Plate 4, Figure 13 1871d, PZS for 1871: 460 (Lanai). Lectotype, here selected, MCZ 161671; paralectotypes MCZ 298475. rubicunda, Helicina maugeriae Plate 5, Figure 12 1865a, PZS for 1864: 676 [nomen nudum]; 1868d, AJC 3: 227: 1871d, PZS for 1871: 466 (Raiatea). Lectotype ANSP 14512 selected by Baker (1964: 162) labeled as from Tahiti; paralectotypes MCZ 314014 from Raiatea. rubida, Catinella Plate 4, Figure 21 1870a, JdeC 18: 97 (Kauai). Lectotype, here selected, MCZ 45252, probable measured type. rubida, Dentiora Plate 7, Figure 7 1863b, PZS for 1862: 240 (Sandwich Islands). Lec- totype, here selected, MCZ 297951, probable mea- sured type; not located in BMNH by Kay (1965: 86). rubida, Neritina 1865d, PZS for 1865: 514 (Islands of the Central Pacific); 1868d, AJC 3: 285, pi. 24, fig. 5 (Tahiti). Lectotype BMNH 1964313 selected by Kay (1965: 82, pi. 14, figs. 15, 16); paralectotypes MCZ 89902; Baker (1964: 160) questions if ANSP 37675a is the figured type since the measurements are not close to those of Pease. rubra, Alcyna 1861b, PZS 28: 436 (Sandwich Islands). Lectotype BMNH 1962828 selected by Kay (1965: 62, pi. 7, figs. 5, 6); paralectotypes MCZ 31720 and 205584. rubra, Odostomia Plate 8, Figure 12 1868d, AJC 3: 291, pi. 24, fig. 18 ([Anaa Island] Pau- motus). Lectotype, here selected, ANSP 19955; para- lectotypes ANSP 391054; paralectotype MCZ 10488. rudis, Neritina 1868d, AJC 3: 285, pi. 24, fig. 4 (Ponape). Holotype ANSP 37543a, teste Baker (1964: 160); paratypes MCZ 73518. rufescens, Helicina 1865a, PZS for 1864: 676 [nomen nudum]. MCZ 297926. rufus, Pleurobranchus 1860a, PZS 28: 25 (Sandwich Islands). Syntype MCZ 297872; not located in BMNH by Kay (1965: 84). rugata. Helix Plate 3, Figure 3 'l866b, AJC 2: 291 (Sandwich Islands); 1871d, PZS for 1871: 474 as Endodonta (Maui). Lectotype, here MoLLUSKS Described by W. H. Pease • Johnson 23 selected, MCZ 17237; paralectotypes MCZ 298479; not located in ANSP by Baker (1963: 233) or else- where by Solem (1976: 377). rugulosa, Amastra \870a, JdeC 18: 95 (Kauai); Crosse, 1876, JdeC 24: 99, pi. 1, figs. 4, 4a. Holotype and fragment of para- type in MNHN, teste Fischer-Piette (1950: 149 as second reference to A. A. rugulosa only); paratypes MCZ 45255. rugulosa, Hclicma Plate 5, Figure 7 1868), AJC 4: 157, pi. 12, fig. 11 (Tahaa). Lectotype, here selected, MCZ 297922; paralectotypes MCZ 298467; not located in ANSP by Baker (1964: 162). rugulosuni, Sistrum Plate 9, Figure 8 'l868g, AJC 4: 93, pi. 11, fig. 7 (Howland [Island]). Lectotype, here selected, ANSP 36740; two para- lectotypes ANSP 391042; paralectotypes MCZ 295581. rustica, Partula 1865a, PZS for 1864: 675 [uomen nudum]; 1866a, AJC 2: 199 (Tahitian Archipelago); 1867a, AJC 3: 81, pi. 1, fig. 5; 1871d, PZS for 1871: 473 (Raiatea). Holo- type ANSP 59480 selected as lectotype by Baker (1963: 205); paralectotypes MCZ 25140 and 25302. rutella, Succinea 1865b, PZS for 1864: 675[nomen nudum]. MCZ 155141. sagitta Tease' Pace, Columbella 1902, Proceedings of the Malacological Society of London 5: 132 [nomen nudum]. salt at a, Mitra Plate 7, Figure 18 1865d, PZS for 1865: 512 (Islands of the Central Pacific). Lectotype MCZ 260613 [not 260605] se- lected by Cernohorsky (1976: 501, pi. 450, fig. 2); two paralectotypes MCZ 298462; not located in BMNH by Kay (1965: 86). sandivicensis, Dapbnella Plate 7, Figure 6 1860b, PZS 28: 148 (Sandwich Islands). Lectotype, here selected, MCZ 49993; paralectotype MCZ 298491; not located in BMNH by Kay (1965: 85). sandivicensis, Erato 1860b, PZS 28: 146 (Sandwich Islands). Lectotype BMNH 1962776 selected by Kay (1965: 31, pi. 9, fig. 10); paralectotypes MCZ 297959. sandwicensis, Marginella 1860b, PZS 28: 146 (Sandwich Islands). Lectotype BMNH 1962778 selected by Kay (1965: 31, pi. 9, fig. 11); paralectotype MCZ 24970. sandivicensis, Oliva 1860b, PZS 28: 145 (Sandwich Islands). Lectotype BMNH 1961188 selected by Kay (1965: 25, pi. 3, figs. 9, 10). sandwicensis, Patella Plate 1, Figure 8 1861b, PZS 28: 437 (Sandwich Islands). Lectotype, here selected, MCZ 304058; paralectotypes MCZ 38803 and 150789; not located in BMNH by Kay (1965: 85). sandivicensis, Pedipes 1860b, PZS 28: 146 (Sandwich Islands). Holotype BMNH 1962775 figured by Kay (1965: 30, pi. 9, fig. 7); paratypes MCZ 74813. sandivicensis, Tornatina 1860a, PZS 28: 19 (Sandwich Islands). Lectotype BMNH 1962751 selected by Kay (1965: 6, pi. 9, fig. 4); paralectotypes MCZ 31712 with the note by Pils- bry "seems to be two species, but they are so worn I cannot be sure." sandwicensis, Turbo 1861b, PZS 28: 436 (Sandwich Islands). Holotype BMNH 1961179, teste Kay (1965: 60, pi. 7, figs 7, 8). sandwichensis, Columbella Plate 6, Figure 22 1861d, PZS for 1861: 244 (Sandwich Islands); 1868h, AJC 4: 122 is Columbella turtunna Lamarck (1822). Lectotype, here selected, MCZ 297950; paralecto- types MCZ 298472; not located in BMNH by Kay (1965: 86). sandwichensis 'Pease' Nevill, Fossar 1884, Hand List of Mollusca in the Indian Museum 2: 165 [nomen nudum]. MCZ 242384. sandwichensis, Vitularia 1861a, PZS 28: 397 (Sandwich Islands). Lectotype BMNH 1961182 selected by Kay (1965: 36, pi. 4, figs. 1, 2). scalariformis, "?" Cyclophorus 1865e, AJC 1: 289 (Polynesia); Tryon, 1866a, AJC 2: 82, pi. 5, fig. 6, as Pupoidea (Hervey Isles); Pease, 187 Id, PZS for 1871: 465 as Palama (Caroline Is- lands). Holotype ANSP 13360, teste Baker (1964: 165). scalariformis, Realia 1865e, AJC 1: 288 (Polynesia); Tryon, 1866a, AJC 2: 82, pi. 5, fig. 3 (Oualan Island); Pease, 1869a, JdeC 17: 159 as Scallinella (Atiu). Holotype ANSP 13362a, teste Baker (1964: 178); paratypes MCZ 74935, 176572, and 187839 all from Atiu. sculpta, Nucula 1860c, PZS 28: 189 (Corea [Korea] Sea). Not men- tioned by Kay (1965). sculptilis, Coralliobia 1865d, PZS for 1865: 513 (Islands of the Central Pacific). Not located in BMNH by Kay (1965: 86). sculptilis. Helix 1865a, PZS for 1864: 669 (Mangier [error for Man- gaia. Cook Islands]); 1867b, AJC 3: 104 as Helix fra- tercula to replace H. sculptilis 1865 non Bland 1868. Lectotype BMNH 1962705 selected by Solem (1976: 423 [not figured]); paralectotypes MCZ 17206. sculptilis, Terebra Plate 8, Figure 18 1869d, AJC 5: 64 ([Honolulu] Oahu). Lectotype MCZ 248804, teste Bratcher and Cernohorsky (1987: 40) presumed here to have been selected by them; para- lectotype MCZ 248805. sculptum, Cerithium 1869d, AJC 5: 77, pi. 8, fig. 8 (Paumotus). Holotype ANSP 17592, teste Houbrick (1992: 49, fig. 31h). scuta, Helix 1865a, PZS for 1864: 675 [nomen nudum]. sectilis, Mitra 1868d, AJC 3: 271 (Hawaii [Island]). Not located in ANSP. 24 Bulletin Museum of Comparative Zoology, Vol. 154, No. 1 semicostata, Rissoa 1868d, AJC 3: 296, pi. 24, fig. 32 (Paumotus). Type lost, label only in ANSP. semkostatus, Turbo 1861b, PZS 28: 435 (Sandwich Islands). Lectotype BMNH 1961202 selected by Kay (1965: 60, pi. 7, figs. 3, 4); paralectotypes MCZ 297920. semiplicata, Melampus (Tralia) 1860b, PZS 28: 146 (Sandwich Islands). Lectotype BMNH 1962773 selected by Kay (1965; 30, pi. 4, fig. 7); paralectotypes MCZ 176566. semiplicata, Rissoina 1863b, PZS for 1862: 242 (Pacific Islands); 1868d, AJC 3: 295, pi. 24, fig. 29 (Rowland Island). Lec- totype BMNH 1964294 selected by Kay (1965: 76, pi. 13, figs. 3, 4). semiplicata, Vanikoro 1861b, PZS 28: 435 (Sandwich Islands). Holotype BMNH 1962819, teste Kay (1965: 57, pi. 6, figs. 7, 8). semistriata, Amphiperas 1863b, PZS for 1862: 241 (Pacific Islands); 1868g, AJC 4: 96, pi. 1 1, fig. 16 (Ponape). Lectotype 1964286 selected by Kay (1965: 74, pi. 12, figs. 5, 6). Cate (1973: 112, fig. 138) apparently unaware that Kay had previously selected a lectotype for this species, or that Pease had later figured it and stated its lo- cality as Ponape Island, refigured ANSP 17042 as the holotype and restricted the type locality to: 73 m, Kii Channel, Japan (24°00'N, 134°48'E). This re- striction is therefore invalid. semistriata, Atys 1860a, PZS 28: 20 (Sandwich Islands). Holotype BMNH 1961459, teste Kay (1965: 10, pi. 1, figs. 7, 8); paratypes MCZ 31716 largest figured by Pilsbry (1917: 217, fig. 5) and 31717. serrata, Lamellina Plate 2, Figure 17 1861c, PZS 28: 439 (Ebon Island). Lectotype, here selected, MCZ 302555; paralectotypes MCZ 28926 and 298484; paralectotype ANSP figured by Pilsbry and Cooke (1915: 165, pi. 33, figs. 1, 2); not located in ANSP by Baker (1963: 199). similaris, Amastra rugulosa Plate 2, Figure 11 1870a, JdeC 18: 96 ([Waimea] Kauai). Lectotype, here selected, MCZ 45253; paralectotypes MCZ 58936 and 298498. similis, Triphons Plate 10, Figure 14 1871a, PZS for 1870: 774 ([Haena] Kauai). Lectotype, here selected, MCZ 50079. simillima, Haminea Plate 4, Figure 24 1868e, AJC 4: 72, pi. 7, fig. 3 (Tahiti). Lectotype, here selected, MCZ 297875; paralectotypes MCZ 303194. simillima, Helix 1865a, PZS for 1864: 669 (Islands of the Central Pacific [Raiatea, teste Garrett, 1884: 19, pi. 2, figs. 32, 32a, bj). Lectotype, here selected, ANSP 49288 is the specimen figured by Garrett; not located in BMNH, teste Mordan (personal communication). simplana. Vertigo 187 Id, PZS for 1871: 461 (Marquesas Islands). Not located in BMNH, teste Mordan (personal com- munication). simplex, Leptachatina 1869c, JdeC 17: 170 (Hawaii [Island]). Measured ho- lotype in MNHN, teste Fischer-Piette (1950: 73); paratype ANSP 57821 figured by Cooke (1910: 38, pi. 1, figs. 8, 9); paratypes MCZ 45176. simplex, Olivella (Callianax) Plate 8, Figure 16 1868d, AJC 3: 281, pi. 23, fig. 24 (Paumotus). Lec- totype, here selected, ANSP 28969; paralectotype ANSP 391047. simplex, Tornatellma Plate 4, Figure 8 1865a, PZS for 1864: 673 (Islands of the Central Pacific); 1871d, PZS for 1871: 473 (Tahaa). Lecto- type, here selected, MCZ 175745; paralectotypes MCZ 298473. simulans, Partula 1865a, PZS for 1864: 675 [nomen nudum]; 1866a, AJC 2: 202 ([Moorea] Tahitian Archipelago); 1867a, AJC 3: 81, pi. 1, fig. 11. Lectotype ANSP 59941 selected by Baker (1963: 205); paralectotypes MCZ 24881, 24955, and 25116. stniatralis 'Pease' Pilsbry, Partula 1909, MofC (2) 20: 185. Manuscript name under the synonymy of Partula otaheitana Bruguiere. sinistrosa 'Pease' Garrett, Partula 1865a, PZS for 1864: 675 [nomen nudum]; Garrett, 1884, JANSP (2) 9: 49 (Papieri, Tahiti). Lectotype ANSP 59548a selected by Baker (1963: 205) figured by Pilsbry (1909: 191, pi. 26, fig. 1); paralectotypes MCZ 24997 and 297864. solid a, Amastra 1869c, JdeC 17: 173 (Oahu; Coll. [of] Pease [and] Crosse). Three syntypes MCZ 23341 and 141338 fig- ured by Pilsbry and Cooke (1914: 28, 31, pi. 7, figs. 1-3); syntype in MNHN, teste Fischer-Piette (1950: 73). solida, Helictna Plate 5, Figure 3 1865a, PZS for 1864: 673 (Islands of the Central Pacific [Tahiti]). Lectotype, here selected, MCZ 297923; paralectotypes MCZ 302379. soliduscida 'Pease' Sowerby, Magilus 1872 [in] Reeve, Conchologia Iconica 18: Magilus, pi. 4, fig. 12 (Sandwich Islands). Holotype in BMNH, teste Sowerby. speciosa, Philinopsis 1860a, PZS 28: 21 (Sandwich Islands). Syntype MCZ 297874; not located in BMNH by Kay (1965: 84). sphaerica, Amastra 1870a, JdeC 18: 94 ([Waimea] Kauai); Crosse, 1876, JdeC 24: 98, pi. 1, figs. 5, 5a. Holotype and paratypes in MNHN, teste Fischer-Piette (1950: 149 as first reference to y4. A. rugulosa); paratypes MCZ 45162. squamosum, Sistrum 1868d, AJC 3: 277, pi. 23, fig. 14 (Kingsmill [Island]). Lectotype ANSP 29910 selected by Cernohorsky (1987: 97, figs. 10, 11) is the figured type; paralec- totypes MCZ 295583. squamosus, Latirus 1863b, PZS for 1862: 240 (Pacific Islands); 1868d, AJC 3: 278, pi. 23, fig. 16 (Baker Island). Lectotype MoLLUSKS Described by W. H. Pease • Johnson 25 BMNH 1964284 selected by Kay (1965: 72, pi. 12, figs. 12, 13); paralectotypes MCZ 2192 and 261183. squamulosa, Fastigiclla Plate 7, Figure 14 1868d, AJC 3: 290, pi. 24, fig. 15 ([Anaa Island] Pau- motus). Lectotype, here selected, ANSP 36702; para- lectotype ANSP 391037; paralectotypes MCZ 297956. stolicia, Partula 1866a, AJC 2: 198 (Tahitian); 1871d, PZS for 1871: 473 (Raiatea). Syntypes MCZ 25311 and 104508 la- beled as from Tahiti; not located in ANSP by Baker (1963: 205). strammca, Helicina 1865a, PZS for 1864: 676 [nomen nudum]. MCZ 176563. striata, Alcytia 1869d, AJC 5: 70 (Hawaii [Island]). Not located in ANSP. striata, Chondrella Plate 2, Figure 19 1871d, PZS for 1871: 477 (Roratonga [Raratonga Is- land, Cook Islands]). Lectotype, here selected, MCZ 187917; paralectotypes MCZ 74942. striata, Etigina 1868d, AJC 3: 275, pi. 23, fig. 10 (Paumotus). Not located in ANSP. striata, Odostomia Plate 8, Figure 14 1868d, AJC 3: 291, pi. 24, fig. 16 (Paumotus). Lec- totype, here selected, MCZ 10491; paralectotypes MCZ 303452 and ANSP 58022. striatula, Rissoina Plate 9, Figure 4 1868d, AJC 3: 296, pi. 24, fig. 31 (Paumotus). Lec- totype, here selected, MCZ 178844 labeled as ho- lotype; two paralectotypes ANSP 19254. striatula Vertigo 1871d, PZS for 1871: 461 ([Kalapana, Puna] Hawaii [Island]). Lectotype, here selected, MCZ 45239 [not 45234] figured by Pilsbry and Cooke (1926: 223, pi. 28, figs. 1, 2); paralectotypes MCZ 151647 and BPBM ex MCZ. striatum, Sistrum Plate 9, Figure 9 1868d, AJC 3: 276, pi. 23, fig. 12 (Kingsmill [Is- lands]). Lectotype, here selected, ANSP 36735; three paralectotypes ANSP 391063; paralectotypes MCZ 178941. striatus, Melampms (Tralia) 1861d, PZS for 1861: 244 (Tahiti); 1868g, AJC 4: 100, pi. 12, fig. 14. Lectotype ANSP 22356a selected by Baker (1964: 151); a lectotype BMNH 1964282 was later also selected by Kay (1965: 71, pi. 12, figs. 8, 10). strigata, Cythara Plate 6, Figure 9 1863b, PZS for 1862: 242 (Pacific Islands [Howland Island]). Lectotype, here selected, MCZ 49990; para- lectotypes MCZ 298490; not located in BMNH by Kay (1965: 86). strigata, Partula 1868J, AJC 4: 1 55, pi. 12, fig. 7 (Marquesas [Islands]). Holotype ANSP 59510a, teste Baker (1963: 205). strigata, Pisania 1863b, PZS for 1862: 241 (Pacific Islands); 1868g, AJC 4: 93, pi. 11, fig. 6 (Ponape). Lectotype BMNH 1964288 selected by Kay (1965: 73, pi. 12, figs. 16, 17). striolata. Helix Plate 2, Figure 2 1861c, PZS 28: 439 (Ebon Island). Lectotype, here selected, MCZ 11563; paralectotypes MCZ 298474. striolata, Partula 1865a, PZS for 1864: 675 [nomcti nudum]; 1866a, AJC 2: 195 (Tahitian Archipelago [Moorea]); 1867a, AJC 3: 81, pi. 1, fig. 4. Lectotype ANSP 59648a selected by Baker (1963: 205) is the figured type; paralec- totypes MCZ 24910, 25293, and 25303. subangulata, Alcyna 1861b, PZS 28: 436 (Sandwich Islands). Holotype BMNH 1962830 figured by Kay (1965: 63, pi. 9, fig. 6); paratype MCZ 31723 is now the figured holotype of Alcyna subangulata flammulata Pilsbry (1917: 213, pi. 15," figs. 5, 6). subangulata, Partula faba 1870c, JdeC 18: 401 (Tahaa); 1871d, PZS for 1871: 458. Syntypes MCZ 24948. subpellucida, Eulima Plate 7, Figure 12 1868g, AJC 4: 94 (Tahiti). Lectotype, here selected, MCZ 31709; paralectotypes MCZ 303450. The type figured by Reeve (1865, Conchologia Iconica 15: Eu- lima, pi. 3, figs. 20a, b); was not located in BMNH by Kay (1965: 86). subrufa, Helicina 1865a, PZS for 1864: 676 [nomen nudum]. subrufa 'Pease' Garrett, Helicina 1884, JANSP (2) 9: 104, pi. 3, figs. 68, 68a, b (Raiatea and Borabora). Lectotype ANSP 14463a selected by Baker (1964: 162). subviridis, Vitrina 1868), AJC 4: 154, pi. 12, fig. 5 (Marquesas [Islands]). Holotype ANSP 49269a, teste Baker (1963: 237). suffusa, Terebra Plate 9, Figure 18 1869d, AJC 5: 67 ([Honolulu] Oahu). Lectotype MCZ 248802 selected by Bratcher and Cernohorsky (1987: 47, pi. 5, fig. 17b); paralectotype MCZ 248803. sulcata, Terebra Plate 8, Figure 1 1869d, AJC 5: 67 ([Honolulu] Oahu). Lectotype, here selected, MCZ 49967; paralectotypes MCZ 248801. sulcifera, Torinia 1869d, AJC 5: 79 (Kauai). Not located in ANSP. sulcosus, Triphoris Plate 10, Figure 3 1871a, PZS for 1870: 774 (Kauai). Lectotype, here selected, MCZ 50080; paralectotype MCZ 298496. suturalis, Partula 1865b, PZS for 1864: 675 [nomen nudum]. MCZ 24825. symmetrica, Scalaria ' 1868d, AJC 3: 290, pi. 24, fig. 14 (Tahiti). Holotype ANSP 19586 refigured by DuShane (1988 (7): 5, fig. [not numbered]) and (1990: 9, fig. 44). tahitensis, Cyclostoma 1861d, PZS for 1861: 243 (Huaheine); 1869b, JdeC 17: 158, pi. 7, fig. 1, as Scalinella. Holotype in MNHN, teste Fischer-Piette (1950: 72); paratypes MCZ 74941, 187858, 187859, and 187860. tahitensis, Dolabrifera 1861d, PZS for 1861: 245 (Tahiti); 1868e, AJC 4: 77, pi. 8, fig. 5; not located in BMNH by Kay (1965: 86). 26 Bulletin Museum of Comparative Zoology, Vol. 154, No. 1 tahitensis, Helicina 1871d, PZS for 1871: 466. New name for Helicma pisum Rousseau 1854 non Philippi 1847. tahitensis 'Pease' Brot, Melama 1877 [;>;] Martini and Chemnitz, Conchylien Cabinet (2) 1 pt. 24, Die Melaniaceen, p. 323 [nomen nudum]. Listed as a synonym of M. mamensis Lea. tahitensis, Tectura Plate 1, Figure 7 1868g, AJC 4: 98, pL 11, fig. 21 (Tahiti). Lectotype, here selected, ANSP 38949; paralectotype MCZ 150757. tenebrosa, Leptachatina 1870a, JdeC 18: 92 ([Waimea] Kauai); Crosse, 1876, JdeC 24: 97, pL 3, fig. 5. Holotype in MNHN, teste Fischer-Piette (1950: 149); paratypes MCZ 45189 and 50110. tenuicostata, Leptachatina 1869c, JdeC 17: 170 (Hawaii Island). Holotype in MNHN figured by Fischer-Piette (1950: 72, pi. 3, fig. 51). tenuiscula, Helicina 1865a, PZS for 1864: 676 [nomen nudum]; 1868b, AJC 3: 226 [nomen nudum] (Tahaa). tenuistriata, Helicma 1865a, PZS for 1864: 676 [nomen nudum]; 1868b, AJC 3: 226 (Tahaa) [nomen nudum]. tenuistriata, Rissoina Plate 9, Figure 5 1868d, AJC 3: 295, pi. 24, fig. 30 (Paumotus). Lec- totype, here selected, ANSP 19253; paralectotypes MCZ 178851. terreslris 'Pease' Garrett, Partula 1884, J ANSP (2) 9: 75 (. . . Opoa Valley on the south- east coast [of Raiatea]). Lectotype ANSP 59450a se- lected by Baker (1963: 205) figured by Pilsbry (1909: 243, pi. 17, fig. 11); paralectotypes MCZ 25296 and 297859. tessellatus, Plcurobranchus 1861d, PZS for 1861: 245 (Pacific Islands) [descrip- tion without name]; named, 1864, PZS for 1863: 510; 1868e, AJC 4: 80, pi. 9, fig. 4; not mentioned by Kay (1965). triangulatum, Sistrum Plate 9, Figure 23 1868d, AJC 3: 278, pi. 23, fig. 15 (Hawaii [Island]). Lectotype, here selected, MCZ 295580; paralecto- types MCZ 295589; not located in ANSP. tricarinatum, Bittium 1861c, PZS 28: 433 (Sandwich Islands). Holotype BMNH 1962806, teste Kay (1965: 49, pi. 5, fig. 14). trigonalis. Pinna Plate 1, Figure 2 1861a, PZS for 1861: 242 (Kingsmill Islands). Lec- totype, here selected, MCZ 225951; not located in BMNH by Kay (1965: 86). trilineata, Partula 1865b, PZS for 1864: 675 [nomen nudum]; 1866a, AJC 2: 195 (Tahiti); 1867a, AJC 3: 81, pi. 1, fig. 1. Lec- totype ANSP 59563a selected by Baker (1963: 205) is the figured type; paralectotypes MCZ 24940 and 25130. triplicata, Auriculella Plate 2, Figure 9 1868f, JdeC 16: 346 ([Oahu]). Lectotype, here se- lected, MCZ 45150; paralectotypes MCZ 298487; not located in MNHN by Fischer-Piette (1950: 71). triticea, Rissoina 1861b, PZS 28: 438 (Sandwich Islands). Holotype BMNH 1962844 figured by Kay (1965: 68, pi. 11, fig. 11); paratypes MCZ 178859. triticea, Tnphoris 1861b, PZS 28: 433 (Sandwich Islands). Holotype BMNH 1962807, teste Kay (1965: 50, pi. 10, fig. 3). tuberculatus, Triphoris Plate 10, Figure 4 1871a, PZS for 1870: 776 (Kauai). Lectotype, here selected, MCZ 50081; paralectotypes MCZ 298497. luberculiferum, Cerithium 1869d, AJC 5: 76 (Paumotus). Figured by Sowerby [(>;] Reeve, 1865, Conchologia Iconica 15: Cerithium, pi. 2, fig. 11 (Island of Annaa, Cuming) as Cerithium adansomi 'Bruguiere' Sowerby non Bruguiere 1792. Holotype BMNH 199021; paratypes MCZ 302380. tuberculosa, Engma 1863b, PZS for 1862: 243 (Pacific Islands); 1867d, AJC 3: 274 (Baker Island). Lectotype BMNH 1964296 selected by Kay (1965: 76, pi. 14, figs. 5, 6); para- lectotypes MCZ 260616. tumida, Clathurella Plate 6, Figure 18 1868a, AJC 3: 218, pi. 15, fig. 14 (Paumotus). Lec- totype, here selected, ANSP 15813 is the figured type; two paralectotypes ANSP 391053; paralecto- types MCZ 231968 and 231969. tumida, Hyalopsis 1871f, AJC 7: 27, pi. 9, fig. 6 (Solomon Islands). Not located in ANSP. turbinatus, Omphalius Plate 8, Figure 20 1869e, AJC 5: 84, pi. 8, fig. 15 (Gulf of [Golfo de] California, La Paz [Baja California Sur, Mexico]). Lectotype ANSP 40820 selected by Pilsbry (1932: 85, pi. 10, figs. 9, 9a, 9b); paralectotype ANSP 31053. turbinella, Helicma 1865b, PZS for 1864: 676 [nomen nudum]. turgidula, Leptachatina i870a, JdeC 18: 89 ([Waimea] Kauai); Crosse, 1876, JdeC 24: 96, pi. 4, fig. 5. Holotype in MNHN, teste Fischer-Piette (1950: 149); paratypes MCZ 45182 and 45183. turgidula, Limnaea Plate 4, Figure 1 1 1870b, AJC 6: 5, pi. 3, fig. 3 (Oahu). Lectotype, here selected, MCZ 298901; paralectotype MCZ 298902; two paralectotypes ANSP 21932 though not located in ANSP by Baker (1964: 154). turgidus, Bulimus [Partula] 1865a, PZS for 1864: 670 (Islands of the Central Pacific); 1871d, PZS for 1871: 473 (Tahiti). Probable syntypes MCZ 25367 and 26354 though both labeled as from Raiatea. turricula, Partula 1865d, PZS for 1864: 675 [nomen nudum]; 1872b, AJC 7: 196 ("?" New Hebrides). Holotype ANSP 59926a, /rsff Baker (1963: 205). turricula, Rissoina 1861b, PZS 28: 438 (Sandwich Islands). Lectotype BMNH 1962845 selected by Kay (1965: 69, pi. 11, fig. 10); paralectotypes MCZ 178858. MoLLUSKS Described by W. H. Pease • Johnson 27 ualanensis 'Pease' Martens and Langkavel, Melania 1871, Donum Bismarcktatium, p. 38. Error for ouala- nensis 'Pease' Tryon, Melania. umbilicata, Cassis Plate 6, Figure 3 1861b, PZS 28: 436 (Sandwich Islands). Lectotype, here selected, MCZ 298907; not located in BMNH by Kay (1965: 85). umbilicata, Partula 1865a, PZS for 1864: 675 [nomen nudum]; 1866a, AJC 2: 200 ([Haamene Valley, Tahaa] Tahitian Archi- pelago); 1867a, AJC 3: 81, pi. 1, fig. 7. Lectotype ANSP 59452a selected by Baker (1963: 205) from one of the syntypes figured by Pilsbry (1909: 230, pi. 21, figs. 13, 14); paralectotypes MCZ 25300 and 25310. umbilicata, Scalaria Plate 9, Figure 12 1869d, AJC 5: 76 (Oahu). Holotype MCZ 187393 figured by DuShane (1990: 4, fig. 9). undatolirata 'Pease' Dall, Patella 1871, AJC 6: 279 (Sandwich Islands) [nomen nudum]; Pease, 1872c, AJC 7: 198 listed as a synonym of Helcionisus exaratus Nuttali. undulata 'Pease' Tryon, Natica 1886, MofC (1) 8: 23 [nomen nudum]. unilineatum, Cerithium 1861b, PZS 28: 432 (Sandwich Islands). Lectotype BMNH 1962798 selected by Kay (1965: 45, pi. 10, fig. 5); paralectotypes MCZ 297946. umplicata, Aunculella 18681, JdeC 16: 344, pi. 14, figs. 7, 7a ([Lahaina] Maui). Two figured syntypes in MNHN, teste Fi- scher-Piette (1950: 71); syntypes MCZ 159563 and 161636. I'aria, Cythara 1860b, PZS 28: 147 (Sandwich Islands). Lectotype BMNH 1962782 selected by Kay (1965: 33, pi. 10, fig. 13); paralectotypes MCZ 50000, 50001, and 50002. See under: peasei, Martens and Langkavel, Colum- bella (Seminella). Z'ariabilis, Carelia 1870c, JdeC 18: 402 (Kauai). Single specimen found by Pease; not mentioned by Fischer-Piette (1950: 76). variabilis, Collonia 1861b, PZS 28: 436 (Sandwich Islands). Lectotype BMNH 1963331 selected by Robertson [m] Kay (1965: 61, pi. 7, figs. 1, 2); paralectotypes MCZ 119269. variabilis, Engina Plate 7, Figure 5 1868d, AJC 3: 275, pi. 23, fig. 6 (Paumotus). Lecto- type MCZ 260618 selected by Cernohorsky (1987: 99, figs. 11, 12); paralectotypes MCZ 260615. variabilis, Partula 1865b, PZS for 1864: 675 [nomen nudum]; 1866a, AJC 2: 203 ([Vaioara Valley, on the west coast of Raiatea, teste Hartmann, 1884: 76] Tahitian Archipelago); 1867a, AJC 3: pi. 1, figs. 12-14; 1871, PZS for 1871: 473 (Raiatea). Lectotype ANSP 59452a selected by Baker (1963: 205) is the syntype represented on pi. 1, fig. 12; paralectotypes MCZ 24994. variabilis, Realia 1865e, AJC 1: 288 (Polynesia); Tryon, 1866a, AJC 2: 82, pi. 5, fig. 2 (Hervey Isles); Pease, 1869a, JdeC 17: 148 as Omphalotropis (Atiu); 1872b, AJC 7: 197 (also Tongan Group). Holotype ANSP 13341a, teste Baker (1964: 179); paratypes MCZ 187844 and 187847 from Atiu; paratypes MCZ 187846 from the Tongan Group. variatis, Pleurobranchus 1860a, PZS 28: 25 (Sandwich Islands). Not located in BMNH by Kay (1965: 84). varicifera, Daphnella 1868a, AJC 3: 221, pi. 15, fig. 21 (Paumotus). Ho- lotype ANSP 15726. variegata, Dolabella 1860a, PZS 28: 22 (Sandwich Islands). Syntype MCZ 297871; not located in BMNH by Kay (1965: 84). venosus. Helix 1866b, AJC 2: 290, pi. 21, fig. 2 (Polynesia); 1871d, PZS for 1871: 475 as Helicopsis (Roratonga). Holo- type ANSP 49163a, teste Baker (1963: 237); paratypes MCZ 11529 and 297935. venusta 'Pease' Pace, Columbella 1902, Proceedings of the Malacological Society of London 5: 150 [nomen nudum]. venusta, Eulima Plate 7, Figure 10 1868d, AJC 3: 294, pi. 24, fig. 24 (Tahiti). Lectotype, here selected, ANSP 19788; paralectotypes MCZ 187748 and 187749. verecunda, Pithys 1870c, JdeC 18: 397 (Marquesas [Islands]); 1871d, PZS for 1871: 454 as Pitys. Not mentioned by Fi- scher-Piette (1950: 76). verticillata, Helix Plate 3, Figure 7 1865b, PZS for 1864: 675 [nomen nudum]; 1868b, AJC 3: 228 as Nanina (Moorea). Holotype ANSP 49284, teste Baker (1963: 237). vescoi 'Dohrn' Pease, Helicina 1865b, PZS for 1864: 676 [nomen nudum]. vexillum, Partula 1865b, PZS for 1864: 675 [nomen nudum]; 1866a, AJC 2: 198 (Tahitian Archipelago); 1867a, AJC 3: 81, pi. 1, fig. 8; 1871d, PZS for 1871: 474 (Moorea). Lec- totype ANSP 59646a selected by Baker (1963: 205); paralectotypes MCZ 24814, 24886, and 24936. violacea, Clathurella Plate 6, Figure 15 1868a, AJC 3: 218, pi. 15, fig. 15 (Paumotus). Lec- totype, here selected, MCZ 231971; paralectotypes MCZ 303190; not located in ANSP. violaceus, Pleurobranchus 1864, PZS for 1863: 510. New name for Pleurobran- chus reticulatus Pease 1860 non Rang 1832 non Ke- laart 1859. See under: reticulatus, Pleurobranchus. virescens, Pachypoma Plate 8, Figure 17 1869d, AJC 5: 73, pi. 8, fig. 10 (Tarawa Island). Lec- totype, here selected, MCZ 302624; paralectotypes MCZ 302625; not located in ANSP. virginea, Partula 1865a, PZS for 1864: 675 [nomen nudum]. virginea 'Pease' Garrett, Partula 1884, JANSP (2) 9: 61, pi. 3, fig. 54 (Vaipiti Valley on the west coast of Tahaa). Lectotype ANSP 59474a selected by Baker (1963: 206) is the figured type; paralectotypes MCZ 25243. 28 Bulletin Museum of Comparative Zoology, Vol. 154, No. 1 virgulata, Partula assimilis 1870c, JdeC 18: 401 (Roratonga). Syntypes MCZ 24939; not mentioned by Fischer-Piette (1950: 76). viridans, Cyclostoma 1865b, PZS for 1864: 676 [nomen nudum]. viridescens, Cyclostoma 1861d, PZS for 1861: 243 (Pacific Islands); 1869a, JdeC 17: 153, pi. 7, fig. 7, as Omphalotropis (Hua- heine). Holotype in MNHN, teste by Fischer-Piette (1950: 72). viridis, "7" Acanthochites 1872a, AJC 7: 194 (Kauai). Not located in ANSP. viridis, Carelia variabilis 1870c, JdeC 18: 402 (east side of Kauai). Pilsbry and Cooke (1914: 16, pi. 9, fig. 11) suggested that this figured specimen MCZ 23343 may be the type of viridis; not mentioned by Fischer-Piette (1950: 76). viridis, Lobiger Plate 4, Figure 18 1864, PZS for 1863: 510 ([Tahiti] Pacific Islands); 1861d, PZS for 1861: 246, description without a name. Lectotype, here selected, MCZ 297869; para- lectotypes MCZ 298463. viridis, Lophcercus 1861d, PZS for 1861: 246 (Pacific Islands); 1868e, AJC 4: 74, pi. 8, figs. 1, 2. Lectotype BMNH 1964324 selected by Kay (1965: 72, pi. 12, figs. 7, 8); para- lectotype MCZ 297932. vitrea, Bullina 1860a, PZS 28: 19 (Sandwich Islands). Lectotype BMNH 1961456 selected by Kay (1965: 5, pi. 1, figs. 1,2). vitrea 'Pease' Dohrn, Tornatellides 1863, Malakozoologische Blatter 10: 162. Not recog- nizable, teste Pilsbry and Cooke (1915: 203). vittata, Partula 1865b, PZS for 1864: 675 [nomen nudum]; 1866a, AJC 2: 194 (Tahitian Archipelago [restricted to the high- er portions of Toloa Valley, on the west coast of Raiatea, teste Garrett, 1884: 75]). Syntypes MCZ 24879. zigzac, Helicina Plate 5, Figure 8 1868b, AJC 3: 229 (Oualan [Island]). Lectotype, here selected, ANSP 14485 is the measured type, teste Baker ( 1 964: 1 63); three paralectotypes ANSP 359504; paralectotypes MCZ 297924. The taxa for which no type material was located has been extracted from the pre- ceding alphabetical list and arranged here according to publications and dates. Proceedings of the Zoological Society of London nebulosa, Borsonia: 1860b, PZS 28: 143 (Sandwich Islands) pallida, Mitra: 1860b, PZS 28: 146 (Sandwich Is- lands) oryza, Marginella: 1860b, PZS 28: 147 (Sandwich Islands) maculosa, Daphnella. 1860b, PZS 28: 148 (Sandwich Islands) sculpta, Nucula: 1860b, PZS 28: 189 (Corea [Korea] Sea) luteostoma, Ranella. 1861a, PZS 28: 397 (Sandwich Islands) neglectus, Conus. 1861a, PZS 28: 398 (Sandwich Is- lands) cancellata, Coralliobia. 1861a, PZS 28: 399 (Sand- wich Islands); "Only a single dead specimen found." lineata, Cohimbella. 1861a, PZS 28: 399 (Sandwich Islands) costellifera, Anachis: 1863d, PZS for 1862: 279 (Pa- cific Islands) ovata, Hydrocena : 1865a, PZS for 1864: 674 [Mangaia Island] sculptilis, Coralliobia: 1865d, PZS for 1865: 513 (Is- lands of the Central Pacific) conica, Torinia: 1865d, PZS for 1865: 514 (Islands of the Central Pacific) multiplicata, Mitroidea. 1865d, PZS for 1865: 514 (Islands of the Central Pacific) punctatus, Triphoris. 1871a, PZS for 1870: 775 (An- naa Island) cylindricus, Triphoris. 1871a, PZS for 1870: 776 (Apaiang Island) filicostata, Pitys: 1871d, PZS for 1871 (Kauai) simplaria. Vertigo: 1871d, PZS for 1871: 461 (Mar- quesas Islands) bacca. Vertigo: 1871d, PZS for 1871; 461 (Kalapana [Puna District] Hawaii) costellifera, Truncatella: 1871d, PZS for 1871: 468 (\'avau [Island]; "Collected at the above locality by Mr. John Brazier" Journal de Conchyliologie extensa, Leptachatina: 1870a, JdeC 18: 92 (Kauai) hicida, Leptachatina: 1870a, JdeC 18: 93 (Kauai) verectinda, Pithys: 1870c, JdeC 18: 397 (Marquesas Islands) American Journal of Conchology distans. Helix: 1866b, AJC 2: 290 (Sandwich Islands) flammulata. Mitra: 1868a, AJC 3: 212 (Sandwich and Paumotus Islands) .sectilis. Mitra: 1868d, AJC 3: 271 (Hawaii) aurantium. Opercidatum: 1868d, AJC 3: 287 (Ha- waii) rosacea. Terebra: 1869d, AJC 5: 65 (Oahu) propinqiia, Terebra: 1869d, AJC 5: 66 (Hawaii) lirata. Pleurotoma: 1869d, AJC 5: 68 (Oahu) monilifera. Pleurotoma: 1869d, AJC 5: 68 (Oahu) neivcombii. Mitra: 1869d, .^JC 5: 69 (Oahu) striata. .Mcijna: 1869d, AJC 5: 70 (Hawaii) balteata, Rissoina: 1869d, AJC 5: 72 (Hawaii) sulcifera, Torinia: 1869d, AJC 5: 79 (Kauai) viridis. "?" Acanthochites: 1872d, AJC 7: 194 (Kauai) MoLLUSKS Described by W. H. Pease • Johnson 29 obesa. Partula: 1868b, AJC 3: 223, pi. 15, fig, 12 (Hab?); "We have but a single specimen" striata. Engina: 1868d, AJC 3: 275, pi. 23, fig. 10 (Paumotus) elongata, Volutella: 1868d, AJC 3: 281, pi. 23, fig. 23 (Fanning Island) semicostata. Rissoa: 1868d, AJC 3: 296, pi. 24, fig. 32 (Caroline Islands) uttcea, Nassa: 1869d, AJC 5: 70, pi. 8, fig. 7 (Caroline Islands) tiimida, Hyalopsis: 1871, AJC 7: 27, pi. 9, fig. 6 (Solomon Islands) ADDITIONAL OR CORRECTED LOCALITY DATA This list was based essentially on the Na- tional Geographic Atlas of the World (6th ed., 1990), but also useful were the Index to the Islands of the Pacific by Brigham (1900), the Hawaiian Gazetteer by Coul- ter (1935), and Pacific Island Names by Motteler (1986). Aitutake Island [Aitutaki Atoll, Hervey Group, Cook Islands] Annaa Island [Anaa, Tuamotu Archipelago] Apaian Island [Abaiang Island, Gilbert Islands, Ki- ribati] Apaiang Island [Abaiang Island, Gilbert Islands, Ki- ribati] Apiana Island [Abaiang Island, Gilbert Islands, Ki- ribati] Ascension Island [Pohnpei, Caroline Islands] Atiu Island [Hervey Group, Cook Islands] Baker Island [0°13'30"N, 176°29'30"W] Bolabola Island [Bora-Bora, Society Islands] Ebon Island [Atoll, Marshall Islands] Fanning Island [Tabuaeran, Kiribati; 3°5r25"N, 159°22"W] Guam [Island, Mariana Islands] Hauheine Island [Society Islands] Hawaii [Island, Hawaiian Islands] Howland Island [0°49'N, 176°40'W] Huaheine Island [Society Islands] Jarvis Island [Line Islands; 0°22'33"S, 159°54'1I"W] Kauai [Island, Hawaiian Islands] Kingsmill Islands [Kingsmill Group, Gilbert Islands, Kiribati] Lanai [Island, Hawaiian Islands] Makaimo Island [Makemo Island, Tuamotu Archi- pelago] Mangaia Island [Hervey Group, Cook Islands] Mangia Island [Mangaia Island, Hervey Group, Cook Islands] Maui [Island, Hawaiian Islands] Molokai [Island, Hawaiian Islands] Niihau [Island, Hawaiian Islands] Niiie Island [south of the Samoa Islands; 19°S, 170°W] Nukahiva [Nuku Hiva Island, Marquesas Islands] Oahu [Island, Hawaiian Islands] Oualan [Kusaie or Kosrae Island, Caroline Islands] Paumotus [Tuamotu Archipelago] Philip Island [south of Norfolk Island] Ponape [Pohnpei Island, Caroline Islands] Raiatea [Island, Society Islands] Roratonga [Rarotonga Island, Hervey Group, Cook Islands] Sandwich [Hawaiian Islands] Tahaa [Island, Society Islands] Tahiti [Island, Society Islands] Tarawa [Island, Gilbert Islands, Kiribati] Tutuila [Island, Samoa Islands] Vavau [Island, Vava'u Group, Tonga Islands] Viti [Fiji] Islands] LITERATURE CITED For the complete bibliography of Pease, see Clench (1975). Austin, Walter. 1921. James Walker Austin. Nor- wood, Massachusetts; Plimpton Press, Privately Printed, ix + 227 pp. Baker, H. B. 1963. Type land snails in the Academy of Natural Sciences of Philadelphia. Part II. Land Pulmonata, exclusive of North America north of Mexico. Proceedings of the Academy of Natural Sciences of Philadelphia, 115(8): 191-259 (No- vember). . 1964. Type land snails in the Academy of Natural Sciences of Philadelphia. Part III. Lim- nophile and Thalassophile Pulmonata. Part IV. Land and fresh-water Prosobranchia. Proceed- ings of the Academy of Natural Sciences of Phil- adelphia, 116(4): 149-193 (October). Bratcher, T., and W. Cernohorsky. 1987. Liv- ing Terebras of the World. Melbourne, Florida: American Malacologists. 240 pp. Brigham, W. T. 1900. Index to the islands of the Pacific Ocean. Memoirs of the B. P. Bishop Mu- seum, 1(2): 1-256 (December). Brodie, G. D., and R. C. Willan. 1993. Rede- scription and ta.xonomic reappraisal of the trop- ical Indo-Pacific nudibranch Siraius nucleola (Pease, 1860) (Anthobranchia: Doridoidea: Dori- didae). The Veliger, 36(2): 124-133 (April). Brot, a. 1872. Notice sur quelques especes de Me- lanies nouvelles ou imparfaitement connues. [in] Materiaux pour servir a I'etude de la famille des Melaniens. Ill: 27-55, pis. 2-4. . 1877. [in] Martini and Chemnitz (Kuester's Edition). Systematisches Conchylien-Cabinet (2) 1: pt. 24, Die Melaniaceen. Signature 264, pp. 273-352. Gate, C. N. 1973. A systematic revision of the Recent Cypraeid family Ovulidae (Mollusca: Gastropoda). The Veliger, 15(Suppl.): 116 pp., 51 pis. (January). Cernohorsky, W. O. 1976. The Mitridae of the World. Part 1. The subfamily Mitrinae. Indo- 30 Bulletin Museum of Comparative Zoology, Vol. 154, No. 1 Pacific Mollusca, 3(17): 273-528, pis. 248-466 (September). . 1984. Systematics of the family Nassariidae (Mollusca: Gastropoda). Bulletin of the Auckland Institute and Museum, 14: 1-356, 51 pis., 173 figs. (July). . 1987. Type specimens of Pacific Mollusca described mainly by A. Garrett and W. Pease with description of a new Morula species (Mol- lusca: Gastropoda). Records of the Auckland In- stitute and Museum, 24: 93-105 (December). 1991. The Mitridae of the World. Part 2. The subfamily Mitrinae concluded and subfam- ilies Imbricariinae and Cylindromitrinae. Mono- graphs of Marine Mollusca, (4): 1-164, 155 pis. (May). Cernohorsky, W. O., and T. Bratcher. 1976. Notes on the taxonomy of Indo-Pacific Terebri- dae (Mollusca: Gastropoda), with description of a new species. Records of the Auckland Institute and Museum, 13: 131-140, text figs. (December). Clench, VV. J. 1932. Diplomorpha coxi (Pease). Nautilus, 46(2): 68-69 (October). . 1964. The genera. Pedipes Bind Laemodon- ta in the Western Atlantic. Johnsonia, 4(42): 117- 128, pis. 76-79 (February). . 1975. Catalogue of species and bio-bibli- ography of William Harper Pease. See under: Kay, E. A., and W. J. Clench (1975). 1979. A biography of Andrew Garrett, ear- ly naturalist of Polynesia: Part 2. Catalogue of molluscan species and bibliography. Nautilus, 93(3): 96-102 (April). Clench, W. J., and R. D. Turner. 1948. A cat- alogue of the family Truncatellidae with notes and descriptions of new species. Occasional Pa- pers on Mollusks, 1(13): 157-212, pis. 23-25 (June). . 1957. The family Cymatiidae in the West- ern Atlantic. Johnsonia, 3(36): 189-244, pis. 110- 135 (December). CoOKE, C. M. 1910, 1911. Leptachatina. Manual of Conchology, (2)21: 1-64, pis. 1-9 (July 1910); 65-92, pis. 10-16 (March 1911). Cooke, C. M., and Y. Kondo. 1960. Revision of Tornatellinidae and Achatinellidae (Gastropoda, Pulmonata). Bernice P. Bishop Museum, Bulle- tin, 221: 303 pp. (December). Coulter, J. W. 1935. A gazetteer of the territory of Hawaii. University of Hawaii, Research Pub- lications, 11: 241 pp. Crampton, H. E. 1916. Studies on the variation, distribution, and evolution of the genus Partula. The species inhabiting Tahiti. Carnegie Institu- tion of Washington, Publication, 228: 311 pp., 34 pis. . 1932. Studies on the variation, distribution, and evolution of the genus Partula. The species inhabiting Moorea. Carnegie Institution of Washington, Publication, 410: vi -I- 335 pp., 24 pis. Crosse, H. 1876. Note complementaire sur quel- ques especes de Mollusques terrestres habitant I'ile Kauai (lies Hawaii). Journal de Conchvliolo- gie, 24: 95-99, pis. 1, 3, 4 (January). Dall, W. H. 1871. On the limpets; with special reference to the species of the West Coast of America, and to a more natural classification of the group. American Journal of Conchologv, 6(3): 227-282, pis. 14-17 (April). DOHRN, H. 1863. Ueber Tornatellina. Malako- zoologische Blatter, 10: 156-162 (August). DuShane, H. 1988. Hawaiian Epitoniidae. Ha- waiian Shell News, 36(4): 1, 7 (April); (5): 9-10 (May); (7): 4-5 (July); (9): 4-5 (September); (10): 5 (October). . 1990. Hawaiian Epitoniidae. Hawaiian Shell News, 38(Suppl. 1): 17 pp. (January). FisCHER-PlETTE, E. 1950. Liste des types decrits dans le Journal de Conchyliologie et conserves dans la collection de ce journal. Journal de Con- chyliologie, 90(1): 8-23, pi. 1 (January); 90(2): 65-82, pis. 2-4 (April). Garrett, A. 1879. Annotated catalogue of the spe- cies of Cypraeidae collected in the S. Sea Islands. Journal of Conchology, 2(4): 105-128 (April). . 1881. The terrestrial Mollusca inhabiting the Cook's or Hervey Islands. Journal of the Academy of Natural Sciences of Philadelphia, (2) 8: 381-411 (March). . 1884. The terrestrial Mollusca inhabiting the Society Islands. Journal of the Academy of Natural Sciences of Philadelphia, (2) 9: 17-114, pis. 2, 3 (July). . 1887. On the terrestrial mollusks of the V'iti Islands. Proceedings of the Zoological Society of London for 1867, pp. 284-316. Greene, K. W. 1960. William Harper Pease— 100 years later. Hawaiian Shell News, 8(6): 1, 4, 5 (April); Pease's catalogue of books on Hawaii. Ibid. 8(7): 4, 5 (May); W. H. Pease's library- some facts — some theories. Ibid. 8(8): 1, 3, 4, 8 (June); William Harper Pease — his 21 years in Hawaii. Ibid. 8(9): 5, 8 (July); W. H. Pease— his last few years and death at 47. Ibid. 8(10): 5, 8 (August); William Harper Pease and his interest in land shells. Ibid. 8(11): 5, 8 (September); Wil- liam Harper Pease and his interest in marine shells. Ibid. 8(12): 5, 6, 8 (October). Hartman, W. D. 1881. Observations on the species of the genus Partula Per. with a bibliographic catalogue. Observations on the duplicates of the genus Partula Fer. contained in the Museum of Comparative Zoology, Cambridge, Mass., for- merly belonging to the collection of the late Wil- liam H. Pease. Bulletin of the Museum of Com- parative Zoology, 9(5): 171-196, 2 maps (November). . 1886. New species of Partula from the New Hebrides and Solomon Islands. Proceedings of the Academy of Natural Sciences of Philadel- phia. 38: 30-35, pi. 2 (April). HouBRiCK, R. S. 1992. Monograph of the genus Cerithium Bruguiere in the Indo-Pacific (Cer- MOLLUSKS Described by W. H. Pease • Johnson 31 ithidae: Prosobranchia). Smithsonian Contribu- tions to Zoology, 510: iv + 211 pp., 145 figs. Hyatt, A., and H. A. Pilsbry. 1911. Amastrinae. Manual of Conchology, (2) 21: 93-387, pis. 14- 56 (March-December). Kay, E. a. 1965. Marine molluscs in the Cuming collection, British Museum (Natural History) de- scribed by William Harper Pease. Bulletin of the British Museum (Natural History), Zoology, Sup- plement 1, 96 pp., 14 pis. Kay, E. a., and W. J. Clench. 1975. A biobibliog- raphy of William Harper Pease, Malacologist of Polynesia. Part 1. A biography of William Harp- er Pease by E. A. Kay, pp. 1-21; Part 2. Catalogue of species and bibliography of William Harper Pease by W. J. Clench, pp. 22-50. Nemouria, Occasional Papers of the Delaware Museum of Natural History, 16: 1-50 (December). KiLIAS, R. 1969. Die typen und typoide der Mol- lusken-Sammlung des Zoologischen Museums in Berlin (5). Mitteilungen aus dem Zoologischen Museum in Berlin, 45(2): 245-258, 2 figs. KoRN, A. 1958. The Victorian Visitors. Honolulu: University of Hawaii Press. Martens, E. von, and B. Langkavel. 1871. Do- num Bismarckianum. Eine sammlung von Siid- see-Conchylien. Berlin: Verlag von Ferdinand Berggold. vi -t- 74 pp., 4 col. pis. (Originally published as a University "Program" earlier in 1871 with the title Beschreibung von Siidsee- conchylien des Donum Bismarckianum von Dr. Langkavel. This edition consisted of 36 pages only, of which pp. 1-24 are identical to those of the V. Martens and Langkavel edition (74 pp.). Each edition has the same 4 colored plates.) Motteler, L. S. 1986. Pacific Island Names. Bish- op Museum Miscellaneous Publication, 34: [vii] + 91 pp. Myers, B. W., and A. D'Attilio. 1989. A new Favartia Jousseaume, 1880 from the Western Pa- cific and a lectotype designated for Murex foveo- latus Pease, 1869 = Murex peasei Tryon, 1880. Venus, The Japanese Journal of Malacology, 48(3): 154-160, figs. 1-12 (October). Nevill, G. 1878. Hand List of Mollusca in the Indian Museum, Calcutta. Part I. Gastropoda, xv + 338 pp. 1884. Hand List of Mollusca in the Indian Museum, Calcutta. Part II. Gastropoda, .x -I- 306 pp. Pace, S. 1902. Contributions to the study of the Columbellidae. Proceedings of the Malacological Society of London, 5(1): 36-112 (April); (2): 113- 154 (July). Paetel, Fr. 1873. Catalog der Conchylien-Samm- lung. Berlin: Verlag von Gebriider Paetel. 172 pp. . 1887, 1888. Catalog der Conchylien- Sammlung. Berlin: Verlag von Gebriider Paetel, 1(1-6): 1-480, 1887; (7, 8): 481-638, 1888. Pease, W. H. 1860a. Descriptions of new species of Mollusca from the Sandwich Islands. Proceed- ings of the Zoological Society of London, 28: 18- 36 (February-May). 1860b. Descriptions of new species of Mol- lusca from the Sandwich Islands. (Part II). Pro- ceedings of the Zoological Societv of London, 28: 141-148 (February-May). 1860c. Descriptions of three species of ma- rine shells from the Pacific Ocean. Proceedings of the Zoological Societv of London, 28: 189- 190 (February-May). . 1861a. Descriptions of seventeen new spe- cies of marine shells, from the Sandwich Islands, in the collection of Hugh Cuming. Proceedings of the Zoological Society of London, 28: 397- 400 (August 1860-March 1861). . 1861b. Descriptions of forty-seven new species of shells from the Sandwich Islands, in the collection of Hugh Cuming. Proceedings of the Zoological Society of London, 28: 431-438 (August 1860-March 1861). 1861c. Descriptions of six new species of land shells, from the Island of Ebon, Marshall's group, in the collection of H. Cuming. Proceed- ings of the Zoological Societv of London, 28: 439-440 (August 1860-March 1861). . 1861d. Descriptions of new species of Mol- lusca from the Pacific Islands. Proceedings of the Zoological Society of London for 1861, pp. 242- 247 (September). . 1862a. Descriptions of two new species of Helicter {Achatinella, Swains.), from the Sand- wich Islands, with a history of the genus. Pro- ceedings of the Zoological Societv of London for 1862, pp. 3-7 (June). 1862b. Catalogue des especes de Rissoina des lies Sandwich et description dune espece nouvelle. Journal de Conchyliologie, 10: 381- 383, pi. 13 (October). . 1863a. Description of a new genus of ma- rine shells from the Sandwich Islands. Proceed- ings of the Zoological Society of London for 1862, p. 240 (April). 1863b. Descriptions of new species of ma- rine shells from the Pacific Islands. Proceedings of the Zoological Society of London for 1862, pp. 240-243 (April). . 1863c. Descriptions of new species of shells from the Pacific Islands. Proceedings of the Zoo- logical Societv of London for 1862, pp. 243-245 (April). . 1863d. Descriptions of new species of ma- rine shells from Pacific Islands. Proceedings of the Zoological Society of London for 1862, pp. 278-280 (April). . 1864. On errors and omissions in former communications. Proceedings of the Zoological Society of London for 1863, p. 510 (April). . 1865a. Descriptions of new species of land shells from the Islands of the Central Pacific. Proceedings of the Zoological Society of London 32 Bulletin Museum of Comparative Zoology, Vol. 154, No. 1 for 1864, pp. 668-676 [p. 676 misprinted as 766] (May). . 1865b. Remarks on the species of Succinea infiabiting the Tahitian Archipelago, with de- scription of a new species. Proceedings of the Zoological Society of London for 1864, pp. 676- 677 (\4ay). . 1865c. Description of a new species of La^i- rus, and remarks on others, inhabiting the Pacific Islands. Proceedings of the Zoological Society of London for 1865, pp. 53-54 (June). . 1865d. Descriptions of new genera and species of marine shells from the Islands of the Central Pacific. Proceedings of the Zoological So- ciety of London for 1865, pp. 512-517 (October). . 1865e. Descriptions of new species of Phaneropneumona, inhabiting Polynesia. Amer- ican Journal of Conchology, 1(4): 287-291 (Oc- tober). . 1866a. Descriptions of new species of Par- tula, inhabiting the Tahitian Archipelago. Amer- ican Journal of Conchology, 2(3): 193-203 (July). . 1866b. Descriptions of new species of land shells, inhabiting Polynesia. American Journal of Conchology, 2(4): 289-293 (July). . 1867a. Illustrations of new species of Par- tula. American Journal of Conchology, 3(1): 81, pi. 1 (April). . 1867b. [under] Scientific Intelligence. American Journal of Conchology, 3(1): 104 (April). . 1868a. Descriptions of marine Gasteropo- dae, inhabiting Polynesia. American Journal of Conchology, 3(3): 211-222, pi. 15 (January). . 1868b. Descriptions of new species of land shells, inhabiting Polynesia. American Journal of Conchology, 3(3): 223-230 (January). . 1868c. Notes on the remarks of Dr. P. P. Carpenter (Published in Proc. Zool. Soc. Lond., 1865) on certain species of marine Gastropoda, named by W. Harper Pease. American Journal of Conchology, 3(3); 231-234 (January). . 1868d. Descriptions of sixty-five new spe- cies of marine Gasteropodae, inhabiting Poly- nesia. American Journal of Conchology, 3(4): 271- 297, pis. 23, 24 (April). . 1868e. Descriptions of marine Gasteropo- dae, inhabiting Polynesia. American Journal of Conchology, 4(2): 71-80, pis. 7-10 (October). . 1868f. Descriptions d'especes nouvelles d'Auriculella, provenant des iles Hawa'i. Journal deConchyliologie, 16:342-347, pi. 14 (October). 1868g. Descriptions of marine Gasteropo- dae, inhabiting Polynesia. American Journal of Conchology, 4(3): 91-102, pis. 11, 12 (Novem- ber). 1868h. Synonymy of marine Gasteropodae inhabiting Polynesia. .American Journal of Con- chology, 4(3): 103-132 (November). . 18681. Description of a new species of ge- nus Latirus, Montf. American Journal of Con- chology, 4(3): 152 (November). . 1868]. Description of a new genus and elev- en species of land shells, inhabiting Polynesia. American Journal of Conchology, 4(3): 153-160, pi. 12 (November). . 1869a. Monographic de la famille des Rea- liea, Pfeiffer. Journal de Conchvliologie, 17: 131- 160 (April). . 1869b. Liste des especes supposees appar- tenir au genre Assiminea de Leach. Journal de Conchyliologie, 17: 161-167, pi. 7 (April). . 1869c. Description d especes nouvelles du genre Helicter, habitant les iles Hawaii. Journal de Conchyliologie, 17: 167-176 (April). . 1869d. Descriptions of new species of ma- rine Gasteropodae inhabiting Polynesia. Ameri- can Journal of Conchology, 5(2): 64-79, pi. 8 (October). . 1869e. Remarks on marine Gastropodae inhabiting the West Coast of America; with de- scriptions of two new species. American Journal of Conchology, 5(2): 80-84 (October). . 1869f. Corrections and additions to "Syn- onomy of Marine Gasteropodae inhabiting Poly- nesia." American Journal of Conchology, 5(2): 85-87 (October). . 1869g. On the classification of the Helic- terinae. Proceedings of the Zoological Society of London for 1869, pp. 644-652 (December). . 1870a. Observations sur les especes de co- quilles terrestres qui habitent {sic) 1 ile Kauai (iles Hawaii) accompagnees de descriptions d'especes nouvelles. Journal de Conchyliologie, 18: 87-97 (January). . 1870b. Remarks on the species of Melania and Limnaea inhabiting the Hawaiian Islands, with descriptions of new species. American Jour- nal of Conchology, 6(1): 4-7, pi. 3 (July). . 1870c. Remarques sur certaines especes de coquilles terrestres, habitant la Polynesie, et de- scription d'especes nouvelles. Journal de Con- chyliologie, 18: 393-403 (October). . 1871a. Remarks on the genus Triphoris (Desh.), with descriptions of new species. Pro- ceedings of the Zoological Society of London for 1870, pp. 773-777 (April). . 1871b. Synonymic de quelques genres et especes de coquilles terrestres habitant la Poly- nesie. Journal de Conchyliologie, 19: 92-97 (April). . 1871c. Remarques sur le genre Dibaphus et sur quelques especes du genre Conus. Journal de Conchyliologie, 19: 98-100 (April). . 187 Id. Catalogue of the land-shells inhab- iting Polynesia, with remarks on their synonymy, distribution, and variation, and descriptions of new genera and species. Proceedings of the Zoo- logical Society of London for 1871, pp. 449-477 (August). . 187 le. Notes on the synonymy and distri- MoLLUSKS Described by W. H. Pease • Johnson 33 bution of marine Gastropoda. American Journal of Conchology 7(1): 20-25 (August). . 187 If. Descriptions of new species of land shells. American Journal of Conchology, 7(1): 26-27, pi. 9 (August). . 1872a. Polynesian Chitonidae. American Journal of Conchology, 7(3): 194-195 (March). . 1872b. Descriptions of four species of land shells inhabiting the Papuan Islands. American Journal of Conchology, 7(3): 196-197 (March). . 1872c. Svnonvmv of Patella exarata, Rve. American Journal of Conchology, 7(3): 198-200 (March). PiLSBRY, H. A. 1895. Bullidae. Manual of Con- chology, (1)15: 326-436, pis. 34-43 (February). . 1896. Trochomorpha fuscata Pease. Nau- tilus, 9(10): 120 (February). . 1909, 1910. Partulidae. Manual of Con- chology, (2) 20: 155-314, pis. 22-36 (September 1909); 315-336, pis. 37-43 (February 1910). . 1917. Marine mollusks of Hawaii, I-III. Proceedings of the Academy of Natural Sciences of Philadelphia, 69: 207-230, pis. 14, 15 (Au- gust). PiLSBRY, H. A., AND C. M. CooKE. 1914-16. Ap- pendix to Amastridae. Tornatellinidae. Manual of Conchology, (2) 23: 1-48, pis. 1-13 (October 1914); 49-128, pis. 14-23 (August 1915); 129- 256, pis. 24-38 (December 1915); 257-302, pis. 39-55 (February 1916). . 1920. Lyropiipa and Nesopupa. Manual of Conchology, (2) 25: 226-277, pis. 19-24; 322- 392, pis. 30-34 (April). . 1922. The identity of Helix depressiformis and H. prostrata Pease. Nautilus, 36(1): 17 (July). 1926. Lyropupa. Manual of Conchology, (2) 27: 223-266, pis. 28-32 (March). PiLSBRY, H. A., AND H. N. LowE. 1932. West Mex- ican and Central American mollusks collected by H. N. Lowe, 1929-31. Proceedings of the Acad- emy of Natural Sciences of Philadelphia, 84: 33- 144, pis. 1-17 (May). Ponder, W. F., and R. de Keyzer. 1992. A re- vision of the genus Diala (Gastropoda: Cerithio- dea: Dialidae). Invertebrate Taxonomy, 6(4): 1019-1075. Pruvot-Fol, a. 1947. Les Opisthobranches de W. Harper Pease revision. Journal de Conchyliolo- gie, 87: 96-114 (June). Reeve, L. A., and G. B. Sowerby. 1843-78. Con- chologia Iconica, Vols. 1-20, 2727 pis. and let- terpress. London. Robertson, R., I. Richardson, G. M. Davis, and A. E. BOGAN. 1981. Catalog of the types of Neontological mollusca of the Academy of Nat- ural Sciences of Philadelphia. Part 2. Gastropoda: Archaeogastropoda: Pleurotomariacea, Fissurel- lacea, Patellacea. Tryonia, (5): 1-47 (December). SCHMELTZ, J. D. E. 1874. Museum Godeffroy, Cat- alog V. Hamburg: L. Friederichsen & Co. xxxvi + 215 pp. (February). Smith, H. H. 1902. An annotated catalogue of the shells of the genus Partula in the Hartman col- lection belonging to the Carnegie Museum. An- nals of the Carnegie Museum, 1(3): 422-485 (March); [pp. 580-582, Index]. SOLEM, A. 1976. Endodontoid Land Snails from Pacific Islands (Mollusca: Pulmonata: Sigmureth- ra): Part I. Family Endodontidae. Chicago: Field Museum of Natural History, xii -t- 508 pp. (Oc- tober). . 1983. Endodontoid Land Snails from Pa- cific Islands (Mollusca: Pulmonata: Sigmurethra): Part II. Families Punctidae and Charopidae, Zoo- geography. Chicago: Field Museum of Natural History, ix -I- 336 pp. (January). SowERBY, G. B. 1842-82. Thesaurus Conchylio- rum. Vols. 1-5. London. Thomas, W. S. 1979. A biography of Andrew Gar- rett, early naturalist of Polynesia: Part 1. Nau- tilus, 93(1): 15-28 (January). Tryon, G. W. 1866a. Note on Mr. Pease's species of Polynesian Phaneropneumona. American Journal of Conchology, 2(1): 82, pi. 5 (January). . 1866b. Descriptions of new species of Mela- niidae and Melanopisdae. American Journal of Conchology, 2(4): 299-301, pi. 20 (October). . 1872. Remarks on the genus Triphoris (Dash.) with descriptions of new species. W, Harper Pease. American Journal of Conchology, 7(3): 206 (March). . 1880. Muricinae. Manual of Conchology, (1) 2(3): 129-192 (July). . 1883. Columbellidae. Manual of Conchol- ogy, (1) 5(2): 100-128, pis. 42-49 (March). . 1884. Pleurotomidae. Manual of Conchol- ogy, (1) 6(3): 151-214 (June); 215-413, pis. 1- 34'(October). . 1885. Terebridae. Manual of Conchology, (1) 7(1): 1-64 (January). 1886. Naticidae. Manual of Conchology, (1) 8(1): 1-64, pis. 1-26 (January). TUDGE, C. 1992. Last stand for Society snails. New Scientist, 1829: 25-29 (July). Turner, R. D. 1946. John Gould Anthony with a bibliography and catalogue of his species. Oc- casional Papers on Mollusks, 1(8): 81-108 (July). War6n, a. 1980. Descriptions of new taxa of Eu- limidae (Mollusca, Prosobranchia), with notes on some previously described genera. Zoologica Scripta, 9(4): 283-306. Zilch, A. 1962. Die typen und typoide des Natur- Museums Senckenberg, 26: Mollusca, Achatinel- lacea. Archiv fiir Molluskenkunde, 91(1-3): 77- 94, pis. 2, 3 (June). 34 Bulletin Museum of Comparative Zoology, Vol. 154, No. 1 INDEX This index lists all of the names introduced by Pease included in the catalog by Clench (1975) as well as those mistakenly attributed to him by other authors. Here the species are placed under the original genera in which they were included. References to all of the numbered notes and notes additionalles made on the opisthobranch species studied by Pruvot-Fol (1947: 107-114) are included in this index, but only those mentioned by Kay (1965) or those that had shells, and were subsequently located, are mentioned in the preceding list of taxa. All taxa for which type material is extant is indicated here by **. A single * indicates that either no type material was located or that the taxon was included for some other reason such as a replacement for a preoccupied one. No * indicates that the taxa are opisthobranchs that occur only in the Clench list; several taxa followed by C were not in the latter's list. The following names mentioned here are not listed by Clench: Amastra rugulosa similaris. Helix fraterciila, Melania neivcombii, Melania oualaensis, Melania valanensis, Partula approximata, and Partula microstoma. Acanthochites "?" armattis** viridis Aclesia areola 16 producta 66 Aeolis parvula 47 semidecora 46 Alcyna lineata** rubra ** striata* stibangulata Amastra porphyrostoma** rugulosa** rugulosa similaris; not in C solida** sphaerica** Amathina bicarinata** Amphiperas semistriata** Anachis costellifera* Assiminea lateritia* lucida ** Atropis Atys costulosa 80** debilis 7** semistriata 6** Auriculella ambusta** expansa** ptdchra** triplicata** uniplicata** Avicula bru tinea** radiata** Bittium tricarinatum** Blauneria gracilis** Bornella arborescens 83 Borsonia bifasciata** corrugata* crassicostata** lutea** nebulosa* Bulimus annectens** argutus** ("?" Borus) coxi** turgidus** Bulla conspersa 79** marmorea** Bullina lauta 2** vitrea 1** Capidus liberatus* Carelia adusta angulata** olivacea** variabilis* variabilis viridis** Cassis umbilicata** Catinella rubida** Cerithium asperum** boeticum** cylindraceum** fucatum** gracile** paxillum** sculptum** tuberculiferum** unilineatum** Chondrella striata** Chromodoria godeffroijana Chromodoris inornata 102 lentiginosa 103 maculosa 99 rufomaculata 100 simplex 101 varians 104 variegata 98 Cirsotrema attenuatum** CAthara. See also Cythara brevis** [Cythara] daedalea** [Cythara] decussata** paucicostata** Citharopsis gracilis** ornata** Clathurella bait eat a** bicarinata** brunnea** buccinoides** canaliculata** cylindrica** elegans** exilis** fuscomaculata** harpa** MoLLUSKS Described by W. H. Pease • Johnson 35 mactilosa** patu'icostata** producta** ptilchella** tiimida** violacea** CoUonia "?" Candida** granulosa** multistriata* picfa** variabilis** ColumheUa cytharoides* fiisiforjins* lineata* maculosa* micans** pallida* peasei* pelliicida** pusilla** sagitta* sandwichensis** venusta* CoTius fusiformis** neglectus* parvus* purus** Coralliobia cancellata* sculptilis* Crypt ophthalmus cylindricus 56, 56 bis Cyclophorus "?" scalariformis** Cyclostonm biangidatum** parvum** tahitensis** viridans* viridescens** Cylindra formosa** Cypraea Candida** compta** fuscomaculata** granulate** Cythara. See also Cithara angiostoma* debilis* garrettii** pusilla ** strigata** varia** Daphnella bella** crenulata** curt a** interrupta** maculosa* sandwicensis** varicifera** Dentiora rubida** Diadema rotella** Distorsio pusilla** Dolabella oariegata 11** Dolabrifera fischeri* fusca 73** marniorea* olivacea 12** tahitensis 57, 57 bis* Doriopsis 100 granulosa 41 scabra 41, 81 viridis 41, 53, 53 bis Doris albopustulosa 35 cinerosa 94 compta Note 5 debilis 91, Note 4 decora 32 echinata 25 excavata 23 foetida 38 fuscescens 97 grandiflora 36 marginata 33 nubilosa 95 nucleola 31* oreosoma 23 papillata papillosa 34 picta 30 pilosa 27 prismatica imperialis 35, 39 prismatica lineata 40 propinquata 29, 35 pulchra 60 reticulata 24 rubrilineata 93 rugosa 37 scabriuscula 26 setosa 22 sordida 96 tmcfa vibrata 28, 35 villosa 90 Drillia exilis** lauta** nodifera** nodulosa** Eburnella Elysia ocellata 48 Emarginula clathrata** Endodonta celsa* Engina albocincta** costata** fusiformis** lineata maculata** monilifera** nodicostata** nodulosa** ovata** parva** striata* tuberculosa** variabilis** Erato sandtvicensis** Euchelus angulatus** corrugatus** 36 Bulletin Museum of Comparative Zoology, Vol. 154, No. 1 fimbriatus** maculosus** Eulima aciculata** exilis** inflexa ** subpelhtcida** venusta** Fastigiella squamulosa** Fissurella granifera** Fossa r garret tii* nuilticostatus** sandwichensis* Gena laevis** rosacea** Goniobranchus p. 106 albomacitlatus 63, 63 bis reticulatus 64 Haminea aperta 70** crocata 4** galba** nigropunctata 67** ovalis 68** pusilla 5** simillima 69** Helicina Bella* hrazieri** calliostoma** cincta* color ata** corrugata** discoidea ** faha** flavesceiis** guppyi* inconspicua* lenticulina* maugeriae albinea** maugeriae rubicunda** "?" multicolor* oceanica** pacifica** parvula** picta* rufescens* rugulosa** solida** straminea* stilmija** tahitensis* tenuiscula* tenuistriata* turbinella* vescoi* zigzac** Helicter hutchinsonii** proximus** Helix acetabulum** aha** capillata** congrua** consimilis** conula** decussatula* depressiformis** distans* fabrejacta** ficta** fratercula; not in C* frivola** ftiscata* laminata** lent a* marquesana** nigritella* normalis** obconica** oualanensis** parvidens** prostrata** retunsa** rugata** sculptilis** scuta* simillima** striolata* venosus** verticillata** Hexabranchus nebidosus 43 pulchellus 42 Hindsia angicostata** Histiophorus p. 99 maculatiis 51 Hyalopsis tumida* Hijdrocaena costata* elongata* robusta* Hijdrocena fragilis** nitida** ovata* Labiella compact a** pachystoma** Laimodonta conica** Lamellina laevis* serrata** Laminella erect a** Lampania baetica* Latirus gibbus** granulosus** liratus** squamosus** Leiostraca distorta** Leptachatina antiqua** balteata** brevicula** costulosa** cylindrata** extensa* laevis** lucida* simplex** tenebrosa** tenuicostata** MoLLUSKs Described by W. H. Pease • Johnson 37 turgidula** Leptothijra costata** Lihraiula plana** Lirunaea amhigiia** compacta** turgidula** Littorina cinerea** Lobifera p. 101 nigricans 61 papillosa 62 Lobiger picta 58, 58 bis viridis 58** Lophocerctis viridis 59, 59 bis* Magilus soliduscula** Margarita marmorea** Marginella cijlindracea** cijlindrica* debilis* orijza* pacifica** paumotensis* peasei* polita* pyriformis** sandwicensis** Melampus cinctus* fusciis* liicidiis** (Tralia) semiplicata** {Tralia) striatus** Melania contigua** kauaiensis** newcombii; not in C* oualaensis., not in C** tahitensis* valanensis; not in C* Melibe pilosa Mitra assimilis** ericea** flammulata* glabra** lubrica* netvcombii* nigricans** ordinata* pallida* pudica** saltata** sectilis* Mitroidea multiplicata* Mitropsis ftisiformis** Mucronalia gracilis** nitidula** ovata** rosea** Murex foveolatus** garettii* Narica delicata** granifera** Nassa approximata** balteata** gracilis** microstoma** nticea* obliqua** Natica undtdata* Neptunea fuscolineata** Nerita maculata** Neritina deshaijesii* dispar* neglecta** rubida** rudis** Neritopsis interlirata** Nucula sculpta* Odostomia debilis** gracilis** polita** rosacea** rubra** striata** Oliva sandwicensis** Olivella (Callianax) simplex* Omphalius turbinatus** Omphalotropis moussoni* nebulosa** producta** robust a** Operculatum aurantium* Pachypoma virescens** Partula abbreviata* ajfinis** affinis dubia* alternata** approximata; not in C** assimilis** attenuata** bella* bicolor** bilineata** brazieri** brevicula** castanea** citrina** clara** cognata** compacta** concinna* crassa** crassilabris** crassilabrun\* elongata** expansa** faba subangulata** fasciata** formosa** fusca ** garrettii** globosa* gracilior** gracilis** imperforata** 38 Bulletin Museum of Comparative Zoology, Vol. 154, No. 1 labiata** lignaria** lineolaia** Itigubris** megastoma** microstoma; not in C** nucleola** obesa* otaheitana dubia** ovalis** peUucida * perplexa** perversa** planilabrum** producta** propinqua** protea* pulchra* pulchra varia* radio t a** recta** robust a** rustica** simulans** sinistralis* sinistrorsa** stolida** strigata** striolata (2)** subangidata** suturalis* terrestris** trilineata** turricula** umbilicata** variabilis** vexillum** virginea** virgulata** vittata** Parttilina compta** Patella sandwicensis** undatolirata* Pedicularia pacifica** Pedipes sandwicensis** Perdicella Perna hatvaiiensis* Philinopsis p. 99 nigra 10* speciosa 9** Phyllidia nigra 78 Pinna trigonalis** Pisania strigata** Pithy s analogica** atien-sis** celsa ** imperforata** paucicostata** rarotongensis* roratongensis** rotellina** verecunda* Pitys filocostata* fratercula* Placobranchus gracilis 85 variegatus 86 Planaxis abbreviata** atra** fasciata** plumbea** Pleurobranchus delicatus 54, 53 bis* grandis 76 marginatus 18* ovalis 11 pelhicidus 17 reticulatus 21* rufus 19** tessellatus* varians 20* violaceiis 21* Pleurotoma lirata* monilifera* Polybranchia p. 100 pelhicida 52* Pterocyclos parva** Pterogasteron p. 104 belliim 50 marginatus 87 nigropunctatus 89 ornatum 49 rujescens 88 Pupoidea Purpura marmorata** Ranella luteostoma* producta** Realia abbreviata** affinis** costata** elongata** laevis** ochrostoma** producta** scalariformis** variabilis** Registoma complanatum** Rhizochilus exaratus** Rissoa flammea** gracilis** semicostata* Rissoina angasii* balteata* cerithiopsis* costulata** granulosa** semiplicata** striatula** tenuistriata** triticea** turricula** Scalaria crenulata** crispata** decussata** MoLLUSKS Described by W. H. Pease • Johnson 39 fucata** millecostata** patimotensis** perplexa** symmetrica** iimhilicata** Scutellina aculeata** cancellata** compressa** granocostata** Seminella Siphonaria depressa** Sistrum affine** rugulosum** squamosum** striatum** triangulatum* Stenodoris p. 99 rubra 65 Strigatella brunnea** fiiscescens** picea** Strombus cancellatus* Stylifer deformis** robustus* Succinea costulosa* elongata** labiata** mammillata** ovata* rubella** rutella* Syphonota p. 104 bipes 13** elongata 15** grandis 14** punctata 75** viridescens 74 Taheitea pallida** Tectura conoidalis** radiata** tahitensis** Terebra assimilis* contigua* costellifera** lauta** propinqua* rosacea * sculptilis** suffusa** sulcata** swainsonii inflexa** Thala alba** angiostoma** Torinia conica** discoidea ** sulci f era* Tornatellides vitrea* Tornatellina aperta** dentata** gracilis** nit Ida** oblonga ** simplex** Tornatina sandwicensis 3** Trevelyana picta 84 Triforis marmorata** Triopa "?" gracilis 82 Triphoris affinis** alternata** bicolor* brunneus** cingulifera** clavata** costatus** cylindricus* flammulata** fucata** gracilis** granosus** incisa** maculatus** minimus** oryza** pallidus** peasei* prefectus** punctatus* pustulosus** robustus** similis** sulcosus** triticea** tuberculatus** Triton cylindricus** intermedius** pusilla** Tritonia hawaiiensis 44 Tritonidea australis* Trivia corrugata** Trochomorpha contigua* fuscata** nigritella oppressa** trochiformis pollens** Trochus conoidalis* exilis** marmoreus** Truella Truncatella concinna** costellifera* cylindracea* cylindrica * pacifica** Tugalia oblonga** Turbo multistriatus* sandwicensis** semicostatus** Turbonilla decussata** elongate** Turcica coreensis** Turricula approximata** bella** 40 Bulletin Museum of Comparative Zoology, Vol. 154, No. 1 fortipHcata** ""^"^ „ modesta** perlonga** nodulosa** simplaria* phcatula** stnatula** putillus** Vexilla fusconigra** Turns monilifera** ^'f" depressiformis* Vanikoro imbricata** jusca semiplicata** subviridis** Vertagus graniferus** Vitularia sandwicensis** Vertigo arnmta** J^« "^^/ « ''"""fT.o.. Ijacca* Volvatella Candida 72** costata** f'^'Si'ts 8, 8 bis** costulosa** pyriformis 71** dentijera** 42 Bulletin Museum of Comparative Zoology, Vol. 154, No. 1 Plate 1 Figure 1. "?" Acanthochltes armatus Pease. [Pauloa] Oahu [Hawaiian Islands]. Holotype MCZ 73452. Length 10 mm, width 6 mm. Figure 2. Pinna trigonalis Pease. Kingsmill Islands [Kiribati]. Lectotype MCZ 225951 . Length 217 mm, width 94 mm. Figure 3. Avicula brunnea Pease. [Moiokai] Sandwich [Hawaiian] Islands. Lectotype MCZ 298465. Length 141 mm, width 73 mm. Figure 4. Avicula radiata Pease. Kingsmill Islands [Kiribati]. Lectotype MCZ 298466. Length 84.5 mm, width 42 mm. Figure 5. Siphonaria depressa Pease. Apaiang [Abaiang] Island [Kiribati]. Lectotype ANSP 22199. Length 18 mm, width 14 mm, height 3 mm. Figure 6. Tectura conoidalis Pease. Roratonga [Rarotonga Island, Cook Islands] labeled as from the Paumotus. Lectotype MCZ 302558. Length 18 mm, width 14 mm, height 7 mm. Figure 7. Tectura tahitensis Pease. Tahiti [Society Islands]. Lectotype ANSP 38949. Length 38 mm, width 12 mm, height 3.5 mm. Figure 8. Patella sandwicensis Pease. Sandwich [Hawaiian] Islands. Lectotype MCZ 304058. Length 37 mm, width 30 mm, height 18.5 mm. Figure 9. Scutellina aculeata Pease. Hawaii [Hawaiian Islands]. Lectotype MCZ 83720. Length 4.5 mm, width 4 mm, height 3.4 mm. Figure 10. Scutellina compressa Pease. Tahiti [Society Islands]. Lectotype MCZ 302552. Length 4 mm, width 1.5 mm, height 1.4 mm. MoLLUSKS Described by W. H. Pease • Johnson 43 Plate 1 44 Bulletin Museum of Comparative Zoology, Vol. 154, No. 1 Plate 2 Figure 1. Helix capillata Pease. Kauai, Sandwich [Hawaiian] Islands. Lectotype ANSP 1975a. Width 4 mm, height 2 mm. Figure 2. Helix striolata Pease. Ebon Island [Atoll], Marshall [Islands]. Lectotype MCZ 11563. Width 3 mm, height 2 mm. Figure 3. Helicina pacifica Pease. Oualan [Kusaie or Kosrae Island, Caroline Islands]. Lectotype ANSP 14443. Width 6 mm, height 4 mm. Figure 4. Helix congrua Pease. Ponape [Pohnpei Island, Caroline Islands]. Lectotype MCZ 12161. Width 11.5 mm, height 8.5mm. Figure 5. Trochomorpha trochiformis pallens Pease. Moorea [Society Islands]. Lectotype MCZ 12188. Width 18 mm, height 12 mm. Figure 6. Trochomorpha nigritella oppressa Pease. Ponape [Pohnpei Island, Caroline Islands]. Lectotype MCZ 12187. Width 15 mm, height 8.5 mm. Figure 7. Pithys paucicostata Pease. Kauai [Hawaiian Islands]. Lectotype MCZ 17271. Width 3.5 mm, height 2 mm. Figure 8. Pithys atiensis Pease. Atiu [Island, Cook Islands]. Lectotype MCZ 17335. Width 3 mm, height 1.5 mm. Figure 9. Auhculella triplicata Pease. Oahu [Hawaiian Islands]. Lectotype MCZ 45150. Width 5 mm, height 10 mm. Figure 10. Auhculella ambusta Pease. [Waianae Mountains, Oahu, Hawaiian Islands]. Lectotype MCZ 45152. Width 5 mm, height 8.5 mm. Figure 11. Amastra rugulosa similahs Pease. [Waimea] Kauai [Hawaiian Islands]. Lectotype MCZ 45253. Width 8 mm, height 13,5 mm. Figure 12. Bulimus ("?" Borus) coxi Pease. [New Hebrides Islands]. Holotype MCZ 86495. Width 15 mm, length 25 mm. Figure 13. Cyclostoma biangulatum Pease. Aitutaki [Atoll], Cook Islands. Lectotype MCZ 141031. Width 2 mm, height 3 mm. Figure 14. Carelia olivacea Pease. Kauai [Hawaiian Islands]. Holotype MCZ 571 14. Width 18 mm, height 41 mm. It is assumed that of the original measurements given by Pease — width 19 mm, height 69 mm — the latter is a misprint. Figure 15. Cyclostoma parvum Pease. [Tahiti, Society Islands]. Lectotype MCZ 74949. Width 2 mm, height 2.5 mm. Figure 16. Lamellina laevis Pease. Hervey Islands [Cook Islands]. Lectotype MCZ 154942. Width 1.5 mm, height 3.15 mm. Figure 17. Lamellina serrata Pease. Ebon Island [Atoll], Marshall [Islands]. Lectotype MCZ 302555. Width 1.5 mm, height 3 mm. Figure 18. Tornatellina aperta Pease. Tahiti [Society Islands]. Lectotype MCZ 175722. Width 2.5 mm, height 3 mm. Figure 19. Chondrella striata Pease. Roratonga [Rarotonga Island, Cook Islands]. Lectotype MCZ 187917. Width 1.5 mm, height 2 mm. Figure 20. Vertigo costata Pease. [Kona] Hawaii [Hawaiian Islands]. Holotype MCZ 45238. Width 1 mm, height 2 mm. Figure 21. Vertigo nitens Pease. Ebon Island [Atoll], Marshall [Islands]. Lectotype MCZ 151650. Width 1 mm, height 2 mm. Figure 22. Vertigo costulosa Pease. Hawaii [Hawaiian Islands]. Lectotype MCZ 45244. Width 1 mm, height 1.5 mm. MoLLUSKS Described by W. H. Pease • Johnson 45 Plate 2 46 Bulletin Museum of Comparative Zoology, Vol. 154, No. 1 Plate 3 Figure 1. Helix depressiformis Pease. Tahiti [Society Islands]. Lectotype MCZ 17342. Width 7 mm, height 2.5 mm. Figure 2. Helix fabrefacta Pease. [Raiatea, Society Islands]. Lectotype MCZ 17238. Width 8 mm, height 4 mm. Figure 3. Helix rugata Pease. Maui [Hawaiian Islands]. Lectotype MCZ 17237. Width 5 mm, height 3 mm. Figure 4. Helix laminate Pease. Kauai [Hawaiian Islands]. Lectotype MCZ 17233. Width 7 mm, height 2.5 mm. Figure 5. Helix marquesana Pease. Marquesas [Islands]. Lectotype MCZ 302557. Width 10 mm, height 7.5 mm. Figure 6. Helix conula Pease. Tahiti [Society Islands]. Lectotype MCZ 297945. Width 6.5 mm, height 5 mm. Figure 7. Helix verticillata Pease. Moorea [Society Islands]. Holotype ANSP 49284. Width 5 mm, height 3 mm. Figure 8. Helix oualanensis Pease. Oualan [Kusaie or Kosrae Island, Caroline Islands]. Lectotype ANSP 47763. Width 5.5 mm, height 2 mm. Figure 9. Helix normalis Pease. Moorea [Society Islands]. Lectotype MCZ 11543. Width 4.5 mm, height 3 mm. MoLLUSKS Described by W. H. Pease • Johnson 47 Plate 3 48 Bulletin Museum of Comparative Zoology, Vol. 154, No. 1 Plate 4 Figure 1. Helicter proximus Pease. Molokai [Hawaiian Islands]. Lectotype MCZ 25823. Width 15 mm, height 29 mm. Figure 2. Helicter hutchinsonii Pease. Maui [Hawaiian Islands]. Lectotype MCZ 45254. Width 7 mm, height 16 mm. Figure 3. Registoma complanatum Pease. Ebon Island [Atoll], Marshall [Islands]. Lectotype MCZ 141018. Width 3 mm, height 6 mm. Figure 4. Omphalotropis nebulosa Pease. [Makela = San Cristobal Island] Solomon Islands. Lectotype MCZ 72347. Width 5.5 mm, height 9 mm. Figure 5. Labiella compacta Pease. [Palauea] Maui [Hawaiian Islands]. Lectotype MCZ 45196. Width 4.5 mm, height 9 mm. Figure 6. Tornatellina oblonga Pease. Tahiti [Society Islands]. Lectotype MCZ 154941. Width 1.5 mm, height 3 mm. Figure 7. Tornatellina nitida Pease. Ebon Island [Atoll], Marshall [Islands]. Lectotype MCZ 28921. Width 2 mm, height 4 mm. Figure 8. Tornatellina simplex Pease. Tahaa [Island, Society Islands]. Lectotype MCZ 175745. Width 2 mm, height 3 mm. Figure 9. Tornatellina dentata Pease. Hawaii [Hawaiian Islands]. Lectotype MCZ 28918. Width 1 mm, height 2.5 mm. Figure 10. Tornatellina gracilis Pease. Kauai [Hawaiian Islands]. Lectotype MCZ 302554. Width 1.5 mm, height 4 mm. Figure 11. Limnaea turgidula Pease. Oahu [Hawaiian Islands]. Lectotype MCZ 298901. Width 8 mm, height 14.5 mm. Figure 12. Succinea mammlllata Pease. Nukahiva [Nuku Hiva, Marquesas Islands]. Lectotype MCZ 155145. Width 7.5 mm, height 12.5 mm. Figure 13. Succinea rubella Pease. Lanai [Hawaiian Islands]. Lectotype MCZ 161671. Width 7.5 mm, height 11.5 mm. Figure 14. Succinea elongata Pease. Kauai [Hawaiian Islands]. Holotype MCZ 161665. Width 9 mm, height 15.5 mm. Figure 15. Succinea costulosa Pease. Tahiti [Society Islands]. Lectotype MCZ 31406. Width 5 mm, height 7 mm. Figure 16. Succinea labiata Pease. Raiatea [Society Islands]. Lectotype MCZ 298909. Width 12 mm, height 19.5 mm. Figure 17. Haminea aperta Pease. Tahiti [Society Islands]. Lectotype ANSP 57575. Width 11 mm, height 15 mm. Figure 18. Lobiger viridis Pease. Tahiti [Society Islands]. Lectotype MCZ 297869. Width 8 mm, height 12 mm. Figure 19. Melania contigua Pease. Kauai [Hawaiian Islands]. Lectotype MCZ 74887. Width 10 mm, height 20 mm. Figure 20. Melania oualanensis 'Pease' Tryon. Oualan [Kusaie or Kosrae Island, Caroline Islands]. Lectotype ANSP 26274. Width 12 mm, height 29 mm. Figure 21 . Catinella rubida Pease. Kauai [Hawaiian Islands]. Lectotype MCZ 45252. Width 9 mm, height 1 1 mm. Figure 22. Melampus lucidus Pease. [Honolulu] Oahu [Hawaiian Islands]. Lectotype ANSP 22284. Width 3.5 mm, height 5 mm. Figure 23. Haminea ovalis Pease. Tahiti [Society Islands]. Lectotype MCZ 297877. Width 8 mm, height 10 mm. Figure 24. Haminea simillima Pease. Tahiti [Society Islands]. Lectotype MCZ 297875. Width 8 mm, height 10 mm. MoLLUSKS Described by W. H. Pease • Johnson 49 Plate 4 50 Bulletin Museum of Comparative Zoology, Vol. 154, No. 1 Plate 5 Figure 1 . Scutellina granocostata Pease. Hawaii [Hawaiian Islands]. Lectotype MCZ 83723. Width 6 mm, length 8 mm, height 3 mm. Figure 2. Libratula plana Pease. Islands of the Central Pacific. Lectotype MCZ 1 1 91 62. Width 3 mm, length 7 mm, height 1 mm. Figure 3. Heliclna solida Pease. [Tahiti, Society Islands]. Lectotype MCZ 297923. Width 8 mm, height 5 mm. Figure 4. Helicina corrugata Pease. Raiatea [Society Islands]. Lectotype MCZ 297925. Width 5 mm, height 3 mm. Figure 5. Helicina calliostoma Pease. Marquesas [Islands]. Lectotype MCZ 74950. Width 9 mm, height 6 mm. Figure 6. Helicina discoidea Pease. Tahaa [Island, Society Islands]. Holotype ANSP 14398. Width 6 mm, height 3 mm. Figure 7. Helicina rugulosa Pease. Tahaa [Island, Society Islands]. Lectotype MCZ 297922. Width 3 mm, height 2 mm. Figure 8. Helicina zigzac Pease. Oualan [Kusaie or Kosrae Island, Caroline Islands]. Lectotype ANSP 14485. Width 6 mm, height 4.5 mm. Figure 9. Helicina oceanica Pease. Kingsmill [Islands, Kiribati]. Lectotype MCZ 176561. Width 4 mm, height 3 mm. Figure 10. Petrocyclos parva Pease. Aitutake [Aitutaki Atoll], Hervey Islands [Cook Islands]. Lectotype ANSP 13406. Width 3 mm, height 3 mm. Figure 1 1 . Helicina brazieri Pease. Niue [Island; 19°S, 170°W]. Lectotype MCZ 74947. Width 6 mm, height 5 mm. Figure 12. Helicina maugeriae rubicunda Pease. [Tahiti, Society Islands]. Lectotype ANSP 14512. Width 13 mm, height 8 mm. MoLLUSKS Described by W, H. Pease • Johnson 51 ^gf JP '^^JI^SJpT-v-K . . »'>5^ V .^iltaai9>^ >' #11 r: i A X 8 # ^iL ^K .\ ^- "^"^ .V 11 12 Plate 5 52 Bulletin Museum of Comparative Zoology, Vol. 154, No. 1 Plate 6 Figure 1. Gena rosacea Pease. Paumotus [Tuamotu Archipelago]. Lectotype ANSP 40756. Length 8.5 mm, width 5 mm. Figure 2. Gena laevls Pease. Tahiti [Society islands]. Lectotype ANSP 40754. Length 9 mm, width 5.5 mm. Figure 3. Cassis umbllicata Pease. Sandwich [Hawaiian] Islands. Lectotype MCZ 298907. Width 43 mm, height 64.5 mm. Figure 4. Alcyna lineata Pease. [Puuioa] Oahu [Hawaiian Islands]. Holotype MCZ 31724. Width 1.5 mm, height 2 mm. Figure 5. Atys costulosa Pease. [Waimalu] Oahu [Hawaiian Islands]. Holotype MCZ 31714. Width 2 mm, height 3 mm. Figure 6. Cerlthium cylindraceum Pease. Paumotus [Tuamotu Archipelago]. Paratype MCZ 297939. Width 10 mm, height 20.5mm. Figure 7. Cithara [Cythara] decussata Pease. [Anaa Island] Paumotus [Tuamotu Archipelago]. Lectotype ANSP 1 5651 . Width 4 mm, height 9 mm. Figure 8. Cithara [Cytt^ara] paucicostata Pease. Tahiti [Society Islands]. Lectotype MCZ 231925. Width 3.5 mm, height 8 mm. Figure 9. Cythara strigata Pease. [Howland Island; 0°49N, 176°40'W]. Lectotype MCZ 49990. Width 3 mm, height 7 mm. Figure 10. Cithara [Cythara] brevis Pease. [Anaa Island] Paumotus [Tuamotu Archipelago]. Lectotype ANSP 15652. Width 3.5 mm, height 7 mm. Figure 1 1 . Cithara [Cythara] daedalea Pease. Paumotus [Tuamotu Archipelago). Lectotype ANSP 15663. Width 2 mm, height 4.5 mm. Figure 12. Citharopsis ornata Pease. Tahiti [Society Islands]. Lectotype ANSP 16919. Width 1.5 mm, height 3.5 mm. Figure 13. Citharopsis gracilis Pease. Paumotus [Tuamotu Archipelago]. Lectotype ANSP 16921 . Width 1 .5 mm, height 4 mm. Figure 14. Clathurella maculosa Pease. Paumotus [Tuamotu Archipelago]. Lectotype ANSP 48693. Width 2 mm, height 5 mm. Figure 1 5. Clathurella violacea Pease. Paumotus [Tuamotu Archipelago]. Lectotype MCZ 231 971 . Width 2 mm, height 4.5 mm. Figure 16. Clathurella canaliculata Pease. Paumotus [Tuamotu Archipelago]. Lectotype MCZ 231978. Width 5 mm, height 10mm. Figure 17. Clathurella bicarinata Pease. Kingsmill Islands [Kiribati]. Lectotype ANSP 15806. Width 5 mm, height 10 mm. Figure 1 8. Clathurella tumida Pease. Paumotus [Tuamotu Archipelago]. Lectotype ANSP 1 581 3. Width 2.5 mm, height 6.5 mm. Figure 19. Columbella micans Pease' Tryon. Paumotus [Tuamotu Archipelago]. Lectotype MCZ 304063. Width 5.5 mm, height 10 mm. Figure 20. Collonia granulosa Pease. Ponape [Pohnpei Island, Caroline Islands]. Lectotype MCZ 245264. Width 3 mm, height 2.5 mm. Photographed by James H. McLean. Figure 21 Collonia picta Pease. [Anaa Island] Paumotus [Tuamotu Archipelago]. Lectotype MCZ 245268. Width 4.5 mm, height 4 mm. Photographed by James H. McLean. Figure 22. Columbella sandwichensis Pease. Sandwich [Hawaiian] Islands. Lectotype MCZ 297950. Width 8 mm, height 1 2 mm. Figure 23. Conus purus Pease. Niihau Island [Hawaiian Islands]. Holotype MCZ 72331. Width 22.5, height 41 mm. MoLLUSKS Described by W. H. Pease • Johnson 53 I % 12 ''* 14 ^ 16 17 19 Plate 6 54 Bulletin Museum of Comparative Zoology, Vol. 154, No. 1 Plate 7 Figure 1. Cylindra formosa Pease. Ascension [Pohnpei Island, Caroline Islands]. Holotype MCZ 260605. Width 6 mm, height 14 mm. Figure 2. Daphnella curta Pease. Paumotus [Tuamotu Archipelago]. Lectotype ANSP 16956. Width 2 mm, height 4 mm. Figure 3. Daphnella crenulata Pease. [Howland Island; 0°49'N, 176°40'W]. Lectotype ANSP 1 5694. Width 2 mm, height 7 mm. Figure 4. Engina ovata Pease. Howland Island [0°49'N, 176°40'W]. Lectotype ANSP 34536. Width 8 mm, height 12 mm. Figure 5. Engina variabilis Pease. Paumotus [Tuamotu Archipelago]. Lectotype MCZ 260618. Width 4 mm, height 7 mm. Figure 6. Daphnella sandwicensis Pease. Sandwich [Hawaiian] Islands. Lectotype MCZ 49993. Width 3.5 mm, height 8 mm. Figure 7. Dentlora rubida Pease. Sandwich [Hawaiian] Islands. Lectotype MCZ 297951. Width 2 mm, height 3 mm (2 views). Figure 8. Drillia lauta Pease. [Anaa Island] Paumotus [Tuamotu Archipelago]. Lectotype ANSP 15692. Width 4 mm, height 8.5mm. Figure 9. Eulima inflexa Pease. Paumotus [Tuamotu Archipelago]. Lectotype ANSP 59334. Width 4 mm, height 10 mm. Figure 10. Eulima venusta Pease. Tahiti [Society Islands]. Lectotype ANSP 19788. Width 2 mm, height 6 mm. Figure 11. Eulima exilis Pease. Paumotus [Tuamotu Archipelago]. Lectotype MCZ 187831. Width 4 mm, height 9 mm. Figure 12. Eulima subpellucida Pease. Tahiti [Society Islands]. Lectotype MCZ 31709. Width 7 mm, height 16 mm. Figure 13. Euchelus angulatus Pease. Annaa [Anaa] Island [Tuamotu Archipelago]. Lectotype ANSP 40671. Width 5 mm, height 5 mm. Figure 14. Fastigiella squamulosa Pease. [Anaa Island] Paumotus [Tuamotu Archipelago]. Lectotype ANSP 36702. Width 1 1 mm, height 27 mm. Figure 15. Latirus liratus Pease. Marquesas [Islands]. Lectotype MCZ 302628. Width 15 mm, height 28.5 mm. Figure 16. Latirus gibbus Pease. Howland Island [0°49'N, 176°40'W]. Lectotype MCZ 261182. Width 8 mm, height 13 mm. Figure 1 7. Latirus granulosa Pease. Paumotus [Tuamotu Archipelago]. Lectotype MCZ 261 1 81 . Width 1 1 mm, height 20.5 mm. Figure 18. Mitra saltata Pease. Islands of the Central Pacific. Lectotype MCZ 260613. Width 3 mm, height 7.5 mm. Figure 19. Leptothyra costata Pease. Maui [Hawaiian Islands]. Lectotype MCZ 245261 . Width 4 mm, height 4 mm. Figure 20. Mitra assimilis Pease. [Huahine [sic] Island, Society Islands]. Lectotype ANSP 28718. Width 7 mm, height 17 mm. Figure 21 . Marginella pyriformis Pease. Paumotus [Tuamotu Archipelago]. Lectotype ANSP 29541 . Width 2 mm, height 4 mm. Figure 22. Marginella pacifica Pease. Paumotus [Tuamotu Archipelago]. Lectotype ANSP 2941 5. Width 2.5 mm, height 4 mm. Figure 23. Marginella paumotensis Pease. Paumotus [Tuamotu Archipelago]. Lectotype ANSP 29497. Width 2.5 mm, height 5 mm. Figure 24. Mucronalia gracilis Pease. Tahiti [Society Islands]. Lectotype MCZ 288506. Width 2 mm, height 4 mm. Figure 25. Engina nodicostata Pease. Paumotus [Tuamotu Archipelago]. Lectotype MCZ 260614. Width 3.5 mm, height 6 mm. Photographed by Kenneth J. Boss. MoLiAJSKS Described by W. H. Pease • Johnson 55 ^m 1 / I 18 Plate 7 56 Bulletin Museum of Comparative Zoology, Vol. 154, No. 1 Plate 8 Figure 1 . Terebra sulcata Pease. [Honolulu] Oahu [Hawaiian Islands]. Lectotype MCZ 49967. Width 5 mm, height 22.5 mm. Figure 2. Narica granifera Pease. Jarvis [Island, Line Islands. 0°22'33"S, 159°54'11"W]. Lectotype ANSP 37301. Width 8.5 mm, height 9 mm. Figure 3. Narica delicate Pease. Paumotus [Tuamotu Archipelago]. Lectotype ANSP 40201. Width 6.5 mm, height 6 mm. Figure 4. Nassa approximata Pease. [Ascension (Pohnpei) Island, Caroline Islands]. Lectotype MCZ 151800. Width 9.8 mm, height 24.5 mm. Figure 5. Nassa gracilis Pease. Ascension [Pohnpei Island, Caroline Islands]. Lectotype MCZ 228822. Width 11 mm, height 19.5 mm. Figure 6. Nassa obliqua Pease. Islands of the Central Pacific. Lectotype MCZ 228823. Width 12 mm, height 15.5 mm. Figure 7. Nehta maculata Pease. Tahiti [Society Islands]. Lectotype ANSP 37490. Width 15.6 mm, height 16.5 mm. Figure 8. Neritina dispar Pease. Roratonga [Rarotonga, Cook Islands]. Lectotype ANSP 37714. Width 8 mm, height 9 mm. Figure 9. Neritopsis interlirata Pease. Annaa [Anaa] Island, [Tuamotu Archipelago]. Lectotype MCZ 73479. Width 10 mm, height 10 mm. Figure 10. Odostomia pollta Pease. Tahiti [Society Islands]. Lectotype MCZ 10562. Width 7 mm, height 7 mm. Figure 1 1 . Odostomia rosacea Pease. Paumotus [Tuamotu Archipelago]. Lectotype ANSP 1 9959. Width 2 mm, height 6 mm. Figure 12. Odostomia rubra Pease. [Anaa Island] Paumotus [Tuamotu Archipelago]. Lectotype ANSP 19955. Width 2 mm, height 8 mm. Figure 13. Planaxis abbreviate Pease. Tahiti [Society Islands]. Paralectotype MCZ 187833. Width 6 mm, height 9.5 mm. Figure 14. Odostomia striata Pease. Paumotus [Tuamotu Archipelago]. Lectotype MCZ 10491 . Width 1 .5 mm, height 4.5 mm. Figure 15. Odostomia debilis Pease. Howland [Island; 0°49N, 176°40W]. Lectotype MCZ 297933. Width 2 mm, height 9 mm. Figure 16. Olivella {Callianax) simplex Pease. Paumotus [Tuamotu Archipelago]. Lectotype ANSP 28969. Width 2 mm, height 4.5 mm. Figure 17. Pachypoma virescens Pease. Tarawa Island [Kiribati]. Lectotype MCZ 302624. Width 26 mm, height 28 mm. Figure 1 8. Terebra sculptilis Pease. [Honolulu] Oahu [Hawaiian Islands]. Lectotype MCZ 248804. Width 4.5 mm, height 24 mm. Figure 1 9. Purpura marmorata Pease. Apaian [Abaiang] Island [Kiribati]. Lectotype MCZ 1 77824. Width 27 mm, height 45 mm. Figure 20. Omphialius turbinatus Pease. Gulf of [Golfo de] California, La Paz [Baja California Sur, Mexico]. Lectotype ANSP 40820. Width 16.5 mm, height 14 mm. MoLLUSKS Described by W. H. Pease • Johnson 57 Plate 8 58 Bulletin Museum of Comparative Zoology, Vol. 154, No. 1 Plate 9 Figure 1. Planaxis atra Pease. Marquesas [Islands]. Lectotype ANSP 18282. Width 5.5 mm, height 9 mm. Figure 2. Planaxis abbreviata Pease. Tahiti [Society Islands]. Lectotype ANSP 18261 . Width 6 mm, height 8 mm. Figure 3. Rissoa gracilis Pease. [Kauai] Sandwich [Hawaiian] Islands. Lectotype MCZ 178853. Width 1 mm, height 2.5 mm. Figure 4. Rissolna striatula Pease. Paumotus [Tuamotu Archipelago]. Lectotype MCZ 178844. Width 4 mm, height 8 mm. Figure 5. Rissolna tenulsthata Pease. Paumotus [Tuamotu Archipelago]. Lectotype ANSP 19253. Width 4 mm, height 9 mm. Figure 6. Rissolna costulata Pease. Paumotus [Tuamotu Archipelago]. Lectotype ANSP 19241. Width 2 mm, height 4 mm. Figure 7. Rissoa flammea Pease. Caroline [Islands]. Lectotype MCZ 178868. Width 1 mm, height 2.68 mm. From Ponder and de Keyzer. Figure 8. Slstrum rugulosum Pease. Howland [Island; 0°49'N, 176°40'W]. Lectotype ANSP 36740. Width 5.5 mm, height 8 mm. Figure 9. Slstrum striatum Pease. Kingsmill [Islands, Kiribati]. Lectotype ANSP 36735. Width 8.5 mm, height 16.5 mm. Figure 10. Strlgatella brunnea Pease. Polynesia [(Paumotus) Tuamotu Archipelago]. Lectotype ANSP 29722. Width 10.5 mm, height 21.5 mm. Figure 11. Scalaria perplexa Pease. Hawaii [Hawaiian Islands]. Lectotype MCZ 181978. Width 12.5 mm, height 27 mm. Figure 12. Scalaria umblllcata Pease. Oahu [Hawaiian Islands]. Holotype MCZ 187393. Width 4 mm, height 9.5 mm. Figure 13. Scalaria chspata Pease. Paumotus [Tuamotu Archipelago]. Lectotype ANSP 19575. Width 7 mm, height 16 mm. Figure 14. Scalaria decussata Pease. Hawaii [Hawaiian Islands]. Lectotype ANSP 19585. Width 4 mm, height 10 mm. Figure 1 5. Styllfera deformis Pease. Paumotus [Tuamotu Archipelago]. Lectotype ANSP 19836. Width 4 mm, height 1 1 .5 mm. Figure 1 6. Scalaria paumotensis Pease. Paumotus [Tuamotu Archipelago]. Lectotype ANSP 1 9581 . Width 5 mm, height 9 mm. Figure 17. Taheltea [sic] pallida Pease. Tahiti; Huaheine [both Society Islands; localities not differentiated on label]. Lectotype ANSP 12659. Width 2.5 mm, height 6.5 mm. Figure 18. Terebra suffusa Pease. [Honolulu] Oahu [Hawaiian Islands]. Lectotype MCZ 248802. Width 10 mm, height 30 mm. Figure 19. Terebra costelllfera Pease. [Honolulu] Oahu [Hawaiian Islands]. Lectotype MCZ 248800. Width 6 mm, height 20 mm. Figure 20. Thala alba Pease. Paumotus [Tuamotu Archipelago]. Lectotype ANSP 28755. Width 2.5 mm, height 8 mm. Figure 21 . Terebra lauta Pease. Oahu [Hawaiian Islands]. Holotype ANSP 33589. Width 5.5 mm, height 20 mm. Figure 22. Thala angiostoma Pease. Paumotus [Tuamotu Archipelago]. Lectotype ANSP 28754. Width 4 mm, height 12 mm. Figure 23. Slstrum triangulatum Pease. Hawaii [Island]. Lectotype MCZ 295580. Width 14 mm, height 24 mm. Photographed by Kenneth J. Boss. Figure 24. Torinia discoldea Pease. Paumotus [Tuamotu Archipelago]. Lectotype ANSP 38804. Width 5.5 mm, height 3 mm. Figure 25. Trochus marmoreus Pease. Paumotus [Tuamotu Archipelago]. Lectotype ANSP 4061 4. Width 4.5 mm, height 6 mm. Figure 26. Trochus conoidalis Pease. Paumotus [Tuamotu Archipelago]. Lectotype ANSP 18868. Width 4 mm, height 4 mm. MoLLUSKS Described by W. H. Pease • Johnson 59 60 Bulletin Museum of Comparative Zoology, Vol. 154, No. 1 Plate 10 Figure 1 . Triphoris costatus Pease. Annaa [Anaa Island, Tuamotu Archipelago]. Lectotype MCZ 273206. Width 8 mm, height 25 mm. Figure 2. Triphoris pustulosus Pease. Kauai [Hawaiian Islands]. Lectotype MCZ 50077. Width 5.3 mm, height 8 mm. Figure 3. Triphoris sulcosus Pease. Kauai [Hawaiian Islands]. Lectotype MCZ 50080. Width 2 mm, height 8 mm. Figure 4. Triphoris tuberculatus Pease. Kauai [Hawaiian Islands]. Lectotype MCZ 50081. Width 2 mm, height 8 mm. Figure 5. Triforis marmorata 'Pease' Martens and Langkavel. [Kauai] Sandwich [Hawaiian] Islands. Lectotype MCZ 50055. Width 3 mm, height 10.5 mm. Figure 6. Triphoris brunneus Pease. Apaiang [Abaiang] Island [Kiribati]. Lectotype MCZ 73922. Width 2 mm, height 7 mm. Figure 7. Triphoris minimus Pease. [Haena] Kauai [Hawaiian Islands]. Lectotype MCZ 50071. Width 1 mm, height 3 mm. Figure 8. Triphoris pallidus Pease. Kauai [Hawaiian Islands]. Lectotype MCZ 50073. Width 1 .5 mm, height 6.5 mm. Figure 9. Triphoris granosus Pease. Tahiti [Society Islands]. Lectotype MCZ 273207. Width 2 mm, height 7 mm. Figure 10. Triphohs gracilis Pease. Kauai [Hawaiian Islands]. Lectotype MCZ 50058. Width 1.5 mm, height 7.5 mm. Figure 11. Triphoris oryza Pease. Kauai [Hawaiian Islands]. Lectotype MCZ 50072. Width 2 mm, height 7.5 mm. Figure 12. Triphoris robustus Pease. Makaimo [Makemo Island, Tuamotu Archipelago]. Lectotype MCZ 73923. Width 2 mm, height 5.5 mm. Figure 13. Triphoris maculatus Pease. Kauai [Hawaiian Islands]. Lectotype MCZ 50069. Width 2 mm, height 4 mm. Figure 14. Triphoris similis Pease. [Haena] Kauai [Hawaiian Islands]. Lectotype MCZ 50079. Width 1.5 mm, height 4.5 mm. Figure 15. Triphoris perfectus Pease. Kauai [Hawaiian Islands]. Lectotype MCZ 302553. Width 1 mm, height 4 mm. Figure 16. Triton intermedius Pease. Oahu [Hawaiian Islands]. Lectotype MCZ 191331 . Width 22.5 mm, height 42 mm. Figure 17. Triton cylindricus Pease. Tahiti [Society Islands]. Lectotype MCZ 239749. Width 4.5 mm, height 1 1 mm. Figure 18. Truncatella concinna Pease. Apaiang [Abaiang Island, Kingsmill Islands, Kiribati]. Lectotype MCZ 178650. Width 3 mm, height 7 mm. Figure 19. Truncatella pacifica Pease. Oualan [Kusaie or Kosrae Island, Caroline Islands]. Lectotype MCZ 59799. Width 3 mm, height 8 mm. Figure 20. Turricula modesta Pease. [Ponape Island, Caroline Islands] Polynesia. Lectotype ANSP 28780. Width 1 2 mm, height 38 mm. Figure 21. Turricula (Costellaria) fortiplicata Pease. Paumotus [Tuamotu Archipelago]. Lectotype ANSP 28844. Width 3 mm, height 7 mm. Figure 22. Turbonilla elongata Pease. Paumotus [Tuamotu Archipelago]. Lectotype MCZ 10537. Width 5 mm, height 18 mm. Figure 23. Turcica coreensis Pease. Corea [Korea] Sea. Lectotype MCZ 104609. Width 21.5 mm, height 25 mm. Figure 24. Turricula (Pusia) nodulosa Pease. Paumotus [Tuamotu Archipelago]. Lectotype ANSP 28713. Width 4 mm, height 9 mm. Figure 25. Trivia corrugata Pease. Paumotus [Tuamotu Archipelago]. Lectotype MCZ 303451 . Width 4.5 mm, height 6 mm. Figure 26. Trochus excilis Pease. Pautomus [Tuamotu Archipelago]. Lectotype MCZ 104617. Width 4 mm, height 6 mm. Figure 27. Turricula (Costellaria) plicatula Pease. Paumotus [Tuamotu Archipelago]. Lectotype ANSP 28845. Width 3 mm, height 7 mm. MoLLusKs Described by W. H. Pease • Johnson 61 (US ISSN 0027-4100) Bulletin OF THE Museum of Comparative Zoology MCZ LIBRARY The Neotropical Orb-Weaver Genus Metazygia (Araneae: Araneidae) HERBERT W.LEVI HARVARD UNIVERSITY CAMBRIDGE, MASSACHUSETTS, U.S.A. VOLUME 154, NUMBER 2 8 MAY 1995 (US ISSN 0027-4100) PUBLICATIONS ISSUED OR DISTRIBUTED BY THE MUSEUM OF COMPARATIVE ZOOLOGY HARVARD UNIVERSITY Breviora 1952- bulletin 1863- Memoirs 1865-1938 JOHNSONIA, Department of Mollusks, 1941-1974 Occasional Papers on Mollusks, 1945- SPECIAL PUBLICATIONS. 1. Whittington, H. B., and W. D. I. Rolfe (eds.), 1963. Phylogeny and Evolution of Crustacea. 192 pp. 2. Turner, R. D., 1966. A Survey and Illustrated Catalogue of the Tere- dinidae (Mollusca: Bivalvia). 265 pp. 3. Sprinkle, J., 1973. Morphology and Evolution of Blastozoan Echino- derms. 284 pp. 4. Eaton, R. J., 1974. A Flora of Concord from Thoreau's Time to the Present Day. 236 pp. 5. Rhodin, A. G. J., and K. Miyata (eds.), 1983. Advances in Herpetology and Evolutionary Biology: Essays in Honor of Ernest E. Williams. 725 pp. 6. Angelo, R., 1990. Concord Area Trees and Shrubs. 118 pp. Other Publications. Bigelow, H. B., and W. C. Schroeder, 1953. Fishes of the Gulf of Maine. Reprinted 1964. Brues, C. T., A. L. Melander, and F. M. Carpenter, 1954. Classification of Insects. (Bulletin of the M.C.Z., Vol. 108.) Reprinted 1971. Creighton, W. S., 1950. The Ants of North America. Reprinted 1966. Lyman, C. P., and A. R. Davve (eds.), 1960. Proceedings of the First Inter- national Symposium on Natural Mammalian Hibernation. (Bulletin of the M.C.Z., Vol 124.) Ornithological Gazetteers of the Neotropics (1975-). Peters' Check-list of Birds of the World, vols. 1-16. Proceedings of the New England Zoological Club 1899-1947. (Complete sets only.) Proceedings of the Boston Society of Natural History. Price list and catalog of MCZ publications may be obtained from Publications Office, Museum of Comparative Zoology, Harvard University, Cambridge, Massachusetts 02138, U.S.A. This publication has been printed on acid-free permanent paper stock. © The President and Fellows of Harvard College 1995. THE NEOTROPICAL ORB-WEAVER GENUS METAZYGIA (ARANEAE: ARANEIDAE) HERBERT W. LEVM Abstract. Metazygia are Neotropical nocturnal orb weavers. Many species are small, less than 5 mm total length. They are found from the southeastern United States to Argentina, but most species occur in the Amazon area and southeastern South America. Al- though lacking a paramedian apophysis in the male palpus, they exhibit other characters that suggest that they be grouped (together with Eustala) close to Al- paida. There are 86 Neotropical species: 68 new (79%) and only 18 previously known (21%). Also, there are two Nearctic species, M. carolinalis and M. calix, making a total of 88 species of Metazygia. Of the 68 new species, 18 are known from both sexes, 16 from the male only, and 34 from the female only. Six names are synonymized for the first time. The females of all species are believed to rest in a retreat at the side of the web during daytime, and many build the orb with an open sector adjacent to the retreat. INTRODUCTION This is one of a series of revisions of Neotropical orb weavers (complete list in Levi, 1993b). These revisions should make it possible for researchers to identify Neo- tropical orb weavers, not possible earlier as some previously described species had never been illustrated and males had not been matched to females. Examining and illustrating the holotype specimens of old names is one of the most important tasks of the revisor. After all the genera of the family have been revised, the relationship of the genera to each other can be studied. METHODS AND MATERIALS The methods have been described in de- tail in Levi (1993b). As in previous papers, ' Museum of Comparative Zoology, Harvard Uni- versity, Cambridge, Massachusetts 02138. eye sizes are expressed as ratios, comparing the diameter of the measured eye (with cornea in profile) to that of the anterior median eyes (Levi, 1993b, figs. 27, 28). Distances between eyes of the anterior row are expressed as diameters of the anterior median eyes (in profile); distances between eyes of the posterior row are given as di- ameters of the posterior median eyes. The height of the clypeus (the distance be- tween the lower edge of the anterior me- dian eyes and the edge of the carapace) is given in diameters of an anterior median eye (Levi, 1993b, fig. 28f). These mea- surements are approximate, as araneid eyes are variable and difficult to measure ac- curately. Secondary Homonyms. The superb spi- der catalogs by Petrunkevitch, Roewer, and Bonnet, which so greatly facilitate the work of systematists, lumped genera. As a result of lumping genera, secondary homonyms are created: specific names that are unique in their own genera turn out to be hom- onyms when placed in the large genera Aranea or Araneus, having been used with Aranea or Araneus previously. Petrunkevitch (1911) and also Roewer (1942) made new names for the secondary homonyms. I have dismissed these new names when returning species to their pre- vious genera. In this I have followed other authors. For example, Petrunkevitch (1911) lumped 18 genera, replacing Singa moesta with metuens, Singa maculata with tusus, and Singa variabilis with varians, among others. These Petrunkevitch replacement names have not been used in Comstock (1912), Gertsch (Comstock, rev. edit. Bull. Mus. Comp. Zool., 154(2): 63-151, May, 1995 63 64 Bulletin Museum of Comparative Zoology, Vol. 154, No. 2 Gertsch 1940), or Kaston (1948). All have tried to avoid name changes. According to Article 59 of the Interna- tional Code of Zoological Nomenclature (ICZN, 1985), a junior secondary hom- onym replaced before 1961 is permanently invalid. Article 59b says that if the re- placement name for such a junior second- ary homonym is the cause of confusion, then the case is to be referred to the Com- mission. Article 59d says "a species group name rejected after 1960 on grounds of junior secondary homonymy is to be re- instated by anyone who believes the two species group taxa in question are not con- generic. . . ." The newer method (59d) has been used by American arachnologists all along and has avoided confusion. Roewer (1942, 1955) replaced some sec- ondary homonyms cited in this revision (e.g., Aranea palloides Roewer for Me- tazygia pallidula, Aranea errans for Me- tazygia erratica). But Roewer had the good judgment not to use Petrunkevitch's re- placement names (e.g., for Singa). Also, Bonnet (1955-59) did not use the replace- ment names of Petrunkevitch when re- turning species to their original genera (e.g., Petrunkevitch changed Zilla guttata to gemellus when he placed it in Araneus, but the name is back to Zilla guttata in Bonnet, 1959). Perhaps the International Commissioners should make the rules more flexible. It is unfortunate that many younger ar- achnologists still give their new species overworked names such as pallida, ma- culata, and variabilis, which are likely to have been used before in the same family or in related families and may be the cause for later discovery of secondary homony- my. Lectotypes. As in previous papers, lec- totypes have been designated when syn- types belonged to different species. They were not indicated routinely as an aspect of the revision; there is no requirement to do so (ICZN, 1985: Art. 74). A decision has to be made on whether to designate a male or a female as the lectotype. This choice may become critical later, if it is found that the presumed species actually consists of sibling species recognizable only in one sex, not in the other. Collections. Specimens from the follow- ing collections were used. I thank the cu- rators for making the material available for this study: AD A. Dean, Texas A and M Uni- versity, College Station, Texas, United States AMNH American Museum of Natural History, New York, United States; N. Platnick, L. Sorkin ANSP Academy of Natural Science, Philadelphia, Pennsylvania, United States BMNH Natural History Museum, London, England; P. Hillyard, F. Wanless CAS California Academy of Sci- ences, San Francisco, United States; W. J. Pulawski, D. Ubick CD C. Deeleman, Ossendrecht, Netherlands cue Cornell University Collection, kept in the AMNH; N. Plat- nick CV Carlos Valderrama A.; Bogota, Colombia DU D. Ubick, San Francisco, Cal- ifornia, United States FSCA Florida State Collection of Ar- thropods, Gainesville, Florida, United States; G. B. Edwards HECO Hope Entomology Collec- tions, Oxford University, Ox- ford, England; J. Lansbury IBNP Inventario Biologico Nacional, San Lorenzo, Paraguay; J. A. Kochalka lELP Instituto de Ecologia, Casilla 20127, La Paz, Bolivia; E. For- na, J. Coddington lESC Instituto Ecologia y Systema- tica, Cuban Academy of Sci- ence, Havana, Cuba, L. Armas METAZYGIA'Levi 65 INPA Institute Nacional de Pesquis- as da Amazonia, Manaus, Est. Amazonas, Brazil; C. Magal- haes IRSNB Institut Royal des Sciences Na- turelles de Belgique, Brussels, Belgium; L. Baert JEC J. Carico, Lynchburg, Virgin- ia, United States JMM J. Maes, Leon, Nicaragua MACN Museo Argentine de Ciencias Naturales, Buenos Aires, Ar- gentina; E. A. Maury MCN Museu de Ciencias Naturais, Funda9ao Zoobotanica do Rio Grande do Sul, Porto Alegre, Rio Grande do Sul, Brazil; E. H. Buckup MCP Pontificia Universidade Cato- lica do Rio Grande do Sul, Por- to Alegre, RS, Brazil; A. A. Lise MCZ Museum of Comparative Zo- ology, Cambridge, Massachu- setts, United States MECN Museo Ecuatoriano de Cien- cias Naturales, Quito, Ecua- dor; L. Aviles MEG M. E. Galiano; Buenos Aires, Argentina MIUP Museo de Invertebrados, Uni- versidad de Panama, Panama City, Panama; D. Quintero A. MLP Museo de Universidad Na- cional, La Plata, Argentina; R. F. Arrozpide MNHN Museum National d'Histoire Naturelle, Paris, France; J. Heurtault, C. Rollard MNHNC Museum Nacional de Historia Natural, Havana, Cuba; G. Alayon MNRJ Museu Nacional, Rio de Ja- neiro, Brazil; A. Timotheo da Costa MNSD Museo Nacional de Historia Natural, Santo Domingo, Re- publica Dominicana; Felix Del Monte MUSM Museo de Historia Natural, Universidad Nacional Mayor de San Marcos, Lima, Peru; D. Silva D. MZSP Museu de Zoologia, Universi- dade de Sao Paulo, Sao Paulo, SP, Brazil; P. Vanzolini, J. L. Neme MZUF Museo Zoologico de "La Spe- cola" Universita di Firenze, Florence, Italy; S. Mascherini NMW Naturhistorisches Museum, Vienna, Austria; J. Gruber NRMS Naturhistoriska Riksmuseet, Stockholm, Sweden; T. Kro- nestedt PAN Polska Akademia Nauk, War- szawa, Poland; A. Riedel, J. Proszynski, A. Slojewska, E. Kierych PMY Peabody Museum, Yale Uni- versity; C. Remington, D. G. Furth REL R. E. Leech; Edmonton, Al- berta, Canada SMF Forschungsinstitut Sencken- berg, Frankfurt am Main, Germany; M. Grasshoff SR Susan Riechert; Knoxville, Tennessee, United States USNM National Museum of Natural History, Smithsonian Institu- tion, Washington, D.C., Unit- ed States; J. Coddington ZMB Zoologisches Museum der Humbolt Universitat, Berlin, Germany; M. Moritz ZMK Zoologisk Museum, Copenha- gen, Denmark; H. Enghoff ZSM Zoologisches Staatssammlung, Munich, Germany I would also like to thank R. Buskirk and W. Eberhard for natural history in- formation and G. B. Edwards and P. Van- zolini for information on collecting sites. C. D. Dondale gave advice on a nomen- clatural problem. Laura Leibensperger helped thoughout and read the manu- script. L. R. Levi improved the wording. E. H. Buckup and two anonymous readers carefully read the manuscript and made 66 Bulletin Museum of Comparative Zoology, Vol. 154, No. 2 numerous corrections and improvements. D. Sherry helped with the final word pro- cessing. The research was started with the help of NSF grant BMS 75-05719. Publi- cation costs of this study were covered in part by the Wetmore Colles Fund. Metazygia F. P.-Cambridge Metazygia F. P.-Cambridge, 1903: 501. Type species by original designation M. wittfeldae (McCook). The gender of the name is feminine. Diagnosis. Metazygia differs from oth- er araneid genera by having a glabrous carapace with posterior median eyes that face up and almost touching each other, and an oval abdomen, that is widest at the middle (Figs. 5, 12). In many species, but not all, the female has a round, laterally compressed scape on the ventral face of the epigynum (Figs. 1- 4). The palpus of the male has only one patellar seta and lacks a paramedian apophysis (Figs. 45, 112). Contiguous posterior median eyes are found also in Larinia and Cyclosa and among some species of Alpaida, Araneus, Aculepeira, and Dubiepeira. Larinia dif- fers from Metazygia by having an elon- gate abdomen. These genera (except Al- paida) differ from Metazygia by having the carapace setose and by the annulate scape. All except Alpaida and Cyclosa have two macrosetae on the palpal patella, and the male palpus has a distal hematodocha. Cyclosa differs by having the carapace narrow in the cephalic region. Both Al- paida and Cyclosa have a paramedian apophysis in the male palpus. Metazygia females can easily be confused with those of Singa, Nuctenea (Larinioides) , and Zygiella because of similar abdomen col- oration and shape. These three genera are not found in the Neotropical region, al- though Zygiella x-notata (Clerck) and Larinioides sclopetaria (Clerck) are com- mon in Chile where they have been intro- duced. No Metazygia species are known from Chile. Zygiella males differ from Me- tazygia by having a cone-shaped palpal tibia, as is common in tetragnathids; Nuc- tenea and Larinioides males have two pal- pal patellar macrosetae. Some Metazygia females have been confused with Chry- someta, but this has tetragnathid charac- ters (Levi, 1986). The species of these sim- ilar genera all make a tube-shaped, silken retreat. It is possible to have doubts and misplace Metazygia species if only a female is avail- able. Relationship. Metazygia is closest to the genus Eustala. Shared apomorphies in- clude the anteriorly projecting scape in some Metazygia females (Figs. 263, 270), the lateral placement of the median apophysis in the male palpus (M in Figs. 44, 45), and, in the palpus of larger species, a semitransparent blister below the distal prong of the terminal apophysis (B in Figs. 45, 46). In spite of these unusual shared characters, Eustala is distinguished by the position of the posterior median eyes, on a slight swelling and facing laterally, and by the shape of the abdomen, subtrian- gular to elongate, widest anteriorly and often with a median white streak on the underside. Also, the carapace of Eustala is setose, while that of Metazygia is glabrous (Table 1). Both Metazygia and Eustala, although lacking a paramedian apophysis in the male palpus (Figs. 45, 112), have to be grouped near Alpaida and other genera having a paramedian apophysis. They mostly have only a single palpal patellar seta and the position of the conductor of the male pal- pus is usually on the inside face of the tegulum (and not on the outside rim of the tegulum as in Araneus-related genera). There is a relative absence of the distal hematodocha in the male palpus, a struc- ture also prominent in Araneus-relaied genera. There is abundant pigment around the eyes in Metazygia species, as there is in Alpaida species. Many of the genera related to Alpaida (but not Alpaida or Metazygia) have abdomens with two or three posterior, median humps on the ab- METAZYGIA'Levi 67 Table 1. Differential characters of Parawixia (PARW), Eriophora (ERIO), Acanthepeira (ACAN), Wagneriana (WAGN), Eustala (EUST), Acacesia (ACAC), Alpaida (ALPA), Ocrepeira (OCRE), Cyclosa (CYCL), and Metazygia (METZ). CYCL EUST PARW ERlO acan WAGN ACAC ALPA OCRE (prov.) METZ (prov) Pattern Carap. glabrous - - - - — [ + *] — _ [ + ] _ Paired spots on carap. + * - + - - — - _ Marks between ME and LE + * _* + - - - _* - — — Black eye rings - - - - - [ + *] - - + * — Sides of thoracic reg. black - - - [ + *] - _* — — — Pattern on sternum [ + *] - - - — — — _ _ Abd. V. black with white streak - - - - - - - - [ + ] Female morphology Eye reg. narrow - - - - [ + ] - — + - — LE on sides of tuber. + - + - - - - - — PME on slight swelling _* - - - + - + * - — + PME touch - - - - - - — + [ + ] — Carap. swollen behind eyes + + + + - + + * - - Abd. with tubers 4-15 0-3 12 + 9-15 - _* [2*] 0-2 - _* Ant. median abd. tuber _* - + - - - — — - Abd. subspherical + * + + - - - — — + + * Abd. oval _* - - + + + _* + [ + ] + * Abd. with tail - - - + * - - — + * + * Three median post, tubers + _* + + — - — — — — Abd. glabrous - - - - - [ + ] - - [ + ] - Epigynum Scape + + + - + — +* + + * + Lat. flattened scape - - - - - — - - [ + *] - Lobe - - - + - + _* — + * — Knob at tip _* - — + - _* - — — — Notch on face - - - - - [ + *] - - — — Scape ant. projection - - - - - - - _* [ + ] Male morphology Ceph. reg. wide - - [ + ] - - - - - - - Hook on coxa I + + - + + + * + +* + + Macrosetae on coxa III, IV + * + * + * — + * + * +* + * + * — Trochanter IV macrosetae + * + * — — - _* - - - - Tibia II modified - - — — + _* - - - - Carapace with lobes - - - - - - - - [ + *] - Fangs clasping - - - - - - - - [ + *] - Palpus Much wider than long - - - - - - - [ + ] - - Patella macrosetae 1** [2**] 1 1 1 1* 1 1 1 1 Y narrow - [ + ] - - - - - - - - PM present + + + + + + + + - + M lateral - - - - - - + - + - SA blister-shaped — - - - - - - - + - M soft - - - - - - - - _* - * There are exceptions. ** P. bistriata and E. nephiloides (Levi, 1971) have a large macroseta and a smaller one on the male* palpal patella. Bracketed characters are autapomorphies for the genus. Abbreviations: abd., abdomen; ant., anterior; carap., carapace; ceph., cephalic; post., posterior(ly); prov., provisionally; reg., region; tuber(s)., tubercle(s); v., venter. LE, lateral eyes; M, median apophysis; ME, median eyes; PM, paramedian apophysis; PME, posterior median eyes; Y, cymbium. Data from Levi, 1971, 1976, 1977, 1988, 1991b, 1992, 1993b. 68 Bulletin Museum of Comparative Zoology, Vol. 154, No. 2 domen (Eriophora, Parawixia, Wagner- iana, Acanthepeira, and also Eustala), a feature uncommon in any other spider. Perhaps the lack of paramedian apophysis and the sometimes intermediate position of the conductor (Fig. 112) indicate an intermediate relationship (Table 1). Description. The cephalothorax is or- ange to orange-brown in alcohol, the legs rarely with dark rings. The abdomen has a characteristic folium pattern consisting of pairs of brackets (Figs. 5, 124) or some- times a Zygiella-\ike pattern (Figs. 58, 110). The smallest species have a white (green or silvery when alive) abdomen with a black band around the anterior margin (Figs. 325, 330). The green color, which washes out in alcohol, is known for M. octama, M. serian, and M. lopez (Eber- hard, personal communication). The black band around the anterior is not found in other Neotropical araneids. The carapace has few setae, and the median eye quad- rangle is always narrower posteriorly (ex- cept in M. vaupes, where it is square). The anterior median eyes are slightly larger than the posterior medians, and the laterals are always the smallest. The height of the clypeus is less than the diameter of an an- terior median eye. The eye region is usu- ally all black but in some species lacks pig- ment (Fig. 336). The cephalic region of the carapace is more than half the width of the thoracic region; in M. uma (Fig. 224), it is almost equal in width. The abdomen is oval, widest in the mid- dle. In some small species, it is almost spherical (Figs. 306, 307); in some large species, almost cylindrical (Fig. 228). Sometimes the abdomen is anteriorly flat- tened (Fig. 317) or anteriorly projecting (Fig. 242) or has an anterior notch (M. vaupes. Fig. 301). The males of three species have modi- fied fangs (see later). Genitalia. In the larger species, the epi- gynum has a laterally flattened scape, which is round in lateral view (Figs. 1-4, 69-71). The greatest diversity of epigyna occurs in the small species, where some even have a long scape (Figs. 322, 328). In many species, the scape appears to be torn off by the male after mating, and it may be unusual to find a female with the scape intact (Figs. 277-280). (Removal of the scape by the male may protect its sperm by preventing additional female matings.) Often it is not known whether there is a scar from a torn scape or the middle area is sculptured (Figs. 285-287, 365, 366). Part of the base of the epigynum is torn off in M. mundulella (Fig. 234) and may be missing also in M. saturnino (Figs. 197, 198) and M. amalla (Figs. 247, 248). While it is common for an araneid male to re- move the scape, destruction of the base of the epigynum is not otherwise known in araneid spiders. In M. erratica, the open- ing of the epigynum is sealed with an amorphous black secretion that is difficult to remove (Figs. 370-372). A similar brown exudate may be present on the epigynum of M. lopez. In still other species, part of the male embolus breaks and plugs the opening of the epigynum (Fig. 40). Internal female genitalia were exam- ined in two pairs of species: M.wittfeldae and M. dubia, and M. zilloides and M. keyserlingi. No differences were found in the similar species that might be useful for determinations. Male. Males of some small species have the carapace margin lobed above the first coxae (Figs. 384, 390), a modification not seen in males of other genera. All males have one macroseta on the palpal patella. Males of all except three species (M. gre- galis, M. benella, and M . yobena) have a tooth on the endite that faces a tooth or tubercle at the proximal end of the palpal femur. These three species also have mod- ified fangs: the fangs have a lobe (Figs. 261, 262, 269, 276), presumably to hold the female during mating. Metazygia gre- galis also has the distal end of the chelic- erae modified as a protecting flange (Fig. 261). All species have a small hook on the distal margin of the first coxa, but if very small, the hook may have moved, to face Metazygia 'Levi 69 the second coxa. The second tibiae are usu- ally thicker than the first and are armed with macrosetae. Males may have macro- setae on the first tibia, also. Some males have a short macroseta on the fourth coxa and sometimes on the third, as in related genera (those with one patellar macroseta). Some species have a large terminal apophysis (A) in the palpus, with a ter- minal prong above a transparent blister (B in Figs. 45, 112); others have lost the ter- minal apophysis and have only one sclerite beyond the embolus (E) or none (Fig. 260). The part remaining with the embolus (Figs. 221, 230, 237) might be a reduced terminal apophysis or the embolus lamella (L of Figs. 45, 54). There is no way at present to determine the homology. Here it is called the "lamella" (in keys and descriptions). None of the Metazygia species has a para- median apophysis. As in Eustala, the me- dian apophysis has moved ventrally to the side of the palpus (M in Figs. 44, 45, 112, 113) and has lost its sclerotization, often becoming soft and white. The conductor (C) has moved in the same direction, and there is no projection from the conductor toward the cymbium, as in Alpaida, to form a paramedian apophysis. The position of the conductor in ara- neids may be on the rim of the tegulum, visible in ventral view as in Neoscona (Levi, 1993a, C in figs. 6, 7, 16), in Dubiepeira (Levi, 1991a, center of fig. 525, C in fig. 526), and in many species of Araneus (Levi, 1991a, figs. 3, 14, the light, round sclerite in center of fig. 22), or it may be on the inside face of the tegulum, closer to the cymbium, as in most species with a para- median apophysis. In the latter case, it is also surrounded on the outside by the wall of the tegulum, as in Micrathena (Levi, 1985, C in figs. 6, 9), Alpaida (Levi, 1988, C in fig. 10), and Wagneriana (Levi, 1991b, fig. 19). In Metazygia, which lack a para- median apophysis, the conductor may be on the rim, as in M. crewi (C in Fig. 113) and M. isabelae (at 8 hr in Fig. 92), or inside, below the rim of the tegulum, as in M. castaneoscutata (center of Fig. 308), or below the rim but hanging over it, as in M. zilloides (C in Figs. 44, 45) and M. gregalis (at 10 hr in Fig. 259, C in Fig. 260). In addition to the blister-shaped subter- minal apophysis and the lateral position of the median apophysis, there are additional peculiarities in the palpus of the larger species. The radix is pushed "down," out of the way, and is in a much "lower" po- sition (R in Figs. 44, 45, 112) than in spe- cies of other genera. Also, there is a stipes (I) in the form of a band that overlaps the dorsal surface of the palpal bulb (bottom third of Fig. 46), to which the embolus and its lamella are attached (bottom left of Fig. 46, and also Fig. 112). Finally, the median apophysis may be in a more common po- sition (at 5 hr in Fig. 243; M in Fig. 245), the radix farther "up" in the palpus (below the embolus in Fig. 308). There are many small species in Me- tazygia. As is common in spiders, the smallest species display the greatest diver- sity in genitalia. Great diversity in geni- talia is also known for Micrathena species, mostly medium-sized. Matching males with females of the same species is difficult because so many species are similar in ap- pearance, differing only in genitalic struc- tures. It has not been possible to clearly group Metazygia species into subgenera because the diversity of characters does not fall into corrolative patterns. The larger species have a pattern of brackets on the abdomen (Fig. 5), have a terminal apophysis in the palpus (A in Fig. 45), and have the median apophysis (M in Figs. 44, 45) soft and to- ward the side. Small species have a black band around the anterior of the abdomen (Figs. 381, 404), a diversity of female epi- gyna, have the male palpus without ter- minal apophysis, and the median apoph- ysis in the more common araneid position at 4-5 hr in the left palpus (Figs. 383, 389). However, the Metazygia curari female (Figs. 144-146) has the characteristic flat- tened, round scape, as does M. wittfeldae (Figs. 1-4), and the male lacks a terminal 70 Bulletin Museum of Comparative Zoology, Vol. 154, No. 2 apophysis (Fig. 148). Metazygia mundu- lella also has this kind of scape (Figs. 231- 234), but the male has a median apophysis with sclerotized points at 5 hr in the left palpus (Fig. 237). Metazygia serian (Fig. 175) and M. adisi (Fig. 141), with a flat- tened round "Metazygia" scape, have the black band of small species around the an- terior of the abdomen. Natural History. All species build a ver- tical orb and have a retreat that is usually above the web and attached to a branch, wall, ceiling, or curled leaf. The spider rests in the retreat during the day and in the center of the web at night. Some webs have a vacant sector in the part of the orb adjacent to the retreat (Plate 1): the vacant sector orb is known for M. wittfeldae, zil- loides, keyserlingi, chicanna, and incerta. The webs of M. keyserlingi and M. lati- ceps are kept up during the day and have a signal line (Eberhard, personal com- munication). "All other species built at night, many (all?) quite early in the eve- ning and not generally have either an open sector or a recognizable retreat" (Eber- hard, personal communication). Metazy- gia incerta rebuilds orbs every two to four days (Buskirk, personal communication). Some species take down their webs during the day (Lubin, 1978). While Metazygia wittfeldae is usually solitary, the web size and structure are the same when they aggregate and the angle of the orb continues to vary from just hor- izonal to vertical at Monteverde, Costa Rica (Buskirk 1986). Eberhard (personal correspondence) writes, I have watched both gregalis and octama build in great hurry (rapidly, with little exploratory behavior) just as the light is failing, and have web photos of chenevo ... at 6 pm; serian at 5 pm, lopez at 7 pm; wittfeldae here [Costa Rica] also builds ear- ly in the evening. Thus I suspect these spe- cies are working on the flush of insects which fly just at dusk. One keyserlingi also had a web up at night and since I saw another web of this species which was rebuilt around noon after rain, I suspect it is like gregalis in having not one but a series of webs dur- ing each 24 hour period. . . . Species occur often in great abundance, females and males together. But because they are difficult to collect by sweeping or beating, many species are present in col- lections only as single individuals. The following observations are excerpt- ed from Eberhard (personal communica- tion): Relatively open habitat (rel. early sec- ondary growth, grass): lopez, gregalis, oc- tama, pallidula, wittfeldae, yobena, be- nella, lazepa, serian, chenevo. I suspect some species at least of preferring to be near water (esp. pallidula), and of liking twigs, barbed wire or other relatively rigid sup- ports for their webs, but have seen yobena and chenevo on webs in tall grass. On build- ings (especially near lights): wittfeldae, gregalis, dubia. Silk retreats, more or less cylindrical (open at both ends — spider will leave on rear if bothered from front side) during the day: octama, gregalis, (in this case, often in cracks or other protected sides); the retreat of this species generally has no connection whatsoever with the web, which is often left intact during the day when the spider is in its retreat, and it is thus generally impossible to associate a giv- en spider with the vestiges of a given web during the day. In contrast, octama re- moves the web completely during the day, I think usually without a single line being left up, and its retreat is at least sometimes on a green leaf in relatively exposed posi- Plate 1 . A, Metazygia chicanna n. sp. B, M. dubia (Keyserling). C, photograph of web of M. crew/ (Banks); horizontal diameter 6 cm. D, photograph of web of M. keyserlingi Banks, horizontal diameter in middle about 12 cm. M £ TAZYCIA • Levi 7 1 72 Bulletin Museum of Comparative Zoology, Vol. 154, No. 2 Map 1 . The number of species of Metazygia in various areas. tion. I don't know much about the retreats of the other nocturnal species (and since retreats may be disassociated from webs, I was unhkely to be aware of the spider's retreat when I found the spider on a web at night). Dynamics of webs: I can only give you details for gregalis and octama. The oc- tama web seems relatively fragile, and the spiders had often torn down the web and were feeding on a ball of prey not more than a couple of hours after dusk. I never saw one of these (they lived in our yard in Cali) put up another web, but might have missed it (espcially if a second web was put up just before dawn and then soon torn down). I had some students do all night projects with gregalis, and they found that the same individual built two to three webs per night. Generally the first web was just at dusk, and the others substantially later at night. M. gregalis is a generalist when it comes to feeding. I have the impression that in- traspecific variation in the form in the Me- tazygia orb is relatively high in M. gregalis. Sizes of prey items are reported in Cas- tillo and Eberhard (1983). Distribution. Metazygia is known only from the Americas. There are several pairs of very similar allopatric species, one north, the other south, without overlap: wittfel- dae and dubia, and zilloides and keyser- lingi. Metazygia species of the southeastern United States, not otherwise cited in this paper, are as follows: M. carolinalis (Arch- er) (the male is unknown); and M. calix (Walckenaer), NEW COMBINATION. Metazygia calix (Levi, 1976, figs. 137-144) was placed in Alpaida but has genitalia similar to those of M. laticeps (Figs. 226, 227, 230) and M. sendero (Figs. 216, 217, 221). Misplaced Species. Metazygia livida Mello-Leitao, 1941: 151, 12. Female from Argentina is a Dictyna (Dictynidae). Metazygia unquiformis: — Valle and Valle, 1972: 33 is Alpaida veniliae (Key- serling) (Levi, 1988: 402). Keys. Keys are difficult to construct for species of which only one or a few indi- viduals are known. With few specimens, one does not know whether or not the col- oration is characteristic, whether or not the epigynum has been torn apart by a male when mating, whether the male has a vir- gin embolus with a cap or if he has mated, and whether all males of the species have a macroseta on the fourth coxa or only the one sampled. For using the key, the female's epigyn- um has to be slightly pulled out with a curved needle to see the posterior face of the structure. Key to Female Metazygia 1. Epigvnum with an anterior projection (Figs. 254, 263, 270) _ 2 Epigynum otherwise 4 2(1). Epigynum in ventral view wider than long (Fig. 254); Central America, West Indies to northern Argentina (Map 3E) gregalis Epigynum in ventral view longer than wide (Figs. 263, 270) 3 3(2). Epigynum with a deep notch on the pos- terior border (Fig. 270); anterior pro- jection without pair of wings (Fig. 270); Amazon area; Sao Paulo (Map 3C) yobena METAZYGIA'Levi 73 Map 2. Distribution of Metazygia species. - Epigynum without notch on posterior 5(4). border (Fig. 263); anterior projection with a pair of wings (Fig. 263); Pan- ama, Colombia (Map 3C) benella 4(1). Epigynum in posterior view with a me- - dian plate forming a septum in an hour glass-shaped depression as in Figure 108; Greater Antilles (Map 21) crewi 6(5). Epigynum with median posterior plate otherwise 5 Epigynum in ventral view with a scape that extends beyond the posterior mar- gin of the base (Figs. 193, 277, 298, 303, 314, 322, 328) 6 Epigynum with scape not extending be- yond posterior margin (Figs. 2, 55, 231) 12 In ventral view scape extending from epigynum's posterior margin (Figs. 298, 314) 7 74 Bulletin Museum of Comparative Zoology, Vol. 154, No. 2 - Scape extending from middle or anterior of basal plate (Figs. 193, 277, 303, 322, 328) 8 7(6). Epigynum with notch on each side (Fig. 298), anterior of abdomen indented (Fig. 301); western Amazon area (Map 4 A) vaupes - Epigynum without notch (Fig. 314), ab- domen oval to subspherical (Fig. 317); Rio de Janeiro State, Brazil (Map 4B) floresta 8(6). Scape attached on anterior of base (Figs. 193, 277, 328) 9 - Epigynum with scape attached in mid- dle of plate (Figs. 303, 322) 11 9(8). Scape thick and with a deep groove (Fig. 193); Bahia State, Brazil (Map 3A) uratron - Scape otherwise 10 10(9). Scape thin and transparent (Fig. 328); Amazon area to Misiones Prov., Ar- gentina (Map 4C) lagiana - Scape thick, opaque (Fig. 277); Amazon area to northern Argentina (Map 3D) voluptifica 11(8). Epigynum with notch on each side in posterior view (Fig. 323); Mexico, Central America (Map 4G) .... nigrocincta - Epigynum with posterior view otherwise (Fig. 305); Amazon area (Map 4B) castaneoscutata 12(5). Abdomen with paired, dark patches (Fig. 90); Goias State, Brazil (Map 2C) redfordi Abdomen marked otherwise (Figs. 58, 175) 13(12). Epigynum with scape round, laterally flattened (Figs. 1-4, 62-64, 158-160, 1 64- 1 66 ) - Epigynum otherwise (Figs. 114, 133, 189, 370) 14(13). Epigynum with bordered depression or opening, visible in ventral view (Figs. 40, 55, 74, 87, 158, 168, 172) Epigynum otherwise (Figs. 2, 35, 129, 1 44, 23 1 ) 22 15(14). Openings oval, very large and on each side of epigynum (Fig. 74); Mexico, to Guianas and Peru (Map 2E) palliduta Openings otherwise (Figs. 40, 48, 55, 121, 1 58, 1 59) 1 6 16(15). Openings along anterior of plate (Figs. 40, 49, 158, 168) 17 - Openings in middle or posterior 20 17(16). In posterior view epigynum longer than wide; with narrow median plate (Fig. 169); northern Amazon region, Guian- as (Map 2K) uraricoera In posterior view epigynum wider than long, median plate wide (Figs. 41, 50, 159) _.._ 18 18(17). Openings anterior, lateral (Figs. 39, 48) 19 13 14 42 19(18). 20(16). Openings near median, indistinct (Fig. 158); western Amazon area (Map 2K) yucumo United States to Honduras, West Indies (Map 2G) zilloides Costa Rica, Trinidad to Colombia (Map 2G) keyserlingi Abdomen with black band around an- terior (Fig. 175); Costa Rica (Map 2) serian Abdomen otherwise 21 21(20). Openings round in center of each side (Fig. 55), scape small, light, indistinct (Fig. 55); southern Mexico to Hon- duras, Jamaica (Map 2B) chicanna - Openings tiny notches toward posterior of base, scape large, distinct (Fig. 121); Mexico (Map 21) taman 22(14). In posterior view of epigynum, width of median plate equal to or less than that of laterals (Figs. 63, 95) or median plate T-shaped with vertical piece narrow (Figs. 139, 165) 23 - In posterior view median plate wider than laterals (Figs. 30, 36, 70, 126) or oth- erwise 23(22). A scale on each side of epigynum as in Figures 94 and 95; southeastern Brazil (Map 2C) rogenhoferi Epigynum without scale (Figs. 62, 139, 1 65 ) 24 24(23). Median plate T-shaped (Figs. 139, 165) 25 26 Median plate otherwise (Fig. 63); west- ern Amazon area (Map 2B) tapa 25(24). Arms of T-shaped median plate con- stricted at base and pointed (Fig. 139); Amazon area (Map 2H) adisi Arms not constricted (Fig. 165); southern Brazil, northern Argentina (Map 2K) ipanga 15 26(22). Base of epigynum on each side with shal- low lateral notch as in Figures 231 and 234; southeastern Brazil (Map 3F) mundulella Base without such notches 27 27(26). Margin of base in ventral view entire, without notches on sides (Fig. 69); Central America (Map 2J) incerta - Margin of base otherwise (Figs. 22, 100, 121, 181) 28 28(27). Folds posterior to scape in ventral view (Fig. 100); Sao Paulo State, Brazil (Map 2F) harueri - Epigynum otherwise 29 29(28). A transverse bar posterior to scape in ventral view as in Figure 181; Colom- bian Amazon area (Map 3A) lazepa - Epigynum otherwise 30 Metazygia 'Levi 75 Map 3. Distribution of Metazygia species. 76 Bulletin Museum of Comparative Zoology, Vol. 154, No. 2 30(29). A dark area on each side of scape in ventral view as in Figure 125; Amazon area (Map 2H) paquisha - Epigynum otherwise 31 31(30). In lateral view scape about twice as long as wide (Figs. 83, 146) 32 - In lateral view scape about as long as wide (Figs, 18, 37, 123, 131) 33 32(31). Abdomen with a pair of dorsal, longi- tudinal white lines (Fig. 147); Amazon area (Map 2H) curari - Abdomen with a pair of dorsal, longi- tudinal dusky bands (Fig. 84); Neblina area of Amazon (Map 2E) enabla 33(31). Abdomen oval, about three quarters as wide as long 34 - Abdomen elongate, about two thirds as wide as long (Fig. 153), epigynum as in Figures 150-152; Bolivia (Map 2F) bolivia 34(33). Posterior median plate with a pair of notches on each side as in Figures 30 and 1 78 35 - Posterior median plate otherwise 36 35(34). In ventral view length of scape about two thirds length of base (Fig. 177); Guy- ana (Map 3A) chenevo - In ventral view length of scape less than half length of base (Fig. 29); Peruvian Amazon to Bahia State, Brazil (Map 2D) patiama 36(34). Median plate in posterior view heart- shaped (Fig. 130); Ecuador (Map 2H) nobas Median plate otherwise (Figs. 23, 36, 122) 37 37(36). Posterior median plate hexagonal (Fig. 122); Mexico (Map 21) taman Posterior median plate otherwise (Figs. 23, 36) 38 38(37). Epigynum in ventral view with a slit depression on each side as in Figure 35; Pernambuco, Brazil (Map 2D) .. ipago - Epigynum otherwise (Figs. 1, 2, 9, 16, 22) 39 39(38). Abdomen with paired spots (Fig. 25); posterior median plate about as long as wide (Fig. 23); Venezuela, Peru (Map 2D) pimentel Abdomen with foHum (Figs. 5, 12, 19); posterior median plate slightly wider than long (Figs. 3, 10, 17) 40 40(39). Bahama Islands (Map 2A) bahama - United States to northern South Ameri- ca, West Indies 41 41(40). Base of epigynum almost twice as wide as long in ventral view (Fig. 2); United States to Costa Rica (Map 2A) wittfeldae Base of epigynum narrower, about three eighths as long as wide (Fig. 16); Costa Rica, West Indies, Galapagos to north- ern Brazil and Peru (Map 2A) dubia 42(13). Epigynum with set off scape (Figs. 133, 185, 208, 309, 378) 43 - Epigynum without scape or with only scars of torn scape (Figs. 280, 285, 292, 360, 370, 391) 49 43(42). Scape with transverse wrinkles (Fig. 208); Cuba (Map 2J) matanzas Scape smooth (Figs. 133, 185, 189) 44 44(43). Scape with dent on each side as in Figure 185; Peruvian Amazon (Map 3 A) atalaya - Scape otherwise 45 45(44). Scape ventrally flattened (Figs. 189, 378) 46 Scape knob-like (Figs. 114, 133, 309) 47 46(45). Scape subtriangular (Fig. 378); Peruvian Amazon region (Map 4E) genaro Scape oval (Fig. 189); Colombian Am- azon region (Map 3B) corima 47(45). Posterior median plate with concave sides (Fig. 310); Panama, Colombia (Map 4B) octama Posterior median plate convex on each side (Figs. 115, 134) 48 48(47). Posterior median plate wider than long (Fig. 115); Guatemala (Map 21) vaurieorum - Posterior median plate almost as wide as long (Fig. 134); Lower Amazon area (Map 2K) goeldii 49(42). Epigynum in posterior view longer than wide (Figs. 240, 338) 50 - Epigynum in posterior view wider than long (Figs. 361, 386) 51 50(49). Posterior median plate narrower dorsal- ly than ventrally (Fig. 338); Bolivian Amazon area (Map 4C) loque - Posterior median plate almost rectan- gular (Fig. 240); coast of southeastern Brazil (Map 3F) genialis 51(49). Epigynum with scars, scape usually torn off' (Figs. 197, 198, 248, 280, 292) 52 Epigynum without distinct scars (Figs. 216, 226, 365, 391) 58 52(51). Scars in midline only (Figs. 203, 280, 285, 289) 54 - Whole venter of epigynum seemingly torn off (Figs. 197, 247) 53 53(52). Posterior median plate wider than long (Fig. 248); southern Brazil (Map 3D) amalla Posterior median plate square (Fig. 198); southern Brazil (Map 3A) saturnino 54(52). Posterior median plate in a depression (Fig. 286); southern Brazil (Map 3A) viriosa - Posterior median plate otherwise (Figs. 204, 281, 290, 293) 55 55(54). Posterior median plate much wider than METAZYGiA'Levi 77 Map 4. Distribution of Metazygia species. 78 Bulletin Museum of Comparative Zoology, Vol. 154, No. 2 long (Fig. 290); Mato Grosso, Brazil (Map 3D) ituari Posterior median plate otherwise (Figs. 204, 281, 293) 56 56(55). Posterior median plate Y-shaped as in Figure 293; Peruvian Amazon (Map 4A) limonal Posterior median plate otherwise 57 57(56). Epigynum in ventral view with a round, shallow depression on each side as in Figure 280; Amazon area to northern Argentina (Map 3D) voluptifica - Epigynum in ventral view with a raised area on each side as in Figure 203; southeastern Brazil (Map 3B) crabroniphila 58(51 ). Cephalic area of carapace almost as wide as thoracic area; abdomen with lon- gitudinal stripes as in Figure 224; Am- azon area (Map 3B) uma Carapace otherwise 59 59(58). Epigynum in ventral view pentagonal as in Figure 333; Colombian Amazon (Map 4C) carimagua - Epigynum otherwise 60 60(59). Epigynum subtriangular in ventral view with shallow median groove and with- out distinct lip as in Figures 216 and 226; cephalic region of carapace rel- atively wide (Figs. 219, 228) 61 Epigynum otherwise (Figs. 346, 360, 391) 62 61(60). Posterior median plate longer than wide (Fig. 226); Panama to Rio de Janeiro State, Brazil (Map 3F) laticeps Posterior median plate wider than long (Fig. 217); Ecuador, Peru (Map 3B) . sendero 62(60). Epigynum in ventral view with posterior margin lobed and median area swollen as in Figures 342 and 351 63 Epigynum otherwise (Figs. 346, 360, 365, 370) 64 63(62). Epigynum with depression on ventral face (Figs. 351, 354), posterior median plate dumb-bell-shaped (Fig. 352); Venezuela to Peru (Map 4D) lopez - Epigynum swollen on ventral face (Fig. 342), posterior median plate triangu- lar (Fig. 343); Hispaniola (Map 4D) ... _ _ cienaga 64(62). In ventral view posterior margin of epi- gynum with a swollen lip as in Figures 360, 365, 370 and 405 65 - Posterior margin of epigynum without swollen lip (Figs. 346, 385, 409) 68 65(64). Lip a horizonatal bar as in Figure 405; posterior median plate T-shaped (Fig. 406); Rio Grande do Sul, Brazil (Map 4F) ._ „ _ valentim Lip V-, U-, or T-shaped (Figs. 360, 365, 370) „ _ 66 66(65). Posterior median plate pentagonal (Fig. 371); black amorphous material on each side posteriorly (Figs. 370, 372); Amazon area (Map 4E) erratica - Posterior median plate not pentagonal (Figs. 361, 366); without black amor- phous material 67 67(66). Posterior median plate triangular (Fig. 361); Peruvian Amazon (Map 4D) samiria - Posterior median plate square, anterior to it a textured area (Fig. 366); Am- azon region, Bolivia (Map 4D) ducke 68(64). Posterior margin with a notch in middle and a lobe extending each side as in Figure 346; Amazon area (Map 4D) .. souza - Posterior margin otherwise, often with a pair of lobes (Figs. 385, 391, 401, 409) 69 69(68). In ventral view a pair of lobes as in Fig- ure 385; posterior median plate dumb- bell-shaped (Fig. 386); Mato Grosso, Brazil (Map 4E) voxanta Ventral view otherwise (Figs. 391, 401, 409) 70 70(69). A median ventral notch in posterior view of epigynum (at 12 hr in Figs. 402, 410) 7 1 No notch visible in posterior view (Figs. 392, 395); Colombia, Ecuador to mouth of Amazon (Map 4F) peckorum 71(70). Posterior median plate constricted ven- trally (at 12 hr in Fig. 402); Ecuador- an, Peruvian Amazon (Map 4F) . moldira Posterior median plate as in Figure 410; Bahia to Sao Paulo States, Brazil (Map 4F) bahia Key to Male Metazygia 1. Cheliceral bases or fangs modified with transparent lobes (Figs. 261, 262, 269, 276) 2 Cheliceral bases or fangs not modified .. 4 2(1). Median apophysis (M in Fig. 260), in mesal view short (Figs. 258, 260); Cen- tral America, West Indies, South America (Map 3E) _ gregalis Median apophysis, in mesal view, longer (Figs. 267, 274) 3 3(2). Median apophysis with a black wall (at 4 hr in Fig. 274, at 6 hr in Fig. 275); Amazon area, Sao Paulo State, Brazil (Map 3C) yobena - Median apophysis without black wall (Figs. 267, 268); Panama, Colombia (Map 3C) benella 4(1). Fourth coxae with a macroseta or point- ed tubercle 5 Fourth coxae without macroseta or tu- bercle 15 METAZYGIA'Levi 79 5(4). 6(5). 7(6). 8(7). 9(8). 10(8). Fourth coxae with a tubercle; embohis of palpus thorn-like (Fig. 357, E in Fig. 359); Venezuela, Amazon area (Map 4D) lopez Fourth coxae with macroseta 6 Carapace with a lobe above first coxae (Fig. 390) 7 Carapace without lobe above first coxae 13 Abdomen posteriorly black (Fig. 388); Mato Grosso, Brazil (Map 4E) voxanta Abdomen posteriorly light In mesal view, median apophysis pro- jecting beyond tegulum (at 3 hr in Fig. 318, at 3 hr in Fig. 423) 9 Median apophysis smaller and not pro- jecting beyond tegulum (Figs. 313, 383, 4 1 8, 425 ) 1 0 Median apophysis very large, facing cymbium at 3 hr in Figure 423; Bahia, Brazil (Map 4H) atama Median apophysis distally rectangular in mesal view at 3 hr in Figure 318; Am- azon area (Map 4C) mariahelenae Embolus S-shaped as seen through trans- parent lamella (between 11 hr and at center in Fig. 313) octama - Embolus otherwise, usually hidden (Figs. 383, 418, 425) 11 11(10). Median apophysis with bulge on side (at 3 hr in Fig. 425, right of center in Fig. 426); coastal Ecuador (Map 41) oro Median apophysis without bulge (Figs. 383, 4 1 8 ) .' 1 2 12(11). Median apophysis subtriangular (at 3 hr in Fig. 418); Rio Grande do Sul, Brazil (Map 4H) cunha Median apophysis distally enlarged (at 4 hr in Fig. 383); Peruvian Amazon (Map 4E) genaro 13(6). Embolus a thread with a transverse loop as in Figures 112 and 113; Greater An- tilles (Map 21) crewi Embolus otherwise (Figs. 327, 332) 14 14(13). Median apophysis distally tapering to a point (at 3 hr in Fig. 327); Mexico, Central America (Map 4G) .... nigrocincta Median apophysis distally bulging as in 4 hr in Figure 332; southern Amazon region to Misiones Prov., Argentina (Map 4C) lagiana 420); eastern Para State, Brazil (Map 41) aldela 18(15). Abdomen with median, transverse light band (Fig. 417); palpus as in Figure 416; Amazon area (Map 41) cazeaca Abdomen and palpus otherwise 19 Palpus with terminal apophysis (A in Figs. 45, 112; top of Figs. 162, 201, 207) 20 Palpus without terminal apophysis, only a lamella (Figs. 142, 148, 213, 221, 225, 230, 237, 243-246) 38 8 20(19). Terminal apophysis with distal straight or slightly curved prong (Figs. 6, 13) 15(4). 16(15). 17(16). Carapace with small lobe above first coxa (Figs. 376, 415, 421) Carapace without lobe 18 Abdomen with ventral black band as in Figure 377; Peru (Map 4E) manu Abdomen marked otherwise 17 Sickle-shaped embolus (Fig. 414); Co- lombian Amazon (Map 4F) rothi Embolus barely visible, hidden by large sclerotized lamella (at 11 hr in Fig. 19(18). 21 - Terminal apophysis with short distal prong (between center and 2 hr in Fig. 91); Mato Grosso, Brazil (Map 2C) _ isabelae 21(20). In mesal view terminal apophysis with two prongs (between center and 2 hr in Fig. 6); southeastern United States to Costa Rica (Map 2A) wittfeldae - Terminal apophysis otherwise (Figs. 13 27) 22(21). A comb projecting beyond prong in mes al view (at 1 hr in Figs. 20, 27, 66) ... No comb-like projection (Figs. 13, 201 22 23 25 23(22). Comb-like projection longer than wide in mesal view (at 1 hr in Fig. 20); te- gulum with pointed spine (at 3 hr in Fig. 21); Costa Rica to Guianas and northern Peru, Galapagos Islands and West Indies (Map 2A) dubia - Comb-like projection in mesal view wid- er than long (at 1 hr in Figs. 27, 66); tegulum without pointed spine (Figs. 28, 67 ) 24 24(23). Comb small (at 1 hr in Fig. 66), most of embolus hidden behind conductor (Figs. 66, 67); northern Peruvian Am- azon (Map 2B) pastaza Comb large (at 1 hr in Fig. 27), most of embolus exposed, only tip of conduc- tor hidden (at 3 hr in Fig. 27); Ven- ezuela to Peruvian coast (Map 2D) ._. pimentel 25(22). A pointed tooth projecting beyond prong of terminal apophysis in mesal view (at 1 hr in Fig. 13, center in Fig. 14); Bahama Islands (Map 2A) bahama _ 26 No such tooth 16 26(25). Median apophysis projecting beyond other sclerites toward 4 hr, conductor toward 3 hr with triangular space be- tween these sclerites in mesal view (Figs. 201, 207); southern Brazil 27 Median apophysis and conductor oth- erwise (Figs. 60, 99, 104, 213) 28 27(26). Conductor with a knob at tip (at 3 hr in Fig. 201); median apophysis straight 80 Bulletin Museum of Comparative Zoology, Vol. 154, No. 2 (at 4 hr in Fig. 201); Rio Grande do Sul, Brazil (Map 3A) saturnino 39(38). Conductor tapering to tip (at 3 hr in Fig. 207); median apophysis with elbow (at 4 hr in Fig. 207); southeastern Brazil (Map 3B) crabroniphila 28(26). Tip of embolus with horseshoe-shaped structure (Fig. 104, center of Fig. 105); Guianas, Amazon area (Map 2F) . jamari 40(38). Tip of embolus otherwise, or hidden by conductor (Figs. 60, 99, 162) 29 - 29(28). Embolus long, saber-shaped, and curved 41(40). up (Fig. 99); southeastern Brazil (Map 2C) rogenhoferi Embolus otherwise (Figs. 73, 119, 162) 30 30(29). Embolus U-shaped (with dark cap as in Fig. 119); Guatemala (Map 21) carrizal 42(40). Embolus otherwise _ 31 31(30). Embolus a small hook as in center of Figure 162; western Amazon area (Map 43(42). 2K ) yiicumo Embolus otherwise 32 32(31). Embolus almost straight structure, tip hidden by terminal apophysis prong as in Figure 85; Neblina area of Am- azon (Map 2E) enabla - Embolus otherwise 33 - 33(32). Embolus thorn-shaped as in Figure 33; Peruvian Amazon to Bahia, Brazil (Map 2D) patiama Embolus otherwise; Mexico, Central 44(42). America 34 34(33). Embolus lamella (L in Figs. 45, 54) cov- ering most of embolus with one large - point and a forked tooth as in Figure 73; Central America (Map 2J) incerta 45(44). Embolus lamella otherwise (Figs. 44, 54, 60, 79) 35 - 35(34). Embolus lamella triangular tip covered by blister-like part of terminal apoph- 46(45). ysis (center in Fig. 60); southern Mex- ico to Honduras, Jamaica (Map 2B) ... chicanna - - Embolus lamella otherwise 36 36(35). Embolus lamella rounded as in Figure 47(45). 44; Florida, Texas to Honduras, Cuba, Jamaica (Map 2G) zilloides Embolus lamella otherwise 37 37(36). Embolus lamella with axis of tip at right angle to axis of cymbium (Fig. 54); 48(47). embolus without lobes (Fig. 54); Costa Rica to Trinidad and Colombia (Map 2G) „ keyserlingi Embolus lamella with tip pointing to- ward 1 hr in Figure 79; embolus with two lobes below (Figs. 79, 80); Mexico 49(48). to Guianas and Peru (Map 2E) . pallidula 38(19). Median apophysis long, thumb-shaped - and projecting beyond tegulum (at 3 hr in Figs. 302, 308) 39 Median apophysis otherwise 40 Embolus lamella projecting beyond cymbium edge as at 11 hr in Figure 308; Amazon area (Map 4B) castaneoscutata Cymbium edge extending beyond la- mella as at 11 hr in Figure 302; Am- azon area (Map 4A) vaupes Tegulum with a large fold having a comb (at 12 hr in Figs. 283, 295) 41 Tegulum otherwise 42 Embolus axis at right angle to axis of cymbium (Fig. 283); Amazon area to southeastern Brazil (Map 3D) voliiptifica Embolus axis forming an acute angle with that of cymbium (Fig. 295); Guyana (Map 4A) tanica Tegulum with a large distal lobe (at 12 hr in Figs, 237, 243) 43 Tegulum otherwise or only small lobe 44 Lamella with subparallel sides distally concave (at 3 hr in Fig. 243, at 10 hr in Fig. 244, L in Fig. 245); median apophysis with one point (at 5 hr in Fig. 243, M in Fig. 245); Bahia State to Rio Grande do Sul, Brazil (Map 3F) genialis Lamella distally narrowing (at 3 hr in Fig. 237); median apophysis with two points (at 5 hr in Fig. 237); southeast- ern Brazil (Map 3F) mundiilella Median apophysis "hanging down"; up- side-down T-shaped (at 6 hr in Fig. 250); Guianas (Map 3D) ikuruwa Median apophvsis otherwise (Figs. 142, 149, 213, 22 i, 230) 45 Lower edge of median apophysis semi- circular (at 5 hr in Figs. 213, 230) 46 Median apophysis otherwise (Figs. 142, 148, 221, 225) 47 Palpus as in Figure 230; Panama to Rio de Janeiro State, Brazil (Map 3F) laticeps Palpus as in Figure 213; Bolivia to Mato Grosso do Sul (Map 3B) corumba Embolus thread-like curving "above" te- gulum (at 12 hr in Fig. 142); Neblina area of Amazon (Map 2H) arnoi Embolus otherwise (Figs. 148, 221, 225) 48 Median apophysis having a "bottom" bulge and tip hidden by conductor in mesal view (Figs. 148, 149); Amazon area (Map 2H) curari Median apophvsis otherwise (Figs. 221, 225) ' 49 Median apophysis as in Figure 225; Am- azon area (Map 3B) uma Median apophysis as in Figure 221; Ec- uador to Amazonian Peru (Map 3B) . sendero Metazygia 'Levi 81 Metazygia wittfeldae (McCook) Figures 1-7; Map 2A Epeira wittfeldae McCook, 1894: 168, pi. 7, figs. 6, 7, 2, 6. Three female, two male, and one imm. male syntypes from Florida in ANSP, examined. Metazygia wittfeldae:— F. P.-Cambridge, 1904: 501, pi. 47, figs. 22, 23, 2, 6. Roewer, 1942: 368. Bonnet, 1957: 2820. Levi, 1977: 92, pi. 6, figs. 90-103, 2, S. Description. Female from Tabasco, Mexico. Carapace orange, cephalic region dark orange. Chelicerae dark orange. La- bium, endites, sternum orange. Coxae, legs orange. Dorsum of abdomen with gray pattern on white pigment spots (Fig. 5); venter light orange-gray, without marks. Posterior median eyes 0.5 diameter of an- terior medians, laterals 0.7 diameter. An- terior median eyes 0.7 diameter apart, 1.3 diameters from laterals. Posterior median eyes 0.2 diameter apart. Height of clypeus equals 0.5 diameter of anterior median eye. Total length 8.0 mm. Carapace 4.2 mm long, 3.1 wide, 1.9 behind lateral eyes. First femur 3.2 mm, patella and tibia 4.1, meta- tarsus 3.0, tarsus 1.2. Second patella and tibia 3.8 mm, third 2.3, fourth 3.1. Male from Tabasco, Mexico. Color as in female but cephalic region lighter. Pos- terior median eyes 0.7 diameter of anterior medians, laterals 0.7 diameter. Anterior median eyes 0.7 diameter apart, 1 diam- eter from laterals. Posterior median eyes 0.3 diameter apart, 2.1 diameters from lat- erals. Height of clypeus equals 0.5 diam- eter of anterior median eye. Second tibia thicker than first, both first and second with macrosetae. Total length 5.2 mm. Carapace 2.9 mm long, 2.1 wide, 1.1 be- hind lateral eyes. First femur 2.7 mm, pa- tella and tibia 3.5, metatarsus 3.0, tarsus 1.1. Second patella and tibia 3.1 mm, third 1.7, fourth 2.1. Note. Males and females are commonly collected together. Variation. Total length of females 7.2 to 11.1 mm, males 4.2 to 7.2. Illustrations were made from specimens from Tabasco State, Mexico. Diagnosis. In ventral view the epigyn- um is twice as wide as long and has pos- terior swellings on each side (Fig. 2); in related species M. dubia (Fig. 16) and M. bahama (Fig. 9), it is narrower and lacks these swellings. Males are separated from M. dubia and M. bahama by the soft prong parallel to the sclerotized prong of the ter- minal apophysis (at 1 hr in Fig. 6), which is absent in the other two. The distribution of M. wittfeldae is allopatric with respect to related species (Map 2A). Natural History. Specimens were col- lected under eaves of buildings and in brush and are commonly found in mud-dauber wasp nests. Distribution. From southeastern United States, Virginia to Costa Rica. Its distri- bution does not overlap that of M. bahama and M. dubia (Map 2A). United States and some Mexican records on Map 2 come from Levi (1977). Specimens Examined. MEXICO Tamaulipas: Ciudad Mante (AMNH); Tampico (AMNH). San Luis Potosi: Tamazuchale (AMNH, CAS); Valles (AMNH). Zacatecas: Tabasco (MCZ). Nayarit: Tepic (AMNH); 27 km S Acaponeta (CAS). Colima: Santiago, NW Manzanillo (AMNH). Veracruz: Acayucan (CAS); Catemaco (AMNH, CAS); 7.5 km W Catemaco; 17 km W Cerro Azul; Cordoba; Fortin de las Flores; Jalapa; La Palma; Lago Catemaca (all AMNH); Mo- cambo (CAS); Papantla; Tecolutla (all AMNH); Ve- racruz (AMNH, MCZ); Orizaba (MCZ). Guerrero: 13 km W Acapulco (AMNH). Oaxaca: Tehuantepec; 3.2 km NE Tehuantepec (all AMNH). Tabasco: Villa Hermosa (AMNH). Yucatan: Chetumal (MCZ). Chia- pas: Palenque Ruins (MCZ); Prusia (AMNH). BE- LIZE Stann Creek: Dangriga (MCZ); 80 km S Stann Creek (MCZ); Twin Cays, W of Swamp Dock (USNM). GUATEMALA Antigua; Moca; Suchitepequez; San Julian; Tiquizate; Variedades; Zacapa (all AMNH). HONDURAS Lancetilla, nr. E Tela (MCZ). NICA- RAGUA Lago Jiloi (SR). COSTA RICA Cartago: Turrialba (MCZ). Guanacaste: 4 km NW Caiias (MCZ); nr. Cafias (MCZ). San fose: San Jose, common in city (AMNH, MCZ). Metazygia baliama new species Figures 8-14; Map 2A Holotype. Male holotype, one male, two female para- types from South Bimini, Bahama Islands, June 1951 (M. A. Cazier), in AMNH. The specific name is a noun in apposition after the type locality. Description. Female paratype. Cara- pace orange-brown with sides of thoracic 82 Bulletin Museum of Comparative Zoology, Vol. 154, No. 2 region orange. Chelicerae red-brown. La- bium, endites dark orange. Sternum, cox- ae, legs orange. Dorsum of abdomen with pairs of black brackets (Fig. 12); venter light gray without marks. Posterior me- dian eyes 0.7 diameter of anterior medi- ans, laterals 0.7 diameter. Anterior median eyes 0.8 diameter apart, 2 diameters from laterals. Posterior median eyes 0.3 diam- eter apart, 4 diameters from laterals. Height of clypeus equals 0.8 diameter of anterior median eye. Total length 8.0 mm. Carapace 3.9 mm long, 2.7 wide, 1.7 be- hind lateral eyes. First femur 2.9 mm, pa- tella and tibia 3.7, metatarsus 2.5, tarsus 1.1. Second patella and tibia 3.4 mm, third 2.1, fourth 2.7. Male holotype. Color as in female. Pos- terior median eyes 0.6 diameter of anterior medians, laterals 0.6 diameter. Anterior median eyes 0.8 diameter apart, 1 diam- eter from laterals. Posterior median eyes 0.3 diameter apart, 2.2 diameters from lat- erals. Height of clypeus equals 0.5 diam- eter of anterior median eye. Second tibia slightly thicker than first, with stronger macrosetae. Abdomen widest in middle. Total length 5.1 mm. Carapace 2.8 mm long, 2.1 wide, 1.1 behind lateral eyes. First femur 2.8 mm, patella and tibia 3.7, meta- tarsus 2.8, tarsus 1.1. Second patella and tibia 3.1 mm, third 1.7, fourth 2.2. Note. Males and females were collected together. Variation. Total length of females 6.7 to 10.7 mm, males 4.8 to 5.7. Illustrations were made from the male holotype and a female paratype collected with it. Diagnosis. Epigynum of the female (Fig. 9) is narrower than that of M. wittfeldae in ventral view (Fig. 2), and the posterior median plate is wider dorsally (at 6 hr in Fig. 10) than that of M. dubia (at 6 hr in Fig. 17) in posterior view. The male differs from both of these species by having a spine on the subterminal apophysis (at 1 hr in Fig. 13 and center of Fig. 14). Natural History. This species probably has habits similar to M. wittfeldae and M. dubia. Distribution. Bahama Islands. The dis- tribution does not overlap that of M. dubia and M. wittfeldae (Map 2A). Paratypes. From type locality: June 1950, 62, 13, imm. (M. A. Cazier, F. Rindge, AMNH); May 1951, 59, 43, imm. (M. Ca- zier, W. J. Gertsch, AMNH); June 1951, 302, 13, imm. (M. A. Cazier; C, P. Vaurie, AMNH); July 1951, 13 (C, P. Vaurie, AMNH). Specimen Examined. BAHAMA ISLANDS Dog Key, N Andros Island, 13 May 1904, 16 (AMNH). Metazygia dubia (Keyserling) Plate 1 ; Figures 15-21 ; Map 2A Epeira dubia Keyserling, 1864: 123, pi. 4, figs. 12, 13, 9. Two female syntypes from Sta. Fe de Bogota, N. Granada [Bogota, Colombia], in BMNH, ex- amined. Keyserling, 1892: 187, pi. 9, fig. 138, 9. Epeira moraballii Kingston, 1932: 363, figs. 53, 54, web. Specimens from the Essquibo River, Guyana, lost [not in BMNH]. NEW SYNONYMY. Aranea dubia: — Roewer, 1942: 841. A. moraballii: — Roewer, 1942: 847. Araneus moraballicus Bonnet, 1955: 546. Metazygia dubia: — Levi, 1991a: 179. Synonymy. Hingston's E. moraballii is synonymized with M. dubia because Hingston described the proximity of the posterior median eyes, the oval abdomen, and the Zygiella x-notata-\ike web. Also, only a few Guianan Metazygia are 11 mm total length. (The other large Metazygia, M. laticeps, has spinnerets anterior of the posterior tip, a fact noticed by Hingston for his Epeira folisecens, but not here.) Figures 1-7. Metazygia wittfeldae (McCook). 1-5, female. 1-4, epigynum. 1, anterior. 2, ventral. 3, posterior. 4, lateral. 5, dorsal. 6, 7, left male palpus. 6, mesal. 7, apical. Figures 8-14. M. bahama n. sp. 8-1 2, female. 8-1 1 , epigynum. 8, anterior. 9, ventral. 1 0, posterior. 1 1 , lateral. 1 2, dorsal. 1 3, 14, male palpus. 13, mesal. 14, apical. METAZYGiA'Levi 83 Figures 15-21, M. dubia (Keyserling). 15-19, female. 15-18, epigynum. 15, anterior. 16, ventral. 17, posterior. 18, lateral. 19, dorsal. 20, 21, male palpus. 20, mesal. 21, apical. Figures 22-28. M. pimentel n. sp. 22-26, female. 22-24, epigynum. 22, ventral. 23, posterior. 24, lateral. 25, dorsal. 26, abdomen, ventral. 27, 28, male palpus. 27, mesal. 28, apical. Scale lines. 1.0 mm, genitalia 0.1 mm. 84 Bulletin Museum of Comparative Zoology, Vol. 154, No. 2 Bonnet (1955: 466) lists the name Epeira dubia erroneously as a synonym of Ara- neus cornutus [= Larinioides cornutus (Clerck 1758)]. Description. Female from Gamboa, Panama. Carapace brown, sides of thorac- ic region orange. Chelicerae brown. La- bium, endites light brown. Sternum or- ange. Coxae light orange, legs orange. Dor- sum of abdomen whitish with gray marks (Fig. 19); venter light gray without marks. Posterior median eyes 0.7 diameter of an- terior medians, laterals 0.7 diameter. An- terior median eyes their diameter apart, two diameters from laterals. Posterior me- dian eyes 0.3 diameter apart. Height of clypeus equals 0.5 diameter of anterior median eye. Total length 9.5 mm. Cara- pace 4.1 mm long, 2.9 wide, 2.2 behind lateral eyes. First femur 3.4 mm, patella and tibia 4.0, metatarsus 3.0, tarsus 1.2. Second patella and tibia 3.8 mm, third 2.1, fourth 3.3. Male from Gamboa. Color as in female, but carapace all orange. Posterior median eyes 0.7 diameter of anterior medians, lat- erals 0.7 diameter. Anterior median eyes their diameter apart, 1.3 diameters from laterals. Posterior median eyes 0.2 diam- eter apart. Height of clypeus equals 0.4 diameter of anterior median eye. Total length 6.7 mm. Carapace 3.5 mm long, 2.5 wide, 1.5 behind lateral eyes. First femur 3.5 mm, patella and tibia 4.4, metatarsus 3.7, tarsus 1.4. Second patella and tibia 3.8 mm, third 2.1, fourth 2.7. Note. Males and females were collected together. Variation. Total length of females 7.2 to 1 1.7 mm, males 4.8 to 6.2. The abdomen may have dark marks or sometimes none at all. The scape is torn off the epigynum in some females. The epigynum of a fe- male from Depto. Huila, Colombia, is heavily sclerotized. A male from Vene- zuela had the tegulum spine short. Illus- trations were made from a female and male from Gamboa, Panama; the specimen on Plate 1 is from Negril, Jamaica. Diagnosis. The Metazygia dubia epi- gynum (Fig. 16) is narrower in ventral view than that of M. wittfeldae (Fig. 2) and lacks the swelling on each side. It has a narrower scape (Figs. 16, 17) than that of M. bahama (Figs. 9, 10). Males can be separated from the more northern M. witt- feldae by the distal comb-like projection of the male terminal apophysis (at 1 hr in Fig. 20, center of Fig. 21) and by the long black spine on the tegulum (at 3 hr in Fig. 21). Natural History. This species was col- lected from disturbed areas outside and sometimes inside buildings. It was found under bark in Cuba and was collected from coral just above the high tide mark in Ja- maica; from pasture and from dense veg- etation in Jamaica; on a wire fence at Coa- mo, Puerto Rico; and under a roof over- hang and on a boat dock on Barro Colorado Island, Panama. Specimens from Galapa- gos were found in seashore vegetation. The spiders are nocturnal and sit in a silk re- treat during the day. There is no signal line to the retreat. The vertical web is re- built every evening and has a partly closed hub (Lubin, personal communication). Some specimens come from mud-dauber wasp nests. Distribution. Costa Rica, West Indies to Brazil and Peru and Galapagos Islands. It does not overlap M. wittfeldae (Map 2A). Specimens Examined. COSTA RICA Limon: Li- mon (DU). PANAMA Herrera: Sarigua (MIUP). Co- de: Rio Hato (MIUP). Colon: Santa Rosa (AMNH). Panama: very common (AMNH, MIUP, MCZ). CUBA Archip. Canarreos: Cayo Cantiles, Bajo Corteza (lESC); Cayo Rosario, Bajo Corteza (lESC); Cayo Avalos (MNHNC). JAMAICA very common (AMNH, MCZ). HISPANIOLA Dominican Republic: Samana: Las Terrenas (MNSD). Independencia: Bano de Zorsa (MNHNC); betw. Neiba and Duverge (MNSD). Dis- trito Nacional: Acuario Nacional, Santo Domingo (MNSD). PUERTO RICO Laguna Cartagena, 10 km SW Lajas (MCZ); Banos de Coamo (MCZ). CURA- gAO Hato (AMNH). VENEZUELA Sucre: Cumani (MCN). Monagas: Caripito (AMNH). Bolivar: Ca- naima [PCanaime] (AD). COLOMBIA Magdalena: Cienaga (IBNP); Pozo Colorado, 11 km W Santa Mar- ta (AMNH); San Pablo (IBNP). Atldntico: Barran- quilla (AMNH, IBNP). Antioquia: Mutata (MCZ). Huila: 10 km E Santa Leticia, Finca Meremberg, 2,300 m (MCZ). ECUADOR Sucumbios: Cuyabeno Metazygia 'Levi 85 Reserv. (MCZ). Pastaza: Rio Pastaza, Rio Verde, Mara Trail, 1,200 m (AMNH). Giiayas: 3 km NE La Lib- ertad (CAS). Galapagos IsL: W coast Albemarle Isl. (AMNH): Bahia Borrero, Santa Cruz (MCZ). PERU Libertad: Pacasma>o (PAN); Guadalupe (PAN). BRAZIL Amazonas: Rio Autas, Sta. Amelia (NRMS). Ceard: Pacajus Guarani, 3 July 1972, 29 (Exped. Acad. Bras. Cienc, MZSP 12408). ' Metazygia pimentel new species Figures 22-28; Map 2D Holotype. Male holotype and four female paratypes from Pimentel, Depto. Lambayeque, Peru, 21 Sept. 1988 (D. Silva D.), in MUSM; one paratype in MCZ. The specific name is a noun in apposition after the type locality. Description. Female paratype. Cara- pace orange, darkest in eye region. Che- licerae dark orange-brown Labium, en- dites, sternum light orange. Coxae light orange, legs orange. Dorsum of abdomen whitish with five pairs of black spots (Fig. 25); venter with indistinct white patch sur- rounded by dusky area (Fig. 26). Posterior median eyes 0.8 diameter of anterior me- dians, laterals 0.8 diameter. Anterior me- dian eyes 0.8 diameter apart, 2.2 diameters from laterals. Posterior median eyes 0.2 diameter apart. Height of clypeus equals 0.7 diameter of anterior median eye. Total length 8.2 mm. Carapace 4.5 mm long, 3.4 wide, 2.1 behind lateral eyes. First femur 3.4 mm, patella and tibia 4.1, metatarsus 2.7, tarsus 1.2. Second patella and tibia 3.8 mm, third 2.5, fourth 3.4. Male holotype. Color as in female but white patch in dusky area on venter of abdomen is more distinct. Posterior me- dian eyes 0.7 diameter of anterior medi- ans, anterior laterals 0.7 diameter, poste- rior laterals 0.6. Anterior median eyes 0.4 diameter apart, 1.2 diameters from later- als. Posterior median eyes 0.2 diameter apart. Height of clypeus equals 0.3 di- ameter of anterior median eye. Total length 5.7 mm. Carapace 3.1 mm long, 2.4 wide, 1.4 behind lateral eyes. First femur 2.9 mm, patella and tibia 3.7, metatarsus 2.7, tarsus 1.2. Second patella and tibia 3.0 mm, third 1.9, fourth 2.3. Note. Males and females were collected together. Variation. Total length of males 5.7 to 6.3. Illustrations were made from the type specimens. Diagnosis. The median, less sclerotized triangular area of the epigynum is more pointed (Fig. 22) than that of M. dubia (Fig. 16). The male has a comb-like pro- jection of the subterminal apophysis (at 1 hr in Fig. 27) but lacks the spine on the tegulum that is present in M. dubia (at 3 hr in Fig. 21). The venter of the abdomen has a white patch (Fig. 26) absent in sim- ilar species. Natural History. Specimens were abun- dant in branches of locust "algarrobos," Prosopis, SL leguminous tree growing in sand dunes in Peru, and dry to very dry tropical forest in Venezuela. Distribution. Venezuela, Peru in arid areas (Map 2D). Specimen Examined. VENEZL^ELA Falcon: Par- aguana Peninsula, 6 km W Nuevo Pueblo, 26 Nov.- 4 Dec. 1960, \i (A. L. Markezich, MCZ). Metazygia patiama new species Figures 29-34; Map 2D Holotype. Female holotype from Fazenda Matiapa, Camacan, Bahia, Brazil, 16 Oct. 1978 (J. S. Santos), in MCN no. 11116; male paratype, same locality and collector, 14 Oct. 1978, in MCN no. 10182. The specific name is an arbitrary combination of letters. Description. Female holotype. Cara- pace orange. Chelicerae orange-brown. Labium, endites dark orange. Sternum or- ange. Coxae light orange; legs dark orange, distal parts of articles darker. Dorsum of abdomen setose, with anterior pair of dark patches on dusky white and posterior quar- ter dark gray (Fig. 32). Venter black, fad- ing toward sides. Posterior median eyes 0.7 diameter of anterior medians, laterals 0.7 diameter. Anterior median eyes 0.5 di- ameter apart, 1.2 diameters from laterals. Posterior median eyes 0.2 diameter apart, 2.4 diameters from laterals. Height of clypeus equals 0.6 diameter of anterior median eye. Total length 5.6 mm. Cara- pace 2.8 mm long, 2.2 wide, 1.4 behind lateral eyes. First femur 2.5 mm, patella 86 Bulletin Museum of Comparative Zoology, Vol. 154, No. 2 and tibia 2.7, metatarsus 1.9, tarsus 0.9. Second patella and tibia 2.6 mm, third 1.6, fourth 2.3. Male paratype. Color light orange, ex- cept for abdomen, which has tiny white pigment spots dorsally and lacks all gray or black marks. Posterior median eyes 0.8 diameter of anterior medians, laterals 0.6 diameter. Anterior median eyes 0.8 di- ameter apart, 0.8 diameters from laterals. Posterior median eyes 0.2 diameter apart, 1.3 diameters from laterals. Height of clypeus equals 0.4 diameter of anterior median eye. Total length 4.4 mm. Cara- pace 2.3 mm long, 1.6 wide, 0.8 behind lateral eyes. First femur 2.1 mm, patella and tibia 2.5, metatarsus 1.7, tarsus 0.7. Second patella and tibia 2.3 mm, third 1.3, fourth 1.6. Note. Males and females were collected at the same locality. Variation. Total length of females 5.6 to 5.7 mm. Illustrations were made from the type specimens. Diagnosis. The epigynum (Fig. 29) is wider than that of M. pimentel in ventral view (Fig. 22), and in posterior view (Fig. 30) the lateral plates appear wider than those of M. pimentel (Fig. 23) and M. dubia (Fig. 17). The male palpus (Figs. 33, 34) lacks the comb-like projection of the terminal or subterminal apophysis found in M. dubia (at 1 hr in Fig. 20) and M. pimentel (at 1 hr in Fig. 28) and also lacks the spine on the tegulum. The pattern on the abdomen (Fig. 32) is not a complete folium as in related species. Natural History. Specimens were col- lected at night in Peru. Distribution. Amazon region of Peru to Bahia State, Brazil (Map 2D). Specimen Examined. PERU Madre de Dios: Zona Reservada de Manu, Puesto de Vigilancia Pakitza, 6 Oct. 1987, 12 (D. Silva D., J. Coddington, USNM). Metazygia ipago new species Figures 35-38; Map 2D Holotype. Female holotype from Igarapeagu, igapo capim flutuante (periodically flooded forest), Est. Pernambuco, Brazil, 12 July 1980, in MNRJ. The specific name is an arbitrary combination of letters. Description. Female holotype. Cara- pace orange, cephalic region dark orange. Chelicerae dark orange. Labium, endites dark orange. Sternum orange. Coxae light orange, legs orange. Dorsum of abdomen with a faint gray folium on white (Fig. 38); venter with some white pigment spots on light gray behind epigynum. Posterior median eyes 0.7 diameter of anterior me- dians, laterals 0.7 diameter. Anterior me- dian eyes 0.6 diameter apart, 1.2 diameters from laterals. Posterior median eyes 0.3 diameter apart. Height of clypeus equals 0.4 diameter of anterior median eye. Total length 6.8 mm. Carapace 3.4 mm long, 2.7 wide, 1.5 behind lateral eyes. First femur 3.1 mm, patella and tibia 4.0, metatarsus 2.7, tarsus 1.1. Second patella and tibia 3.5 mm, third 1.9, fourth 2.7. Diagnosis. In ventral view there is a groove to each side of the median area of the epigynum (Fig. 35). The groove is ab- sent in the epigynum of M. patiama (Fig. 29) and similar species. Metazygia zilloides (Banks) Figures 39-47; Map 2G Epeira zilloides Banks, 1898: 255, pi. 15, fig. 2, 2, 6. Three female, one male, and one juvenile syntype from Tepic, Nayarit, Mexico, in MCZ, examined. Aranea dilatata F. P.-Cambridge, 1904: 513, pi. 49, Figures 29-34. Metazygia patiama n. sp. 29-32, female. 29-31, epigynum. 29, ventral. 30, posterior. 31, lateral. 32, dorsal. 33, 34, left male palpus. 33, mesal. 34, apical. Figures 35-38. M. ipago n. sp.. female. 35-37, epigynum. 35, ventral. 36, posterior. 37, lateral. 38, dorsal. Figures 39^7. M. zilloides (Banks). 39-43, female. 39^2, epigynum. 39, anterior. 40, ventral. 41, posterior. 42, lateral. 43, dorsal. 44-46, male palpus. 44, mesal. 45-47, pulled apart. 45, mesal. 46, dorsal, cymbium removed. 47, emtxjius and lamella. METAZYCiA'Levi 87 Figures 48-54. M. keyserlingi Banks. 48-53, female. 48-51 , epigynum. 48, anterior. 49, ventral. 50, posterior. 51 , lateral. 52, dorsal. 53, atxiomen, ventral. 54, male palpus, mesal. Abbreviations. A, tenninal apophysis; B, blister-like subterminal apophysis; C, conductor; E, embolus; H, hematodocha; I, stipes; L, embolus lamella; M, median apophysis; P, paracymbium; R, radix; Y, cymbium. Scale lines. 1.0 mm, genitalia 0.1 mm. Bulletin Museum of Comparative Zoology, Vol. 154, No. 2 fig. 9, $. Male lectotype designated Levi, 1977: 92 from Guatemala, in BMNH, examined. Roewer, 1942: 841. Synonymized by Levi, 1977. Metazygia a/foonigrc— Bryant, 1940: 339, figs. 107- 109, 111,2, (5. Erroneous determination, noi Larinia albonigra Franganillo. Aranea zilloides: — Roewer, 1942: 857. Araneus pallidulus:—KTaus, 1955: 24, fig. 66, S. Er- roneous determination. Araneus dilatatus: — Bonnet, 1955: 497. Araneus zilloides: — Bonnet, 1955: 632. Metazygia zilloides: — Levi, 1977: 92. Description. Nontype female from Te- pic, Nayarit, Mexico. Carapace orange, dusky in midline. Chelicerae, labium or- ange-brown. Endites, sternum orange, sides darker. Coxae, legs orange, distal ends of femora and tibiae darker. Dorsum of ab- domen with white pigment spots and an- terior black marks (Fig. 43); venter with white pigment spots. Eyes subequal. An- terior median eyes 0.8 diameter apart, 1.2 diameters from laterals. Posterior median eyes 0.3 diameter apart, 2 diameters from laterals. Height of clypeus equals 0.4 di- ameter of anterior median eye. Total length 7.4 mm. Carapace 3.0 mm long, 2.6 wide, 1.3 behind lateral eyes. First femur 3.6 mm, patella and tibia 4.6, metatarsus 3.1, tarsus 1.1. Second patella and tibia 3.7 mm, third 2.1, fourth 3.0. Male from Tepic, Mexico. Color as in female, but abdomen darker with poste- rior transverse bars, and venter with a transverse white patch behind genital groove, which is surrounded by black. Pos- terior median eyes same diameter as an- terior medians, laterals 0.9 diameter. An- terior median eyes 0.9 diameter apart, 1 diameter from laterals. Posterior median eyes 0.2 diameter apart, 1.8 diameters from laterals. Height of clypeus equals 0.4 di- ameter of anterior median eye. Total length 4.8 mm. Carapace 2.6 mm long, 2.1 wide, 1.0 behind lateral eyes. First femur 3.2 mm, patella and tibia 4.5, metatarsus 3.6, tarsus 1.1. Second patella and tibia 3.4 mm, third 1.7, fourth 2.4. Note. Males and females were collected together. Variation. Total length of females 3.8 to 6.7 mm, males 3.0 to 5.0. Illustrations were made from nontype specimens from Tepic, Mexico. Diagnosis. Females are difficult to sep- arate from M. keyserlingi. The embolus part stuck in the opening is smaller (Fig. 40), and there is a median ventral groove in posterior view of the epigynum (arrow in Fig. 41). The male differs from that of M. keyserlingi (Fig. 54) by the round shape of the embolus lamella (L in Fig. 44). Natural History. Specimens were col- lected in second-growth forest edge in Mexico and from beach grape, in a hotel, and in a citrus orchard and pasture in Ja- maica. Others came from a Sceliphron wasp nest in Jamaica. Distribution. Florida, central Texas to Honduras, Bahamas, Cuba, Cayman Is- lands, and Jamaica (Map 2G). United States and some Mexican records of Map 2 come from Levi (1977). Specimens Examined. MEXICO Tamaulipas: Li- mon (AMNH); Mante (AMNH). Nuevo Leon: Linares (AMNH): Los Cristales (AMNH). San Luis Potosi: Tamazunchale (AMNH); Valles (AMNH). Nayarit: 3 km N Compostela (AMNH); Tepic (AMNH). Jalisco: Chapala (CAS); Puerto Vallarta (AMNH); Tizapan (AMNH). Veracruz: Catemaco (AMNH); Lago Ca- temaco (AMNH); Rio Blanco (MCZ); 4 km N Son- tecomapan (REL); Veracruz (USNM). Hidalgo: Ixmi- quilpan, Rio Tula (AMNH). Distrito Federal: (AMNH). Michoacan: Jiquilpan (AMNH); Lago Cha- pala (AMNH); Ciudad Michoacan (AMNH). Morelos: Cuernavaca (AMNH, MCZ); Tehui.xtla (AMNH). Oa- xaca: Temascal (MCZ); Tolosa (AMNH). Tabasco: 3 km NE Comalcalco (AMNH); Villa Hermosa (AMNH). Campeche: Ciudad del Carmen (AMNH). Yucatan: Chetumal (MCZ); Chicxulub (CAS). Chia- pas: N Arriaga Mtns. (AMNH); Cacahuatan (AMNH); 24 km SW Cintalapa (AD); 45 km SE Comitan (AMNH); Las Cruces (AMNH); Mapastepec (AMNH); Prusia (AMNH); Tonala (AMNH). GUATEMALA Guatemala: Amatitlan (AMNH). Quiche: Chichicas- tenango (AMNH). Sacatepequez: Antigua (AMNH); Capetillo, 1,500 m (AMNH). Suchitepequez: Moca (AMNH); Nebaj (AMNH); San Julian (AMNH); Var- iedades, 300 m (AMNH). Chimaltenango: Yepocapa (AMNH); San Pedro (AMNH). EL SALVADOR Can- delaria (AMNH), HONDURAS Copan (AMNH); 27 km S Tegucigalpa (MCZ). BAHAMA ISLANDS An- dros Island: Coakley Town (AMNH), CUBA Pinar del Rio: Cabafias (AMNH); S Pinar del Rio (AMNH); San Vicente (AMNH), La Habana: Habana (MCZ, USNM), Matanzas: Cienaga de Zapata (MCZ); Ma- tanzas (AMNH). Villa Clar-a: Vega Alta (MCZ). Cien- Metazygia 'Levi 89 fuegos: Soledad (MCZ); Trinidad Mtns., Mina Carlota (MCZ); Topes de Collantes (AMNH). Camagiiey: Agramonte (AMNH); San Bias (MCZ). Holguin: Banes (AMNH): Valle de Maybe (MNHNC). Santiago: Si- bonev (AMNH); Santiago (AMNH); coast below Pico Turquino (MCZ). CAYMAN ISLANDS Grand Cay- man (MCZ). JAMAICA Christiana (AMNH); Clare- mont (MCZ); Evanton (MCZ); Fort Henderson (AMNH); Hope Gardens (AMNH); Kingston (MCZ); Long Mtn. (MCZ); Lucea (AMNH); Mandreville (AMNH); Mona (MCZ); Negril (MCZ); Old Harbour (MCZ); Port Henderson (MCZ); 1 km E Reading (MCZ); Spanish Town (MCZ). Metazygia keyserlingi Banks Plate 1 ; Figures 48-54; Map 2G Metazygia keyserlingi Banks, 1929: 94, pi. 4, fig. 63, 9. Five female syntypes from Barro Colorado Is- land, Canal Zone [Lago Gatun, Panama Prov., Pan- ama], in MCZ, examined. Roewer, 1942: 868. Bon- net, 1957: 2820. Synonymy. Banks designated two vials with females as types, two females col- lected on 20-24 June and three females on 13 July (both without year). This species had been erroneously synonymized with M. zilloides (Levi, 1977). When describing M. keyserlingi. Banks compared it to M. pallidula but not to his own M. zilloides, which is more similar to M. keyserlingi than is M. pallidula. Description. Female from Barro Colo- rado Island. Carapace orange. Chelicerae, labium, endites orange. Sternum orange. Legs orange. Dorsum of abdomen whitish with two indistinct black longitudinal bands (Fig. 52). Venter with some white pigment behind epigynum, dark dusky on each side between epigynum and spin- nerets, and with a faint white line on each side (Fig. 53). Posterior median eyes same diameter as anterior medians, laterals 0.8 diameter. Anterior median eyes 0.8 di- ameter apart, 1 diameter from laterals. Posterior median eyes 0.3 diameter apart, 1.3 diameters from laterals. Height of clypeus equals 0.4 diameter of anterior median eye. Total length 4.5 mm. Cara- pace 1.9 mm long, 1.5 wide, 0.9 behind lateral eyes. First femur 2.1 mm, patella and tibia 2.5, metatarsus 1.7, tarsus 0.7. Second patella and tibia 1.9 mm, third 1.1, fourth 1.7. Male from Barro Colorado Island. Color as in female. Eyes subequal. Anterior me- dian eyes 1.2 diameters apart, 1 diameter from laterals. Posterior median eyes 0.5 diameter apart. Height of clypeus equals 0.6 diameter of anterior median eye. Total length 3.6 mm. Carapace 1.8 mm long, 1.3 wide, 0.7 behind lateral eyes. First femur 2.0 mm, patella and tibia 2.5, metatarsus 2.0, tarsus 0.8. Second patella and tibia 2.0 mm, third 1.0, fourth 1.4. Note. Males and females were collected together. Variation. Total length of females 3.5 to 5.7 mm, males 2.3 to 3.6. Illustrations were made from specimens collected from the type locality: Barro Colorado Island, Gatun Lake, Panama. The web photo- graph (Pi. 1) is also from Barro Colorado Island, Panama. Diagnosis. The white patch on the ven- ter of the abdomen (Fig. 53) is more dis- crete and the groove present in M. zilloides (arrow in Fig. 41) is smaller or absent (Fig. 50). The male can be told from M. zilloides (Fig. 44) by the embolus lamella (L in Fig. 54), which is pointed "above" the embolus. The females are difficult to separate from M. zilloides, but all the ones collected with males had the central swollen part of the epigynum 0.32 mm wide, whereas those of M. zilloides had the swollen part of the epigynum 0.40 mm wide. The epigynum of M. keyserlingi has a slightly smaller scape than that of M. zilloides. Natural History. This species was col- lected in moist tropical forest in Costa Rica, in leaf litter in Panama, and in a garden in Call, Colombia. Males are uncommon in collections. Distribution. Costa Rica to southern Co- lombia, Trinidad (Map 2G). Specimens Examined. COSTA RICA Limon: Ca- huita, 31 Mar. 1979, 25 (J. Coddington, MCZ). Pun- tarenas: Osa, Parque Nacional Corcovado, 15 Aug. 1978, 12 (J, Coddington, MCZ); Manuel Antonio Na- tional Park, 24-26 Mar. 1983, 29, 1<5 (D. Ubick, DU); Osa Peninsula, Llorona Station, 6 Aug. 1980, 19 (J. Coddington, USNM). PANAMA Colon: Fort Sher- man, Aug. 1939, 19 (A. M. Chickering, MCZ); Fort Gulick, 23 Feb. 1980, IS (Harlan, AMNH). Panama: 90 Bulletin Museum of Comparative Zoology, Vol. 154, No. 2 Barro Colorado Island, Lago Gatun, very common, 19, 13 (AMNH, MCZ); Pipeline Road, Soberania Natl. Park, June 1978, 59 (F. Vollrath, MCZ), 7 Aug. 1983, 19 (H., L. Levi, MCZ); Toboga Island, 23 Aug. 1946, 29 (N. L. H. Krauss, AMNH). TRINIDAD St. George Co.: Simla, 6.4 km N Arima, 2 May 1967, 19 (C. T. Collins, AMNH); Arima Valley, 10-22 Feb. 1964, IS (P. Wygodzinsky, AMNH). COLOMBIA Valle: Cali, 19 Oct. 1969, 19, 2 Oct. 1969, 19; Rio Jamundi, 1,000 m, 18 km S Cali, 9 July 1969, 19, 14 Jan. 1970, 19, 17 June 1970 (all W. Eberhard, MCZ); 4 Mar. 1973, 39 (W. Eberhard, H. Levi, MCZ); 10 km N Piendamo, 1,700 m, Feb. 1974, 19, 16 (W. Eberhard, MCZ). Narino: nr. Barbacoas, 20 m, 20 Mar. 1974, 19 (W. Eberhard 727, MCZ). Metazygia chicanna new species Plate 1 ; Figures 55-61 ; Map 2B Holotype. Female holotype and female and two male paratypes from Chicanna Ruins, 8 km W Xpujii, ca. 18°32'N, 89''31'W, Campeche, Mexico, 12-14 July 1983 (W. Maddison), in MCZ. The specific name is a noun in apposition after the type locality. Description. Female holotype. Cara- pace orange, with median dusky band. Chelicerae, labium, endites, sternum or- ange. Coxae, legs orange. Dorsum of ab- domen with two pairs of black bands, sides black (Fig. 58); venter with a central white patch (Fig. 59). Posterior median eyes 0.8 diameter of anterior medians, laterals 0.8 diameter. Anterior median eyes 0.8 di- ameter apart, 1 diameter from laterals. Posterior median eyes 0.2 diameter apart, 1.6 diameters from laterals. Height of clypeus equals 0.6 diameter of the anterior median eyes. Total length 5.5 mm. Cara- pace 2.2 mm long, 1.5 wide, 0.8 behind lateral eyes. First femur 2.1 mm, patella and tibia 2.7, metatarsus 1.8, tarsus 0.7. Second patella and tibia 2.3 mm, third 1.3, fourth 1.9. Male paratype. Color as in female, but legs indistinctly ringed darker, and ab- domen with two pairs of longitudinal bands. Posterior median eyes 0.9 diameter of anterior medians, laterals 0.8 diameter. Anterior median eyes 0.9 diameter apart, 0.8 diameter from laterals. Posterior me- dian eyes 0.2 diameter apart, 1 .8 diameters from laterals. Height of clypeus equals 0.6 diameter of anterior median eye. Total length 3.5 mm. Carapace 1.9 mm long, 1.4 wide, 0.7 behind lateral eyes. First femur 2.1 mm, patella and tibia 2.6, metatarsus 2.1, tarsus 0.9. Second patella and tibia 2.1 mm, third 1.1, fourth 1.6. Note. Males and females were collected together. Variation. Total length of females 2.5 to 5.8 mm, males 2.5 to 4.2. Illustrations were made from paratypes; the photo- graph (Pi. 1) was made in Negril, Jamaica. Diagnosis. All specimens have a white spot on the venter of the abdomen (Fig. 59). Females can be distinguished by the ventral view of the epigynum, which has a cone-shaped scape with openings on each side (Fig. 55). The scape also has a kink in lateral view (Fig. 57). Unlike other species, the male has a cone-shaped embolus la- mella (at center of Fig. 60). Natural History. The holotype came from short tropical rain forest. Other spec- imens came from moist forest border in Quintana Roo, Mexico, border of forest road; from roadside, on top of woody shrubs without leaves; and from beach grape in Jamaica. Distribution. From Yucatan Peninsula, Mexico, to Honduras, Jamaica (Map 2B). Figures 55-61 . Metazygia ctiicanna n. sp. 55-59, female. 55-57, epigynum. 55, ventral. 56, posterior. 57, lateral. 58, dorsal. 59, atxjomen, ventral. 60, 61 , left male palpus. 60, mesal. 61 , apical. Figures 62-65. W. tapa n. sp., female. 62-64, epigynum. 62, ventral. 63, posterior. 64, lateral. 65, dorsal. Figures 66-68. M. pastaza n. sp., male. 66, 67, male palpus. 66, mesal. 67, apical. 68, dorsal. Figures 69-73. M. incerta (O. P.-Cambridge). 69-72, female. 69-71 , epigynum. 69, ventral. 70, posterior. 71 , lateral. 72, dorsal. 73, male palpus. METAZYCiA'Levi 91 Figures 74-80. M. pallidula (Keyserling). 74-78, female. 74-76, epigynum. 74, ventral. 75, posterior. 76, lateral. 77, dorsal. 78, abdomen, ventral. 79, male palpus. 80, embolus, with and without cap. Abbreviations. C, conductor; L, embolus lamella; M, median apophysis. Scale lines. 1.0 mm, genitalia 0.1 mm. 92 Bulletin Museum of Comparative Zoology, Vol. 154, No. 2 Specimens Examined. MEXICO Campeche: Be- can, 18°33'N, 89°30'W, 31 July 1991, 16 (W. Piel, G. S. Bodner, MCZ); Ceiba Playa [?], 2 Aug. 1949, 12 (C. J. Goodnight, AMNH). Yucatan: Chichen Itza, Nov. 1945, 19 (H. Wagner, AMNH). Quintana Roo: Chetumal, 28 June 1975, 15 (W. C. Sedgwick, MCZ); Chancanah Cozumel, 8 Aug. 1949, 19 (C. J. Good- night, AMNH); Reserva de Sian Ka'an, km 5, 4 June 1991, many 9, 1 pe and male paratype from Barueri, Est. Sao Paulo, Brazil, 8 Sept. 1965 (K. Lenko), in MZSP no. 4026. The specific name is a noun in apposition after the t\pe locality. Description. Female holotype. Cara- pace light orange, cephalic region orange. Chelicerae brown. Labium, endites or- Metazygia 'Levi 99 ange. Sternum light orange. Coxae light orange, legs dusky orange. Dorsum of ab- domen with folium outlined by dusky brackets and with a median longitudinal dusky line (Fig. 103). Venter light, without marks. Posterior median eyes 0.6 diameter of anterior medians, laterals 0.7 diameter. Anterior median eyes 0.8 diameter apart, 1 diameter from laterals. Posterior median eyes 0.4 diameter apart. Height of clypeus equals 0.5 diameter of anterior median eye. Total length 5.5 mm. Carapace 2.4 mm long, 1.9 wide, 1.1 behind lateral eyes. First femur 2.1 mm, patella and tibia 2.3, meta- tarsus 1.7, tarsus 0.9. Second patella and tibia 2.1 mm, third 1.3, fourth 1.9. Diagnosis. Unlike the epigynum of oth- er species, the area behind and on the side of the scape is wrinkled (Figs. 100-102). Metazygia jamari new species Figures 104-106; Map 2F Holotype. Male holotype from Jamari, Rondonia, Brazil, 23 Jan. 1989 (Equipe Operacjao Jamari), in MCN no. 18550. The specific name is a noun in apposition after the type locality. Description. Male holotype. Carapace orange. Chelicerae, labium, endites or- ange. Sternum orange. Legs orange. Dor- sum of abdomen white with a pair of lon- gitudinal dark bands (Fig. 106); venter gray. Carapace rebordered above first coxa. Posterior median eyes 0.8 diameter of an- terior medians, laterals 0.7 diameter. An- terior median eyes 0.8 diameter apart, 0.6 diameter from laterals. Posterior median eyes 0.2 diameter apart, 1.4 diameters from laterals. Height of clypeus equals 0.6 di- ameter of anterior median eye. Second tib- ia thinner than first. Total length 2.5 mm. Carapace 1.5 mm long, 1.1 wide, 0.5 be- hind lateral eyes. First femur 1.7 mm, pa- tella and tibia 2.1, metatarsus 1.6, tarsus 0.7. Second patella and tibia 1.6 mm, third 0.8, fourth 1.2. Note. I was not successful matching this male to a female. Variation. Total length of males 2.4 to 2.8. The Porto Velho male differed slightly in the shape of the embolus. It is not known if the horse shoe-shaped structure above the embolus (Figs. 104, 105) breaks off when mating. Illustrations were made from the holotype. Diagnosis. Metazygia jamari differs from other species by the U-shaped tip of the embolus and by the straight terminal apophysis (at 12 hr to 2 hr in Fig. 104, left in Fig. 105). Distribution. Amazon region (Map 2F). Specimens Examined. SURINAM Marowijne: Lawa River, Anapaike Village, 8-29 Nov. 1963, 16 (B. Malkin, AMNH). BRAZIL Roraima: Ilha de Mar- aca, Rio Uraricoera, 29 Mar. 1987, IS (A. A. Lise, MCN 20060). Rondoina: Porto Velho, Rio Tapirape, Feb., Mar. 1963, 16 (P. Pinheiros, AMNH). Metazygia crew! (Banks) Plate 1; Figures 107-113; Map 21 Singa crewi Banks, 1903: 342, pi. 15, fig. 8, 9. Female holotype from Haiti, lost (not in AMNH, ANSP, cue, MCZ, USNM). Roewer, 1942: 877. Larinia coamensis Petrunkevitch, 1930: 335, figs. 221- 224, 9. Female holotype from Coamo Springs, Puer- to Rico, in PMY, examined. Roewer, 1942: 771. Bonnet, 1957: 2348. First synonymized by Bryant 1945. Aranea crewi: — Bryant, 1945: 364, figs. 1-3, S. Araneus crewi: — Bonnet, 1955: 471. Metazygia crewi: — Harrod, Levi, and Leibensper- ger, 1991: 245. Description. Female from Puerto Plata, Dominican Republic. Carapace dusky or- ange, black between eyes, margin of tho- racic region darker. Chelicerae dusky or- ange. Labium, endites orange. Sternum light yellow-brown, black on each side, black fusing posteriorly to form a V shape. Legs light orange, distally with some dark rings. Dorsum of abdomen with dusky me- dian band, a white cardiac mark, a white band on each side and black on sides (Fig. 110). Venter black with a square white patch posterior to genital groove, black surrounding spinnerets, sides light (Fig. 111). Carapace narrow in front without thoracic depression. Posterior median eyes 0.6 diameter of anterior medians, laterals 0.7 diameter. Anterior median eyes 0.6 di- ameter apart, 0.7 from laterals. Posterior median eyes 0.4 diameter apart, 2 diam- 100 Bulletin Museum of Comparative Zoology, Vol. 154, No. 2 eters from laterals. Height of clypeus equals 0.4 diameter of anterior median eye. Total length 5.0 mm. Carapace 2.1 mm long, 1.6 wide, 0.7 behind lateral eyes. First femur 2.3 mm, patella and tibia 2.5, metatarsus 1.9, tarsus 0.8. Second patella and tibia 1.9 mm, third 1.2, fourth 1.8. Male specimen from Puerto Plata, Do- minican Republic. Color as in female. Thoracic depression a longitudinal line. Posterior median eyes 0.7 diameter of an- terior medians, anterior laterals 0.7, pos- terior 0.5. Anterior median eyes 1 diam- eter apart, slightly less from laterals. Pos- terior median eyes 0.4 diameter apart, 2 diameters from laterals. Height of clypeus equals 0.8 diameter of anterior median eye. First coxa with hook, fourth with a small macroseta. Total length 3.7 mm. Carapace 1.9 mm long, 1.5 wide, 0.6 behind lateral eyes. First femur 2.3 mm, patella and tibia 2.9, metatarsus 2.2, tarsus 0.9. Second pa- tella and tibia 2.0 mm, third 1.1, fourth 1.7. Note. Males and females were collected together. Variation. Total length of females 4.0 to 5.5 mm, males 3.3 to 3.7. Some females have three teeth on the anterior margin of the chelicerae, two on the posterior; others have four and three. The epigynum's transparent scape can be short (Fig. 107) or long and overhanging the anterior bor- der of the epigynum; often it appears torn off with only a dark round scar. Illustra- tions were made from specimens collected in Puerto Plata, Dominican Republic; the web (Pi. 1) was photographed near Mar- icao, Puerto Rico. Diagnosis. The female is separated from other species by having an epigynum that, when viewed posteriorly, has two long slits, separated by a narrow median plate (Fig. 108). The male is distinguished by having a palp with filamentous embolus and over- hanging terminal apophysis (A) (Figs. 112, 113). Natural History. Specimens were col- lected sweeping in forest on St. Johns, Vir- gin Islands, and in a coffee plantation at Jayuya, Puerto Rico. Many specimens came from an abandoned, dry, sunny road on a south-facing slope near Maricao, Puerto Rico, at 800 m elevation. The spiders had retreats in the heads of grass above a small transparent web (Pi. 1). These spiders could not be dislodged with a sweep net but had to be collected individually. Distribution. Greater Antilles, Virgin Islands (Map 21). Specimens Examined. CUBA [no local], 19 (R. V. Chamberlin, AMNH). HAITI Cap Haitien, Mar. 1934, 29 (Utowana Exped., MCZ); Dame Marie, 1941, 29, IS (A. Audant, MCZ); Port-au-Prince, July 1941, 139, 2(5 (A. Audant, MCZ); 18-21 July 1955, 49, 2S, 8 imm. (A. F. Archer, AMNH); 21, 22 Mar. 1969, 29; 30 Aug- 7 Sept. 1969, 29, U (L. Reynolds, MCZ), DOMINI- CAN REPUBLIC Balneario Saladilla, S Barahona, 8 Aug. 1958, 19 (A. F. Archer, AMNH); 5 km NW Las Matas de Farfan, 25 Aug, 1970, 19 (B, Patterson, MCZ); Loma de Los Pinos, Colonia Ramfis, T. Valdez, 700-900 m, 7 Aug, 1958, 29 (A, F, Archer, E, Boynie Moya, AMNH); Puerto Plata, July-Aug, 1941, 239, 2(5 (D. Hurst, MCZ); La Gran Chorra, Altagracia, 11 Apr, 1992, 39 (F, Del Monte, MNSD), PUERTO RICO Jayuya, 1,000 m, 20-26 Mar, 1986, 29, imm, (H., L. Figures 107-113. Metazygia crew/ (Banks), 107-111, female. 107-109, epigynum, 107, ventral, 108, posterior, 109. lateral, 110, dorsal. Ill, atxjomen, ventral, 112, 113, left male palpus. 111, mesal, 112, ventral. Figures 114-118, M. vaurieorum n, sp,, female, 114-116, epigynum, 114, ventral, 115, posterior, 116, lateral, 117, dorsal, 118, abdomen, ventral. Figures 119, 120. M. carrizaln. sp., male. 119, palpus. 120, dorsal. Figures 121-124, M. taman n. sp., female. 121-123, epigynum. 121, ventral. 122, posterior. 123, lateral. 124, dorsal. Figures 125-128, M. paquisha female. 125-127, epigynum. 125, ventral. 126, posterior. 127, lateral, 128, dorsal. Figures 129-132. M. nobasn. sp., female. 129-131, epigynum, 129, ventral, 130, posterior, 131, lateral, 132, dorsal. METAZYCIA'Levi 101 Figures 133-137. M. goeldiin. sp., female. 133-136, epigynum. 133, ventral. 134, posterior. 135, lateral. 136, paratype, ventral. 137, dorsal. Abbreviations. A, terminal apophysis; C, conductor; M, median apophysis; R, radix. Scale lines. 1.0 mm, genitalia 0.1 mm. 102 Bulletin Museum of Comparative Zoology, Vol. 154, No. 2 Levi, MCZ); Reserva Forestal, Maricao, Monte de Estado, 800 m, 4-6 Apr. 1989, 132, 16 imm. (H., L. Levi, MCZ). U.S. VIRGIN ISLANDS St. Johns, forest above Cinnamon Bav, 17 Mar. 1970, 12, 13 (H., L., F. Levi, MCZ), Metazygia vaurieorum new species Figures 114-118; Map 21 Holotype. Female holotype and one paratype from \ariedades, H-Ol'N, 90°30'W, Suchitepequez, 120- 170 m, Guatemala 27-31 Aug. 1947, and two fe- male paratypes, 1-4 July 1947 (C. and P. Vaurie), in AMNH. The species is named after the collectors. Note. The location of the type locaUty is pubhshed in Vaurie and Vaurie (1949). Description. Female holotype. Cara- pace orange, cephalic region darker. Che- licerae dark orange-brown. Labium, en- dites, sternum orange. Coxae legs orange. Dorsum of abdomen with gray folium pat- tern on white (Fig. 117). Venter with trans- verse white patch behind epigynum and semicircular white line in front of spin- nerets (Fig. 118). Posterior median eyes 0.7 diameter of anterior medians, laterals 0.7 diameter. Anterior median eyes 0.7 di- ameter apart, 1.3 diameters from laterals. Posterior median eyes 0.2 diameter apart. Height of clypeus equals 0.5 diameter of anterior median eye. Total length 6.0 mm. Carapace 3.5 mm long, 2.7 wide, 1.5 be- hind lateral eyes. First femur 3.4 mm, pa- tella and tibia 4.1, metatarsus 3.2, tarsus IT. Second patella and tibia 3.6 mm, third 2.0, fourth 2.7. Diagnosis. In ventral view the epigyn- um (Fig. 114) resembles that of M. crewi (Fig. 107), but the scape is a large knob rather than a flat plate (perhaps a piece has broken off). It differs in posterior view by having a wide median plate (Fig. 115). Metazygia carrizat new species Figures 119, 120; Map 21 Holotype. Male holotype from Mataquescuintla, El Carrizal, Depto. Jalapa, Guatemala, beating foliage at river, 25 Apr. 1982 (S. Fend), in CAS. The spe- cific name is a noun in apposition after the type locality. Description. Male holotype. Carapace dusky orange to black. Chelicerae dusky orange. Labium black, endites orange. Sternum black. Coxae light orange, legs orange with dark rings. Dorsum of abdo- men with median longitudinal light band and gray band to the side, whitish color to the sides of gray (Fig. 120). Venter dark with indistinct white pigment spots in cen- ter. Posterior median eyes 0.7 diameter of anterior medians, laterals 0.7 diameter. Anterior median eyes 0.8 diameter apart, 0.8 diameter from laterals. Posterior me- dian eyes 0.3 diameter apart, 2 diameters from laterals. Height of clypeus equals 0.3 diameter of anterior median eye. Second tibia thicker than first, with long macro- setae on both first and second tibiae. Total length 4.2 mm. Carapace 2.1 mm long, 1.5 wide, 0.8 behind lateral eyes. First femur 2.3 mm, patella and tibia 3.1, metatarsus 2.3, tarsus 0.9. Second patella and tibia 2.3 mm, third 1.2, fourth 1.8. Note. This was collected near a M. vau- rieorum female. However, it is considered to represent another species, as it has a different coloration: black sternum in male, light in female; white above cardiac area in male (Fig. 120), dark in female (Fig. 117). Diagnosis. This species differs from all others by the terminal apophysis hanging "down" with its axis at an acute angle to the axis of the cymbium and by the shape of the sclerotized embolus lamella (at cen- ter of Fig. 119) and by the U-shaped em- bolus with a pointed cap (Fig. 119). Metazygia taman new species Figures 121-124; Map 21 Holotype. Female holotype from near Taman, ca. 16 km SW of Tamazunchale on Highway 85, 21°irN, 98°53'W, ca, 300 m, San Luis Potosi State, Mexico, 11 June 1983 (W, Maddison, R. S. Anderson), in MCZ. The specific name is a noun in apposition after the t\pe locality , Description. Female holotype. Cara- pace orange, cephalic area dusky. Chelic- erae, labium, endites orange-brown. Ster- num dusky orange. Coxae, legs orange, dis- tal articles darker. Dorsum of abdomen with paired dark gray marks on orange- Metazycia 'Levi 103 gray (Fig. 124); venter uniformly dark gray. Posterior median eyes 0.8 diameter of anterior medians, laterals 0.7 diameter. Anterior median eyes 0.7 diameter apart, 1.4 diameters from laterals. Posterior me- dian eyes 0.3 diameter apart. Posterior me- dian eyes oval, the long diameter from median posterior, to lateral and anterior. Height of clypeus equals 0.6 diameter of anterior median eye. Total length 7.7 mm. Carapace 3.9 mm long, 3.0 wide, 1.6 be- hind lateral eyes. First femur 3.3 mm, pa- tella and tibia 3.9, metatarsus 2.5, tarsus 1.1. Second patella and tibia 3.4 mm, third 2.1, fourth 3.1. Diagnosis. This species has a thinner scape (Figs. 121-123) than does M. pa- quisha (Figs. 125-127), but the scape is rounded (Fig. 123) and the posterior me- dian plate is wider (Fig. 122) than that of M. paquisha (Fig. 126). Metazygia paquisha new species Figures 125-128; Map 2H Holotype. Female holotype from Alto Rio Comaina, Puesto de Vigilancia 22, "Falso Paquisha, " Cor- dillera del Condor, border with Ecuador, 05°02'S, 78°5rW, Depto. Amazonas, Peru, night collecting, 23 Oct. 1987 (D. Silva D.), in MUSM. The specific name is a noun in apposition after the type locality. Description. Female holotype. Cara- pace orange, darkest anterior. Chelicerae orange. Labium, endites, sternum dusky orange. Legs dusky orange, darkest dis- tally. Abdomen gray without marks (Fig. 128). Carapace with a median line of setae. Posterior median eyes 0.8 diameter of an- terior medians, laterals 0.7 diameter. An- terior median eyes 0.6 diameter apart, 1.1 diameters from laterals. Posterior median eyes 0.3 diameter apart. Height of clypeus equals 0.6 diameter of anterior median eye. Total length 7.2 mm. Carapace 3.1 mm long, 2.3 wide, 1.3 behind lateral eyes. First femur 2.5 mm, patella and tibia 3.3, meta- tarsus 2.3, tarsus 0.9. Second patella and tibia 3.1 mm, third 2.1, fourth 2.5. Illustration. Figures 125-128 were made from the holotype. Diagnosis. The female differs from oth- ers by the wide massive scape (Figs. 125- 127) and from M. taman (Figs. 121-124) by the more pointed scape and narrower posterior plate (Figs. 126, 127). Distribution. Amazon Region (Map 2H). Specimen Examined. VENEZUELA Amazonas: Cerro de la Neblina, base camp, 140 m, 0°50'N, 66°10'W, low foliage, 21-28 Feb. 1985, 12 (W. E. Steiner, USNM). Metazygia nobas new species Figures 129-132; Map 2H Holotype. Female holotype from Bafios, 1,600 m, Tungurahua Prov., Ecuador, July 1938, and para- type from Banos, 2,000 m, July-Aug. 1938 (W. C. Macintyre), in MCZ. The specific name is an ar- bitrary combination of letters. Description. Female holotype. Cara- pace orange. Chelicerae, labium, endites orange-brown. Sternum dark orange. Legs orange. Dorsum of abdomen light with a faint folium and spots (Fig. 132); venter gray without marks. Posterior median eyes 0.8 diameter of anterior medians, laterals 0.8 diameter. Anterior median eyes 0.8 di- ameter apart, 1.7 diameters from laterals. Posterior median eyes 0.2 diameter apart, 2.5 diameters from laterals. Height of clypeus equals 0.6 diameter of anterior median eye. Total length 8.6 mm. Cara- pace 4.0 mm long, 2.9 wide, 1.9 behind lateral eyes. First femur 3.0 mm, patella and tibia 3.6, metatarsus 2.3, tarsus 1.1. Second patella and tibia 3.2 mm, third 2.2, fourth 2.9. Diagnosis. The epigynum (Figs. 129- 131) differs from that of the Mexican M. taman (Figs. 121-123) and the Colombian M. chenevo (Figs. 177-179) by having the median anterior edge of the base sclero- tized (at 12 hr in Fig. 129) and the pos- terior median plate heart-shaped (Fig. 130). Metazygia nobas differs from M. pa- quisha (Fig. 126) in the wider posterior median plate (Fig. 130). The posterior margin of the epigynum base in ventral view has a notch on each side (at 7 hr and 5 hr in Fig. 129) resembling that of M. voluptifica (Fig. 280). 104 Bulletin Museum of Comparative Zoology, Vol. 154, No. 2 Metazygia goeldii new species Figures 133-137; Map 2K Holotype. Female holotype from Goeldi Museum, Belem, Est. Para, Brazi'l, 10 Feb. 1959 (A. M. Nad- ler), in AMNH. The species is named after the BraziUan naturaUst E. Goldi. Description. Female holotype. Cara- pace orange. Chelicerae, labium, endites dark orange. Sternum orange. Coxae light- est orange, legs dusky orange. Dorsum of abdomen light with gray folium (Fig. 137); venter light dusky. Posterior median eyes 0.7 diameter of anterior medians, laterals 0.7 diameter. Anterior median eyes 0.6 di- ameter apart, 1.2 diameters from laterals. Posterior median eyes 0.2 diameter apart, 2 diameters from laterals. Height of clyp- eus equals 0.3 diameter of anterior median eye. Total length 6.0 mm. Carapace 2.8 mm long, 2.3 wide, 1.2 behind lateral eyes. First femur 2.4 mm, patella and tibia 2.9, metatarsus 2.0, tarsus 0.9. Second patella and tibia 2.5 mm, third 1.7, fourth 2.3. Variation. Total length of females 4.7 to 6.0 mm. The shape of the scape in the two specimens differs (Figs. 133, 136). Il- lustrations (Figs. 133-135, 137) were made from the female holotype. Diagnosis. The abdomen is oval, widest in anterior half (Fig. 137). Metazygia goel- dii differs from others by the knob-shaped scape (Figs. 133-135), which appears curved in lateral view (Fig. 135). Specimen Examined. BRAZIL Pard: Belem, Inst. Agron., 11 Feb. 1959, 19 (A. Nadler, AMNH). Metazygia adisi new species Figures 138-141; Map 2H Holotype. Female holotype from Lago do Jose, Ma- naus, Amazonas State, Brazil, 9 Aug. 1987 (J. Adis at al), in MCN no. 20058. The species is named after the collector. Description. The female holotype has lost all white pigment including the silver tapetum; it apparently was collected in a preservative other than alcohol (Levi, 1989). The specimen is all orange, except for a black band around anterior of ab- domen which is in the middle (Fig. 141). The eye region is black. Posterior median eyes same diameter as anterior medians, laterals 0.7 diameter. Anterior median eyes 0.7 diameter apart, 0.5 diameter from lat- erals. Posterior median eyes 0.2 diameter apart, 1 diameter from laterals. Height of clypeus equals 0.3 diameter of anterior median eye. Abdomen slightly flattened anteriorly (Fig. 141). Total length 3.7 mm. Carapace 1.8 mm long, 1.3 wide, 0.7 be- hind lateral eyes. First femur 1.9 mm, pa- tella and tibia 2.5, metatarsus 1.8, tarsus 0.6. Second patella and tibia 2.1 mm, third 1.2, fourth 1.8. Diagnosis. Unlike other species, Meta- zygia adisi has framed bulges on each side of the scape of the epigynum (Figs. 138, 139). Metazygia arnoi new species Figures 142, 143; Map 2H Holotype. Male holotype from Morro dos Seis Lagos, Parque Nacional do Pico da Neblina, Est. Ama- zonas, Brazil, 3 Oct. 1990 (A. A, Lise), in MCP. The species is named after the collector. Description. Male holotype. Carapace orange. Chelicerae orange. Labium, en- dites, orange. Sternum orange, slightly dusky. Coxae; legs orange. Dorsum of ab- domen light gray with faint dusky marks, pairs of spots and outline of folium (Fig. 143); venter light gray. Carapace with me- dian longitudinal line. Posterior median eyes 0.8 diameter of anterior medians, lat- erals 0.7 diameter. Anterior median eyes 0.6 diameter apart, 0.4 diameter from lat- erals. Posterior median eyes 0.3 diameter apart, 1.5 diameters from laterals. Height of clypeus equals 0.5 diameter of anterior median eye. Total length 5.1 mm. Cara- pace 2.6 mm long, 2.1 wide, behind lateral eyes 0.8 wide. First femur 2.4 mm, patella and tibia 2.7, metatarsus 2.0, tarsus 0.8. Second patella and tibia 2.3 mm, third 1.4, fourth 1.9. Note. This might be the male of M. paquisha. Diagnosis. The palpus (Fig. 142) lacks a terminal apophysis and differs from all others by the loop of the thread-shaped embolus (at 12 hr in Fig. 142). Met AZYGiA* Levi 105 149 Figures 138-141. Metazygia adisin. sp., female. 138-140, epigynum. 138, ventral. 139, posterior. 140, lateral. 141, dorsal. Figures 142, 143. M. arnoin. sp., male. 142, left male palpus. 143, dorsal. Figures 144-149. M. curarin. sp. 144-147, female. 144-146, epigynum. 144, ventral. 145, posterior. 146, lateral. 147, dorsal. 148, 149, male palpus. 148, mesal. 149, ventral. Figures 150-153. M. bolivia n. sp., female. 150-152, epigynum. 150, ventral. 152, posterior. 152, lateral. 153, dorsal. Figures 154-163. M. yucumon. sp., 154-161, female. 154-156, 158-160, epigynum. 154, 158, ventral. 155, 159, posterior. 156, 160, lateral. 157, 161, dorsal. 154-157, (Colombia). 158-161, (Bolivia). 162, 163, male. 162, palpus. 163, left femur, subventral. Scale lines. 1.0 mm, genitalia 0.1 mm. 106 Bulletin Museum of Comparative Zoology, Vol. 154, No. 2 Metazygia curari new species Figures 144-149; Map 2H Holotype. Female holotype from Ilha de Curari, Ma- naus, Est. Amazonas, Brazil, 3 Aug. 1987 (J. Adis et al), in MCN no. 20055. The specific name is a noun in apposition after the type locality. Description. Female holotype. Cara- pace yellowish white, little black between eyes. Chelicerae, labium, endites yellowish white. Sternum, legs yellowish white. Dor- sum of abdomen with two longitudinal white lines on dusky white (Fig. 147); ven- ter yellowish white. Posterior median eyes 0.8 diameter of anterior medians, laterals 0.7 diameter. Anterior median eyes 1.2 di- ameters apart, 0.4 diameter from laterals. Posterior median eyes 0.3 diameter apart, 1.3 diameters from laterals. Height of clypeus equals 0.4 diameter of anterior median eye. Total length 2.7 mm. Cara- pace 1 .25 mm long, 0.97 wide, 0.49 behind lateral eyes. First femur 1.38 mm, patella and tibia 1.61, metatarsus 1.11, tarsus 0.49. Second patella and tibia 1.32 mm, third 0.82, fourth 1.28. Male. Color as in female but distal ar- ticles of legs dusky. Posterior median eyes 0.9 diameter of anterior medians, laterals 0.8 diameter. Anterior median eyes 0.4 di- ameter apart, 0.2 diameter from laterals. Posterior median eyes 0.1 diameter apart, 1.2 diameters from laterals. Height of clypeus equals 0.6 diameter of anterior median eye. Endite with pointed tooth. Second tibia thicker than first, proximally swollen with macrosetae. Total length 2.5 mm. Carapace 1.32 mm long, 1.05 wide, 0.66 behind lateral eyes. First femur 1.30 mm, patella and tibia 1.61, metatarsus 1.13, tarsus 0.59. Second patella and tibia 1.32 mm, third 0.84, fourth 1.09. Note. Males and females were matched because both have similar median white lines on the abdomen and little black pig- ment between eyes. Their collecting lo- calities are within 10 km of each other. A recently collected female had a green ab- domen. Diagnosis. The pair of longitudinal white lines on the abdomen (Fig. 147) is a distinctive character. In ventral view of the epigynum, the median plate overlaps the lateral plates along their posterior mar- gins (at 8 hr, and at 4 hr in Fig. 144). The palpus lacks a terminal apophysis and the embolus has a distinctive shape (Fig. 148). Specimens Examined. BRAZIL Amazonas: Ma- naus. Canal Januari, mixed water forest, 16, 17 June 1987, IS (H. Hofer, IXPA); Ilha de Marchantaria, Rio Solimoes, 3°15'S, 59°58'W. 2 Sept. 1992, 19 (J. Adis et al., INPA). Metazygia botivia new species Figures 150-153; Map 2F Holotype. Female holotype from Est. Biologico Beni, Zone 3, ca. 14°47'S, 66°15'W, ca. 225 m, Depto. Beni, Bolivia, 8-14 Nov. 1989 (J. Coddington, S. Larcher, C. Griswold, D. Silva D., E. Panaranda), in lELP. The specific name is a noun in apposition after the type locality. Description. Female holotype. Cara- pace orange, sides of thoracic region light- est. Chelicerae orange-brown. Labium, endites dark orange. Sternum orange. Cox- ae light orange, legs dusky orange. Dorsum of abdomen with white pigment spots and pairs of gray brackets (Fig. 153); venter light gray. Posterior median eyes 0.7 di- ameter of anterior medians, laterals 0.7 diameter. Anterior median eyes 0.8 di- ameter apart, 1.3 diameters from laterals. Posterior median eyes 0.4 diameter apart. Height of clypeus equals 0.3 diameter of anterior median eye. Total length 6.7 mm. Carapace 3.0 mm long, 2.0 wide, 1.5 be- hind lateral eyes. First femur 2.2 mm, pa- tella and tibia 2.5, metatarsus 1.8, tarsus 0.8. Second patella and tibia 2.3 mm, third 1.5, fourth 2.0. Diagnosis. The elongate abdomen (Fig. 153) suggests that this species may belong to another genus, as yet unnamed, whose species have a cylindrical abdomen. How- ever, the epigynum is of the characteristic shape found in Metazygia, although the flat scape is larger than in other species (Figs. 150-152). Metazygia yucumo new species Figures 154-163; Map 2K Holotype. Female holotype, female and male para- types from 26.9 km SW Yucumo, 500 m, ca. 15°23'S, 66°59'W, Depto. Beni, BoHvia, 15-19 Nov. 1989 (J. METAZYGIA'Levi 107 Coddington, C. Griswold, D. Silva D., S. Larcher, E. Panaranda), in USNM. The specific name is a noun in apposition after the type locafity. Description. Female holotype. Cara- pace light orange-yellow. Chelicerae, la- bium, eridites orange-yellow. Sternum, coxae orange-yellow. Legs orange-yellow, except distal half of first femora with brown ring, underside black, and first two tibiae and distal articles brown, underside darker. Dorsum of abdomen white without pat- tern (Figs. 157, 161); vente^ light gray. Posterior median eyes 0.8 diameter of an- terior medians, laterals 0.8 diameter. An- terior median eyes their diameter apart, 0.7 diameter from laterals. Posterior me- dian eyes 0.3 diameter apart, 2 diameters from laterals. Height of clypeus equals 0.4 diameter of anterior median eye. Total length 4.2 mm. Carapace 2.0 mm long, 1.4 wide, 0.8 behind lateral eyes. First femur 2.3 mm, patella and tibia 2.7, metatarsus 1.8, tarsus 0.7. Second patella and tibia 2.2 mm, third 1.1, fourth 1.8. Male. Color as in female. Posterior me- dian eyes 0.8 diameter of anterior medi- ans, laterals 0.8 diameter. Anterior median eyes their diameter apart, 0.5 diameter from laterals. Posterior median eyes 0.3 diameter apart, 1 diameter from laterals. Height of clypeus equals 0.4 diameter of anterior median eye. First femur with many macrosetae (Fig. 163). Total length 4.2 mm. Carapace 1.9 mm long, 1.3 wide, 0.6 behind lateral eyes. First femur 2.3 mm, patella and tibia 2.9, metatarsus 2.3, tarsus 0.9. Second patella and tibia 2.3 mm, third 1.1, fourth 1.7. Note. Males and females were collected together. Variation. Total length of females 3.7 to 4.5 mm. The openings of the epigynum may be indistinct and hard to see (Fig. 154); some have a lip laterally (Fig. 158). Illustrations were made from the female holotype (Figs. 158-161) and a female from Mitu, Colombia (Figs. 154-157), and the male paratype. Diagnosis. The female is separated from others by the epigynum, which has an in- distinct pair of anterior median openings on each side of a thin scape (at 11 hr and at 1 hr in Figs. 154, 158). The male has a brush of ventral setae on the underside of the femur (Fig. 163), and the tip of the embolus is hook-shaped (Fig. 162). Distribution. Upper Amazon region (Map 2K). Specimens Examined. COLOMBIA Vaupes: Mitii, 200 m, Feb. 1975, 19 (P. A. Schneble, MCZ). PERU Hudnuco: Divisoria, 1,700 m, 23 Sept. to 3 Oct. 1946, 19 (F. Woytkowski, AMNH). Madre de Dios: El Li- monal, Alto Rio Madre de Dios, 21 July 1988, night collecting, 19 (P. Lozada, MUSM). Metazygia ipanga new species Figures 164-167; Map 2K Holotype. Female holotype from Museum Park, Ipi- ranga, Sao Paulo, Est. Sao Paulo, Brazil, 6 Dec. 1960 (J. Luiz), in MZSP no. 7642. The specific name is an arbitrary combination of letters. Note. Ipiranga is the district of the city of Sao Paulo in which the MZSP is located. Description. Female holotype. Cepha- lothorax very light orange, only eyes with black pigment. Dorsum of abdomen whit- ish with faint black band around anterior (Fig. 167). Venter whitish without pig- ment. Posterior median eyes same diam- eter as anterior medians, laterals 0.8 di- ameter. Anterior median eyes 0.7 diame- ter apart, 0.5 diameter from laterals. Pos- terior median eyes 0.2 diameter apart, 1 diameter from laterals. Height of clypeus equals 0.3 diameter of anterior median eye. Total length 4.5 mm. Carapace 1.9 mm long, 1.4 wide, 0.8 behind lateral eyes. First femur 2.2 mm, patella and tibia 2.5, meta- tarsus 2.0, tarsus 0.6. Second patella and tibia 2.0 mm, third 1.2, fourth 1.8. Variation. The holotype lacks white pigment; it appears washed out by the pre- serving fluid (Levi, 1989). The specimen from the Chaco Prov., Argentina, had both sides of the abdomen white and white pig- ment spots anterior to the black band. The Bolivian specimen had black pigment be- tween the eyes and a square white pigment area on the venter of the abdomen. The holotype was illustrated. Diagnosis. The abdomen is spherical (Fig. 167). This species differs from M. 108 Bulletin Museum of Comparative Zoology, Vol. 154, No. 2 uraricoera by the shape of the posterior median plate of the epigynum (Fig. 165). Distribution. Sao Paulo State, northern Argentina to Bolivia (Map 2K). Specimens Examined. BOLI\'IA Beni: Est. Biol. Beni, HMT'S, 66°15'W, ca. 225 m, 8-14 Nov. 1989, 19 (J. Coddington et al., USNM). ARGENTINA Cha- co: Selvas del Rio Oro, 27 Jan. 1965, 19 (M. E. Galiano, MEG). Metazygia uraricoera new species Figures 168-171; Map 2K Holotype. Female holotype from llha de Maraca, Rio Uraricoera, Est. Roraima, Brazil, 20 Mar. 1987 (A. A. Lise), in MCN no. 20064. The specific name is a noun in apposition after the type locality. Description. Female holotype. Cara- pace orange, eye region black. Chelicerae dark orange. Labium, endites dark orange. Sternum orange. Legs orange. First femur with a distal black spot on underside, distal end of first tibia dark. Dorsum of abdomen white with an anterior pair of black spots (Fig. 171). Venter with a white longitu- dinal band, starting at the side of book- lung covers. Posterior median eyes 0.8 di- ameter of anterior medians, laterals 0.7 diameter. Anterior median eyes 0.3 di- ameter apart, 0.2 diameter from laterals. Posterior median eyes 0.2 diameter apart, 1.5 diameters from laterals. Height of clypeus equals 0.3 diameter of anterior median eye. All patellae with a posterior, blunt tubercle. Total length 3.1 mm. Car- apace 1.4 mm long, 1.1 wide, 0.5 behind lateral eyes. First femur 1.4 mm, patella and tibia 1.6, metatarsus 1.2, tarsus 0.5. Second patella and tibia 1.3 mm, third 0.8, fourth 1.1. Variation. Total length of females 3.1 to 4.4 mm. The specimen from Surinam had the epigynum slightly shorter in pos- terior view and the abdomen more elon- gate. Illustrations were made from the ho- lotype. Diagnosis. The epigynum of this spe- cies, unlike that of others, has a sclerotized notch anteriorly on each side (at 10 hr and at 2 hr in Fig. 168) and a long narrow posterior median plate (Fig. 169). Natural History. The specimen collect- ed in Guyana came from forest savanna. Distribution. Guianas to northern Brazil (Map 2K). Specimens Examined. GUYANA Canje River, Iku- ruvva, 5°50'N, 57°50'W, Aug.-Sept. 1961, 19 (G. Bent- ley, AMNH). SURINAM Marowijne: Lawn River, Anapaike Village, Nov, 1963, 19 (B. Malkin, AMNH). BRAZIL Amazonas: Morro dos Seis Lagos, Parque Nacional do Pico da Neblina, 5 Oct. 1990, 19 (A. A. Lise, MCP). Metazygia serian new species Figures 172-176; Map 21 Holotype. Female holotype from La Selva, 4 km SE of Puerto Viejo, Heredia Prov., Costa Rica, 7 Aug. 1980, probably from wasp nest (R. Coville), in MCZ. The specific name is an arbitrary combination of letters. Description. Female holotype. Cara- pace orange, cephalic region brown, eye region black. Chelicerae dark brown to black. Labium, endites brown. Sternum orange. Coxae, legs orange. Dorsum of ab- domen with dense white pigment spots and transverse black band around anterior (Fig. 175). Venter white with a large white square between epigynum and spinnerets (Fig. 176). Posterior median eyes 0.9 di- ameter of anterior medians, laterals 0.8 diameter. Anterior median eyes 0.8 di- ameter apart, 0.3 diameter from laterals. Posterior median eyes 0.3 diameter apart, 1.2 diameters from laterals. Height of Figures 164-167. M. ipanga n. sp., female. 164-166, epigynum. 164, ventral. 165, posterior. 166, lateral. 167, dorsal. Figures 168-171. M. uraricoera n. sp.. female. 168-170, epigynum. 168, ventral. 169, posterior. 170, lateral. 171, dorsal. Figures 172-176. M. serian n. sp., female. 172-174, epigynum. 172, ventral. 173, posterior. 174, lateral. 175, dorsal. 176, abdomen, ventral. Metazygia • Levi 109 Figures 177-180. M. chenevon. sp., female. 177-179, epigynum. 177, ventral. 178, posterior. 179, lateral. 180, dorsal. Figures 181-184. M. lazepa n. sp., female. 181-183, epigynum. 181, ventral. 182, posterior. 183, lateral. 184, dorsal. Figures 185-188. M. atalaya n. sp., female. 185-187, epigynum. 185, ventral. 186, posterior. 187, lateral. 188, dorsal. Figures 189-192. M. corima n. sp., female. 189-191, epigynum. 189, ventral. 190, posterior. 191, lateral. 192, dorsal. Figures 193-196. M. uratron n. sp., female. 193-195, epigynum. 193, ventral. 194, posterior. 195, lateral. 196, dorsal. Scale lines. 1 .0 mm, genitalia 0.1 mm. 110 Bulletin Museum of Comparative Zoology, Vol. 154, No. 2 clypeus equals 0.7 diameter of anterior median eye. Total length 5.0 mm. Cara- pace 1.8 mm long, 1.5 wide, 0.9 behind lateral eyes. First femur 2.2 mm, patella and tibia 2.6, metatarsus 1.8, tarsus 0.6. Second patella and tibia 2.2 mm, third 1.2, fourth 1.9. Variation. Total length of females 4.8 to 5.0 mm. Living specimens are green (Eberhard, in letter). Illustrations were made from the holotype. Diagnosis. The abdomen is subspheri- cal, widest in anterior half (Fig. 175). This species resembles the tetragnathid genus Chrysometa in general appearance and epigynum (but not the coloration). In ven- tral view, the epigynum has openings close to the posterior margin (Fig. 172), unlike any other Metazygia species. Paratype. COSTA RICA Heredia: La Selva, Jan. 1978, 1$ (W. Eberhard no. 1279, MCZ). Metazygia chenevo new species Figures 177-180; Map 3A Holotype. Female holotype from Finca Chenevo, 175 m elev., 20 km S El Porvenir, 20 km N Rio Muco, Depto. Meta, Colombia, 1978 (W. Eberhard, no. 1386), in MCZ. The specific name is a noun in apposition after the type locality. Description. Female holotype. Cara- pace orange, cephalic region darkest. Che- licerae orange-brown. Labium, endites or- ange. Sternum orange. Coxae lighter or- ange, legs dusky orange. Dorsum of ab- domen with three longitudinal dusky bands on white (Fig. 180), sides and venter gray. Posterior median eyes 0.6 diameter of an- terior medians, laterals 0.6 diameter. An- terior median eyes 0.8 diameter apart, 1.5 diameters from laterals. Posterior median eyes 0.2 diameter apart, 3 diameters from laterals. Height of clypeus equals 0.5 di- ameter of anterior median eye. Total length 6.0 mm. Carapace 2.9 mm long, 2.2 wide, 1.3 behind lateral eyes. First femur 2.7 mm, patella and tibia 3.1, metatarsus 2.1, tarsus 0.9. Second patella and tibia 2.7 mm, third 1.7, fourth 2.2. Diagnosis. Unlike the Ecuadorian M. nobas (Fig. 130), the posterior median plate of the epigynum has a lateral constriction in posterior view (Fig. 178). Distribution. Amazon drainage, Colom- bia to Guyana (Map 3A). Specimen Examined. GUYANA Rupununi Savan- na, swamp, Sept. -Oct. 1989, 12 (S. Djojosudharmo, F. Mees, CD). Metazygia lazepa new species Figures 181-184; Map 3A Holotype. Female holotype from Hacienda Mozam- bique, 15 km SW Puerto Lopez, 200 m, Depto. Meta, Colombia, Aug. 1978 (W. Eberhard, no. 1812), in MCZ. The specific name is an arbitrary com- bination of letters. Description. Female holotype. Cara- pace orange, cephalic region orange- brown. Chelicerae brown. Labium, en- dites orange brown. Sternum, coxae or- ange; legs dark orange. Dorsum of abdo- men light gray with pairs of gray brackets (Fig. 184); venter gray without markings. Posterior median eyes 0.7 diameter of an- terior medians, laterals 0.7 diameter. An- terior median eyes 0.5 diameter apart, 1 diameter from laterals. Posterior median eyes 0.2 diameter apart, 2.5 diameters from laterals. Height of clypeus equals 0.6 di- ameter of anterior median eye. Total length 8.8 mm. Carapace 3.8 mm long, 3.1 wide, 2.0 behind lateral eyes. First femur 3.4 mm, patella and tibia 3.9, metatarsus 2.9, tarsus 1 .2. Second patella and tibia 3.4 mm, third 2.1, fourth 3.1. Diagnosis. Unlike other Metazygia, the posterior plate of the epigynum is wide and completely covers the lateral plates (Figs. 181, 182). The holotype was illus- trated. Distribution. Colombia, Venezuela (Map 3A). Specimen Examined. VENEZUELA Carabobo: San Esteban, 26 Jan. 1940, 19 (P. Andruze, AMNH). Metazygia atalaya new species Figures 185-188; Map 3A Holotype. Female holotype from Atalaya, Rio Car- bon, night collecting, Depto. Cuzco, Peru, 23 Sept. Metazygia 'Levi 111 1987 (D. Silva D.), in MUSM. The specific name is a noun in apposition after the type locality. Description. Female holotype. Cara- pace orange. Chelicerae, labium, endites dark orange. Sternum orange. Legs dusky orange. Abdomen gray without any marks (Fig. 188). Posterior median eyes 0.7 di- ameter of anterior medians, anterior lat- erals 0.7 diameter, posterior 0.6. Anterior median eyes 0.7 diameter apart, 1.1 di- ameters from laterals. Posterior median eyes 0.2 diameter apart. Posterior median eyes slightly oval. Height of clypeus equals 0.3 diameter of anterior median eye. The abdomen is slightly pointed anteriorly (Fig. 188). Total length 7.5 mm. Carapace 3.2 mm long, 2.4 wide, 1.4 behind lateral eyes. First femur 2.9 mm, patella and tibia 3.4, metatarsus 2.5, tarsus 1.2. Second patella and tibia 3.1 mm, third 1.9, fourth 2.7. Diagnosis. The abdomen is slightly pointed anteriorly (Fig. 188). The epigyn- um of this species differs from that of oth- ers by the unusual shape of the scape (Figs. 185-187). Metazygia corima new species Figures 189-192; Map 3B Holotype. Female holotype from Carimagua, 100 m, Depto. Meta, Colombia, grass, brush along fence, Oct. 1973 (W. Eberhard), in MCZ. The specific name is an arbitrary combination of letters. Description. Female holotype. Cara- pace orange, cephalic region brown. Che- licerae brown. Labium, endites orange. Sternum dark orange. Coxae, legs orange. Dorsum of abdomen with a median lon- gitudinal light band that is bordered by darker ones (Fig. 192). Sides, venter gray. Posterior median eyes 0.7 diameter of an- terior medians, anterior laterals 0.7 di- ameter, posterior 0.6. Anterior median eyes 0.6 diameter apart, 1 diameter from lat- erals. Posterior median eyes 0.4 diameter apart, 2 diameters from laterals. Height of clypeus equals 0.4 diameter of anterior median eye. Total length 5.5 mm. Cara- pace 2.7 mm long, 1.7 wide, 1.1 wide be- hind lateral eyes. First femur 2.3 mm, pa- tella and tibia 2.7, metatarsus 1.9, tarsus 0.8. Second patella and tibia 2.4 mm, third 1.5, fourth 2.1. Diagnosis. The epigynum and scape re- sembles that of Araneus tiganus (Cham- berlin) (Levi, 1991a, figs. 9-11); the scape is oval in ventral view (Fig. 189). Metazygia uratron new species Figures 193-196; Map 3A Holotype. Female holotype from Fazenda Santo An- tonio, Uruguca, Bahia, Brazil, 24 Oct. 1978 (J. S. Santos), in MCN no. 1097. The specific name is an arbitrary combination of letters. Description. Female holotype. Cara- pace orange. Chelicerae dark orange. La- bium, endites orange. Sternum light or- ange, with irregular dusky spots. Coxae light orange, legs orange. Dorsum of ab- domen dusky with darker gray folium out- line (Fig. 196). Venter gray without marks. Posterior median eyes 0.8 diameter of an- terior medians, laterals 0.7 diameter. An- terior median eyes 0.7 diameter apart, 1.5 diameters from laterals. Posterior median eyes 0.2 diameter apart. Height of clypeus equals 0.3 diameter of anterior median eye. Total length 7.0 mm. Carapace 3.4 mm long, 2.5 wide, 1.5 behind lateral eyes. First femur 3.1 mm, patella and tibia 3.3, meta- tarsus 2.3, tarsus 1.1. Second patella and tibia 2.7 mm, third 1.9, fourth 2.7. Diagnosis. This species differs from oth- er Metazygia by the shape of the large scape (Figs. 193-195). Metazygia saturnine new species Figures 197-202; Map 3A Holotype. Male holotype and immature male from Barragem Saturnine de Brito, Santa Maria, Rio Grande do Sul, Brazil, 7 July 1982 (M. Rosenau), in MCN no. 10596. The specific name is a noun in apposition after the type locality. Description. Female from Rio Grande, Rio Grande do Sul. Carapace with cephalic region brown, thoracic region light orange. Chelicerae brown. Labium, endites, brown. Sternum dusky orange. Coxae light or- ange; legs orange to brown, first two legs 112 Bulletin Museum of Comparative Zoology, Vol. 154, No. 2 and distal articles darkest. Dorsum ot ab- domen with gray pattern on light gray (Fig. 200); venter light gray. Posterior me- dian eyes 0.6 diameter of anterior medi- ans, laterals 0.6 diameter. Anterior median eyes 0.7 diameter apart, 0.7 diameter from laterals. Posterior median eyes 0.3 diam- eter apart. Height of clypeus equals 0.6 diameter of anterior median eye. Total length 5.3 mm. Carapace 2.5 mm long, 2.1 wide, 1.2 behind lateral eyes. First femur 2.1 mm, patella and tibia 2.5, metatarsus 1.7, tarsus 0.8. Second patella and tibia 2.3 mm, third 1.3, fourth 1.8. Male holotype. Color as in female but sternum and coxae are both orange. Pos- terior median eyes 0.8 diameter of anterior medians, laterals 0.8 diameter. Anterior median eyes 0.7 diameter apart, 0.7 di- ameter from laterals. Posterior median eyes 0.2 diameter apart, 2 diameters from lat- erals. Height of clypeus equals 0.7 diam- eter of anterior median eye. First coxa with minute hook. Second tibia as thick as first. Total length 4.5 mm. Carapace 2.3 mm long, 1.9 wide, 0.8 wide behind lateral eyes. First femur 2.7 mm, patella and tibia 3.3, metatarsus 2.5, tarsus 1.1. Second patella and tibia 2.8 mm, third 1.4, fourth 1.9. Note. Males and females were matched because of similar size and dark cephalic region of carapace. Variation. The epigynum of the female is probably broken off (Figs. 197-199). Diagnosis. The base of the epigynum appears broken. The epigynum has a long, wide posterior median plate (Fig. 198). The male differs from that of M. crabroniphila (Fig. 207) by the shape of the conductor and the much larger radix (Fig. 201). Specimen Examined. BRAZIL Rio Grande do Siil: Rio Grande, Est. Ecologica do Taim, 15 Oct. 1985, 19 (H. Buckup, MCN 13559). Metazygia crabroniphila Strand Figures 203-207; Map SB Aranea (Metazygia) crabroniphila Strand, 1915: 117. Female and male syntypes from Joinville, Est. Santa Catarina, Brazil, in SMF, examined. Metazygia crabroniphila: — Roewer, 1942: 867. Bon- net, 1957: 2819. Description. Female from Jurubatuba, Est. Sao Paulo. Carapace light orange, ce- phalic region orange. Chelicerae brown. Labium, endites dark orange. Sternum or- ange. Coxae light orange; legs orange, dis- tal articles darker. Dorsum of abdomen with a pair of anterior dusky patches and pairs of diagonal marks (Fig. 206); venter light, without marks. Posterior median eyes 0.8 diameter of anterior medians, laterals 0.8 diameter. Anterior median eyes 0.8 di- ameter apart, 1 diameter from laterals. Posterior median eyes 0.3 diameter apart. Height of clypeus equals 0.6 diameter of anterior median eye. Total length 5.5 mm. Carapace 2.4 mm long, 1.8 wide, 1.1 wide behind lateral eyes. First femur 2.1 mm, patella and tibia 2.6, metatarsus 1.6, tarsus 0.5. Second patella and tibia 2.3 mm, third 1.3, fourth 1.8. Male from Pinhal, Santa Catarina, Bra- zil. Color as in female. Posterior median eyes 0.8 diameter of anterior medians, an- terior laterals 0.8 diameter, posterior lat- erals 0.6. Anterior median eyes 0.8 di- ameter apart, 0.6 diameter from laterals. Posterior median eyes 0.3 diameter apart. Height of clypeus equals 0.6 diameter of anterior median eye. First coxa without hook. Total length 4.0 mm. Carapace 2.5 mm long, 1.8 wide, 0.9 wide behind lateral eyes. First femur 2.4 mm, patella and tibia 3.1, metatarsus 2.2, tarsus 0.9. Second pa- tella and tibia 2.7 mm, third 1.3, fourth 1.8. Variation. Total length of females 4.5 to 8.0 mm, males 4.0 to 4.3. It is uncertain whether the scape of the epigynum is torn off in all examined or is a minute bulge (Fig. 203). Illustrations were made from a female and male from Pinhal. The dorsal view is of a female from Jurubatuba. Diagnosis. The female can be separated from others by the two oval sclerotized plates of the epigynum (Fig. 203). The male can be distinguished from M. satur- nino by having a straight conductor (at 3 hr in Fig. 207) and a longer median apoph- ysis (at 4 hr and at 5 hr in Fig. 207). Natural History. Most examined spec- Metazygia • Levi 113 Figures 197-202. Metazygia saturnino n. sp. 197-200, female. 197-199, epigynum. 197, ventral. 198, posterior. 199, lateral. 200, dorsal. 201, 202, left male palpus. 201, mesal. 202, apical. Figures 203-207. M. crabroniphila Strand. 203-206, female. 203-205, epigynum. 203, ventral. 204, posterior. 205, lateral. 206, dorsal. 207, male palpus. Figures 208-21 2. M. matanzas n. sp., female. 208-21 0, epigynum. 208, ventral. 209, posterior. 21 0, lateral. 21 1 , dorsal. 21 2, abdomen, ventral. Figures 213-215. M. corumba n. sp., male. 213, palpus. 214, dorsal. 215, abdomen, ventral. Scale lines. 1.0 mm, genitalia 0.1 mm. imens are in poor condition and may all have come from wasp nests. Specimens Examined. BRAZIL Sdo Paulo: Juru- batuba, 6 July 1941, 19 (F. Lane, MZSP 9630); Jardim Botanico, Agua Funda, 7 July 1962, 19, 13 (A. F. Archer, AMNH). Parana: Curitiba, 2 Feb. 1988, 3$ (D. H. Habeck, FSCA). Santa Catarina: Pinhal, Dec. 1947, 79, 16; Jan. 1948, 19; Dec. 1948, 29, 16; Jan. 1949, 39 (A. Mailer, AMNH). Metazygia matanzas new species Figures 208-212; Map 2 J Holotype. Female holotype from Pan de Palenque, Matanzas, Cuba, 11 Aug. 1955 (A. F. Archer), in AMNH. The specific name is a noun in apposition after the type locality. Description. Female holotype. Cara- pace dusky orange, darkest in eye region. 114 Bulletin Museum of Comparative Zoology, Vol. 154, No. 2 Chelicerae, labium, endites dusky orange. Sternum orange with sides black. Coxae orange, legs dusky orange. Dorsum of ab- domen with black bands on sides, bordered by white bands, and an indistinct dusky folium containing anterior black marks (Fig. 211). Venter with white patch on black (Fig. 212). Posterior median eyes 0.8 diameter of anterior medians, laterals 0.8 diameter. Anterior median eyes 0.8 di- ameter apart, 0.6 diameter from laterals. Posterior median eyes 0.2 diameter apart, 1.7 diameters from laterals. Height of clypeus equals 0.4 diameter of anterior median eye. Total length 3.9 mm. Cara- pace 1.6 mm long, 1.2 wide, 0.7 wide be- hind lateral eyes. First femur 1.4 mm, pa- tella and tibia 1.8, metatarsus 1.1, tarsus 0.7. Second patella and tibia 1.5 mm, third 0.9, fourth 1.2. Diagnosis. This species may belong to Araneus because it has an annulate, flat, round scape (Fig. 208). It is placed here because of the shape of the abdomen and the position of the posterior median eyes (Fig. 211). Metazygia corumba new species Figures 213-215; Map 3B Holotype. Male holotype and male paratype (with one female of M. voluptifica) from Corumba, Mato Grosso do Sul, Brazil, 28-29 May 1960 (B. Malkin), in AMNH. The specific name is a noun in appo- sition after the type locality. Description. Male holotype. Cephalo- thorax orange. Dorsum of abdomen white with sides dusky. Venter with white patch behind genital groove (Fig. 215). Posterior median eyes 0.8 diameter of anterior me- dians, anterior laterals 0.8 diameter, pos- terior laterals 0.7. Anterior median eyes 0.6 diameter apart, 0.6 diameter from lat- erals. Posterior median eyes 0.3 diameter apart, 1.7 diameters from laterals. Height of clypeus equals 0.5 diameter of anterior median eye. Endite without tooth. First coxa with small hook. Total length 5.2 mm. Carapace 2.5 mm long, 2.0 wide, 1.0 wide behind lateral eyes. First femur 2.6 mm, patella and tibia 3.2, metatarsus 2.4, tarsus 0.9. Second patella and tibia 2.7 mm, third 1.5, fourth 2.0. Note. The two males from Corumba were collected with a female of M. mun- dula. Variation. Total length of males 3.7 to 5.2 mm. Illustrations were made from the holotype. Diagnosis. This species differs from M. crabroniphila by having the two edges of the embolus straight and parallel (Fig. 213). Specimen Examined. BOLIVIA Smita Cruz: Ma- taral, 14 Dec. 1984, 16 (L. Pena, AMNH). Metazygia sendero new species Figures 216-221 ; Map SB Holotype. Female holotype from Sendero Campa- mento, Laguna Grande, PUCE Field Station, Re- serva Faunistica Cuyabeno, 00°00'N, 76°10-irW, 31 July to 5 Aug. 1988, Sucumbios Prov., Ecuador, 31 July to 5 Aug. 1988 (W. Maddison), in MECN. The specific name is a noun in apposition after the type locality; sendero is the Spanish word for path. Description. Female holotype. Cara- pace orange and black. Chelicerae brown- ish black. Labium, endites brown. Sternum brown. Coxae dusky orange, legs dark or- ange-brown. Dorsum of abdomen with a series of pairs of black brackets (Fig. 219). Venter with a black square (Fig. 220). Pos- terior median eyes 0.7 diameter of anterior medians, laterals 0.7 diameter. Anterior median eyes 0.5 diameter apart, 2.2 di- ameters from laterals. Posterior median eyes 0.2 diameter apart. Height of clypeus equals 0.5 diameter of anterior median eye. Total length 11.0 mm. Carapace 4.6 mm long, 3.2 wide, 2.5 behind lateral eyes. First femur 3.5 mm, patella and tibia 4.5, meta- tarsus 3.1, tarsus 1.4. Second patella and tibia 3.9 mm, third 2.2, fourth 3.4. Male. Coloration as in female, but ven- ter of abdomen with scattered white pig- ment spots. Posterior median eyes 0.7 di- ameter of anterior medians, laterals 0.7 diameter. Anterior median eyes 0.4 di- ameter apart, 1.6 diameters from laterals. Posterior median eyes 0.4 diameter apart, 3.5 diameters from laterals. Height of clypeus equals 0.7 diameter of anterior Metazygia • Levi 115 Figures 216-221. Metazygia sendero n. sp. 216-220, female. 216-218, epigynum. 216, ventral. 217, posterior. 218, lateral. 219, dorsal. 220, abdomen, ventral. 221, left male palpus. Figures 222-225. M. uma n. sp. 222-224, female. 222, 223, epigynum. 222, ventral. 223, posterior. 224, right, dorsal; left, abdomen, ventral. 225, male palpus. Figures 226-230. M. laticeps (O. P.-Cambridge). 226-229, female. 226-228, epigynum. 226, ventral. 227, posterior. 228, dorsal. 229, abdomen, ventral. 230, male palpus. Figures 231-238. M. mundulella Strand. 231-236, female. 231-235, epigynum. 231 , ventral. 232, posterior. 233, lateral. 234, broken ventral. 235, broken posterior. 236, dorsal. 237, 238, palpus. 237, mesal. 238, ventral. Scale lines. 1.0 mm, genitalia 0.1 mm. 116 Bulletin Museum of Comparative Zoology, Vol. 154, No. 2 median eye. Total length 6.5 mm. Cara- pace 3.2 mm long, 2.3 wide, 1.5 behind lateral eyes. First femur 2.5 mm, patella and tibia 3.1, metatarsus 2.3, tarsus 1.2. Second patella and tibia 2.7 mm, third 1.5, fourth 2.1. Note. Males and females were matched because of similar coloration, a similar wide cephalic region and swelling behind the eyes, and similarly shaped abdomen. The only other species with these characters is M. laticeps. Male and female come from the opposite slopes of the Andes. Variation. Total length of females 8.0 to 11.0 mm. Illustrations were made from the female holotype and a male from Tin- alandia, Ecuador. Diagnosis. The abdomen is elongate, oval, slightly overhanging spinnerets (Fig. 220). This species differs from M. laticeps by having the epigynum with a narrower triangle in ventral view (Fig. 216); in pos- terior view, the median plate is as wide as long (Fig. 217) while that of M. laticeps is narrow (Fig. 227). The median apoph- ysis of the palpus (at 4 hr in Fig. 221) has two tips. Distribution. Ecuador, Peru (Map 3B). Specimens Examined. ECUADOR Pichincha: Tinalandia, 12 km E Santo Domingo de los Colorados, 750 m, beating vegetation, 11-17 May 1986, 1 imm., IS, 26 (G. B. Edwards, FSCA). PERU Ucayali: Co- lonia Calleria, Rio Calleria, 15 km from Ucayali, 10- 30 Sept. 1961, 12 (B. Malkin, AMNH). Metazygia uma new species Figures 222-225; Map 3B Holotype. Female holotype from Puesto de Vigilan- cia Pakitza, Zona Reservada de Manu, Depto. Ma- dre de Dios, 11°58'S, 7ri8'W, Peru, inundated for- est, 30 Sept. 1987 (D. Silva D., J. Coddington), in MUSM. The specific name is an arbitrary combi- nation of letters. Description. Female holotype. Cara- pace orange. Chelicerae, labium, endites orange. Sternum orange. Coxae orange; legs orange but distal tips of tibiae, meta- tarsi and tarsi black. Dorsum of abdomen with longitudinal gray lines (Fig. 224); venter with a black rectangle (Fig. 225). Posterior median eyes 0.7 diameter of an- terior medians, laterals 0.7 diameter. An- terior median eyes 0.5 diameter apart, 3 diameters from laterals. Posterior median eyes 0.4 diameter apart. Height of clypeus equals 0.6 diameter of anterior median eye. Total length 8.2 mm. Carapace 3.8 mm long, 2.5 wide, 2.3 behind lateral eyes. First femur 2.6 mm, patella and tibia 3.4, meta- tarsus 2.3, tarsus 1.1. Second patella and tibia 2.9 mm, third 1.7, fourth 2.7. Male. Coloration as in female, but ven- ter of abdomen all black. Posterior median eyes 0.8 diameter of anterior medians, lat- erals 0.5. Anterior median eyes 0.8 diam- eter apart. Posterior median eyes 0.3 di- ameter apart. Coxal hook very small. Sec- ond tibia thicker than first with two long macrosetae in a line. Height of clypeus equals 1 diameter of anterior median eye. Total length 3.9 mm. Carapace 2.0 mm long, 1.4 diameters wide, 0.9 wide behind lateral eyes. First femur 1.4 mm, patella and tibia 1.8, metatarsus 1.2, tarsus 0.7. Second patella and tibia 1.5 mm, third 0.8, fourth 1.1. Note. Male was matched with the fe- male because of similar coloration and the relatively wide cephalic region of the car- apace. Diagnosis. This species is distinct in that the eye region of the carapace is almost as wide as the thoracic region (Fig. 224). The male differs from M. laticeps by the curved embolus lamella with a short curved em- bolus "above" it (Fig. 225). This strange- looking species is apparently a Metazygia and closest to M. laticeps, which has also the carapace relatively wide anteriorly (Fig. 228). Natural History. The male was collect- ed in forest interior. Distribution. Amazon region, Peru to Brazil (Map 3B). Specimens Examined. PERU Madre de Dios: Zona Reservada Tambopata, 290 m, 8 July 1987, 19 (D. Silva D., MUSM). BRAZIL Amazonias: Reserva Flo- restal, 80 km from Manaus, 19 Feb. 1991, IS (H. Fowler, E. Venticinque, R. S. Vieira, MCZ). Metazygia • Levi 117 Metazygia laticeps (O. P.-Cambridge), new combination Figures 226-230; Map 3F Epeira laticeps O. P.-Cambridge, 1889: 18, pi. 4, fig. 16, 2. Female holotype from Bugaba, Panama, in BMNH no. 1890.7.1.5020, examined. Keyserling, 1892: 175, pi. 8, fig. 129, 2. Aranea laticeps: — F. P.-Cambridge, 1904: 516, pi. 49, fig. 21, 2. Roewer, 1942: 845. Araneus laticeps: — Bonnet, 1955: 527. Note. The type specimen belonged to Keyserling. The vial containing the type also contains Keyserling's original, toothed, blue-bordered label, which is still faintly legible when dried. The first line reads Guatemala and not Bugaba. Description. Female from Pipeline Road, Panama. Carapace dark orange with median double line, black on each side of thoracic region. Chelicerae, labium, en- dites orange. Sternum dusky orange-brown. Coxae light orange, legs dusky orange- brown. Dorsum of abdomen with pairs of brackets (Fig. 228); venter with median dark gray patch (Fig. 229). Posterior me- dian eyes 0.8 diameter of anterior medi- ans, laterals 0.6 diameter. Anterior median eyes 0.8 diameter apart, 2 diameters from laterals. Posterior median eyes 0.6 diam- eter apart. Height of clypeus equals 0.6 diameter of anterior median eye. Total length 10.0 mm. Carapace 3.7 mm long, 3.1 wide, 2.2 behind lateral eyes. First fe- mur 3.4 mm, patella and tibia 4.4, meta- tarsus 2.7, tarsus 1.3. Second patella and tibia 3.7 mm, third 2.2, fourth 3.2. Male from Barro Colorado Island, Pan- ama. Color as in female. Posterior median eyes 0.7 diameter of anterior medians, lat- erals 0.7 diameter. Anterior median eyes 0.8 diameter apart, 2.5 diameters from lat- erals. Posterior median eyes 0.3 diameter apart. Height of clypeus equals 0.4 di- ameter of anterior median eye. Abdomen as in female, but smaller. Total length 6.3 mm. Carapace 3.4 mm long, 2.5 wide, 1.7 behind lateral eyes. First femur 2.9 mm, patella and tibia 3.8, metatarsus 2.7, tarsus 1.2. Second patella and tibia 3.1 mm, third 1.7, fourth 2.5. Note. Males and females were matched because both have a dusky patch on the venter of the abdomen (Fig. 229) and be- cause both have the same wide distribu- tion. Males are much less common in col- lections than females. Variation. Total length of females 8.0 to 11.0 mm. Females and male from Mato Grosso, Brazil, differed from those of other regions: the female epigynum is more rounded posteriorly and in posterior view there is a round depression ventrally, but the lateral and median sclerites are of the same width as the one illustrated (Fig. 227). The male from Mato Grosso has a more elongate median apophysis, and the two parallel prongs of the embolus are of sim- ilar width. Illustrations were made from a female from Pipeline Road, Panama, and a male from Barro Colorado Island, Panama. Diagnosis. The abdomen is elongate oval (Fig. 228). Metazygia laticeps female has a pointed, wide, triangular epigynum (Fig. 226) with the posterior median plate slightly narrower than the lateral plates on each side (Fig. 227). The male has a large, semicircular median apophysis (at 5 hr in Fig. 230). Natural History. Females are collected in tropical forest by unrolling rolled-up leaves, their retreat, at a height of about 150 cm. Specimens from Mato Grosso all came from gallery forest; Guyanas from forest savanna and swamp forest; and near Iquitos, Peru, from rain forest. Distribution. Panama to Rio de Janeiro and northern Bolivia (Map 3F). Specimens Examined. PANAMA Colon: Fort Sherman, 12 (MCZ). Panama: nr. Gamboa, edge of Canal, 1948, 12 (W. Eberhard, MCZ); Soberania Natl. Park, Pipeline Road, 8 km NW Gamboa, 12 (MCZ); Barro Colorado Isl,, Lago Gatun, 2S, 22, MCZ. TRINIDAD Port of Spain, 13 (MCZ). GUYANA Canje Ikuruwa River 05°70'N, 57°50'W, 12 (AMNH). FRENCH GUIANA nr. Placer Tresor, Roura Mtns., 12 (MCZ); nr, Sautero, Matouri, 12 (MCZ). COLOM- BIA Satander: Rio Suarez, 800-1,000 m, 12 (AMNH). PERU hereto: Iquitos, 12 (AMNH); Explorama Lodge, 80 km NE Iquitos, 12 (FSCA). San Martin: 32 km SE Moyabamba, \$ (AMNH). Junin: Amable Maria, 118 Bulletin Museum of Comparative Zoology, Vol. 154, No. 2 12 (PAN). Madre de Dios: Zona Reservada Tambo- pata, 290 m, 19 (MUSM). BRAZIL Pard: Canind6, Rio Gurupi, 12 (AMNH). Amazonas: Tefe, 12 (S. Par- rish, MCZ); Rio Negro, Umarituba, 12 (NRMS); Maturaca, 12 (MCP). Espirito Santo: Linhares, Par- que Souterrana, 12 (MZSP). Rio de Janeiro: Rio de Janeiro, Jardim Botanico, 12 (MCZ); Parque da Ci- dade, 12 (MCZ). Mato Grosso: 260 km N Xavantina, 400 m, 12''49'S, 51°46'W, 32, U (MCZ). BOLIVIA Beni: Chacobo Indian Village, Rio Benicito, 12 (AMNH); Est. Biologica Beni, 12 (USNM). Metazygia mundulella Strand Figures 231-238; Map 3F Aranea (Metazygia) mundulella Strand, 1915: 114. Ten female, two male, and two immature syntypes from mud-dauber wasp nest, Joinville, Santa Ca- tarina State, Brazil, in SMF no. 4010, examined. Aranea mundulella: — Roewer, 1942: 848. Larinia mundula: — Bonnet, 1957: 2350. Note. Strand (1915: 114) wrote, "Since there is no certainty from the hterature as to which species the specimens belong, I make some descriptive remarks and pro- pose, if necessary, the name mundulella." Description. Female syntype. Carapace orange, cephalic region darker. Chelicer- ae, labium, endites dark orange. Sternum orange. Coxae, legs orange. Dorsum of ab- domen light with a pair of dark anterior patches and four pairs of dark lines (Fig. 236). Venter light. Posterior median eyes same diameter as anterior medians, lat- erals 0.8 diameter. Anterior median eyes 0.5 diameter apart, 0.9 diameter from lat- erals. Posterior median eyes 0.3 diameter apart, 2.1 diameters from laterals. Height of clypeus equals 0.5 diameter of anterior median eye. Total length 5.9 mm. Cara- pace 2.8 mm long, 2.1 wide, 1.4 behind lateral eyes. First femur 2.3 mm, patella and tibia 2.6, metatarsus 1.8, tarsus 0.8. Second patella and tibia 2.3 mm, third 1.4, fourth 2.0. Male syntype. Color, including marks on abdomen, as in female. Posterior me- dian eyes 0.9 diameter of anterior medi- ans, laterals 0.5 diameter, posterior laterals 0.5. Anterior median eyes 0.7 diameter apart, 0.7 diameter from laterals. Posterior median eyes 0.2 diameter apart, 1.2 di- ameters from laterals. Height of clypeus equals 0.7 diameter of anterior median eye. Second tibia as thick as first, with macro- setae. Total length 4.5 mm. Carapace 2.4 mm long, 2.0 wide, 0.9 wide behind lateral eyes. First femur 2.1 mm, patella and tibia 2.7, metatarsus 1.9, tarsus 0.7. Second pa- tella and tibia 2.3 mm, third 1.3, fourth 1.8. Note. The syntypes have lost all white pigment and also the silver pigment of the eyes, perhaps from having been in a buf- fered formaldehyde solution (Levi, 1989). All except for one specimen (Figs. 231, 232) have the epigynum broken. Each side is broken off (Figs. 234, 235), apparently the result of mating. Males and females were collected together. Diagnosis. The epigynum, unlike that of M. genialis (Fig. 239), has a concave margin on each side and a flat scape (Figs. 231, 233). In posterior view, it has a ventral pocket on each side (at 11 hr and 2 hr in Fig. 232). The male palpus, like that of M. genialis (Figs. 243, 244), has a lobe on the tegulum (at 12 hr in Figs. 237, 238) but differs in the shape of the embolus lamella and median apophysis (center and at 5 hr in Fig. 237). Natural History. All specimens came from a mud-dauber wasp nest. Metazygia genialis (Keyserling) Figures 239-246; Map 3F Epeira genialis Keyserling, 1892; 156, pi. 8, fig. 114, 2. Two female syntypes from Rio Grande do Sul, Brazil, one has the epigynum broken, the other covered by secretions, in BMNH, examined. Epeira mundula: — Keyserling, 1892: 179, pi. 9, fig. 132, S (not female lectotype). Aranea genialis: — Roewer, 1942: 843. Araneus genialis: — Bonnet, 1955: 507. Metazygia genialis: — Levi, 1991a: 179. Note. The male paralectotype of E. mundula belongs with the female of M. genialis. Description. Female from Santa Vitoria do Palmar, Rio Grande do Sul. Cephalic region of carapace dark brown, thoracic region yellowish. Chelicerae dark brown. Labium, endites dark brown. Sternum light brown. Coxae yellowish; legs with proxi- METAZYCIA'Levi 119 239 Figures 239-246. Metazygia genialis (Keyserling). 239-242, female. 239-241, epigynum. 239, ventral. 240, posterior. 241, lateral. 242, dorsal. 243-246, left male palpus. 243, mesal. 244, ventral. 245, 246, palpus pulled apart. Figures 247-249. M. amalla n. sp., female. 247, 248, epigynum. 247, ventral. 248, posterior. 249, dorsal. Figures 250-252. M. ikuruwa n. sp., male. 250, 251, palpus. 250, mesal. 251, ventral. 252, dorsal. Abbreviations. C, conductor; H, hematodocha; E, embolus; L. emtwius lamella; M, median apophysis; P, paracymbium; R, radix; T, teguium; Y, cymbium. Scale lines. 1.0 mm, genitalia 0.1 mm. mal articles yellowish, distal brown. Dor- sum of abdomen with dusky outline of fo- lium (Fig. 242); venter gray, without marks. Posterior median eyes 0.6 diameter of anterior medians, laterals 0.6 diameter. Anterior median eyes 0.7 diameter apart, 1 diameter from laterals. Posterior median eyes 0.3 diameter apart, 3 diameters from laterals. Height of clypeus equals 0.3 di- ameter of anterior median eye. Total length 6.3 mm. Carapace 3.0 mm long, 2.3 wide, 1.5 wide behind lateral eyes. First femur 2.2 mm, patella and tibia 2.7, metatarsus 1.8, tarsus 0.8. Second patella and tibia 2.5 mm, third 1.7, fourth 2.1. Male from Santa Vitoria do Palmar, Rio Grande do Sul. Color as in female, but cephalic region yellowish. Posterior me- 120 Bulletin Museum of Comparative Zoology, Vol. 154, No. 2 dian eyes 0.5 diameter of anterior medi- ans, laterals 0.5 diameter. Anterior median eyes 0.4 diameter apart, 0.5 diameter from laterals. Posterior medians 0.2 their di- ameter apart, 2.2 diameters from laterals. Height of clypeus equals 0.4 diameter of anterior median eye. Third femur with three short macrosetae; each only three times as long as wide. Total length 4.8 mm. Carapace 2.6 mm long, 2.2 wide, 1.1 be- hind lateral eyes. First femur 2.2 mm, pa- tella and tibia 2.7, metatarsus 2.2, tarsus 0.7. Second patella and tibia 2.4 mm, third 1.4, fourth 1.7. Note. Males and females were collected together. Variation. Total length of females 6.3 to 7.7 mm, males 4.8 to 5.8. Illustrations were made from Santa Vitoria, Rio Grande do Sul, Brazil. Diagnosis. The epigynum has a knob (Figs. 239, 240) rather than a flat scape as in M. mundulella (Figs. 231-235). The male, like M. mundulella (Figs. 237, 238), has a tegulum (T in Fig. 245) with a lobe (at 12 hr in Figs. 243, 245) and a distinctive shape of the embolus lamella (L in Fig. 245) and median apophysis (at 5 hr in Fig. 243). Distribution. Bahia to Rio Grande do Sul States, Brazil (Map 3F). Specimens Examined. BRAZIL Bahia: Mucuri, Fa- zenda Farol, 11 Apr. 1979, 1i - - -- + +---- - - -- abd.shape [ + ] - [ + ] [ + ] [ + ] [ + ] [ + ] - - [ + ] [ + ] [ + ] abd.pttrn. + - - - - --[ + ][ + ]- - - - - abd. spins. + — *— - — _____ _ _ __ epig.lobe + + ++ + -[ + ]+ + - - - -- epig.scp. - - — — — ~[ + ]~~+ + + + + scape pnt. — — — — — ~[ + ]~~~" + + ~~ scape pck. ________ — —_ — + + 3 dwarf - - ?- +*+- + -- -*- -*- 3abdom - - ?- + +-[ + ]-- " - -- patella 1 1* ? 1 1 1 1 1 1 2 2* 2 2* 2 M tooth -- ?--- + --+ + + + + Oagella __?-_- + --- + + -- PM + + ? + +*?-? + ----- C on edge - - ?- _*__-_+ + + + + DH __?____-_+ + + + + A + + ? + + --- + + + + + + A cplx. - + ?_-_---+ + + +- Genera: ACA, Acacesia; ACT, Actinosoma; ALP, Alpaida; AM, Amazonepeira; ARA, Araneus; HIN, Hingstepeira, MAD, Madrepeira, MCA, Micrathena; MEP, Micrepeira; NEO, Neoscona; PRO, Pronous; SPI, Spinepeira; SPL, Spilasma, TAT, Tatepeira. Revised genera are set in italic. Abbreviations: PME, posterior median eyes modified; head, cephalic region modified; IV > I, leg 4 longer than leg 1; abd.shape, abdomen shape modified; abd.pttrn., abdomen pattern modified; abd. spins., median posterior spines or tubercles present; epig.lobe, epigynum a lobe; epig.scp., epigynum with scape; scape pnt., scape tip pointed; scape pck., scape tip with distal pocket; 6 dwarf, male dwarfed and may lack endite tooth, coxal hook; S abdom, male abdomen modified with ventral scutum; patella, palpal patella with 1 or 2 setae; M tooth, median apophysis with tooth; flagella, median apophysis with flagella; PM, paramedian apophysis; C on edge, conductor on edge of tegulum; DH, distal hematodocha present; A, terminal apophysis present; A cplx., terminal apophysis branched. +, present; [ + ], an autapomorphy of genus; -, absent; *, variable in some species. veloped distal hematodocha (Fig. 224); the female has an epigynum with an annulat- ed scape (Figs. 217-219) and a pair of humps on the abdomen (Fig. 220). These characters all are synapomorphies with Ar- aneus and Aculepeira. Other genera here, Pronous, Micrepeira, and Spilasma, have some intermediate characters (Table 1). Readers of manuscripts often suggest that more information on apomorphies should be provided for construction of cladograms. But while some help can be given, the judgment about what is a good apomorphy is subject to reconsideration and relies on knowledge of other genera, some not yet revised. A study of relation- ships and useful apomorphies will be in order at the end of the revisions. Some examples of recent changes follow: Earlier it was thought that the presence of one or two macrosetae on the male palpal patella was useless in phylogeny and was not used. Now, however, despite many exceptions, it appears that the Araneus group of gen- era generally has two macrosetae, while those close to Alpaida have one (Table 1). In another example, the usual armature of araneid males, such as an endite tooth, a hook on the first coxa, or the modified sec- ond tibia seems to be lost in males that are very small compared to the female [with many exceptions, e.g., Acanthepeira Marx (Levi, 1976)]. At present, such armature is of little use in phylogeny. Another case of changed value concerns the pointed scape of the epigynum and the paired flagella of AcTiNOSOMA, Spilasma, Micrepeira, Pronous • Levi 155 the median apophysis in the male palpus, formerly considered a synapomorphy for about five genera close to Araneus. But now these features are also noted in Mi- crepeira (Fig. 162, M in Fig. 179), which otherwise is close to Alpaida O. P. -Cam- bridge (Levi, 1988). At this stage the best information that can be provided is limited to the autapomorphies of individual re- vised genera. MATERIALS AND ACKNOWLEDGMENTS Specimens of the following collections were used, and I thank their curators for making the spiders available: AD A. Dean, College Station, Texas, United States AMNH American Museum of Natural History, New York, United States; N. Platnick, L. Sorkin BMNH Natural History Museum, Lon- don, England; P. Hillyard, F. Wanless CAS California Academy of Sciences, San Francisco, California, United States; W. J. Pulawski, D. Ubick, C. Griswold DU D. Ubick, San Francisco, Califor- nia, United States FMLT Fundacion Miguel Lillo, Tucu- man, Argentina; J. A. Corronca FSCA Florida State Collection of Ar- thropods, Gainesville, Florida, United States; G. B. Edwards HECO Hope Entomology Collections, Oxford University, Oxford, En- gland; I. Lansbury, M. Atkinson INPA Instituto Nacional de Pesquisas da Amazonia, Manaus, Est. Amazo- nas, Brazil; C. Magalhaes IRSNB Institut Royal des Sciences Natu- relles de Belgique, Brussels, Bel- gium; L. Baert JAK J. A. Kochalka, Ciudad Universi- taria, Paraguay MACN Museo Argentino de Ciencias Na- turales, Buenos Aires, Argentina; E. A. Maury MCN Museu de Ciencias Naturais, Fun- da^ao Zoobotanica do Rio Grande do Sul, Porto Alegre, Rio Grande do Sul, Brazil; E. H. Buckup MCP Museu de Ciencias, Pontificia Universidade Catolica do Rio Grande do Sul, Porto Alegre, Rio Grande do Sul, Brazil; A. A. Lise MCZ Museum of Comparative Zoolo- gy, Cambridge, Massachusetts, United States MECN Museo Ecuatoriano de Ciencias Naturales, Quito, Ecuador; L. Aviles MEG M. E. Galiano; Buenos Aires, Ar- gentina MHNG Museum d'Histoire Naturelle, Ge- neve, Switzerland; V. Mahnert MIUP Museo de Invertebrados, Univer- sidad de Panama, Panama; D. Quintero A. MNHN Museum National d'Histoire Na- turelle, Paris, France; J. Heur- tault, C. Bollard MUSM Museo de Historia Natural, Univ- ersidad Nacional Mayor de San Marcos, Lima, Peru; D. Silva D. MZCR Museo Zoologico de Universidad de Costa Rica, San Jose, Costa Rica; C. E. Valerio MZSP Museu de Zoologia, Universidade de Sao Paulo, Sao Paulo, SP, Bra- zil; P. Vanzolini, J. L. Leme MZUF Museum Zoologico de La "Spe- cola," Universita di Firenze, Flor- ence, Italy, L. Bartolozzi, S. Mas- cherini NMB Naturhistorisches Museum, Basel, Switzerland; E. Sutter, A. Hanggi NRMS Naturhistoriska Riksmuseet, Stock- holm, Sweden; T. Kronestedt PAN Polska Akademia Nauk, Warsza- wa, Poland; J. Proszynski, A. Slo- jewska, E. Kierych REL R. E. Leech, Edmonton, Alberta, Canada RLCB R. L. C. Baptista, Rio de Janeiro, Brazil SMF Forschungsinstitut Senckenberg, Frankfurt am Main, Germany; M. Grasshoff 156 Bulletin Museum of Comparative Zoology, Vol. 154, No. 3 UCVC Universidad Central de Venezue- la, Caracas, Venzuela; J. Racenis USNM National Museum of Natural His- tory, Smithsonian Institution, Washington, D.C., United States; J. Coddington, S. F. Larcher ZMK Zoologisk Museum, Copenhagen, Denmark; H. Enghoff ZSM Zoologische Staatssammlung, Mu- nich, Germany In addition, I am grateful to J. A. Cor- ronca, who sent specimens and photo- graphs of Actinosoma habitat, W. Eber- hard for habitat and behavior information, I. Lansbury for information on O. P. -Cam- bridge types, T. Kronestedt for type in- formation on Actinosoma species, P. Van- zolini and S. F. Larcher for locality infor- mation, H. G. Fowler and his students for making their large Amazonian collections available, and R. L. C. Baptista, J. A. Cod- dington, W. Eberhard, and H. Hofer for photographs of webs. A. Johnston, L. Lei- bensperger, L. R. Levi, and W. Piel read the paper and made many improvements in the writing, and two anonymous readers also suggested changes. Especially helpful were H. Hofer, E. H. Buckup, and M. A. L. Marques, who read a draft of the manu- script meticulously and made me aware of many misprints. National Institutes of Health and the National Science Foun- dation grants supported the start of this research. Publications costs were covered, in part, by the Wetmore-Colles Fund. METHODS Most methods used in these studies are described in Levi (1993a). Internal genitalia of Pronous specimens were examined by removing the epigyn- um using minutennadeln (used otherwise for mounting small insects) mounted on a wooden stick. The epigynum was placed in Hoyer's medium between two cover- slips so that either side could be examined. An outline drawing was made with a dis- secting microscope, at the same magnifi- cation as the illustration of the epigynum. Details were added later. Finally, the epi- gynum was removed from the mount and placed in a 2-by-4-mm glass vial stoppered with cotton and kept with the female spec- imen. Palpi were spread out by pulling apart rather than by expanding, because pulling apart keeps all sclerites in similar positions and makes it easier to match and illustrate the labeled sclerites with those of the con- tracted palpus. This is of special impor- tance in genera related to Araneus that have well-developed distal hematodocha. The sizes of the eyes are expressed as diameters of the anterior median eyes; dis- tances between eyes of the anterior row are expressed as diameters of the anterior median eyes (in profile), and distances be- tween eyes of the posterior row are given as diameters of the posterior median eyes (in profile). The height of the clypeus, the distance between the anterior median eyes and the edge of the carapace, is expressed as diameter of the anterior median eyes (Levi, 1993a, fig. 28f). These measure- ments are approximate. To examine for the presence of a ta- petum, the eyes of Pronous specimens were first checked with reflected light. When no tapetum was found in the posterior me- dian eyes, the eye region of the carapace (of a female P. quintana) was cut off, cleared with methyl benzoate, and ex- amined more carefully. Actinosoma Holmberg Actinosoma Holmberg, 1883: 239. Type species Ac- tinosoma pentacanthum (Walckenaer) designated by Holmberg and by monotypy. The gender of the name is neuter (Bonnet, 1955: 156). Diagnosis. Actinosoma is readily sep- arated from all other Neotropical araneid genera by the five large sclerotized, orange spines on the abdomen of both female and male (Figs. 4-6, 10), an autapomorphic character. The genitalia are very similar to those of Alpaida, but the terminal apophysis of the male palpus is a simple structure (A in Fig. 9); in Alpaida it is subdivided. AcTiNosoMA, Spilasma, MiCREPEiRA, Pronous • Levi 157 pentacanthum Map 1 . Distribution of Actinosoma pentacanthum. Actinosoma differs from Micrathena, with which it is commonly confused, by having the first pair of legs longer than the fourth, by lacking stridulating structures on the book-lungs (present in most Mi- crathena), and by lacking a square to rect- angular abdomen of the male. The abdo- men of male Actinosoma has spines (Fig. 10) just as in the female; male Micrathena lack them. Description. Female. Abundant black pigment between median eyes (Fig. 7). Width of cephalic region behind eyes slightly more than half thoracic width. Epigynum with projecting, rounded shelf, its tip a rounded lobe (Figs. 1-3) as in Alpaida. Male. Width of eye region behind eyes about half the width of the thoracic region (Fig. 11). Palpus with conductor in middle of tegulum (Fig. 8, C in Fig. 9); tegulum without sclerites on distal, ventral side (right, in left palpus at 3 hr in Fig. 8, T in Fig. 9). Paramedian apophysis (PM in Fig. 9) is a separate, lightly sclerotized, L-shaped sclerite attached by soft stalk to conductor. Median apophysis cup-shaped (M in Fig. 9; Levi, 1988, fig. 10), as it often is in Alpaida. Embolus supported by para- median apophysis; its tip on conductor (E in Fig. 9). Terminal apophysis a narrow lobe above embolus (A in Fig. 9). Relationship. The lobe of the epigyn- um of the female resembles that of Al- paida (see Levi, 1988) and Micrathena (see Levi, 1985); it is a synapomorphy. The palpus has the conductor in the center of the tegulum (C in Fig. 9) as in Alpaida and Micrathena, and there is a parame- dian apophysis (PM in Fig. 9) as in these two genera [it may be missing in some Micrathena (Levi, 1985)], another syna- pomorphy of various genera. Unlike Ara- neus, Neoscona, Madrepeira, Tatepeira, and others (Figs. 213, 224), the median apophysis lacks sharp spines or flagella-like extensions. Natural History. Actinosoma penta- canthum lives on emergent aquatic veg- etation above the water surface in the mid- dle of ponds and puddles. Distribution. Only one species is wide- spread in South America (Map 1). Misplaced. Actinosoma heteracantha 158 Bulletin Museum of Comparative Zoology, Vol. 154, No. 3 Mello-Leitao, 1943, is Wagneriana het- eracantha (Mello-Leitao) (Levi, 1991c). Actinosoma riscoi Archer (1971: 158, figs. 3, 4, 9) is Rubrepeira rubronigra (Mel- lo-Leitao) (Levi, 1991a). Actinosoma pentacanthum (Walckenaer) Figures 1-12; Map 1 Plectana pentacantha Walckenaer, 1841: 170. Lo- cality of specimens not known but may have been from Cayenne (French Guiana); specimens lost. Acrosoma stelligeriim Thorell, 1860: 301; 1868: 26. Origin of specimen unknown and type specimens not in NRMS, lost. Specimens determined by Tho- rell from Corumba, Mato Grosso [Mato Grosso do Sul], Brazil, in NRMS, examined. Similarity of de- scription of stelligerum and pentacantha first no- ticed by Butler (1873). Name first synonymized by Holmberg (1883). Acrosoma pentacanthum: — Butler, 1873: 428. Holmberg, 1883: 237. Acrosoma pulcherrima Holmberg, 1876: 143 (p. 21 in a separate publication, not seen). Specimens from Puerto Obligada [Vuelta de Obligado, Prov. Buenos Aires, 33°35'S, 59°49'W (USDI, 1968)], Argentina, lost. First synonymized by Holmberg (1883). Actinosoma pentacanthum: — Holmberg, 1883: 239. Roewer, 1942: 778. Bonnet, 1955: 156. Cyrtarachne quinquespinosa Keyserling, 1892: 55, pi. 3, fig. 44, 9, S. Five immature and five female syntypes from Espirito Santo, Brazil, in USNM, examined. Goldi, 1892: 227. First synonymized by Simon (1895). Araneus pentacanthus: — Simon, 1895: 819, fig. 869, 2. Badcock, 1932: 27. Araneus pentacantha: — Zapfe, 1957: 29, fig. 3, 9, 5. Islote. Acrosoma Perty, 1833, with the type species A. swainsoni, is a subjective synonym of Micrathena (Levi, 1985). Cyr- tarachne Thorell (1868: 10) was a name to replace Cyrtogaster Keyserling (1864: 80) [the name preoccupied by Walckenaer (1833) for a hymenopteran]. The genus has the type Cyrtogaster grubei, by mono- typy, from Mauritius, Indian Ocean, for an araneid with the carapace high in tho- racic region and abdomen triangular, wid- er than long, pointed behind, and having four dorsal tubercles. According to Holmberg (1883: 227), the information of 1876 was published on page 143, 1876 (date from Holmberg); it is page 21 on a photocopy of the journal that was available. The syntypes of Cyrtarachne quin- quespinosa were unexpectedly found in unsorted collections of the USNM. The USNM specimens were recognized as Key- serling specimens because they were la- beled "Espirito Santo, Cyrtarachne 5-spinosa Keys.", on a label, 24 by 30 mm, having toothed perforations on all four sides and a blue frame, a short distance inside from the white teeth. This is a typical Key- serling label. A second plain old label in the vial read "Espirito Santo, Rio Minas". Description. Female from Puerto Napo, Ecuador. Carapace orange, eye region black. Chelicerae orange, distally black. Labium, endites black. Sternum, coxae or- ange; legs without rings, dusky black, fem- ora darkest. Dorsum of abdomen with or- ange and black, the black including paired white patches (Fig. 4); venter black with a pair of white lines (Fig. 6). Eyes sub- equal. Anterior median eyes their diam- eter apart, 2 diameters from laterals. Pos- terior median eyes 0.8 diameter apart, 2.5 diameters from laterals. Ocular quadran- gle slightly narrower behind than in front, slightly longer than wide. Height of clyp- eus equals 1.1 diameters of anterior me- dian eye. Abdomen with five prominent spines (Figs. 4-6). Total length 8.0 mm. Carapace 3.2 mm long, 2.7 wide, 1.5 be- hind lateral eyes. First femur 3.4 mm, pa- tella and tibia 3.9, metatarsus 2.7, tarsus 1.5. Second patella and tibia 3.3 mm, third 2.3. Fourth femur 3.4, patella and tibia 3.7, metatarsus 2.5, tarsus 1.1. Male from Puerto Napa, Ecuador. Col- oration as in female (Fig. 10). Posterior median eyes 0.8 diameter of anterior me- dians, laterals 0.8 diameter. Anterior me- dian eyes 0.7 diameter apart, 1.4 diameters from laterals. Posterior median eyes 0.6 diameter apart, 2 diameters from laterals. Ocular quadrangle narrower behind, slightly longer than wide. Height of clyp- eus equals 1.2 diameters of anterior me- dian eye. Endite with tooth facing a tu- bercle on the coxa and a tooth on the prox- imal end of the palpal femur. Palpal pa- tella with one macroseta. First coxa with hook. Second tibia thicker than first but ACTINOSOMA, Spilasma, MiCREPEiRA, Pronous • Levi 159 Figures 1-12. Actinosoma pentacanthum (V\Ja\ckenaer). 1-7, female. 1-3, epigynum. 1 , ventral. 2, posterior. 3, lateral. 4, dorsal. 5, lateral. 6, abdomen ventral. 7, eye region and chelicerae. 8-12, male. 8, left palpus, mesal view. 9, palpus pulled apart. 10, dorsal. 11, carapace. 12, eye region, chelicerae and right palpus. Figures 13-20. Spinepeira schlingeri n. sp., female. 13-15, epigynum. 13, ventral. 14, posterior. 15, lateral. 16, dorsal. 17, lateral. 18, carapace. 19, carapace and chelicera. 20, eye region and chelicerae. Abbreviations. A, terminal apophysis. C, conductor. E, embolus. M, median apophysis. PM, paramedian apophysis. R, radix. T, tegulum. Scale lines. 1.0 mm, genitalia 0.1 mm. 160 Bulletin Museum of Comparative Zoology, Vol. 154, No. 3 without macrosetae. Abdomen as in fe- male (Fig. 10). Total length 6.7 mm. Car- apace 3.1 mm long, 2.3 wide; 1.1 behind lateral eyes. First femur 3.4 mm, patella and tibia 3.8, metatarsus 2.4, tarsus 1.1. Second patella and tibia 3.1 mm, third 2.0. Fourth femur 3.3 mm, patella and tibia 3.4, metatarsus 2.1, tarsus 1.0. Note. Males and females were collect- ed together. Variation. Total length of females 7.0 to 9.7 mm, males 6.0 to 7.0 mm. The bases of the five spines vary in width in different specimens. Specimens from the Napo Riv- er, Ecuador, were used for the illustrations, except for the dorsal view of the female (Fig. 4), which was made from a specimen from Mato Grosso, Brazil (NRMS). Re- cently collected specimens had bright or- ange, black, and white coloration. Diagnosis. The abdomen with five sclerotized spines in both males and fe- males (Figs. 4, 5, 10), and the orange, white, and black coloration, separates the species from all other araneids. Natural History. Holmberg (1876) captured the spiders along the sides of a brook. Goldi (1892) found this red spider in earthen organ-pipe tubes of wasps, which were attached to walls and window frames in Santo Antonio de Padua, Est. Rio de Janeiro, Brazil. The Colombian specimens were taken from a Sceliphron wasp nest. All others were from ponds or streams. Specimens from the Napo River, Ecuador, were collected by sweeping be- low water and water surface in middle of a swamp. A female from Manaus, Brazil, came from Paspalum grass, which forms vast floating meadows in Amazonian riv- ers. Badcock (1932: 27) recorded the spe- cies as "common everywhere on surface of swamps" in Paraguayan Chaco. Bad- cock's locality is described by Carter (1928: 97), the collector, as an island surrounded by swamps with daily maximum temper- atures of 90° F (42° C), mean 85° F (30° C), with surface layer of water at 108° F (43° C). A specimen from Paraguay (MHNG) was collected in a swamp. Cor- ronca (personal correspondence) reported webs in flooded grasslands, in tropical Ban- ado areas. When the web is disturbed, the spider dives or walks on the water surface. This species is probably just as common in rain forest ponds as in wet areas of dry regions, but collectors are more likely to be attracted to marshes and ponds in dry areas. Galiano reported in collecting data from Santa Fe Province, Argentina, that the spider makes a horizontal web on float- ing plants. Distribution. Amazon area to Buenos Aires Province, Argentina (Map 1). Specimens Examined. COLOMBIA Amazonas: 25 km N Leticia, 29 (MCZ). ECUADOR Napo: 10 km E Puerto Napo, S side Rio Napo, 69, 6(5, 14 imm. (MCZ). PERU "northern PERU", Reanrer [local. ?], 49 (AMNH). Loreto: Parinari Canyon, Rio Samiria, 19, 1(5, imm. (AMNH). Hudnuco: Monzon Valley, Tingo Maria, 19, 13 (CAS). BRAZIL Amazonas: Porto America [05°10'S on Rio Madeira or OySQ'S on Rio Purus, Vanzolini, personal communication], 19 (AMNH); Rio Autaz, Santa Amelia, 19 (NRMS); Rio Autaz, Cururuzinho, \$ (NRMS); Manaus, Bilhares, 26 (NRMS); Manaus, Flores, 19 (NRMS); Rio Negro, 19 (MZSP); Manaus, Igarape, km 30, M-Caracarai, 19 (INPA); Manaus, 19 (INPA). Mato Grosso: Pocone, km 11 Transpantaneira, imm. (MCZ). Rio de Janeiro: Santo Antonia de Padua (Goldi, 1892). Sao Paulo: Reprisa, Sao Jose de Rio Preto, 19, U (MZSP). BO- LIVIA El Beni: Yacuma, Espirito, 19 (ZSM); Rosario, 29 (USNM); 19 (M. R. Lopez, USNM). La Paz: Reves, 14°03'S, 68°01'W (Zapfe, 1957); upper lake on c'ha- caltaya, 19 (ZSM). PARAGUAY San Pedro: 2 km NW Lima, imm. (MHNG). Presidente Hayes: Transcha- co, km 320, 29 (IRSNB); Makthlawaiya (Badcock, 1932). Paraguari: nr. Ybytymi, 119, 5(5 (MCZ). Gtiaird: As. Tacuara. Central: Esterus del San Lorenzo, 19, 1(5 (CAS); Asunci6n, 19 (MCZ); Villa del Maestro, San Lorenzo, 29, 1(5, 2 imm. (MHNG). Caaguazu: 20 km N Cnl. Oviedo, 19 (MHNG). ARGENTINA Formosa: Reserva Ecologica El Bagual, 59, 1,5 (FMLT, MCZ). Chaco: Selva del Rio de Oro, 19, 1(5 (MEG). Santa Fe: Arroyo El Toba, 19, 13 (MEG); Tortagal, 59, 1(5 (FMLT). Buenos Aires: San Nicolas, Baradero, Pilar (Holmberg, 1883); Buenos Aires, 19 (MCZ); Delta de Parana, Arroyo Espera, 13 (MEG); Escobar, 1(5 (MACN); Campana, 29 (CAS); Tigre, 29 (BMNH). Spinepeira new genus Type species. Spinepeira schlingeri. The name is an arbitrary combination of letters attached to "epeira". The name of the genus is feminine. AcTiNOSOMA, Spilasma, MicREPEiRA, Pronous • Levi 161 Diagnosis. This genus, which is close to Alpaida, differs by having a pair of long projections on the abdomen, one anterior median tubercle, and two posterior me- dian tubercles (Figs. 16, 17). Relationship. The two posterior me- dian tubercles represent a synapomorphy found in numerous genera including Al- paida, Acanthepeira, Eriophora, Eustala, and Wagneriana. With the exception of Eustala, these genera have males with a paramedian apophysis in the palpus. Ad- ditional characters placing the genus close to Alpaida are the black pigmentation be- tween the median eyes (Figs. 18-20) and lack of setae on the carapace. schli Map 2. Distribution of Spinepeira schlingeri. Spinepeira schlingeri new species Figures 13-20; Map 2 Holotype. Female holotype from Monzon Valley, Tingo Mana, Depto. Huanuco, Peru, 18 Dec. 1954 (E. I. Schlinger, E. S. Ross), in CAS. The species is named after the collector, entomologist E. I. Schlin- ger. Description. Female holotype. Cara- pace yellow-white, with a longitudinal black line through the middle, area be- tween median eyes black (Fig. 18). Che- licerae, labium, endites, sternum black. Legs yellow-white. Dorsum of abdomen with pairs of black and white patches (Figs. 16, 17); venter with genital area and spin- nerets black and a large black patch be- tween (Fig. 17). Posterior median eyes 1.5 diameters of anterior medians, anterior laterals 1 diameter, posterior laterals 1.5 diameters. Anterior median eyes 1.3 di- ameters apart, 2 diameters from laterals. Posterior median eyes their diameter apart, 1.5 diameters from laterals. Ocular quad- rangle wider behind than in front. Height of clypeus equals 1 diameter of anterior median eye. Abdomen with a pair of long projections and one pair of lateral tuber- cles, an anterior median tubercle and two posterior median tubercles (Figs. 16, 17). Total length 5.5 mm. Carapace 1.7 mm long, 1.5 wide, 0.9 wide behind lateral eyes. First femur 2.2 mm, patella and tibia 2.7, metatarsus 2.0, tarsus 0.7. Second patella and tibia 2.0 mm, third 1.3, fourth 1.8. No additional specimens have been found. Hingstepeira new genus Type species. Epeira folisecens Hingston, 1932. The name is an arbitrary combination of letters linking part of explorer Hingston's name to "epeira". The gender of the name Hingstepeira is feminine. Diagnosis. Hingstepeira is separated from most other araneid genera by the shape of the abdomen: elongate, oval, usu- ally widest at the posterior half (Fig. 24), with the spinnerets overhung by an ante- rior to the posterior tip of the abdomen (Figs. 25, 39, 44, 52). It differs from Singa by having less black coloration in the me- dian eye region (Figs. 26-28) and by the genitalia, especially the palpus, which in Hingstepeira has a prominent paramedian apophysis (PM in Fig. 30). In the close spacing of the posterior median eyes (Fig. 38) and the shape and color of the abdo- men, Hingstepeira is similar to Metazy- gia, but the paramedian apophysis of the male palpus (PM in Fig. 30) separates the genera. Unlike other male araneids, except the males of Metazygia gregalis (O. P.- Cambridge), M. benella Levi, and M. yob- ena Levi, the Hingstepeira male has the 162 Bulletin Museum of Comparative Zoology, Vol. 154, No. 3 distal end of the chelicerae modified and with a frontal tooth (Fig. 33). The Meta- zygia species have a widened fang faced by a wide anterior tooth (Levi, 1994, fig. 261). There are other genera in which the females have an elongate abdomen ex- tending beyond the spinnerets. Hingste- peira lacks the thick first pairs of legs of the southeast Asian Perilla Thorell, 1895, and Hingstepeira's abdomen is wider. [Perilla has been placed by Brignoli (1983) in Nephilinae, which belongs in the family Tetragnathidae.] Hingstepeira differs from the Mediterranean and African Nemos- colus Simon, 1895, by having a parame- dian apophysis in the palpus. Hingstepeira differs from Arachnura Vinson, 1863, which has a similar web, by lacking the narrow tail; also, the male Hingstepeira lacks a sclerotized shield on the abdomen. Hingstepeira differs from the Indopacific Milonia Thorell, 1890, in lacking Milonia's silver coloration. (Mil- onia may be a tetragnathid.) Description. Cephalothorax glabrous, orange to orange-brown, median eye re- gion black, without marks; cephalic area darker than thorax. Cephalic region in both sexes wide behind lateral eyes, more than half width of thoracic region (Figs. 26, 32). Median ocular quadrangle narrower be- hind than in front. Abdomen oval, longer than wide, widest in middle or posterior (Fig. 24). Abdomen coloration variable; all but one species (H. isherton) with trans- verse black rectangle between epigynum and spinnerets (Figs. 25, 44, 52). Epigyna variable, often with a smooth, bulbous pro- jection (Figs. 21, 35). The male of only one species (H. foli- secens) is known. Male carapace slightly smaller than that of female (Figs. 32-34). Endite with tooth (Fig. 34) facing tubercle on palpal femur. First coxa with hook on distal margin (Fig. 34). Palpal patella with one macroseta (Fig. 34). First and second legs with macrosetae. Relationship. The presence of the paramedian apophysis of the male palpus (PM in Fig. 30), the absence of a distal hematodocha, the presence of the conduc- tor (C) in the middle of the tegulum, and the free section of the tegulum on the up- per right in the left palpus (T in Fig. 30) place the genus close to Alpaida. Natural History. The Hingstepeira web resembles that of Arachnura by hav- ing a coiled up leaf as retreat in the upper vertical radius (Pi. 1). Distribution. There are four species, all South American, found in the Amazon area and Guyanas. Separating Species. The species differ in the markings on the abdomen. Males and females of the same species have sim- ilar markings. Key to Species of Hingstepeira The only male known is that of H. foHsecens (Figs. 29-34), 1. Epigynum with median knob-shaped structure (Figs. 21-23, 35-37) 2 Epigynum otherwise, without median knob- shaped structure 3 2(1). In ventral view of epigynum, round struc- ture facing posteriorly (Fig. 21); abdo- men with a dorsal, median, posterior black patch (Fig. 24) folisecens In posterior view of epigynum, round structure facing anteriorly (Fig. 35); ab- dominal pattern is a series of facing brackets (Fig. 38) isherton 3(1). In ventral view of epigynum, the median area, anterior of a transverse bar, is con- cave (Fig. 40); abdomen with black bands on sides (Fig. 43) arnolisei In ventral view epigynum convex (Figs. 45, 48); abdomen with two dorsal black bands (Fig. 51 ) dimona Plate 1 . Hingstepeira folisecens (Hingston). Upper web with 35 cm diameter, hub about 140 cm above forest floor. Lower web about 35 cm diameter, hub 160 cm above forest floor (photos, H. Hofer). AcTiNOSOMA, Spilasma, Micrepeira, Pronuus • Levi 163 164 Bulletin Museum of Comparative Zoology, Vol. 154, No. 3 folisecens SURINAM '>\>\^r^^^'^'^ ' '-^GUYANA 'XJ~X'^~ '' VENEZUELA y ) O arnolisei # dimona A isherton Map 3. Distribution of Hingstepeira species. Hingstepeira folisecens (Hingston) Plate 1; Figures 21-34; Map 3 Epeira folisecens Hingston, 1932; 364, figs. 51, 52, webs. Female from Moraballi Creek, 2 mi [3.2 km] east of Essequibo River, 12 mi [19 km] south of Bartica, Guyana, lost. Not in BMNH, HECO. Larinia bristowei Mello-Leitao, 1940: 180, figs. 6, 7, S. Male holotype from Moraballi Creek, Essequibo River, Guyana, in BMNH, examined. Brignoli, 1983: 272. NEW SYNONYMY. Aranea folisecens: — Roewer, 1942: 842. Araneus folisecens: — Bonnet, 1955: 504. Note. Hingston's (1932, figs. 51, 52) il- lustration of the web of this species resem- bles the web recently photographed by H. Hofer (Pi. 1). Also, Hingston described the "abdomen . . . grey-brown darkening to apex which is nearly black, . . . with spin- nerets at about the centre, area behind spinnerets brown, a quadrate black patch in front of spinnerets. . . . Total length 5 mm" (p. 364). The type localities of Epeira folisecens and Larinia bristowei are the same. Description. Female from Taruma Mirim, Amazonas State, Brazil. Carapace orange, lightest posteriorly, area between eyes black. Chelicerae, labium, endites or- ange. Sternum, coxae light orange. Legs dusky orange. Dorsum of abdomen light gray, posterior end with a black patch (Fig. 24); venter with a dark, dusky trapezoidal patch between spinnerets and epigynum (Fig. 25). Posterior median eyes 0.8 di- ameter of anterior medians, laterals 0.5 diameter. Anterior median eyes 0.7 di- ameter apart, 1.3 diameters from laterals. Posterior median eyes 0.2 diameter apart, 2.5 diameters from laterals. Height of clypeus equals 0.4 diameter of anterior median eye. Abdomen elongate, widest behind middle (Fig. 24). Total length 5.7 mm. Carapace 2.1 mm long, 1.4 wide, 1.1 behind lateral eyes. First femur 1.3 mm, patella and tibia 1.8, metatarsus 0.7, tarsus 0.5. Second patella and tibia 1.6 mm, third 1.4, fourth 1.6. Male holotype of Larinia bristowei. Carapace orange, black around eyes. Che- licerae, labium, endites, sternum, legs or- ange. Dorsum of abdomen whitish with three pairs of dusky patches and a black patch above spinnerets (Fig. 31); venter light gray. Posterior median eyes 0.6 di- ameter of anterior medians, laterals 0.5 diameter. Anterior median eyes 0.6 di- ameter apart, slightly more than their di- ameter from laterals. Posterior median eyes 0.6 diameter apart, 3 diameters from lat- erals. Height of clypeus slightly less than diameter of anterior median eye. Abdo- men oval (Fig. 31). First and second legs armed with macrosetae; second thicker than first. Total length 3.7 mm. Carapace 1.9 mm long, 1.4 wide. First femur 1.4 mm, patella and tibia 1.9, metatarsus 1.2, tarsus 0.6. Second patella and tibia 1 .6 mm, third 1.1, fourth 1.5. AcTiNOSOMA, Spilasma, Micrepeira, Pronous • Levi 165 Note. Males and females have been col- lected together and share similar markings on the abdomen, although males are darker than females. Variation. Total length of females 4.7 to 7.4 mm, males 3.1 to 4.6. The sides of the abdomen in ventral view may be dark. The posterior patch on the abdomen and ventral markings may be almost absent, or gray to black. Illustrations were made from a female from near Manaus, Brazil (Figs. 21-28), from the male holotype of Larinia bristowei (Fig. 29), and from specimens from near Manaus, Brazil (Figs. 30-34). Diagnosis. The elongate abdomen and the characteristic markings, a black patch on the posterior end (Figs. 24, 25, 31), sep- arate this species from all other araneids. Natural History. Hingstepeira folise- cens makes a vertical web with a curled leaf retreat attached to an upper radius (Kingston, 1932, figs. 51, 52; Pi. 1). Webs have been found in the interior and border of forests. They were found in igapo (pe- riodically flooded forests) near Manaus and on the Manaus-Itacotiara in campina for- est highway and in terra firma forest near Manaus. Adult males were common only during February and March. Distribution. Guy anas and Brazilian Amazonian area (Map 3). Specimens Examined. SURINAM Saramacca: Voltzberg-Raleighvallen Nature Reserve, 04°32'N, 56°32'W, 8 Feb. 1982, 19 (D. Smith Trail, MCZ). Merowijne: Anapaike Village, Lawa River, 8-29 Nov. 1963, 12 (E. Malkin, AMNH). Brokopondo: Browns Berg, 5°N, 55''27'W, 20 Feb. 1982, 12 (D, Smith Trail, MCZ). FRENCH GUIANA Cayenne: Montagnes Kaw, nr. Camp Caiman, ca. 27 km SE Paura, 04°33'N, 52°09'W, 100-300 m, 25 Aug. 1988, 12; 1 Sept. 1988, 1(5 (S. Marshall, USNM). BRAZIL Amazonas: Taruma Mirim, igapo, Manaus, 24 Apr. 1984, 12 (J. Adis, INPA); 25 Feb. 1987, 12 (H. Hofer, INPA); 5 Dec. 1987, 12 (H. Hofer, INPA); km 192, highway Manaus to Itacotiara, 18 Apr. 1987, 12 (H, Hofer, INPA); 80 km N Manaus, 9 Mar. 1989, 16 (H. Fowler, MCZ); Colosso Reserve, 80 km N Manaus, 18 Jan. 1989, 12, 1(3; Jan. 1989-June 1991, 632, 12(5; Reserva Cabo Frio, 80 km N Manaus, Oct. 1989-June 1991, 132, 3(5; 13 May 1992, 12; Reserva Dimona, 80 km N Manaus, 1989-1992, 32; Mar.-June 1991, 182; Reserva Km 41, 80 km N Manaus, Mar.-May 1991, 72; Reserva C. de Powell, 80 km N Manaus, 32; Reserva Florestal, 80 km N Manaus, 19 Feb. 1991, 32; Reserva Gaviao, 21 Feb.-21 Oct. 1991, 32; (all H. Fowler, R. S. Vieira, E. Venticinque, MCZ); Reserva Porto Alegre, 80 km N Manaus, 1989-1992, 12; 27 May 1992, 12 (H. G. Fowler, MCZ). Hingstepeira isherton new species Figures 35-39; Map 3 Holotype. Female holotype from Isherton, Guyana, 10 Nov. 1937 (W. G. Hassler), in AMNH. The spe- cific name is a noun in apposition after the locality. Description. Female holotype. Cara- pace orange, cephalic region darkest. Che- licerae orange-brown. Labium, endites dark brown. Sternum orange, sides dusky. Legs orange. Dorsum of abdomen with white pigment spots and five pairs of brackets (Fig. 38); venter black around spinnerets fading into surrounding area, except for clear border posteriorly (Fig. 39). Posterior median eyes 0.8 diameter of anterior medians, laterals 0.6 diameter. Anterior median eyes 0.8 their diameter apart, 1.8 diameters from laterals. Poste- rior median eyes 0.2 diameter apart, 4 di- ameters from laterals. Height of clypeus equals 0.2 diameter of anterior median eye. Abdomen elongate oval, widest posteriorly (Fig. 38). Total length 7.0 mm. Carapace 3.4 mm long, 2.1 wide, 1.5 behind lateral eyes. First femur 2.3 mm, patella and tibia 3.0, metatarsus 2.0, tarsus 0.8. Second pa- tella and tibia 2.7 mm, third 1.6, fourth 2.3. Note. This female has a dorsal abdom- inal pattern similar to Metazygia, and its eyes resemble those of Metazygia (Levi, 1994). The epigynum also resembles that of Metazygia goeldii Levi. The main Hingstepeira character is the elongate ab- domen (Fig. 38). Only finding a male will ascertain the placement. Diagnosis. The shape of the epigynum (Figs. 35-37) and the dorsal abdominal pattern (Fig. 38) separate this species from other Hingstepeira. Hingstepeira arnolisei new species Figures 40-44; Map 3 Holotype. Female holotype from Ilha de Maraca, Rio Uraricoera, Roraima Stae, Brazil, 5 Dec. 1987 166 Bulletin Museum of Comparative Zoology, Vol. 154, No. 3 (A. A. Lise), in MCN no. 20062. The species is named after the collector. Description. Female holotype. Cara- pace light orange-yellow. Chelicerae, la- bium, endites, sternum, legs yellowish white. Dorsum of abdomen with four lon- gitudinal white bands, sides black anteri- orly and posteriorly (Fig. 43); venter with a tranverse black rectangle between epi- gynum and spinnerets, sides black (Fig. 44). Posterior median eyes 0.8 diameter of anterior medians, anterior laterals 1.2 di- ameters, posterior laterals 0.8 diameter. Posterior median eyes oval. Anterior me- dian eyes their diameter apart, 2.3 diam- eters from laterals. Posterior median eyes their narrow diameter apart, 5 diameters from laterals. Height of clypeus equals di- ameter of anterior median eye. Abdomen elongate, 1.8 times as long as wide (Fig. 43). Total length 8.0 mm. Carapace 3.1 mm long, 2.3 wide, 1.8 behind lateral eyes. First femur 1.9 mm, patella and tibia 3.0, metatarsus 1.8, tarsus 0.9. Second patella and tibia 2.5 mm, third 1.7, fourth 2.5. Diagnosis. Hingstepeira arnolisei dif- fers from other Hingstepeira species by having a concave area on the anterior of the epigynum (Fig. 40) and by the lateral, black banding of the abdomen (Figs. 43, 44). Hingstepeira dimona new species Figures 45-52; Map 3 Holotype. Female holotype from Reserva Dimona, 80 km north of Manaus, Amazonas State, Brazil, in forest, 15 May 1991 (H. Fowler, R. S. Vieira, E. Venticinque), in MCN no. 25543. The specific name is a noun in apposition after the type locality. Description. Female holotype. Cara- pace orange, darkest in midline. Chelic- erae dark orange. Labium, endites, ster- num orange. Legs orange with indistinct dark orange rings. Dorsum of abdomen with two black bands, midline and sides of bands white (Fig. 51); venter with a black transverse rectangle between epi- gynum and spinnerets, sides and posterior black (Fig. 52). Posterior median eyes 0.8 diameter of anterior medians, laterals 0.6 diameter. Anterior median eyes 0.6 di- ameter apart, 2 diameters from laterals. Posterior median eyes 0.3 diameter apart, 2.7 diameters from laterals. Height of clypeus equals 0.5 diameter of anterior median eye. Abdomen oval (Fig. 51). Total length 5.2 mm. Carapace 2.5 mm long, 1.8 wide, 1.3 behind lateral eyes. First femur 1.8 mm, patella and tibia 2.3, metatarsus 1.4, tarsus 0.7. Second patella and tibia 1.9 mm, third 1.1, fourth 1.8. Variation. Total length of females 4.7 to 6.0 mm. The borders of the posterior lateral plates of the epigynum (Figs. 46, 49) are variable. Figures 45-47, 51, and 52 were made from the female holotype, and Figures 48-50 from the specimen from Reserva Ducke. Diagnosis. The domed structure of the epigynum (Figs. 45-50) and the dorsal black bands of the abdomen (Fig. 51) sep- arate this species from other Hingstepeira. The species may belong to Metazygia and be close to M. laticeps (Kevserling) (Levi, 1995, figs. 226-230). Natural History. All females were col- lected in forest. Paratypes. BRAZIL Amazonas: Reserva Dimona, 80 km N Manaus, 26 Mar. 1991, 22; 17 Mav 1991, Figures 21-34. Hingstepeira folisecens (Hingston). 21-28, female. 21-23, epigynum. 21 , ventral. 22, posterior. 23, lateral. 24, dorsal. 25, abdomen ventral. 26, carapace. 27, carapace and chelicera. 28, eye region and chelicerae. 29-34, male. 29, left palpus. 30, palpus pulled apart. 31 , dorsal. 32, carapace. 33, carapace and chelicera. 34, eye region, chelicerae, and right palpus. Figures 35-39. H. istierton n. sp., female, 35-37, epigynum. 35, ventral. 36, posterior. 37, lateral. 38, dorsal. 39, abdomen, ventral. Figures 40-44. ventral. H. arnolisei n. sp., female. 40-42, epigynum. 40, ventral. 41, posterior. 42, lateral. 43, dorsal. 44, abdomen. ACTINOSOMA, Spilasma, Micrepeira, Pronous • Levi 167 Figures 45-52. H. dimona n. sp., female. 45-50, epigynum. 45, 48, ventral. 46, 49, posterior. 47, 50. lateral. 45^7, (holotype). 48-50, (paratype). 51, dorsal. 52, abdomen, ventral. Abbreviations. A, terminal apophysis. C, conductor. E, embolus. M, median apophysis. PM, paramedian apophysis. Scale lines. 1.0 mm, genitalia 0.1 mm. 168 Bulletin Museum of Comparative Zoology, Vol. 154, No. 3 19; 25 June 1991, 22 (H. Fowler, R. S. Vieira, E. Venticinque, MCZ). Specimens Examined. BRAZIL Amazonas: Re- serva Ducke, Manaus, 26 July 1973, 19 (L. P. Albu- querque, MCN 20050); 27 Feb. 1974, 29 (L. P. Al- buquerque, MCN 23561); Reserva Km 41, 80 km N Manaus, 21 May 1991, 19 (H. Fowler, R. S. Vieira, E. Venticinque, MCZ); Cabo Frio, 80 km N Manaus, 13 Mar. 1989, 19; 27 May 1989, 19 (H. Fowler, R. S. Vieira, E. Venticinque, MCZ). Pronous Keyserling Pronous Keyserling, 1881: 547. Type species Pronous tuberculifer Keyserling, 1881, by monotypy. The gender of the name is masculine (Bonnet, 1958: 3778). Paphlagon O. P.-Cambridge, 1893: 117. Type species Paphlagon beatus O. P.-Cambridge, 1893, by monotypy. Species (erroneously) and genus syn- onymized with P. tuberculifer by O. P.-Cambridge (1898: 281). Zigana Chamberlin and Ivie, 1936: 53. Type species Zigana wixoides Chamberlin and Ivie, 1936, by monotypy. NEW SYNONYMY. Note. The synonymy of Zigana with Pronous was recognized by Chickering, judging from specimen labels, but was never published. Diagnosis. The large posterior median eyes, which have lost their tapeta, are about twice the diameter of the anterior median eyes, face laterally, and are closer to the lateral eyes than to each other (Figs. 53- 56, 58-62). This apomorphy separates Pronous from all other araneid genera in- cluding Micrathena. Micrathena and Pronous share two apomorphies: the fourth legs are longer than the first, and the ab- domen of the male is almost rectangular (Figs. 61,62). Pronous differs from Hypsosinga by having the clypeus height equal to the di- ameter of the anterior median eye (in Hypsosinga the clypeus is higher) and by having the fourth legs longer than the first. Pronous may be confused with liny- phiid genera (e.g., Dubiaranea Mello-Lei- tao, 1943 = Paranesticus Mello-Leitao, 1944: 333 = Hormembolus Millidge, 1991), which have similar large posterior median eyes. The linyphiids, however, have a high clypeus and thinner legs and may have a sclerotized stridulating ridge laterally on the chelicerae. Description. Female. Species all with similar coloration and pattern: cephalo- thorax orange to orange-brown, sides of thoracic region sometimes darker. Chelic- erae, labium, and endites orange. Sternum light orange-red. Coxae and legs dusky or- ange to brown, rarely black. Abdomen light orange with six black patches, two on mid- dle tubercles and two pairs on posterior tubercles (Figs. 56, 57). Some specimens have scattered white pigment spots, es- pecially on anterior median tubercle, and gray to black shading on posterior and sides (Fig. 55). Venter usually without pigment, sometimes with a median black streak. Living specimens are all bright orange-red with black marks. Carapace punctate. Anterior median eyes always slightly larger than laterals and posterior medians about 2 diameters of an- terior medians (Figs. 53-55). Median eye quadrangle wider behind than in front (Fig. 53). Height of clypeus about equal to diameter of anterior median eyes. Legs with femur almost as long as patella and tibia. Unlike Micrathena, Pronous lacks a stridulating area on the book-lung covers. Carapace width behind lateral eyes about half maximum diameter of carapace (Fig. 53). Posterior median eyes about 2 diam- eters of anterior medians. Anterior eyes their diameter apart, about 2 diameters from laterals. Posterior median eyes 1 to 1.5 diameters apart, 0.5 to 0.8 diameter from laterals. Abdomen with a small an- terior median tubercle and three pairs of tubercles, one pair in middle, two pairs posterior (Pi. 2, Figs. 56, 57). Anterior me- dian tubercle usually absent in females from South America (Fig. 57). Epigynum (Fig. 63) without scape, sclerotic areas transparent and difficult to delineate. Male. Slightly smaller than female. Car- apace punctate. Width of carapace behind lateral eyes narrower than half of width in thoracic region (Fig. 58). Posterior me- dian eyes 1.5 to 2 diameters of anterior AcTiNOSOMA, Spilasma, MiCREFEiRA, Fronovs • Levi 169 median eyes. Anterior median eyes 0.6 to 1 diameter apart, 1 to almost 2 diameters from laterals. Posterior median eyes 1.1 to 1.5 their diameter apart, 0.5 to 1 diameter from laterals. Difference in length be- tween first legs and longer fourth legs is always less in males than in females. Ab- domen with a rectangular, lightly sclero- tized, dorsal scutum, widest in anterior half, with only faint indications of tubercles (Figs. 60-62). Male endite with tooth (Fig. 59), palpal femur with facing tubercle. Palpal patella with one macroseta (Fig. 59). First coxa with a small hook posteriorly on distal rim of margin, second femur with opposing small groove (Fig. 59). Second tibia with macrosetae on anterior face. Palpus with median apophysis within frame of tegul- um in median view (M in Figs. 64, 65). Conductor positioned centrally on the te- gulum (C in Figs. 64, 65). A lobe of te- gulum overhangs conductor (at 12 hr in Figs. 64, 65). Embolus a curved structure, variable in thickness and shape in different species (E in Figs. 64, 65). Palpus without paramedian apophysis. Variation. All species are remarkably similar in coloration, shape, and size (Pi. 2, Figs. 56, 57). Also, the variation of spec- imens of different species collected in the same area is similar: most South American females have smaller tubercles on the ab- domen (Fig. 57) than those from Central America (Fig. 56), and the two specimens collected in Guyana (P. nigripes and P. intus) have black legs. Relationship. The long fourth legs and nearly rectangular male abdomen of Pron- ous are characters shared with Micrath- ena; both are believed to be synapomor- phies. However, Micrathena uses the long fourth legs to hold its abdomen upside down, horizontally, at a right angle to the web, making the sky and leaves the back- ground for the bright dorsal coloration and the ground the background for the dull ventral view. Pronous, on the other hand, adopts a resting position more like that of other araneids (Eberhard, personal com- munication). Thus, the homology of the long legs in the two genera is uncertain. Natural History. The species makes small vertical webs, about 7 to 10 cm di- ameter, above leaf litter in forests (per- sonal observations in Panama; Pi. 2). The web has few supporting threads and may catch walking insects and insects flying above ground level (Lubin, 1978). At the slightest disturbance, the spider falls from the web and disappears in leaf litter. Spec- imens are thus difficult to collect. About half of the available Pronous specimens were collected by A. M. Chickering, who laboriously sifted leaf litter. Distribution. Fourteen species are American, most coming from Mexico and Central America (Map 4). The four species described by Simon (Roewer, 1942: 967) from Asia and Madagascar are probably all misplaced. The holotypes of only two of the four were found in the MNHN. Pronous laevisternis Simon, 1908, from Tonkin, the area around Hanoi, Vietnam, and P. taprobanicus Simon, 1895, from Ceylon (Sri Lanka), were examined and found to belong to Hypsosinga. The first is H. pygmaea (Sundevall, 1831), NEW SYNONYMY (see Levi, 1975). Another two, more recently described, Pronous minutus Saito, 1939 (in Yaginuma, 1986), and Pronous tetraspinulus Yin, Wang, Xie and Peng, 1990 (Platnick, 1993: 462) from China and Japan are probably also mis- placed. The fourth legs are much shorter than the first in P. minutus (Song, personal communication); P. minutus may also be Hypsosinga. Pronous chelifer Hasselt (1882: 24), from Sumatra, is Gea spinipes C. L. Koch, 1843 (Levi, 1983). Separating Species. Males are readily separated by the structure of their palpi, the shape of the embolus, the median apophysis, conductor, and tegular lobe. Females are difficult to separate: the epi- gyna of females of different species are similar. The borders of structures in the epigynum are transparent and difficult to delineate. The cleared epigynum showing ducts (cleared in Hoyer's medium) is help- 170 Bulletin Museum of Comparative Zoology, Vol. 154, No. 3 ACTINOSOMA, Spilasma, Micrepeira, Pronous • Levi 171 ful, but the dorsal view of the epigynum (vulva) is less so. Key to Female Pronous The terms used in this key are illustrated in Figure 63. 1. 2(1). 3(2). 4(3). 5(4). 6(5). 7(1). Females from South America (Maps 4B, D, F, G) 2 Females from Mexico or Central Amer- ica (Maps 4A, B, C, E, F, G) 7 Epigynum with V-shaped dusky patch (Figs. 63, 131, 135); lateral and pos- terior overhang of epigynum wide (Figs. 131, 135); in posterior view long a.xis of opening longitudinal (Figs. 134, 138); South America (Map 4D) tuberculifer Epigynum otherwise (Figs. 98, 107, 117, 1 22, 1 27 ) 3 Epigynum with posterior median plate narrow ventrally, wider dorsally (at center of Fig. 120); lateral overhang with a pointed projection in ventral view (Fig. 117); Colombia (Map 4G) Sides of posterior median plate straight dorsally (Figs. 101, 106, 125, 130) 4 Anterior lip of epigynum visible in ven- tral view (Figs. 103, 107, 122, 127) 5 Anterior lip not visible in ventral view (Fig. 98); Colombia (Map 4F) wixoides Epigynum with anterior lip small (Fig. 127); dusky patches with a pair of round spots (Fig. 127); Guyana (Map 4G ) n igripes Epigynum with anterior lip larger (Figs, 103, 107, 122); dusky patches other- wise Anterior lip swollen in median (Figs. 103, 107); posterior lip curved anteriorly in median (Figs. 103, 107); Costa Rica to Ecuador, Venezuela, northern Brazil (Map 4B) intus Anterior lip not swollen in median (Fig. 122); Colombia (Map 4G) pance Mexico (Maps 4A, C, E) 8 Honduras to Panama (Maps 4A, B, C, E, F, G) _ _ 10 Epigynum with a median notch (Fig. 71) quintana Epigynum without notch (Figs. 66, 75) 9(8). 10(7). 11(10). 12(11) Median area of epigynum with anterior margin of epigynum V-shaped (Fig. 66) beatus Median area of epigynum with margins parallel forming a small bordered lon- gitudinal groove (Fig. 75) felipe Epigynum in ventroposterior view with a transverse, oval median plate having a small median longitudinal ridge (Fig. 114); margin of epigynum rebordered (Fig. 112); Panama (Map 4G) shanus Epigynum otherwise (Figs. 79, 81, 83 85, 88, 90) 11 In posterior view, epigynum median plate with wide median split (Fig. 96); Costa Rica (Map 4A) golfito Epigynum otherwise (Figs. 69, 82, 101) 12 Posterior lips of epigynum almost tri- angular in shape with a posterior me- dian bulge (Fig. 79); Honduras (Map 4E) lancetilla Epigynum otherwise (Figs. 66, 83, 88 98) 13 13(12). No distinct anterior lip visible in ventral illg view (Figs. 66, 98) 14 Anterior lip visible in ventral view, 4 sometimes only in median area (Figs. 83, 88) 15 14(13). Posterior median plate wider dorsally than ventrally and only barely notched ventrally (at 12 hr in Figs. 68, 69); Costa Rica (Map 4A) beatus - Sides of posterior median plate about parallel and with a long median di- vision (Fig. 101); in cleared view with a pair of median wide duct curves (Fig. 99); Panama (Map 4F) wixoides Q 15(13). Margin of epigynum in ventral view with a median longitudinal bordered groove (Fig, 83); Costa Rica (Map 4C) peje Epigynum otherwise (Figs. 88, 103, 107) 16 16(15). Ventroposterior view of epigynum with two curved openings, lateral plates constricting median plate; median plate with long groove (Fig. 90); Costa Rica (Map 4 A) colon - Epigynum otherwise, lateral plates only slightly constricting median plate (Figs. 105, 109); median plate with short ventral split (Figs. 106, 1 10); Costa Rica to Colombia (Map 4B) _ intus Plate 2. Pronous tuberculifer, bright orange-red coloration with black patches and dusky legs, total length 5 mm. Web of Pronous wixoides about 10 cm wide (lower photo. W. Eberhard). 172 Bulletin Museum of Comparative Zoology, Vol. 154, No. 3 Key to Male Pronous The males of P. colon, P. lancetilla, and P. nigripes are unknown. 1. Outer, upper edge of median apophysis (M in Figs. 64, 65) with a square notch, transverse spine of median apophysis thick (Fig. 97); Costa Rica (Map 4A) ... golfito Median apophysis without square corner notch (Figs. 64, 70, 74, 116) 2 2(1). Distal, outer edge of median apophysis with an outer spine and an elongate notch between it and edge (Figs. 70, 74, 116, at 3 hr in Fig. 70); Mexico to Pan- ama (Map 4) 3 Outer edge of median apophysis entire (Figs. 64, 78, 87, 121) 5 3(2). Embolus (E in Figs. 64, 65) a semicircular disk (Fig. 70); Mexico, Costa Rica (Map 4A) beatus Embolus filamentous (Figs. 74, 116) 4 4(3). Transverse spine of median apophysis close to its "upper" edge (Fig. 74); Mexico (Map 4C) quintana - Transverse spine of median apophysis in middle of median apophysis (Fig. 116); Panama (Map 4G) shanus 5(2). Embolus a long fine thread, whose length (if stretched) is longer than diameter of palpus (Fig. 102); Panama to Colombia, Ecuador (Map 4F) wixoides Embolus otherwise (Figs. 87, 121, 139) 6 6(5). Embolus a semicircular disk (partly trans- parent) (Fig. 139); South America (Map 4D) _ tuberculifer Embolus more or less filamentous (Figs. 78, 111, 121, 126) 7 7(6). Spine of median apophysis annulate (Fig. 121); Colombia (Map 4G) valle - Spine of median apophysis smooth (Figs. 87, 97, 126) 8 8(7). Median apophysis pointed distally "above" spine (at 3 hr in Fig. 78); Mexico (Map 4E) felipe - Median apophysis rounded "above" spine (Figs. 87, 111, 126) 9 9(8). Embolus thick, wider than opening of notch, the "upper" area enclosed by embolus (Fig. 87); Costa Rica (Map 4C) peje Embolus a finer thread (Figs. Ill, 116) 10 10(9). Embolus a fine thread (Fig. 126); tegulum lobe truncate (at 12 hr in Fig. 126); Colombia (Map 4G) pence Embolus thick, tegulum lobe rounded (Fig. Ill); Costa Rica to Ecuador (Map 4B) intus Pronous beatus (0. P.-Cambridge) Figures 66-70; Map 4A Paphlagon beatus O. P.-Cambridge, 1893: 117, pi. 14, fig. 10, 2. Female from Teapa [Tabasco State], Mexico, in BMNH, lost. Erroneously synonymized with P. tuberculifer by O. P.-Cambridge (1898: 281). Note. The holotype female, without male of P. beatus, could not be found in the BMNH. Cambridge later received more specimens and cited P. beatus as a synonym of P. tuberculatus (in error). These later females, together with males, were found in the BMNH to be P. quin- tana. There are two specimens in the HECO collection, one male and one fe- male. The only label in the vial was made out for my loan, with my loan number, the Paphlagon beatus name, and locality, Teapa. However, the specimens also be- longed to P. quintana. From O. P. -Cam- bridge's illustration, it is clear that the Hope Collection specimens cannot be the types. This was verified by a letter from I. Lans- bury, stating that the holotype should be in the Natural History Museum in London. I suspect that when O. P.-Cambridge syn- onymized Paphlagon beatus (erroneously) the type label of P. beatus was changed, although it will always remain the type of this name, whether synonymized with an- other name or not. Description. Female from Huehuetan. Figures 53-65. Morphology of Pronous. 53-57, 63, female. 53, carapace. 54, eye region and chelicerae. 55, lateral. 56, 57, dorsal. 63, epigynum, diagrammatic. 58-62, 64, 65, male. 58, carapace. 59, eye region, chelicerae, right palpus, left first coxa and second trochanter, and proximal part of femur. 60, lateral. 61, 62, dorsal. 63, epigynum, diagrammatic. 64, 65, left male palpus. 64, palpus pulled apart. 65, ventral. 53, 54, 56, 58-61 , 64, 65, P. intus. 55, P. wixoides. 57, 62, P. tuberculifer Figures 66-70. Pronous beatus (O. P.-Cambridge). 66-69, female epigynum. 66, ventral. 67, ventral, cleared. 68, ventroposterior. 69, posterior. 70, male left palpus, mesal. AcTiNOSOMA, Spilasma, Micrepeira, Prunous • Levi 173 U$^ j /jl^'H*^" Figures 71-74. P. quintana n. sp. 71-73, female epigynum. 71, ventral. 72, ventral, cleared. 73, posterior. 74, male palpus. Figures 75-78. P. felipe n. sp. 75-77, female epigynum. 75, ventral. 76, ventral, cleared. 77, posterior. 78, male palpus. Abbreviations. C, conductor. E, embolus. M, median apophysis. R, radix. T, tegulum. Scale lines. 1.0 mm, genitalia 0.1 mm. 174 Bulletin Museum of Comparative Zoology, Vol. 154, No. 3 Total length 4.5 mm. Carapace 2.0 mm long, 1.7 wide, 1.1 behind lateral eyes. First femur 2.2 mm, patella and tibia 2.2, meta- tarsus 1.7, tarsus 0.9. Second patella and tibia 1.9 mm, third 1.5. Fourth femur 2.9 mm, patella and tibia 2.8, metatarsus 2.1, tarsus 0.7. Male from Huehuetan. Abdomen elon- gate, intermediate between P. intus (Fig. 61) and P. tuberculifer (Fig. 62). Total length 4.3 mm. Carapace 1.9 mm long, 1.6 wide, 0.8 behind lateral eyes. First femur 2.1 mm, patella and tibia 2.0, metatarsus 1.8, tarsus 0.9. Second patella and tibia 1.7 mm, third 1.3. Fourth femur 2.7 mm, pa- tella and tibia 2.4, metatarsus 2.2, tarsus 0.9. Note. Males and females were collect- ed together. Variation. Total length of females 4.5 to 4.8 mm, males 4.3 to 4.7. The specimens illustrated came from near Huehuetan, Chiapas. Diagnosis. The epigynum of the fe- male differs from P. quintana (Figs. 71- 73) in lacking a median notch in ventral view (Fig. 66). In posterior view the epi- gynum differs from that of P. felipe (Fig. 77) by having a median plate that is nar- row ventrally and widest dorsally (Fig. 69). The male has a round semicircular em- bolus (Fig. 70), as does the South American P. tuberculifer (Fig. 139). The outer edge of the median apophysis has a notch sep- arating off an outer spine (Fig. 70); the transverse spine has nearly parallel sides and, unlike P. quintana (Fig. 74), a large lobe above (Fig. 70). Distribution. Mexico, perhaps to Costa Rica (Map 4A). Specimens Examined. MEXICO Nayarit: San Bias, 4-5 Aug. 1947, IS (B. Malkin, C, M. Goodnight, AMNH). Jalisco: Rio Pitillal, Playa Grande, 5 km E Puerto Mallarta, 12 July 1989, 19; 8 July 1992, 39 (R. West, MCZ); Chamela, Sept, 1989, 19 (W. Eberhard, MCZ). Oaxaca: 3.2 km SE Niltepec, 16°32'N, 94°33'W, 16 Aug. 1966, 19 (J., W. Ivie, AMNH); 8 km E Ta- panatepec, 230 m, 28 Aug. 1967, 19 (R. E. Leech, REL). Chiapas: Esquintla, 19 (Crawford, MCZ); 10 km N Arriaga, 305 m, 23 Aug. 1972, 19 (C. Mullinex, K. Lucas, CAS); Finca Santa Marta, nr. Huehuetan, 31 July-1 Aug. 1950, 59, 2$ (C, M. Goodnight, AMNH); 16 km SE Tierra y Libertad [?], 1,000 m, 23 Aug. 1972, 13 (C. MulHnex, K. Lucas, CAS). COS- TA RICA Cartago: El Cedral, Navarro Orosi-Car- tago, 29 Nov. 1979, 19, doubtful determination (C. E. Valeric, MZCR). Pronous quintana new species Figures 71-74; Map 4C Pronous tuberculatus: — O. P. -Cambridge, 1898: 281, pi. 36, fig. 13, S (misidentification). Holotype. Male holotype, male paratype, and two female paratypes from 31 km NE of Felipe Carrillo Puerto, on Highway 307 toward Tulum, ca. 19°48'N, 87°52'W, Quintana Roo, Mexico, 17 July 1983 (W, Maddison, R. S. Anderson), in MCZ. The specific name is a noun in apposition after the locality. Note. Cambridge specimens (1898, but not 1893), both females and males, are all P. quintana. Description. Female paratype. Color- ation as in other species, legs dark orange, dusky marks on sides of carapace. Total length 5.4 mm. Carapace 2.1 mm long, 1.9 wide, 1.1 wide behind lateral eyes. First femur 2.0 mm, patella and tibia 1.9, meta- tarsus 1.7, tarsus 0.7. Second patella and tibia 1.8 mm, third 1.5. Fourth femur 2.8 mm, patella and tibia 2.7, metatarsus 2.1, tarsus 0.9. Male holotype. Abdomen relatively short, intermediate between P. intus (Fig. 61) and P. tuberculifer (Fig. 62), with soft, bulging sides. Black patches small. Total length 4.3 mm. Carapace 1.9 mm long, 1.5 wide, 0.7 behind lateral eyes. First femur 1.2 mm, patella and tibia 1.9, metatarsus 1.3, tarsus 0.7. Second patella and tibia 1.5 mm, third 1.1. Fourth femur 2.1 mm, pa- tella and tibia 2.0, metatarsus 1.6, tarsus 0.6. Note. Males and females were matched on the basis of two collections that includ- ed both sexes. Variation. Total length of females 3.6 to 5.4 mm, males 3.6 to 4.3. The illustration of the male palpus (Fig. 74) was made from the holotype; the female (Figs. 71- 73) was drawn from a specimen from near La Palma, Veracruz. Diagnosis. The female differs from other Mexican species by having a deep AcTiNOSOMA, Spilasma, Micrepeira, Fronovs • Levi 175 Map 4. Distribution of Pronous species. 176 Bulletin Museum of Comparative Zoology, Vol. 154, No. 3 median notch on the posterior margin of the epigynum (Fig. 71). The male differs by having a short, thread-hke embolus and by the median apophysis having a lateral notch, separating off an outer spine (Fig. 74). Natural History. The male holotype was found hanging by a thread from veg- etation. The female was found close to the ground in forest, along the edge of a road in a small orb web with its hub open. Distribution. Southern Mexico (Map 4C). Specimens Examined. MEXICO Veracruz: Es- tacion de Biologi'a Tropical "Los Tuxtlas", nr. La Palma, 15 km N Catemaco, ca. 18°36'N, 95°07"W, 1, 2 Aug. 1983, 19, 16 (W. Maddison, AMNH ex MCZ); Aug. 1986, 2$ (W. Eberhard 3363, 3366, MCZ). Chia- pas: Pichucalco, 17 Julv 1947, 12 (C, M. Goodnight, AMNH); Palenque, July 1981, 12 (C. Gold, CAS). Tabasco: Teapa, 22, 2$ (BMNH, HECO). Pronous felipe new species Figures 75-78; Map 4E Holotype. Male holotype from 31 km NE of Felipe Carrillo Puerto on Highway 307 to Tulum, 19°48'N, 87°52'W, Quintana Roo, Mexico, 17 July 1983 (W. Maddison, R. S. Anderson), in MCZ. The specific name is a noun in apposition after the locality. Description. Female from Coba Ruins, Quintana Roo. Total length 4.9 mm. Car- apace 2.2 mm long, 1.9 wide, behind lat- eral eyes 0.9 wide. First femur 2.1 mm, patella and tibia 2.1, metatarsus 1.6, tarsus 0.7. Second patella and tibia 2.0 mm, third 1.6. Fourth femur 2.9 mm, patella and tibia 2.6, metatarsus 2.3, tarsus 0.9. Male holotype. Coloration light, abdo- men intermediate in length between P. intus (Fig. 61) and P. tuberculifer (Fig. 62). Posterior dorsal black patches small. Posterior dorsal tubercles distinct. Total length 4.3 mm. Carapace 1.8 mm long, 1.5 wide, 0.7 wide behind lateral eyes. First femur 1.7 mm, patella and tibia 1.7, meta- tarsus 1.3, tarsus 0.7. Second patella and tibia 1.4 mm, third 1.1. Fourth femur 2.1 mm, patella and tibia 2.0, metatarsus 1.7, tarsus 0.7. Note. Males and females were matched by elimination, both being the third Mex- ican species after males and females of the other two species were associated. Variation. Total length of females 4.9 mm, males 3.8 to 4.3. The illustrations were made from specimens from the Coba Ru- ins. Diagnosis. The female differs from P. beatus (Figs. 66-69) and P. quintana (Figs. 71-73) by having a median longitudinal bordered groove in ventral view of the epigynum (Fig. 75). In the male, the pal- pus has a short embolus enclosing a shallow notch, and the median apophysis has a dis- tal, "upper" point and no outer spine (Fig. 78). Distribution. Eastern Mexico (Map 4E). Specimens Examined. MEXICO San Luis Potosi: Tamazunchale, 6, 7 Julv 1941, 12 (L. I. Davis, AMNH); 19 April 1963, 12 (W. J. Gertsch, W. Ivie, AMNH); 26 June 1947, 12 (B. Malkin, AMNH). Veracruz: Cor- doba, 20-27 July 1976, 12 (C. H., H. Seevers, AMNH); Tecolutla, 13 Oct. 1947, 16 (H. M. Wagner, AMNH). Quintana Roo: Coba Ruins, ca. 20°30'N, 87°42'W, 18 July 1983, 12 (W. Maddison, MCZ). Pronous lancetilla new species Figures 79-82; Map 4E Holotype. Female holotype from Lancetilla, Depto. Atlantida, Honduras, July 1929 (A. M. Chickering), in MCZ. The specific name is a noun in apposition after the locality. Description. Female holotype. Total length 4.4 mm. Carapace 2.0 mm long, 1.6 wide, 0.9 behind lateral eyes. First femur 1.9 mm, patella and tibia 1.9, metatarsus 1.3, tarsus 0.7. Second patella and tibia 1.7 mm, third 1.2. Fourth femur 2.5 mm, pa- tella and tibia 2.1, metatarsus 1.8, tarsus 0.8. Diagnosis. Pronous lancetilla differs from P. golfito (Figs. 93-96) and other Pronous by having in ventral view of the epigynum a semicircular anterior margin, no anterior lips visible, and a transverse posterior lip forming a triangle on each side (Fig. 79). In posterior view, the me- dian plate lacks the deep split (Fig. 82) present in P. golfito (Fig. 96). The internal AcTiNOSOMA, Spilasma, MiCREPEiRA, Pronuus • Levi 177 genitalia are simple (Fig. 80) as in P. gol- fito (Fig. 94). Pronous peje new species Figures 83-87; Map 4C Holotype. Male holotype from La Selva, near Puerto Viejo, Heredia Prov.', Costa Rica, 1-3 Dec. 1981 (J. Coddington), in MCZ. The specific name is a noun in apposition after the name of a stream near the type locaHty. Description. Female paratype. Color- ation as in other species but legs brown. Total length 4.3 mm. Carapace 1.9 mm long, 1.5 wide, 0.9 behind lateral eyes. First femur 1.9 mm, patella and tibia 1.9, meta- tarsus 1.3, tarsus 0.7. Second patella and tibia 1.6 mm, third 1.3. Fourth femur 2.4 mm, patella and tibia 2.1, metatarsus 1.8, tarsus 0.7. Male holotype. Coloration as in other species but legs black. Abdomen elongate as in P. intus (Fig. 61). Total length 3.2 mm. Carapace 1.5 mm long, 1.1 wide, 0.6 behind lateral eyes. First femur 1.4 mm, patella and tibia 1.3, metatarsus 1.0, tarsus 0.6. Second patella and tibia 1.1 mm, third 0.8. Fourth femur 1.6 mm, patella and tibia 1.4, metatarsus 1.2, tarsus 0.6. Note. Males and females were collect- ed at the same locality. Variation. Total length of females 4.1 to 4.4 mm. The illustrations (Figs. 83-87) were made from the holotype and para- type. Diagnosis. Pronous peje differs from other species found in Costa Rica by the large circular margins of the epigynum and the presence of a bordered median longitudinal groove with parallel sides (Fig. 83). The embolus of the male encloses a small space with a diameter smaller than the diameter of the embolus at its widest (Fig. 87). Natural History. A female from Limon Province was collected in a wet forest. An- other from La Selva was collected with six unwrapped eggs. Specimens Examined. COSTA RICA Heredia: La Selva, nr. Puerto Viejo, Feb. 1986, 29 paratypes (W. Eberhard, MCZ). Liynoii: Cerro Tortuguero, Tortu- guero, 110 m, 6 Jan. 1986, 19 (J. Coddington, USNM). PANAMA Chiriqui: La Fortuna, 5 Apr. 1984, 29 (W. Eberhard, MCZ). Pronous colon new species Figures 88-92; Map 4A Holotype. Female holotype from near Villa Colon, San Jose Prov., Costa Rica, Nov. 1990 (W. Eber- hard), in MCZ. The specific name is a noun in apposition after the locality. Description. Female holotype. Total length 4.8 mm. Carapace 2.1 mm long, 1.7 wide, 1.1 wide behind lateral eyes. First femur 2.3 mm, patella and tibia 2.3, meta- tarsus 1.8, tarsus 0.9. Second patella and tibia 2.1 mm, third 1.6. Fourth femur 2.7 mm, patella and tibia 2.7, metatarsus 2.1, tarsus 1.0. Diagnosis. Pronous colon differs from P. golfito (Fig. 93) and P. peje (Fig. 83) by the wide overhangs and almost circular openings on each side of the anterior lip as seen in the ventral view of the epigynum (Fig. 88). In posterior view, the epigynum of this species differs from others by hav- ing the median plate constricted by a shal- low lobe of the lateral plates (Figs. 90, 91). Pronous golfito new species Figures 93-97; Map 4A Holotype. Male from 3 km northeast of Golfito, 100 m, collected on and under logs, Puntarenas Prov., Costa Rica, 22, 23 May 1987 (D. Ubick), in CAS. The specific name is a noun in apposition after the locality. Description. Female paratype. Total length 5.0 mm. Carapace 2.1 mm long, 1.7 wide, 0.8 behind lateral eyes. First femur 2.1 mm, patella and tibia 2.1, metatarsus 1.5, tarsus 0.7. Second patella and tibia 1.8 mm, third 1.4. Fourth femur 2.7 mm, pa- tella and tibia 2.5, metatarsus 2.0, tarsus 0.8. Male holotype. Abdomen elongate as in P. intus (Fig. 61). Total length 3.8 mm. Carapace 1.7 mm long, 1.3 wide, 0.6 be- hind lateral eyes. First femur 1.9 mm, pa- tella and tibia 1.8, metatarsus 1.3, tarsus 178 Bulletin Museum of Comparative Zoology, Vol. 154, No. 3 0.7. Second patella and tibia 1.4 mm, third 1.1. Fourth femur 2.0 mm, patella and tibia 1.9, metatarsus 1.6, tarsus 0.7. Note. Males and females were collect- ed in the same area. Variation. Total length of females 4.8 to 5.0 mm. The illustrations (Figs. 93-97) were made from the holotype and para- type. Diagnosis. The female differs from P. wixoides by having only one pair of dark patches between seminal receptacles and anterior margin of the epigynum (Fig. 93) and by having the ventroposterior open- ings narrowly oval (Fig. 95). In posterior view it differs by the wide median notch (Fig. 96). The male differs from all others by the small nick on the "upper right" corner of the median apophysis of the left palpus (Fig. 97). Natural History. Specimens were col- lected in forest vegetation. Paratypes. COSTA RICA Puntarenas: Golfito, 24 July 1981, 2$ paratypes (G. B. Edwards, FSCA); 24 July 1981, 19 (B. K.' Dozier, FSCA). Specimens Examined. COSTA RICA Puntaren- as: Osa Peninsula, 4 km SW Rincon, 8-12 Mar. 1967, 19 (Organiz. for Tropical Studies, MCZ). Pronous wixoides (Chamberlin and Ivie) new combination Figures 55, 98-102; Map 4F Pronous beatus: — Banks, 1929: 96 (misidentifica- tion). Zigana wixoides Chamberlin and Ivie, 1936: 53, pi. 16, figs. 137, 138, 9. A female lectotype, here des- ignated, and four female and one immature male paralectotypes (all type specimens in poor condi- tion), from Barro Colorado Island [Lago Gatun, Panama Prov.], Panama, in AMNH, exannined. Roewer, 1942: 883. Bonnet, 1959: 4962. Description. Female from Barro Col- orado Island, Panama. Total length 3.7 mm. Carapace 1.6 mm long, 1.4 wide, 0.9 behind lateral eyes. First femur 1.7 mm, patella and tibia 1.8, metatarsus 1.3, tarsus 0.6. Second patella and tibia 1.5 mm, third 1.1. Fourth femur 2.1 mm, patella and tibia 2.0, metatarsus 1.7, tarsus 0.7. Male from Barro Colorado Island, Pan- ama. Abdomen relatively short (Fig. 62). Total length 3.6 mm. Carapace 1.6 mm long, 1 .3 wide, eyes 0.6 behind lateral eyes. First femur 1.7 mm, patella and tibia 1.6, metatarsus 1.4, tarsus 0.7. Second patella and tibia 1.3 mm, third 1.0. Fourth femur 1.8 mm, patella and tibia 1.8, metatarsus 1.5, tarsus 0.6. Note. The sexes were matched on the basis of the connecting ducts in the epi- gynum of the female (Fig. 99), which cor- respond to the long embolus of the male. These ducts are longer in P. wixoides than in P. intus and P. shanus. The distribution of male specimens from Panama to Ec- uador matches that of the females (Map 4F). Variation. One female individual of doubtful determination from Serra Nueva Granada, northern Colombia, has the ab- domen almost completely black. Total length of females 3.6 to 5.0 mm, males 3.4 to 4.1. The illustrations (Figs. 98-102) were Figures 79-82. Pronous lancetilla n. sp., female epigynum. 79, ventral. 80, ventral, cleared. 81, ventroposterior. 82, posterior. Figures 83-87. P. pe/'e n. sp. 83-86, female epigynum. 83, ventral. 84, ventral, cleared. 85, posterior. 86, lateral. 87, male left palpus. Figures 88-92. P. colon n. sp. 88-91 , female epigynum. 88, ventral. 89, ventral, cleared. 90, ventroposterior. 91 , posterior. 92, lateral. Figures 93-97. P. golfito n. sp. 93-96, female epigynum. 93, ventral. 94, ventral, cleared. 95, ventroposterior. 96, posterior. 97, male palpus. Figures 98-102. P. wixoides (Cfiamberlin and Ivie). 98-101, female epigynum. 98, ventral. 99, ventral, cleared. 100, ventro- posterior. 101, posterior. 102, male palpus. AcTiNOSOMA, Spilasma, Micrepeira, Pronovs • Levi 179 Figures 103-111. P. intus n. sp. 103-110, female epigynum. 103, 107, ventral. 104, 108, ventral, cleared. 105, 109, ventro- posterior. 106, 110, posterior. Ill, male palpus. 103-106, 111, (Panama). 107-110, (Ecuador). Scale lines. 0.1 mm. 180 Bulletin Museum of Comparative Zoology, Vol. 154, No. 3 made from specimens from Barro Colo- rado Island, Panama. Diagnosis. This species is slightly smaller than P. intus and P. shanus. The female Pronous wixoides differs from re- lated Central American species by having four (rather than two) dark patches on the anterior of the epigynum. The anterior pair of patches are in the shape of a tranverse hne (Fig. 98). The internal genitalia differ by having the duct loops facing each other in the median (Fig. 99). The female is dif- ficult to separate from P. intus, which has an anterior lip showing in the median (Figs. 103, 107); P. wixoides does not (Fig. 98). The epigynum may have to be cleared for diagnosis. In contrast, the male differs from all other species by having a much longer filamentous embolus (Fig. 102). Natural History. Specimens were col- lected in second growth forest near Bue- naventura, Colombia. Distribution. Panama, Colombia to southern Ecuador (Map 4F). Specimens Examined. PANAMA Code: El Valle, Jan. 1936, 29 (J. A. Griswold, MCZ); July 1936, U (A. M. Chickering, MCZ); Oct. 1954, 19 (W. E. Lundy, AMNH). Colon: Fort Davis, 14 Aug. 1936, 12 (A. M. Chickering, MCZ); Porto Bello [Portobelo], 12 Aug. 1936, 19 (A. M. Chickering, MCZ), Panama: Barro Colorado Island, very common. COLOMBIA Mag- dalena: Serra Nueva Granada, S. N. de Santa Marta, 1,311 m, 24 May 1975, 19 (doubtful determ., J. A. Kochalka, JAK). Meta: ca. 15 km SW Puerto Lopez, Hacienda Mozambique, 200 m, 19 (W. Eberhard 1748, MCZ). Valle: Hydroelectric Dam on Rio Anchicaya, 400 m, 1978, 15 (W. Eberhard, MCZ); no date, 19 (W. Eberhard, 851, MCZ); 1975, 19 (W. Eberhard, MCZ); 28 km E Buenaventura, 20 Jan, 1970, 19 (W. Eberhard 247, MCZ), Narino: nr, Barbacoas, 20 m, 20 Mar, 1974, 19 (W, Eberhard 750, MCZ), ECUA- DOR "Peru, Palmal" [Palmal, Ecuador, see Palmales, 03°41'S, 80°00'W, El Oro Prov,, 93 m (Stephens and Traylor, 1983)], \6 (K, Jelski, J. Sztolcman, PAN), Pronous intus new species Figures 53, 54, 56, 58-61, 64, 65, 1 0S- Ill; Map4B Holotype. Male holotype and three female para- types from Barro Colorado Island, Lago Gatun, Prov, Panama, Panama, June 1950 (A. M. Chick- ering), in MCZ. The specific name is an arbitrary combination of letters. Description. Female paratype. Total length 4.7 mm. Carapace 1.9 mm long, 1.6 wide, 1.1 behind lateral eyes. First femur 1.9 mm, patella and tibia 2.1, metatarsus 1.5, tarsus 0.7. Second patella and tibia 1.8 mm, third 1.3. Fourth femur 2.5 mm, pa- tella and tibia 2.5, metatarsus 1.8, tarsus 0.7. Male holotype. Abdomen elongate (Fig. 61). Total length 4.0 mm. Carapace 1.9 mm long, 1.5 wide, 0.7 behind lateral eyes. First femur 1.8 mm, patella and tibia 1.9, metatarsus 1.4, tarsus 0.7. Second patella and tibia 1.4 mm, third 1.2. Fourth femur 2.0, patella and tibia 1.9, metatarsus 1.7, tarsus 0.7. Note. Males and females were matched on the basis of their distributions: both have been found from Costa Rica to northern South America. Also, clearing the epigyn- um of P. intus reveals a shorter duct than in P. wixoides, corresponding to a shorter embolus in the male palpus. Variation. The abdomen of a male from Trinidad is shorter than that of other spec- imens. Total length of females 4.6 to 5.6 mm, males 3.5 to 4.7. Figures 53, 54, 56, 58-61, 103-106, and 111 were made from specimens from Barro Colorado Island, Panama, and Figures 107-110 from Ecu- adorean specimens. Diagnosis. In ventral view the open- ings of the epigynum have a wider anterior sclerotized lip at the median (Figs. 103, 107) than has P. wixoides (Fig. 98), and when cleared, the connecting ducts appear shorter (Figs. 104, 108) than those of P. wixoides (Fig. 99). The male palpus has a median apophysis (Fig. Ill) as in P. wix- oides but, unlike P. wixoides (Fig. 102), has a shorter, thicker embolus, lightly col- ored on the inside (Fig. 111). Natural History. Specimens have been found in premontane forest and by sweep- ing meadows in Curumani, Colombia. A male was collected in a pitfall trap near Fortuna, Panama. A male from Luepa, Venezuela, came from a Malaise trap in cloud forest. Distribution. Costa Rica to Venezuela ACTiNOSOMA, Spilasma, Micrepeira, Pronous • Levi 181 and Ecuador, northern Amazonas State, Brazil (Map 4B). Paratypes. PANAMA Panama: Barro Colorado Island, 16-20 July 1924, 15 (N. Banks); July 1934, 69; Jan. 1936. 49; June 1936; 1 9; July 1936; \$ (USNM ex MCZ); June 1939, 19, 1<5; July 1939, 1<5; Aug. 1939, 59; Sept. 1939, 29, 26; July 1950, 179, 2<5; Aug. 1950, 39; Julv 1954, 99, 2(5; Aug. 1954, 13; Jan. 1958, 89, 53; Feb. 1958. 163, 369 (MCZ, 13 in AMNH ex MCZ) (all A. M. Chickering, MCZ); 29 (K. W. Cooper, AMNH); Nov. -Dec. 1939, 19 (G. C. Wood, AMNH). Specimens Examined. COSTA RICA Limon: Ba- taan [Batan], 16 June 1951, 19 (O. L. Cartwright, USNM). Puntarenas: Golfito, 16 July 1981, 13 (B. K. Dozier, FSCA); Osa Peninsula, Sirena, 10 m, Feb. 1984, 13 (W. Eberhard, MCZ). PANAMA Bocas del Toro: Fortuna, Chiriqui Grande Road, 8°47'N, 82''11'W, 16-18 July 1987, 13 (D. Olson, MCZ). Chi- riqui: David, 26 Nov. 1975, 19 (D. Quintero, MCZ). Code: Valle, July 1936, 13 (A. M. Chickering, MCZ). Panama: Summit, Sept. 1946, 19 (N. L. H. Krauss, AMNH); Gamboa, Jan. 1958, 19, 13 (A. M. Chick- ering, MCZ); Forest Reserve, 28 July 1954, 13 (A. M. Chickering, MCZ). Darien: Villa Darien, 12-18 Feb. 1984, 13 (M. N. Garcia, MIUP). WEST INDIES Trin- idad: nr. Port of Spain, 1913, 23 (R. Thaxter, MCZ). VENEZUELA Monagas: Caripito, Mar. 1942, 19 (W. Beebe, AMNH). Bolivar: 10 km N Luepa, Gran Sa- bana, 26 June-11 July 1987, 1,500 m, 13 (S., J. Peck, AMNH). Distrito Federal: "La Moka" [probably a villa near Caracas; H. Enghoff, personal communi- cation], 1891, 19 (ZMK). COLOMBIA Bolivar: Car- tagena, 21 Dec. 1964, 19 (P. R. Craig, CAS). Cesar: Curumani, 22 July 1968, 13 (B. Malkin, AMNH). Meta: ca. 20 km N Rio Muco, ca. 20 km S El Porvenir, Finca Chenevo, 170 m, no date, 13 (W. Eberhard, MCZ); 15 km SW Puerto Lopez, 200 m, 1972, 19 (W. Eberhard, MCZ). Antioquia: Cancheras, Mutata, July 1963, 13 (P. B. Schneble, MCZ). Valle: 50 km S Bue- naventura, Mar. 1973, 29 (W. Eberhard, MCZ). EC- UADOR Sucumbios: bridge over Rio Cuyabeno, on road betw. Tarapoa & Tipishca, 00°01'S, 76°18'W, 25-30 June 1988, 19, 13 (W. Maddison, MCZ). BRA- ZIL Amazonas: Parque Nacional do Pico da Neblina, 5 Oct. 1990, 13 (A. A. Lise, MCP). Pronous shanus new species Figures 112-116; IVIap 4G Holotype. Male holotype and three female para- types from Barro Colorado Island, Gatun Lake, Panama Prov., Panama, June 1950 (A. M. Chick- ering), in MCZ. The specific name is an arbitrary combination of letters. Description. Female paratype. Total length 5.0 mm. Carapace 2.0 mm long, 1.8 wide, 0.9 behind lateral eyes. First femur 2.1 mm, patella and tibia 2.2, metatarsus 1.7, tarsus 0.8. Second patella and tibia 2.0 mm, third 1.4. Fourth femur 2.8 mm, pa- tella and tibia 2.6, metatarsus 2.1, tarsus 0.9. Male holotype. Abdomen elongate as in P. intus (Fig. 61). Total length 4.1 mm. Carapace 1.8 mm long, 1.4 wide, 0.6 be- hind lateral eyes. First femur 2.1 mm, pa- tella and tibia 2.0, metatarsus 1.5, tarsus 0.7. Second patella and tibia 1.5 mm, third 1.3. Fourth femur 2.2 mm, patella and tibia 2.1, metatarsus 1.8, tarsus 0.7. Note. Males and females were matched on the basis of similar geographical distri- bution and because both male and female genitalia differed strongly from genitalia of other species of Pronous. (Females have been collected with males of the other two Panamanian species, P. wixoides and P. intus.) Variation. Some females have the ven- ter and sides of the abdomen black. Total length of females 4.3 to 5.6 mm, males 3.8 to 4.8. The illustrations (Figs. 112-116) were made from the holotype and para- types. Diagnosis. Females are easily separat- ed from other species by the rebordered margin of the epigynum (Fig. 112) and, in ventroposterior view, by the transverse, oval median plate having a slight median, longitudinal ridge (Fig. 114). The male differs from the other two sympatric spe- cies, P. intus and P. wixoides, by having a short, thin embolus and the median apophysis with a narrow transverse spine as well as an outer spine at a right angle to the first (at 4 hr in Fig. 116). There is a narrow notch between the outer spine and the median apophysis, best seen in a more median view of the palpus. Distribution. Panama Canal area (Map 4G). Paratypes. PANAMA Panama: Barro Colorado Island, Aug. 1936, 13 (AMNH ex MCZ); June, 1 Aug. 1939, 29, 33; July 1950, 19, 13; Aug. 1950, 19 (USNM ex MCZ); Jan. 1958, 13 (all A. M. Chickering, MCZ or ex MCZ). Specimens Examined. PANAMA Code: El Valle, 182 Bulletin Museum of Comparative Zoologij, Vol. 154, No. 3 July 1936, 13 (A. M. Chickering, MCZ). Panama: Arraijan, Aug. 1936, IS (A. M. Chickering, MCZ); Forest Preserve, 29 Jan. 1958, 1<5 (A. M. Chickering, MCZ); Summit, Aug. 1950, 29, IS (A. M. Chickering, MCZ); nr. Gamboa, July 1981, IS (W. Eberhard, USNM ex MCZ); Julv 1984, 12 (W. Eberhard 2669, MCZ); Cocoli, 24 Nov. 1954, 12 (W, Lundy, AMNH); Farfan, 9 Jan. 1958, 12 (A. M. Chickering, USNM e.x MCZ); Fort Kobbe, 3 Aug, 1983, 1<5 (H., L. Levi, H. Stockwell, MCZ); Miraflores Locks, 3 Jan. 1958, 12 (A. M. Chickering, MCZ). Pronous valle new species Figures 117-121; Map 4G Holotype. Male holotype from near Saladito, 1,700 m, Depto. Valle, Colombia, Mar. 1976 (W. Eber- hard 1051), in MCZ. The specific name is a noun in apposition after the locality. Description. Female paratype. Abdo- men lacks an anterior median tubercle, other tubercles small. Total length 5.6 mm. Carapace 2.3 mm long, 1.9 wide, 1.1 be- hind lateral eyes. First femur 1.9 mm, pa- tella and tibia 2.0, metatarsus 1.5, tarsus 0.8. Second patella and tibia 1.8 mm, third 1.3. Fourth femur 2.5 mm, patella and tibia 2.4, metatarsus 1.9, tarsus 1.0. Male holotype. Abdomen length be- tween P. intus (Fig. 61) and P. tubercu- lifer (Fig. 62). Total length 3.8 mm. Car- apace 1.7 mm long, 1.4 wide, 0.7 behind lateral eyes. First femur 1.8 mm, patella and tibia 1.5, metatarsus 1.2, tarsus 0.7. Second patella and tibia 1.4 mm, third 1.1. Fourth femur 1.9 mm, patella and tibia 1.8, metatarsus 1.4, tarsus 0.7. Note. Males and females were matched on the basis of having been collected from the same locality. Variation. The male holotype is newly molted and the palpus was preserved when still soft. An older male might have other areas of the palpus sclerotized and darker than in the one illustrated (Fig. 121). Diagnosis. The female of P. valle dif- fers from females of P. pance and P. tub- erculifer species by having the angle of the epigynum margin on each side more obtuse (Fig. 117) and, in posterior view, the median plate widens dorsally (at 6 hr in Fig. 120). The male differs from other species by having a lightly colored lobe on the outside curvature of the embolus (in center of Fig. 121) and by having an an- nulate transverse spine on the median apophysis (Fig. 121). Paratype. COLOMBIA Valle: Saladito, 1,700 m, Apr. 1977, 12 (W. Eberhard 1162, MCZ). Pronous pance new species Figures 122-126; Map 4G Holotype. Male holotype and one male and five female paratypes from Rio Pance, near Cali, 1,100 m, Colombia, 23 Mar. 1970 (W. Eberhard 1-290), in MCZ. The specific name is a noun in apposition after the locality. Description. Female paratype. The sides of the carapace are dusky, and there is a median, dorsal gray line on the abdomen. Total length 4.5 mm. Carapace 2.1 mm long, 1.7 wide, 0.9 behind lateral eyes. First femur 1.9 mm, patella and tibia 1.8, meta- tarsus 1.3, tarsus 0.7. Second patella and tibia 1.7 mm, third 1.3. Fourth femur 2.1 mm, patella and tibia 2.1, metatarsus 1.7, tarsus 0.7. Male paratype. Abdomen shape be- tween P. intus (Fig. 61) and P. tubercu- lifer (Fig. 62). Total length 3.8 mm. Car- apace 1.7 mm long, 1.4 wide, 0.7 behind lateral eyes. First femur 1.7 mm, patella and tibia 1.7, metatarsus 1.3, tarsus 0.7. Figures 112-116. Pronous stianus n. sp. 112-115, female epigynum. 112, ventral. 113, ventral, cleared. 114, ventroposterior. 115, posterior. 116, male left palpus. Figures 117-121. P. valle n. sp. 117-120, female epigynum. 117, ventral. 118, ventral, cleared. 119, ventroposterior. 120, posterior. 121, male palpus. Figures 122-126. P. pance n. sp. 122-125, female epigynum. 122, ventral. 123, ventral, cleared. 124, ventroposterior. 125, posterior. 126, male palpus. AcTiNOSOMA, Spilasma, Micrepeira, Pronous • Levi 183 Figures 127-130. P. nighpes Capohacco, female epigynum. 127, ventral. 128, ventral, cleared. 129, ventroposterior. 130, posterior. Figures 131-139. P. tuberculifer n. sp. 131-138, female epigynum. 131, 135, ventral. 132-136, ventral, cleared. 133, 137, ventroposterior. 134-138, posterior. 139, male palpus. 131-134, 139, (Peru). 135-138, (Guyana). Scale lines. 0.1 mm. 184 Bulletin Museum of Comparative Zoology, Vol. 154, No. 3 Second patella and tibia 1.5 mm, third 1.1. Fourth femur 1.8 mm, patella and tibia 1.8, metatarsus 1.5, tarsus 0.5. Note. Males and females were collect- ed together. Variation. Total length of females 4.5 to 4.8 mm. Diagnosis. The female differs from other species by the semicircular openings and the wide anterior lip of the epigynum (Fig. 122). The male has a short filamen- tous embolus (Fig. 126), narrower than that of P. intus (Fig. Ill) and P. valle (Fig. 121) and a larger transverse spine on the median apophysis (Fig. 126) than in P. shanus (Fig. 116) and P. valle (Fig. 121). Paratype. COLOMBIA Valle: Rio Pance nr. Call, 1,100 m, 22 June 1970, 19 (W. Eberhard 290, MCZ). Specimens Examined. COLOMBIA Cauca: Rio Palace, Totoro, 1,800 m, oak forest, 24 Mar. 1967, 19, uncertain determination (R. Root, W. L. Brown, MCZ). Pronous nigripes Caporiacco Figures 127-130; Map 4G Pronous nigripes Caporiacco, 1947: 25; 1948: 662, figs. 71, 72, 9, S. Female lectotype, immature par- alectotype, designated by Levi (1985: 542), from Port Diamond (perhaps Great Diamond, or Dia- mond Plantation, both 6°43'N, 58°11'W), Guyana, in MZUF, examined. Male paralectotype is Mi- crathena acuta (Walckenaer) (Levi, 1985). Description. Female lectotype. Color- ation as in other species, except sternum light; coxae dark or brown, legs black. Car- apace slightly rugose. Total length 4.8 mm. Carapace 1.9 mm long, 1.6 wide, 0.9 be- hind lateral eyes. First femur 2.0 mm, pa- tella and tibia 2.0, metatarsus 1.5, tarsus 0.7. Second patella and tibia 1.8 mm, third 1.3. Fourth femur 2.7 mm, patella and tibia 2.4, metatarsus 2.1, tarsus 0.8. Variation. Total length of females 4.7 to 4.8 mm. The illustrations (Figs. 127- 130) were made from the lectotype. Diagnosis. The pair of black patches within the dusky patch of the epigynum (Fig. 127) separate this species from P. tuberculifer, P. intus, and other South American species. Unlike P. tuberculifer, P. nigripes has the openings in posterior view of the epigynum with their long axis transverse (Fig. 130). The epigynum lacks overhang above the openings (Fig. 127). Specimens Examined. GUYANA Tumatumari, 21 July 1936, 19 (MZUF). Pronous tuberculifer Keyserling Plate 2, Figures 57, 62, 131-139; Map 4D Pronous tuberculifer Keyserling, 1881: 548, pi. 16, fig. 1, 9, S. Male lectotype and two female para- lectotypes from Amable Maria [Depto. Junin], Peru, in PAN, examined; Keyserling, 1892: 35, pi. 2, fig. 31, 9, 6. Note. Simon (1895: 863, figs. 922-924), Bonnet (1958: 3778) and Roewer (1942: 967) erroneously considered all American specimens, and all citations of the genus Pronous, to refer to P. tuberculifer. Description. Female from Tingo Mar- ia, Peru. Abdomen lacking anterior me- dian tubercle; other tubercles small (Fig. 57). Total length 5.4 mm. Carapace 2.2 mm long, 1.8 wide, 1 .0 behind lateral eyes. First femur 1.9 mm, patella and tibia 2.0, metatarsus 1.5, tarsus 0.7. Second patella and tibia 1.8 mm, third 1.3. Fourth femur 2.5 mm, patella and tibia 2.3, metatarsus 1.9, tarsus 0.8. Male from Tingo Maria, Peru. Abdo- men as in Figure 62. Total length 3.7 mm. Carapace 1.8 mm long, 1.5 wide, 0.7 be- hind lateral eyes. First femur 1.7 mm, pa- tella and tibia 1.7, metatarsus 1.3, tarsus 0.7. Second patella and tibia 1.3 mm, third 1.1. Fourth femur 2.0 mm, patella and tibia 1.8, metatarsus 1.7, tarsus 0.7. Note. Males and females were paired in the original description, and both be- long to the only species of Pronous col- lected so far in Peru. Variation. Total length of females 3.9 to 5.4 mm, males 3.3 to 4.6. The female illustrated (Figs. 131-134) came from Tin- go Maria, Peru, the male (Fig. 139) from the Depto. Amazonas, Peru, and Figures 135-138 from Guyana. Diagnosis. The epigynum differs from that of other Pronous by having a V-shaped I AcTiNOSOMA, Spilasma, Micrepeira, Pronous • Levi 185 dusky patch and by the lateral overhang covering the openings (Figs. 131, 135). In posterior view the longest diameter of the opening is longitudinal (Figs. 131, 135); in others it is transverse (Fig. 130). The em- bolus is almost semicircular (it may be transparent; Fig. 139); a similar embolus is found only in the Mexican P. beatus. Also, unlike all other species, P. tubercu- lifer has the conductor with a small cylin- drical projection directed to the right (in a left palpus pointing toward 3 hr in Fig. 139) and a short tegulum lobe (Fig. 139). Natural History. The male from Alto Rio Comaina came from secondary veg- etation at the border of a swamp. Distribution. Guyana, Amazon area to Misiones Prov., Argentina. Male specimens (readily determined) have been found in Minas Gerais, Brazil, and in Paraguay and Bolivia (Map 4D). Specimens Examined. GUYANA Mackenzie [06°00'N, 58°ITW], 10 km below Three Friends on Demerara River, 55 m el. (Stephens and Traylor, 1985), Sept. 1931, 12 (MZUF). COLOMBIA Putu- mayo: Rio Putumayo, nr. Puerto Asis, no date, 19 (W. Eberhard 434, MCZ). ECUADOR Sucumbios: bridge over Rio Cuyabeno on road betw. Tarapoa & Tipish- ca, O'Ol'S, TenS'W, 8-9 Aug. 1988, 19 (W. Maddison, MCZ); Reserva Faunistica Cuyabeno, Laguna Gran- de, Sendero La Hormiga, 00°00', 76°10'W, 2-5 Aug. 1988, 29 (W. Maddison, MCZ). Napo: Pampeya, Rio Napo, May 1965, 29 (L. Pena, MCZ). Pichincha: Cay- ambe, 2,300 m, June 1965, 1<5 (L. Pefia, MCZ). Za- mora Chinchipe: Rio Jumbue, 1 June 1965, 1<5 (L. Pena, MCZ). PERU Amazonas: Alto Rio Comaina, Puesto de Vigilancia, 22, Falso Paquisha, 850-1,150 m, 21 Oct.-3 Nov. 1987, 1<5 (D. Silva D. MUSM). Hudnuco: Cucharas, Huallaga Valley, Feb.-Apr. 1954, 19 (F. Woytkowski, CAS); Tingo Maria, 26 May 1947, 19, 1$ (J, C. Pallister, AMNH), 2 June 1967, 19 (A. F. Archer, S. Risco, AMNH); Monzon Valley nr. Tingo Maria, 10 Nov. 1954, 29 (E. S. Ross, E. I. Schlinger, CAS). BRAZIL Amazonas: Reserva Porto Alegre, 80 km N Manaus, 27 May 1992, 19 (H. G. Fowler, MCZ). Paraiba: Independencia, 1911, 19 (W. M. Mann, MCZ). Pernambuco: Pernambuco [Recife], (SMF). Minas Gerais: Mina Serinha, Diamantina, Dec. 1944, 16 (E. Cohn, AMNH). Parana: Cataratas do Igua?u, 18-25 Mar. 1985, 29 (H., L. Levi, MCZ). PARAUGAY Con- cepcion: Apa, Jan.-Feb. 1909, U (AMNH). Cordil- lera: San Bernardino, 19 (E. Reimoser, MCZ). BO- LIVIA El Beni: Estac. Biol. Beni, 14°47'S, 65°15"W, 225 m, 8-14 Nov. 1989, 15 (J. Coddington et al, USNM). ARGENTINA Misiones: Montecarlo, Jan. 1966, 19 (M. E. Galiano, MEG). Spilasma Simon Spilasma Simon, 1895; 794, fig. 856, 3 abdomen, fig. 857 9 eye region. Type species Spilasma artifex Simon, 1895 (=S. duodecimguttata (Keyserling, 1880)] by original designation and monotypy. The gender of the name is feminine (Bonnet, 1958: 4121). Diagnosis. Spilasma has a narrow ce- phalic region of the carapace (Figs. 146, 154) as in Cyclosa and Acacesia. Spilasma differs from Cyclosa by having few setae on the carapace and lacking the dark rings on the legs (Fig. 145). Spilasma differs from both Cyclosa and Acacesia by having an oval abdomen, widest in the middle, with six dorsal pairs of white patches, and dif- fers from many araneids by having no markings in the midline of the abdomen and no folium pattern (Fig. 145). Description. Cephalothorax glabrous, orange to orange-brown without marks. Abdomen oval, widest in middle, with dor- sal paired patches of white (Fig. 145) and no ventral markings. Posterior median eyes subequal to anterior medians and their di- ameter or less apart. The epigynum has a ventrally extending lobe (Figs. 140, 142) with posterior sculpturing (Fig. 143). Male smaller than female, with ventral sclerotized area from sides of pedicel to genital groove (Fig. 153). Endite without tooth, no coxal hook. Palpal patella with two weak setae. Second tibia thinner than first. Relationship. Spilasma can be placed with the Alpaida group of species, on the basis of the following characters: the epi- gynum is lobe-shaped (Figs. 140-144); the tegulum is free in the upper right corner of the left palpus (T in Fig. 151); the con- ductor is in the middle of the tegulum, not the upper right in the left palpus (C in Fig. 151); the median apophysis is without spines or flagella (except distally) and pro- jects beyond bulb of palpus (M in Fig. 151); in the upper left of the conductor (in the left palpus) there is a pocket (C in Fig. 161), which may be a homolog with the paramedian apophysis. The narrow ce- phalic region of the carapace is probably a synapomorphy with Acacesia Simon 186 Bulletin Museum of Comparative Zoology, Vol. 154, No. 3 ^^^^^^^ ^'fV llfi ^EHH^t^^^^^S^^^^'^" i^^n^ IS wSmi B t^i^i^BSriWmmm ^^^^5c^ ' 1^ ■ •^'■'la ■ Plate 3. Spilasma duodecimguttata. Web diameter of upper web about 25 cm, lower one 10.4 cm {upper photos, R. L. C. Baptista; lower photo, J. A. Coddington). (Glueck, 1994) and perhaps also with Cy- closa. Natural History. The web is horizon- tal, pulled up at the hub, with sticky threads (W. Eberhard, personal communication) and, above the hub, a median granular retreat, with sand grains, having a lateral flap (Pi. 2). The flap pulls in when dis- turbed. The retreat is also used to house the eggsacs (Lubin, 1978). Distribution. There are three Ameri- can species, with only one specimen each available for two of them, from Honduras to Rio de Janeiro State, Brazil. Spilasma africana Simon, 1903, from Spanish Guin- ea, is the only described species not from America. The specimens of S. africana are lost, and the species was synonymized with Prasonica seriata Simon, 1895, by Gras- shoff (1971). Prasonica is a genus related to M angora. Key to Spilasma Species The female of S. utaca and the male of S. baptistai are not known. Males Females 2(1). 3(1). A pair of deep longitudinal grooves on car- apace (Fig. 161); base of embolus with duct making two loops (Fig. 160); Peru (Map 5) utaca Carapace without grooves (Fig. 152); em- bolus duct with one loop (Figs. 149-151); from Honduras to Bolivia and southern Brazil (Map 5) duodecimguttata In posterior view of epigynum, slit-shaped openings face laterally (Fig. 157); Per- nambuco, Brazil (Map 5) baptistai AcTiNOSOMA, Spilasma, Micrepeira, Pronous • Levi 187 D baptistai # duodecimguttata A utaca Map 5. Distribution of Spilasma species. In posterior view of epigynum, slit-shaped openings face anteromedially (Figs. 141, 143); Honduras to Bolivia and south- eastern Brazil duodecimguttata Spilasma duodecimguttata (Keyserling) Plate 3; Figures 140-155; Map 5 Singa duodecimguttata Keyserling, 1880: 302, pi. 4, fig. 6, 9. Female holotype from New Granada [old name for Colombia], in BMNH, examined. Key- serling, 1893: 286, pi. 14, fig. 211, 9. Roewer, 1942: 877. Epeira lamentaria Keyserling, 1883: 199, pi. 15, (?fig. 5); 1892: 174, pi. 8, (?fig. 128). Female holotype from Amazon Prov., Brazil at HECO, examined. NEW SYNONYMY. Spilasma tredecimguttata Simon, 1895: 789, figs. 856, 857, 6 abdomen, 9 eyes; 1897: 476. Many male and female syntypes from the Amazon in MNHN no. 1010, examined. Roewer, 1942: 776. Bonnet, 1958: 4121. NEW SYNONYMY. Spilasma artifex Simon, 1895: 794; 1897: 477, pi. 12, figs. 1-5, pi. 13, fig. 1, web. Female holotype from San Esteban, Venezuela, MNHN no. 10127, ex- amined. Roewer, 1942: 776. Bonnet, 1958: 4121. NEW SYNONYMY. Epeira davisi Hingston, 1932: 365, fig. 40, web. Spec- imens from Guyana, in BMNH, lost. NEW SYN- ONYMY. Aranea lamentaria: — Roewer, 1942: 845. Araneus duodecimguttatus: — Bonnet, 1955: 449. Araneus lamentarius: — Bonnet: 1955: 526. Spilasma coccineum Caporiacco, 1955: 344, fig. 29, (3. Male holotype from Rancho Grande, Venezuela, in UCVC, not examined. Brignoli, 1983: 280. NEW SYNONYMY. Note. The small illustration of Keyser- ling (1883, fig. 5; and the same in 1892, fig. 128) does not look like the epigynum of this species. Keyserling may have illus- trated an artifact or debris. Caporiacco's (1955) illustration of the male palpus of S. coccineum is good. Lopez (1985: 86) described a new spe- cies, Spilasma richei, collected along the path from Roura to Kaw, southeast of Cay- enne, French Guyana. He placed it in the Museum de Beziers, France, but it has been lost (Lopez, 1986, personal communica- tion). Few diagnostic characters are pro- vided for S. richei except for noting the difference in "la spinulation des pattes et 188 Bulletin Museum of Comparative Zoology, Vol. 154, No. 3 la form precise de I'epigyne", but a de- tailed description is forthcoming (Lopez, 1985: 87). This is considered a nomen nu- dum. It is also not listed in Platnick's cat- alogs (1989, 1993). The genitalia of S. duodecimguttata are quite variable. When many specimens from different locations are available, one readily sees the reason for the many syn- onymies cited here. Epeira davisi was erroneously placed in Cyrtophora by Levi (1991b: 179). Description. Female from Depto. Meta, Colombia. Carapace, chelicerae, labium, endites, sternum, legs, bright orange. Dor- sum of abdomen orange, with a gray cast, with six pairs of symmetrical white patches (Fig. 145); venter darker gray. Posterior median eyes same diameter as anterior medians, laterals 0.7 diameter. Anterior median eyes 0.3 diameter apart, 0.8 di- ameter from laterals. Posterior median eyes 0.5 diameter apart, 1.1 diameters from lat- erals. Ocular rectangle, slightly longer than wide. Height of clypeus equals 0.4 diam- eter of anterior median eye. Sternum slightly wider than long. Abdomen oval, widest anterior of middle (Fig. 145). Total length 4.5 mm. Carapace 2.0 mm long, 1.9 wide, 0.8 behind lateral eyes. First femur 2.3 mm, patella and tibia 2.3, metatarsus 1.3, tarsus 1.0. Second patella and tibia 2.0 mm, third 1.2, fourth 2.0. Male from Depto. Meta, Colombia. Col- or as in female. Distal leg articles darkest. Posterior median eyes 0.7 diameter of an- terior medians, laterals 0.5 diameter. An- terior median eyes 0.2 diameter apart, 0.6 diameter from laterals. Posterior median eyes 0.7 diameter apart, 1.2 diameters from laterals. Height of clypeus equals the di- ameter of an anterior median eye. Abdo- men widest in middle, with a soft, but sclerotized shield in the area anterior to the genital groove (Fig. 153). Total length 3.2 mm. Carapace 1.6 mm long, 1.3 wide, 0.5 behind lateral eyes. First femur 1.6 mm, patella and tibia 1.6, metatarsus 1.0, tarsus 0.7. Second patella and tibia 1.4 mm, third 0.9, fourth 1.3. Note. Males and females have been col- lected together. Variation. The genitalia of no two specimens look quite alike, and each new specimen makes one think they have a new species. Most variation is in the protuber- ance of the epigynum: some having a semi- circular keel (Figs. 140-142), some not (Figs. 143, 144). The male palpus also var- ies in structure. The tip of the median apophysis and the curvature of the em- bolus are unusually variable (Figs. 149, 150). Some individuals lack white patches on the abdomen. Total length of females 3.0 to 7.0 mm, males 2.2 to 3.6. Most il- lustrations were made from specimens from Depto. Meta, Colombia, but Figures 143-145 and 150 were made from speci- mens from 80 km north of Manaus, Brazil. Diagnosis. The first femur and patella- tibia are of about the same length, unlike those of most other araneids. The epigyn- um differs from S. baptistai by having the slit openings face anteriomedially (Figs. 141, 143); the male differs from S. utaca by lacking the pair of grooves on the car- apace (Fig. 161). The male palpus super- ficially resembles that of Ocrepeira yaelae Levi (1993a, fig. 359). Natural History. Spilasma duodecim- guttata is found in foliage of wet forest in Guapiles, Costa Rica; in lowland forest. El Dorado, Venezuela; in rain forest near Le- ticia, Colombia; in secondary forest by a lake 15 km southwest of Puerto Lopez, Colombia; in forest interior in reserves Figures 140-155. Spilasma duodecimguttata (KeyserWng). 140-148, female. 140-144, epigynum. 140, 143, ventral. 141, pos- terior. 142, 144, lateral. 145, dorsal. 146, carapace. 147, carapace and cheliera. 148, eye region and chelicerae. 149-155, male. 149-151, left palpus. 149, 15C, mesal. 151, pulled apart. 149, 151, (Depto. Meta, Colombia). 151, (N of Manaus, Brazil). 152, dorsal. 153, abdomen ventral. 154, carapace. 155, eye region, chelicerae and right palpus. Figures 156-159. S. baptistai, female. 156-158, epigynum. 156, ventral. 157, posterior. 158, lateral. 159, dorsal. ACTINOSOMA, Spilasma, Micrepeira, Pronous • Levi 189 Figures 160, 161. S. utaca, male. 160, palpus. 161, dorsal. Abbreviations. C, conductor. E, embolus. M, median apophysis. R, radix. T, tegulum. Scale lines. 1 mm, genitalia 0.1 mm, except Figures 146-148, 153-155, 0.5 mm. 190 Bulletin Museum of Comparative Zoology, Vol. 154, No. 3 north of Manaus, Brazil. In Rio de Janeiro, Brazil, a web was found to be 40 cm above ground. Distribution. Honduras to Rio de Ja- neiro State, Brazil, and Bolivia (Map 5). Specimens Examined. HONDURAS Atlantida: Lancetilla, 9 (MCZ). COSTA RICA Limon: 5.5 km E Guapiles, $ (DU). Heredia: Finca La Selva, 9, S (CAS, MCZ). Puntarenas: Reserva Carara nr. Tar- coles, 9 (MCZ). Cartago: Turrialba, S (AD). PANAMA Panama: Barro Colorado Island, 9, S (MCZ); Cerro Galero, S (MCZ), WEST INDIES Trinidad: Old Gold Mine, Arima Ward, 9 (USNM). VENEZUELA Ara- gua: Rancho Grande, <5 (AMNH), 9 (MCZ), 9, S (USNM). Bolivar: 22 km S El Dorado, S (AMNH). Falcon: 3 km S El Hondo de Uria, 15 km E Curi- magua, 9 (MCZ). GUYANA Shudikar River, Upper Essequibo River, $ (AMNH); Upper Essequibo River, Onoro Region, 9 (AMNH); Essequibo River opposite Twasinki Mtns., imm. (AMNH); Upper Essequibo River, nr. Akaramukra Rapids, 3 (AMNH). SURI- NAM Saramacca: Voltzberg Raleighvallen Nature Reserve, 9 (MCZ). FRENCH GUIANA nr. Camp Cai- man, Montagnes Kaw, 9 (USNM). COLOMBIA San- tander: Carare, Opon Capote 6°38'N, 73°55'W, 6 (MCZ). Meta: 15 km SW Puerto Lopez, 9, <5 (MCZ). Valle: 28 km E Buenaventura, <3 (MCZ); Central An- chicaya, 9 (MCZ). Putumayo: Rio Putumayo, nr. Pto. Asis, 9 (MCZ). Amazonas: Amacayacu, 48 km NW Leticia, S (MCZ); 18 km N Leticia, S (AMNH); Le- ticia, 9, 6 (CAS). ECUADOR Sucumbios: Reserva Faunistica Cuyabeno, 9 (MCZ, MECN). Napo: Rio Coca and Napo, <5 (MCZ); 48.6 km NE Baeza, S (MCZ). PERU Hudnuco: Panguana, Rio Pachitea, 9°37'S, 74°56'W, 9 (MCZ); Monzon Valley, Tingo Maria 9, <5 (CAS); Dantas La Molina, Quebrada Sapote, SW Puerto Inca, 270 m, 9°38'S, 75°00'W, 19 (MUSM). Madre de Dios: Zona Reserva de Manu, 9 (USNM); 15 km E Puerto Maldonado, 6 (CAS); Res. Tambo- pata, 12°50'S, 69°17'W, 9, S (USNM). BRAZIL Ro- raima: Ilha de Maraca, 9, $ (INPA). Amazonas: Cabo Frio Reserve, 80 km N Manaus, 9, $ (INPA, MCN, MCZ); Colosso Reserve, 80 km N Manaus, 9, 3 (INPA, MCN, MCZ); Dimona Reserve, 80 km N Manaus, 9, & (MCZ); Gaviao Reserve, 80 km N Manaus, 3 (MCZ); Km 41 Reserve, 80 km N Manaus, 9, 6 (MCZ); Fa- zenda Esteio, Manaus, $ (INPA); Belem, periodically flooded forest, nr. confluence with Rio Solimoes, S (AMNH); Manaus, 9, $ (MACN); Reserva Porto Ale- gre, 80 km N Manaus, 9 (MCZ); Reserva Ducke nr. Manaus, 9 (MCN, INPA); Parque Nacional do Pico da Neblina, Maturaca, 9 (MCP). Pard: Caninde, 9 (AMNH). Bahia: Encruzilhada, $ (AMNH); Fazenda Matiapa, Camacan, $ (MCN). Mato Grosso: Barra de Tapirape, 3 (AMNH, MCZ); Chapada dos Guimaraes, 9 (AMNH); Sinop, S (AMNH); Villa Vera, 55°30'S, 12°46'W, 6 (AMNH). Rio de Janeiro: Mangaratiba, S (AMNH); Barra da Tijuca, 9 (RLCB); Morro da Urea, 9 (RLCB). BOLIVIA El Bent: 27 km SW Yucumo, 500 m, 15°50'S, eg-H'W, 6 (USNM). Spitasma baptistai new species Figures 156-159; Map 5 Holotype. Female from Dois Irmaos Reservation, Recife, Pernambuco State, Brazil, 25 Jan. 1989 (R. L. C. Baptista, A. P. Chaves 2654) in MZSP. The species is named after the collector. Description. Female holotype. Cara- pace, chelicerae, sternum, labium, and en- dites orange. Legs orange, distal articles darkest. Dorsum of abdomen with pairs of white patches (Fig. 159); venter gray. Eyes small. Posterior median eyes 0.8 diameter of anterior medians, laterals 0.8 diameter. Anterior median eyes 0.2 diameter apart, 1 diameter from laterals. Posterior median eyes 1 diameter apart, 1.3 diameters from laterals. Ocular quadrangle slightly longer than wide, slightly wider behind. Height of clypeus equals 0.5 diameter of the an- terior median eye. Abdomen oval (Fig. 159). Total length 3.0 mm. Carapace 1.40 mm long, 1.19 wide, 0.57 behind lateral eyes. First femur 1.30 mm, patella and tibia 1.32, metatarsus 0.67, tarsus 0.52. Sec- ond patella and tibia 1.19 mm, third 0.71, fourth 1.10. Diagnosis. The eyes are smaller than those of S. duodecimguttata (Fig. 159) and in posterior view of the epigynum, the me- dian plate (Fig. 157) is relatively wider than that of S. duodecimguttata (Figs. 141, 143), and the openings face laterally (Fig. 157). Natural History. The specimen was collected with its retreat. Vestiges of the web were projecting from refuge between leaves of a bush, 1.5 m above ground in secondary forest. No other specimens were found. Spitasma utaca new species Figures 160, 161; Map 5 Holotype. Male holotype from 1,600 to 2,200 m, Utcuyacu [above Merced, 1,465 m, 11°12'S, 75°28'W; Stephens and Traylor, 1983], Depto. Ju- nin, Peru, 4 Apr. 1948 (F. Woytkowski), in AMNH. The specific name is an arbitrary combination of letters. Description. Male holotype. Carapace dark orange-brown. Chelicerae, labium. AcTiNOSOMA, Spilasma, MiCREi'EiRA, Pronovs • Levi 191 endites, sternum dark brown. Coxae brown, except posterior half of fourth coxae, which are hght yellowish; legs dark brown. Dor- sum of abdomen black with two pairs of white spots, the first round, the second di- amond-shaped, and with transverse pos- terior dark bands (Fig. 161); venter black. Posterior median eyes 0.7 diameter of an- terior medians, anterior laterals 0.7 di- ameter, posterior 0.6. Anterior median eyes 0.2 diameter apart, 0.4 diameter from lat- erals. Posterior median eyes 0.4 diameter apart, 1.2 diameters from laterals. Ocular quadrangle narrower behind than in front. Height of clypeus equals 0.7 diameter of anterior median eye. Sternum corniculate. Abdomen oval, widest behind middle (Fig. 161). Total length 3.0 mm. Carapace 1.59 mm long, 1.19 wide, 0.53 wide behind lat- eral eyes. First femur 1.00 mm, patella and tibia 1.72, metatarsus 0.87, tarsus 0.69. Sec- ond patella and tibia 1.56 mm, third 0.93, fourth 1.45. Diagnosis. The carapace of S. utaca has a pair of grooves (Fig. 161) not present in S. duodecimguttata (Fig. 152), and the duct loops twice in the proximal end of the embolus (Fig. 160). No other specimens were found. Micrepeira Schenkel Micrepeira Schenkel, 1953: 26. Type species M. al- bomaculata Schenkel by monotypy. The gender of the name Micrepeira is feminine. Diagnosis. Micrepeira differs from most other araneid genera, including Ar- aneus and Singa by the domed sternum (Figs. 186, 191) and a subspherical abdo- men with bold, white dorsal markings: of- ten a wide median, longitudinal band and a wide transverse band or a few pairs of transverse patches, but no white markings on the venter (Figs. 165, 176, 185, 190, 195, 204). Also in Micrepeira, the femora are the same length as the adjacent patella and tibia; in other genera, the femora are shorter. Micrepeira resembles the Indo- pacific Anepsion Strand, 1929 (Chrysan- thus, 1969), but has a narrower cephalic region. It resembles the African Pherenice Thorell, 1899, but has a low clypeus, less than the diameter of the anterior median eye. Micrepeira differs from all similar genera by the following: the epigynum has a wide shelf, resembling that of Alpaida, with a tiny, soft, distal, pointed scape, eas- ily overlooked (Figs. 162, 192, 200), by the male palpus, which has a large median apophysis bearing two flagella (Figs. 168, 177, M in Fig. 179), and a conductor either on the edge or inward on the tegulum, and no terminal apophysis or embolus lamella (Figs. 179, 180). The median apophysis flagella are, as usual, associated with a pointed scape of the female epigynum. Description. Female. Cephalothorax glabrous, orange to brown without distinct marks, cephalic region wide (Figs. 166, 167). Eyes subequal. Median ocular quad- rangle almost square. Height of clypeus 0.6 to 0.8 diameter of the anterior median eye. Legs relatively short and thick (Fig. 204). Males much smaller than females (Fig. 181), width behind eye region less than half width of carapace. Endite with small tooth (M. fowleri) or none (M. hoeferi), no tubercle or tooth on palpal femur. No hook on first coxa, palpal patella with one macroseta. Second tibia as thick as or thin- ner than first, without any large macro- setae. Median apophysis of palpus has two proximal flagella associated with a pointed scape of the epigynum. Relationship. The broad lobe of the epigynum, the palpus with a bare tegulum (at 1-2 hr in Fig. 168), the lack of distal hematodocha, the lack of terminal apoph- ysis or embolus lamella, and the single macroseta on the palpal patella place the genus close to Alpaida. But the tiny, point- ed scape of the epigynum (Figs. 162-164) and the two flagella of the median apoph- ysis (Fig. 168, M in Fig. 179) suggest re- lationship with Kaira, Aculepeira, Ama- zonepeira, and Metepeira, all genera clos- er to Araneus. Although the pointed scape and flagella on the median apophysis were previously considered synapomorphies, evidence now indicates that this may be a homoplasious character, as the make-up of 192 Bulletin Museum of Comparative Zoology, Vol. 154, No. 3 Plate 4. Micrepeira hoeferi n. sp. Diameter of web 20 cm, hub of web 150 cm above forest floor (photo, H. H6fer). the genitalia of Micrepeira resembles so much more that of Alpaida. The position of the conductor (C in Figs. 179, 180, at 1 hr in Fig. 177 and at 11 hr in Fig. 178) is on the edge of the tegulum but is at the distal end and not above the median apophysis as in Araneus. Natural History. Webs are known of three species (M. hoeferi, M. tubulofa- ciens, and M. velso). All have the shape ACTiNOSOMA, Spilasma, MiCREPEiRA, Pronous • Levi 193 f \GUYANA (•k>'-rT>-~ FR. GUIANA D f o w I e r i \'>{^. I. ▲ pachitea \ PERU # tubulofaciensV^^ '[' V velso ^-c^^/f^ SURINAM FR. GUIANA n albomaculata # hoeferi Vsmithae BOLIVIA 'r-- ---\\ \ l'^ Map 6. Distribution of Micrepeira species. of the lower half of a circle hanging on a line with a retreat above the hub (Pi. 4). The retreat is made of detritus and silk and has a lid that can be closed (Hofer note on label with specimen of M. hoeferi). At its pointed end, the retreat has two strong silk support lines covered by some detritus. Distribution. All seven species are Cen- tral and South American (Map 6). Key to Micrepeira Species Males of M. albomaculata, M. pachitea, M. smi- thae, M. tubulofaciens, and M. velso are unknown. 1 . Male 2 Female 3 2(1). Middle of median apophysis with a lobe (Fig. 168); Amazonian Ecuador to Ma- naus region of Brazil (Map 6) fowleri Middle of median apophysis without lobe (M in Figs. 177-180); Guianas to Peru, Mato Grosso, Brazil (Map 6) hoeferi 194 Bulletin Museum of Comparative Zoology, Vol. 154, No. 3 3(1). Sternum with an anterior longitudinal groove (Fig. 191); Peruvian Amazon (Map 6) pachitea Sternum without such groove (Fig. 186) 4 4(3). Epigynum in ventral view a convex lobe (Fig. 200); Falcon State, Venezuela (Map 6) _ _ albomaculata Epigvnum flat (Figs. 162, 171, 182, 187, 192, 196) 5 5(4). Epigynum posterior median plate over- hanging laterals as in Figure 197; Costa Rica (Map 6) velso Posterior median plate otherwise; South America 6 6(5). Epigynum with posterior margin straight in ventral view and with narrow scape (Figs. 171, 174); Guianas, to Peru, and Mato Grosso, Brazil (Map 6) hoeferi - Epigynum more or less triangular in ven- tral view or a triangular scape (Figs. 162, 182, 1 92 ) 7 7(6). Triangle formed by epigynum taking up most of posterior margin of epigynum (Figs. 162, 182) 8 Triangle formed by epigynum making up less than half of posterior margin of epi- gynum (Fig. 192); Surinam (Map 6) smithae 8(7). Epigynal lobe marked as in Figure 182; Guianas (Map 6) tuhulojaciens Epigynum without dark marks (Fig. 162); Amazonian Ecuador to Manaus region of Brazil (Map 6) _ fowleri Micrepeira fowleri new species Figures 162-170; Map 6 Holotype. Female holotype from 80 km north of Manaus, Km 41 Reserve, forest interior, Amazonas State, Brazil, 14 Apr. 1991, 200 m (H. G. Fowler, E. V'enticinque, R. S. Vieira), in MCN no. 25536. The species is named after the collector. Description. Female holotype. Cara- pace orange. Chelicerae, labium, endites orange. Sternum dusky orange. Coxae or- ange; legs dusky orange, darker distally. Dorsum of abdomen dark dusky with a longitudinal, median band of white pig- ment patches and a pair of white patches; posterior black with a pair of small pos- terior patches (Fig. 165); venter dusky with a median darker band from epigynum to spinnerets. Posterior median eyes 0.8 di- ameter of anterior medians, laterals 0.6 diameter. Anterior median eyes 0.4 di- ameter apart, 1 diameter from laterals. Posterior median eyes 0.7 diameter apart, 1.5 diameters from laterals. Total length 3.0 mm. Carapace 1.36 mm long, 1.24 wide, 0.75 behind lateral eyes. First femur 1.18 mm, patella and tibia 1.18, metatarsus 0.62, tarsus 0.54. Second patella and tibia 1.09 mm, third 0.65, fourth 1.03. First tibia 0.22 mm thick. Male paratype. Color as in female ex- cept lacking anterior abdominal white pig- ment patches, having only the posterior pair. Posterior median eyes 0.8 diameter of anterior medians, laterals 0.6 diameter. Anterior median eyes 0.7 diameter apart, 0.7 diameter from laterals. Posterior me- dian eyes 0.8 diameter apart, 1.2 diameters from laterals. Endite with small tooth. To- tal length 1.8 mm. Carapace 1.07 mm long, 0.88 wide, 0.44 behind lateral eyes. First femur 0.91 mm, patella and tibia 0.81, metatarsus 0.49, tarsus 0.40. Second patella and tibia 0.72 mm, third 0.45, fourth 0.65. Note. Males and females were collect- ed together; both have the posterior of the abdomen black. Variation. Total length of females 2.4 to 3.5 mm, males 1.8 to 2.1. The anterior, median, transverse groove of the epigyn- um (center of Fig. 162) may be absent or different in width from that illustrated. Illustrations were made from specimens collected 80 km north of Manaus. Diagnosis. Micrepeira fowleri, unlike other species, has its abdomen black pos- teriorly, contrasting with the lighter an- terior three-quarters (Fig. 165). Females are smaller than M. hoeferi and are sep- arated from them by having a large tri- angular lobe of the epigynum with a cen- tral transverse groove or dark mark be- tween two darker areas (Fig. 162) and by having concave sides of the posterior me- dian plate (Fig. 163). The male differs from that of M. hoeferi by having a lobe in the middle of the median apophysis (Fig. 168). It is possible that all specimens considered M. fowleri are variants of M. tubulofa- ciens. Natural History. Specimens of M. fow- leri were collected in forest border and ACTINOSOMA, Spilasma, Micrepeira, Pronovs • Levi 195 Figures 1 62-1 70. Micrepeira fowleri n. sp. 1 62-1 67, female. 1 62-1 64, epigynum. 1 62, ventral. 1 63, posterior. 1 64, lateral. 1 65, dorsal. 1 66, carapace. 1 67, eye region and chelicerae. 1 68-1 70, male. 1 68, left palpus. 1 69, carapace. 1 70, eye region, chielicerae, and righit palpus. Figures 171-181. Micrepeira hoeferin. sp. 171-176, female. 171-175, epigynum. 171, 174, ventral. 172, 175, posterior. 173, lateral. 171-173, (Mato Grosso State). 174, 175, (Amazonas State). 176, dorsal. 177-181, male. 177-180, left palpus. 177, mesal. 178, ventral. 179, 180 pulled apart. 181, dorsal. Abbreviations. C, conductor. E, embolus. M, median apopfiysis. R, radix. T, tegulum. Scale lines. 1.0 mm, genitalia 0.1 mm, except Figures 166, 167, 169, 170, 0.5 mm. 196 Bulletin Museum of Comparative Zoology, Vol. 154, No. 3 forest interior. The retreat has parallel sides, about 4.0 to 4.5 mm wide and 10 to 20 mm long. Distribution. Amazonian Ecuador, Peru to Manaus Brazil (Map 6). Paratypes. BRAZIL Amazonas: Km 41 Reserve, 25 May 1991, 19 (H. G. Fowler, E. Venticinque, R. S. Vieira). Specimens Examined. ECUADOR Sucumhnos: Cuyabeno Reserve, Laguna Grande, 13 Feb. 1984, 12 (L. Aviles, MECN). PERU Madre de Dios: Zona Reservada Tambopata, 290 m, 12°50'S, 69''17"W, 14 June 1988, 12 (D. Silva D., MUSM). BRAZIL Ama- zonas: 80 km N Manaus, 2''24'S, 59'"52'W, 26 Feb. 1989, 6, 9 Nov. 1989, 1<5 (H. G. Fowler, MCZ); Cabo Frio Reserve, 80 km N Manaus, 1989, 1990, 52, 26; Colosso Reserve, 80 km N Manaus, 1989, 1990, 1012, 16serling, am) Meta- phidippus digitatus F. P. -Cambridge are newly syn- onymized with P. galathea; Dendrijphantes uteanus Chamberhn & Gertsch with P. aeneola; and Dendry- phantes mimus Chamberhn with P. furcata. Eu- ophrys concolor Banks is removed from synonymy with P. proterva and considered a senior synonym of Sittaciis cursor Barrows, yielding the new combina- tion Sitticus concolor. Identification keys are pre- sented for all Pelegrina males and for females from restricted geographical regions. All species are de- scribed and illustrated. Male/female associations were achieved for all species north of Mexico. Courtship behavior is described for 22 species of Pelegrina, karyotypes for 10 species, and habitat information for most species. The genus Pelegrina may be closely related to the Meiaphidippus mannii group, Nagairia and/or Eris. Nagaina incunda G. & E. Peckham is described and illustrated; Dendryphantes vegettts G. & E. Peck- ham, Meiaphidippus flavolineatus F. P. -Cambridge, and Meiaphidippus expallidatus F. P.-Cambridge are synonymized with N. incunda. The species of the mannii group (temporarily retained in Meiaphidip- pus) that occur in the United States are also described and illustrated; two new combinations, Meiaphidip- pus chera (Chamberlin) and Meiaphidippus car- meriensis (Chamberlin), are established; one species, Meiaphidippus emmilius, is described as new; Den- dryphanies versicolor G. & E. Peckham is synony- mized with Meiaphidippus mannii (G. & E. Peck- ham), and Meiaphidippus franciscanus Schenkel with Meiaphidippus diplacis (Chamberlin). INTRODUCTION For about 50 years after the Peckham's (1909) revision of the jumping spider spe- cies north of Mexico, taxonomic work on North American representatives of this large family consisted mostly of scattered species descriptions by Chamberlin, Gertsch, Ivie, and others. Some generic re- visions consolidating and clarifying the previous work began appearing in the 1950s (Gertsch and Ivie, 1955; Barnes, 1955, 1958), but most genera remained untouched, including the three largest genera, Habronattus* Phidippiis, and Meiaphidippus, which together include Authors of scientific names are given in the index. about half of the nearly 300 species of sal- ticids occurring north of Mexico (accord- ing to the count of Richman and Cutler, 1978). In the last three decades, increased interest in the family has resulted in re- visions of Habronattus (Griswold, 1987), Phidippus (Edwards, in preparation), and other genera (Proszynski, 1968, 1971a, 1973a, 1980; Cutler, 1981a, 1987; Rich- man, 1981, 1989). However, except for works by Kaston (1973) on some eastern species and by Cutler and Jennings (1985) on the arizonensis group, Metaphidippus has remained unre vised, perhaps because its poorly defined limits have made the scope of any revision potentially trouble- some. When 1 first began to revise Meta- phidippus, I knew that I would have to restrict the revision to only some of the disparate groups placed there. The largest group placed in Metaphidippus, including the species most commonly collected in northern and eastern North America, was chosen for revision and is here moved to the genus Pelegrina Franganillo. The jumping spiders placed in Pelegri- na are medium-sized dendryphantines distributed throughout North America, with some species extending as far south as Panama. The 38 species include the well- known P. galathea, P. proterva, P. fla- vipedes, and P. aeneola. Males of Pele- grina are generally brown with white stripes (Fig. 1), and most can be distin- guished from other dendryphantines by the wide embolus with two rami retrola- teral to the opening (Fig. 3). The spotted females (Fig. 2) have large thickened flaps over the epigynal openings (Fig. 4). Al- though the eastern species were well stud- ied by Kaston (1973), most of the species occur in the western United States, Mex- ico, and Central America, and they re- ceived their last comprehensive treatments by G. & E. Peckham (1909) and F. O. Pickard-Cambridge (1901). Many of the western species have been inadequately described and illustrated, often from only one sex, making identification almost im- possible by anyone other than an araneol- 218 Bulletin Museum of Comparative Zoology, Vol. 154, No. 4 ogist familiar with the group. Many spe- cies in the southwest were undescribed, and for most species there is Httle pub- Hshed information on natural history. The present revision has as its main goal to make the species known, by describing and illustrating them, their courtship displays, and their habitats. Although much prog- ress has been made in distinguishing spe- cies and matching males to females, many problems of geographical variation and uncertain male-female matching remain for future work, especially among Mexi- can Pelegrina. In addition to the Pelegrina species, Na- gaina incunda and the U.S. species of the Metaphidippus mannii group are de- scribed because they could very well be confused with species of Pelegrina and be- cause their taxonomy is in need of revision. This work addresses the phylogeny of Pelegrina and the subfamily Dendry- phantinae, but it has no pretensions of be- ing a comprehensive or modern phyloge- netic treatment. My phylogenetic goals are to propose some characters that might pro- vide an outline of dendryphantine rela- tionships, focusing on the question of the monophyly of Pelegrina and a few other groups formerly placed in Metaphidippus. I hope that this and the basic exploratory, species-level taxonomic work will provide the groundwork for future phylogenetic treatments. ACKNOWLEDGMENTS This paper formed part of a Ph.D. thesis for the Department of Organismic and Evolutionary Biology, Harvard Universi- ty. The work described in it has roots many years deep, and during these years I re- ceived the help of many friends and col- leagues. As always, David Maddison has been foremost among my collaborators in formulating ideas, on collecting expedi- tions, on behavioral observations, and in many other activities. My thesis advisor, Herbert W. Levi, provided a well-orga- nized laboratory in which to work, a model of dedication to arachnology, and a helpful encouragement that allowed me to grow on my own. My other friends and col- leagues at Harvard and in our spider lab- oratory— Leticia Aviles, Anna Weitzman, Mark Moffett, Jonathan Coddington, Mark Stowe, Jackie Palmer, Cecile Villars, John Hunter, Caty Sibble, Ardis Johnston, Laura Leibensperger, Dee Woessner, and Ann Blum — made a very supportive commu- nity in which to work. Leticia Aviles and Christopher Maddison gave me much aid and inspiration during the most difficult stages of the project. H. W. Levi, L. Aviles, D. Maddison, G. B. Edwards, P. F. Stevens, and E. O. Wil- son read and commented on this paper. Two anonymous reviewers gave useful suggestions. For discussion of jumping spi- der issues, I thank B. Cutler, G. B. Ed- wards, D. B. Richman, J. Proszynski, F. Wanless, and P. Wijesinghe. B. Cutler, G. B. Edwards, A. Moldenke, and D. B. Rich- man have sent live specimens to me for courtship observations. Brent Opell helped with trypsin clearing. H. D. Cameron gave me helpful advice on the formation of names, though he is not to be held re- sponsible for any ill formed, especially since I did not always follow his no doubt proper advice. Curators at various institutions lent me specimens; a list can be seen under Materials and Methods. I thank these cu- rators for their help. I would especially like to thank Luis F. de Armas of the Instituto de Ecologia y Sistematica, Havana, Cuba, for sending the Franganillo collection of salticids. This work was supported in part by a NSERC (Canada) Postgraduate Fellow- ship. For allowing me to participate in a major collecting trip to Mexico, I thank S. B. Peck and R. S. Anderson. The trip was funded by NSERC and had assistance from the Universidad Nacional Autonoma de Mexico and Harvard University. The De- partment of Organismic and Evolutionary Biology of Harvard supported trips to Cal- ifornia, Arizona, and Spain. Publication costs of this study were covered in part by the Wetmore Colles fund. Pelegrina Jumping Spiders • Maddison 219 MATERIALS AND METHODS Collections Examined. The taxonomic revision is based on specimens in the fol- lowing collections. The abbreviation for the collection is followed by the name of the collection and the curator and others responsible for aiding in loaning the ma- terial, to whom many thanks are due: AMNH American Museum of Natural History, New York (N. Platnick, L. Sorkin) BMNH The Natural History Museum, London (P. Hillyard) CAS California Academy of Sci- ences, San Francisco (W. Pu- lawski, D. Ubick) DU Darrel Ubick personal collec- tion lESC Instituto de Ecologia y Siste- matica, Havana (Luis F. de Ar- mas) MCZ Museum of Comparative Zool- ogy, Cambridge (H. Levi) MSUW Midwestern State University, Wichita Falls, Texas (N. Hor- ner) TXAM Texas A&M University, College Station, Texas (A. Dean) UCB University of California, Berke- ley (E. Schlinger, C. Griswold) UWBM Burke Museum, University of Washington, Seattle (R. Craw- ford) WPM W. Maddison personal collec- tion ZMB Zoologishes Museum Berlin (M. Moritz, S. Fischer) Note that Canadian specimens are, in gen- eral, underrepresented in this revision be- cause two major collections of Canadian spiders, the Canadian National Collection at the Biosystematics Research Centre, Ot- tawa, and the Royal Ontario Museum col- lection, were not examined due to time limitations. Routine Examination and Illustra- tions. Specimens were examined in a glass dish with a bottom layer of half black, half white silicone rubber (bathtub caulking). The silicone rubber is superior to paraffin for most purposes, for it can hold even minuten pins firmly and later heal, and it offers the advantage over sand of allowing appendages to be pinned open. Palpi were mounted on Vaseline in an alcohol-filled depression slide with a coverslip and drawn at 100 X and 200 x under an Olympus BH-2 compound microscope using inci- dent fiber-optics illumination and a cam- era lucida. Not only did the use of a com- pound microscope allow higher resolution, but also the axial light path prevented the drawing difficulties caused by the side-to- side shifting of the image that occurs when focusing on a stereo dissecting microscope. For the external (ventral) view, epigyna were examined using the same technique, without clearing. Epigyna were dissected off of the specimen to allow for the small working distance of the compound micro- scope. The Vaseline on which they were mounted was made opaque by mixing with chalk dust, in order to simulate the cream- colored muscles and glands that would un- derly the epigynum on an intact specimen. After examination of the specimen. Vas- eline was removed by a xylene rinse. The oblique drawings of the male carapace and chelicerae were made mostly under the compound microscope, at 40 x . Most drawings of the female abdomen were made under a Zeiss stereo dissecting mi- croscope with a camera lucida. The draw- ings of the male face and female abdomen show the appearance in alcohol. Most drawings were done on coquille board with ink, a Conte drawing pencil, and white paint. Small labels with my initials (WPM) and the year drawn (e.g., 84) were placed in vials of specimens illustrated. Photo- graphs of living specimens were made with a standard 55-mm lens reversed on exten- sion tubes to yield approximately 2.5 x magnification on Kodak Technical Pan or Kodachrome film, using illumination by flash. Measurements of carapace length, carapace width, and body length (Galiano, 1963; Wanless, 1978) were made from the 220 Bulletin Museum of Comparative Zoology, Vol. 154, No. 4 dorsal view using an eyepiece reticle on a Leitz stereo dissecting microscope. Gen- erally only about five specimens of each sex were measured, because little reliance on these measurements was made in this study. They are intended to give only a general idea as to the size and proportions of the species. The results are presented as follows: minimum (median) maximum. Thus, if five measurements for carapace length for the females of one species are 2.0, 2.1, 2.2, 2.3, and 2.3, these would be reported as follows: 2.0(2.2)2.3, n = 5. Descriptions. A species description was originally written from a sample of five males and five females or more (if avail- able). During subsequent identifications, the description was periodically checked to ensure that it covered the range of vari- ation within the species. Two characters described that were less thoroughly sam- pled are the exact region of contact of the forehead band with the anterior median eyes, which was examined in only four to eight males, and the details of the internal epigynal ducts of the female, which in many species was examined in only one to three females. Clearing. Clearing was used for detailed examination of the integument, especially to observe external and internal structures of the genitalia and mouthparts. The var- ious body parts were cleared by placing them in warm trypsin solution for 1-2 days to digest internal tissues. When the tryp- sin-clearing procedure is successful, it re- veals palp morphology to a level of detail not previously published (Figs. 3, 16-27). Trypsin was used instead of potassium or sodium hydroxide because it damages the cuticle very little and has shown no ten- dency to expand the palp except perhaps when hematodochae are tightly coiled as in Ashtabula and Bagheera. If the tissues of the specimen are well fixed and firm, then digestion will be very slow. Hence, for best results the specimen should be fresh-killed (not fixed) or "fixed" in a poor fixative with a low concentration of alco- hol. Fresh-killed, dilute ethanol-triton fixed, 80% ethanol fixed, and Kahle's fixed specimens were all used. Specimens not fresh-killed were first rinsed in water for several hours before digestion. The body parts were separated to allow penetration of the trypsin. The trypsin solution was made from about 1 ml purified trypsin (Fisher #T-360) in 10 ml water, filtering after mixing. The trypsin was warmed un- der a light bulb during digestion. After digestion, the parts, especially dark, heavi- ly sclerotized palps or epigyna, were bleached. They were first rinsed in water and then 80% ethanol, then moved to a solution of 1 part ethanol : 1 part 10% Aero- sol for a day, and then for bleaching moved into a solution of 1 part 30% hydrogen peroxide : 1 part ethanol : 2 parts 10% Aero- sol. Aerosol (Fisher) was added to the bleaching solution to inhibit the formation and buildup of bubbles, which otherwise can fill the body part and make it unusable. After about 1 day of bleaching, the body part was moved to 80% ethanol and then transferred to 95-100% ethanol. The body part was then mounted temporarily in clove oil or permanently in Euparal. For most specimens, the transferral to Euparal was accomplished by placing the body parts in Euparal thinned with either eth- anol or Euparal Essence and letting the ethanol or essence evaporate off, thus grad- ually taking the specimen through a series of stronger Euparal solutions. This was most easily done directly on the microscope slide, which was then let to dry partially in a dust-free area. Drying thickened the Eu- paral, allowing final positioning of the parts before adding the coverslip. After the cov- erslip is added, the body parts, especially the palpus, may move. Thus, the palp was placed near the edge of the coverslip to allow repositioning using a microneedle slipped under the coverslip. Expansion of Palps. Palps were best ex- panded permanently by boiling the spec- imen alive and then fixing it in dilute Kahle's fluid to harden it in an expanded position followed by gradual dehydraton to 80% alcohol. This technique is much Pelecrima JiMi'iNG SPIDERS • MaddisoTi 221 like tl.-it described by Sadana (1971). Palps so prepared are resistant to contraction and can be critical-point-dried successfully (Figs. 7, 9). A few palpi that were already preserved in alcohol were expanded in a few minutes b\ placing them in a hot mix- ture of 15% hydrogen peroxide, 40% wa- ter, and 45% ethanol. Chromosomes. Chromosomes were ex- amined using the Feulgen technique as described by Maddison (1982). The results are given under the description of each species examined. Courtship. Courtship observations were obtained for Pelegrina species and nu- merous other dendryphantines. Specimens were examined within a few days after collecting. A male and a female were placed on a cotton beating sheet, usually not in full sunlight, and manipulated until the male faced the female and began dis- play. Behavior of the male was observed by eye, and notes and still photographs were taken. Female behavior was not re- corded. For most species, no filming or videotaping was done. As salticid behavior can be fast, it is difficult to take notes ac- curately, and there are likely errors in ob- servation, especially of subtle differences in timing and positions. Still, consistency of observations and videotape confirma- tion of my own previously taken notes for Pelegrina dithalea, Metaphidippus man- nii, Metaphidippus chera, and Phanias watonus all indicate that the descriptions should be generally accurate. There are a number of observations, such as the alter- nate palp waving in the Pelegrina fla- vipedes group and the triangular crouch pose of Pelegrina aeneola, that have been repeated a number of times and in which I have strong confidence. In the descrip- tions of the displays, the sample sizes for the observations are indicated by listing the number of observations for each fea- ture of the display. For instance, in the description of the courtship of P. galathea, the following sentence occurs: "First legs flicker (n = 12, 63) on each series (n = 4, 23) up and down (n = 4, 23) and alternately back and forth at tips (n = 1), vigorously (ca. 5 c/s) (n = 1) but at low amplitude (n = 5, 33)." The parenthetical comments in- dicate the number of observations (n) and the number of males in which the feature was noted. The number of observations was counted as follows: a male was ob- served doing a bout of courtship display; any features of position or motion were considered to have thus been observed once during this bout. If the male stopped dis- playing because the female left or rejected him, and if he began again later (perhaps after I had moved them back together), then the next display was counted as a separate observation. While the more ob- vious features of the display may have been observed many times (for instance, that the legs were flickered was observed 12 times in six males in the preceding example), some of the more subtle details of the dis- play were not noticed in most displays and, thus, would have been observed only a few times (for instance, that the legs flickered "alternately back and forth at tips" was observed only once). Where there is vari- ation, the description lists each of the al- ternatives with an indication of sample sizes. For instance, if the legs were usually flickered at low amplitude, but occasion- ally at high amplitude or not at all, the description might read like this: "On each series legs flickered (n = 9, 13) noticeably (n = 1) or with fairly low amplitude (n = 7, 33) or perhaps not at all (n = 3, 13)." Sample sizes are in general small. It was felt that it is presently more important to obtain a broad survey, including many species, than a deep analysis of a few spe- cies. Explanations of terms used to describe courtship are given in the section Expla- nation of Morphological and Behavioral Terms and in the description of behavioral characters (item 7) supporting the mono- phyly of Pelegrina. Phylogenetic Analysis. Though this work is primarily concerned with the ge- nus Pelegrina, an attempt was made to outline the broader structure of the family and the relationships of dendryphantines, 222 Bulletin Museum of Comparative Zoology, Vol. 154, No. 4 partly for their own sake and partly to set the context for the genus Pelegrina. The general discussion of phylogeny within the family is presented separately (Maddison, 1988, unpublished manuscript). My phy- logenetic proposals are presented in a nar- rative discussion of groups and characters; no numerical phylogenetic analysis was done. Table 2 shows the distribution of some of the more important characters, but it will be noted that some of these do not perfectly support the groups proposed. Problems with some of the characters are noted in the phylogenetic discussion. Fur- thermore, the presumption of ancestral state for a given character is usually not accompanied by rigorous outgroup anal- ysis (e.g., Maddison et al., 1984). We are still making only preliminary sketches of the phylogenetic structure of this large and poorly known family, and it would not be productive to delay phylogenetic hypoth- eses until they can be rigorously defended. The suggestions made should have at least a glimmer of truth and will, I hope, stim- ulate future work. Species Distinctions. Populations were considered distinct species if several con- sistent and discrete morphological differ- ences could be found among them, but when there were few differences, apparent intergradation, or little material, the de- cision as to whether one or more species are present was sometimes difficult. In sev- eral cases, only a single species was rec- ognized despite geographical variation, because the geographical variation was too confusing at present (furcata-mimus) , or because the differences were slight and possibly not consistent {aeneola / uteanus, northern and Floridian tillandsiae, north- ern and southern carmenensis, mannii/ versicolor) . In other cases, allopatric pop- ulations that are similar but that differ con- sistently in a number of features were rec- ognized as distinct (sabinema/pervaga, bi- cuspidata /volcana, neoleonis / tristis, proxima /galathea/ dithalea,chera/ tricol- or / diplacis) , though in each of these cases the decision was difficult. Pelegrina fla- vipedes, flaviceps, and exigua were main- tained distinct despite apparent hybrid- ization because they differ in numerous features. The two sympatric forms of ex- igua were left under one specific name until they can be better studied. Male /Female Matching. Care had to be taken in proposing which males and fe- males belonged together in the same spe- cies, as in other spiders (Levi, 1985). The problem was especially bad in Pelegrina species from the southwestern United States, Mexico, and Central America, where collecting has been limited. Despite the strong sexual dimorphism, similarity in body form and markings could often be used as evidence. Other criteria used were expected correlations in genitalia (e.g., wide, robust embolus with strong epigynal flaps), co-collecting in same geographical region, locality, or microhabitat, and sim- ilarity of the male and female each to those of another well-matched species. Com- ments regarding evidence used to match males and females are given in those spe- cies descriptions where it seems needed. Among the less certain matchings are those for Pelegrina sandaracina, pallidata, chai- mona, huachuca, and morelos. EXPLANATION OF MORPHOLOGICAL AND BEHAVIORAL TERMS Markings in General. Color patterns are generated by integument coloring and by covering of setae, though pale setae usually overlie pale integument and dark overlie dark. The terms hair and scale are used to refer to a thin, more or less cylindrical seta and a broad, flattened seta, respectively (cf. Hill, 1979). Markings of the Carapace. Male den- dryphantines are commonly dark brown with bands of white to yellow scales, usu- ally including a major longitudinal band on either side of the carapace and abdo- men. The following names are used to re- fer to the bands of pale scales on the car- apace (Fig. 1): side bands: Longitudinal bands on either Pelegrina Jumping Spiders • Maddison 223 side of carapace, beginning beside the anterior lateral eyes (ALEs) and pro- ceeding just beneath the small eyes and posterior eyes and beyond, onto the tho- rax. cheek band: Oblique band on the side of the face, starting beneath the ALEs and proceeding down and posteriorly to the carapace margin, in Pelegrina and the mannii group. forehead band: A V-shaped marking on the dorsal cephalic area just behind the anterior median eyes (AMEs). marginal band: On the lower margin of the sides of the carapace, often extended from the cheek band in Pelegrina. Fe- males of most species show none of these bands distinctly; the carapace is instead often covered more or less uniformly with pale scales including a dense white covering on the clypeus. Setae Surrounding Anterior Median Eyes. These are of various colors, from white to black. In the descriptions of Pe- legrina males, the colors of setae around the circumference of the left anterior me- dian eye (AME) are indicated using a no- tation derived from hours on a clock's face. Usually, the colors on only the dorsal part of the AMEs are described, for those ven- trally are more variable. Thus, "white forehead band contacts the AME dorsally 10:30-12:30" means that as one looks from the front at the left AME there are white setae from 10:30 o'clock to 12:30 o'clock that are continuous with the forehead band, and dark setae on either side of this. Markings of the Legs. The legs are yel- low to dark brown, but there are often annulate markings (Fig. 1), especially in males, so that each leg has pale portions covered with white scales alternating with darker portions lacking white. Markings of the Abdomen. The abdo- men is usually brown above in males, ringed by white side bands (Fig. 1). The dorsum of the abdomen often shows traces of the paired pale spots seen in females (Fig. 2). These pale spots are central and paired: the first pair just anterior to the muscle attachment of the second dorso- ventral muscle (number 86viii -(- ix of Whitehead and Rempel, 1959), the second pair anterior to the attachments of the third dorsoventral muscle, and the third, fourth, fifth, and sixth pairs of spots behind this. The fifth and sixth pairs are very small. Often the fourth through sixth pairs are each connected medially and thus form small chevrons. Between these pale spots may be spots of dark brown, sometimes very dark (e.g.. Fig. 353). Male Palpus (Figs. 3, 6-9). The descrip- tions generally assume that the left palp is being viewed from the ventral. The ad- jectives basal and apical when unqualified refer to the appearance in a contracted palp (i.e., basal = toward the tibia and apical = toward the tip of the cymbium). In contrast, "anatomically basal" and "api- cal" refer to the cymbium-embolus axis of connections of the palp's bulb (ie., an- atomically basal = toward the basal he- matodocha and anatomically distal = to- ward the tip of the embolus). In the subfamily Dendryphantinae, the shoe-shaped cymbium holds a bulb con- sisting of (from anatomically basal to api- cal) a fully expandable basal hematadocha, a small subtegulum, a much reduced me- dian hematodocha, the large tegulum, a fully expandable distal (embolic) hema- todocha, and the embolus. The basal he- matodocha expands so as to give a clock- wise rotation to the subtegulum and te- gulum in ventral view of the left palp, as in other salticids. This has been observed by artificial expansion on many dendry- phantines and during copulation of Den- dryphantes nigromaculatus. The subte- gulum is small and contains little more than the basal portion of the sperm duct reservoir (Figs. 3, 8). From the subtegulum, the sperm duct loops up, down, and around the tegulum (Fig. 3) in a clockwise direction to the em- bolus. The distal retrolateral portion of the tegulum, where the sperm duct enters the tegulum from the subtegulum, is generally 224 Bulletin Museum of Comparative Zoology, Vol. 154, No. 4 extended distally and contains a loop of the sperm duct pressed against its wall. This portion I call the shoulder of the te- gulum (Fig. 3). Just proximal to the em- bolus and shoulder is a fold extending across the surface of the tegulum, the tegular ledge (Figs. 3, 6, 7), which serves as a pock- et to hold the proximal part of the embolic base. Wanless (1984) suggested that the tegular ledge (his M3) is absent from sal- ticids other than spartaeines, in which he described it. However, the fold that I am here calling the tegular ledge may be ho- mologous with his M3, for it appears as an extension of the embolic hematodocha cut- ting across the face of the tegulum and occurs in many groups of salticids. The tegulum is filled with tegular glands, which empty into the sperm duct via a series of pores in the sperm duct (see Osterloh, 1922; Bhatnagar and Rempel, 1962). As in other spiders, these pores are aligned into a band (Schult, 1980), which, when narrow, ap- pears as if it were a seam along the length of the sperm duct. Along the narrower part of the sperm duct toward the embolus, the glands are connected to the sperm duct via long ducts (Fig. 3; also known from other spiders; Osterloh, 1922: figs. 20, 26). These pores, ducts, and glands have been largely ignored in systematics but appear to be a potential source of good systematic char- acters (for instance, in some salticids the pores are arranged in a broad ribbon, whereas in others the band is very narrow, and in Sitticus the pores are arranged into prominent craters). The embolic hematodocha arises on the back side of the tegulum (Fig. 8), prolater- al to the subtegulum. Its wrinkles sweep apically up toward the embolus. The exact arrangement of the folds of the hemato- docha is probably of systematic value but is very difficult to untangle, for especially near the base of the embolus the folds are twisted and confusing even in a well- cleared palpus. During expansion, the em- bolic hematodocha expands fully to move the embolus counterclockwise (i.e., pro- laterally) and back (i.e., toward the cym- bium), as indicated by artificial expansions (Figs. 7, 9). The counterclockwise move- ment of the embolus itself probably ex- plains why the counterclockwise-coiled embolus can still engage the epigynal opening despite the clockwise movement of the tegulum (in salticids with the em- bolus fixed to the tegulum, the embolus has a clockwise curve, which thus coin- cides with the clockwise thrust of the te- gulum). The embolus of dendryphantines usually consists of a basal portion, which is transversely directed, and an apical por- tion, which is usually thin and erect and has the opening of the sperm duct at its tip (Figs. 40-47; also Figs. 64b, d). The embolic base consists of a more or less sclerotized portion of the embolic hema- todocha that is exposed to the ventral sur- face and that rests betwee the tegular ledge and the embolus (Figs. 3, 6, 7). Its wrinkles suggest that it should be considered part of the hematodocha. As noted, the embolus is coiled counterclockwise, a feature much more readily apparent in the euophryines and other subfamilies than it is in the den- dryphantines. In dendryphantines, the spi- ral is hidden and best seen in expansions or dissections (e.g., Fig. 35). The dendry- phantine embolus arises prolaterally and moves across toward the retrolateral side (the transverse basal portion of the em- bolus) and then folds back toward the pro- lateral and abruptly rises as the erect apical portion (Figs. 20-23). In many genera, the erect apical portion is fused against the transverse portion so that there is no open, freely coiling spiral. A suture on the back side of the embolus (the embolic suture), between the transverse portion and the erect portion, is often present and indicates where the folded-back spiral has not com- pletely fused (Figs. 3, 8, 20-23, 31-35). In cleared palpi, a slight bend in the sperm duct can be seen at this point. The coiling of the embolus and the embolic suture sug- gesting it are clearly visible in Dendry- phantes, Eris, Pelegrina, Phidippus, Tu- telina, Phanias, and many other genera. In Metaphidippus chera, the coiling is eas- Pklegrina Jumping Spidehs • Maddison 225 ily seen (Fig. 35); in Pelegrina proterva, the only trace is the small suture (Fig. 34); in Pelegrina flavipedes, the suture is some- times visible and sometimes not. No trace of the suture or past coiling can be found in Terralonus cf. uniciis and Poultonella alboimmaculata, but Terralonus myloth- nis and Tiitelina elegans, which are, re- spectively, their close relatives, show them well. The loss of a trace of coiling may have evolved several times in the dendry- phantines. In some genera such as Zygo- hallus, Hentzia and Mabellina, the coil is not so compact and, instead, is more open (e.g.. Figs. 24-27, 37, 38, 50, 51, 58-63, 64c, f). Figure 64 summarizes some of the coiling patterns seen in dendryphantines. The erect portion of the embolus is some- times a simple spike with the opening ter- minal, but in many dendryphantines there are prolongations that will be referred to as rami, and often small denticles occur on the surface of the embolus. In Pelegri- na, in particular, there are two rami re- trolateral to the opening of the sperm duct, the prolateral ramus very near the opening and the retrolateral ramus some distance away (Fig. 3). Epigynum. In most dendryphantines, the openings are well separated and S-shaped (Figs. 65-70), with entry toward the lateral in the anterior half and toward the medial in the posterior half (Fig. 5). The lateral rim of the opening is often thickened to yield a more or less convex teardrop-shaped area, which will be called the epigynal flap (Fig. 4). The inner mar- gins of the left and right flaps may be parallel to each other (e.g.. Figs. 262, 274, 322) or be divergent from anterior to pos- terior (e.g., Figs. 346, 352, 398) or be con- vergent (e.g., Figs. 280, 297, 332, 363). In some Pelegrina, the flaps converge to such an extent that their posterior halves are rotated 90° and become transverse (e.g., P. kastoni, Fig. 317). Even more extreme ro- tations are seen in P. arizonensis (180°, Fig. 424) and P. helenae (270°, Fig. 430). The surface of the epigynum between and behind the openings varies in topography among the species of Pelegrina (Figs. 236- 255). In some, the entire surface behind the openings is raised into a mound (e.g., P. clemata. Fig. 246); in others, it is much more concave (e.g., P. ftircata. Figs. 249, 250). At the posterior margin of the epi- gynum is the notch, or guide (Figs. 4, 5), into which fits the male tibial apophysis, although in some dendryphantine groups (Poultonella, Hentzia) the guide has moved anteriorly as in pellenines (Pel- lenes, Hahronattus) and Bianor. The different parts of the copulatory ducts of Pelegrina species are named as follows (Fig. 5). From the anterior half of the openings, the ducts proceed first lat- erally and posteriorly (the first curve), then medially (the second curve), and then pos- teriorly (the third curve), and then twist a number of times and proceed dorsally to the fertilization ducts. The inner surface of the duct is relatively smooth in the first and second curves, smooth or rough in the third curve depending on the species, and rough with projections in the twisted area posteriorly. The flower-shaped openings of the accessory glands occur on the second curve, usually near the junction with the first curve (Fig. 5). The pathway from the copulatory opening through the copula- tory ducts toward the fetilization ducts is almost straight, lacking the separate sper- mathecal reservoir seen in, for instance, some euophryines. Courtship. Some introduction to the for- mat of descriptions of courtship has al- ready been given under the Materials and Methods section. Males, especially early in display, often walk not straight toward the female but instead sidle so that they ap- proach obliquely or not at all. The sidling may be first to the left and then to the right, and so on, to form a zigzag dance (Jackson, 1978). The male usually walks sporadically, taking a series of steps then pausing, another series of steps then paus- ing, and so on. Because the male's pose and motions of the first legs, palp, and abdomen often differ depending on whether he is walking or paused, a dis- 226 Bulletin Museum of Comparative Zoology, Vol. 154, No. 4 tinction is made between the walking phase, which is called the series (of steps), and the pauses. While the male is walking, the first legs are generally raised, spread, and/or extended forward. The first legs may be waved or flickered up and down or forward and backward, one or repeat- edly. The distinction between a wave and a flicker is not precise; in general, "wave" is used when the motion is of low speed or frequency, "flicker" for high-frequency repeated motions. The abdomen may be twitched (Jackson, 1978) down then up, which in many species produces a sound that may function in courtship (Maddison and Stratton, 1988). Explanations of "crouch" and "raisedspread" stages of courtship are given in the discussion of monophyly (item 7) in the description of the genus Pelegrina. THE SUBFAMILY DENDRYPHANTINAE Pelegrina and Metaphidippus are den- dryphantines, which are salticids. Maddi- son (1988) reviewed the phylogeny of the family Salticidae. The majority of salticid species are considered to form a mono- phyletic group, called the Salticine divi- sion (Maddison, 1988), which are distin- guished from the remaining groups (Lys- somaninae, Spartaeinae, and the Cocal- odes group) by eye structure (Wanless, 1984; Eakin and Brandenburger, 1971; Blest, 1983; Blest and Sigmund, 1984), ab- sence of a tarsal claw on the female palp, medial displacement of the gnathocoxal glands (Figs. 14, 15), asymmetrical tarsal claws, a mound of slit sense organs with an associated seta on the medial edge of the chelicera (Figs. 12, 13), and a small intercheliceral sclerite (Figs. 12, 13), each of which may be considered derived with- in the family. Within the Salticine divi- sion, there are some prominent subfamilies that have the embolus fixed immovably to the tegulum, including the Heliophaninae (delimited by an apparent stridulatory ap- paratus and a bump on the tegulum just clockwise [left palp] of the embolus; Mad- dison, 1987), the Plexippinae (including Plexippus, Hyllus, Evarcha, Thyene, Te- lamonia, Harmochirus, and part of Bia- nor, delimited by a modified serrula on the male endite and a bump on the te- gulum just counterclockwise [left palp] of the embolus), and many other familiar groups such as Pellenes, Salticus, Sitticus, Phiale, Myrmarachne, Amycus, and their respective relatives. However, the Eu- ophryinae, Dendryphantinae, and several smaller groups are united by an embolus that is free to move, being separated from the tegulum by a fully expandable he- matodocha (Figs. 29-31). In these groups, the embolus is also coiled counterclockwise (left palp). The subfamily Dendryphantin- ae itself is delimited by the derived con- ditions of a carina on the underside of the male chelicera, by the coil of the embolus folded back so as to be hidden behind the base of the embolus, and by S-shaped epi- gynal openings. The Dendryphantinae is a subfamily of several hundred species, most of these in the New World. Males are usually striped, with longitudinal bands of pale scales on either side of the carapace and abdomen, while females usually have paired spots on the abdomen. The chelicerae of males are often enlarged or elongate as is the first pair of legs. The posterior eyes are smaller than in most euophryines. Table 1 lists the generic names (includ- ing those now considered synonyms) that I refer at least tentatively to the subfamily. Some of these genera are included for the first time. Some genera are included with hesitation, either because they are obscure and their type species were not examined (e.g., Anamosa, Homalattus) or because they exhibited none of the supposed den- dryphantine synapomorphies listed later but have the general body form and mark- ings much like those of other dendry- phantines (e.g., Mabellina). Other genera are excluded with hesitation — for in- stance, those genera related to Ballus (Co- laxes, Marengo, PadiUa, Pachyballus, and perhaps Admestina) and to Synageles (Consingis, Descanso, Peckhaniia, and Pelegrina Jumping Spiders • Maddison 227 Table 1. Generic names ok jumi'inc; Dendryphantinak SPIDERS lEN lA Il\ K1,V REEERRED Synonymies are not indicated. TO THE SUBEAMIEY Admirala G. & E. Peckhani, 1901 Agassa Simon, 1901 Amerotritte Mello-Leitao, 1944 Anamosa G. & E. Peckham, 1895 Atiicins Chamberlin, 1925 Anoka G. & E. Peckham, 1893 Ashtalntia G & E. Peckham, 1894 Avitus G. & E, Peckham, 1896 Bagheera G & E. Peckham, 1896 Beata G & E. Peckham, 1895 BeUota G. & E. Peckham, 1892 Bryantella Chickering, 1946 Cerionesta Simon, 1901 (n. nov. for Cydonia G & E Peckham, 1893, preoccupied) Chirothecia Taczanowski, 1878 Dendryphantes C. L. Koch, 1837 Donaldius Chickering, 1946 Dryphias Simon, 1901 Eris C. L. Koch, 1846 Gastromicans Mello-Leitao, 1917 Ghelna new genus Hentzia Marx, 1883 Homalattoides F. P. -Cambridge, 1901 Homalaitus White, 1841 Lurio Simon, 1901 Mabellina Chickering, 1946 Maevioheata di Caporiacco, 1947 Megatimus Thorell, 1891 Messita G. & E. Peckham, 1896 Metaphidipptis E. P. -Cambridge, 1901 Nagaina G. & E. Peckham, 1896 Oscricta Simon, 1901 Paradamoetas G. & E. Peckham, 1885 Parahentzia Bryant Paraphidippus F. P.-Cambridge, 1901 Parnaenus G. & E. Peckham, 1896 Partona Simon, 1902 Pelegrina Franganillo, 1930 Phanias F. P.-Cambridge, 1901 Phidippus C. L. Koch, 1846 Poultonella G. & E. Peckham, 1909 Ramboia Mello-Leitao, 1944 Rhene Thorell, 1869 (n. nov for Rhanis C. L. Koch, 1848, preocc.) Rhetenor Simon, 1902 Rtidra G. & E. Peckham, 1885 Sassactis G. & E. Peckham, 1895 Sebastira Simon, 1901 Selimus G. & E. Peckham, 1901 Semora G. & E. Peckham, 1892 Tacuna G. & E. Peckham, 1901 Terralonus new genus Thammaca Simon, 1902 Tulpius G. & E. Peckham, 1896 Tutelina Simon, 1901 Uluella Chickering, 1946 Wala Keyserling, 1884 Zeuxippus Thorell, 1891 Zygoballus G. & E. Peckham, 1885 perhaps Cheliferoides and Leptorchestes) may very well be derived dendryphan- tines (see Maddison, 1988). Simon (1901, 1903) placed many of these dendryphantine genera in his group Den- dryphanteae or Rheneae, though a num- ber of them were scattered among other groups: Beata in the Simaetheae, BeUota in the Synageleae, Nagaina in the Belli- eneae, Rudra in the Rudreae, Tutelina in the Chrysilleae, and Zygoballus in the Zyg- oballeae. G. & E. Peckham (1901b) in- cluded in their Phidippus group of genera Phidippus, Paraenus, Dendryphantes, Se- limus, and Admirala, here considered dendryphantines, but as well many other genera now considered to belong to the Euophryinae and other subfamilies. Pro- szynski (1976: 15, 148-150) listed the gen- era Cheliferoides, Dendryphantes, Eris, Metaphidippus, Paradamoetas, Phidip- pus, Rhetenor, Sassacus, Thiodina, and Wala, in the subfamily Dendryphantinae; all except Cheliferoides and Thiodina are here considered dendryphantines. Specifying distinct apomorphies to jus- tify my concept of the subfamily is diffi- cult, for no characters both universal throughout and unique to the group are known; homoplasy must therefore be in- voked if the subfamily is to be accepted as proposed. Three characters derived within the family are proposed as syna- pomorphies of the subfamily: 1. Carina on underside of the male che- licerae (Table 2, character 1). On the ventrolateral edge of the basal segment of the chelicera there is a fold or carina (Fig. 10). This carina occurs in almost 228 Bulletin Museum of Comparative Zoology, Vol. 154, No. 4 C C C C JJ ij (ii a d ci. d 0; oj 0; 0; 0; 0; u ^ u CU Cl, Cu c c ? c c -^ c a! « oooooooo OOOOOOOO^'" OOOOOOOO cccccccccc O^OJD1J1>Q-'OUO 1>C^Q^1;0000^0 _c J= J= _c o o o o o o o o k« u ^ ^ o o o o c c c c c c c 0~. 0-. 0-- 0-- J2 c c c c c J2 ^ _D JD c3 cd nj cd aja.oa;oa;a;a;a;oooa;_S_^_^ ^^ ^^^^^^^ ^ C C C C o.. JJ ^ a-, j! 3^ , c c s^ i: 0-. c^. a. c c c ^ c c O O O 3^ 0) OJ , c« v5 « x; « (« ' cd cd cd Pi ^ ^ cdcd cd cd cd cd n^ p^ cd cd OOOOOOOOOOOOOOO cdcdcdcdcdcdcdcdcdcd OOOOOOOOQJQJOOOOO ooobbbbbpbbbbbo^.ooocL. T3 -a -o -o red lade lade OOOOO OO 02 -o -o -a ~ .u = c S: .3 h :£ t; «^ H ^ o -^ ^ -^ 3-S Q,cx;a^(i,u,[x;a,QH!x;a;cCa;cx;§;§ <.<. &C <3 "^ ^ a ^ CQ CQ CQ CQ : : io ^ C 3 ,^ O 3 =:: CA ^ a ■- OJ ^ N o c U c o o IT- OD U > " c t^ cca, ?.J 2 -c; 2 S c h ^ — =.1:2 QP >~ ? 3 C ' •' „ . . a -c -c w W y hj O O o r Pelegrina Jumping Spidehs • Maddison 229 e I' S so H _Q _D ^ ^ _Q ^^ _Q _D cv. o-. c^. c c C C 0; c 0; c 1; -Q -Q _Q OJ 1) ti rt oi c« a, & a cccccccc P C C C C C ^ U Qj q^ U 0^ c c c c c c c c c c c c c c c c c CV. CV. iTL. iTL. r^ r^ [r^ tn tn '^ * en Ln l/^ ^ tyi ty^ tyi lyi '- ir\ ir\ tr> '- irt tn c c c c c c c c c c c c cccccccc OOOOOOOOOOOrL. OOOrL. OOOcv. OO 0) ^. _Q 0-- OJ D (TV- T3 13 a a a a ^ ^ Qj 1> 1; Qj Ij C C C C C C C _-,'T3T3T3-T3T3'T3"0 aaaaaaa 0; Oj OJ C C CCCC CCCCCCCC -o-o-o-Oo^. -o-oTJ aj-73-a-o-Oo-. "O TJ "O "O -O T3 "T3 S<"0 "O 'O "O c c c £ c c c jj u u 5^ 1) 1; u c c c c c Qj 1> 1; ^ q; Q^cfl cs cL,™ 03 cs aaaaa a "* CcCCCcCCCc Q,rt aaa^ aaa^ _! .^ so Oj (u n o c > ■> -c Ir ~ c i: a >- c2?'c2^Q-0-Qh'^ 2^ -J s U Dh U ^^£^1-5 i^-2 2 S <^ c a. = J I £^ cc Oh ^ § §^ _^^ 13 cU X 3 Q.aScu'^ij^^ a; a. a, cC 3 "*■-=: S 2 co= 5 ■5~ s: ? S c ^ -5 J5 *- 2 '3 a. a - ^ 2 ~ -. ^ S? c: c o a, -3 a ^ r^ o o a o a ^ ^ ^ in 03 ^ bCJC -C c o o 3 O a3 o;^''^ 2 P O P o! a ■" 4^ _c -P -o S ^ . .. " O CQ - a- -a CO c '^ ^ CO _c Oj 3 CL I ^^ ■f o o; 5 - r, o -S K t; P : ^ 2 >- Si § Cj — -~ ~ -c - U g c 13 o •P ° c -a - ^ £ 2 "C ^ -2 &^ O 1^ OJ _c a <£ -a •p =«3-2 -P 3 3> 3 (J t1 c^ 2 Cj ? c ? a 3 ^ <^ b p^ ^^ £ I ScC' a< i — c a p I2.£i - 3^^ I c-5 ^ -2, " :q . ^ p P « W < c aj o i; P o a So 3 ^ '^ .2 £ a -i t: r? P 3 n 03 c ■•^ = J2 .^ ^ J -5 ■5 w 23 -S " 4* C 05 c 3 .2 _D - 0; £ cfl "53 t*- ^ 4J 6c a J:i I' :c p "5 s;:: ^O c cAi ^ s p ^ bC"o c ■p ^ J2 — J3 • en CO ■>< 3 -£ _ M -o ii a. — a; rt 03 "ts ;_ *j , c oi_ 2 o OJ .b ^ :r" -3 a "-" o !« "rt > ^ a- 0; J2 *^ i2 -Q c ' ^S '^ P ,. ^ C^ -^ l-" 03 p 3 0 5 a 0^ - .ii t3 0; -a CQ o< --4 ■« -c ■— _ m u, c 4j .0; a^ 03 c o ^ lb 0^ ■ ~ ~ ^ Ji a. •= •- >."2 '*3 5 ^ 'S § _Q -c P 'r; cc -p *^ — aPf^i 5"* I' S 02- =-^ o-^~ a '" - ^ .ta^ "^ ts i«'^^-. — -2 '3 i; 0; lb 30 CO C -g ■"^'"C3a^ £ 5 ^ ^ .§ ■- 2 S -o - - >,^ -s ^ Q "O • - ■" ^-o P « (b — -o • P *:; > JD 03 : ' SSa: .^. § °J5 -^O'C 3 "^ 2 c 2 a 230 Bulletin Museum of Comparative Zoology, Vol. 154, No. 4 all genera considered dendryphantines. It is absent in some groups in which the chelicerae are much elongate (Messua, Hentzia, Rudra, Paradamoetas), though this may be related to the elon- gation, as Hentzia mitrata (with short chelicerae) does have the carina. More troublesome are those dendryphantine groups (some Phanias, Tutelina, "Eris' nidicolens, "Beata" octopunctata) in which the chelicerae are not elongate and yet the carina is lacking. In at least one of these, Phanias, there are some species with the carina, and its absence in others is presumably a secondary loss. The carina is lacking in all other salti- cids I have seen, except for Simaetha paetula (see Prosynski, 1984: 132) and Simaetha tenuior, in which a similar carina is present. These species have a Sitticus-\ike palpus, with a round te- gulum and immovable embolus, and appear to be distantly related to den- dryphantines. 2. Coil of embolus compressed, so that the embolus folds back sharply on itself and superficially appears not to be coiled (Figs. 20-25, 31-35, 37-39, 64b-d, g; Table 2, character 2). Typically, the embolus arises prolaterally and moves across toward the retrolateral side (the transverse basal portion of the embolus) and then folds back toward the prolat- eral and abruptly rises as the erect api- cal portion. A suture on the back side of the embolus, between the transverse portion and the erect portion, is often present and indicates where the folded- back spiral has not completely fused. The coils of the embolus are not folded but rather open and exposed (Fig. 19), in the other groups, such as the eu- ophryines, that have a counterclock- wise-coiled embolus. Two main prob- lems with the use of this character are posed by various dendryphantines: in some, the embolus appears not even coiled, and in others the embolus ap- pears coiled but the coils are exposed (Fig. 64 summarizes some of the vari- ation seen among the dendryphan- tines). Despite the first problem, coiling is considered primitive for the subfam- ily (see earlier discussion under the sec- tion Explanation of Morphological and Behavioral Terms). More troublesome is the second problem: the occurrence of dendryphantine groups such as Par- adamoetas, the South American "Sas- sacus," Mahellina, Dendryphantes tro- picus, and Bryantella in which the coils of the embolus are exposed (Figs. 26, 27, 58-63, 64e, f). However, three lines of evidence suggest that the exposure of the coils is not homologous to that of the euophryines and, instead, is sec- ondarily derived from the hidden con- dition. First, the exposed spiral of Den- dryphantines is always of a different form from that of the euophryines, be- ing placed more retrolaterally and not so small and tightly coiled. Second, oth- er characters suggest that these trou- blesome dendryphantines do indeed belong with other dendryphantines, such as the lack of a concave retrolateral loop on the sperm duct in the tegulum, which may be apomorphic as noted lat- er with respect to Pha7iias, the more basally placed tegular ledge, which may be apomorphic or plesiomorphic, and the occurrence of the cheliceral carina in at least some species (e.g., Metaphi- dippus vitis). Third, there is an appar- ent morphocline between the hidden spiral of Eris aurantia (Fig. 64b), through the more marginally open spi- ral of Eris militaris and a species near Zygoballns incertus (Figs. 64d, 53, 57), to the more open spiral of Zygoballns incertus and Paradamoetas (Figs. 64e, 58), to the open and well-coiled spiral of Mahellina and Dendryphantes tro- picus (Figs. 62, 63, 64f). The transition would require merely a retrolateral shift of the erect portion of the embolus and a loss of sclerotization of the basal (lat- erally directed) portion of the embolus to leave the erect part of the embolus free, at which point it could coil as in Pelegrina Jumping Spiders • Maddison 231 Dendryphantes tropicus and Mabelli- na prescotti (Figs. 62, 63). 3. Epigynal openings S-shaped, with en- try toward the lateral in the anterior half and toward the medial in the pos- terior half (Figs. 5, 65, 67-70; Table 2, character 3). To my knowledge, this sinuate opening is unique among sal- ticids. Though most dendryphantines have this character, it has similar prob- lems to the preceding one and is cor- related with it. A number of dendry- phantines (e.g., Eris militaris, Tutelina elegans, Phidippiis clarus, Hentzia, Paradamoetas, Anicius, Zygoballus, Messua, Bagheera) have C-shaped or simple cavernous openings. In some of these, however, ridges descending into the opening are presumably remnants of the teardrop-shaped flaps associated with the sinuate openings (Eris mili- taris. Fig. 66; Tutelina elegans). For some of these, there are related species that have the sinuate openings {Eris floridana, Tutelina hartii, Phidippus spp., Anicius sp., Zygoballus tibialis), but there remain other genera with no remnant of the flaps or obvious close relatives with sinuate openings. In gen- eral, the species lacking a compact em- bolic spiral also lack the sinuate open- ings, perhaps because a retrolateral shift of the erect portion of the embolus is correlated with an expansion of the epi- gynal openings and weakening of the teardrop-shaped flaps. Such a correlat- ed change seems to mark each of Eris militaris, Pelegrina kastoni, Dendry- phantes nigromaculatus, and Phidip- pus octopunctatus from its close rela- tives. RELATIONSHIPS WITHIN THE DENDRYPHANTINAE The limits and interrelationships of gen- era within the subfamily Dendryphantin- ae (see Table 1 for a list) are at present poorly understood. The following discus- sion will make an attempt to resolve only a small part of the confusion, for I will focus on those phylogenetic questions that are most important to resolve the generic placement of the galathea group of spe- cies, which is the subject of the species revision that makes up the bulk of this paper. The galathea group has resided in the genus Metaphidippus, but, as ex- plained next, this genus is polyphyletic. Here I will begin (but not complete) the task of dismantling Metaphidippus. The galathea group will be moved to the genus Pelegrina, the harfordii group to Phanias, the mylothrus group to Terralonus, the castaneus group to Ghelna, some other species to Messua, but some species groups (mannii group, vitis group, various neo- tropical species) will remain temporarily in Metaphidippus for want of a better al- ternative. Table 3 lists the proposed re- classification of Metaphidippus and some other dendryphantine species. Metaphidippus was described in 1901 by F. O. Pickard-Cambridge, who gave no clear justification for its limits. Many of the North American species that had been placed in Dendryphantes were subse- quently transferred to Metaphidippus, in part because of Bryant's (1941) conclusion that these species did not belong with the Old World Dendryphantes. In fact, her cited evidence was mistaken: the Euro- pean specimens she compared were ac- tually "Eris" nidicolens misidentified as D. hastatus (Maddison, 1988). Regardless, the more or less wholesale transfer of North American species from Dendryphantes re- sulted in a Metaphidippus that has been desperately polyphyletic, being nothing more than a catch-all genus of unremark- able North and Central American dendry- phantines spanning much of the diversity of the subfamily. Because the type species of Metaphidippus (M. mandibulatus F. O. Pickard-Cambridge) is not closely related to most species placed in the genus, in- cluding the galathea group, many species of Metaphidippus should be placed else- where. The relationships of the true Me- taphidippus will first be considered, after which the limits of Dendryphantes and 232 Bulletin Museum of Comparative Zoology, Vol. 154, No. 4 Table 3. Summahy of placements of various New World dendryphantine species apart from Pelegrima species, discussed in text. ? indicates placement considered possible but unconfirmed. Authors of names are given in index. Bagheera Bagheera kiplingi B. prosper ?Metaphidippus nigropictus ?M. bicavatus Beat a Beata hispida B. inconcinna B. longipes B. maccunii B. magna B. rust tea Dendryphantes (partial) Dendryphantes nigrotnaculatus D. chiildensis D. fusconotatus D. hastatus D. rudis Gastromicans Gastromicans albopilosa G. hondurensis G. levispina G. noxiosa G. vigens ?Hasarius lisei Messua Messua desidiosa M. centralis M. dentiger M. donalda M. lata M. laxa M. limbata M. niunia M. octonotata M. pura M. tridentata ?M etaphidippus cupreus ?Metaphidippus ovatus ?M etaphidippus iridescens ?M etaphidippus inflatus ?M etaphidippus quadrinotatus ^Dendryphantes felix Phanias Phanias albeolus P. concoloratus P. doniinatus P. flavostriatus P. furcifer P. furcillatus P. harfordii P. monticola P. neomexicanus P. watonus ?P. salvadorensis Terralonus Terralonus californicus T. niylothrus T. unicus T. shaferi T. versicolor T. vittatus T. fraternus Ghelna Ghelna castanea G. barrowsi G. canadensis G. sexmaculata Metaphidippus niannii group Metaphidippus mannii M. carmenensis M. chera M. diplacis M. emniiltus M. tricolor Metaphidippus vitis group Metaphidippus vitis M. texanus M. mathetes Dendryphantes melanomerus Beata will be discussed. Then, the correct placement of various groups placed in Me- taphidippus will be considered. The Bagheera Group of Genera It is within the Bagheera group, com- mon in the Neotropics, that the true Me- taphidippus appear to fall. Members of this group, which includes such common species known as ^^Eris' limbata, ^'Meta- phidippus" prosper, "Beata" albopilosa, have a distinctive embolus, which appears, at first glance, to be coiled or curved clock- wise in the left palpus, opposite the coun- Pelegrina Jumping Spiders • Maddison 233 terclockwise coiling I have said character- izes the dendry phantines and related sub- families (Figs. 99-101). This clockwise coiling is apparently superimposed upon the normal coiling by a twisting of the embolus. The main axis of the embolus has twisted counterclockwise (as viewed from tip of embolus in left palpus), thus winding the hematodocha and sperm duct around it (Figs. 64h, 99-101). The tip of the em- bolus, though, seems to have been left be- hind by the twisting, so that the apical part of the embolus takes on a clockwise curl- ing. In some species, this clockwise coiling is visible in the uncleared palpus (Figs. 81, 84, 87, 93), but in others it is not (Fig. 91), and the palpus is little modified from the typical compressed counterclockwise spi- ral of dendryphantines. The most extreme clockwise coiling is seen in "Metaphidip- pus" prosper (Fig. 99) and ''Beata' al- bopilosa (see Fig. 101). Species in the group can be sorted pro- visionally into three subgroups, which may or may not be monophyletic, for each of which there exists an available generic name that has been mostly ignored in the literature: Bagheera G. & E. Peckham, Messua G. & E. Peckham, and Gastromi- cans Mello-Leitao. A fourth genus sharing a twisted embolus is Ashtabula G. & E. Peckham, though whether or not it belongs with the group is unclear. The twisting of the embolus in Ashtabula (Fig. 102) is hid- den beneath the tegulum and is more ex- treme, though similar to, that in Bagheera, Messua, and Gastromicans. However, there are several features that cast doubt on the placement of Ashtabula with these genera. Ashtabula has an extra concave loop on the sperm duct in the palpus, pos- sibly placing it near the base of the sub- family as noted later in connection with Phanias. The body in Ashtabula is not nearly so large and robust as in the Bag- heera group, instead resembling that of Hentzia or Anicius. More work is needed before the place of Ashtabula can be set- tled. Species of Ghelna and the arizonensis group of Pelegrina (discussed later) both have a twisted embolus, but in each it takes a very different form from the twisting in the Bagheera group of genera. Sebastira, Thammaca, Lurio, and Parnaenus may also belong with the Bagheera group of genera, though their emboli do not so ob- viously possess the twisting. A brief ac- count of Bagheera, Messua, Gastromicans and Metaphidippus is here given. Bagheera {Figs. 80-85). Bagheera males have elongate, horizontal, parallel chelic- erae (Fig. 80); the retromarginal teeth are near the base of the chelicera; in all but one species there is distally, near the fang, what appears to be a large retromarginal tooth but actually is not (it does not have the terminal canal through the cuticle that seems to characterize all true teeth), and most species have tubercles bearing setae on the inner margin of the basal cheliceral segment. Included in Bagheera are the type species, B. hiplingi G. & E. Peckham, 1901 (type species by monotypy; holotype ex- amined; Figs. 80-83), and Bagheera pros- per (G. & E. Peckham) (NEW COMBI- NATION; Figs. 84, 85, 99) and at least two undescribed species. Metaphidippus ni- gropictus F. P. -Cambridge and M. bica- vatus F. P. -Cambridge may also belong in Bagheera. Messua (Figs. 86-92). Males of this ge- nus have elongate divergent chelicerae (Fig. 86) with a long and sickle-shaped retromarginal tooth near the fang; the pro- marginal teeth are near the base, well sep- arated from the retromarginal tooth; on the anterior distal margin of the basal seg- ment of chelicera near the fang is a flange. Included in Messua are the type species M. desidiosa G. & E. Peckham, 1896 (type species by monotypy; holotype and collec- tions of males and females from San Jose, Costa Rica examined; Figs. 86-89), Mes- sua centralis (G. & E. Peckham) (lectotype here designated, a male from Chiriqui), Messua dentiger (F. P. -Cambridge) (see Fig. 91), Messua donalda (Kraus), Messua lata (Chickering), Messua laxa (Chicker- ing), Messua limbata (Banks) (Figs. 90, 100, 117), Messua moma (F. P. -Cam- 234 Bulletin Museum of Comparative Zoology, Vol. 154, No. 4 bridge), Messua octonotata (F. P. -Cam- bridge), Messua pura (Bryant), and Mes- sua tridentata (F. P. -Cambridge). All these except M. desidiosa are NEW COMBI- NATIONS. Metaphidippus cupreus F. P.- Cambridge, M. ovatus F. P. -Cambridge, M. iridescens F. P. -Cambridge, M. injla- tus F. P. -Cambridge, and M. quadrino- tatus F. P. -Cambridge may also belong in Messua. Dendryphantes felix G. & E. Peckham might be considered either a Bagheera or Messua depending on any future lectotype designation: the body (G. & E. Peckham, 1901b: fig. 6a) in the type vial and its attached palpus are of a Bag- heera species, probably B. prosper, while the separate palpus in a microvial (G. & E. Peckham, 1901b: fig. 6) is of a Messua species. Gastromicans (Figs. 93-95). This genus is distinguished from Bagheera and Mes- sua in having short and vertical but very robust male chelicerae. Included are Gas- tromicans albopilosa (G. & E. Peckham), Gastromicans hondurensis (G. & E. Peck- ham), Gastromicans levispina (F. P. -Cam- bridge) (Figs. 93-95, 101), Gastromicans noxiosa (Simon), and Gastromicans vigens (G. & E. Peckham). All these are NEW COMBINATIONS. Hasarius lisei Bauab- Vianna & Soares probably also belongs in Gastromicans. Galiano (1980) synony- mized Gastromicans Mello-Leitao with Beata G. & E. Peckham because its type species Gastromicans squamulata Mello- Leitao (type species by monotypy) is syn- onymous with "Beata" albopilosa. But in- sofar as Beata albopilosa does not belong in the genus Beata, Gastromicans is avail- able as a generic name for albopilosa and its relatives. Metaphidippus (Figs. 96-98). Though the placement of the true Metaphidippus with these genera of the Bagheera group is to some extent problematical, such a placement is the best supported at present. Before discussing the uniting characters, it would be valuable to give the following brief description of the type species of Me- taphidippus, M. mandibulatus F. P. -Cam- bridge (type species by original designa- tion), whose single known male (Costa Rica, BMNH, examined), is strikingly unlike most other jumping spiders that have been placed in Metaphidippus. Palpus (Figs. 97, 98): Embolus reminiscent of that of Eris species but with the longitudinally directed apical portion not fully erect, in- stead reclined to the prolateral (Fig. 98). The embolic base bears a flange covering the basal part of the embolus. Chelicerae: Long and cylindrical, horizontal and di- verging (Fig. 96), with two promarginal teeth near the base and one retromarginal tooth near the fang. The fang is forked near its base (Fig. 96). Markings (Fig. 96): Carapace brown, lacking side bands ex- cept one patch of white scales on either side of fovea. Wide white band along mar- gin. At least some metallic green-blue scales on cephalic area. Abdomen with thin white side bands broken basally; just anterior to each of the main dorsoventral muscle at- tachments is a small white patch of scales; in the posterior half of the dorsum are two pairs of small lateral white bars. The dor- sum has some metallic green-blue scales. Measurements: Body length 5.4 mm; car- apace length 2.4 mm, carapace width/ length 1.93/2.37. Two features that can be proposed as synapomorphies for the group of Bag- heera, Messua, Gastromicans, and Me- taphidippus are the following: 1. The tibial apophysis is erect and at its base parallel-sided, shaped like a knife blade (Figs. 71-74; Table 2, character 4). Almost all other dendryphantines have an apophysis tapering throughout its length (Figs. 75-79), including Fhanias and other genera that appear to be near the base of the subfamily (see later), and usually the apophysis points at least somewhat ventrally. The only other dendryphantines known to me with a similar knife-shaped apoph- ysis are Ashtabula and Poultonella. Poultonella does not belong with these genera; rather, its peculiar chelicerae I Pelegrina Jumping Spiders • Maddison 235 assure a relationship with Tutelina. A few species of Messua, inchiding M. lata, have a more tapering apophysis. 2. The tegular ledge runs longitudinally (Figs. 87, 90-93, 98; Table 2, character 5), instead of obliquely at 0-60° from the transverse as seen in other dendry- phantines and other salticids with a teg- ular ledge (Figs. 40-46, 50, 52, 53). While this is unusual among dendry- phantines, it is not unique: it also occurs in Phanias, Anicius, and Zygoballus in- certus. Additional features that suggest a rela- tionship between Metaphidippus and Messua in particular are the long, tubular divergent chelicerae with a near-terminal retromarginal tooth and a distal anterior flange beside the fang base. Other den- dryphantines have long divergent chelic- erae, but in all that I have seen except some South American "Sassacus," the tooth ar- rangement is different than in Metaphi- dippus and Messua, with the retromar- ginal tooth remaining near the base or the promarginal teeth near the apex. This dif- ferent tooth placement may indicate that the elongation occurred in different por- tions of the chelicerae in these other den- dryphantines, and thus the elongation is not homologous. As well, the general body form and occurrence of greenish reflective scales are also suggestive of a close rela- tionship between Metaphidippus and Messua, though these characters are loose- ly defined and not necessarily unique. The only feature that would exclude M. man- dibulatus from Messua is the apparent lack of the reverse twisted embolus in M. man- dibulatus. However, one undescribed spe- cies from Costa Rica represented by a sin- gle male specimen appears very closely related to M. mandibulatus in having sim- ilar body form and markings and in having a slightly forked fang, and yet it has a slightly twisted embolus (Fig. 92). If these two species form a monophyletic group, then the lack of twisting in mandibulatus may be a secondary loss, which is not un- reasonable given that other species such as Messua octonotata have little trace of a twisted embolus, and the embolus of M. mandibulatus shows unusual folds and does recline to the prolateral as in Messua. Per- haps more detailed study of its peculiar embolus, when more specimens become available, will allow a more definitive an- swer to the question of twisting in M. man- dibulatus. If the genus Metaphidippus is only an offshoot of Messua, then Meta- phidippus would fall as a synonym of the older name Messua. However, I am re- luctant to effect such a synonymy at pres- ent given the number of Metaphidippus species that would be left homeless, and so Metaphidippus will be left standing for the moment. Regardless, the best-support- ed conclusion at present is that the name "Metaphidippus" properly applies to a small group of neotropical dendryphan- tines related to Messua, Bagheera, and Gastromicans. Two Genera That Have Exchanged Species with Metaphidippus: Dendryphantes and Beata Because most species that have been placed in Metaphidippus do not belong to the Bagheera group of genera, they cannot follow M. mandibulatus and, thus, need to be placed elsewhere. The first place we might look for a possible home for Me- taphidippus species are two genera, Den- dryphantes and Beata, which have in the past exchanged many species with Meta- phidippus. Many "Metaphidippus" were formerly placed in Dendryphantes and several Beata were formerly placed in Me- taphidippus. To discuss the proper place- ment of species now in Metaphidippus more clearly, it would be valuable to re- consider the limits of these two genera. Dendryphantes (Figs. 65, 103-108, 120). The genus Dendryphantes, described last century, has over the years accumulated many species, mostly on the basis of their being unremarkable dendryphantines. Many species were since moved to genera such as Metaphidippus, while others re- 236 Bulletin Museum of Comparative Zoology, Vol. 154, No. 4 maining in the genus will probably even- tually be placed elsewhere (see the com- ments of Edwards, 1977). Among the New World species, there is only one species for which there is presently good evidence for a placement in Dendryphantes: D. nigro- maculatus (Keyserling), most recently placed in Eris (Kaston, 1973). Several Old World species placed in Dendryphantes, including D. fusconotatiis and D. chuld- ensis, appear very closely related to D. nigromaculatus (see figures of Proszyhski, 1971b, 1982). Like the Old World D. has- tatus (the type species) and D. rudis, D. nigromaculatus has a slightly elongate body dully marked. Perhaps the best char- acter that strictly delimits Dendryphantes is the presence of a fold of embolic he- matodocha that lies across the basal part of the embolic base, covering the wrinkles there (Figs. 103-108). If this character is used to delimit the genus, then it would be a small genus of mostly Palearctic spe- cies. The placement of nigromaculatus in Dendryphantes is further supported by its sharing with D. rudis, D. fusconotatus, and D. chuldensis a much elongated prong coming off of the base of the embolus and curving toward the cymbium (Figs. 106, 108). The embolus therefore appears to have two rami, much as in Pelegrina fla- vipedes, though not homologous according to reasoning given below. Species of the other two major groups of Old World den- dry phantines, the "Eris" nidicolens group (Fig. 61) and the genus Rhene (Fig. 52), lack the fold across the embolic base seen in Dendryphantes. Though Rhene species often have a prong arising from the base of the embolus, in Rhene it is not curled toward the cymbium and, instead, is erect as in Beata and the mannii group. Rhene has been considered a close relative and possibly a synonym of Dendryphantes (Proszynski, 1973b), but a number of other features of Rhene such as the presence of epiandrous fusules and the concave retro- marginal loop of the sperm duct of the palpus also cast some doubt on this place- ment. The option of returning a number of groups from Metaphidippus to Dendry- phantes has little merit at present. Moving these back to Dendryphantes would be useful only if they are likely to stay there, that is, if they are closely related to the type species of Dendryphantes. Other- wise, we would merely be worsening Den- dryphantes' status as a catch-all genus and adding to it the confusion of changing the generic placement of many common spe- cies. As noted, only D. nigromaculatus among New World species is a strong can- didate to stay in Dendryphantes. Beata (Figs. 77, 109-112). The limits of the genus Beata have been greatly over- estimated (Simon, 1903; Chickering, 1946). Because the type species of Beata is fissi- dent (it has a bifid retromarginal cheliceral tooth), it has not only been removed from the dendryphantines (Simon, 1903) but has also been burdened with diverse dendry- phantines that happen to have a similarly bifid tooth (Simon, 1903; Chickering, 1946). Note that the tooth is better consid- ered a single bifid tooth rather than two fused teeth because the inner boundary of the cuticle does not extend to the tip of the second cusp. The second cusp shows all gradations of development in the den- dryphantines, with most lacking it, some showing a slightly swollen margin (e.g., Pelegrina proterva, Fig. 10), and others having a well-developed cusp. It is there- fore best to place far less emphasis on this character. Beata magna G. & E. Peckham (Fig. 109), the type species of Beata (by monotypy), bears few resemblances to most of the other fissident dendryphantines, in- stead having many more resemblances with the other robust-bodied dendryphantines previously placed in Dryphias, Homalat- toides, and Anamosa. The following char- acters, which appear derived within the subfamily, delimit this group containing Beata magna: 1. Tibial apophysis narrow and bent to- ward the ventral, almost paralleling a ridge on the tibia below it (Fig. 77). Pelegrina Jumping Spiders • Maddison 237 This tibial ridge is similar to that in Pelegrina (Fig. 78), but it is longer and sharper in Beata. 2. First leg tibia dark and enlarged at least slightly compared to patella, even in females. 3. Carapace distinctively wide and high, higher than in Sassacus, Agassa, and Rhene, wider than in Sassacus and Agassa. Unlike Zygoballus, but like Sassacus and Agassa, the carapace is wide well past the posterior eyes before it abruptly drops and narrows. 4. Carapace scales erect (in at least some but perhaps not all species). 5. Retromargin of base of embolus with prong rising parallel to apical erect por- tion of embolus (Fig. 110). Such a prong is also seen in Rhene (Fig. 52) and the Metaphidippus mannii group (Fig. 499). The following species are placed in Bea- ta and NEW COMBINATIONS therefore established: Beata hispida (G. & E. Peck- ham) (Figs. 77, 110-112), Beata inconcin- na (G. & E. Peckham), Beata maccunii (G. & E. Peckham), and Beata rustica (G. & E. Peckham). Also included is Beata longipes F. P. -Cambridge, which may be the male of B. magna. Dryphias (type spe- cies maccuni by original designation) is a NEW SYNONYM of Beata. The genus Beata as here delimited excludes B. digi- tata (= Pelegrina galathea) and B. var- iegata (= Pelegrina variegata), B. albopi- losa (= Gastromicans albopilosa), B. fla- volineata (= Nagaina incunda), B. ce- phalica F. P. -Cambridge, B. jubata (C. L. Koch), B. munda Chickering, B. pernix (G. & E. Peckham), B. venusta Chicker- ing, B. wickhami (G. & E. Peckham), and B. zeteki Chickering. No new placements are suggested for the last-mentioned seven species. Other species placed in the genus but probably not belonging there are B. cinereonitida Simon, B. germaini Simon, B. lineata (Vinson), and B. striata Pe- trunkevitch. The Proper Placements of Metaphidippus Species Now that the relationships of the true Metaphidippus and the limits of Dendry- phantes and Beata have been reconsid- ered, we are in position to discuss how the various groups within Metaphidippus might be dispersed. Some of these conclu- sions are summarized in Table 3. Four groups are here removed from Metaphidippus, as discussed in subsequent sections. The galathea group is transferred to the genus Pelegrina (and its species re- vised), the harfordii group to the genus Phanias, the mylothrus group to the new genus Terralonus, and the castaneus group to the new genus Ghelna. Two species groups occurring in the United States are retained temporarily in Metaphidippus pending further study: the mannii group and the vitis group. The mannii group is discussed later in connec- tion with the revision of its U.S. species. The status of the vitis group (Figs. 27, 59), including Metaphidippus vitis (Cocker- ell), M. texanus (Banks), M. mathetes (Chamberlin), and Dendryphantes me- lanomerus Chamberlin, is not clear. These species have a characteristic hooked em- bolus (Figs. 27, 59) and are small, some- what elongate, and brown to black and shiny. Metaphidippus vitis was placed in Sassacus by Hill (1979) on the basis of scale morphology and courtship, but scale mor- phology is known in only few dendry- phantines, and a similar courtship is also seen in many other dendryphantines (raisedforward, Table 2, character 9). Fur- thermore, the genitalia of M. vitis are very different from those of the true Sassacus and, instead, are similar to those of the neotropical species placed in Sassacus such as S. arcuatus, which may better be placed in the genus Ramboia (see Bauab-Vianna and Soares, 1982). The vitis group, here retained in Metaphidippus, may eventu- ally find its place in a primarily neotropical genus. Some species placed in Metaphidippus 238 Bulletin Museum of Comparative Zoology, Vol. 154, No. 4 belong in Bagheera or Messua, as already noted. Metaphidippus paiutus Gertsch is a Sassacus (Sassacus paiutus (Gertsch), NEW COMBINATION), possibly a syn- onym of S. papenhoei. Other species of Metaphidippus require further study be- fore their placement can be settled. Among the neotropical species listed under Me- taphidippus by Bonnet (1957), M. longi- palpus F. P. -Cambridge, M. nitidus (G. & E. Peckham), and perhaps M. taylori (G. & E. Peckham) seem to belong to Parna- enus, M. pallens F. P. -Cambridge in Eris, M. perfectus (G. & E. Peckham) (Fig. 60) in Selimus, and M. tropicus (G. & E. Peck- ham) (Fig. 62) in Bryantella. For the re- maining neotropical species (see Proszyn- ski, 1990), I have no placement to suggest. Phanias F. P.-Cambridge, 1901 Type species, Phanias flavostriat us F. P.-Cambridge, 1901, by monotypy. The unusual, elongate dendryphantines of the harfordii group (Figs. 20, 36, 47, 70) include several species in the western United States but many more in the high- lands of Mexico. A generic name is avail- able for this group, Phanias F. P.-Cam- bridge, 1901, based on Phatiias flavostria- tus, described from two females from Om- ilteme, Mexico (BMNH, examined), and previously considered to belong in the Marpissinae. The species of the Metaphi- dippus harfordii group are therefore transferred to Phanias. The members of the genus Phanias share these characters, which may be apomorphies within the subfamily: 1. Tegular ledge expanded so as to cover the tegular shoulder (Fig. 47). The teg- ular ledge of other dendryphantines and other salticids is not so expanded. 2. Embolic hematodocha reduced and sclerotized prolaterally and basally. In at least some species, much of the ex- pansion occurs from out of the tegular ledge instead of from the prolateral dorsal surface (back side) of the tegul- um (Fig. 36), but this feature may be primitive as it is also found in Synageles and Admestina (Figs. 29, 30). 3. Courtship with first legs raised, for- ward, and parallel (Fig. 118) and waved asymmetrically so that the leading leg on sidles is waved exclusively or more strongly (seen by me otherwise only in Anicius). 4. Small blunt teeth on the embolus (Fig. 20). Many other dendryphantines (e.g., Pelegrina montana. Fig. 204) have teeth on the embolus, but their teeth are sharp. The blunt teeth are lacking, however, in some species of Phanias (e.g., P. watonus). 5. Longitudinal bands of white scales, in- stead of passing below and beside the posterior eyes as in other dendryphan- tines, pass around and directly poste- riorly from the posterior eyes. The dis- tribution of this character is not well known. The following NEW COMBINATIONS are established: Phanias albeolus (Cham- berlin & Ivie) (Figs. 20, 70), Phanias con- coloratus (Chamberlin & Gertsch), Phan- ias dominatus (Chamberlin & Ivie), Phan- ias furcifer (Gertsch), Phanias furcillatus (F. P.-Cambridge), Phanias harfordii (G. & E. Peckham) (Fig. 47), Phanias mon- ticola (Banks), Phanias neomexicanus (Banks), and Phanias watonus (Chamber- lin & Ivie). Phanias marginalis Banks (type specimen examined) is a Menemerus, not a Phanias, while Phanias salvadorensis Kraus may be either a Phanias or an An- icius. Also included in Phanias are at least 15 undescribed species from Mexico and the southwestern United States. Phatiias may be placed near the base of the sub- family, for it has two features that are ar- guably ancestral for the subfamily, name- ly, the presence of epiandrous gland fu- sules (Machado, 1951; Table 2, character 6; see also Maddison, 1988), which it shares with Hentzia, "Beata" wickhami, ^^Eris" nidicolens, Rhene, and groups apparently related to dendryphantines (euophryines, synagelines, ballines, Mopsus, Itata, Phle- Pelegrina Jumping Spiders • Maddison 239 gra, though not Neon), and a concave sperm duct loop along the retromargin of the tegulum (Fig. 20), which it shares with some Hentzia, "Eris" nidicolens (Fig. 61), Rhene (Fig. 52), some Tutelina, Anicius, euophryines (Fig. 19), and Admestina. Terralonus new genus Type species, Dendryphantes mylothrus Chamber- lin, 1925, here designated. Name treated as mas- culine. Description and Diagnosis (Figs. 22, 44, 68). These western North American spe- cies are unusual among dendryphantines in being ground-dwelling, usually on ground more or less barren of vegetation, often under rocks. The body shape and markings are distinctive and uniform throughout the group. They are somewhat elongate and have relatively low-contrast mottled markings of coarse brown or gray pubescence. The embolus is long and its base is more longitudinally directed than is usual in dendryphantines (Fig. 22), ex- cept for T. calif ornicus, which has a more typical embolus (Fig. 44). Included Species. The following species are moved to Terralonus: Maevia calif or- nicus G. & E. Peckham, Dendryphantes mylothrus Chamberlin, Dendryphantes unicus Chamberlin & Gertsch, Metaphi- dippus shaferi Gertsch & Riechert, Icius versicolor G. & E. Peckham, Menemerus vittatus Banks, and Menemerus fraternus Banks (type specimens of last-mentioned three species examined). This establishes the NEW COMBINATIONS: Terralonus calif ornicus (G. & E. Peckham), Terra- lonus mylothrus (Chamberlin), Terralon- us unicus (Chamberlin & Gertsch), Ter- ralonus shaferi (Gertsch & Riechert), Ter- ralonus versicolor (G. & E. Peckham), Ter- ralonus vittatus (Banks), and Terralonus fraternus (Banks). Discussion. By appearance, these are not typical dendryphantines — two of the spe- cies were described in the genus Mene- merus and have remained there to this day. Their relatives are unclear, but almost certainly they do not belong in the genus Metaphidippus, because they lack the characters of the Bagheera group of gen- era. I have chosen to describe a new genus for the group to remove it from its uneasy placement in Metaphidippus. 1 do this with some hesitation, given the overabundance of obscure genera in salticids, but the my- lothrus group apparently does not reach the Neotropics where it might have found an available generic name, and so it is un- likely to be synonymized soon. Describing a genus allows easier discussion of this dis- tinctive group. The three species associ- ated with the group for the first time {fra- ternus, vittatus, and versicolor) can there- fore be moved directly into Terralonus without being temporarily sentenced to Metaphidippus. Ghetna new genus Type species, Atttis castaneiis Hentz, 1846, here des- ignated. Name treated as feminine. Description and Diagnosis. These east- ern North American species, like the spe- cies of Terralonus, are ground-dwelling, though in more mesic habitats. They share a dark granulate carapace with fine golden scales, posterior lateral spines on the first tibia displaced anteriorly, reduced spines on the first femora, first coxae nearly touching, and the female palpus slightly swollen. The embolus, at least in the first two species mentioned and perhaps in all, is twisted so as to wind the embolic he- matodocha around the embolus much as in the Bagheera group of genera, though the twisting takes a very different form. Included species. The species Attus cas- taneus Hentz, Metaphidippus barrowsi Kaston, Icius sexmaculatus Banks, and Icius canadensis Banks are here moved to Ghelna to establish the following NEW COMBINATIONS: Ghelna castanea (Hentz), Ghelna barrowsi Kaston, Ghelna sexmaculata (Banks), and Ghelna cana- densis (Banks). Discussion. As with Terralonus, the rel- atives of Ghelna are unclear but, likewise, 240 Bulletin Museum of Comparative Zoology, Vol. 154, No. 4 are not near the genus Metaphidippus. The justification for a new generic name is sim- ilar to that for Terralonus. THE GENUS PELEGRINA FRANGANILLO, 1930 Pelegrina Franganillo, 1930: 44. Type species by original designation and monotypy Pelegrina gen- iculata Franganillo (= P. proxima). Notes on Synonymy. The problem of the generic name to be given to the gal- athea group was not an easy one to solve. In my thesis (Maddison, 1988), I concluded that a new generic name was needed for the group, because (1) it clearly does not belong with Metaphidippus and (2) the group is arguably monophyletic and of re- spectable size for a genus, and (3) it ap- peared that there was no genus whose type species fell within the group. However, subsequent investigation has indicated that an obscure Franganillo name, Pelegrina, should be applied to the galathea group. Although the revival of obscure names is often undesirable, in Pelegrina's case there is little harm because no other published name is available for the group. In 1930, Franganillo described Pelegrina and based it on Pelegrina geniculata Franganillo, which he placed in the section Unidentati, subfamily Heliophaninae. As discussed un- der the description of P. proxima, P. gen- iculata is here considered a junior syn- onym of Dendryphantes proximus G. & E. Peckham; thus, the name Pelegrina is applicable to the galathea group. Description and Diagnosis. Small to medium-sized dendryphantines distribut- ed throughout North and Central Ameri- ca. Males (Fig. 1) typically brown with longitudinal bands of white scales on either side of the carapace and abdomen. The inverted white V-shaped marking on the forehead that contacts the AMEs distin- guishes Pelegrina from most other den- dryphantines, though it is not present in all Pelegrina. Legs often with annulate markings. The relatively wide embolus with the tip expanded retrolateral to its opening and bearing two rami (Figs. 3, 190-216, 220-224) is generally a good di- agnostic feature for the genus, but it is absent in a number of species. Tibial apophysis stout; just ventral to apophysis is usually a ridge (Fig. 78), developed into a second apophysis in some species (Jur- cata group) or a wide flange in other spe- cies (arizonensis group). Females gray, yellow, or brown with mottled markings of four prominent pairs of pale spots on the abdominal dorsum (see, e.g., Figs. 2, 263, 269, 358, 382). Epigynal openings rel- atively long. Among small dendryphan- tines, the species of Pelegrina have per- haps the best-developed epigynal flaps (Figs. 236-255), which are the teardrop- shaped lateral rims of the openings. The flaps are usually convex and overlap the medial rim of opening. All species of Pe- legrina examined have the same chro- mosome complement, 2n6 = 26-1- XXO, as is prevalent throughout the family. Monophyly. Thirty-eight species are in- cluded in Pelegrina. Most of these species can be easily recognized as belonging to the genus by an experienced identifier on the basis of body form, size, and markings, but to articulate precisely characters that could serve as evidence for monophyly is more difficult. The following characters support the monophyly of the genus, though none provides a simple, strict de- limitation. Some of the characters delimit a group slightly smaller than the genus, others a group slightly larger. Thus, each character provides only indirect and par- tial evidence for monophyly. 1. Embolus with two terminal rami retro- lateral to opening (Figs. 3, 190-216, 220-224). The sperm duct opening lies on the prolateral side of the embolus, often below the tip. Retrolateral to the opening are two rami, one just distal to the opening; the other often elongate (see especially P. tristis. Fig. 197) and forming the retrolateral tip of the em- bolus. While other dendryphantines such as Tulpius, Phanias, and Tutelina I Pelegrina Jlimping Spiders • Maddison 241 have accessory rami emerging from the embohis, none have such rami in the position or form seen in Pelegrina. This is perhaps the clearest character dehm- iting the genus, but some Pelegrina ap- pear to lack it (vereciinda, tillandsiae, btinites, orestes, arizonensis, helenae; Figs. 217-219, 225-227), while others have the rami but in a much modified form (flavipedes and related species; Figs. 201-203). These problems are dis- cussed later. 2. Hematodocha of embolus bulges as far distally as the base of the erect portion of the embolus (Fig. 3; Table 2, char- acter 7). This feature is present throughout Pelegrina, including P. or- estes and P. bunites. In other dendry- phantines examined, including the mannii group and Eris, the hemato- docha joins the retrolateral edge of the embolic base more basally so that the hematodocha fails to bulge as far dis- tally (Figs. 22, 23). 3. Epigynal flaps well developed, long, and wide and not descending into the open- ing posteriorly. This character is diffi- cult to assess, for there are other den- dryphantines with well-developed flaps, though in these the flaps do not exactly match those of Pelegrina, being either much shorter (Phidippus, Fig. 67; Bel- lota), narrower {Beata; Figs. 109, 112), or less convex (e.g., "Pseudicius'^ siti- culosus). The flaps of most Pelegrina differ from those of the mannii group and Eris, which have weak flaps that descend into the opening at posterior end (Figs. 66, 256, 257); the flaps in the two most problematical Pelegrina spe- cies, P. bunites and P. orestes, are somewhat like those of the mannii group. 4. Wrinkles present on anterior margin of male cheliceral fang (Fig. 11; Table 2, character 8). Running parallel to the serrate edge of the fang (Hill, 1977b), just anterior to it, is a line of transverse wrinkles, which appears like an irreg- ular secondary serrate edge. In Pele- grina, except P. orestes, it reaches dis- tally to the fang opening. This contrasts with Dendryphantes, the mannii group, and most other dendryphantines, which lack such wrinkles. The only other sal- ticids seen with such a "secondary ser- rula" are (a) Phanias (four species ex- amined), in which the wrinkles are re- stricted to a depression that does not reach the opening; (b) Selimus sp. and '^Beata" octopunctata, in which the wrinkles are long and regular; and (c) Eris militaris, in which the wrinkles do not quite reach the opening. Except for E. militaris, the wrinkles are of differ- ent form, suggesting they are not ho- mologous. 5. Distinct cheek bands on the male face (Figs. 1, 258, 264, and so on). Though other dendryphantines may have pale scales on the side of the face under the ALEs, in most these pale scales do not form a distinct band separated from the side bands by a dark area. Such a sep- aration of the cheek and side bands is also seen in some species of the mannii group (Figs. 493, 514, 534). It is lacking in other dendryphantines except one species from Venezuela examined, pos- sibly an Admirala sp., though the char- acter has not been surveyed intensively. Some species of Pelegrina (P. varie- gata. Fig. 447), however, do not have a distinct cheek band. 6. An inverted V-shaped mark of pale scales on the forehead, contacting the AMEs (Figs. 1, 258, 264, and so on). Most Pelegrina males show this fore- head band, though some lack it (e.g., P. aeneola). While the character has not been surveyed thoroughly, it is lack- ing in other dendryphantines except for a few species (e.g., Tutelina hartii). 7. Male courtship with prolonged "crouch" display. In the courtship of Pelegrina, Eris militaris, the Metaphi- dippus mannii group, and Nagaina in- cunda, there is a crouch display, in which the first legs are held low and forward, usually horizontal and below 242 Bulletin Museum of Comparative Zoology, Vol. 154, No. 4 the level of the body (Figs. 125-128, 134, 140, 162 for Pelegrina; Figs. 124, 178, 184 for Eris and mannii group; Peckham and Peckham, 1889; Rich- man, 1982: fig. 3; Table 2, character 9). The more distal segments may be raised (e.g., Pelegrina kastoni, Fig. 140), but at least the femur is low. The male pro- ceeds toward the female in this pose, waving the first legs at low amplitude if at all. The body is held horizontal, in many species close to the substrate (hence the name "crouch"). This pose contrasts with that seen in most other dendryphantines (Table 2; Peckham and Peckham, 1889, 1890; Crane, 1949a; Richman, 1982; Jackson, 1978; Hill, 1977a; and my own unpublished observations on about 50 species in which the first legs are generally raised and spread (the raisedspread display. Figs. 113-115, 121; e.g., Phidippus, Paradamoetas, Zygoballus) or raised and held forward (the raisedforward display. Figs. 118-120; e.g., Phanias, Sassacus papenhoei, Dendryphantes nigromaculatus) , or held horizontal and spread very wide (the lowspread dis- play. Figs. 116, 117; e.g., Messua lim- bata, Tulpius). The distinctions among these poses are, of course, vague. It should be noted that Pelegrina does have a raisedspread display, generally performed with the carapace held high and the abdomen low, but this is per- formed usually when the male is far from the female. A Pelegrina male thus often begins with a raisedspread display and then proceeds to a crouch display as he approaches the female. Insofar as other dendryphantines often reach with the legs parallel, forward, and low just before mounting and copulation (stage II courtship. Crane, 1949b), as in the crouch display, it may be claimed that the crouch display is merely stage II courtship and, thus, not restricted to Pe- legrina and relatives. Indeed, the crouch display may represent a prolonged stage II. If so, however, it is still distinctive from the brief stage II seen in other salticids, in being much more pro- longed, performed farther from the fe- male, and having a more rigid appear- ance. The crouch display has been ob- served in all Pelegrina and mannii group species examined, with the fol- lowing exceptions: in P. furcata a strik- ingly different display (the semaphore display) is seen, whereas in Pelegrina arizonensis, P. chalceola, and Meta- phidippus diplacis only a raisedspread display was seen. The lack of observed crouch display in these latter species calls attention to the difficulty of using the character. When seen, the crouch display is distinctive and can be con- sidered present. However, when not seen, it may still be characteristic of the species but not observed because the male simply failed to perform it, re- maining longer than usual in the raised- spread display. Nevertheless, the crouch display was scored as absent in these Pelegrina species because the males were observed for a reasonable sample of displays. Another problem with the character is the occurrence of a similar first leg pose in Tutelina and Hentzia (Figs. 122, 123). However, the exact leg poses and motions in these other genera differ in a number of respects from those of Pelegrina. 8. Ridge under tibial apophysis (Figs. 78, 389, 421, 427; Table 2, character 10). Under the tibial apophysis is a ridge that in extreme cases can make the apophysis appear bifid (e.g., Pelegrina bicuspidata). The ridge is present throughout Pelegrina (except fla- vipedes and similar species) but is also present in some species of the mannii group (mannii, diplacis), Beata (Fig. 77), and in Tulpius hilarus. It is poorly developed in Eris militaris (Fig. 79), Bellota wheeleri, and Metaphidippus vitis. It is lacking in other dendryphan- tines including Dendryphantes (Figs. 75, 76). Pelegrina Jumping Spiders • Maddison 243 The preceding characters give reasonable support to the monophyly of the genus Pelegrina, though none provides a strict dehmitation. Even if the genus as consti- tuted here is not monophyletic, the char- acters provide good evidence that it is at least mostly monophyletic, to the extent that monophyly could probably be achieved by including or excluding only a few troublesome species. The following discussion regards the Pelegrina species that fail to show some of the characters supposedly delimiting the genus and why I have chosen to include these species in the genus. Whether or not species in the mannii group might be included in Pe- legrina is discussed in the introduction to the revision of the species of the mannii group. Pelegrina arizonensis and P. helenae. These species lack the two rami on the embolus. Instead, the embolus is blade- shaped, shifted retrolaterally, and concave on the exposed (ventral) surface (Figs. 217, 218, 422, 428). However, there are a num- ber of other characters that would other- wise place the two species within Pelegri- na. These species have distinct cheek and forehead bands, a ridge just ventral to the tibial apophysis, and the male fang with wrinkles as in other Pelegrina. The pe- culiar embolus might be derived from that of a typical Pelegrina embolus by twisting about its longitudinal axis so as to reverse pro- and retrolateral edges and to present the embolus's concave surface, normally facing inward to the cymbium, outward toward the front. This is indicated by the presence of a ridge cutting across the face of the embolus joining the prolateral sur- face of the base with the retrolateral edge of the embolus, the position of the opening on the retrolateral side, the concave ex- posed face of the embolus, and a pro- nounced furrow on the embolis hemato- docha as if folded inward. Cutler and Jen- nings (1985) noted that "internal epigynal structure of his [Proszynski, 1982:1] D[endryphantes]. czekanowskii bears a close resemblance to the internal epigynal structures seen in the M. arizonensis group." Perhaps the similarity they no- ticed was the looping of the duct just inside the opening. The epigynum of D. czeka- nowskii is much like that of D. nigroma- culatus and D. fusconotatus. This looping in both Dendryphantes and the arizonen- sis group is related to the rotation of the epigynal flaps medially at the posterior end. In Dendryphantes, the rotation reaches 90°; in the arizonensis group, it is much more extreme, 180-270°. This rotation must not be considered a critical character; similar rotations are seen in the Pelegrina pervaga group (P. kastoni), in Phidippus octo- punctatus, and in Agassa (as compared to Sassacus). Indeed, in other respects the epigyna of the arizonensis group and Den- dryphantes are rather different, with the flaps being on the surface in the former (as in other Pelegrina, Fig. 251), whereas they descend beneath the opening as a sim- ple ridge in the latter (Fig. 65). Pelegrina flavipedes, P. flaviceps, and P. exigua. The biramous embolus of this group (Figs. 201-203) might be interpret- ed either as arising from an embolus like that of Pelegrina sabinema (Fig. 198) by more deeply splitting the two terminal rami of the embolus or as arising from an em- bolus like that of Dendryphantes rudis and D. nigromaculatus (Figs. 106, 108) by pro- longation of the retrolateral projection on the shoulder of the embolus. The fla- vipedes group lacks a ridge under the tibial apophysis, which otherwise seems char- acteristic of Pelegrina, thus supporting the interpretation that the flavipedes group does not belong with Pelegrina. However, there is more compelling evidence that the flavipedes group is derived from within Pelegrina. Like the pervaga group, the fla- vipedes group members are conifer dwell- ers with yellow chelicerae. Like Pelegrina, they have wrinkles anterior to the fang serrula and an embolic hematodocha bulg- ing distally, cheek bands on the male face, and a crouch display in male courtship (Fig. 127). Like Pelegrina and a few other dendryphantines, the embolic base is well 244 BtiUetin Museum of Comparative Zoology, Vol. 154, No. 4 sclerotized, with few wrinkles over most of its exposed surface. Finally, if the fla- vipedes group were derived from those Dendryphantes with long retrolateral pro- longations on the embolus shoulder (rudis, nigromaculatus) , the group should have the synapomorphy for Dendryphantes, the fold across the embolic base. The fla- vipedes group lacks this fold. Pelegrina furcata. This species has two terminal rami on the embolus, robust epi- gynal flaps, and cheek bands, but its court- ship display is unlike that of any others in the genus. The first legs are held wide and high, unlike Pelegrina, Eris militaris, and the mannii group, but like most other den- dryphantines, and waved in a distinctive semaphore-like fashion (Fig. 121). Pelegrina verecunda. Arizonan speci- mens lack the two distinct terminal rami on the embolus (Fig. 219), but Chihuahuan specimens identified as this species have the rami. Pelegrina tillandsiae. This species lacks the two terminal rami on the embolus (Fig. 225) and is in many respects atypical. It is tentatively included in Pelegrina because its epigynum shows strong flaps that, as in Pelegrina, do not descend into the open- ings (Fig. 254). Pelegrina orestes and P. hunites. These species present the greatest problems with inclusion in Pelegrina, and I might have treated them as belonging to the mannii group or elsewhere. They lack the two ter- minal rami on the embolus characteristic of Pelegrina (Figs. 226, 227), though at least in P. orestes the embolus is obliquely truncated distal to the opening and has one ramus well separated from the opening. On the other hand, the epigynal flaps of hunites are flatter and narrower than in other Pelegrina, more as in the mannii group (Figs. 225, 481). Pelegrina hunites, though, lacks the bulge above the tibial apophysis characteristic of the mannii group and has three characters that might argue for the placement of hunites in Pe- legrina: the occurrence of wrinkles on the cheliceral fang, the bulging of the embolic hematodocha to the shoulder of the em- bolus, and the forehead band contacting the AMEs. On balance, then, a case can be made for tentatively describing hunites in Pelegrina. The situation with orestes is more difficult. Pelegrina orestes lacks the cheliceral fang wrinkles of Pelegrina, and one character, the presence of a ridge on the chelicera (Fig. 483), gives positive ev- idence to place orestes in the mannii group, though the ridge is especially weak and not present in all males. However, orestes lacks the bulge just dorsal to the tibial apophysis characteristic of the mannii group and does have the bulging embolic hematodocha characteristic of Pelegrina. Because of this, orestes will be described in Pelegrina, though it may eventually have to be moved. Natural History. Species of Pelegrina are found in various habitats from the Arc- tic to the tropical lowlands of Central America. While most species in Mexico and Central America appear to occur in the highlands (cloud forest and oak-pine zones), there are some lowland tropical species (P. sandaracina and P. yucate- cana). All species in the genus are pri- marily dwellers on foliage, being only oc- casionally found on the ground. Most other dendryphantines are also foliage dwellers, though some dendryphantines, in partic- ular those with more elongate and dully marked gray or brown bodies, are ground or bark dwellers (Terralonus, Ghelna, some Phanias species). A number of species of Pelegrina appear to be most common on or restricted to particular sorts of plants: the flavipedes group to various confiers, the pervaga group and P. halia to junipers, P. clemata and P. helenae to sagebrush (Artemisia tridentata), and P. tillandsiae to Spanish moss (Tillandsia usneoides). Other species do not appear so specialized to particular plants, yet in my collecting they do seem to prefer certain habitats: P. proterva occurs in forest understory, at least in the south of its range; P. galathea, in fields; P. variegata, in desert scrub; P. montana, in streamside vegetation; and P. Pelegrina Jumping Spiders • Maddison 245 insignis, on low plants in fields and bogs. A number of southwestern species are most commonly collected from oaks. One gen- eralist species is P. aeneola, which is found on trees and herbs of various sorts in the Pacific Northwest, though not usually in the arid regions. The silken retreats and egg sacs are constructed among the foliage on which the adults are collected. PHYLOGENY WITHIN PELEGRINA While insufficient evidence was found to indicate the basal divisions of Pelegrina, the delimitation of a number of smaller groups can be made (see cladogram given in Fig. 129). One clearly delineated group is the flavipedes group, whose three mem- bers (flavipedes, flaviceps, and exigua) share the following characters derived within the genus: 1. Embolus deeply bifid (Figs. 201-203). All other species of Pelegrina have the division between the terminal rami of the embolus not nearly so deep as in the flavipedes group. Other dendry- phantines have either a simple embolus or one with accessory rami different from those in Pelegrina and the fla- vipedes group. 2. Chelicerae of male yellow with dark spot in medial concavity (Figs. 319, 324, 329, 334). The dark concavity is unique to this group. 3. Asymmetrical circling of palps in male courtship. During the crouch display, the palps are waved fairly slowly at high amplitude in circles such that as one is rising the other is falling (Fig. 127). Though other Pelegrina may wave their palps out of phase, in none examined is this asymmetrical waving made so obvious by the large slow waves. No other dendryphantines examined have such waving, except some Phidippus. Another group similarly well defined is the pervaga group, consisting of pervaga, sabinema, and kastoni, which share the following: 1. Distinct markings on yellowish male palpus, consisting of prominent patches of white scales on the femur, tibia, and cymbium, interrupted by dark hairs on the patella. Other Pelegrina have no white scales on the tibia, or fewer on the tibia than on the patella. 2. Cheek band extended posteriorly par- allel to the side band, separated from the side band by dense band of dark hairs (Figs. 304, 309, 314). This yields the appearance of white-black-white carapace side bands. Other Pelegrina have dark setae between side and cheek bands, but in none is the cheek band so horizontal and the dark hairs so dense. One species in the mannii group, Me- taphidippus emmiltus, has a superfi- cially similar pattern. These two groups, the flavipedes and pervaga groups, may together form a monophyletic group, as delimited by the following: 1. Chelicerae yellow in males. Other Pe- legrina have dark brown chelicerae ex- cept southern males of P. tillandsiae. Brown chelicerae are present in other similar dendryphantines such as Eris, Nagaina, Beata, and the mannii group except Metaphidippus emmiltus, which also has yellow chelicerae. 2. Conifer dwelling. All members of the flavipedes group, and P. pervaga and P. kastoni, are known to dwell more or less exclusively on conifers. The habitat of P. sabinema is unknown. This con- trasts with the habitat of most other Pelegrina, which inhabit broadleaf trees, shrubs, and herbs. However, P. proterva, P. aeneola, and P. furcata are known to frequent conifers, whereas P. halia appears restricted to conifers. Also, the polarity of this character is unclear. Outside the genus, Metaphidippus em- miltus is a juniper dweller, whereas Eris species are often collected from coni- fers. The neoleonis group, including tristis 246 Bulletin Museum of Comparative Zoology, Vol. 154, No. 4 from Arizona and neoleonis from Mexico, is distinguished by the broad, dark rotated epigynal flaps. Though their appearance is much more Hke that of typical brown- legged Pelegrina, they share the following characters with the preceding two groups of yellow-legged species: 1. Retrolateral ramus of embolus much elongate and curled to the prolateral (Figs. 196-203). The only other Pele- grina with such a long ramus is the furcata group, though in that group all but one species have it curling to the retrolateral. The exception is morelos, which has a prolateral curl (Fig. 215), but because it seems very close to fur- cata the prolateral curl is probably con- vergent. 2. Embolus very broad. Such a broad em- bolus is not found in other Pelegrina except peckhamorum. 3. First curve of duct of epigynum broad. Though unusual, a duct as broad is found in P. huachuca and P. morelos. The relationships of this proposed fla- vipedes-pervaga-neoleonis clade are not altogether clear, though the sickle-shaped retrolateral ramus of the embolus of pro- terva, galathea, edrilana, and pallidata may be viewed as a preliminary version of the very long ramus of this clade. Most of these species together with dithalea, proxima, and peckhamorum have an angle on the embolus just basal to the tip of the retrolateral ramus (Figs. 259, 265, 271, 277, 283, 289, 310) which may be a synapo- morphy for a large clade (Fig. 129), but whether or not the angle is absent from all other Pelegrina is unclear (e.g., see Fig. 215). Another clearly delineated group is the arizonensis group (Cutler and Jennings, 1985), including the two species P. ari- zonensis and P. helenae. The apomorphies supporting the group are as follows: 1. Epigynal flaps far rotated, at least 180° (Figs. 424, 430). No other dendryphan- tines known to me have a similar ro- tation; flap rotations in Terralonus and Ghelna, for example, are in the oppo- site direction. 2. Erect portion of embolus displaced to retrolateral side and tegulum rotated somewhat clockwise (Figs. 422, 428). The embolus displacement is seen in some other dendryphantines, but the tegulum rotation is perhaps unique. 3. Embolus twisted about longitudinal axis so as to reverse pro- and retrolateral edges and to present the back side for- ward. 4. Ridge under tibial apophysis uniquely developed into flange (Figs. 421, 427). 5. Markings of female abdomen some- what lineate (Figs. 425, 431). Lineate markings are also seen in P. tillandsiae and the Metaphidippus mannii group. The furcata group of Central America and southwestern North America includes the widespread and common P. furcata as well as a number of rare species (huachu- ca, morelos, bicuspidata, volcana, ochra- cea). Apomorphies supporting the group are as follows: 1. Ridge under tibial apophysis unusually strongly developed, so as to form a sec- ond apophysis (Fig. 389; even stronger in the other species in the group). This is found in all species, although it is lacking in some specimens of furcata. 2. Wrinkles on the retrolateral basal edge of the embolic base either transverse or ascending apically toward the retrola- teral (Figs. 390-394, 404, 406, 411, 416). This contrasts with the wrinkles of other Pelegrina and similar dendryphan- tines, which have wrinkles descending basally toward the retrolateral. Pele- grina insignis may have wrinkles sim- ilar to those in the furcata group. 3. Epigynal flaps very convex, unlike oth- er dendryphantines except Pelegrina proxima and peckhamorum. 4. Epigynum concave behind flaps, unlike the case in Eris, the mannii group, and Nagaina, though also seen in Pelegrina proxima, peckhamorum, and balia. Pelegrina Jumping Spiders • Maddison 241 Pelegrina furcata itself has a very pe- culiar courtship display. Whether or not this feature is shared by other members of this group awaits their examination. The montana group includes three spe- cies: montana, msigriis, and chaimona. These species share the following: 1. Concavity on back of embolus restrict- ed to distal half of erect portion. In other Pelegrina species, the concavity on the back of the embolus (Figs. 7, 34, 35) extends from the base of the erect portion to its tip. Some Pelegrina, how- ever, have no clear concavity at all (e.g., P. aeneola, P. chalceola). 2. Small, sharp denticles on front surface of embolus (Figs. 204-206). Other den- dryphantines have denticles on the em- bolus, but they are usually on the re- trolateral surface or are of different form. This character has not been well surveyed, however. Among the remaining species of Pele- grina are many that have a narrow em- bolus with small rami {aeneola, clemata, chalceola, halia, variegata, verecunda, sandaracina, and others). However, this may be the primitive condition for the genus. No clear subgroups were found among these species. IDENTIFYING SPECIES OF PELEGRINA AND THE METAPHIDIPPUS MANNII GROUP In general, the genitalia provide the best means of identifying species, but facility in recognizing the distinguishing features may require some experience. Males are much more easily identified than females, as the palpus and face markings provide more readily described and interpretable differences than the differences in epi- gyna. Take note especially of the width of the erect portion of the embolus and the size and orientation of the two terminal rami. Indeed, it is usually possible to iden- tify males simply by referring to the two pages of Figures 190-235. A single key for all species is given for males. Females, however, pose many more problems for identification. Though the abdominal markings and epigyna of fe- males vary in a number of features, the differences can be subtle and difficult to describe. One might think of the abdom- inal markings as falling in two major cat- egories: those in which the paired white spots dominate the dorsum (Figs. 263, 275, 293, 382, 441, 451) and those in which the dark patches between and beside the white spots dominate the dorsum (Figs. 281, 287, 347, 358, 364, 377, 387). Epigynal features to note are the topography of its surface and the size, convexity, color, and place- ment of the teardrop-shaped flaps cover- ing the openings. Even once experienced with these characters, an identifier can still have difficulties with some specimens. The problems are lessened within a given geo- graphical region. Because of this, separate female keys are given for five regions: the eastern United States and Canada, the Great Plains, Pacific Coast Untied States and western Canada, Arizona and Mexico, and Central America. Parts of the south- western United States are therefore with- out a key to females of Pelegrina, namely, Texas and the Rocky Mountain states. For Montana and Wyoming, the Pacific Coast key can be used (except possibly for prairie species). For Colorado, Utah, and New Mexico, most identifications can be accom- plished using the Pacific Coast and Ari- zona keys, though the Great Plains key will be needed occasionally. For Texas, the Ar- izona and Great Plains keys will usually suffice, but the eastern United States key will be needed on occasion. In general, mannii group females are not included in the keys. Metaphidippus mannii is in- cluded in the Pacific Coast key, but five other species in the group that occur in southern California are not included; man- nii and chera are in the Arizona key, but carmenensis is excluded. The keys are written for adult speci- mens, but the keys for females will be of some aid to identifying immatures based on markings. A key for immature Pele- 248 Bulletin Museum of Comparative Zoology, Vol. 154, No. 4 grina from Minnesota has been given by Cutler (1981b). Figures 258-538 provide the most com- prehensive set of illustrations of the spe- cies, but important aid can be obtained from Figures 130-189, which show living specimens. Figures 190-235, which sum- marize the male emboli, and Figures 236- 257, which show the surface topography of the epigyna. Key to the Males of All Species of Pelegrina and Those Metaphidippus mannii Group Species Occurring in the United States* 1. Erect portion of embolus extremely nar- row or tapers to point, lacking two ter- minal rami (Figs. 226-235); lateral margin of chelicera usually with ridge near base of fang (Figs. 483, 493, 503, 509, 514, 529); at least small patch of white or orange scales on chelicerae; western United States and southwest- ern Canada, south into Central Amer- ica {Metaphidippus mannii group, in part, and some Pelegrina) 2 Erect portion of embolus in most species wide at tip and with two terminal rami (Figs. 190-225); chelicera lacking ridge near base of fang (e.g.. Figs. 258, 264); scales on chelicerae varied; widely dis- tributed 10 2(1). Long patch of white scales on chelicerae, longer than V2 length of chelicerae (Figs. 478, 493, 514, 529) 3 White or orange patch small (Figs. 483, 503, 509) 7 3(2). Embolus wide at base of erect portion (Figs. 226-231) 4 Embolus very narrow (Figs. 516-523, 535) '. 6 4(3). Forehead white band lacking so that se- tae above AMEs are dark (Figs. 493, 503); retrolateral edge of base of em- bolus with prolongation (Figs. 494, 499, 505) 5 White forehead band present and con- tacting AMEs above (Fig. 478); retro- lateral edge of base of embolus lacking * The mannii group species in Me.xico and Central America are not included; they can be distinguished from Pelegrina by their narrower embolus tip, which lacks the two rami. Nagaina incunda, described later, is brown and yellow striped with the first legs brown and the posterior legs yellow (Fig. 174). prolongation (Fig, 479); central Ari- zona south to Oaxaca .. 37. biinites (part) 5(4). Carapace and abdomen shiny dark or coppery brown contrasting strongly with dense white cheek band and chel- iceral patches, with small or no white side bands (Figs. 178, 180, 493); pro- longation on retrolateral edge of base of embolus blunt or small (Fig. 494) where sympatric with diplacis; British Columbia to California east to Idaho and central Arizona 40. mannii Carapace and abdomen with more ex- tensive white side bands (Figs. 182, 503); cheek band not distinct from side band (Fig. 503); prolongation of retro- lateral edge of base of embolus distinct and long (Figs. 504, 505); along Pacific Coast of southern California and Baja California 41. diplacis (part) 6(3). Erect portion of embolus straight (Fig. 233); face dark under eyes (Fig. 514) 43. chera Erect portion of embolus curves ven- trally (Fig. 234); face extensively cov- ered with white scales (Fig. 529) ex- cept in Baja California Sur 44. carmenensis 7(2). Erect portion of embolus obliquely trun- cated, broad at base (Figs. 227, 484); carapace and abdomen dusted with beige to light brown scales (Figs. 172, 483); Arizona and Mexico 38. Orestes (part) Erect portion of embolus not so broad or truncated (Figs. 505, 510); markings dark brown with white (Figs. 182, 184, 503, 509) or mostly yellow (Fig. 176) 8 8(7). Legs yellow, first legs fringed with white (Fig. 176); anterior median eyes ringed with red; chelicerae vertical and rel- atively weak (Fig. 534); erect portion of embolus very thin (Fig. 535); south- ern California to New Mexico 45. emmiltus (part) Legs with dark brown markings (Figs. 182, 184, 503, 509); anterior median eyes ringed with dark setae; chelicerae at least slightly divergent; embolus thick or thin; Pacific Coast of Califor- nia and Baja California 9 9(8). Embolus wide at base of erect portion (Figs. 231, 504, 505); scales on chelic- erae white; body with bronze reflec- tions (Fig. 182); prolongation on re- trolateral edge of base of embolus large and distinct (Figs. 504, 505); southern California and Mexico 41. diplacis (part) Embolus thin (Figs. 232, 510); scales on Pelegrina Jumping Spiders • Maddison 249 Maps 1-7. Distributions of Pelegrina species. 1 . Pelegrina galathea in North and Central America. 2. P. proxima in the Caribbean. 3. P. dithalea in Arizona. 4. P. edrilana in Mexico. 5. P. proterva in North America. 6. P. neoleonis in Mexico. 7. P. fnsf/s in Arizona. 250 Bulletin Museum of Comparative Zoology, Vol. 154, No. 4 Maps 8-15. Distributions of Pelegrina species. 8. P. peckhamorum in the eastern United States. 9. P. sabinema in the southwestern United States. 10. P. pervaga in the central United States. 11. P. kastonitn the southwestern United States. 12. P flavipedes in North America. 13. P. flaviceps in eastern North America. 14. P. exigua in the eastern United States. 15. P. clavator in Mexico. Pelecrina Jumping Spiders • Maddison 251 chelicerae orange; body dark and dull (Fig. 184); prolongation on retrolateral edge of base of embolus small or absent (Fig. 510); central and northern Cali- fornia 42. tricolor 10(1). Clypeus with patch of white or yellow scales between AMEs (not merely hairs overhanging chelicerae or circling an- terior median eyes) 11 - Clypeus without white or yellow setae between AMEs, except perhaps for hairs surrounding anterior median eyes and overhanging chelicerae 17 11(10). Chelicerae with large white or yellow patch of scales medially, at least for Vz length of chelicerae (Figs. 159, 437, 442, 457); Mexico and Central Amer- ica _ _ 12 i2(ii; 13(12) 14(11) 15(14) 16(14) 17(10). 18(17) Chelicerae lacking pale scales (Figs. 282, 288, 319, 334); United States east of Rocky Mountains and Canada 14 Retrolateral ramus of embolus much lon- ger than prolateral (Fig. 221) 32. pallidata Retrolateral ramus of embolus small, subequal to prolateral (Figs. 220, 224) 13 Pale markings usually yellowish; embo- lus small and tapers to tip (Fig. 224); lowland 35. sandaracina Pale markings white; embolus broadly truncate (Fig. 220); montane 31. elevator (part) Chelicerae yellow (Figs. 319, 334); em- bolus deeply divided (Figs. 320, 330); cymbium lacks white scales; dwellers on conifers 15 Chelicerae brown (Figs. 282, 288); em- bolus not deeply divided (Figs. 260, 266); cymbium with dorsal patch of white scales; habitat varies 16 First tibia yellow or with thin black stripe; retrolateral ramus of embolus thick, only slightly thinner than prolateral (Figs. 201, 320) 12. flavipedes First tibia dark; retrolateral ramus much thinner than prolateral (Figs. 203, 330) 14, exigua (part) Embolus tapers to tip (Fig. 284), with long hooked retrolateral ramus (Fig. 283) 5. proterva Embolus very broad at tip (Fig. 290), retrolateral ramus short (Fig. 289) 6. peckha moru m Chelicerae yellow (Figs. 304, 309, 314, 324, 329, 534); dwellers on conifer and Spanish moss Chelicerae brown (e.g.. Figs. 258, 264) 24 Medial black spot on chelicerae (Figs. 324, 329); embolus deeply divided (Figs. 325, 330); eastern and central United States and Canada _ 19 - No medial black spot on chelicerae (Figs. 304, 309, 314, 534); southern United States and northern Mexico 20 19(18). Forehead flat (Fig. 329); forehead dark brow n in alcohol; bod\ and legs brown (Fig. 146); chelicerae yellow laterally (Fig. 329); southeastern United States north to Massachusetts and New York 14. exigua (part) Forehead bulbous (Fig. 324); forehead yellow in alcohol; body and legs pale (Fig. 144); chelicerae with dark spot laterally (Fig. 324); northeastern Unit- ed States and southeastern Canada 13. flavieeps 20(18). Erect portion of embolus very thin (Fig. 535); AMEs ringed with red; first legs fringed with white (Fig. 176) 45. emmiltus (part) - Erect portion of embolus thicker (Figs. 305, 310, 315, 474); AMEs ringed with white or brown; legs not fringed 21 21(20). Cymbium yellow; band of dark setae un- der carapace side band (Figs. 304, 309, 324); embolus wide at tip (Figs. 305, 310, 315); dwelling on conifer; Kansas west to Arizona 22 Cymbium dark distally; no band of dark setae under carapace side band (Fig. 472); embolus tapers to narrow tip (Fig. 474); dwelhng on Spanish moss, Flor- ida and North Carolina west to Texas 36. tillandsiae (part) 22(21). Clypeus brown (Figs. 304, 309); retro- lateral ramus of embolus long (Figs. 198, 199, 305, 310); embolus broad at base of erect portion 23 Clypeus with white band except cen- trally (Fig. 314); retrolateral ramus of embolus short (Fig. 200); embolus rectangular, narrow, and displaced re- trolaterally (Fig. 315) 11. kastoni 23(22). Abdomen brown above; embolus (Fig. 305) wider than in pervaga 9. sabinema Abdomen with central longitudinal pale stripe as in females (Fig. 313); embolus (Fig. 310) narrower than sabinema 10. pervaga 24(17). Ridge under tibial apophysis usually de- veloped into acute second apophysis (Fig. 389); wrinkles on embolic base transverse or ascending apically to- ward the retrolateral edge (Figs. 390, 404, 406, 411, 416); southwestern United States to Panama (Jurcata group) 25 At most small ridge or broad flange un- der tibial apophysis (Figs. 78, 421, 427); wrinkles on embolic base descending basally toward the retrolateral edge 252 Bulletin Museum of Comparative Zoology, Vol. 154, No. 4 Pelegrina Jumping Spiders • Maddison 253 29 26 27 (e.g.. Figs. 315, 438); widely distrib- uted _ 25(24). Rami of embolus small and subequal (Figs. 213, 214) Retrolateral ramus of embolus long, much longer than prolateral (Figs. 212, 215, 216) 26(25). White patches on chelicerae extend to at least 1/2 their length (Fig. 403); embolus more or less straight (Figs. 213, 404); Panama 23. volcano - White patches on chelicerae small (Fig. 405); embolus bent to retrolateral (Figs. 214, 406); Guatemala and Mexico 24. bicuspidata 27(25). Retrolateral ramus curls to prolateral (Figs. 215, 411) 26. morelos Retrolateral ramus points retrolaterally or distally (Figs. 212, 216, 390-394, 416) 28 28(27). Retrolateral ramus points more distally, about twice as long as prolateral ramus (Figs. 212, 390-394); widely distrib- uted 22. fiircata (part) - Retrolateral ramus points more retrolat- erally, more than four times longer than prolateral ramus (Figs. 216, 416); southern Arizona 27. huachuca 29(24). Large white patches on chelicerae at least to V2 their length (Figs. 437, 447, 452, 478); embolus twists to tip; Arizona, Mexico, and Central America 30 - Chelicerae with at most a small medial- basal patch of scales (e.g.. Figs. 258, 348); embolus varied; widely distrib- uted geographically 33 30(29). Embolus broad and truncated (Figs. 220, 438); dorsum of abdomen mostly brown between side bands (Fig. 164); montane 31. clavator (part) - Embolus narrower, tapers to tip; dorsum of abdomen may have large white spots (Fig. 166); varied habitats 31 31(30). Embolus tapers to narrow tip with ter- minal opening (Fig. 226); abdominal dorsum brown between white side bands (Fig. 170); montane habitats 37. bunites (part) - Embolus tip wider, opening subterminal (Figs. 223, 224); abdominal dorsum with mixed pale and dark spots (Fig. 166) as in female; deserts and tropical lowlands „ _.... 32 32(31). Side and cheek bands fused (Fig. 447); chelicerae robust; embolus appears to taper in ventral view but in an oblique view the two small subecjual rami are easily seen (Fig. 222); abdomen with strong white spots (Fig. 166); arid regions of Mexico and Central Amer- ica _ 33. variegata (part) Side and cheek bands separate (Fig. 452); chelicerae narrower; embolus weaker with two rami not easily visible; ab- domen with unusual transverse bands (Fig. 169); seasonal tropical forests of Yucatan Peninsula 34. yucatecana 33(29). Embolus with long retrolateral ramus (e.g., Figs. 196, 197, 209); western United States and Mexico _ 34 Embolus with short retrolateral ramus (e.g.. Figs. 190, 206, 210), or rami not distinct (e.g.. Figs. 219, 225); widely distributed 38 34(33). Retrolateral ramus curled prolaterally (Figs. 196, 197); embolus very broad at base 35 Retrolateral ramus erect or pointing pro- laterally (Figs. 193, 209, 212); embolus narrower at base _. 36 35(34). Prolateral ramus of embolus obtuse (Figs. 196, 259); retrolateral ramus blunt and with bump (Fig. 295); embolus nar- rower at base than in iritis (Fig. 196); Mexico 8. neoleonis Prolateral ramus acute (Figs. 197, 300); retrolateral ramus sharp; embolus broader at base (Fig. 197); Arizona - 7. tristis 36(34). Side and forehead bands on carapace re- duced or absent (Figs. 156, 365); west- ern United States and Canada 19. aeneola (part) - Side and forehead bands on carapace well developed (Figs. 158, 276, 388); south- western United States, Mexico, and Central America 37 37(36). Retrolateral ramus of embolus longer and diverging from prolateral (Fig. 212); widely distributed 22. jurcata (part) - Retrolateral ramus of embolus vertical (Fig. 193); embolus narrows abruptly near tip; central Mexico _ __ 4. edrilana (part) 38(33). Erect portion of embolus arises on retro- lateral side (Figs. 422, 428); flange un- der tibial apophysis (Figs. 421, 427; arizonensis group) 39 Maps 16-20. Distributions of Pelegrina species. 16. P. montana in North America. 17. P. insignis in Nortti America. 18. P. chaimona in Mexico and Arizona. 19. P. ciemata in western North Amehca. 20. P. balls and P. chalceola in western North America. 254 Bulletin Museum of Comparative Zoology, Vol. 154, No. 4 Map 25 y Map 27 Pelegrina Jumping Spiders • Maddison 255 Erect portion of embolus arises centrally; ridge under tibial apophysis not de- veloped into flange (Fig. 78) 40 39(38). Embolus tip sharp (Fig. 422); tibial, apophysis flange broad and short (Fig. 421) 28. arizonensis Embolus tip blunt (Fig. 428); tibial apoph) sis flange narrow and elongate (Fig. 427) 29. helenae 40(38). Chelicerae lacking pale scales 41 Chelicerae with small patch of white or yellow scales 46 41(40). Embolus narrows abruptly just basal to opening (Fig. 190); retrolateral ramus is small hook (Fig. 259); southern On- tario, eastern United States south to Central America 1. galathea Embolus parallel-sided or widens near tip; retrolateral ramus is small angle or apparently absent; Canada and mountainous areas of United States 42 42(41). Retrolateral ramus of embolus longer than prolateral and leaning retrolater- ally (Figs. 208, 209); forehead band absent 19. aeneola (part) Both rami of embolus small; forehead band present or absent 43 43(42). Embolus swollen near tip (Figs. 204, 344) 15. montana Embolus with sides parallel or slightly tapering near tip (Figs. 207, 210, 211) _ 44 44(43). Forehead band well developed and con- tacting AMEs (Figs. 152, 359); tegul- um with prominent prolateral bump (Fig. 361) 18. clemata (part) - Forehead band absent or if present then not contacting AMEs (Figs. 378, 383); tegulum with at most small prolateral bump (Figs. 379, 384) 45 45(44). Cheliceral fang with flange (Fig. 378); carapace side bands broad (Fig. 378), embolus bends slightly (Figs. 210, 379); California and northern Arizona north to Washington 20. balia Cheliceral fang lacking flange (Fig. 383); side bands narrower (Fig. 383); em- bolus straight (Figs. 207, 211); south- ern Arizona to southern Illinois 2 1 . chalceola 46(40). Abdomen with striking lineate markings as in female (Fig. 477); embolus ta- pering to sharp tip in ventral view; living on Spanish moss 36. tillandsiae (part) - Abdomen brown or spotted above; em- bolus usually broad at tip though var- ies; habitat varied 47 47(46). Erect portion of embolus very thin (Fig. 535); anterior median eyes ringed with red; first legs yellow fringed vvitli white (Fig. 176) 45. emmiltus (part) Erect portion of embolus thicker; eyes ringed with white or brown; legs not fringed 48 48(47). Abdomen with paired black spots on brown dorsum; pale markings yellow- ish; embolus long and rectangular, leaning slightly retrolaterally (Figs. 205, 349, 350) truncate at tip and with retrolateral ramus apparently absent (Fig. 205); Canada and northeastern United States 16. insignis Abdomen lacking distinct black spots; pale markings white or yellowish; em- bolus shorter (Fig. 219) or if long, then straight and with more prominent rami (e.g.. Figs. 192, 206, 207, 222) 49 49(48). Abdomen with large and distinct paired white spots as in female (Fig. 166); side and cheek bands fused (Fig. 447); che- licerae robust (Fig. 447); erect portion of embolus parallel-sided and with two subequal rami (Fig. 222); deserts of Mexico and Central America 33. variegata (part) - Abdomen with only small (Figs. 130, 132, 152) or indistinct (Figs. 162, 172) paired pale spots, side and cheek bands separate; chelicerae not so robust; em- bolus varied 50 50(49). Side bands of carapace and abdomen weak or absent and abdomen mottled (Figs. 162, 172); embolus small, lack- ing two distinct rami (Figs. 219, 227) 51 - Side bands well developed and abdomen not mottled (Figs. 130, 132, 152) 52 51(50). Pale markings white or gray; erect por- tion of embolus widens gradually on prolateral Iside as it contacts basal por- tion (Fig. 433); tip of embolus rounded (Fig. 219) 30. vereciinda - Pale markings orange or tan; erect por- tion of embolus widens abruptly on prolateral side as it contacts basal por- Maps 21-27. Distributions of Pelegrina species. 21. P. aeneola in western North America. 22. P. furcata in Mexico and the southwestern United States. 23. P. furcata group members in Mexico and Central America. 24. P. huachuca in Arizona. 25. P. arizonensis and P. helenae in western North America (see Cutler and Jennings, 1985, for additional records). 26. P. verecunda in western North America. 27. P. tillandsiae in southeastern North America. 256 Bulletin Museum of Comparative Zoology, Vol. 154, No. 4 Pelegrina Jumfinc; Spidehs • Maddison 257 tion so as to make a distinct corner (Fig. 484); tip of embolus pointed ret- rolateral to opening (Fig. 227) _ 38. orestes (part) 52(50). Retrolateral ramus longer than prolater- al (Figs. 190, 193) 53 Rami subequal (Figs. 191, 192, 206, 207) 54 53(52). Embolus sides parallel below opening (Fig. 190); retrolateral ramus narrow (Fig. 259); widespread 1. galathea (part) - Embolus inflated below opening (Fig. 193); retrolateral ramus wide (Fig. 277); central Mexico 4. edrilana (part) 54(52). Abdomen with two central broken lon- gitudinal pale stripes in addition to side bands as in female (Fig. 269); Carib- bean 2. proxirna Abdomen lacking central longitudinal bands; western North America 55 55(54). Embolus parallel-sided or tapers slightly to tip (Figs. 207, 360, 361); basal to opening the prolateral side is straight (Fig. 207); sagebrush of western Un- tied States and Canada _ 18. clemata (part) Embolus widens slightly near tip (Figs. 192, 206); just basal to opening on pro- lateral side is angle (Figs. 192, 206); Arizona and Mexico 56 56(55). Rami of embolus well separated (Fig. 192); side bands have extensions join- ing between posterior eyes (Fig. 132); no denticles on exposed surface of em- bolus (Fig. 192) 3. dithalea Rami close together (Fig. 206); side bands without extensions; surface of embolus with denticles (Fig. 206) 17. chaimona Key to the Female Pelegrina of the Eastern United States and Canada (East of the Mississippi River and Manitoba) 1. Posterior margin of epigynal flap truncated so as to be transverse, and standing high over surface behind it (Figs. 236, 262); epigynal flaps convex, parallel; legs dis- tinctly annulate (Fig. 131); abdomen marked with four pairs of prominent white spots with small black spots behind them (Figs. 131, 263) _ 1- galathea - Posterior margin of epigynal flap rounded, not transverse, or if transverse then flaps flat and flush with surface Ix'hind them (Figs. 238, 239, 241-245, 254); legs not distinctly annulate; abdomen w ith white spots smaller, often thinner and elongate (e.g.. Figs. 287, 323, 353) 2 2(1). Abdomen with prominent paired black spots on orange-brown background (Figs. 161, 353); epigynal flaps divergent (Fig. 352); epigynal surface rises dramatically from low area around flaps to high pos- terior margin (Fig. 245); legs and face yellowish; mostly northern 16. insignis - Abdomen lacking prominent paired black spots though may have brow n or reddish patches; epigynal flaps parallel, conver- gent, or divergent; epigynal surface in most species with little relief (e.g., Figs. 238, 241); legs and face varied; locality varied 3 3(2). Epigynum with ridge just behind each flap (Figs. 244, 346); posterior notch often rectangular; body large and dark with very small paired white spots on dark abdominal dorsum (Fig. 347); Canada and mountains of United States _ _ _ 15. montana Epigynum lacking ridges behind flaps; pos- terior notch triangular; body smaller; ab- domen varied 4 4(3). Abdomen strongly striped longitudinally yellow and brown (Fig. 477); epigynal flaps pale; living on Spanish moss of the southeastern United States 36. tillandsiae Abdomen not striped yellow and brown longitudinally, usually spotted; epigynal flaps varied _ - 5 5(4). Epigynal surface and flaps very flat (Figs. 24i-243); flaps not much darker than rest of epigynum except for narrow rim (Figs. 322, 327, 332); carapace often with shiny scales and pale spot above and between anterior median eyes (e.g., Fig. 143); co- nifer dwellers {flavipedes group) 6 Epigynal surface and flaps with more relief (Figs. 237, 238); flaps usually distinctly darker than rest of epigynum (Figs. 286, 292); carapace lacking shiny scales .._ 8 6(5). Forehead dark above and between .\MEs; head often bulbous; legs pale yellow, usu- ally with thin longitudinal dark lines on Maps 28-37 Distributions of species of Pelegrina. the Metaphidippus mannll species group, and Nagainaincunda^ 28. P. variegata in Mexico and Central America. 29. P. pallidata, P. yucetecana. and P. sandaracina in Mexico and Central America. 30 Pbunites and P. orestes in Arizona and western Mexico. 31 . M. mannim western North Amenca. 32. M. tricolor n Califom a. 33' M. diplacis in California and Baja California. 34. M. carmenensis in Mexico and the southwestern United States. 35^ M. ctiera in Mexico and the southwestern United States. 36. P. emmlltus in California and New Mexico. 37. Nagaina incunda in Mexico and Central America. 258 Bulletin Museum of Comparative Zoology, Vol. 154, No. 4 femora; epigynum with second curve of spermathecal duct wider than in fla- vipedes but no so wide as in exigua (Figs. 340, 341); northeastern United States bordering Canada and southeastern Can- ada 1 3. flaviceps - Forehead with pale spot above and be- tween anterior median eyes; head not bulbous; legs generally lacking longitu- dinal lines or if persent then wide and mostly on anterior legs; epigynum oth- erwise; distribution generally farther north or farther south than flaviceps 7 7(6). Epigynal flaps parallel (Figs. 241, 322); sec- ond curve of spermathecal duct narrow and oblique (Figs. 321, 338, 339); cara- pace narrow; mostly northern (Canada and northern United States) though found occasionally on southern mountains 12. flavipedes - Epigynal flaps convergent (Figs. 243, 332); second curve of spermathecal duct very broad and transverse (Figs. 331, 336, 337) carapace broader; mostly southern Unit- ed States north to Massachusetts and New York 14. exigua 8(5). Abdomen marked with large square brown spots on or between paired pale spots (Figs. 2, 135, 287); epigynal flaps con- vergent and fairly flat, short (Figs. 238, 286); surface rises quickly behind flaps to broad mound (Fig. 238) 5. proterva Abdomen uniformly light brown with small white spots (Figs. 137, 293); epigynal flaps long and fairly convex (Figs. 239, 292); surface rises gradually behind flaps to mound at posterior (Fig. 239) 6. peckhamorum Key to the Female Pelegrina of the Great Plains (between the Rocky Mountains and the Mississippi River)* 1. Epigynal flaps rotated 180° (Fig. 424); ab- domen with strongly lineate markings (Fig. 425) 28. arizonensis - Epigynal flaps rotated at most 45°; abdom- inal markings not so clearly lineate 2 2(1). Legs distinctly annulate, and abdomen marked with four pairs of prominent white spots with small black spots behind them (Figs. 131, 263); epigynal flaps con- vex, parallel, posterior margin truncated so as to be transverse and standing high over surface (Figs. 236, 262) 1. galathea - Legs not distinctly annulate; abdomen with more prominent dark areas on either side of smaller pale spots (though pervaga with pale spots coalesced into single large spot); epigynal flaps varied, but posterior mar- gin not truncated 3 3(2). Area behind epigynal flaps raised into high mound (Figs. 245, 246); carapace densely covered with white or yellow scales (Figs. 153, 161 ) 4 - Area behind epigynal flaps more or less flat (Figs. 238, 240); carapace thinly covered with pale scales (e.g.. Fig. 135) 5 4(3). Scales on carapace white; legs beige and brown; abdomen with large dark patches on either side of central paired spots but lacking strong black spots (Fig. 364), an- teriormost pale spots fused into short lon- gitudinal bands; epigynal surface behind flaps raised into broad dark shiny round mound (Fig. 246); flaps convergent (Figs. 246, 363); usually collected from sage- brush 18. clemata Scales on carapace yellowish; legs yellow; abdomen with paired black spots (Fig. 353); epigynal surface behind flaps raised gradually but steeply into high mound along posterior of epigynum (Fig. 245); flaps divergent (Figs. 245, 352); low herbs in fields and bogs 16. insignis 5(3). Abdomen marked with large square brown spots between small paired pale bands (Figs. 135, 287); epigynal flaps conver- gent and fairly flat, short and dark (Figs. 238, 286); surface rises quickly behind flaps to broad mound (Fig. 238); cara- pace narrow; widespread (5) proterva - Abdomen with large central pale spot (Fig. 313); epigynal flaps large and flat, fairly pale (Figs. 240, 312); carapace very wide; Kansas to Texas 10. pervaga Key to the Female Pelegrina of the Pacific Coast of the United States and Western Canada* Epigynal flaps rotated 270° so that flaps are transverse (Fig. 430); abdomen with li- neate markings (Figs. 155, 431); com- monly found on sagebrush 29. helenae Epigynal flaps rotated less than 45°; abdo- * Not included are some tree-dwelling species whose ranges reach into the Great Plains: Pelegrina fla- vipedes and exigua, which occur on conifers in the north and east; P. peckhamorum, on oaks in the south- east; and chalceola, in Texas to extreme southern Illinois. * Includes California, Nevada, Oregon, Washing- ton, Idaho, British Columbia, and Alberta. Included is Metaphidippus mannii, as well, but not the other mannii group species, which are restricted to the southern part of the area of the key. Not included is P. verecunda (see Arizona key). Pelegrina Jumping Spidkrs • Maddison 259 men with markings not lineate except occasinally in cleniata, habitat varied 2 2(1). Epigynal surface and flaps very flat (Figs. 241, 256); body with shiny bronze or cop- per scales (Figs. 143, 179) 3 - Epigynal flaps more convex and epigynal surface with more relief (Figs. 238, 244- 248); flaps usually distinctly darker than rest of epigynum; body usually without metallic sheen; habitat varied 4 3(2). Orange scales bet\\een and beside AMEs just above clypeus; body fairK smooth with shiny coppery scales (Fig. 179); usu- ally on oaks, holly, Arctostaphijlos, and other shrubs and trees with leathery leaves 40. mann ii - White or dark scales around eyes; carapace often with pale spot above and between AMEs (e.g., Fig. 143); body with rougher appearance; on conifers 12. flavipedes 4(2). Epigynal flap angled where flap bends down toward opening (Figs. 247, 374, 376); surface rises immediately behind flap to broad plateau covering posterior of epi- gynum (Fig. 247); carapace thinly cov- ered with white scales that often form an inverted T behind the AMEs (Fig. 157); abdomen often with anterior me- dial paired spots coalesced into one large white spot (Figs. 157, 377); common on various plants including conifers 19. aeneola Epigynal flaps not angled; surface of epi- gynum varied; carapace lacking T-shaped marking on head; abdomen with anterior medial paired spots separate (Figs. 347, 364, 382); habitat varied 5 5(4). Scales on carapace yellowish; legs yellow; abdomen with paired black spots (Fig. 353); epigynal surface behind flaps raised gradually but steeply into high mound along posterior of epigynum (Fig. 245); flaps divergent (Figs. 245, 352); low herbs in fields and bogs 16. insignis Scales on carapace white, beige, or tan; epi- gynal surface behind flaps either more or less flat (Figs. 244, 248) or raised quickly behind flaps into mound (Figs. 238, 246); habitat varied 6 6(5). Area behind epig>nal flaps raised into a broad mound (Figs. 238, 246); flaps con- vergent; carapace with whitish scales 7 Area behind flaps more nearly flat or con- cave (Figs. 244, 248); flaps divergent or convergent; carapace dark or covered with yellowish scales ^ 7(6). Area behind epigynal flaps strongly raised into round dark shiny mound (Figs. 246, 363); flaps convergent; carapace densely covered with white scales (Fig. 153); ab- domen with large dark patches on either side of central paired spots (Fig. 364), anteriormost pale spots fused into short longitudinal bands; commonly found on sagebrush 18. clemata - Area behind flaps only moderately raised into broad mound (Fig. 238); carapace not densely covered with white; abdo- men marked with large square brown spots between paired pale spots (Figs. 2, 135, 287); found on various shrubs and trees 5. proterva 8(6). Body dark, with very small pale spots on abdomen (Fig. 347); white scales be- tween AMEs; epigynum dark; surface rising immediately behind flap to ridge (Fig. 244); collected from waterside shrubs and trees 15. montana Body mottled beige and tan, with large yellowish spots on abdomen (Fig. 382), orange scales between AMEs; epigynum pale; surface rising gradually behind flaps (Fig. 248); juniper-dwelling 20. balia Key to the Female Pelegrina and mannii GROUP OF Arizona* 1. Epigynal flaps thin and rotated 90°, lying in cavity (Fig. 317); markings gold and beige; junipers of southern mountains 1 1 . kastoni Epigynal flaps rotated less than 60° (or, if rotated 90°, rarely in tristis, then flaps very broad); markings varied 2 2(1). Epigynal flaps broad and flat (Figs. 302, 307); epigynal surface more or less flat 3 Epigynal flaps narrower; epigynal sur- face varied 4 3(2). Anterior end of epigynal opening deep, with the surface there pale and de- scending deeply under flap; flaps dark brown; southern Arizona 7. tristis Anterior end of epigynal opening shal- low, with the surface not so shallow nor descending so deeply as in tristis; flaps usually light brown; northern Ar- izona and New Mexico 9. sabinema 4(2). Epigynal surface flat or convex behind flaps (Figs. 247, 252, 255-257); flaps often flat 5 Epigvnal surface concave behind flaps (Figs. 248-250); flaps convex 14 5(4). Epigynal flaps narrow and flat (Fig. 257), often transparent and difficult to see * Not included in the key are northern species that may occur in Arizona but have been at most rarely collected there; Pelegrina montana, flavipedes, in- signis, and clemata. Metaphidippus carmenensis is a species similar to chera with one known specimen from Arizona. It is not included in the key. 260 Bulletin Museum of Comparative Zoology, Vol. 154, No. 4 (Figs. 525, 526); flaps divergent; com- mon in desert vegetation though oc- curs at higher elevations 43. chera - Epigynal flaps wider and more pig- mented, usually more robust than in chera, flaps parallel, convergent or di- vergent; generally found in oak-coni- fer habitats above 1,200 m elevation 6 6(5). Epigynal flaps narrow, flat and mostly parallel e.xcept for sharp bend inward near posterior end (Figs. 255, 481); body vellow, sometimes with paired dark spots on abdomen almost as in insignis (Fig. 171) 37. bunites - Epigynal flaps without sharp posterior bend; if abdomen yellow then lacking large paired dark spots 7 7(6). Abdomen very pale, yellowish, with markings consisting of little more than small dark speckles (Figs. 436, 487, 502) 8 - Abdomen more darkly marked with brown or gray (Figs. 275, 358, 377, 387 ) _.._ _ 1 0 8(7). Epigynal flaps divergent and narrow (Fig. 501) _ 40. mannii (part) Epigvnal flaps parallel or convergent (Figs. 435, 486) _ _ _ .... 9 9(8). Epigynal flaps pale (Fig. 486) 38. orestes Epigynal flaps small and dark (Fig. 435) _ 30. verecunda 10(7). Epigynal flap angled about midway along its length where flap bends down to- ward opening (Fig. 247); surface rises immediately behind flap to broad pla- teau covering posterior of epigynum (Fig. 247); carapace thinly covered with white scales that often form an inverted T behind the AMEs (Fig. 157); abdomen often with anterior medial paired spots coalesced into one large white spot (Fig. 157) 19. aeneola - Epigynal flaps not angled so abruptly in middle; epigynal surface varied; car- apace lacking T-shaped marking on head; abdomen with paired spots sep- arate 11 11(10). Epigynal flaps divergent and narrow (Fig. 501) _ 40. mannii (part) Epigynal flaps parallel or convergent, not so narrow (Figs. 274, 357, 386) 12 12(11). Abdomen marked much as in galathea, with four pairs of prominent white spots with small black spots behind them (Figs. 133, 275); epigynal flaps short, fairly flat, and parallel (Fig. 274) 3. dithalea Abdomen dark areas more prominent than paired white spots (Figs. 358, 387); epigynal flaps varied 13 13(12). Carapace covered with reflective scales; abdomen brown with large paired darker brown spots (Fig. 387), setae around AMEs darkest dorsally 21. chalceola (part) - Carapace covered with white scales; ab- domen with pale longitudinal side bands enclosing brown dorsum with paired white spots (Fig. 358), setae around AMEs all white . ._ _ _ 17. chaimona (part) 14(4). Epigynal flaps dark, long, narrow, and convex (Fig. 418); epigynal surface with strong relief consisting of raised bumps just medial to each flap, a con- cavity behind flaps rising to posterior edge (Fig. 418); first curve of epigynal ducts broad and long (Fig. 417) _ 27. huachuca Epigynal flaps not so long, nor is epig\ nal surface so strongly sculptured; first curve of ducts narrower 15 15(14). Epigynal flaps strongly convex (Figs. 249, 250); posterior end rounded and stand- ing high abo\e surface (Figs. 249, 250); second curve of duct broad (Figs. 397, 400) _ 22. fiircata - Epigynal flaps less convex (Figs. 248, 357, 381, 386); posterior end not standing high above surface (Fig. 248); second curve of duct narrower (Figs. 356, 380, 385 ) _ 1 6 16(15). Abdomen marked with large round white spots (Fig. 382); carapace wide; epi- gynal flaps narrow and pale (Fig. 381) _ 20. balia - Abdomen with small white spots if any (Figs. 358, 387); carapace varied; epi- gynal flaps broader and shorter (Figs. 357, 386 ) 1 7 17(16). Carapace covered with brown reflective scales; abdomen brown with large paired darker brown spots (Fig. 387); setae around anterior median eyes darkest dorsally ..„ 21. chalceola (part) - Carapace covered with white scales, ab- domen with pale longitudinal side bands enclosing brown dorsimi with paired white spots (Fig. 358); setae around anterior median eyes all white 17. chaimona (part) Key to the Pelegrina and Nag.mna FEMALES OF ME.XICO AND Central America* 1. Body and legs mostly yellow (Fig. 163, 173, 175, 436, 461, 487, 492), with small * Females of the Metaphidippus mannii groups are not included. These can usually be distinguished from Pelegrina females by their weaker epigynal flaps, which descend into the openings posteriorly. Pelegrina Jumping Spiders • Maddison 261 dark markings if anv; epigvnum and flaps mostly flat (Figs, 252,' 253, 255) 2 - Bod\ and legs well marked with brown and gray (e.g.. Figs. 298, 318, 409, 414, 425); epigynal surface flat or more or less concave (e.g.. Fig. 250) 9 2(1). First femur, patella, and/or tibia with small subterminal dark transverse bar (e.g., Peckham and Peckham, 1896: fig. 10); clypeus covered with yellow scales except for barren patch beneath AMEs, beneath which on chelicera is dark line; epigynal flaps weak (Fig. 491); dis- turbed lowland habitats 39. Nagaina incunda Legs uniform in color or if annulate, with dark annulae more extensive; clypeus densely covered with pale scales even below AMEs; habitat varied 3 3(2). Epigynal surface more or less flat except for longitudinal ridge between flaps (Fig. 253); flaps convergent, narrow, only slightly convex (Figs. 253, 450) .. 33. variegata (part) - Epigynal surface lacking central ridge (Figs. 252, 255); flaps varied 4 4(3). Body and legs uniformly orange-yellow except sometimes for discrete small dark spots on abdomen; epigynum transparent so that spermathecae eas- ily visible without dissection (Figs. 460, 463); flaps convergent; southern Mex- ico and Central America 35. sandaracina Body and legs pale yellowish beige, not so orange; epigynum varied 5 5(4). Abdomen uniformly yellowish, with small discrete dark speckles only (Figs. 436, 487 ) ' 6 Abdomen mostly yellow but any dark markings are larger spots and patches (e.g.. Figs. 171, 409, 446) 7 6(5). Epigynal flaps pale, transparent (Fig. 486), convergent _ _.... 38. orestes Epigynal flaps dark (Fig. 435), conver- gent to divergent 30. verecunda 7(5). Epigynal flaps strongly convex (Fig. 408); epigynum concave behind flaps 25. ochracea Epigynal flaps flat (Figs. 255, 482, 445); epigynal surface more or less flat 8 8(7). Epigynal flaps narrow, with abrupt bend near posterior end (Figs. 255, 481); Ar- izona to Oaxaca 37. bunites (part) - Epigynal flaps wider, convergent, but without abrupt bend (Fig. 445); Chia- pas to Nicaragua 32. pallidata (part) 9(1). Epigynal flaps rotated 180° (Fig. 424); abdomen with strong lineate markings (Fig. 425) 28. arizonensis - Epigynal flaps rotated at most 90°; ab- dominal markings not so clearly li- neate 10 10(9). Epigynal flaps dark, wide, flat, and strongly convergent (Fig. 297); mon- tane 8. neoleonis Epigynal flaps not so dark and wide; hab- itat varied 1 1 11(10). Epigynal flaps rotated 90° and in pits (Fig. 317); body yellowish (Fig. 141); northern Mexico, on junipers 1 1 . kastoni Epigynal flaps rotated less than 60°; body varied; distribution varied 12 12(11). Abdomen with peculiar transverse markings (Figs. 169, 456); fourth pair of spots in particular a transverse stripe; legs strongly annulate (Fig. 169); face thinly covered with pale scales; epi- gynal flaps pale and convergent (Fig. 455); Yucatan Peninsula . 34. ijucatecana - Abdomen without such transverse mark- ings; fourth pair of spots not a trans- verse stripe; legs, face and epigynal flaps varied 13 13(12). Epigynal flaps with abrupt bend near posterior end (Figs. 255, 481); epigyn- al flaps and surface more or less flat .. 37. bunites (part) - Epigynal flaps without abrupt bend near posterior end; epigynal surface varied 14 14(13). Epigynal surface more or less flat except for longitudinal ridge between flaps (Fig. 253); flaps convergent, narrow, only slightly convex; abdomen marked with large white spots (Figs. 167, 451) 33. variegata (part) Epigynal surface usually rises to poste- rior edge; if flat then lacking longi- tudinal ridge; flaps varied; markings varied 15 15(14). Epigynal flaps strongly convex (e.g.. Figs. 236, 249, 250), parallel or slightly con- vergent 16 - Epigynal flaps flat or only slightly convex (similar to those in Figs. 247, 252); may be strongly convergent _ 20 16(15). Epigynal surface concave behind flaps (Figs. 249, 250), rising gradually to posterior margin (Jurcata group) 17 Epigynal surface rises quickly behind flaps to mound covering most of pos- terior (Fig. 236) 19 17(16). Epigynal flaps fairly short, pale (Fig. 408); southern Mexico and Guatemala 25. ochracea Epigynal flaps generally longer, dark (Figs. 298, 413) 18 18(17). First curve of duct narrow, second curve very broad (Figs. 397, 400); abdominal markings shiny, pale spots generally small (Figs. 396, 398, 402) 22. jurcata 262 Bulletin Museum of Comparative Zoology, Vol. 154, No. 4 - First curve of duct wide, second curve narrow (Fig. 412); strong pale spots on dark abdominal dorsum (Fig. 414) 26. morelos 19(16). Epigynal flaps long and convergent, not truncate posteriorly (Figs. 4, 280); lon- gitudinal dark bands on abdomen prominent (Fig. 281) 4. edrilana - Epigynal flaps shorter, parallel, truncate posteriorly (Figs. 236, 262); abdomen marked with prominent white spots, without prominent dark bands (Figs. 131, 263) 1. galathea (part) 20(15). Epigynal flaps convergent, higher me- dially and tilted down laterally (Fig. 440); epigynal surface high between and behind teardrops, lower lateral to this (Fig. 440) 31. clavator - Epigynal flaps not so tilted; epigynal sur- face flat or only slightly higher me- dially than laterally 21 21(20). Epigynal flaps truncate posteriorly (Fig. 236), high above surface at posterior end 1. galathea (part) Epigynal flaps not truncate posteriorly nor so high above surface 22 22(21). Abdomen marked much as in galathea, with four pairs of prominent white spots with small black spots behind them (Figs. 133, 275); epigynal flaps fairly flat and parallel (Fig. 274) 3. dithalea Abdomen with smaller white spots (Figs. 358, 446); epigynal flaps usually con- vergent (Figs. 357, 445) 23 23(22). Epigynum very flat (as in verecunda. Fig. 252); dark band along inner margin of epigynal opening very wide (Fig. 444) 32. pallidata (part) Epigynum with some relief; dark band along inner margin of epigynal open- ing narrow (Fig. 356) 17. chaimona* DESCRIPTIONS OF THE SPECIES OF PELEGRINA The Pelegrina species of Canada and the northern and eastern United States can be considered reasonably well known, but the same cannot be said for the species of Arizona, Mexico and Central America. In Arizona are many species, some poorly col- lected such as P. huachuca, P. chaimona, P. tristis, P. chalceola, and P. dithalea. * A number of unmatched females from Mexico may be P. chaimona or a species easily confused with it. Even if no additional species are discov- ered in Arizona, there is the danger that males and females of some of the known species have been mismatched. In Mexico and Central America, the situation is worse, where there are probably several species that will remain undescribed for some time to come. Already there are known some female Pelegrina from southern and cen- tral Mexico that apparently represent spe- cies not described here. I shall not give names to them here so as to avoid making more species names based on difficult to determine females and because with ad- equate collecting we may discover that they are females of already-described males. I do, however, give figures of some of them (Figs. 464-471). Figures 464-466 show a single female from Neriaco, Mexico (state unknown), which may represent an extreme southern form of P. chalceola. Figures 467 and 468 show a form from Guerrero, Jalisco, and Michoacan that may be a southern form of P. dithalea. Figures 469-471 show a form occurring in collec- tions from Durango. The descriptions follow a more or less consistent format except that occasionally a feature is noted in a few species that is not noted in any others: for instance, strongly annulate legs are noted under P. yucatecana, but leg annulation is usually not even mentioned, and in P. balia the flange on the cheliceral fang is noted but the fang is ignored in most other descrip- tions. In the case of leg annulation and male abdominal markings, the species should be assumed to be characterized by the usual Pelegrina condition (legs annu- late, but fairly indistinctly, and male ab- domen brown above, with at most small white spots, and ringed by white side bands) unless otherwise mentioned. In the case of the other characters, such as the flange in balia, the distribution of the fea- ture in all species is not fully known. Such a character is described to aid in separating the species from similar species that are known to lack it (in this example, chalceola lacks the flange). Pelecrina Jumping Spiders • Maddison 263 Information on the labels of type ma- terial is cited, and, where possible, the au- thor of handwritten labels is identified. Banks, Chamberlin, Kaston, and Levi types still have with them the author's original labels, handwritten except those of Kaston, whose typewriter was distinctive. F. Pick- ard-Cambridge's and some of the Peck- hams' types no longer have their original labels. F. P. -Cambridge's labels have been replaced by labels handwritten in pencil, perhaps by Pocock or Browning (Levi, personal communication). Some of the Peckhams' labels were rewritten by Bry- ant, but most labels of Pelegrina types are apparently original. Some are in George Beckham's handwriting, but most are in a handwriting that is probably that of Eliz- abeth Peckham, for it occurs in other orig- inal labels in the Peckham Collection and in some of George Peckham's correspon- dence to Henshaw. 1 . Pelegrina galathea (Walckenaer, 1837) new combination Figures 5, 10, 11, 13, 35, 78, 125, 130, 131, 190, 236, 258-263; Map 1 Attus galathea Walckenaer, 1805; 23 (cites Bosc's MS figure, pi. 1, fig. 4, 9) (nomen nudum). Attus galathea Walckenaer, 1837: 456, sp. 100. Type material lost or destroyed. Walckenaer (1837) cited Bosc's MS pi. 1, fig. 4, and also Abbott's fig. 405 (9), but as Walckenaer (1805) refered only to Bosc's figure, this is to be taken as figure of type. Insofar as A. galathea is such a common and well-known species, and Bosc's ambiguous figure could be in- terpreted as another species, a NEOTYPE is here designated, 1 ton, Hancock, Winnebago, Woodbury; MISSOURI: Boone, Cole, Jackson, John- son, St. Louis; NEBRASKA: Lancaster, Loup, Saline; KANSAS: Decateur, Rile\; TEXAS: Anderson, Den- ton, Hardin, Sabine, San Jacinto; MONTANA: Ra- valli, Stillwater; COLORADO: Fremont, Larimer. Natural History. Found on various trees and shrubs, usually in or near forests; less often found in fields and on herbs than is P. galathea. Dondale (1961) describes the life history of P. proterva in Nova Scotia. 6. Pelegrina peckhamorum (Kaston, 1973) new combination Figures 126, 136, 137, 195,239, 288-293; Map 8 Metaphidippus peckhamorum Kaston, 197.3: 115, figs. 39-42, (52. Holotype 6 and paratype $ in AMNH with labels "Holotype S + allotype 9, Metaphidip- pus peckhamorum n. sp., det by B. J. Kaston (1949)" and "col. by B. Malkin, Lakehurst, N. J. 25 May 1941," e.xamined. BrignoU, 1983: 643. Diagnosis. A relatively rare eastern spe- cies with male body form and markings very much like proterva but outstanding for its very broad embolus. The female is best distinguished from other eastern spe- cies by the indistinct markings and large, slightly concave epigynum, with flaps that are more convex than in proterva and that have the posterior edge not truncate as in galathea. Male. Palpus (Figs. 195, 289, 290): Em- bolus very broad, tapering but still broad at tip. Rami well separated; retrolateral ramus not elongate as in proterva. Mark- ings (Figs. 136, 288): Cheek band dense and discrete. Clypeus with prominent di- amond of white scales between AMEs and overhanging chelicerae; lateral to this the clypeus and hairs overhanging chelicerae are dark. White forehead band contacts AMEs dorsally; setae ringing AMEs white 7:00-12:30 and 2:00-4:00. Chelicerae lacking pale scales. Femur of palpus dis- tinctly paler than more distal segments. Cymbium with white scales centrally. Leg femora distinctly paler basally. Measure- ments: Body length 3.0(3.6)3.7 mm; car- apace length 1.4(1.7)1.8 mm, width/length 0.74(0.76)0.78; n = 53 from Barnstable County, Massachusetts. Female. Epigynum (Figs. 239, 291, 292): Large. Flaps long, fairly convex, usually convergent. Surface rises very gradually behind flaps; most of posterior area con- cave. First curve of duct broad; second Pelecrina Jumping Spiders • Maddison 273 curve proceeds obliquely anteriorly. Markings (Figs. 137, 293): Carapace cov- ered with yellowish scales. Clypeus cov- ered thinly with yellowish white scales. Abdomen more uniform in color than pro- terva, light bro\\'n with pale spots. Mea- surements: Body length 3.6(4.0)5.4 mm; carapace length 1.7(1.9)2.1 mm, width/ length 0.73(0.77)0.78; n = 52 from Mas- sachusetts and Arkansas. Male/Female Matching. Males and fe- males have been co-collected in New Jer- sey, Arkansas, and Massachusetts; other- wise, they are the only unmatched $9 in the northeast. Courtship (45 observed from Cape Cod, Massachusetts). Has the crouch display with body high and first legs low as in P. pro- terva. Raisedspread (n = 17, 35). Crouch (Fig. 126; n = 12, 33): Body held normal- high (n = 6, 16) or high (n = 6, 26). First legs held horizontal (n = 12, 3<3) and lower than body (n = 3, 16), waved little if at all (n = 9, 26). Palpi held down (n = 12, 33); still on pause (n = 2, 16), waved on series (n = 4, 26) up and down (n = 5, 15), spe- cifically from down to forward (n = 2, 26), with medium-high amplitude (n = 5, 13). Abdomen still on series (n = 4, 23) but twitched on pause (n = 4, 23). Repertoires: 13 raisedspread only; 23 raisedspread and crouch; 13 crouch only. Distribution (Map 8). Known from Mas- sachusetts, New York, New Jersey, Ohio (Kaston, 1973), Indiana (Kaston, 1973), Tennessee, Arkansas, and Texas. Records. UNITED STATES: MASSACHUSETTS: Barnstable Co.: Chatham (IS 3 9, MCZ), South Chat- ham (6<3, MCZ), nr. North Truro at junction of Hwy 6 and Head of the Meadow Road (10^ 32, MCZ); Dukes Co.: Oak Bluffs (1<3, MCZ); NEW YORK: Dav- isville; Suffolk Co.: Riverhead {2S, AMNH), Coram (2.3 12, AMNH); NEW JERSEY: BurUngton Co.: 11 km W of New Gretna [46 22; AMNH), Lebanon State Forest {16, AMNH); Middlesex Co.: Old Bridge (1<5, AMNH); Morris Co.: Chatham, Great Swamp (12, AMNH); Ocean Co.: Lakehurst {226 112, AMNH), Lake Horicon nr. Lakehurst (32, AMNH), 6 km W of Lakehurst (23 32, AMNH); TENNESSEE: Knox Co.: University of Tennessee farm 3 {16. AMNH); ARKANSAS: Washington Co.: 24 km S of Prairie Grove in Cove Creek Valley of the Boston Mtns. (28.3 62, MCZ), 24 km W of Prairie Grove {56 12, MCZ); TEXAS: Leon Co.: SW of Oakwood {16, AMNH). Natural History. May specialize on oaks. On Cape Cod, Massachusetts, collected by beating oaks and cranberries in understory of pine forest (1 record), sweeping oak- pitch pine (2 records), and beating oaks (1 record). 7. Pelegrina neoleonis new species Figures 138, 196, 294-298; Map 6 Holotype male and paratype female in MCZ with label "MEXICO: NUEVO LEON: Chipinque Mesa just S of Monterrey, ca. 4500 ft. [1,370 m]; ca. 100.4°W 25.6°N, 2 Jun 1983 W. Maddison & R. S. Anderson 83-034, beating and sweeping forest un- derstory." Etymology. After the state from which most known specimens come. Diagnosis. A Mexican species similar to tristis with a distinctive long, curved retro- lateral ramus on the embolus. The erect portion of the embolus is narrower than in tristis. No characters have yet been found to distinguish the female from that of tris- tis, except locality. Male (from Nuevo Leon). Palpus (Figs. 196, 295): Embolus distinctive; broad, with retrolateral ramus extended into long hook, much as in tristis, but ramus bears small bump and is blunt at tip; prolateral ramus obtuse or only slightly acute. Markings (Figs. 138, 294): Cheek band weak. Clyp- eus brown; hairs overhanging chelicerae dark. White forehead band contacts AMEs dorsally 10:30-12:30. Chelicerae lack pale scales. Femur of palpus distinctly pale than more distal segments. Cymbium with none to few white scales. Femur of third leg pale on basal Vs. Measurements: Body length 3.6, 3.7 mm; carapace length 1.7, 1.8, 1.9 mm, width/length 0.75, 0.75, 0.76; n = 33 from Nuevo Leon and San Luis Potosi. Female. Epigynum (Figs. 296, 297): Flaps large and dark, flat and inwardly rotated. Surface gently convex, highest medially behind flaps, except for surface diving deeply under flaps. First curve of duct very broad, expanded to the side and 274 Bulletin Museum of Comparative Zoology, Vol. 154, No. 4 posterior so that second curve begins well posteriad of posteriormost portion of flap (in contrast with the sympatric clavator); second curve proceeds anteriorly. Inner surface of third curve rough, with numer- ous projections. Markings (Fig. 298): Car- apace dark above, covered with transpar- ent reflective scales; sides covered thinly with white scales. Clypeus densely covered with white scales. Legs brown. Abdomen fairly dark, with only small pale spots. Measurements: Body length 4.3(5.9- 6.1)6.2 mm; carapace length 1.9(2.2- 2.3)2.3 mm, width/length 0.74(0.77- 0.79)0.83; n = 4$ from Nuevo Leon, Hi- dalgo, and Oaxaca. Male/Female Matching. This associa- tion is indicated by co-collecting in Nuevo Leon; by the large epigynal flaps, which would be expected in a species with such a robust embolus; and by the similarity of male and female with those of tristis. Geographical Variation. The single male from San Luis Potosi differs from those of Nuevo Leon in having a sharper retrolater- al prong on the embolus, more extensive white scales surrounding AMEs, a small patch of pale scales medially on chelicerae, and the femur of palpus relatively dark. Courtship (33 observed from Chipinque Mesa, Nuevo Leon; Cerro Potosi, Nuevo Leon; and Xilitla, San Luis Potosi). Raised- spread (n = 3, 2(5). Crouch (n = 6, 33): Body held in normal to low position (n = 1). First legs bowed and forward (n = 4, 23), raised to ca. 30° (n = 2, 23), or hori- zontal (n = 1), or femora low but tips curl upward (n = 1). Leg tips not touching (n = 2, 23), apparently not waved (n = 3, 23), or waved only slightly (n = 2, 13) on series (n = 1). Palpi down (n = 6, 33), over che- licerae (n = 1) or curled under tips of che- licerae (n = 1), waved (n = 6, 33) up and down (n = 1) or outward (n = 1) on series (n = 3, 23) ca. 5-7 c/s (n = 1). Abdomen depressed a bit on series (n = 1), or at end of series (n = 1). Repertoires: 13 crouch only; 23 raisedspread and crouch. Distribution (Map 6). Northeastern Mexico south to Oaxaca. Records. MEXICO: SAN LUIS POTOSI; 21 km W of Xilitla on Hwy 120, 99°05'W, 21°18'N, 12 June 1983 (15, MCZ); NUEVO LEON: Chipinque Mesa just S of Monterrey, 100. 4°W, 25.6°N, 2 June 1983 {IS 12, MCZ); Cerro Potosi, ca. 100°14'W, 24°52'N, 4 June 1983 (15 12, MCZ); HIDALGO: Pachuca (12, MCZ); OAXACA: 50 km NW of Oaxaca, 97°00'W, 17°14'N, 6 August 1983 (22, MCZ). Natural History. Beating oaks and pines in oak-pine area (3 records); sweeping shrubs, cloud forest (1 record). Elevations at four locations in Nuevo Leon and San Luis Potosi from 1,400 to 2,900 m. 8. Pelegrina tristis new species Figures 197, 299-303; Map 7 Holotype male and paratype female in AMNH with labels "ARIZONA: Cochise Co., Round Park, Chir- icahua Mtns., June 28, 1967. 9300 ft. [2,840 m], Gertsch, Hastings." Etymology. Latin adjective for "sad," referring to the large size of the teardrop- shaped flaps over the epigynal openings. Diagnosis. A large, dark, plainly marked species known from southern Arizona, similar in genitalia to neoleonis and sa- binema. The erect portion of the embolus is broader than in either of those species, and the rami are sharper than in neoleonis. Females are generally not so yellow as in sabinema, and the epigynal openings are deeper, in that the surface descends more deeply under the anterior part of the flaps. Male. Palpus (Figs. 197, 300): Embolus extremely broad, so that retrolateral mar- gin joins without angle to retromargin of embolar base. Both rami sharply pointed; retrolateral ramus extended into long hook, lacking subterminal bump. Markings (Fig. 299): Carapace dark, with reduced fore- head band. Cheek band very weak to ab- sent. Clypeus brown, with dark hairs over- hanging chelicerae. White forehead band absent or much reduced, fails to contact AMEs, which are ringed with dark above. Chelicerae lacking pale scales. Palpus al- most uniformly brown, femur not distinct- ly paler. Cymbium lacking white scales. Legs relatively uniform brown, femora en- tirely dark. Measurements: Body length 3.7(4.3)4.6 mm ; carapace length 1 .8(2. 1)2.1 Pelegrina Jumping Spiders • Maddison 275 mm, width/length 0.75(0.76)0.82; n = 4<5 from Chiricahua and Santa Catalina Mtns., Arizona. Female. Epigyniim (Figs. 301, 302): Flaps large, dark, and convergent, often far rotated, sometimes as far rotated as in neoleonis (Fig. 297). Just medial to the flap at the anterior end the surface is pale and descends deep under flap (Fig. 302, ar- row). Except for this concavity, the epi- gynal surface is gently convex, highest me- dially behind flaps. First curve of duct very broad; second curve proceeds anteriorly. Markings (Fig. 303): Carapace covered above thinly with white to dark transpar- ent reflective scales. Clypeus densely cov- ered with white scales. Abdomen light to medium brown with small central pale spots. Narrow dark brown spots beside these pale spots form longitudinal dark stripes. Measurements: Body length 5.0(5.7)6.2 mm; carapace length 1.8(2.0)2.3 mm, width/length 0.77(0.78)0.82; n = 59 from Chiricahua, Huachuca, and Santa Rita Mtns., Arizona. Male/Female Matching. This match- ing is indicated by microsympatry in Chir- icahua Mtns., by robust embolus and flaps, by similar large size, and by similarity of genitalia to male and female of sabinema, which are reasonably surely matched. Distribution (Map 7). Southern Arizona. Records. UNITED STATES: ARIZONA: Santa Rita Mtns.: Madera Canyon (49, AMNH, MCZ); Huachuca Mtns.: Garden Canyon (19, AMNH); Santa Catalina Mtns.: Bear Wallow to Mt. Lemmon (19, AMNH), Chiricahua Mtns.: Round Park, Southwestern Re- search Station 8 km W of Portal, Barfoot Park, and Rustler's Park (3<3 49, AMNH). Natural History. Collected at 1,500- 2,800 m elevation (3 records). Females have been collected in June (2 records), July (4 records), and August (3 records). 9. Pelegrina sabinema new species Figures 198, 304-308; Map 9 Holotype male in AMNH with label "ARIZONA, Showlow, July 1967, W. J. Gertsch. " Etymology. An arbitrary combination of letters, to be treated as an adjective. Notes on Specific Distinctness. Pele- grina sabinema is much like pervaga, and indeed I long considered it only the west- ern form of pervaga, but the more strongly developed white-black-white carapace stripes and narrower embolus of pervaga suggest that pervaga may be the sister spe- cies to kastoni, with sabinema the sister to those two. The embolus and markings of P. sabinema are slightly more like those of tristis and neoleonis, which may be con- sidered outgroups. Diagnosis. Differs from pervaga in hav- ing less swollen carapace sides, an abdo- men lacking the pale central stripe on the abdomen, wider embolus, weaker male cheek band, less dense band of dark hairs beneath male carapace side bands, darker and more robust epigynal flaps, and yellow female legs. Differs from tristis in having yellow legs, yellow male chelicerae, nar- rower embolus, dense covering of pale scales on female carapace, and shallower epigynal openings. Male. Palpus {Figs. 198, 305): Embolus very wide at base of erect portion, thought still with a distinct angle between retro- margins of erect portion and base; retro- lateral ramus long and blunt. Carapace often broad though sides not swollen as in pervaga. Markings (Fig. 304): Cheek band weak, runs horizontally and posteriorly be- neath band of dark hairs beneath white side bands. Clypeus brown, hairs over- hanging chelicerae dark. White forehead band contacts AMEs rather far medially, from 9:00 to 12:00. Chelicerae yellow, lacking pale scales. Palpus yellow with white scales on femur, tibia and cymbium interrupted by dark hairs on patella and base of cymbium. Legs uniformly yellow. Abdomen brown centrally with white side bands, showing trace of paired dark spots of female. Measurements: Body length 3.3(3.6)3.8 mm; carapace length 1.7(1.7)1.9 mm, width/length 0.78(0.80)0.83; n = 53 from New Mexico and Arizona. Female. Epigynum (Figs. 306, 307): Flaps flat and large, often far rotated, as in tristis and neoleonis, though usually not 276 Bulletin Museum of Comparative Zoology, Vol. 154, No. 4 so dark as in those species. Openings shal- lower than in tristis; that is, just medial to the flap at the anterior end the surface is not so pale and does not dive deep under flap. Epigynal surface flat. First curve of duct very broad; second curve proceeds less anteriorly than in tristis. Markings (Fig. 308): Carapace well covered with white to yellowish scales. Clypeus very densely covered with white scales. Legs yellow. Abdomen yellowish with brown centrally, paired white spots. Usually paired dark brown spots in posterior half are beside white spots. Measurements: Body length 4.0(4.2)5.5 mm; carapace length 1.6(1.7-1.8)1.9 mm, width/length 0.78(0.80-0.81)0.86; n = 6$ from New Mexico and Arizona. Male/Female Matching. Male and fe- males were matched by similar yellow col- or, by robust embolus and flaps, and by co-collecting and common distribution in New Mexico and northern Arizona, where no other unmatched females and males are known. Distribution (Map 9). New Mexico, northern Arizona, southern Colorado, and western Texas. Records. UNITED STATES: TEXAS: Jeff Davis Co.: 24 km NW of Fort Davis (19, AMNH); COL- ORADO: Montezuma Co.: Mesa Verde National Park (1<5, AMNH); NEW MEXICO: Bernalillo Co. (29, AMNH); Lincoln Co.; nr. Ruidoso, Ruidoso Cmpgd. (22, AMNH); Los Alamos Co.: nr. Los Alamos (19, AMNH); Sandoval Co.: (19, AMNH), Sandia Mtns., Juan Tabo area (29, AMNH); Santa Fe Co. (2ntype 3., Gua- temala (Sarg)," examined. Despite the type label on the latter specimens, the holotype may be better considered to be among the former, given Cam- bridge's indication of the distribution as Orizaba. Bonnet, 1957: 2813. Dendryphantes furcatus: — G. & E. Peckham, 1909: 473. Roewer, 1954: 1203. Dendryphantes mimus Chamberlin, 1925b: 135, figs. 53, 54, 3. Holotype in MCZ 13 with label "Den- dryphantes mimus Chamb., 3 holotype, N. M.: Pe- cos, R. V. Chamberlin Coll. 1047, " examined. Roewer, 1954: 1212. Bonnet, 1956: 1396. NEW SYNONYMY. Diagnosis. A species common in the Mexican highlands, with a striking court- ship display and distinctive embolus hav- ing two blunt rami. The epigynum, with convex flaps, concave surface, and wide second curve of the ducts, is distinctive. Male. Palpus (Figs. 212, 389-394): Em- bolus heavy and slanting, with retrolateral ramus extended and truncate. Markings (Figs. 158, 88): dark brown with distinct sheen and contrasting white side bands. Carapace side bands usually connect to white scales over anterior eye row to make a continuous band of white encircling the front of the carapace (though not seen in male drawn. Fig. 388). Cheek band mod- erately weak. Clypeus brown; hairs over- hanging chelicerae brown to white me- dially, brown laterally. White forehead Pelecrina Jumping Spiders • Maddison 293 band contacts AMEs dorsally 10:30-12:30 or 1:00. Chelicerae with some pale scales medially. Cymbium lacking pale scales. Measureinents: Southern Arizona: Body length 3.5(4.2)4.5 mm; carapace length 1.7(2.1)2.2 mm, width/length 0.74(0.77) 0.77; n = 53 from Santa Rita Mtns., Ari- zona. Northern Arizona: body length 3.5, 3.6, 3.9 mm; carapace length 1.7, 1.8, 1.8 mm, width/length 0.75, 0.78, 0.79; n = 33 from Yavapai Co., Arizona. Female. Epigijnum (Figs. 249, 250, 395, 397, 398, 400, 401): Flaps strongly convex and often dark. Surface concave behind flaps, without mound, rising gradually to lip at back edge. First curve of duct nar- row; second curve broad initially but nar- rows as it proceeds medially. Markings (Figs. 159, 396, 399, 402): Body often with slight bronze sheen; variable in markings. Carapace covered with brassy reflective scales, sometimes dark, sometimes mixed with white. Clypeus only thinly covered with white scales except in northernmost populations (form mimus). Measure- ments: Southern Arizona: Body length 4.5(4.7-4.9)5.4 mm; carapace length 2.0(2.1)2.2 mm, width/length 0.75(0.78- 0.79)0.79; n = 6$ from Santa Rita Mtns., Arizona. Northern Arizona: Body length 4.0(4.6)5.7 mm; carapace length 1.8(1.8)2.0 mm, width/length 0.72(0.75)0.77; n = 59 from Yavapai Co., Arizona. Geographical Variation. Four geo- graphical forms might be recognized. (1) The most widespread form occurs from Guatemala north to northern Mexico, with narrower embolus and thinner epigynal flaps that are divergent or parallel (furcata s.s.,; Fig. 393). The retrolateral ramus of the embolus is truncated obliquely. Most females through this range are well marked with pale spots, as in form mimus (Fig. 396). (2) A second form, very similar to the widespread form, occurs in the Santa Rita, Santa Catalina, and Chiricahua Mountains of southern Arizona and prob- ably in northern Mexico (Figs. 392, 397- 399). The embolus is also narrow and the flaps divergent or parallel, but the retro- lateral ramus of the embolus is truncated transversely, so that its distal tip makes a line perpendicular to the axis of the palpus. Females are dark. (3) A third form occurs in northern Arizona (Yavapai Co.), Colo- rado, and New Mexico, having a wider embolus and convergent flaps (mimus: Figs. 390, 391, 395, 396). The retrolateral ramus of the embolus is truncated trans- versely, as in form (2). The difference be- tween the northern (mimus) and southern (furcata) Arizona specimens is rather strik- ing, for the females are also smaller and paler in the north. Though mimus might be considered a distinct species, specimens in New Mexico present a confusing mix- ture of characteristics of forms (1), (2), and (3). (4) A fourth form is found in western Oaxaca (43 159, 31 km N or Guelatao de Juarez, ca. 96.5°W, 17.5°N, 2,600 m el., 3 August 1983, W. Maddison & R. S. An- derson, MCZ), with very wide embolus, dark females, and extremely robust flaps (Figs. 388, 400-402). This form occurs within 50 km of the widespread form. In total, the variation among these popula- tions is confusing, and though several spe- cies may be present, only one will be rec- ognized until better studied. Chromosomes. 2n3 = 26 acrocentrics + XXO (23 from Madera Canyon, Arizona). Courtship (103 observed from Nuevo Leon, Hidalgo, Queretaro, Puebla, Oaxa- ca, Chiapas, and the Santa Rita Mountains of Arizona). Very unusual for the genus, with vigorous leg waving and body jerking in a stage I will call the semaphore stage. Semaphore (n = 24, 103): Body high to very high (n = 24, 103). Male walked si- dling (n = 19, 83) in series (n = 14, 53). First legs wide, nearly 180° apart, approx- imately horizontal (n = 21, 83) or below horizontal (n = 1), though occasionally not much more than 90° apart and raised to 60° (n = 1), waved vigorously up and down almost to vertical (n = 21, 93), though sometimes only to ca. 40° (n = 1), at ca. 3-4 c/s (n = 2, 23) or 5 c/s (n = 1) on each sidle (n = 9, 43). The leg wave is vertical and slightly posterior to bring the legs up 294 Bulletin Museum of Comparative Zoology, Vol. 154, No. 4 and back (n =7, 43); the left and right sides wave in unison (n = 4, 33), or occasionally asynchronously (n = 1). Palpi hanging down (n = 18, SS) and parallel (n = 9, 46) and a bit forward (n = 7, 33). Palpi wave with low-medium amplitude (n = 5, 33) on each sidle (n = 4, 23), up and down like pushing and pulling motion (n = 1), or largely still (n = 1, 23). Abdomen trails a bit on sidles (n = 5, 23), ca. 10-30° (n = 3, 13), but more or less horizontal (n = 16, 53). Occasionally 3 pauses from vigorous leg waving and jerks whole body (n = 11, 63) approximately 4-5 times (n = 3, 33) or 3 times (n = 1) while the first legs are spread wide and horizontal (n = 11, 63). These jerks came after a few sidles (n = 1). The body may be lowered for the jerks (n = 1). Reach (n = 3, 23): The male pro- ceeded directly from this semaphore stage into the reach to touch the female (n = 3, 23). During reach, body jerked a few times (n = 1). First legs held parallel and forward (n = 2, 23). Palpi held parallel and forward (n = 2, 23). This description is from the widespread form (1) with the following exceptions. Displays of 43 from the distinctive popu- lation from Guelatao de Juarez, Oaxaca (form (4)), showed the same form of sem- aphore display with legs spread wide, vig- orously waved up during series (n = 11, 43). No whole-body jerks were noted, how- ever. One male from southern Arizona (form (2)) showed the same semaphore dis- play (n = 4), though no whole-body jerks were noted. Distribution (Map 22). Throughout the highlands of Mexico, extending north to Colorado and south to Guatemala. Records. Mostly in MCZ and AMNH: UNITED STATES (county records): COLORADO: Boulder (4 / a i-ni embolus 64 g Figure 64. Hypothetical transformations among embolus types from the euophryine type (a) to the dendryphantine types (b- h). At the left or lower left of each of a-h is the base of the embolus; the tip of the embolus is at the top of each figure. Examples with these types are Metaphidippus chera, Ens aurantia, and Bellota wheeleri (b); Hentzia and Zygoballus {c); Eris militaris (d); Paradamoetas (e); Dendryphantes tropicus and Mabellina (f); Metaphidippus mandibulatus (g); and Messua, Bagheera, and Gastromicans (h). Figures 40-63. Left palpi of various salticids of the subfamily Dendryphantinae. 40. Eris cf. aurantia (Chiapas: San Cristobal). 41. Phidippus audax (Ontario: Burlington). 42. Bellota wheeleri {Oaxaca: SW of Valle Nacional). 43. "Pseudicius" siticulosus (Arizona: Yavapai Co.). 44. Terralonus ca/zfom/cus (California: Santa Cruz Co.). 45. Sassacus papenhoei (Nevada: Lander Co.). 46. Tulpius hilarus (Quintana Roo: Kohunllch ruins). 47. Phanias harfordii (California: San Mateo Co.). 48. Ghelna castanea (Virginia: Falls Church). 49. Anicius dolius Chamberlin (holotype; Jalisco; Guadalajara). 50. Hentzia mitrata (Hentz) (Minnesota: Washington Co.). 51. Zygoballus rufipes (Ontario: Essex Co.). 52. Rhene cf. flavigera (China: E. Kwantung: Yim Na San). 53. Eris militaris (Ontario: Port Elgin). 54. Rudra geniculata (Panama: Canal Zone). 55. Parnaenus recurvus (paratype; Panama: Barro Colorado Island). 56. " Paraphidippus" validus (paratype; Panama: Barro Colorado Island). 57. Species near Zygoballus /ncertus (Quintana Roo: Kohunlich ruins). 58. Paradamoetas fontana (Ontano: Hastings Co.). 59. Metaphidippus ct wf/s(Puebla: nr. Xicotepec de Juarez). 60. Dendryphantes perfectus G. & E. Peckham (holotype; Brazil: Para). 61 . "Eris" nidicolens (France: Marseille). 62. Dendryphantes tropicusG. & E. Peckham (holotype; Brazil: Chapoda). 63. Mai)e///naprescoff/Chickering (paratype; Panama: El Valle). Scale bars. 0.1 mm. 334 Bulletin Museum of Comparative Zoology, Vol. 154, No. 4 1 Pelegrina Jumping Spiders • Maddison 335 Figures 71-79. Tibial apophyses of left palpi of dendryphantines. 71. Bagheera kiplingi (Oaxaca: nr. Tuxtepec). 72. Messua desidiosa (Costa Rica: San Jose). 73. Gastromicans levispina (Panama: El Valle). 74. Metaphidippus mandibulatus (holotype; Costa Rica). 75. Phidippus audax (Ontario: Halton Co.). 76. Dendryphantes hastatus (Poland: Smogorzew). 77. Beats hispida (Quintana Roo: Kohunlich ruins). 78. Pelegrina galathea (Texas: Bexar Co.). 79. Eris militaris (North Carolina). Scale bars. 0.1 mm. Figures 65-70. Scanning electron micrographs of epigyna of dendryphantines, showing teardrop-shaped flaps over openings. View is mostly ventral, slightly oblique lateral. 65. Dendryphantes rudis (U.S.S.R.: Buzjatia). 66. Eris militaris (Michigan: Emmet Co.). 67. Phidippus audax (Minnesota: Rochester). 68. Terralonus mylothrus (Colorado: Pitl0 Figures 304-308. Pelegrina sabinema (sp. 9). 304. 6 face (New Mexico: Santa Fe Co.). 305. Palp (Arizona: Coconino Co.). 306, 307. Epigynum, dorsal (306) and ventral (307) (New Mexico: Santa Fe Co.). 308. 2 abdomen (New Mexico: nr. Edgewood). Figures 309-31 3. Pelegrina pervaga (sp. 1 0). 309. 6 face (Texas: Kerr Co.). 31 0. Palp (Texas: Kerr Co.). 311,312. Epigynum, dorsal (311) and ventral (312) (Texas: Val Verde Co.). 313. 5 abdomen (Texas: Val Verde Co.). Figures 314-318. Pelegrina kastoni (sp. 11). 314. 3 face (Arizona: Sycamore Canyon). 315. Palp (Arizona: Chiricahua Mtns.). 316, 317. Epigynum, dorsal (316) and ventral (317) (Arizona: nr. Cienega Lake). 318. 2 abdomen (Arizona: Madera Canyon). Figures 31 9-323. Pelegrina flavipedes (sp. 1 2). 31 9. <; face (Alberta: Cypress Hills). 320. Palp (Ontario: Muskoka District). 321 , 322. Epigynum, dorsal (321) and ventral (322) (Ontario: Bruce Co.). 323. 2 abdomen (Michigan: Crawford Co.). See also Figures 338 and 339. Scale bars. 0.1 mm, except for 6 face and 2 abdomen 0.5 mm. Pelecrina Jumping Spiders • Maddison 359 338 Figures 324-328. Pelegrina flaviceps (sp. 13; Ontario: Kingston, except 324 New Hampshire: Surry). 324. s face. 325. Palp. 326. Epigynum, dorsal. 327. Epigynum, ventral. 328. 2 abdomen. See also Figures 340 and 341. Figures 329-335. Pelegrina exigua (sp. 14). 329-333. Dull form (Virginia: Shenandoah Co., except 333 Kentucky: Rowan Co.). 329. 6 face. 330. Palp. 331. Epigynum, dorsal. 332. Epigynum, ventral. 333. 5 abdomen. 334, 335. Striped form. 334. 6 face (Maryland: Montgomery Co.). 400. s abdomen (Missouri: Jefferson City). See also Figures 336, 337, and 342. Figures 336-342. Spermathecal ducts of cleared epigyna, dorsal view, of flavipedes group species. 336, 337, 342. P. exigua (336, North Carolina: Durham Co.; 337, Kentucl40% of total length [> 13% difference between means of proportional tail length of the two sexes]), and (2) male superiority in body size and ventral counts. The first character is unknown in snakes outside the L. sexlineatus group. In contrast, the stumpffi group lacks these unusual features. A phy- logenetic hypothesis supported by characters of ex- ternal morphology, hemipenes, skulls, and behavior suggests the following relationships: {{({dolicocercus, grandidieri), rhadinaea new species), sexlineatus), pinguis); and ({stumpffi, epistibes new species), in- frasignatus, lateralis). Although the species groups of Liopholidophis ap- pear to be monophyletic, no strong evidence supports the monophyly of Liopholidophis sensu lato. How- ever, revision of the generic concept is not warranted until broader relationships among Malagasy colubrids are better understood. Problems concerning the ge- neric status are highlighted by similarities among Lio- pholidophis rhadinaea, new species, several species in the genus Liophidium, and a specimen resembling L. rhadinaea, new species, in external and hemipenial characters, but whose generic and species placement is enigmatic. Species of Liopholidophis are diurnal and terres- trial, except for sexlineatus, which is semiaquatic (no observations for pinguis and stumpffi). Species of Liopholidophis consume primarily frogs (one cha- maeleon record; no data for pinguis, grandidieri, and stumpffi). Most dietary items for injrasignatus, ep- istibes, new species, and dolicocercus were terrestrial microhylid frogs {Plethodontohyla spp.); two records for rhadinaea, new species, were clutches of frog eggs; sexlineatus consumed Heterixalus (Hyperoli- idae) and Ptychadena (Ranidae); lateralis consumed Mantidactijlus and Ptychadena (Ranidae), and Boo- phis (Rhacophoridae). Species of Liopholidophis are probably oviparous with the exception of sexlineatus, which appears to be viviparous (data somewhat equivocal for dolicocercus; no data for stumpffi and pinguis). INTRODUCTION The Madagascar! snake genus Liophol- idophis (Colubridae) as presently defined (Mocquard, 1904; Guibe, 1958) includes rather generalized terrestrial to semi- aquatic snakes. Several nominal species, including grandidieri Mocquard, dolico- cercus Peracca, sexlineatus Giinther, and a new species described herein, exhibit the unusual characteristic of considerable sex- ual dimorphism in tail length, a character used by Mocquard (1904) in defining the genus. In these species the tail of males averages >40% of total length, whereas in females the tail is usually <30% of total length (>50% and 35% for males and fe- males, respectively, in grandidieri) (see additional comments herein). In other col- ubrids, including other species of Lio- pholidophis, the tails of males and females do not show such exaggerated differences in length, and the sexes overlap in the rel- ative proportion of tail to total length. The monophyly of Liopholidophis sensu lato has never been explicitly justified and will be considered in detail later in this paper. Most nominal taxa of Liopholidophis were described in the first half-decade of the twentieth century or earlier. Their no- menclatural history is summarized in the species accounts. Present understanding of Liopholidophis stems primarily from the generic summaries of Parker (1925), Guibe (1954, 1958), and Domergue (1969, 1973). Parker (1925) described a new species (pinguis) and informally recognized two species groups within Liopholidophis: a ^'sexlineatus group" including grandidi- eri, dolicocercus, and sexlineatus, based on the shared characters of extreme sexual dimorphism in tail length and 17 midbody scale rows; and a "stumpffi group," in- cluding lateralis and stumpffi, which lack the extreme tail dimorphism and have 19 midbody scale rows. Parker left pinguis, which shares 17 midbody scale rows with the sexlineatus group but has reduced sex- ual dimorphism in tail length, unplaced in either group. Subsequently, Guibe (1958) synonymized dolicocercus with sexlinea- tus, and stumpffi was first synonymized with lateralis (Guibe, 1954), and then res- urrected (Domergue, 1973). Domergue (1973) described a new species, thieli, and also (Domergue, 1969) recognized that L. pseudolateralis Guibe (1954) was a syn- onym of Dromicodryas bernieri (Dumeril, Bibron, and Dumeril, 1854). These changes have resulted in the presently recognized species of Liopholidophis: grandidieri LlOPHOLIDOPHIS (COLIBRIDAE) FROM MADAGASCAR • Cadle 371 Mocquard (1904), sexlineatus (Giinther, 1882), pinguis Parker (1925), lateralis (Dumeril, Bibron, and Dumeril, 1854),- stumpffi (Boettger, 1881a,b), and thieli Domergue (1973) (e.g., Glaw and Vences, 1994). At least one undescribed species from northern Madagascar is known (Rax- worthy and Nussbaum, 1994a). A general herpetological survey of the recently established Ranomafana National Park in eastern Madagascar (hereafter, RNP; Fianarantsoa Province, Ifanadiana fivondronana; Fig. 3) has resulted in dis- covery of a number of new species of am- phibians and reptiles (e.g., Cadle, 1995). The primary aim of this paper is to clarify the status of species of Liopholidophis from the RNP. In doing so, I describe two new species, resurrect two old names from syn- onymy, and summarize data for the other species. A full-scale revision of Liopholi- dophis is beyond the scope of this report, but I have undertaken revisionary steps pertinent to the nomenclature of species occurring in the RNP, which includes all species recognized herein except pinguis and stumpffi (Boettger) (see later). Some questions concerning species limits within Liopholidophis, especially in the broadly distributed species lateralis and sexlinea- tus, clearly need to be examined anew with more detailed geographic comparisons than undertaken here. 1 summarize knowledge of all species, present illustrations (except stumpffi and pinguis) and descriptions of hemipenes of all species, and hypothesize relationships within the species groups based on external morphology, color patterns, hemipenes, skull morphology, and behavior. Revised synonymies are given for all species. No- menclaturally relevant actions taken here- in include the following. (1) Dromicus dol- icocercus Peracca (1892) is resurrected ^ Virtually all authors, apparently beginning with Jan (1863) and Boulenger (1893), have cited author- ship of this name as "Dumeril and Bibron," but the species is described in volume 7 of the Erpetologie Generale, authored bv Dumeril, Bibron, and Du- from the synonymy of Liopholidophis sex- lineatus (Giinther), where it was placed by Guibe (1958). (2) A lectotype is desig- nated for Dromicus stumpffi Boettger, a species known only from northern Mad- agascar (at least Nossi-be, the type locality, and the vicinity of Montague d'Ambre); a wide-ranging species of the eastern forests previously confused with stumpffi sensu Boettger is described as new. (3) A lecto- type is designated for Ptyas infrasignatus Giinther (1882), and that name is recog- nized as a senior synonym of Liopholi- dophis thieli Domergue (1973), as used widely in current literature (e.g., Glaw and Vences, 1994). MATERIALS AND METHODS My study of Liopholidophis is based pri- marily on specimens resulting from a her- petofaunal survey of the RNP. In review- ing the species of the RNP, 1 incorporate data from other specimens (Appendix) and from the literature as necessary. 1 have not attempted a comprehensive survey of mu- seum specimens or a thorough study of geographic variation in any species, al- though I comment where appropriate on apparent geographic patterns. I have prob- ably seen most known specimens of doli- cocercus, grandidieri, pinguis, and rhad- inaea, new species. Distributional summaries are based on specimens examined (Appendix), Do- mergue (1973), and Parker (1925). How- ever, I have not verified the identity of specimens at the limits of the ranges for the widespread species epistibes, new spe- cies, lateralis, and infrasignatus; the lit- erature and localities documenting those limits are cited in the species accounts. Most natural history observations are from the RNP region, although for the wide- spread species L. sexlineatus and L. la- teralis, I have included observations from other localities. Such instances are identi- fied in the text. Comments on general mac- rohabitats of prey items (e.g., "arboreal") are from personal observations and, unless otherwise stated, are from the RNP and of active animals; of course, the snakes could 372 Bulletin Museum of Comparative Zoology, Vol. 154, No. 5 likely have captured inactive prey, whose retreat sites are less well known. For convenience I refer to the two spe- cies groups erected by Parker (1925), mod- ified to reflect my view of their composi- tion, as follows: (a) the sexlineatus group: sexlineatus, dolicocercus, grandidieri, pinguis, and rhadinaea, new species; and (b) the stumpffi group: stumpffi, infrasig- natus, lateralis, and epistibes, new species. Additional justification for these groups is given later (see "Monophyly of the Species Groups of Liopholidophis'^) . Hemipenial terminology follows Myers (1973, 1974), Myers and Campbell (1981), and Myers and Cadle (1994). All everted organs were inflated with colored jelly to enhance the surface ornamentation prior to description. In several cases identified below I used a modification of the method of Pesantes (1994) for preparing everted organs from hemipenes originally pre- served partially or wholly inverted. How- ever, rather than neutralization of the po- tassium hydroxide (KOH) treatment with hydrochloric acid, as in Pesantes (1994), I exhaustively soaked the organs in several changes of water followed by several changes of alcohol. This was to avoid pos- sible damage to calcified structures by the acid treatment. The method works well, although it is easy to puncture small, del- icate organs. However, as a cautionary note, hemipenes everted in this way may not assume precisely the same form as organs everted from fresh specimens unless the KOH treatment is sufficient to assure com- plete expansion of the soft tissue (see de- scription of Liopholidophis stumpffi hem- ipenis, later). In my limited experience, the method works better for larger organs than for smaller ones, which are inherently more delicate (hence, I tended to be con- servative in application of the KOH treat- ment). Descriptions of hemipenial mor- phology and skull osteology are relegated to comprehensive sections apart from spe- cies accounts. Inferences of reproductive mode were confirmed, where possible, according to criteria and terminology outlined by Blackburn (1993, 1994). Museum abbre- viations used in the text are given at the beginning of the Appendix. Translations from French and Italian are my own; Mal- agasy names for snakes are translated when their meaning seems evident. Coordinates for localities are given in the text where pertinent and for all local- izable localities in the Appendix. Unless otherwise stated, coordinates were derived from three principal sources: (1) for lo- calities in the vicinity of the RNP, the se- ries of 1:50,000 maps published by the Foi- ben-Taosarintanin'i Madagasikara, Anta- nanarivo (FTM); (2) a series of four 1:1,000,000 maps of Madagascar, also pub- lished by the FTM; and (3) the Defense Mapping Agency (1989) gazetteer. Spe- cific localities within the RNP are mapped in Cadle (1995). A useful discussion of some historical Malagasy collections and locali- ties is given by Carleton and Schmidt (1990), and Claw and Vences (1994:ap- pendix 7) give an abbreviated list of her- petological localities. Malagasy place names are notoriously redundant and highly variable in their spellings (e.g., Nossi-be, Nosy Be, and No- sibe for the island properly referred to as Nosy Be ["Big Island"]). Most names of the colonial period are now reverting to their traditional ones (e.g., Diego Suarez = An- tsiranana; Tamatave = Toamasina). In quoting localities from original sources (e.g., publications, museum catalogs), I use the spelling variants in those sources but give a modern equivalent at least upon the first use; localities are heavily annotated in the Appendix to facilitate cross-referenc- ing. DESCRIPTIONS OF TWO NEW SPECIES The first new species to be described is a member of the genus Liopholidophis Mocquard (1904:302-304) by virtue of having strong sexual dimorphism in tail length (>35% of total length in males, <30% in females), 17 midbody scale rows (reducing to 15 posteriorly), hypapophyses present on posterior trunk vertebrae, max- illary teeth 23-28 -I- 2 ungrooved fangs. LlOPHOLIDOPHIS (COLL'BRIDAE) FROM MADAGASCAR • Cadlc 373 Figure 1. Liopholidophis rhadinaea, holotype (MCZ 180395, male). Approximately xO.9. smooth scales without apical pits, deeply bilobed hemipenis ornamented with spines, and a deeply bifurcate centrolineal sulcus spermaticus. The relative tail lengths in the two sexes, in combination with having 17 midbody scale rows, ally the new spe- cies to the sexlineatus group of Liopholi- dophis (Parker, 1925). However, as sug- gested later (see "Monophyly of Liophol- idophis"), little evidence supports the monophyly of Liopholidophis broadly conceived, and future reevaluation of the status of all included species is warranted. Liopholidophis rhadinaea, new species Figures 1-2, 4-5 Liophidium sp.: Domergue (1988:144, specimen 2). Holotype. Museum of Comparative Zoology (MCZ) 180395 (field number JEC 11466), an adult male in good condition (Figs. 1-2) from Talatakely, Ranomafana National Park, 950-1,000 m, Fivondron- ana Ifanadiana, Fianarantsoa Province, Madagascar [21°16'S, 47°25'E]. Specimen obtained by John E. Cadle 20-26 Decem- ber 1991. Paratypes. Eighteen specimens, 17 in the Museum of Comparative Zoology ob- tained by J. E. Cadle, one in the Museum National d'Histoire Naturelle, Paris (MNHN). All specimens in the MCZ are paratopotypes; data for the MNHN spec- imen are given below: MCZ 180385 (field number JE Cadle 9644), adult female, 26 October 1990; MCZ 180386 (JEC 9649), adult female, 24 October 1990; MCZ 180387 (JEC 9932), hatchling female,^ 19 November 1990; MCZ 180388 (JEC 9933), subadult female, 19 November 1990; MCZ 180389 (JEC 10087), aduh male, 25-28 November 1990; MCZ 180390 (JEC 10115), adult male, 4 December 1990; MCZ 180391 (JEC 10152), aduh female, 9 De- cember 1990; MCZ 180392 (JEC 10610), 'Specimens <135 mm SVL were considered hatchlings. 374 Bulletin Museum of Comparative Zoology, Vol. 154, No. 5 Figure 2. Liopholidophis rhadinaea. head of MCZ 180395 (holotype). Approximately x5. adult male, 15 November 1990; MCZ 180393 (JEC 11180), adult female, fluid + cleared and stained skull, 17 December 1991; MCZ 180394 (JEC 11223), adult male, 18 December 1991; MCZ 180396 (JEC 11564), adult male, 2 January 1992; MCZ 180397 (JEC 11575), adult female, 3 January 1992; MCZ 180398 (JEC 11576), hatchling female, 3 January 1992; MCZ 180399 (JEC 11891), adult female, 11 De- cember 1992; MCZ 180400 (JEC 12344), adult female, 1 January 1993; MCZ 180401 (JEC 12385), adult female, 5 Jan- uary 1993; MCZ 180402 (JEC 12388), adult male, 5 January 1993. MNHN 1988-333 (field number 717/ S), collected 14 January 1966 by M[ichel] Vincke^ "north of Bevatraka, and 22 km north of the terminus of the Perinet for- estry railroad" [Toamasina Province, Fi- vondronana Moramanga] (data translated from field tag attached to specimen). Per- inet (=Andasibe) is at about 900 m on the eastern escarpment [18°56'S, 48°25'E]. This specimen was discussed as Liophidium sp. by Domergue (1988:144, specimen 2), who gave the identical locality except that the initial phrase was reported as "foret de Bevotaka." I have been unable to locate either Bevotaka or Bevatraka in gazetteers or on maps, although Perinet itself is well ^ Listed as "M. Vincke" on the field tag, this is assumed to be the Michel Vincke who collected the type of Geodipsas vinckei, as reported by Domergue (1988:140). known. Domergue (1988) erroneously re- ported the midbody dorsal scale count for this specimen as 15, rather than the 17 that it has. Distribution. Known only from the type locality, Talatakely, within the RNP (21°16'S, 47°25'E), and from near "Beva- traka," 22 km N of Perinet (=Andasibe; 18°56'S, 48°25'E) (Fig. 3). The known ele- vational range is approximately 950-1,100 m at the type locality. Etymology. The specific epithet is a noun in apposition referring to the Neo- tropical snake genus Rhadinaea, many species of which are strikingly similar to Liopholidophis rhadinaea in habitus, col- oration, pattern, and montane forest hab- itat. The name also alludes to the char- acteristic slenderness of both L. rhadinaea and species of Rhadinaea (from the Greek proper name Rhadine, itself derived from rhadinos [=slender, lithe; see Myers, 1974: 16, 19]). Diagnosis. Liopholidophis rhadinaea differs from all other members of the ge- nus by the following combination of fea- tures: dorsal scales in 17-17-15 rows; tail 37-43% of total length in males, 24-27% in females; small size and slender habitus (largest known male 749 mm total length, largest known female 424 mm total length); ventrals 170-179 in males, 150-160 in fe- males; subcaudals 126-135 in males, 69- 77 in females; usually 8 upper labials (but high frequency of 7); 8 or 9 lower labials; and pattern consisting of three light yel- LlOPHOLIDOPHIS (COLUBRIDAE) FROM MADAGASCAR • Ccdle 375 0 L dolicocercus A L. sexlineatus B L. ping u is D L. grandidieri * Ranomafana National Park 200 Kilometers Figure 3. Distribution of species of tiie Liopholidophis sexlineatus group; shaded areas are above 1 ,000 m. All species indicated except pinguls are known from the RNP (locality 5). Liopholidophis rhadlnaea, new species, is known from localities 5 (type locality) and 3. All known localities for species other than sexlineatus are indicated (see text for known distribution, and Glaw and Vences [1994:338] for a more comprehensive map of sexlineatus localities). Localities referred to in the text and Appendix are numbered as follows: (1 ) "Antsihanaka," type locality for L. pinguls Parker; (2) Andrangoloaka, type locality for L. dolicocercus (Peracca); (3) Perinet [Andasibe]; (4) Ambohimitombo; (5) RNP; and (6) Saint Augustine Bay, type locality [in error] for L. grandidieri Mocquard. 376 Bulletin Museum of Comparative Zoology, Vol. 154, No. 5 lowish brown nape spots (Fig. 2), broad dark brown stripe occupying middorsal 3 to 5 scale rows, narrow light yellowish brown dorsolateral stripes centered on scale rows 6 (anteriorly) or 5 (posteriorly), dark brown line on dorsal row 1, and light ven- ter (pink to vermilion in life) (see Figs. 1, 4,5). Liopholidophis rhadinaea differs from species of the stumpffi group in having 17- 17-15 scale rows and strong sexual di- morphism in tail length (19-19-17 scale rows and no strong dimorphism in tail length in the stumpffi group). Four previously described valid nomi- nal species of Liopholidophis {dolicocer- cus, grandidieri, pinguis, sexlineatus; cf. Table 1) have 17-17-15 rows, but all are larger and more robust than L. rhadinaea. Liopholidophis dolicocercus (to 928 mm total length in males, 992 mm in females) has fewer ventrals in both sexes (156-157 in males, 143-150 in females) and more subcaudals in males (140-164), has a dis- tinctively patterned black venter bordered laterally with white stripes, and lacks dis- crete stripes on the dorsum. Liopholido- phis grandidieri (to 1,636 mm total length in males, 674 mm in females) has a black venter, lacks distinct middorsal dark and dorsolateral light stripes (yellowish brown to yellow middorsal area heavily suffused with black or dark brown; lateral dark stripe on rows 2 + 3), and has a longer tail with more subcaudals in both sexes (tail >50% of total length and >200 subcaudals in two males; 35% of total length and 98^- 113 subcaudals in two females). Liophol- idophis pinguis (to 890 mm total length in males, 685 mm in females) has an olive dorsal ground color with dark stripes (sometimes indistinct) and lacks light nape spots; males of pinguis have fewer ventrals (151-154), a shorter tail (33% of total length), and fewer subcaudals (91-98) than males of rhadinaea. Liopholidophis sex- lineatus (to 1,338 mm total length in males, 726 mm in females) differs from L. rhad- inaea in having fewer ventrals (148-163 in males, 139-148 in females), having an olive dorsal ground color with black stripes. lacking light nape spots, and having a whitish belly that may be heavily suffused or mottled with black. Liophidium rhodogaster is sympatric with Liopholidophis rhadinaea at the two known localities for the latter and is very similar in overall appearance, including dorsal pattern and (in life) pink venter (this resemblance was noted by Domergue [1988] in discussing MNHN 1988-333, which he considered an undescribed spe- cies of Liophidium; cf. Figs. 1 and 5 with Glaw and Vences [1994:pl. 339]). Liophi- dium rhodogaster differs from Liopholi- dophis rhadinaea in lacking dorsal scale row reductions (17-17-17), having more ventrals (184-212 in the RNP), lacking ex- treme sexual dimorphism in relative tail length, and having a shorter tail in general (18-23% of total length, sexes combined). Additional comparisons of rhadinaea with Liophidium are given later (Discussion). Data on the Holotype (MCZ 180395). The holotype is an adult male with everted hemipenes. Total length 720 mm; tail length 308 mm (43% of total length). Greatest head width (parietal region) 5.85 mm, head length 11.4 mm from tip of snout to end of mandibles. Dorsals 17-17- 15, the reduction occurring by fusion of rows 3 + 4 at the level of ventral 105. Three preventrals, 179 ventrals, divided anal plate, 126 pairs of subcaudals. 8-8 supralabials (4-5 touching eye), 9-9 in- fralabials, 1+2 temporals on each side. Weight in life 15 g. Description. Measurements, propor- tions, and scutellation are summarized in Table 1. Largest specimen a male (MCZ 180392), 749 mm total length, 320 mm tail length; largest female (MCZ 180399) 424 mm total length, 115 mm tail length. Tail length strongly sexually dimorphic: 37- 43% of total length in males, 24-28% of total length in females. Dorsal scales smooth, lacking apical pits, in 17-17-15 rows. Scale row reduction from 17 to 15 rows usually by fusion of rows 3 + 4 (oc- casionally appearing as loss of row 4) at the level of ventrals 86-113 (males, N = 7) or 78-96 (females, N = 11) (2 individ- LlOPHOLIDOPHIS (COLUBRIDAE) FROM MADAGASCAR • Codle 377 Figure 4. Liopholldophis rhadinaea, MCZ 1 80394 (male), in life. uals had unilateral reduction of 4 + 5 on one side and 3 + 4 on the other). Ventrals 170-179 in males, 150-160 in females. Anal plate divided. Subcaudals 126-137 in males, 69-88 in females (88 subcaudals in MNHN 1988-333; maximum of 77 in the RNP series). Rostral slightly visible from above, about 2 times wider than high. Paired internas- als, each slightly wider than long, about 80% as long as prefrontals. Paired pre- frontals, each wider than long, in contact with each other and with frontal, supra- ocular, preocular, loreal, postnasal, and in- ternasal. Frontal roughly pentagonal (sometimes with a slightly angulate ante- rior border, producing a more hexagonal shape), 1.1-1.3 times longer than its great- est width (at frontal/prefrontal suture), 1.1-1.2 times longer than distance from its anterior edge to tip of snout. Parietals about 1.5 times longer than broad; interparietal suture about 70% length of frontal plate. Nasal divided ventral to nostril, in contact with rostral, internasal, prefrontal, loreal, and first 2 supralabials. Loreal rectangular to pentagonal, usually higher than wide, separated from eye by single preocular (unilateral transverse division of preocular in 3 specimens). Two postoculars; tempor- als 1+2. Supralabials usually 8 with 4-5 touching eye (11 specimens), or 7 with 3- 4 touching eye (5 specimens) (1 specimen each with 7-8 and 7-6). Infralabials 8-8 (6 specimens), 8-9 (2 specimens), or 9-9 (10 specimens), the first pair in contact behind the mental, 1-4 touching an an- terior genial, 4-5 touching a posterior gen- ial (1 specimen with 1-3 and 3-4, respec- tively). Anterior genials approximately equal to, or slightly shorter than, posterior genials. Scattered minute pits or tubercles visible on head plates of some specimens under high magnification, especially on circumorbital series, prefrontals, and na- sals. Overall body form slender, gracile (Fig. 4). Body higher than wide; ventrolateral edge of body angulate. Head very slightly wider than neck. Pupil round. Eye mod- erate, its diameter 60-65% of the distance from anterior edge of eye to tip of snout; eye diameter 1.2 times the distance from eye to posterior edge of nostril. 378 Bulletin Museum of Comparative Zoology, Vol. 154, No. 5 Posterior hypapophyses, examined in situ, appear to be a more or less rectan- gular vane, with a posterior projection that does not, or only barely, overlap the cen- trum of the next vertebra. Dentition. Maxillary teeth 22-28 + 2 (x = 23.9 ± 1.81; N = 16). Modal number of prefang teeth 23 (N = 7), followed by 24 (N = 3), 22 or 25 (N = 2 each), and 28 (N = 2). Essentially no diastema. The un- grooved fangs are about twice as large as the posteriormost maxillary teeth and have a rounded anterior surface and a flattened knifelike posterior surface; their tips are slightly compressed. The ultimate fang is very slightly offset laterad, but the fangs are essentially in line with the tooth row. A cleared and stained skull (MCZ 180393, female) has 16-15 palatine teeth, 25 right pterygoid teeth, and 28 right dentary teeth (left pterygoid and dentary damaged). Hemipenis (see Fig. 30). Deeply bi- lobed, noncapitate, acalyculate (ornamen- tation consists entirely of spines), with small, nude, cylindrical awns at the tips of the lobes. Stalk of organ proximal to lobes moderately long (about 40% the length of the organ). Sulcus spermaticus deeply bi- furcate, centrolineal, with the tips funnel- shaped and opening at the base of the awns. The awns are a very unusual feature of the hemipenes, which are described in de- tail later (see "Hemipenial Morphology in Liopholidophis" ) . Coloration in Life. Two similar but distinct color morphs are evident. Most specimens from the RNP, and the speci- men from Perinet, are a "light" morph; three RNP specimens (MCZ 180385-86, 180388) are a "dark" morph described separately. The two forms differ primarily in the width of the dorsal dark brown stripe and in the shade of the brown flank col- oration. "Light" morph in life, based on MCZ 180392 (male) (Fig. 5)— Dorsum brown, including broad dark brown stripe occu- pying median 3 dorsal rows + V2 of ad- jacent rows, bordered by narrow yellowish brown dorsolateral stripe (centered on row 6 + approximately Vs or less of adjacent scale rows); flanks medium yellowish brown; dark brown line on lower portion of scale row 1 . Top of head brown, without darker patterns. A median and a pair of dorsolateral yellowish brown nape spots. Upper and lower labials and throat whitish with some darker stippling. Anterior 10- 15 ventrals whitish. Remainder of ventrals, anal plates, and subcaudals salmon pink with a few scattered dark brown specks. Dorsal and ventral patterns continue to tail tip. "Dark" morph in life, based on MCZ 180385-86 (females) (Fig. 5)— Broad dark brown stripe occupying median 5 dorsal rows, bordered by narrow yellowish brown stripe from nape to tail tip (centered on row 6 + approximately V3. or less of adja- cent scale rows). Flanks dark brown, of a shade somewhat lighter than the middor- sal dark stripe. A somewhat irregular thin dark brown line on lower half of scale row 1 (manifested posteriorly in MCZ 180385 as a series of irregular spots at the juncture of the ventral plates and scale row 1). Three yellowish brown spots on nape. Top and sides of head brown. Upper and lower la- bials white, speckled with dark grayish or brownish. Throat and anterior ventrals white with some dark pigment on edges of scales. Most ventrals vivid salmon pink with some dark specks laterally. Subcau- dals bright salmon pink. In the "dark morph," the dark flank col- oration occupies the lower 4y2 dorsal rows anteriorly, dropping to the lower SV2 rows posteriorly. Under magnification, these scales are heavily stippled with dark brown, giving a uniform appearance when viewed by eye. In the "light morph," the flanks (first 5 dorsal rows anteriorly, first 4 pos- teriorly) are medium to light brown; under magnification, these scales are light brown, lightly stippled with dark brown. Two specimens, MCZ 180391 and 180400, are somewhat intermediate between the "light" and "dark" morphs: they have a narrow dark middorsal stripe (i.e., 3 rows wide), but their flanks (in preservative) are of a brown shade intermediate between typical specimens of the "light" and "dark" LioPHOLiDOPHis (Colubridae) FROM MADAGASCAR • Cadle 379 morphs. The light dorsolateral stripes vary somewhat in width, the variation due to the proportional involvement of rows 5 and 7 in the stripe (from marginal involvement to approximately V2 of each scale). Although ventral pinkish pigmentation is often variably present in many snakes with otherwise immaculate venters, the pink coloration on the venter of Liophol- idophis rhadinaea is a constant, usually vivid, feature of all specimens. The hue varies from a rather plain pink to brilliant Vermillion. In its most vivid manifestation, the ventral color of L. rhadinaea does not match the brilliant electric hue of the sim- ilar sympatric species, Liophidium rho- dogaster. Most specimens have a small dot of dark brown pigment at the extreme lat- eral edges of the ventral plates; this pig- ment is more extensive in MNHN 1988- 333, in which the ventral plates have dis- tinct darkened borders, than in the RNP sample. Many specimens have additional irregular scattered dark brown flecks on the venter, occasionally arranged in a pair of lines flanking the ventral midline on part of the belly. Coloration in Preservative. Rostral and upper labials mostly white (some fine dark brown stippling, especially on rostral and anterior supralabials). Thin dark blackish line separating whitish upper labial color from the brown head cap; beginning at tip of snout about midlevel on the rostral, ex- tending across upper border of suprala- bials 1-4, thence across lower edge of ven- tral postocular and anterior temporal, and across upper VS-V2 of last two upper labials, ending at corner of mouth. In some some specimens the lower portion of the su- pralabials are also stippled with dark pig- ment, so that the white of the upper lip is essentially sandwiched between dark lines. Top of head brown, slightly lighter than middorsal dark stripe, lightly stippled with dark under magnfication, but essentially patternless. Throat immaculate. Three light nape spots (Fig. 2); lateral ones usually separated from light color of throat by surrounding brown pigment (brown head cap laterally continuous with Figure 5. Liopholidophis rhadinaea, two color morphs. Top: Specimen of the "light" morph (MCZ 180400). Bottom: Spec- imen of the "dark" morph (MCZ 180385). Note the darker flanks in the latter and its narrower dorsolateral light stripe. brown color of flanks); lateral nape spots confluent with light color of throat in six specimens. Nape spots bordered complete- ly or incompletely by thin dark brown line. Middorsal stripe dark brown. Dorsolat- eral light stripes dirty whitish to dirty yel- lowish brown, bordered with thin dark brown line (sometimes incomplete along ventral edge). In some specimens of both color morphs, dorsolateral light stripes es- sentially restricted to scale row 6 (e.g., MCZ 180386, 180396). Light stripes in line with. 380 Bulletin Museum of Comparative Zoology, Vol. 154, No. 5 but separated from, lateral nape spots by a brown collar about 2-3 scales wide. The light stripes continue to the tail tip and are not interrupted in the region of the vent. Dark brown line on lower half of dorsal row 1, sharply separating dorsal and ven- tral ground colors, occasionally indistinct; interrupted briefly at the vent, then con- tinuing at extreme lateral edge of subcau- dal scales to tail tip (subcaudals otherwise immaculate). Venter dull whitish to yel- lowish white, depending on time in pre- servative (pink pigmentation lost), except for the persistent brown dots. Natural History. Liopholidophis rhad- inaea is diurnal and terrestrial. Most spec- imens were encountered while actively crossing trails or (occasionally) apparently sunning on trails. Specimens were collect- ed from selectively logged rainforest that, however, still had a closed canopy and was deeply shaded (essentially as primary for- est) in most parts. These are inoffensive little snakes and do not attempt to bite. One specimen en- countered on a trail used immobility as a defense, flattening its body against the trail and maintaining rigidity; it did not even move initially (even remaining rigid) when prodded or picked up. One specimen at- tempted to take refuge inside a broken bamboo stem close to the ground. Two diet records are available for Lio- pholidophis rhadinaea, both frog eggs. A male collected early in the afternoon of 18 December 1991 (MCZ 180394) regurgi- tated a freshly consumed mass of frog eggs, including 16 more or less intact, plus frag- ments of 1-3 others. The eggs were non- pigmented with yellowish yolk and a ge- latinous capsule. Capsule diameters of the formalin-preserved eggs were 10-12 mm, with the ova 3-3.5 mm. These eggs ap- peared similar to those of Plethodontohyla inguinalis (Microhylidae) observed in the RNP. That species lays clutches in tree holes (Altig and Cadle, unpublished data), often close to the ground, where they might be accessible to a terrestrial snake such as L. rhadinaea. Of course, the identity of the egg clutch remains uncertain, but it seems most likely to be one of the larger cophyline microhylids {Platypelis, Pleth- odontohyla), because these seem to be the only frogs with such large eggs in the RNP (personal observations). Frogs, especially microhylids, appear to be primary dietary items of other forest species of Liopholi- dophis in the RNP (see species accounts). Another male collected 6 December 1990 (MCZ 180390) at 1100 hr. contained four intact egg yolks similar in color, size, and consistency to those described for the pre- vious specimen. Liopholidophis rhadinaea is oviparous. Females apparently begin yolking follicles late in the dry season in the RNP: two females collected 24 and 26 October (MCZ 180385-86) had small yolking follicles. All adult females collected during the rainy season (actual dates 9 December to 14 Jan- uary, including MNHN 1988-33 from Per- inet) had two (four females) or three (four females) well-yolked eggs; eggs in females collected 9-17 December were unshelled oviductal eggs, whereas those collected 1- 14 January all contained shelled eggs. One embryo from MCZ 180401 (collected 5 January) was in Zehr (1962) stage 21-22. Females with yolking follicles or eggs were 262-313 mm SVL. Three small juveniles with umbilical scars (162-225 mm total length; 122-170 mm SVL) were collected on 19 November and 3 January. In the RNP, Liopholidophis rhadinaea is broadly sympatric with the following species of Liopholidophis: lateralis, epis- tibes, new species, infrasignatus {^Hhie- li"), grandidieri, dolicocercus, and sexli- neatus. Of these, all except lateralis, gran- didieri, and sexlineatus are known to be microsympatric with rhadinaea (i.e., to oc- cur in the closed-canopy forest habitat where all specimens of rhadinaea have been collected). In the RNP, lateralis tends to occur in more open habitats, whereas sexlineatus prefers marshy to aquatic hab- itats, and is especially common in rice pad- dies; grandidieri is known from the RNP by a single specimen collected atop a gran- ite massif with rather open habitats (ad- ditional comments later). At Perinet, rhad- LioPHOLiDOPHis (Colubridae) FROM MADAGASCAR • Cadle 381 inaea is broadly sympatric with at least the following species of Liopholidophis: epistibes, new species, lateralis, infrasig- rmtus {''thielf'), sexlineatus, and pinguis (Domergue, 1973; Glaw and Vences, 1994; Appendix). Discussion The superficial similarity of Liopholi- dophis rhadinaea to some species of Lio- phidium is striking, leading Domergue (1988:specimen 2, p. 144) to refer one of the paratypes of rhadinaea to '^Liophi- diiim sp." In fact, Liopholidophis rhadi- naea superficially resembles some species of Liophidiiini (e.g., rhodogaster, torqua- tum) much more than it does other species of Liopholidophis. Hence, it seems worth- while to explore more fully the characters that rhadinaea shares with both genera. A detailed consideration of the relationships of rhadinaea within Liopholidophis is de- ferred until species accounts and detailed hemipenial descriptions of other species are given. Strong sexual dimorphism in tail length, an unusual and unquestionably derived character within colubrids, is the most ob- vious characteristic indicating the rela- tionship of rhadinaea to Liopholidophis (specifically, to the sexlineatus group, for which the character is here interpreted as a synapomorphy; additional comments lat- er). Of the more than 65 species of Mal- agasy colubrids, only species of the Lio- pholidophis sexlineatus group show no overlap between the sexes in the relative tail length compared to the total length; in all species of the sexlineatus group ex- cept pinguis, the tail of males is >35% of total length (averages >40%; see Table 1). In addition, Liopholidophis rhadinaea shares other osteological, scutellational, and pattern characteristics with members of the sexlineatus group (see "Monophyly of the Species Groups of Liopholidophis''). On the other hand, hemipenial morphol- ogy is rather heterogeneous in the Lio- pholidophis sexlineatus group (see "Hem- ipenial morphology in Liopholidophis'); the hemipenis of rhadinaea is no more dissimilar to other members of that group than, for example, are the organs of dol- icocercus compared to either grandidieri or sexlineatus. In contrast to the tail synapomorphy shared between rhadinaea and species of the Liopholidophis sexlineatus group, no special similarities are obvious between Liopholidophis rhadinaea and Liophi- dium. Although synapomorphies for Lio- pholidophis sensu lato have not been iden- tified, some species of Liophidium have derived skull and dentitional characters as- sociated with feeding on hard-bodied liz- ard prey such as skinks and cordylids (Sav- itzky, 1981, 1983). Liopholidophis rhadi- naea shows none of these derived features, which include the following (contrasting characteristics of L. rhadinaea, based on the cleared and stained skull of MCZ 180393, in parentheses): (1) basal hinge allowing teeth to fold toward the back of the mouth (teeth firmly ankylosed to jaws); (2) teeth short, blunt, and often spatulate (teeth sharp, curved, and not short); (3) compound bone of lower jaw strongly curved and articulating far forward, near the anterior end of the dentary (compound bone curved only at tip of the mandible, articulating on the posterior half of the dentary); and (4) long, free posterior den- tigerous process on the dentary (posterior dentigerous process not especially long). Morgan (1973) reviewed Liophidium and compared skulls of four species (may- ottensis, rhodogaster, vaillanti, torqua- tum). An unusual feature of the premaxilla shared by these species was the presence of long lateral processes that overlap the anterolateral surfaces of the maxillae (con- firmed by my study of a skull of L. rho- dogaster [JEC 11571] and photographs of skulls of mayottensis, rhodogaster, and vaillanti in Morgan [1973]). In contrast, the premaxilla and maxillae of Liopholidophis rhadinaea are separated by a moderate gap, which seems to be the common con- dition in Liopholidophis (eight other spe- cies examined, of which the premaxilla and maxilla overlapped in sexlineatus only; 382 Bulletin Museum of Comparative Zoology, Vol. 154, No. 5 see "Osteological Comparisons" for spec- imens examined). I also compared everted hemipenes of LiophoHdophis rhadinaea to those of Lio- phidium rhodogaster and Liophidium torquatum. The hemipenes of these Lio- phidium species are deeply bilobed and spinose, as is the hemipenis of rhadinaea, but otherwise no special resemblances are shared between organs of rhadinaea and the other two. However, as a cautionary note, the organs of L. rhodogaster and L. torquatum are rather different from one another (e.g., basal naked pocket and lobes in torquatum, absent in rhodogaster; per- sonal observations), and they are different from descriptions and figures of three oth- er species given by Domergue (1983). Since hemipenial variation in Liophidium sensu lato remains unstudied, the significance of such differences will only be understood as the hemipenial morphology of it and other Malagasy colubrid genera is com- prehended. Finally, Domergue (1969:15) suggested another "key" character to distinguish spe- cies of Liophidium from LiophoHdophis: venter violaceous, red, or pinkish with reg- ular spotting in the former; yellowish to whitish with irregular spotting in the lat- ter. Clearly, LiophoHdophis rhadinaea, in having a pink to vermilion venter, is an exception to this generality and exception- al among species of LiophoHdophis in this characteristic. Given current definitions and limits for Malagasy colubrid genera, LiophoHdophis is the most appropriate genus for rhadi- naea. Nevertheless, this is not an unequiv- ocal generic placement. Despite a long list of similarities, some of them putatively de- rived, between rhadinaea and the sexli- neatus group of LiophoHdophis (see "Monophyly of the Species Groups of Lio- phoHdophis'), the disturbing lack of clear synapomorphies for LiophoHdophis sensu lato (discussed later) makes resolution of this question problematic. Furthermore, the diversity within LiophoHdophis, the disparity in general habitus between rhad- inaea and the other species, and the re- semblances (albeit superficial) between rhadinaea and species of Liophidium, all convene to raise questions concerning the relationships of rhadinaea. My cursory comparisons of Liophidium species in con- nection with this study raise similar ques- tions for that genus, especially concerning variation in hemipenes and some of the dentitional and cranial characteristics al- ready alluded to. The possibility of a close relationship between Liophidium and Lio- phoHdophis, or parts thereof, should be evaluated as knowledge of species in each genus improves. (See also the subsequent section on MNHN 1988-331.) The next species described has been con- fused with LiophoHdophis stumpffi (Boettger, 1881a,b) in previous literature (Boulenger, 1893; Boettger, 1913; Do- mergue, 1973; Claw and Vences, 1994). Domergue (1973) properly resurrected Dromicus stumpffi Boettger (1881a, b) from the synonymy of LiophoHdophis la- teralis, where it had been placed in pre- vious general reviews of lateralis (Guibe, 1954, 1958). However, Domergue, as had others before (e.g., Boulenger, 1893; Boett- ger, 1913; Kaudern, 1922; Parker, 1925; Angel, 1936), confused a wide-ranging species of eastern and northern Madagas- car with L. stumpffi (Boettger) (Do- mergue, 1973:fig. 1; followed by Glaw and Vences, 1994:336 [map]). My examination of type material of Dromicus stumpffi Boettger, other topotypic specimens, and specimens from eastern Madagascar re- ferable to LiophoHdophis stumpffi sensu Domergue (1973) convinces me that two taxa are involved. Accordingly, LiophoH- dophis stumpffi (Boettger) is here consid- ered a species of the type locality (Nosy Be) and extreme northern Madagascar (Fig. 6). Populations previously confused with stumpffi Boettger (i.e., from the east- ern escarpment and lowlands, the vicinity of Mahajanga in northwestern Madagas- car, and Montagne d'Ambre in extreme northern Madagascar) are described as a new species. Distributional relationships between the new species and true stumpffi are unclear (see "Distribution"). LlOPHOLIDOPHIS (COLUBRIDAE) FROM MADAGASCAR • Cudle 383 % L epistibes A L. infrasignatus M L. lateralis O L. stumpffi * Ranomafana National Park 200 Kilometers Figure 6. Localities for specimens examined of species of the Liopholldophis stumpffi group; these distributions are not comprehensive (see text for known distributions); shaded areas are above 1,000 m. All species indicated except stumpffi are knov\/n from the RNP, which is also the type locality for epistibes. Numbered localities referred to In the text and Appendix are (1) Nosy Be, type locality for L. stumpffi (Boenger); (2) Perinet [Andaslbe], type locality for L. thieli Domergue [= infrasignatus (GiJnther)]; (3) Ankafana, type locality for L. Infrasignatus (Gunther); and (4) RNP. 384 Bulletin Museum of Comparative Zoology, Vol. 154, No. 5 The new species has 19 midbody scale rows and lacks extreme sexual dimorphism in tail length. Hence, it is a member of the Liopholidophis stumpffi group sensu Par- ker (1925) and is to be known as Liopholidophis epistibes, new species Figures 7-1 1 Tropidonotus stumpffi (not of Boettger, 1881a,b), part: Boulenger, 1893:247-248 (specimens f-g, h-i), 1915: 374. Boettger, 1913:312 (specimen from Mora- manga; identity inferred on basis of locality), 1913: 322 (specimen from Nosy Sainte Marie, examined). Kaudern, 1922:445 (identity inferred on basis of locality; see "Distribution"). Liopholidophis stumpffi (not of Boettger, 1881a, 1881b), part: Parker, 1925. Angel, 1936:127 (spec- imens from Tsianovoha; identity inferred on basis of locality). Domergue, 1973:1401; Glaw and Vences, 1994:338 (specimens from eastern Mada- gascar, as discussed later [see "Distribution"]). ? Liopholidophis lateralis (Dumeril, Bibron, and Du- meril), part: Guibe, 1954, 1958. See footnote 10. Holotype. Museum of Comparative Zo- ology (MCZ) 180322 (field number JEC 11460), an adult female in good condition (Fig. 7) from Talatakely, Ranomafana Na- tional Park, 950-1,000 m, Fivondronana Ifanadiana, Fianarantsoa Province, Mad- agascar [2ri6'S, 47°25'E]. Specimen ob- tained by John E. Cadle 20-26 December 1991. Paratypes. The following specimens in the Museum of Comparative Zoology (MCZ) obtained by J. E. Cadle, identical locality data as for the holotype except the elevational range is 970-1,100 m: MCZ 180312 (field number JE Cadle 9646), adult female, 24 October 1990; MCZ 180313 (JEC 9802), adult female, 5 No- vember 1990; MCZ 180314 (JEC 9972), adult female, 24 November 1990; MCZ 180315 (JEC 11078), adult female (skin + complete skeleton), 6-10 January 1992; MCZ 180316 (JEC 11224), subaduh male, 18 December 1991; MCZ 180317 (JEC 11572), subaduh female, 3 January 1992; MCZ 180318 (JEC 11817), aduh male, 7 December 1992; MCZ 180319 (JEC 11836), adult male, 8 December 1992; MCZ 180320 (JEC 1 1890), adult female, 1 1 De- cember 1992; MCZ 180321 (JEC 10609), adult female, 15 November 1990. MCZ 180323 (JEC 11427), adult female (fluid + skull), 27 December 1991, Trail between Ranovao and Menarano, approx- imately 3.5-5 km SSW (airline) Ranoma- fana, approximately 600 m, Fivondronana Ifanadiana, Fianarantsoa Province, Mad- agascar [21°17'S, 47°28'E]. MCZ 180324 (JEC 11797), aduh female, 6-11 January 1992, Trail between Tsaratanana and Am- bohipo, approximately 400-500 m, Fivon- dronana Ifanadiana, Fianarantsoa Prov- ince, Madagascar [2ril'S, 47°37'E]. SMF 57164, adult female, Majunga, NW Madagascar, [Fivondronana Mahajan- ga: Mahajanga Province; 15°43'S, 46°19'E], 7 March 1960, K. L. Koch. SMF 17579, subadult female, St. Marie, E. Madagascar [=Nosy Sainte Marie], [Fivondranana Am- bodifotatra, Toamasina Province; 16°50'S, 49°55'E], about 1905, A. Voeltzkow (see Boettger, 1913:322). SMF 32526-28, adult male and two subadult females, re- spectively. Col [colline (Fr.) = hill] Pierre Radama, Prov. Maroantsetra, [Fivondran- ana Maroantsetra, Toamasina Province; 15°17'S, 50°03'E] [part of H. Bluntschih collection, collected 1931; 1,000 m eleva- tion fide Mertens, 1933; =Vozontanin-d Radama ("Radama Pass") as listed by the Defense Mapping Agency, 1989]. BMNH 89.8.1.8-9 (specimens f-g of Boulenger, 1893:247 [as Tropidonotus stumpffi]), adult males, Tamatave [=Toa- masina], [Fivondranana Toamasina, Toa- masina Province; 18°10'S, 49°23'E], M. Ma- jastre. BMNH 92.3.7.15-16 (specimens h-i of Boulenger, 1893:248 [as Tropidon- otus stump ffi]) , aduh female and male (not individually tagged), Sahambendrana, Central Madagascar^ [Toamasina Prov- ince; 19°24'S, 48°09'E], M. Majastre. ^ Not located on recent maps or in gazetteers, Sa- hambendrana is a type locality for several anurans (e.g., Mantidactylus tornieri [Ahl, 1928]). Ahl (1928: 316-317) gives the locality as "Ankoraka Sahamben- drana (Zentral-Madagaskar)" ("Anhoraka," presum- ably as a misspelling, elsewhere). The coordinates given are for "Ankoraka" listed in the Defense Map- ping Agency (1989) and correspond to this locality as used bv Glaw and Vences (1994). LioPHOLiDOPHis (Colubridae) FROM MADAGASCAR • Cadle 385 USNM 150593-94, adult female and subadult (probably female), respectively, Mt. d'Ambre [=Ambohitra; Fivondronana Antsiranana: Antsiranana Province; 12°30'S, 49°10'E], 1963, Howard E. Uible. Distribution. Known from scattered lo- calities on the eastern escarpment and low- lands, from at least the RNP in the south- east to the Masoala Peninsula in the north- east; Montague d'Ambre and vicinity at the northern tip of Madagascar; the island of Nosy Sainte Marie off the east coast; and the vicinity of Mahajanga in northwestern Madagascar (Fig. 6). A record of ''stump ffi" from Behara (24°57'S, 46°23'E) in extreme southeastern Madagascar (Domergue, 1973:1404; Glaw and Vences, 1994:336) may represent epistibes. The distribution of epistibes in northern Madagascar and its distributional relation- ship with stumpffi are poorly understood. I suspect that all records for ''stumpffi" given by Domergue (1973) and Glaw and Vences (1994), except for the island of Nossi-be (type locality for stum,pffi), ac- tually represent epistibes; the same is probably true for records of "stumpffi" from Fandrarazana (16°45'S, 49°44'E) re- ported by Kaudern (1922:445), from Tsi- anovoha'(=Tsianovoho; 2r57'S, 47°21'E) reported by Angel (1936:127), and from Moramanga (18°56'S, 48°12'E) reported by Boettger (1913:312). Nonetheless, speci- mens of "stum,pffi" from northern Mad- agascar (e.g., Marojezy, as listed by Do- mergue, 1973, and Glaw and Vences, 1994) will have to be reexamined to determine whether or not they are referable to ep- istibes. For example, the specimen from Marojezy just mentioned (Domergue, 1973: table 1) has an unusually high subcaudal count for epistibes females (102; cf. Table 2) but a rather typical one for stumpffi females; it may represent stumpffi sensu stricto (i.e., of Boettger, 1881a,b). All specimens I examined from eastern Madagascar (Masoala Peninsula south) that would be referred to Liopholidophis stumpffi sensu Domergue (1973) and Glaw and Vences (1994) are referred to epis- tibes. I have seen specimens of stumpffi Figure 7. Liopholidophis epistibes, new species, holotype (MCZ 180322, female), in dorsal and ventral views. Approximately xO.46. sensu stricto only from Montague d'Ambre (Antsiranana) and vicinity in extreme northern Madagascar and from the island of Nosy Be (type locality; Fig. 6 and Ap- pendix). It is unclear whether the popu- lations of epistibes around Mahajanga and Montague d'Ambre (Fig. 6) are isolated from the eastern part of the range or whether the distribution of epistibes is continuous throughout northern Madagas- car. Within the RNP, Liopholidophis epis- tibes is apparently widespread, with a 386 Bulletin Museum of Comparative Zoology, Vol. 154, No. 5 known elevational range of approximately 500-1,100 m. One confirmed locality, Ta- matave (=Toamasina), is near sea level, assuming that Tamatave was not simply the shipping point. Etymology. The specific epithet is an adjective meaning "on the trail," modeled after the Greek aTroaTLl3r]s ("off the path"; Liddell and Scott, 1968). From epi ("upon") + stibos ("trail"), in reference to the usual circumstances in which I en- countered this species. Diagnosis. A species of Liopholidophis distinguished from all other species by the following combination of features: scales in 19 rows at midbody, reducing to 17 posteriorly; relatively long tail and high number of subcaudals (27-34% of total length and 86-104, respectively, sexes combined); dorsolateral light stripe on rows 5-7 or 5-6 on neck, 5-6 on anterior part of body, present or absent posteriorly (rows 4-5 when present); dark post ocular stripe more or less continuous with series of dark blotches on side of neck, separating dor- solateral light stripe from light color of throat; anterior 10-30 ventral plates with series of black spots, usually elongate, inset 20-25% the width of the plate from lateral edge (venter otherwise may be more or less immaculate, but usually heavily spot- ted and/or suffused with black or dark gray, especially posteriorly). Liopholidophis epistibes differs from members of the sexlineatus group in hav- ing 19-19-17 dorsal scale rows (vs. 17-17- 15). It differs from other members of the stumpffi group, stumpffi, lateralis, and in- frasignatus, primarily in aspects of color pattern. Liopholidophis epistibes and L. stumpf- fi are separable by the disposition of the dorsolateral light stripes and other pattern characteristics (see "Remarks" for more detailed comparison of specimens of both species from the region of sympatry). In epistibes, the light stripe occupies scale rows 5-6 or 5-7 anteriorly, 5-6 at midbody and, when present posteriorly, rows 4-5; the stripes are not continuous with the light color of the throat (separated by conflu- ence of postocular dark stripe and dark pigment on lateral surface of neck; Fig. 8). In stumpffi, the dorsolateral stripe oc- cupies rows 4-5 anteriorly and at midbody (posteriorly indistinct in adults I have seen, but appears to be restricted to row 4; see "Remarks" in species account for stum,pf- fi); it is confluent with the light coloration of the throat (Fig. 8). The dark postocular stripe is comparatively broad in epistibes, is confluent with dark blotches on the side of the neck (occasionally briefly inter- rupted), and occupies the middle to lower half of the ultimate supralabial (Fig. 8); the postocular stripe in stum,pffi is narrow- er, occupies the upper portion and/or su- ture line of the ultimate and penultimate supralabials, and does not continue pos- terior to the jaw angle (Fig. 8). The dark head cap does not extend well below the jaw line in epistibes, whereas in stum,pffi the dark head cap curves around the angle of the jaw (Fig. 8). The two species also differ in ventral pattern (cf. Figs. 7, 11, and 24): virtually immaculate in stum,pffi except for encroachment of dark flank pig- ment laterally, and usually a series of in- discrete punctations at extreme anterolat- eral edge of ventral plates (not inset from edge); usually heavily spotted or suffused with dark gray or black in epistibes, es- pecially posteriorly, and with series of dis- crete, elongate black spots on each side of anterior 10-30 ventral plates, inset 20-25% from the lateral edges of the plates (Fig. 8). Liopholidophis epistibes averages about 10 more ventral plates in both sexes than stumpffi, and the ranges in the two species are virtually nonoverlapping (Table 2). Hemipenes of epistibes and stumpffi (de- scribed in detail later) also differ: epistibes has about three rows of enlarged spines on the outer surface at the base of each hem- ipenial lobe, whereas stumpffi has only a single row; in addition, nude areas be- tween the lobes are more extensive in ep- istibes than in stumpffi. Liopholidophis epistibes differs from L. lateralis in the position of the lateral stripes: in epistibes on dorsal rows 5-6 or 5-7 an- teriorly, 5-6 at midbody, usually fading LioPHOLiDOPHis (Colubridae) FROM MADAGASCAR • Cadle 387 Figure 8. Comparison of Liopholidophis epistibes, new species (A, C; MCZ 180319), and Liopholidophis stumpffi (Boenger) (B, D; SMF 17577), siiowing the following distinguishing features of head and neck. (1) Postocular bar extending diagonally across ultimate supralabial, intersecting lower edge of ultimate supralabial at the mouth line (epistibes); restricted to upper edge and suture line of ultimate supralabial (stumpffi). (2) Postocular bar confluent with black pigment on side of neck (epistibes); postocular bar short, ending on last supralabial (stumpffi). (3) Dorsolateral light stripe separated from light gular coloration by confluence of postocular bar and black spots on side of neck (epistibes); continuous with light coloration of gular region (stumpffi). (4) No dark wedge from head cap extending below mouth line at angle of jaw (epistibes); dark wedge from head cap extends below mouth line (stumpffi). (5) Discrete, elongate dark spots laterally on anterior ventrals, but inset from lateral edge of ventral plates (epistibes); spots, when present, diffuse, rounded, and present at extreme anterolateral edges of ventral plates (stumpffi). posteriorly (and nearly always indistinct on tail); and anteriorly separated from the light color of the throat by dark pigment on the side of the neck (Figs. 8, 23). In lateralis, the lateral stripes are centered on row 4, with adjacent rows usually involved; the stripes are distinct the length of the body, continue to the tail tip, and are con- tinuous with the light color of the throat (occasionally barely separated by a narrow extension of the dark flank color; Figs. 23, 26; see also Glaw and Vences, 1994:fig. 505). The venter of epistibes may or may not be generally speckled with black (Figs. 7, 11), whereas that of lateralis never ap- pears to be (black spots sometimes present at lateral edges of ventral plates). Liopholidophis epistibes has a longer tail (27-34% of total length, sexes combined) and more subcaudals (91-104, males; 86- 96, females) than infrasignatus (tail 21- 27%; subcaudals 66-81, males; 62-73, fe- males) (see Table 2, including footnote 1 for possible amplification of ranges for ep- istibes). Liopholidophis epistibes is also of more gracile habitus than infrasignatus, and the anterior dorsal colors are predom- inately contrasting black and yellow (browns, olive browns, to olive gray in in- frasignatus). The orientation of the post- ocular dark bar also differs somewhat in epistibes and infrasignatus. In epistibes, the bar extends more or less horizontally posterior to the eye, passing across the up- per portion of the penultimate supralabial (Figs. 8, 23); in infrasignatus, the bar ex- tends at an angle downward across the penultimate supralabial, usually having a somewhat separated portion on the lower portion of the ultimate supralabial (Figs. 23, 28). Data on the Holotype (MCZ 180322). 388 Bulletin Museum of Comparative Zoology, Vol. 154, No. 5 The holotype is a gravid adult female with five eggs, as determined by palpation. To- tal length 753 mm; tail length 218 mm (29% of total length). Greatest head v\/idth (parietal region) 9.75 mm, head length 18.95 mm from tip of snout to end of man- dibles. Horizontal diameter of eye 3.91 mm; anterior edge of eye to posterior edge of nostril 2.87 mm. Dorsals 19-19-17, the reduction occurring by fusion of rows 3 + 4 at the level of ventral 108. Three pre- ventrals, 166 ventrals, divided anal plate, 90 pairs of subcaudals. 8-8 supralabials (4- 5 touching eye), 10-10 infralabials, 2 + 2 temporals on each side, 26 + 2 right max- illary teeth with essentially no diastema. Weight in life 44 g. Description. The following description is based on the 16 females and 8 males in the type series. Measurements, propor- tions, and scutellation are summarized in Table 2 (see footnote 1 in Table 2 for pos- sible extreme values for some statistics re- ported here). Largest specimen a female, 829+ mm total length, tail 195+ mm; largest male 709 mm total length, 238 mm tail length. Tail length not sexually di- morphic, 28-34% of total length in males, 27-31% in females. Dorsal scales smooth, in 19-19-17 rows; 0-2 apical pits present (see "Remarks"). Scale row reduction from 19 to 17 rows by fusion of rows 3 + 4 (occasionally appearing as loss of either row 4 or 3) at the level of ventrals 87-108 (N = 12). Ventrals 157-166 in males, 151- 167 in females. Anal plate divided. Sub- caudals 91-104 in males, 83-96 in females. Eight upper labials (rarely seven or nine) with 4-5 touching eye. Lower labials usu- allv 10-10 (14 specimens), with 8-8 (1), 9-iO (5), 10-11 (1), or 11-11 (1) being uncommon variants; first pair in contact behind the mental, 1-4 or 1-5 touching an anterior genial, 4-5 or 5-6 touching a pos- terior genial. Anterior genials shorter than posterior genials. Loreal present. Preocu- lar single. Temporals usually 2 + 2 (rarely 1 anterior or posterior temporal; in one instance, 3 posterior temporals). Body form slender (Figs. 9-10), slightly higher than wide; ventrolateral edge of body slightly angulate to rounded. Head slightly wider than neck. Pupil round. Eye large (Figs. 8, 23), its diameter greater than the distance between eye and posterior edge of nostril (x = 1.36 ± 0.18; range 1.10-1.69; N = 13). Scattered pits and tu- bercles present on head plates. Hypapophyses (MCZ 180315, complete skeleton) on posterior trunk vertebrae keellike, with a low projecting vane, and a bluntly pointed posterior projection ex- tending beneath the centrum of the next posterior vertebra. Dentition. Maxillary teeth 22-29 + 2 (N - 16; X = 26.4 ± 2.06 prefang teeth). Diastema absent; gap <1 tooth width sep- arating tooth row from enlarged fangs. Ungrooved fangs not offset from tooth row, 2 times as large as the posteriormost max- illary teeth; having a rounded anterior sur- face (except for distal portion, which has a cutting edge) and a flattened knifelike posterior surface. The tips of the fangs are slightly compressed. Two skulls (MCZ 180315 and 180323, both females) have the following numbers of teeth, respec- tively: 17-20, 19-17 palatine teeth; 34-36, 37-38 pterygoid teeth; and 31-30, 32-31 dentary teeth. Domergue (1973) reported 16-17 pre- fang maxillary teeth in Liopholidophis ep- istibes ("stumpffi"), which, in comparison to my counts, suggests failure to count empty sockets or otherwise erroneously low counts. Nevertheless, the range of maxil- lary tooth counts in my series is broad. Any geographic pattern is, however, difficult to discern because most of my counts are from the RNP series, where the range is 24-29 prefang teeth. Hemipenis (see Fig. 34). Deeply bi- lobed, noncapitate, acalyculate (ornamen- tation consists entirely of spines), with a very short basal stalk. Sulcus spermaticus deeply bifurcate, centrolineal. The lobes diverge strongly from one another, lying at essentially right angles to the stalk. Tips of lobes with a central depression ("um- belliform", as described later [see "Hem- ipenial Morphology in Liopholidophis^']) . Coloration in Life (see Claw and I LioPHOLiDornis (Colubridak) from Madagascar • Cadle 389 Figure 9. Liopholidophis epistibes, new species. Specimen from the RNP, MCZ 1 80319, showing typical fading of dorsolateral stripes about midbody. Figure 1 0. Liopholidophis epistibes, new species. Specimen from the RNP, MCZ 1 80323. This specimen has unusually complete and vivid dorsolateral stripes, extending not only the length of the body, but the tail as well. 390 Bulletin Museum of Comparative Zoology, Vol. 154, No. 5 Vences, 1994:pl. 347 of "stumpffi"). MCZ 180312 (female) — Dorsum greenish brown anteriorly, grading to rich medium brown by midbody, to olive brown posteriorly. Yellow dorsolateral stripes begin on nape, turning to yellowish brown but becoming indistinct by midbody. Top of head olive greenish. Black postorbital stripe extend- ing across middle portion of posterior su- pralabial, and continuous with lateral stripe or series of blotches on side of neck. Upper labials, lower labials, and throat pale yel- low. Venter olive greenish with black speckling laterally. MCZ 180313 (female)— Anteriorly, middorsum and flanks black, gradually fading to dark olive brown by Vs of the way along the body. Posterior dorsum dark olive brown with vague obscure black spot- ting and flecks. Dorsolateral pale yellow stripe begins on nape and rapidly fades to yellowish brown, widening and fading posteriorly so that it is barely distinct from middorsal and lateral areas. Top of tail dark brown. Top of head olive brown. Pu- pil round. Iris dark brown. Black postoc- ular stripe extends diagonally across the middle of the ultimate supralabial to angle of jaw. Upper labials anterior to eye olive; posterior ones pale yellow. Lower labials and gular region pale yellow. Anteriorly, ventrals and first scale row pale yellow; black spot inset from lateral edge of ven- trals. All ventrals posterior to approxi- mately the first 10 heavily flecked and spotted with black. Ventral ground color pale yellow anteriorly, fading to brownish yellow, with orangish wash on posterior % of body. Subcaudals dirty white with very light orange tinge, unmarked except for a few scattered black specks. The color plate of Liopholidophis "stumpffi" given by Glaw and Vences (1994:pl. 347; specimen from Perinet [=Andasibel) is of L. epistibes. That spec- imen is similar to coloration of some spec- imens from the RNP, but the dorsolateral stripes in RNP specimens tend to be more intensely yellow. Most specimens have a paired series of discrete black blotches (usually slightly offset) between the dor- solateral stripes on the neck and anterior body and a series along the flanks below the dorsolateral stripes in the same area. Blotches in the lateral series are large (4- 7 scales in diameter), squarish, connected anteriorly with the postocular stripe, and sometimes more or less fused with one an- other. Both the dorsal and lateral blotches become smaller posteriorly, either remain- ing as small punctations the length of the body or disappearing altogether. Coloration in Preservative. Freshly preserved specimens retain much of the original pattern, although the dorsal ground colors become brown to olive brown (blackish anteriorly), and the dorsolateral stripes become greenish yellow, fading to light grayish posteriorly. The stratum cor- neum is lost easily from the dorsal scales, giving such specimens a grayish cast. The venter becomes dirty yellowish or whitish, with dark gray or black markings. The amount of black pigment on the venter varies considerably in the RNP sample — from almost none to heavy spotting or gen- eral suffusion with dark pigment, most prominent posteriorly (Figs. 7, 11). Consistent features of the pattern in Lio- pholidophis epistibes include the dorso- lateral light stripe involving scale rows 5- 6 or 5-7 anteriorly, separated from the light gular coloration by extension of the dark postocular stripe along the neck (Figs. 8, 23). The dorsolateral stripes usually broaden on the nape (Figs. 7, 10), giving the appearance of a pair of light nape spots connected to the stripes. Otherwise, the dorsolateral stripes vary considerably in length and discreteness; in most specimens they fade (but are still evident) by mid- body (Fig. 9), but in others they are dis- crete well onto the tail (Fig. 10). In most specimens, scattered dorsal scales on the anterior body have bright white borders. The lower portion of scale row 1 is lighter (yellowish in life) than the other dorsal rows. A series of discrete, elongate black spots on the anterior 10-30 ventral plates, and inset 20-25% from the lateral edge of the plate, is a constant feature (Fig. 8). Otherwise, the venter is highly variable in I LlOPHOLIDOPHIS (CoLURRIDAE) FROM MADAGASCAR • Ccdle 391 pattern: more or less immaculate (Fig. 11), having an additional median series of spots or continuous line that may run the length of the body (Fig. 7), having irregular dark splotches of varying densities (Fig. 11), be- ing generally suffused with dark pigment and spotting, or some combination of the preceding. Ventral pigmentation is nearly always denser posteriorly than anteriorly. The ventral pigmentation does not seem to develop ontogenetically, as some small juveniles (e.g., MCZ 180316; SVL 197 mm) already show extensive development of posterior spotting on the venter (as well as the anterior ventral spots characteristic of epistibes). Natural History. Liopholidophis ep- istibes is diurnal and terrestrial. Most spec- imens from the RNP were collected ac- tively crossing trails during morning hours, occasionally sunning in leaf litter or bare earth on trails. Most specimens were col- lected from primary montane rainforest, 950-1,100 m elevation; two specimens were from degraded secondary growth at lower elevations in the RNP area. Liopholidophis epistibes dorsoventrally flattens the neck in defensive display, highlighting the white borders to some of the dorsal scales and exposing white patch- es of skin between; it bites readily. The white patches of skin are generally adja- cent to scales with white borders; other- wise, the skin is dark grayish, enhancing the effect of the white patches when the neck is inflated. Two specimens of Liopholidophis ep- istibes contained food. MCZ 180319 (SVL 421 mm) contained one Platypelis polli- caris, a small, nocturnal, arboreal micro- hylid frog, swallowed head first. MCZ 180318 (SVL 390) contained one Pletho- dontohyla alluaudi, a small terrestrial (leaf- litter) microhylid, swallowed head first. Three females from the RNP contained eggs: MCZ 180313 (SVL 522 mm; collect- ed 5 November; 6 unshelled oviductal eggs), MCZ 180314 (SVL 558 mm; 24 No- vember; 6 unshelled oviductal eggs), and MCZ 180322 (SVL 535 mm; 20-26 De- cember; 5 shelled oviductal eggs). No em- Figure 1 1 . Variation in ventral pigmentation in Liopholidophis epistitDes from the RNP. Top: MCZ 180313. Bottom: MCZ 1 80324. The latter specimen had the least ventral pigmentation of any specimen of epistibes examined, but even so still had the elongate spots on anterior ventrals. See also Figure 7. bryos were detected in the first two; MCZ 180322 (holotype) was not dissected. Do- mergue (1973) reported gravid females of Liopholidophis epistibes ("stumpffi") (SVLs 605-675 mm; all from Perinet) con- taining 3-6 eggs in November and Decem- ber, and a female (SVL 567 mm) from Perinet that laid a clutch of 4 eggs (2 ad- ditional eggs unpassed) on 19 December. Thus, Liopholidophis epistibes is ovipa- rous. In the RNP area, Liopholidophis epis- 392 Bulletin Museum of Comparative Zoology, Vol. 154, No. 5 tibes is microsympatric with the following species of the genus: rhadinaea, infrasig- natus, dolicocercus, and lateralis — and more broadly sympatric with grandidieri and sexlineatus. Liopholidophis epistibes is broadly sympatric with L. stumpffi in the vicinity of Montagne d'Ambre in northern Madagascar (see "Remarks"). Remarks. From the region of broad sympatry between epistibes and stumpjfi in the vicinity of Montagne d'Ambre (=Antsiranana; Fig. 6), I examined two specimens of epistibes (USNM 150593-94, adult female and small juvenile, probably female, respectively) and three specimens of stumpffi (MCZ 54368, adult female; MNHN 1893-211, adult male; and USNM 150595, small juvenile male). All of these specimens are readily identified by char- acters given in the diagnosis, with no in- termediacy apparent. The two specimens of epistibes from the Montagne d'Ambre area are somewhat distinguishable from the other specimens of epistibes examined in the following ways that may indicate geographic variation: (1) their ventral counts (151-152) are lower than the range for other specimens (157- 167) (a trend weakly suggested by their low subcaudal counts as well), and (2) the black spots on the anterior ventrals are less elongate, somewhat smaller, and more ir- regular than those of other specimens. Nonetheless, the ventral spotting is typical of epistibes, and none of the specimens of stumpffi showed any approach to this type of patterning. USNM 150593 (adult fe- male) has extensive irregular spotting on the posterior venter, as is characteristic of many specimens of epistibes, but unknown in stumpffi. Domergue (1973) reported apical pits as lacking in Liopholidophis epistibes ("stumpffi"), with the exception of one specimen having two apical pits. My ob- servations revealed that the number of api- cal pits varies from 0 to 2 within an in- dividual, often with only widely scattered dorsal scales having pits. Some individuals had much greater frequency of pits than others; when present, the pits seemed to be more frequent on midbody and pos- terior dorsal scales than on the anterior scales. The Malagasy name tsiririatra is used in the RNP region for Liopholidophis ep- istibes, similar to the name antsiririatra reported by Domergue (1973:1405) for this species in east central Madagascar. SYNOPSES OF OTHER SPECIES OF LIOPHOLIDOPHIS Liopholidophis Mocquard, 1904 Type Species. Liopholidophis grandi- dieri Mocquard, 1904 (designated by Wil- liams and Wallach, 1989:87). Content. Nine recognized species, as follows: Liopholidophis dolicocercus (Per- acca), Liopholidophis epistibes Cadle, Liopholidophis grandidieri Mocquard, Liopholidophis infrasignatus (Giinther), Liopholidophis lateralis (Dumeril, Bibron, and Dumeril), Liopholidophis pinguis Parker, Liopholidophis rhadinaea Cadle, Liopholidophis sexlineatus (Giinther), Liopholidophis stumpffi (Boettger). One undescribed species is recorded from Mon- tagne d'Ambre at the northern tip of Mad- agascar (Raxworthy and Nussbaum, 1994a), from where stumpffi and epistibes are also known (see later). Distribution. Madagascar. Key to Species Most species of Liopholidophis can be distinguished by details of color pattern, tail proportions, and certain scale char- acters (especially dorsal scale rows and subcaudal counts). With the possible ex- ception of separating sexliiieatus and pin- guis in the last couplet, the following key should allow easy identification of speci- mens. Males of sexlineatus and pinguis are easily separated on the basis of relative tail lengths and subcaudal counts, but females are not (and males of the latter, lacking extremely elongate tails, are easily mistak- en for females of the former); previous keys (Parker, 1925; Guibe, 1958; Glaw and Vences, 1994) reliably identify only males. Compounding the difficulty is the fact that LioPHOLiDOPHis (Coli'bridae) FROM Madagasc:ar • Cadle 393 I have seen only 11, mostly rather old, specimens of pinguis, and the extent of variation in mensural and meristic char- acters is unknown in that species; in any case, the variation in such features overlaps considerably when females of sexlineatus and pinguis are compared (cf. Table 1). Parker (1925) commented that pinguis was "of rather stouter habit than its allies" {pinguis [L.] = fat), but that seems clearly true only when comparing larger speci- mens of pinguis to such rather gracile spe- cies as rhadinaea and stump ffi (Parker had only a single specimen of pinguis, and that specimen is exceptionally large, perhaps giving a misleading impression of body form). Similarly, the head plate propor- tions used by Parker (1925:key) to char- acterize pinguis would not likely stand rig- orous scrutiny with additional specimens and statistical comparisons. The pattern characters used in the following key ap- pear to work well for the specimens ex- amined (see also "Species Accounts"), but users of the key should be aware that the characteristics used in couplet 8 may not hold once pinguis is better understood. Much variation also exists in pattern and scalation in the nominal taxa lateralis and sexlineatus, and these taxa may be found to be composites once that variation is more thoroughly studied. Following the key, the species accounts treat the five species of Liopholidophis in addition to rhadinaea and epistibes oc- curring within the RNP and, in addition, provide brief notes on pinguis and stumpf- fi- 1. Dorsal scales in 19 rows at midbody, reducing to 17 posteriorly 2 Dorsal scales in 17 rows at midbody, reducing to 15 5 2. Dorsal ground color black to grayish black; light dorsolateral stripe centered on row 4 (with parts of adjacent rows also involved), continuous and vivid from neck to tail tip, anteriorly confluent with light color of throat; venter usually immaculate except for spots at extreme lateral edges of ventrals in some specimens Liopholidophis lateralis (Dumeril, Bibron, and Dumeril) Dorsal ground color brown, olive, or blackish (may be grayish in preservative); light dor- solateral stripe anteriorly on rows 4-5, 5-6, or 5-7 (often indistinct on posterior body and tail, but usually on rows 4 or 4-5 pos- teriorly when present); dorsolateral stripe anteriorly confluent or not with light color of throat; venter often heavily marked with blackish pigment, which may tend to form midventral line 3 3. Tail short; 23-27% of total length in males, 21-24% of total length in females; fewer than 85 subcaudals in males, fewer than 75 in females; venter with or without dark pig- ment, which may tend to form broken lon- gitudinal lines; dorsolateral light stripe an- teriorly on rows 5-6 Liopholidophis infrasignatus (Giinther) Tail rather long: 28-34% of total length in males, 27-34% of total length in females; more than 90 subcaudals in males, more than 80 in females; venter with or without dark pigment; dorsolateral light stripe an- teriorly on rows 5-6, 5-7, or 4-5 4 4. Dorsolateral light stripe anteriorly on scale rows 5-6 or 5-7, separated from light color of throat by dark pigment; black postocular stripe extending diagonally across middle or lower portion of posteriormost suprala- bial, and continuous with black pigment on side of neck; venter immaculate or (usually) with dark spots or general dark suffusion, especially posteriorly; anterior ventrals with elongate black spots inset 20-25% from edg- es of plates; dark wedge of head cap does not extend ventral to mouth line at jaw an- gle Liopholidophis epistibes Cadle Dorsolateral light stripe anteriorly on scale rows 4-5, continuous with light throat pig- ment behind angle of jaws; black postocular stripe on posteriormost supralabial restrict- ed to uppermost part of scale and/or dorsal suture line, ending at posterior supralabial (not continuous with dark lateral neck pig- ment); venter essentially immaculate ex- cept for dark dorsal pigment narrowly en- croaching laterally; dark spots on anterior ventrals, when present, at extreme lateral edges of plates; dark head cap extends as a wedge ventral to mouth line at jaw angle Liopholidophis stunipffi (Boettger) 5. V'entral scutes (except for anteriormost) solid black, or black with regular creamy white border (forming cream stripe at lateral edge of ventral plates). Dorsum uniform brown- ish; with black reticulations, blotches, or chevrons (especially posteriorly); or with general black suffusion. Lateral or ventro- lateral black stripe may be present on rows 2-1-3, and /or row 1 -I- adjacent edge of ventrals 6 Ventral scutes never solid black, or black with bordering cream-colored stripe (may be im- 394 Bulletin Museum of Comparative Zoology, Vol. 154, No. 5 maculate to heavily, but irregularly, splotched or patterned with dark pigment). Dorsum distinctly striped or not (when present, consisting of dark and light brown stripes, or lateral black stripes); never with dark reticulations, paired blotches, or chev- rons 7 6. Venter solid black (no white or cream stripe at lateral edge of ventral scutes); lateral black stripe on rows 2 + 3; subcaudals more than 200 in males, more than 100 in females . . Liopholidophis grandidieri Mocquard Venter black with cream or white stripe at lateral edges of ventral scutes; black stripe at juncture of ventral plates and first dorsal row, but no lateral black stripe on rows 2 + 3; subcaudals less than 200 in males (high- est observed, 164), less than 100 in females (highest observed, 88) Liopholidophis dolicocercus (Peracca) 7. Size diminutive and slender (maximum <750 mm total length in males, <500 mm in fe- males); 3 light (yellowish in life) nape spots; striped pattern consisting of contrasting shades of brown with broad median dorsal dark brow n stripe 3-5 scales wide, bordered by narrow light yellowish brown stripe cen- tered on row 6; venter pink to red in life, unmarked except for occasional fine dark peppering . . .Liopholidophis rhadinaea Cadle Size larger and more robust (maximum >850 mm total in males, >650 mm in females); no light nape spots; dorsal pattern striped or not, but stripes black when present; ven- ter not pink to red in life, often densely marked with dark pigment 8 8. Relative tail length not strongly sexually di- morphic (in males 30-35% of total length, in females approximately 25% of total); sub- caudals in males <110; stripes usually in- distinct at least on anterior part of body, more distinct on posterior body and tail (when present, consisting of blackened su- ture line between ventrals and dorsal row 1, and lateral stripe or series of spots or dashes on row 3 anteriorly, dropping to su- ture line between 2 and 3 posteriorly; venter not heavily marked with black (some spec- imens with edges of ventrals marked with black, or with lateral or median series of small spots) . . . Liopholidophis pinguis Parker Relative tail length strongly sexually dimor- phic (in males >40% of total length, in fe- males 24-30%); subcaudals in males >120; stripes distinct entire length of body and tail (consisting of black stripe at border be- tween ventrals and dorsal row 1, and lateral stripe involving rows 2-1-3, occasionally 4; indistinct stripe sometimes present at the suture between rows 6 and 7; occasionally rows 1-3 and adjacent venter are entirely black); venter more or less immaculate, to heavily and irregularly marked with black Liopholidophis sexlineatus (Giinther) The sexlineatus Species Group (Parker, 1925) Figures 12-22 (see also Figs. 1-2, 4-5); Table 1 Content. Dronncus sexlineatus Giinther, 1882:264. Dromicus dolicocercus Peracca, 1892:1-3. Liopholidophis grandidieri Mocquard, 1904:304. Liopholidophis pinguis Parker, 1925:390. Liopholidophis rhadinaea Cadle, present work. Liopholidophis dolicocercus (Peracca) Figures 12-17 Dromicus dolicocercus Peracca, 1892:1-3, fig. la-d (Type locality, "Valle dellUmbi (Andrangoloka)" [Valley of the Umbi River (Andrangoloka)] [=An- drangoloaka]. Holotype, Museo Regionale di Scienze Naturali, Torino (MZUT) 796 (Fig. 13). Peracca (1892:3) was explicit about basing his specific ep- ithet on the Greek word doXixos but incorrectly transliterated the name as dolicos, rather than cor- rectly as dolichos. Under Article 32b-c of the In- ternational Code, however, his name stands as the correct original spelling, despite having been un- justifiably emended by all subsequent authors ex- cept Parker (1925). The name is here resurrected from the synonymy of Liopholidophis sexlineatus (e.g., Guibe, 1958; see "Remarks"). Liopholidophis dolichocercus [unjustified emenda- tion] (Peracca): Mocquard, 1904:302, 1909:43, 97; Werner, 1929:11. (Elsewhere in Mocquard's 1904 paper, the incorrect spelling dolischocercus is found.) Tropidonotus dolichocercus [sic] (Peracca): Boettger, 1898:25, 1913:312; Boulenger, 1893:246, 1896:607, 1915:373. Boulenger's (1893, 1896) listing of two females in the British Museum under this name are based on misidentified specimens of the then-un- described and very similar species Liopholidophis grandidieri Mocquard (1904) (these specimens are discussed under the species account ior grandidieri, later). Boulenger listed both dolichocercus [sic] and grandidieri in the 1915 paper, but his concept of the former seems to have been based on the same misidentified specimens he had cited earlier. Liopholidophis dolicocercus (Peracca): Parker, 1925: 392. Liopholidophis sexlineatus (Giinther), part: Guibe, 1958:216; Brygoo, 1983:39 (footnote 29) {Dromicus dolichocercus [sic] Peracca listed as synonym). Holotype (Fig. 13). Museo Regionale di Scienze Naturali, Torino (MZUT) 796, a male in fair condition (probably subadult based on size), 427 mm total length, 162 LioPHOLiDOPHis (Colubridae) FROM MADAGASCAR • Cadle 395 Figure 12. Heads of four species of the Liopholidophis sex- lineatus group, in dorsolateral view (see also Figs. 2 and 19 for L. rhadinaea and L. grandidieri, respectively). A. L. dolico- cercus (Peracca) (MCZ 180405). B. L. grandidieri Mocquard (BMNH 95.7.4.32). C. L. sex//nea fus(Gunther) (MCZ 180331). D. L pinguis Parker (USNM 149242). mm tail length (38% of total length), with 160 ventrals and 164 subcaudals (Peracca, 1892; see "Remarks" concerning subcau- dals). I examined color transparencies of the type, including details of head and Figure 13. Liopholidophis dolicocercus (Peracca), dorsal and ventral views of the male holotype (MZUT 796). Photographs by Dr. Franco Andreone. body. Peracca's (1892) description of the type is excellent. Diagnosis. A species of Liopholidophis having 17-17-15 dorsal scale rows; tail 39- 44% total length and 140-164 subcaudals in males; and a black venter with a mar- ginal white stripe, a black stripe at the suture between the ventral plates and dor- sal row 1, but no lateral stripe involving rows 2-3. Liopholidophis dolicocercus differs from sexlineatus, where it has been synony- 396 Bulletin Museum of Comparative Zoology, Vol. 154, No. 5 en S 0 2 o 3 z £ fc 0 < a: H J z >- < < Q Z > H U a < u X c/5" en Qh en en U 0 K CC ' H w lO C/5 Z o 0 u CC y -^ U < 3 0 ~ — 0. 0- CC C/) U. z 0 0 i CC g en CC < en U 0 Oh Q t/i 03 Z Z 5 C/) 0 < w H u < CC O z < CC < J CQ e32 00 < X K < < > u > U < CC H 'w N u '.« N en Dh So J S J 0- Q CL. < Z < en < en '~' < Q C/5 CC C/5 D +1 +1 Z 'H iH Z o Z ^ O r H 5 O 0. O O tu CC , Q Z O < CQ U.. ^ + ^ - — — - ^-^ .^.^ , , ^-^ ^^ + 5? CD in CO -^ CO o t- CO O 00 •-I -T -H CD o ^ (M ^ CO ^ -^ 05 (N t^ lo in O CO O (N CD d d o 1 + +id .^ 1 •f 1 tJ- +1 d> -H 00 +1 tJ- + +1 ci -H in '^ CO 00 _ cr> o ■^ ^ <^ rr, "^ S2 I-- 00 ^ ^ — ( CO di 05 CD 00 CD ??oSd + -i eg Od in -^ 05 tD d d O o 2 ° ^ Ss^o^ '^p 7 oi 1 c^ 1 CO 1 t^ , , "^ ,, t^ , , t^ -H ^ -l-l T). -H (N -f-l CD ■^ ^^ f^ — '■ o -^t- CO in d CO in -^ in t^ CO ^ eg _( +1 CD +1 O r^ in • CO ■^ ^ ^ ^ o> ^ Tf cn -l-l CO -H 05 ^ eg CD d '^ CO r^ in in CO ^ 032 0.51 015 00 O) CO -^ CO CD o ^ o +1 ^ +1 00 eg eg 05 CT> Oi 05 eg o TT CO + 00 CD 5 O eg eg d^ O CO +1 2. in -co in CO CD ^^in — . eg Tf o o O ^ O CO d d d d 6g ^ I + <^in + ^ eg eg + So eg eg eg I + + d 05 ■"* Q > bO *2 c ,^ V o -i -C s C^ a; at) 3 5 o H "c3 c _4J X c^ S 2 H 6 = o ■<1" -l-l ■^ o 1 ■^ ^ 05 d O to^ d lo eg +1 ^ 50% of total length in males (35% in females); venter (except for anterior ventral plates) entirely black, in- cluding the anal plate, and not bordered by a marginal white stripe; lateral black stripe on dorsal rows 2-4 anteriorly and 2-3 posteriorly, but dark dorsal stripes oth- erwise lacking. These features distinguish Liopholidophis grandidieri from all spe- cies of Liopholidophis, none of which have such proportionally long tails; all species but dolicocercus have dorsal stripes (light or dark) on scale rows other than 2 + 3. Liopholidophis grandidieri is most easily confused with L. dolicocercus; distinguish- ing features are given in the account for that species. Distribution (Fig. 3). Known definitely from the RNP (Mt. Maharira) and from Ambohimitombo Forest, a locality well known from specimens collected by For- syth Major (e.g., Boulenger, 1896; Major, 1896). The type locality, "I'embouchure du Saint-Augustin" (mouth of the Saint- Augustin River), is in the arid southwest- ern sector of the country (23°33'S, 43°46'E; Fig. 3) and almost certainly in error. All other specimens are from the eastern rain- forest belt (Appendix). The type of gran- didieri was the only specimen recognized until recently. The two documented lo- calities are approximately 70 km apart in the central part of the eastern escarpment (Fig. 3). The descriptor "eastern Imerina" (lo- cality for BMNH 95.10.29.52) refers to the territory on the eastern edge of the es- carpment between approximately paral- lels 18° and 21°, the Imerina being one of the indigenous peoples of the central pla- teau (see, e.g., Gallieni, 1908:pl. 6). I have tried, without success, to verify the doc- umentation and localities for the records listed by the UICN/PNUE/WWF (1990: 223) as "three new specimens from the eastern forests." The single specimen from the RNP was collected near the highest point in the park, 1,375 m (Mt. Maharira; see later). BMNH 95.7.4.32 is from Ambohimitombo Forest, presumably near the town of that name, which is at approximately 1,200 m ele- vation. Description. The following description is based on examinaton of two males (in- cluding the holotype) and two females. Measurements, proportions, and scutella- 404 Bulletin Museum of Comparative Zoology, Vol. 154, No. 5 tion are summarized in Table 1. Largest specimen the male holotype, 1,636 mm total length, 904 mm tail length; largest female (BMNH 95.7.4.32) 674 mm total length, 238 mm tail length. Tail length strongly sexually dimorphic, 54-55% of to- tal length in males, 34-35% in females (Figs. 18, 20). Dorsal scales smooth, lacking apical pits, in 17-17-15 rows. Scale row reduction from 17 to 15 rows by fusion of rows 3 + 4 at the level of ventrals 92-112 (N = 3). Ventrals 169-171 in males, 147- 161 in females. Anal plate divided. Sub- caudals 215-221 in males, 113 in female with complete tail (98+ in female with in- complete tail). Eight upper labials with 4- 5 touching eye; 9 lower labials, the first pair in contact behind the mental, 1-4 touching an anterior genial, 4-5 touching a posterior genial. Anterior genials shorter than posterior genials. Two postoculars; temporals 1+2. Body slightly higher than wide; ventro- lateral edge of body angulate. Head slight- ly wider than neck. Pupil round. Eye rel- atively large, its diameter greater than dis- tance between eye and posterior edge of nostril (x = 1.25 ± 0.1; range 1.16-1.38; N = 3), its diameter 60-65% of the distance from anterior edge of eye to tip of snout. Scattered minute pits and tubercles appear to be present on the anterior head plates. Dentition. Maxillary teeth 20-23 + 2 (N = 4). Diastema short or absent, one tooth width or less. The ungrooved fangs are less than twice as large as the poster- iormost maxillary teeth, have a flattened knifelike posterior surface, and have a rounded anterior surface except for the distal tip, which is slightly compressed and has a short cutting edge. The fangs are essentially in line with the prefang teeth (i.e., not offset). A prepared skull (MCZ 180297, male) has 20-23 maxillary teeth, 16-18 palatine teeth, 25-29 pterygoid teeth, and 30 right dentary teeth (left dam- aged); the diastema in this specimen is about the width of the preceding teeth, and the fangs are not offset. Hemipenis (Fig. 32). Deeply bilobed, noncapitate, acalyculate (entirely spinose). with a basal stalk comprising slightly less than V2 the length of the organ. Sulcus sper- maticus deeply bifurcate, centrolineal. Coloration in Life. I have not seen Lio- pholidophis grandidieri in life. However, given the overall exceedingly similar pat- terns of grandidieri and dolicocercus, I suspect that the two species have similar coloration in life (see species account for dolicocercus). Coloration in Preservative (MCZ 180297). The specimen is perhaps some- what excessively darkened as a preserva- tion artifact; its pattern is less distinct than that of the holotype. Anteriorly, a mid- dorsal series of irregular blotches or paired spots, separated by whitish interspaces. The dorsum rather abruptly darkens shortly af- ter the neck region, and most of the dor- sum appears blackish. Many dorsal scales of all rows except the first have distinctly white borders; these are more evident an- teriorly, posteriorly becoming obscured by increasing black pigment. An indistinct brownish streak is present on the flanks, and a brownish gray streak on scale rows 1-2. Black lateral stripe continuous with postocular stripe, on rows 3-5 immediately behind head, soon dropping to rows 2-4 for much of the body, and to adjacent por- tions of rows 2-3 posteriorly; the lateral stripe is indistinct and disrupted anteri- orly, but very distinct, continuous, and with regular borders posteriorly. Throat and several anterior ventrals white, but re- mainder of venter and lower part of scale row 1 solid black; anal plates black. Top of head brownish; black postocular stripe continuous with lateral black stripe. Some black pigment on upper edges of supra- labials, preoculars, loreals, nasals, and ros- tral; supralabials, infralabials, and gular region otherwise white. Tail with a mid- dorsal, midventral, and a pair of ventro- lateral black stripes (at border between subcaudals and dorsal caudal scales); the midventral stripe becomes thinned and broken toward the tip, but stripes other- wise continue to the tip. Subcaudals oth- erwise white; dorsal caudals otherwise brownish. LioPHOLinoPHis (Colubridae) from Madagascar • Cadle 405 ^'-^^^^\ ^'^^i^S^^' \\\\ ^^^s Bll^ ytk^^^i/^Si .%v yr^^n^^- -^jii|flW ^P^ B / ^^(Bd^B wK^^^ Figure 18. Liopholidophis grandidieri Mocquard, dorsal and ventral views of the male holotype (MNHN 02-103). The two BMNH females (Fig. 20) and the holotype (Figs. 18-19) of grandidieri are similar to MCZ 180297, except in hav- ing somewhat more distinct lateral stripes and in tending to have the black ventral coloration broken up somewhat more an- teriorly and posteriorly. Many of the dorsal scales of these specimens also have vivid white borders, as in MCZ 180297. The black stripe on the subcaudal scales is 406 Bulletin Museum of Comparative Zoology, Vol. 154, No. 5 Figure 19. Liopholidophis grandidieri Mocquard, dorsal and lateral views of the head of the holotype (MNHN 02-103). somewhat variable, being more or less con- tinuous to the tail tip (MCZ 180297), es- sentially absent (BMNH 95.7.4.32, 95.10.29.52; a few blackened suture lines) or continuous anteriorly but absent pos- teriorly (MNHN 02-103). Natural History. The RNP specimen was collected 13 April 1992 during the day near the summit of Mt. Maharira, a gra- nitic massif with expanses of bare rock, grassy areas, and low scrubby forest (prob- ably resulting from thin soil over bedrock). Two females, BMNH 95.7.4.32 (SVL 436 mm) and 95.10.29.52 (SVL 412 mm), have large oviductal eggs (five and four, re- spectively, as ascertained by palping) cov- ered by a thickened leathery shell. BMNH 95.7.4.32 was obtained at Ambohimitom- bo by Forsyth Major, who collected there 12-24 January 1895 (Carleton and Schmidt, 1990:table 1). One egg from BMNH 95.7.4.32 contained an embryo in Zehr (1962) stage approximately 23-24. The relatively advanced embryo sur- rounded by a leathery shell suggests ovi- parity in grandidieri according to criteria outlined by Blackburn (1993). Remarks. Liopholidophis grandidieri was described along with a heterogeneous assortment of amphibians and reptiles from Africa and South America (Mocquard, 1904), and neither a collector nor donor of the type was stated. The only other Mal- agasy species described in the same paper, Pseudoxyrhopus dubius {=tritaeniatus\ cf . Raxworthy and Nussbaum, 1994), was said to have been "sent to the [Paris] Museum, without indication of locality, by M. Rous- son, explorer" (Mocquard, 1904:306). Lio- pholidophis grandidieri was described during a period of accelerated French ex- pansion and exploration in Madagascar (Gallieni, 1908), and the type may have been obtained by any number of French political administrators, explorers, or med- ical or military personnel on the island at the time." The type locality. Saint Augus- tine Bay, was a major port and shipping point during the period (see, e.g.. Bastard, 1898), and the type was probably sent to Paris with the "locality" recorded as the shipping point. Two females of Liopholidophis gran- didieri were erroneously referred to L. dolicocer cushy Boulenger (1893:246-247, 1896:607): BMNH 95.10.29.52, collected bv the Reverend R. Baron in "East Im- erina," and BMNH 95.7.4.32 (Fig. 20), col- lected by Dr. Forsyth Major in the "Am- bohimitombo Forest." That these are the specimens Boulenger had in hand is sug- gested by the associated collectors, locality data, measurements (for 95.10.29.52), sex, ventral and subcaudal counts, and ventral pattern (described in detail later), as re- ported by Boulenger (1893, 1896). My ' The Bulletin du Museum d'Histoire Naturelle during this period contains numerous references to collections received from Madagascar. For example, volume 4 (1898, no. 2, p. 4), includes the following: ". . . the arrival of a crate sent from Tamatave [=Toa- masina] by Captain Ardouin and containing some reptiles, diverse arthropods, several molluscs, and two Hova skulls" (the Hova being one of the indigenous peoples). Rarely, it seems, were these notices sufficient in themselves to relate to specific collections or spec- imens. LlOPHOLIDOPHIS (COLIIBRIDAE) FROM MADAGASCAR • Cadlc 407 Figure 20. Liopholidophis grandldieh Mocquard. Dorsal and ventral views of a female from "Ambohimitombo Forest" (BMNH 95.7.4.32). This specimen was referred to (Liopholidophis) dollcocercus by Boulenger (1896:607; specimen b). counts for ventrals and subcaudals differ from Boulenger's by at most two (Table 1), and those differences may be accounted for by a somewhat damaged tail in BMNH 95.7.4.32 and the inclusion of preventrals in the ventral counts given by Boulenger. Boulenger's erroneous referral of these specimens to dolicocercus is understand- 408 Bulletin Museum of Comparative Zoology, Vol. 154, No. 5 able in view of the fact that he was com- paring two females with the description of the male holotype of dolicocercus, in a genus known to exhibit exaggerated sexual dimorphism in subcaudal counts. Boulen- ger failed to note Peracca's explicit de- scription of one of the diagnostic pattern differences between dolicocercus and grandidieri: "... [in L. dolicocercus] a black line . . . runs in part along the mar- gins of the ventrals, in part on the inferior series of the scales of the body. This line is bordered with yellowish white along its inferior margin" [i.e., at the lateral edges of the ventral scutes; emphasis added] (Peracca, 1892:2). The two BMNH speci- mens have the alternative pattern char- acteristic of grandidieri, viz., an entirely black venter (no white stripes), separated by a light interspace from lateral black stripes on dorsal rows 2 + S: "belly black, separated from the lateral streak by a yel- lowish interspace or streak" (Boulenger, 1893:247). Significantly, Boulenger does not mention the two ventral pattern fea- tures diagnostic of dolicocercus (see "Di- agnosis" in the dolicocercus account for other differences). Liopholidophis sex ti neat us (Gijnther) Figures 12, 21 Dromicus sexlineatus Giinther, 1882:264 (Type lo- cality, "Eastern Betsileo"). Syntypes, The Natural History Museum, London (BMNH) 1946.1.13.17- 19 [examined]. Dromicus macrocercus Giinther, 1882:265 (Type lo- cality, "Eastern Betsileo"). Syntypes, BMNH 1946.'l. 13.28-30 [.28 not seen]. Peracca, 1892:2. Boulenger, 1893:246 (synonym of Tropidonotus sexlineatus). Mocquard, 1909:95 (synonym of Lio- pholidophis sexlineatus). Leptophis varius Fischer, 1884:36 (Type locality, "Madagascar"). Syntypes, five specimens in the Natural History Museum in Hamburg, nos. 1174- 75 fide Fischer (1884:38) [not seen]." Boulenger, * I am blindly following Boulenger (1893) in rel- egating varius to the synonymy of sexlineatus, which seems likely for some or all ot the syntypes of varius. In that case, all of the types would be females or else males with incomplete tails, given their subcaudal counts and relative tail proportions (Fischer, 1884:38; 1893:246 (synonym of Tropidonotus sexlineatus). Guibe, 1958:246 (synonym of Liopholidophis sex- lineatus). Dromicus dolicocercus Peracca, 1892:1 (Type local- ity, "Valle deirUmbi [Andrangoloka] "): Guibe, 1958 (synonym of Liopholidophis sexlineatus). Here considered a valid taxon. Tropidonotus sexlineatus (Giinther): Boulenger, 1893: 246, 1896:607, 1915:373. Jourdran, 1903:34. Liopholidophis sexlineatus (Giinther): Mocquard, 1904:303, 305; 1909:43, 95; Boettger, 1913:372; Parker, 1925:392; Werner, 1929:11; Guibe, 1958: 216; Domergue, 1969:19; Brygoo, 1983:55, 1987: 24; UICN/PNUE/WWF, 1990:223; Glaw and Vences, 1992:266, 1994:338; Nicoll and Langrand, 1989:135. Syntypes. BMNH 1946.1.13.17-19 (old numbers 82.5.8.3, 82.5.8.2, and 82.5.8.4, respectively), all three adult females ob- tained by Rev. W. D. Cowan. Scale counts, measurements, and other data for the syn- types are, respectively, as follows: ventrals: 146, 145, 144 (preceded by 1, 2, and 1 preventrals); subcaudals: 74, 73, 77; anal divided; total lengths (mm): 605, 620, 589; tail lengths (mm): 175, 172, 175 (29%, 28%, 30%, respectively, of total length); maxil- lary teeth: 23 + 2 in each case. Diagnosis. Liopholidophis sexlineatus is distinguished from members of the stumpffi group in having 17-17-15 scale rows (vs. 19-19-17). It differs in details of color pattern from dolicocercus, grandi- dieri, and rhadinaea: venter largely solid black, or solid black with marginal white stripe, in grandidieri or dolicocercus, re- spectively; whitish to heavily, but irregu- larly, splotched with black (never solid black) in sexlineatus; three light nape spots cf. Table 1). Nonetheless, given the difficulty of sep- arating females of sexlineatus from both sexes of pinguis, some of the syntypes of varius could be the same as the later-described taxon pinguis. In partic- ular, Fischer's "specimen c" has 92 subcaudals, an unusually high count for sexlineatus females, but a typical one for pinguis males (Table 1) (assuming a complete tail and correct counts); other details given for this specimen conform to either pinguis or sex- lineatus. Fischer (1884) also alludes to varying de- velopment of the lateral and ventrolateral stripes ("often beginning at the middle of the body"), and paired ventral spots, in his series of varius, these char- acteristics also suggest pinguis (see species account). LlOPHOLIDOPHIS (COLUBRIDAE) FROM MADAGASCAR • Cadle 409 and dorsolateral light stripe in rhadinaea (absent in sexlineatus). Liopholidophis sexlineatus is most sim- ilar to L. pUiguis in overall habitus and color pattern. Males of pinguis have short tails ( <35% total length; < 1 10 subcaudals) compared to males of sexlineatus (>40% total length; >120 subcaudals), and the relative eye size of pitiguis is smaller than that of sexlineatus (Table 1). Most speci- mens of pitiguis have less distinct stripes on the anterior body than sexlineatus, and the venter in the former is usually light- colored (sometimes with edges of ventrals blackened or with a median series of small paired spots); in sexlineatus, the stripes are distinct the entire length of the body (flanks may be entirely darkened) and, although the venter may be immaculate, it is often heavily and irregularly splotched with black. Distribution (Fig. 3). Scattered locali- ties on the eastern escarpment and low- lands from the southeastern tip of Mada- gascar (Glaw and Vences, 1994:336 [map]), north to at least Toamasina (Toamasina Province); a few localities on the high pla- teau as documented in the Appendix ("Manjakatompo") and as shown by Glaw and Vences (1994:336 [map]). Although seemingly locally abundant where it oc- curs (personal observations), Liopholido- phis sexlineatus is recorded from relative- ly few localities in the literature. In the RNP region, Liopholidophis sex- lineatus appears ubiquitous in rice paddies and marshy areas and is known from ap- proximately 500 to 1,130 m elevation. Description. The following description is based on examinaton of 18 females and 15 males, including types of Dromicus sex- lineatus and D. macrocercus. Measure- ments, proportions, and scutellation are summarized in Table 1. Largest specimen a male (MCZ 11604), 1,338 mm total length, 663 mm tail length (50% of total); largest female (MCZ 11701) 726 mm total length, 238 mm tail length (33% of total). Tail length strongly sexually dimorphic, 41-51% of total length in males, 24-33% in females. Dorsal scales smooth, lacking apical pits, in 17-17-15 rows. Scale row reduction from 17 to 15 rows by fusion of rows 3 + 4 at the level of ventrals 90-115 (N = 17; 1 specimen showed fusion of 4 + 5 at the level of 92-95). Ventrals 147- 163 in males, 139-148 in females. Anal plate divided. Subcaudals 127-160 in males,^ 67-91 in females. Eight upper la- bials with 4-5 touching eye. Lower labials 9-9 (N = 12), 9-10 (8), 10-10 (11), or 11- 12 (1), the first pair in contact behind the mental, 1-4 or 1-5 touching an anterior genial, 4-5 or 5-6 touching a posterior genial. Anterior genials shorter than pos- terior genials. Loreal present. Preocular usually single (occasionally 2). Temporals 1 + 2 (rarely 3 secondary temporals). Body slightly higher than wide; ventro- lateral edge of body rounded in females, slightly more angulate in males. Head wid- er than neck. Pupil round. Eye relatively small, its diameter approximately equal to or somewhat greater than distance be- tween eye and posterior edge of nostril (x = 1.1 ± 0.16; range 0.95-1.44; N = 16). A few scattered pits on anterior head plates. Dentition. Maxillary teeth 17-26 + 2 (N = 25; X = 23.9 ± 1.9 prefang teeth). Diastema essentially absent; gap <1 tooth width separating tooth row from enlarged fangs. Ungrooved fangs not offset from tooth row, twice as large as the posterior- most maxillary teeth; having a rounded anterior surface (except for distal portion, which has a cutting edge) and a flattened knifelike posterior surface. The tips of the fangs are slightly compressed. The skull from a prepared skeleton (MCZ 180332, female) has 15-15 palatine teeth, 34-35 pterygoid teeth, and 34-34 dentary teeth. Hemipenis (Fig. 33). Deeply bilobed (nearly half the length of the organ), non- ^ One male, MCZ 11605, has unusually low sub- caudal (127) and maxillary tooth (17) counts (next highest values 140 and 20, respectively). This possibly represents normal variation within sexlineatus but needs clarification with a thorough study of geo- graphic variation in this taxon. MCZ 11605 also has an unusual coloration (see "Remarks") but, unfor- tunately, lacks precise locality data. 410 Bulletin Museum of Comparative Zoology, Vol. 154, No. 5 capitate, acalyculate (entirely spinose), with a deeply bifurcate centrolineal sulcus spermaticus that stops short of the tips of the lobes. Distal sulcate tips of the lobes with 8-10 papillae, each surmounted by a spine. Coloration in Life (RNP Region), Based on MCZ 180326-35 (see Glaw and Vences, 1994:pl. 348). Middorsal area five scales wide dark olive brown or medium brown, three rows either side of this olive brown; rows 1-3 grayish brown, but with variable amount of black (generally much black on 2-3, especially anteriorly, with black in- creasing on 1-3 posteriorly). Edges of scale row 3 and 7-9 often with striking white border (very thin); this more prominent on anterior body, and on dorsal edge of row 3. Extreme outer edge of ventrals black, forming black stripe in conjunction with black pigment on dorsal row 1. Remainder of venter with ground color of dull cream, but often heavily invested with black, es- pecially toward posterior end of body. Top of head dark olive brown. Postorbital bar black. Upper and lower labials, and throat dull whitish. The pattern of Liopholidophis sexlinea- tus in the RNP area is similar to the color plate in Glaw and Vences (1994:pl. 348), but the colors are more subdued: median dorsal area darker brown, and dorsolateral ground color dull olive brown rather than yellowish brown. Coloration in Preservative. Although the specific epithet refers to six stripes, most specimens I examined have only four dis- tinct stripes, and general darkening of the flanks may obscure the lateral and ventro- lateral stripes entirely so that the snake appears to be a brown snake with black flanks. Giinther (1882:265) indicated sim- ilar variation, stating that sexlineatus has "six black longitudinal bands, of which, . . . two or more may be indistinct or dis- appear altogether." The entire range of variation is seen within the sample from the RNP, and I detected no geographic trend. When the complete complement of six stripes is present, they are disposed as fol- lows: (1) a black border on the suture be- tween the ventral plates and dorsal row 1; (2) a facial stripe beginning on the upper edge of the supralabial row, widening as a postocular stripe, continuing onto body, where it usually occupies the lower % of row 3 + upper Vs of row 2; occasionally involving lower portion of row 4; (3) a stripe, usually indistinct and often absent, on the suture between dorsal rows 6 and 7. In some specimens the entire lower 3 dorsal rows are blackened, or blackened with only a central spot of light pigment in each scale, with the black extending a variable distance onto the ventral plates (Fig. 21). The median 5 dorsal rows are darker brown than more lateral rows (1- 6), which are grayish brown. The venter is whitish, but with a highly variable in- vestment of black: most specimens from the RNP have at least the lateral edges of the ventrals black, but often black is the predominant ventral coloration. Addition- al comments are given in the "Remarks." Natural History. Liopholidophis sex- lineatus is diurnal and semiaquatic. The species is abundant in rice fields, especially those somewhat overgrown around the edges of paddies, and with a covering of Azolla or duckweed. It seems most char- acteristic of sluggish or standing water, but the species was abundant in tall (0.5 m) grass along one whitewater river with rocky substrate next to rice fields, and two snakes in the same area were in water at the edge of the river. I never observed L. sexlinea- tus in primary or secondary forests, in- cluding aquatic habitats therein (small for- est streams and pools, larger rivers). Lio- pholidophis sexlineatus was observed in apparently natural habitats only near Sa- havondrona, within the RNP. Here, the species was associated with meadows, bogs, and marshes, which, during the rainy sea- son, fill with standing water to depths of up to 0.5 m. The meadows occupied de- pressions of varying sizes surrounded by higher ground supporting forest and are possibly maintained as meadows by sea- sonal flooding during part of the year. The meadows near Sahavondrona were filled LlOPHOLIDOPHIS (COLI'BRIDAE) FROM MADAGASCAR • Cadlc 411 with a grass/sedge association and were breeding and/or retreat sites for species of Heterixalus spp. (Hyperoliidae), Ptychad- ena nmscareniensis (Ranidae), and Boo- phis spp. and Aglyptodactylus madagas- cariensis (Rhacophoridae). Liopholido- phis sexUneatus is inoffensive and does not bite in defense. Glaw and Vences (1994:338), citing per- sonal communication from C. Domergue, reported L. sexlineatus as ovoviviparous (=viviparous fide terminology of Black- burn, 1994), but the basis for the inference was not stated. Given the long egg reten- tion times of many oviparous squamates (Shine, 1983), and the rather stringent cri- teria that must be met to be assured of correct inference of reproductive mode (Blackburn, 1993), the report of viviparity in sexlineatus needs confirmation. I can offer only partial corroboration. Five fe- males in the RNP sample collected 8 De- cember were gravid, with 4-10 embryos (determined bv palpation and inspection), as follows: MCZ 180325 (SVL 400 mm; 7 embryos), MCZ 180329 (SVL 452 mm; 10), MCZ 180330 (SVL 333 mm; 4); MCZ 180331 (SVL 408 mm; 7), and MCZ 180334 (SVL 475 mm; 8). In all cases, the devel- oping embryos were surrounded by fetal membranes, but without thickened shell membranes or leathery shell. Two embry- os removed from MCZ 180334 were in Zehr (1962) stage 25-26; one removed from MCZ 180329 was approximately stage 24. Because all gravid females were collected at the same time, and none showed any apparently more advanced embryos upon casual inspection, no other embryos were examined directly. Embryos of other spe- cies of Liopholidophis of comparable stages of development (see species accounts) are invariably surrounded by leathery shells; absence of such a shell in L. sexlineatus is taken to confirm the presence of viviparity in this species (but see cautionary notes in Blackburn, 1993). All dietary items for Liopholidophis sexlineatus were frogs. I was drawn to a specimen in a marsh by the loud release calls of a Ptychadena mascareniensis. Figure 21. Liopholidophis sexlineatus (GiJnther). Top: Spec- imen from the RNP (MCZ 180331). Bottom: Specimen from near Midongy du Sud (MCZ 1 80379). Note the darkening along the suture lines of dorsal rows 1-2 in the former and the com- plete darkening of the flanks in the specimen from Midongy du Sud, rendering the lateral stripe indiscrete except on the an- terior part of the body. which the snake was eating at 1350 hr on 22 November 1990. The snake was in a relatively open boggy area under a small clump of vegetation; much of the marsh had tall (80-100 cm), dense grass. MCZ 180376 (SVL 439 mm) contained 14-16 Heterixalus cf. madagascariensis. MCZ 180338 (SVL 472 mm) contained uniden- tifiable remains of a small frog. Four spec- imens from Ambatolahy near the RNP col- 412 Bulletin Museum of Comparative Zoology, Vol. 154, No. 5 lected 8 December 1990 and held in a common collecting bag regurgitated frogs (Ptychadena mascareniensis, Heterixalus betsileo, and H. alboguttatus). The frogs recorded in the diet of Lio- pholidophis sexlineatus, Heterixalus and Ptychadena, are the most commonly en- countered frogs in the marshes and rice fields that are the major habitats of sexli- neatus. Remarks. Giinther (1882) described fe- males and males of Liopholidophis sexli- neatus (as Dromicus sexlineatus and D. macrocercus, respectively) in the same pa- per, failing to realize that the extraordi- nary differences in tail length and subcau- dal counts manifested sexual dimorphism (of a nature hitherto unknown in snakes). Curiously, he thought he had males and females within the series of syntypes he described as sexlineatus (p. 265, comment referring to dimorphism in ventral color). The error was caught by Boulenger (1893: 246), who correctly identified types of sex- lineatus as females and those of macro- cercus as males. Boulenger (1893) synon- ymized the two species and, as first revisor, fixed the name of the taxon as sexlineatus (International Commission on Zoological Nomenclature, 1985:article 24[a]). The type locality, "eastern Betsileo" re- fers to territory on the eastern edge of the plateau and the adjacent escarpment be- tween approximately parallels 21° and 22°30'S (the Betsileo being one of the in- digenous peoples inhabiting this region; see, e.g., Gallieni, 1908:pl. 6). Thus, the syn- types of sexlineatus come from the general region of the RNP, but probably from the adjacent plateau rather than from the es- carpment itself. Aside from the variable distinction of the dorsal stripes (see Coloration), the amount and distribution of black pigment in Liopholidophis sexlineatus varies con- siderably. Unfortunately, samples have been insufficient to fully characterize the variation (possibly geographic). Future studies should comprehensively survey the taxon throughout its range to discern whether or not more than one taxon is involved. The following comments high- light patterns I discerned. Specimens from Toamasina Province (MCZ 11602-06) have little black pigment on the venter or on dorsal rows 1-2, except for the upper portion of row 2 involved in the lateral stripe and some darkening or spotting along the suture lines between ventral plates. In MCZ 11605 the lateral and ventrolateral stripes are indistinct (re- stricted to suture lines), and the snake is nearly plain brown. Specimens from far- ther south (RNP sample and Midongy du Sud) have varying degrees of black, some- times extensive, on the venter and rows 1- 2. In some RNP specimens, rows 1-2 are blackened so that only a central light spot of each scale remains (Fig. 21), whereas lateral stripes are rather distinct in most specimens from the RNP. In adults from Midongy du Sud (and one specimen from the RNP region, MCZ 180338), rows 1-3 (i.e., including the lateral stripe) and the adjacent venter are entirely blackened (Fig. 21); one near-hatchling (MCZ 180378; SVL 180 mm) from Midongy du Sud shows no general darkening of rows 1-3, suggesting that the black flanks in adults develop on- togenetically. The extent of ventral pigmentation is highly variable within a locality. For ex- ample, in MCZ 180376 (Midongy du Sud) only the outer 12-15% of each ventral plate is black, with the rest of the venter im- maculate whitish, whereas in MCZ 180379 from the same locality most of the ventral plates are obscured by black. Two speci- mens from a relatively high elevation in the RNP region (MCZ 180336-37, 1130 m) have relatively immaculate venters, whereas specimens from lower elevations in the same area (e.g., MCZ 180325-35, 850 m) have heavily pigmented venters. Giinther (1882:265) commented that the venter was darker in females than in males, but that trend does not hold in the series from the Ranomafana region when other sources of variation are considered. The Malagasy names mandodrano and anakanify are used for Liopholidophis sexlineatus in the RNP area. LlOPHOLIDOPHIS (COLUBRIDAE) FROM MADAGASCAR • Ccdle 413 Liopholidophis pinguis Parker Figures 12, 22 Liopholidophis pinguis Parker, 1925:390 (Type lo- cality: "Antsihanaka"). Holotype, BMNH 1946.1.7.66 (formerly 1925.8.25.7), 'an adult male [not seen], Werner, 1929:11; Guibe, 1958:216; Do- mergue, 1973:1397; Brygoo, 1983:55, 1987:24; UICN/PNUE/WWF, 1990:223; Glaw and Vences, 1992:266, 1994:338. Liopholidophis pinguis Parker (1925) is not known to occur in the RNP. I have no field experience with the species and know nothing of its natural history. Holotype. BMNH 1946.1.7.66 (not seen), an adult male obtained by W. F. H. Rosenberg; 890 mm total length, 300 mm tail length, with 151 ventrals and 91 sub- C3iuda\s fide Parker (1925). Diagnosis. A species of Liopholidophis having 17-17-15 dorsal scale rows, but lacking sexual dimorphism in tail length as extreme as in other members of the sex- lineatus group. The number of midbody scale rows distinguishes pinguis (17) from members of the stumpffi group (19). The short tail (<35% total length) and corre- sponding low numbers of subcaudals (<110) in males distinguish pinguis from males of other members of the sexlineatus group (tail >35% total length and >120 subcaudals in males). Liopholidophis pin- guis is most easily confused with L. sex- lineatus (see "Key to Species"), and char- acters reliably separating females of the two species are subtle. The relative dis- tinctness of the lateral stripes seems to be the most reliable feature (see "Key to Spe- cies" and species account for sexlineatus). Other species of the sexlineatus group have higher numbers of subcaudals (Table 1) and are either striped with distinctive nape spots (rhadinaea) or have distinctively pat- terned venters {dolicocercus and grandi- dieri) (see species accounts). Distribution. Known from the vicinity of the type locality, Antsihanaka, and the nearby Lake Alaotra, and from the Perinet (=Andasibe) reserve (Appendix; Glaw and Vences, 1994:336 [map], 472); all are in the eastern forest region (Fig. 3). The locality for one specimen (SMF 61909) is recorded as "Nord-Madagascar" (northern Mada- gascar), and the UICN/PNUE/WWF (1990) records "Moramanga ' [18°56'S, 48°12'E] without documentation. See "Re- marks." Description. The following description is based on examinaton of 6 females and 5 males but incorporates data for the ho- lotype (Parker, 1925). Measurements, pro- portions, and scutellation are summarized in Table 1. Largest specimen the male ho- lotype (BMNH 1946.1.7.66), 890 mm total length, 300 mm tail length (34% of total; Parker, 1925); largest female (BMNH 1936.3.3.94-97, largest of two females in the series), 664-1- mm total length, incom- plete tail 86-1- mm. Proportional tail length moderately sexually dimorphic, 30-34% of total length in males, 25-26% in females. Dorsal scales smooth, lacking apical pits, in 17-17-15 rows. Scale row reduction from 17 to 15 rows by loss of row 4 (N = 6 sides), fusion of 3 + 4 (N = 4 sides), or fusion of 4 -I- 5 (N = 4 sides) at the level of ventrals 87-102. Ventrals 147-154 in males, 139-147 in females. Anal plate di- vided. Subcaudals 88-99 in males, 67-71 in females. Eight upper labials with 4-5 touching eye (unilateral presence of 9 in one specimen). Lower labials 10-10 (N == 5) or 9-10 (N = 5), the first pair in contact behind the mental, 1-4 or 1-5 touching an anterior genial, 4-5 or 5-6 touching a pos- terior genial. Anterior genials shorter than posterior genials. Loreal present. Preocu- lar usually single (unilaterally divided in two specimens). Postoculars 2. Temporals 1 + 2. Body about as high as, or slightly higher than, wide; ventrolateral edge of body slighly angulate. Head slightly wider than neck. Pupil round. Eye very small, its di- ameter less than the distance between the eye and posterior edge of nostril (x = 0.88 ± 0.06; range 0.77-0.94; N = 5). Scattered pits on head plates. Dentition. Maxillary teeth 20-24 + 2 (N = 9; X = 21.7 ± 1.3 prefang teeth). Diastema absent; gap <1 tooth width sep- arating tooth row from enlarged fangs. Ungrooved fangs not offset from tooth row, 414 Bulletin Museum of Comparative Zoology, Vol. 154, No. 5 twice as large as the posteriormost max- illary teeth; having a rounded anterior sur- face (except for distal portion, which has a cutting edge) and a flattened knifelike posterior surface. The tips of the fangs are slightly compressed. A skull (MCZ 11701, male) has 13-13 palatine teeth, 28-30 pterygoid teeth, and 29-28 dentary teeth. Hemipenis. Deeply bilobed (somewhat less than half total length), noncapitate, acalyculate (entirely spinose), with a deep- ly bifurcate centrolineal sulcus spermati- cus. Coloration in Preservative (Based on USNM 149242 and SMF 61909; AMNH 60692 is similar; see Fig. 22). These are the most recently collected and best-pre- served specimens I have seen. No distinct stripes on most of the body, but thin dark- ened edges to many dorsal scales gives a somewhat braided appearance. Dorsal ground color grayish brown. Thin dark postocular bar extending from extreme lower edge of lower postocular and upper edge of supralabial 5, across lower edge of anterior temporal/upper edge of labials 6- 7, ending on anterior portion of labial 8. Except for thin upper blackened border, supralabials dirty white finely peppered with dark. Minute tubercles and pits on anterior head plates. Blackened suture line between ventrals and dorsal row 1, broad- ening on posterior body and tail to form a distinct stripe at subcaudal/dorsal caudal suture that continues to the tail tip. Similar blackened border between rows 2 and 3 on posterior V2 of body (ending at vent or on anterior part of tail), forming a distinct stripe on posterior 30% of body in SMF 61909. Scale row 3 of SMF 61909 high- lighted with white dots on anterolateral portion of each scale (more evident ante- rior to lateral stripe and dots more consis- tently present on upper edge of scales); similar, but less distinct, dots present in USNM 149242. Venter grayish white with most ventrals (especially posteriorly) having thin black- ened anterior border. Subcaudals immac- ulate grayish white, except for the lateral blackened edge. A series of Liopholidophis pinguis from "Lake Alaotra" (BMNH 1936.3.3.94-97) and another series probably from close to there (MCZ 11698-701; see Appendix for comment) are similar to those just de- scribed but have more distinct lateral stripes. The stripe along the suture be- tween the ventrals and dorsal row 1, man- ifested by black pigment at the extreme lateral edges of the ventrals, is obvious pri- marily on the posterior body and on the tail. A lateral stripe is manifested by a se- ries of dashes or small dots on row 3 an- teriorly (pigment at anterior-posterior junction of adjacent scales), or on the su- ture between rows 2 and 3 posteriorly; an- teriorly, it is invariably a "dotted" line; posteriorly, it varies from bare shading of the suture line in a zigzag pattern to a distinct lateral stripe involving more of the adjacent scales. The lateral stripe either stops at the vent or on the anterior part of the tail or merges with the ventrolateral stripe; tail with a black stripe at lateral edges of the subcaudals continuous with the ventrolateral body stripe. In the MCZ series, dorsal row 3 on the anterior ¥2-% of the body is highlighted by a pair of white dashes on the anterolateral portions of each scale; row 7 is partially similarly highlighted in one specimen. The venter is either immaculate, has obscure irregular grayish markings, or has suture lines be- tween adjacent ventrals outlined indis- tinctly in black. Two specimens have a series of irregular dashes laterally on each ventral (distinct only anteriorly in one of the two). The supralabials are largely im- maculate; a dark postocular bar extends across the top of the last 3 supralabials from the ventroposterior edge of the eye. The gular region and infralabials are im- maculate. Parker (1925) reported the type as hav- ing a distinct black lateral stripe from the eye to the vent on scale row 3 (2 + 3 posteriorly), black spots on the outer ends of the ventrals and subcaudals, and a series of indistinct black dots on either side of the midventral line. Remarks. The type locality, Antsihan- LioPHOLiDOPHis (Coluhridae) FROM MADAGASCAR • Cadle 415 WM^' Figure 22. Liopholidophis pinguis Parker. Specimen (USNM 149242) from Perinet [=Andasibe]. See also Figure 12. aka, is the name for a region in the vicinity of Lake Alaotra, a large freshwater lake at the edge of the central plateau. The Si- hanaka are one of Madagascar's indige- nous peoples inhabiting the area around the lake; "Antsihanaka" literally means "land of the Sihanakas." See, for example, the maps of indigenous peoples in Gallieni (1908:pl. 6) and Grandidier (1893). Sihan- aka can also mean simply "lake," from which the name of the people and the region may derive. The locality was dis- cussed by Carleton and Schmidt (1990:9) as "Sihanaka Forest." Parker (1925:390) stated that the Antsihanaka country was "situated between Lake Alaotra and the first belts of the eastern forest." A series of pinguis in the MCZ from the "eastern Forest" was heretofore identified as sexlineatus (MCZ 11698-701 collected by Frederick R. Wulsin, June to Septem- ber, 1915). According to Barbour (1918: 479), the portion of Wulsin's collection la- beled as coming from the "Eastern Forest" was collected "at a point about half way between Tamatave and Tananarive" (=Toamasina and Antananarivo, respec- tively). Unfortunately, the data are no more precise. Wulsin collected at Andaingo Figure 23. Heads of species of the Liopholidophis stumpffi group, in dorsolateral view. A. L. epistibes, new species (MCZ 180324). B. L lateralis (Dumeril, Bibron, and Dumeril) (MCZ 180349). C. L. infrasignatus (Gunther) (MCZ 180359). See Fig- ures 8 and 25 for L. sfumpffi (Boettger) and also Figure 28 for L. infrasignatus (Gunther). (18°12'S, 48°17'E; Barbour, 1918:478), just south of Lake Alaotra, from where most specimens of pinguis, including the type, seem to have come. Wulsin's specimens of pinguis could be from this region, whose location is consistent with Barbour's more vague description of the locality. The stumpffi Species Group (Parker, 1925) Figures 23-29 (see also Figs. 7-11); Table 2 Content. Dromicus stumpffi Boettger, 1881a:358, 1881b:441. Leptophis lateralis Dumeril, Bibron, and Dumeril, 1854:544. Ptyas infrasignatus Gunther, 1882:263. (Senior syn- onym of Liopholidophis thieli Domergue, 1973, of recent authors, as shown later). Liopholidophis epistibes, described herein. 416 BiiUetin Museum of Comparative Zoologtj, Vol. 154, No. 5 The stump ffi group includes two broad- ly distributed species, infrasignatus and lateralis, and I have not undertaken a study of their geographic variation. A more thor- ough investigation may show these to be composites. Liophotidophis stumpffi (Boettger) Figures 8, 24-25 Leptophis lateralis Dumeril, Bibron, and Dumeril, 1854 (part): (Type locality, Madagascar). Syntypes, MNHN 7312 (1 <5, 3 2) fide Guibe (1958:214) [not seen]. Gunther, 1890:70. Boulenger, 1893:247. Mocquard, 1904:302,'" Thamnosophis lateralis Jan and Sordelli, 1879: Bou- lenger, 1893:247. Dromicus stumpffi Boettger, 1881a:358, 1881b:441, pi. 1, fig. 2 (Type locality, Nossi-Be). Syntypes: three specimens collected by Antonio Stumpff and originally in the Senckenberg Museum; presumably three of four adults listed under catalog number 7247a by Boettger (1898:25). SMF 17576 is here designated the lectotype; see remarks. Ptyas infrasignatus Gunther, 1882: Gunther, 1890: 70 (synonym of Dromicus stumpffi Boettger). Bou- lenger, 1893:247 (synonym of Tropidonotus stumpffi). Ptyas infrasignatus is here recognized as a valid senior synonym of Liopholidophis thieli Domergue, 1973 (see later). Dromicus baroni Boulenger, 1888:104: (Type local- ity, Madagascar). Holotype, BMNH 1946.1.7.67 (old number 87.12.22.38) [examined], Gunther, 1890:70 (synonym of Dromicus stumpffi Boettger). Boulen- ger, 1893:247 (synonym of Tropidonotus stumpffi). Here recognized as a synonym of the resurrected Ptyas infrasignatus Gunther. Tropidonotus stumpffi (Boettger): Boulenger, 1893: 247. Boettger, 1898:25, 1913:312; Mocquard, 1895a: 102, 1895b {Tropidonotus stumpfei); Jourdran, 1903:32 (T. stumpfii). Boulenger, 1915:373-374, Kaudern, 1922:445 (cited specimen probably = L. epistibes, see species account). '" Liopholidophis lateralis has, since Boulenger (1893), been assigned to the synonymy, in part, of stumpffi Boettger, But the general confusion of stumpffi, epistibes, and infrasignatus in the literature suggests a reevaluation, Guibe (1954:242) gave 166 as the ventral count for the male syntype of lateralis and stated that it has "a median black spot on each ventral" (Guibe, 1958:214). Both statements conform more to epistibes than to other members of the stumpffi group (see Table 2 and other species ac- counts). Nonetheless, the syntypes of lateralis must be reexamined to correctly place the synonymy. Liophidium gracile Mocquard, 1908:261: (Type lo- cality, Montague d'Ambre and Nossi-Be), Syntypes, MNHN 1893.211, an adult male collected May- July, 1893 by Alluaud and Belly at Montagne d'Ambre (Mocquard, 1895:123) [examined]; and MNHN 84-595, a juvenile, probably female, col- lected at Nossi-Be [examined], Boulenger, 1915:374 (questionably listed as synonym of Tropidonotus stump ffii). Both of the syntypes of Liophidium gracile are here recognized as the same taxon as Dromicus stumpffi Boettger, Liopholidophis lateralis (Dumeril, Bibron, and Du- meril) (part): Mocquard, 1909:89; Werner, 1929: 11; Guibe, 1954:243, 1958:213, {Dromicus stumpffi Boettger listed as synonym), Liopholidcrphis stumpffi (Boettger): Parker, 1925:391, Domergue, 1973:1401; Nicoll and Langrand, 1989: 44, 72, 130; Brygoo, 1983:55, 1987:24; UICN/ PNUE/WWF, 1990:223; Glaw and Vences, 1992: 226, 1994:338, As noted in the description of L. epistibes herein, most of Domergue s (1973) spec- imens of "stumpffi" from eastern Madagascar (fol- lowed by subsequent authors) probably are epis- tibes. Liopholidophis infrasignatus (Gunther): Parker, 1925: 391 (synonym of [Droiyncus] stumpffi Boettger), Notes on Types and Designation of Lectotype. Boettger (1881a,b) described Liopholidophis stumpffi from three spec- imens collected by Antonio Stumpff, con- sul to Madagascar, on the island of Nosy- Be (the former paper is a brief description in Latin; the latter paper repeats verbatim the Latin description, followed by a de- tailed description in German). Boettger (1881b) gave detailed measurements and scale counts for the three syntypes. I ex- amined eight specimens collected by Stumpff at the type locality (BMNH 1946.1.23.51, FMNH 18291, SMF 17576 [listed as "typus" in SMF records], SMF 17580-84). With the exception of SMF 17576, my scale counts and measurements do not correspond well with the details given by Boettger (1881b), and SMF 17576 (Figs. 24-25) is hereby designated the lec- totype of Dromicus stumpffi Boettger. SMF 17576 apparently is specimen "No. 1" in Boettger (1881b). Details on this specimen are as follows (Boettger's data in parentheses): A gravid adult female, total length 711 mm (750), tail length 236 mm (237), tail as a proportion of total length 33%; 2 preventrals + 151 ventrals (153), LlOPHOLIDOPHIS (CoLUBRIDAE) FROM MADAGASCAR • Cadlc 417 Table 2. Variation in mensural and meristk; characteristics of species of the Liophoiadophis STVMPFFI GROUP. SCALE COUNTS AND BODY PROPORTIONS ARE X ± SD (SAMPLE SIZE) WITH RANGES BELOW IN PARENTHESES; MAXILLARY TOOTH COUNTS ARE PREFANG COUNT RANGES ( + 2 FANGS), FOLLOWED BY X ± SD (SAMPLE size). Tabulations for infrasignatus include data on ventral and subcaudal counts AND relative TAIL PROPORTIONS OF L. THIELI FROM TABLE II OF DOMERGUE (1973:1405). stinnpffi epistibes' lateralis injrasignalus Dorsals 19-19-17 19-19-17 19-19-17 19-19-17 Ventrals Males Females 150.7 ± 1.37 (6) (149-153) 149.4 ± 4.02 (7) (145-157) 162.3 ± 3.55 (7) (157-166) 160.6 ± 4.63 (17) (151-167) 154.0 ± 4.83 (28) 144-1652 159.5 ± 4.86 (20) 151-166 149.0 ± 3.18(17) (141-156) 152,7 ± 3.86 (35) (144-161) Subcaudals Males Females 97.0 ± 6.00 (6) (89-104) 98.3 ± 5.47 (7) (91-109) 96.7 ± 4.35 (7) (91-104) 88.8 ± 3.53 (16) (83-96) 89.5 ± 6.26 (22) (80-98) 87.9 ± 6.31 (15) (76-97) 73.4 ± 4.27(15) (66-81) 67.7 ± 2,64 (30) (62-73) Maximum length (mm) Total (SVL) Males Females 627 (416) 711(475) 709 (471) 829+ (634) 729 (517) 820 (586) 727+ (606) 920 (712) Tail length/total Males Females 0.32 ± 0.02 (6) 0.29-0.34 0.33 ± 0.01 (7) 0.31-0.34 0.31 ± 0.02 (7) 0.28-0.34 0.29 ± 0.01 (15) 0.27-0.31 0.29 ± 0.01 (22) 0.27-0.31 0.27 ± 0.01 (15) 0.25-0.29 0.25 ± 0.01 (16) 0.23-0.27 0.23 ± 0.01 (29) 0.21-0.24 Maxillary teeth 25 + 2—31 + 2 27.8 + 2 (6) 22 + 2—29 + 2 26.4 + 2 (16) 25 + 2—30 + 2 26.6 + 2 (18) 20 + 2—25 + 2 22.3 + 2 (28) Eye diameter/eye-nostril distance 1.22 ± 0.07 1.13-1.31 (7) 1.38 ± 0.17 1.16-1.69(12) 1.1 ± 0.09 0.97-1.3 (24) 1.21 ± 0,10 1,06-1.44 (19) ^ As noted in the text, most specimens referred to L. stump ffi by Domergue (1973:table I) probably are epistibes. Extreme values for meristic and mensural statistics represented by Domergue 's specimens, where different from those reported here, are ventrals: S, 151; subcaudals: 6, 105; 9, 80; total length: 3, 798 (548); 2, 945 (675); tail length/total: 2, 0.32. In these I have excluded Domergue 's specimen from Marojezy, which may represent true stumpffi (see text, "Distribution" in the epistibes species account). 2 Domergue (1972:table III) reported the following ranges for meristic counts for lateralis, ventrals 154- 170 ((3), 152-174 (2); subcaudals 86-99 (6), 58-98 (2). I suspect low values for the female subcaudal counts reflect incomplete tails. As Domergue did not give values for individual specimens, these data were not included in calculations provided here. divided anal plate, subcaudals 96 (98), 1 preocular, 2 postoculars, 2 + 2 temporals; 8-8 supralabials, 4-5 in contact with eye; 10-10 infralabials. Dorsals in 19-19-17 rows, the posterior reduction occurring by fusion of rows 3 + 4 at the level of ventrals 92-90. Horizontal eye diameter 4.0 mm. Anterior edge of eye to posterior edge of nostril 3.1 mm. Diagnosis. Liopholidophis stumpffi dii- fers from members of the sexlineatus group in having 19-19-17 dorsal scale rows (vs. 17-17-15). It differs from other species of the stumpffi group primarily in color pat- tern and a few body proportions, including the following: relatively long tail and high number of subcaudals (31-34% of total length and 91-109, respectively, sexes combined); dorsolateral light stripe on rows 4-5 on neck and anterior part of body, row 4 or 4-5 when present posteriorly; dark postocular stripe separated from dark 418 Bulletin Museum of Comparative Zoology, Vol. 154, No. 5 blotches on side of neck; dorsolateral light stripe continuous with light color of throat; venter mostly immaculate except for pig- ment encroaching laterally from flanks (small spots may be present on extreme anterolateral edge of anterior ventral plates, but these are not inset from edge of the plates). Liopholidophis stumpffi is most easily confused with epistibes, and their distin- guishing characteristics are given in the account for the latter. Liopholidophis stumpffi differs from L. lateralis in the position of the lateral stripes: in stumpffi on dorsal rows 4-5 on neck and anterior body, usually fading posteriorly (indistinct on tail); in lateralis on rows 3- 5 (occasionally only row 4), very distinct the length of the body, continuing to the tail tip. The species also differ in color pattern: indiscrete dark spots on neck and anterior body, and brownish posterior body with light stripes indistinct or absent in stumpffi; continuous dark middorsal stripe and flanks, separated by vivid light stripes the length of the body in lateralis. Liopholidophis stumpffi differs from infrasignatus in the orientation of the postocular dark bar. In stumpffi the bar extends horizontally posterior to the eye, paralleling the upper border of the pos- terior supralabials (Figs. 8, 25); in infra- signatus the bar extends at an angle down- ward across the penultimate and ultimate supralabials (Figs. 23, 28). In stumpffi the dorsolateral light stripe anterior is on scale rows 4-5 (5-6 in infrasignatus). Liophol- idophis stumpffi also has a longer tail than infrasignatus (31-34% of total length vs. 21-27%, sexes combined; see Table 2), is of more gracile habitus, and has more dis- tinct spots on the neck (present or not in infrasignatus, but not conspicuous). Distribution. Liopholidophis stumpffi is here considered a species of extreme northern Madagascar. Most specimens ex- amined are from the island of Nosy-be, the type locality; three specimens (includ- ing one syntype of Liophidium gracile Mocquard) are from near the northern tip of mainland Madagascar in the vicinity of Montagne d'Ambre (Fig. 6 and Appendix). As pointed out in the account for Lio- pholidophis epistibes, most literature re- cords of "stumpffi'^ from eastern Mada- gascar probably represent epistibes, and the distributional relationships between the two species in northern Madagascar are unclear. Description. The following description is based on examination of seven females and six males, including the lectotype and other topotypical material (see preceding comments) of Dromicus stumpffi Boettger and the two syntypes of Liophidium grac- ile (Mocquard). Measurements, propor- tions, and scutellation are summarized in Table 2. Largest specimen a female, 711 mm total length, tail 236 mm; largest male 627 mm total length, 211 mm tail length. Tail length not sexually dimorphic, 32- 34% of total length in males, 31-33% in females. Dorsal scales smooth, in 19-19- 17 rows; 0-2 apical pits on different scales within an individual. Five of seven spec- imens showed posterior scale reduction by fusion of rows 3 + 4 at the level of ventrals 84-95; two of six specimens from the type locality for which this character was de- termined had fusion of rows 4 -H 5 at ven- trals 92-93." Ventrals 149-153 in males, 145-157 in females. Anal plate divided. Subcaudals 96-104 in males, 91-109 in fe- males. Eight upper labials (rarely seven or nine) with 4-5 touching eye. Lower labials usually 10-10 (eight specimens), with 9- 10 (1) and 10-11 (1) being uncommon var- iants; first pair in contact behind the men- tal, 1-5 touching an anterior genial, 5-6 touching a posterior genial. Anterior ge- nials shorter than posterior genials. Loreal present. Preocular single. Temporals 2 + 2. Body slightly higher than wide; ventro- lateral edge of body slightly angulate. Head " The type locality is an island, and the high fre- quency of an "unusual" scale reduction pattern (fu- sion of 3 + 4 seems to be the common mode of reduction in Liopholidophis ) could reflect the isolated nature of the population. No other scale anomalies were detected in these specimens. LlOPHOLinOPHIS (COLl'BRIDAF.) FROM MADAGASCAR • Cadlc 419 Figure 24. Liopholidophls stumpffi (Boettger), lectotype (SMF 17576, female) from "Nossi-Be.' Dorsal and ventral views. slightly wider than neck. Pupil round. Eye range 1.15-1.31; N = 7). Scattered minute large (Figs. 8, 23, 25), its diameter greater pits on head plates, especially the supra- than the distance between eye and pos- oculars, prefrontals, and nasals, terior edge of nostril (x = 1.22 ± 0.07; Dentition. Maxillary teeth 25-31 + 2 420 Bulletin Museum of Comparative Zoology, Vol. 154, No. 5 (N = 6; X = 27.8 ± 2.23 prefang teeth). Diastema absent; gap <1 tooth width sep- arating tooth row from enlarged fangs. Ungrooved fangs not offset from tooth row, 2 times as large as the posteriormost max- illary teeth; having a rounded anterior sur- face (except for distal portion, which has a cutting edge) and a flattened knifelike posterior surface. The tips of the fangs are slightly compressed. Hemipenis. Very deeply bilobed (di- viding at the base of the organ and having an extremely short stalk), acalyculate (en- tirely spinose), and with a deeply divided centrolineal sulcus spermaticus. Tips of the lobes with a central "umbelliform" de- pression. Coloration in Life. Unknown. Coloration in Preservative. Most spec- imens I studied had lost most of the stra- tum corneum, and appear grayish with a black network the length of the body (formed by black borders to many scale rows), black irregular spots on the neck, and dorsolateral light stripes that vary in extent and discreteness. Two adult topo- types (FMNH 18291, SMF 17581) retain the stratum corneum. These are more or less brown snakes with an indistinct dark network on the dorsal scales, indistinct dark spots on the neck (generally 2-4 dorsal scales in size), and indistinct dorsolateral light stripes; top of the head brown to gray- ish brown; supralabials, infralabials, and gular region dirty whitish; blackish post- ocular bar; venter yellowish white, with dark encroaching pigment from flanks on lateral edges of ventral scales (plus dark punctations at lateral edges of anterior ventrals, as described in the diagnosis). The dorsolateral light stripes are anteriorly confluent with light color of the throat (Figs. 8, 25); they occupy rows 4-5 ante- riorly, usually fading by midbody but on rows 4 or 4-5 when present posteriorly. In several individuals, including juveniles and adults (e.g., MCZ 54368, MNHN 1893- 211), the light stripes continue to the tail tip and are bordered ventrally at the sub- caudal/dorsal caudal suture by a blackish streak. Three small juveniles (SMF 17582-84; total lengths 195-308 mm) are similar to adults in pattern, and the dorsolateral stripes also vary in discreteness and length, as in adults. Natural History. Liopholidophis stumpffi presumably is diurnal and ter- restrial like other members of the stumpffi group. The lectotype (SMF 17576; SVL 475 mm; month of collection unknown) is a gravid female with four large eggs, as de- termined by palpation. Remarks. The FMNH and the BMNH have specimens of Liopholidophis stumpffi collected by Stumpff on Nosy-Be and ex- changed with the Senckenberg Museum in the 1880s (FMNH 18291 and BMNH 1946.1.23.51, respectively). The FMNH records indicate their specimen as a "para- type" (see, e.g., Marx, 1958:480), an im- possible designation because Boettger's se- ries consisted of three syntypes. Boulenger (1893:247) noted the BMNH specimen "As typical of D. stumpffi"; such a designation would be unlikely if the BMNH specimen were really a syntype, because in such cases Boulenger routinely used the word "type." In any case, my measurements and scale counts for these specimens do not corre- spond to any of the three syntypes of stumpffi, as reported by Boettger (1881a). Boettger (1898:25) listed eight specimens (catalog number 7247a) from Nosy-Be col- lected by Stumpff in the Senckenberg Mu- seum at that time. In addition to the FMNH and BMNH specimens, the SMF now has several specimens collected by Stumpff on Nosy Be (Appendix). Boettger (1881a,b) stated that Liophol- idophis stumpffi has "two distinct apical pits." My observations revealed that the number of apical pits varies from 0 to 2 within an individual, even considering only those scale rows that occasionally had pits. Liopholidophis lateralis (Dumeril, Bibron, and Dumeril) Figures 23, 26 Leptophis lateralis Dumeril, Bibron, and Dumeril, 1854:544, part (Type locality, "Madagascar"). LioPHOLinopins (Colubridae) from Madagascar • Cadle 421 Figure 25. Liopholidophis stumpffi (BoeXtger), lectotype (SMF 17576, female). Dorsolateral view of head. Dromiciis melanotus, var. ? Giinther, 1858:133; Bou- lenger, 1893:248 (synonym of Tropidonotus later- alis). Thamnosophis lateralis (Dumeril, Bibron, and Du- meril): Jan, 1863:133. Jan and Sordelli, 1879:liv. 49, pi. II. Boulenger, 1893:248 (synonym of Tropidon- otus lateralis). Guibe, 1954:243, 1958:213 (syn- onym of Liopholidophis lateralis). Dromicus madagascariensis Giinther, 1872:22, pi. V, fig. A.: (Type locality, "Madagascar"). Syntypes, BMNH 1946.1.15.19 (female), collector unknown, and BMNH 71.6.28.17 (male), obtained by Mr. Bar- lett [both examined]. The latter specimen is here recognized as the male syntype upon which Giinth- er based his description. Boulenger, 1893:248 (syn- onym of Tropidonotus lateralis). Guibe, 1954:243, 1958:213 (synonym of Liopholidophis lateralis). Ahaetulla lateralis (Dumeril, Bibron, and Dumeril): Boettger, 1877:33. Boulenger, 1893:248 (synonym of Tropidonotus lateralis). Guibe, 1954:243, 1958: 213 (synonym of Liopholidophis lateralis). Philothamnus lateralis (Dumeril, Bibron, and Du- meril): Boettger, 1881b:526. Boulenger, 1893:248 (synonym of Tropidonotus lateralis). Guibe 1954: 243, 1958:213 (synonym of Liopholidophis later- alis). Dromicus stumpffi Boettger, 1881a:358, 1881b:441, pi. 1, fig. 2: Mocquard, 1904:302, 1909:89; Guibe, 1954:243, 1958:213. (synonym of Liopholidophis lateralis). Here considered a valid taxon. Ptyas irtfrasignatus Giinther, 1882: Guibe, 1954:243, 1958:213 (synonym of Liopholidophis lateralis). Here recognized as a valid taxon. Dromicus haroni Boulenger, 1888:104: Guibe, 1954: 243, 1958:213 (synonym of Liopholidophis later- alis). Here considered a synonym of the resurrected Liopholidophis infrasignatus (Giinther). Tropidonotus lateralis (Dumeril, Bibron, and Du- meril): Boulenger, 1893:248, 1915:374. Boettger, 1898:25, 1913:312. Jourdran, 1903:32. Kaudern, 1922:444. Figure 26. Liopholidophis lateralis (Dumeril, Bibron, and Dumeril). Specimen from the RNP, tvICZ 180353. 422 Bulletin Museum of Comparative Zoology, Vol. 154, No. 5 Liopholidophis lateralis (Dumeril, Bibron, and Du- meril): Mocquard, 1904:303, 1909:43; Parker, 1925: 391; Werner, 1929:11; Guibe, 1954:243, 1958:213; Domergue, 1973:1398; Brvgoo, 1983:55, 1987:24; NicoU and Langrand, 1989:48, 88; UICN/PNUE/ WWF, 1990:223; Glaw and Vences, 1992:266, 1994: 337. Liophidium gracile Mocquard, 1908:261: (Type lo- cality, Montagne d'Ambre and Nossi-Be). Syntypes, MNHN 1893.211 and MNHN 84-595 (see synon- ymy of Liopholidophis stumpffi Boettger for data). Guibe, 1958:213 (synonym of Liopholidophis la- teralis). Here considered a synonym of Dromicus stumpffi Boettger. Syntypes. MNHN 7312 (1 6, 3 2) fide Guibe (1958:214) [not seen]. Guibe (1954, 1958) gave some meristic counts and other descriptive data on the types. Diagnosis. Liopholidophis lateralis dif- fers from members of the sexlineatus group in having 19-19-17 dorsal scale rovi's (vs. 17-17-15). It is the only species of Lio- pholidophis with vivid dorsolateral light stripes (white to yellowish in life) centered on row 4 the entire length of the body and tail (adjacent parts of rows 3 and 5 usually also involved) (Fig. 26; Glaw and Vences, 1994:pl. 346). The dorsolateral light stripes are on rows 5-6 or 5-7 anteriorly in ep- istibes and infrasignatus and are indis- tinct posteriorly in stumpffi (see species account for other differences). Superficially, Liopholidophis lateralis is similar to Dromicodryas hernieri. These two species can be distinguished in life by the brown (vs. black) dorsal ground color of D. hernieri as compared to lateralis (compare Glaw and Vences, 1994:pls. 342, 346) and by the anterior disposition of the dorsolateral light stripe: confluent or near- ly so with the light gular coloration in L. lateralis, separated by dark flank colora- tion in D. hernieri (cf. Glaw and Vences, 1994:figs. 505-507). Dromicodryas has en- larged anterior mandibular teeth and dif- fers in fundamental hemipenial charac- teristics from Liopholidophis (Guibe, 1958; personal observations). Distribution. Liopholidophis lateralis, as presently understood, has an extraor- dinary geographic and macroenvironmen- tal range, including the eastern lowlands and montane rainforest belt, scattered lo- calities on the central plateau, and dry for- ests of western Madagascar (Fig. 6; for more comprehensive distribution maps, see Glaw and Vences, 1994:336; Domergue, 1973:1398). Domergue (1973:1401) re- corded localities from sea level to more than 2,000 m elevation. Such an ecological amplitude for a snake species is rare and warrants a thorough assessment of geo- graphic variation. In the vicinity of the RNP, Liopholidophis lateralis is known from approximately 500 to 900 m eleva- tion. Description. The following description is based on examinaton of 20 females and 28 males. Measurements, proportions, and scutellation are summarized in Table 2. Largest specimen a female (MCZ 11663), 820 mm total length, 234 mm tail length (29% of total); largest male (MCZ 180345) 729 mm total length, 212 mm tail length (29% of total; tip of tail missing). Tail length not strongly sexually dimorphic, 27-31% of total length in males, 25-29% in fe- males. Dorsal scales smooth, in 19-19-17 rows (one individual each with 21-19-17 and 17-19-17); usually two apical pits on scales of all rows between the dorsolateral stripes (see "Remarks"). Scale row reduc- tion from 19 to 17 rows by fusion of rows 3 4-4 (occasionally appears as loss of row 4) at the level of ventrals 85-105 (N = 11; two specimens with unilateral fusion of 4 + 5). Ventrals 144-165 in males, 151-166 in females. Anal plate divided. Subcaudals 80-97 in males, 76-97 in females. Eight upper labials with 4-5 touching eye. Low- er labials usually 10-10 (N = 19), other variants being 8-9 (2), 9-9 (5), 9-10 (12), and 10-11 (4), the first pair in contact be- hind the mental, 1-5 (rarely 1-6) touching an anterior genial, 5-6 (rarely 6-7) touch- ing a posterior genial. Anterior genials shorter than posterior genials. Loreal pres- ent. Preocular single. Temporals usually 2 -I- 2 (occasionally 1 anterior temporal or, less frequently, 1 posterior temporal; rath- er high frequency of azygous temporal LioPHOUDOPHis (Coiaibridae) FROM MADAGASCAR • Cadle 423 scales, fragmentation of scales in temporal region, or fusion of a temporal with a su- pralabial). Bod\ slightly higher than wide; ventro- lateral edge of body slightly angulate. Head distinctly wider than neck. Pupil round. Eye moderately large, its diameter equal to or slightly greater than the distance be- tween the eye and posterior edge of nostril (x = 1.1 ± 0.09; range 0.97-1.30; N = 24). Scattered pits and tubercles on circumor- bital and anterior head plates. Dentition. Maxillary teeth 25-30 + 2 (N = 17; X = 26.6 ±1.6 prefang teeth). Diastema absent; gap <1 tooth width sep- arating tooth row from enlarged fangs. Ungrooved fangs not offset from tooth row, twice as large as the posteriormost max- illary teeth; having a rounded anterior sur- face (except for distal portion, which has a cutting edge) and a flattened knifelike posterior surface. The tips of the fangs are slightlv compressed. Two skulls, MCZ 180350 and AMNH 60676 (both females) have the following tooth counts, respec- tively: 16-?, ?-21 palatine teeth; ?-35, 31+- 34 pterygoid teeth; 30-31, ?-34 dentary teeth. Domergue (1973) reported 13-15 max- illary teeth and 15-23 dentary teeth in Liopholidophis lateralis, about half the tooth number I counted (25-30 prefang maxillary teeth and 30+ dentary teeth); I assume that Domergue failed to count empty sockets. Hemipenis (Fig. 35). Deeply bilobed, noncapitate, acalyculate (entirely spinose), with a deeply bifurcate centrolineal sulcus spermaticus. Sulcus spermaticus centroli- neal, dividing near the base of the organ. Tips of the lobes with a central "umbel- liform" depression. Coloration in Life and Preservative. In life, Liopholidophis lateralis appears as a black snake with whitish to yellowish lateral stripes (see Glaw and Vences, 1994: pi. 306). The lateral light stripes usually occupy row 4 and adjacent halves of rows 3 and 5 (occasionally rows 4-5, and in MCZ 180345 essentially restricted to row 4 an- teriorly, 3-4 posteriorly); the stripes are continuous from the nape to the tip of the tail (uninterrupted at vent). Anteriorly, the stripes are usually confluent with the light (yellowish to whitish) color of the throat (occasionally separated by a narrow line of dark pigment; see Glaw and Vences, 1994:fig. 505). Dorsal rows below the lat- eral stripe are blackish, except for row 1, which tends to have only a stippling of blackish pigment (appears dirty white to grayish). The venter and underside of the tail are immaculate whitish to pale yellow, usually with outer edges of ventrals stip- pled with dark pigment and/or with small rounded black dots. In preservative, the light stripes are whitish and the dorsal ground color gray- ish black to brownish. Upon loss of the stratum corneum, the scales become gray- ish or grayish brown. Natural History. Liopholidophis later- alis is diurnal and terrestrial. It occurs in relatively open, often disturbed, areas (sec- ondary growth and rice fields). I have nev- er observed it in closed-canopy forest, ei- ther primary or moderately dense second- ary forest. Domergue (1973:1401) and Glaw and Vences (1994:337) reported L. lateralis as being semiaquatic, but other than occasional (and, in Madagascar, in- evitable) association with flooded rice fields, this species in my experience does not appear to be especially associated with water, certainly not to the extent of L. sexlineatus. Liopholidophis lateralis is abundant in appropriate open microhab- itats in the vicinity of the RNP and seems especially active on very hot days. These snakes often raise the head and anterior '/, of the body off the ground as an intruder approaches. They bite rather ineffectively (small teeth) when captured and often flatten the neck and body for about % of its length, exposing white skin between scales and broadening the body stripes. Domergue (1973:1401) also re- ported body inflation and neck flattening in Liopholidophis lateralis, exposing white markings on the scales. One individual 1 424 Bulletin Museum of Comparative Zoology, Vol. 154, No. 5 observed extended the tongue while slowly flicking it up and down, or held the tongue extended with little movement except at the tips for extended periods. Domergue (1973) reported frogs in the diet of Liopholidophis lateralis, and that is confirmed by all my observations. Three lateralis in the RNP sample contained food: MCZ 180344 (SVL 547 mm) contained one Mantidactylus hetsileanus (Ranidae; a terrestrial, diurnal frog) swallowed tail first; MCZ 180345 (SVL 517 mm) contained one Boophis madagascariensis (Rhacophori- dae; at least sometimes terrestrial when inactive diurnally) swallowed head first; MCZ 180350 (SVL 543 mm) contained re- mains of one Ptychadena mascareniensis (Ranidae; terrestrial/semiaquatic, diur- nal) swallowed head first. AMNH 60675 (SVL 537 mm) contained one Ptychadena mascareniensis swallowed head first and one other small unidentified frog swal- lowed tail first. In contrast to other species of terrestrial Liopholidophis with record- ed food items, lateralis seems to consume terrestrial microhylids infrequently (the semiaquatic L. sexlineatus is another ex- ception). This probably reflects the more open habitats frequented by lateralis and the absence of microhylids in those habi- tats. The frogs recorded in the diet are frequently encountered in open or sec- ondary habitats, as is L. lateralis. Domergue (1973) reported clutch sizes of 6-13 in Liopholidophis lateralis and ob- served several clutches at the end of No- vember/beginning of December (locality not given). Hence, the species is oviparous. Two specimens from the RNP, MCZ 180348 (SVL 583 mm) and 180344 (SVL 547 mm) collected 6-11 January, con- tained nine and seven enlarged ovarian eggs, respectively. MCZ 180375 (SVL 491 mm), collected 13 January near Midongy Atsimo (Appendix), contained seven en- larged ovarian eggs. Remarks. Most specimens of lateralis have two apical pits on all scale rows be- tween the dorsolateral light stripes. Oc- casional specimens appeared to have no apical pits (e.g., MCZ 180380), and in still others the number of pits and their con- sistency varied. When present, the pits continue onto the dorsal caudal scales to the tail tip. Liopholidophis infrasignatus (Gunther) Figures 23, 27-29 Ptyas infrasignatus Gunther, 1882: 263 (Type lo- cality, "Arkafana, Eastern Betsileo" [corrected to "Ankafana, Betsileo" by Boulenger, 1893:247; see "Remarks" and Cowan, 1883:147]). Lectotype by present designation, BMNH 1946.1.7.57, collected by Reverend W, D. Cowan. Dromicus baroni Boulenger, 1888:104 (Type locality, "Madagascar"). Holotype, BMNH 1946.1.7.67 (old number 87.12.22.38), collected by R. Baron, [ex- amined] New synonymy. Tropidonotus stiimpffii (Boettger), part: Gunther, 1890:70; Boulenger, 1893:247-248 {Ptyas infrasig- natus Gunther listed as a synonym; specimens b- d[=BMNH 1946.1.7.56-58, types oi infrasignatus] and k-m [=BMNH 95-10.29.53-55]). Liopholidophis lateralis (Dumeril, Bibron, and Du- meril), part: Mocquard, 1909:95 {Ptyas infrasig- natus Gunther listed as synonym). Guibe, 1958:243 {Ptyas infrasignata [sic] Gunther listed as a syn- onym). Liopholidophis stumpffi (Boettger), part: Parker, 1925: 391 (L. infrasignatus (Gunther) listed as a synonym in footnote). Liopholidophis thieli Domergue, 1973:1405 (Type locality, "fish ponds of the Perinet Tropical For- estry Station, 900 m elevation). Holotype, MNHN 1973-332. New synonymy. Brygoo, 1983:55, 1987: 24; Nicoll and Langrand, 1989:117, 130; UICN/ PNUE/WWF, 1990:223; Glaw and Vences, 1992: 266, 1994:338. Notes on Types and Designation of Lectotype. Ptyas infrasignatus Gunther is here considered the valid name for the species referred in recent literature to Lio- pholidophis thieli Domergue (references cited in synonymy). The type locality of infrasignatus is about 25 km ENE of the RNP, whereas the type of thieli is from Perinet (=Andasibe). Based on comparison of the types of infrasignatus and thieli, 1 conclude that they and the series from the RNP assigned to infrasignatus represent the same taxon. Additional comments on the type of thieli are given later (see "Notes on Type Specimens of Junior Synonyms"). The syntypes of Ptyas infrasignatus are BMNH 1946.1.7.56-58, two adult females and adult male, respectively [old numbers LioPHOLiDOPHis (Colubridae) FROM MADAGASCAR • Cadle 425 Figure 27. Liopholidophis infrasignatus (GiJnther), lectotype (BMNH 1946.1.7.57) from ■Ankafana, eastern Betsileo." Dorsal and ventral views. 426 Bulletin Museum of Comparative Zoology, Vol. 154, No. 5 82.2.25.59-64]), collected by Rev. W. D. Cowan. The largest of the two females, BMNH 1946.1.7.57, is here designated the lectotype (Figs. 27-28). Characteristics of these specimens are reported here in order of the series, 1946.1.7.56-58. These are the largest specimens of infrasignatus report- ed (Domergue, 1973, as "thielV') or stud- ied herein, with measurements (mm) and proportions as follows (total length, tail length, tail length as a percentage of total): two females (904, 215, 24%; 920, 208, 23%), male (727 + , 121 + ). Ventrals 155, 156, 152.5, in each case preceded by two prev- entrals. Subcaudals 71, 67, 46 + . Dorsal scales in 19-19-17 rows. One preocular, two postoculars, and 2-2 temporals. Eight supralabials (4-5 touching eye); 10-10 (fe- males) or 9-10 (male) infralabials. Divided anal. Maxillary teeth 21, 23, 23, followed by two enlarged, ungrooved fangs. General dorsal color light brownish with occasional scattered darker flecks (no- where dense). Indistinct light dorsolateral lines (most evident under fluid) on neck and anterior 20% of body. Thin dark post- ocular bar from posteroventral corner of eye, across penultimate supralabial, and ending on anteroventral corner of ultimate supralabial. Otherwise, supralabials light, immaculate (dorsal ground color en- croaches onto ultimate one). Infralabials and gulars immaculate yellowish white. Venter yellowish white with dense series of dark spots and markings, increasing pos- teriorly (similar to variation within the RNP sample), tending to form mid ventral series or line in females; dark encroach- ment of dorsal color onto lateral edge of ventrals, and an indistinct series of spots laterally on ventrals (except male). Diagnosis. Liopholidophis infrasig- natus differs from members of the sexli- neatus group in having 19-19-17 dorsal scale rows (vs. 17-17-15). It has a rela- tively short tail and low numbers of sub- caudals compared to other members of the stump ffi group (Table 2). The dorsolateral light stripes are anteriorly on rows 5-6, by which infrasignatus differs from lateralis (rows 3-5) and stump ffi (rows 4-5). Ep- istibes differs from infrasignatus in hav- ing a relatively long tail and higher ventral and subcaudal counts (Table 2 and epis- tibes account). Distribution. Scattered localities on the eastern escarpment and eastern edge of the high plateau, as shown on maps (for Lio- pholidophis "thieli") presented by Do- mergue (1973:1398) and Claw and Vences (1994:336). From at least the vicinity of Midongy Atsimo (23°35'S, 47°0TE) in the south (Appendix) to Antongil Bay (Nosy Mangabe) in the north (Domergue, 1973: 1409). Most localities appear to be upland sites, 600-1,200 m elevation. The Nosy Mangabe locality is <100 m (Domergue, 1973:1409), whereas the type locality for infrasignatus is possibly as high as 1,600 m (see "Remarks"). Liopholidophis infra- signatus appears to be widespread within the RNP, and turns up at most forested localities with sufficient sampling (known elevational range within the park approx- imately 800-1,150 m). Description. The following description is based on examinaton of 19 females and 11 males, including syntypes of Ptyas in- frasignatus Giinther and the holotypes of Dromicus baroni Boulenger and Liophol- idophis thieli Domergue; ranges of vari- ation for size, tail proportions, and ventral and subcaudal counts incorporate data for L. thieli given by Domergue (1973:table II). Measurements, proportions, and scu- tellation are summarized in Table 2. Larg- est specimen the female lectotype (BMNH 1946.1.7.57), 920 mm total length, 208 mm tail length (23% of total); largest male (BMNH 1946.17.58, a paralectotype), 727+ mm total length, incomplete tail 121+ mm. Proportional tail length not strongly sexually dimorphic, 23-27% of to- tal length in males, 21-24% in females. Dorsal scales smooth, in 19-19-17 rows; 0-2 apical pits on different scales within an individual. Scale row reduction from 19 to 17 rows by fusion of rows 3 + 4 (occasionally loss of row 4, and one in- stance of 4 + 5 fusion) at the level of ventrals 78-94 (N = 18). Ventrals 146-156 in males, 144-161 in females. Anal plate LioPHOUDOPHis (Colubridaf) from Madagascar • Cadle 427 divided. Subcaudals 66-81 in males, 62- 73 in females. Eight (rarely seven) upper labials with 4-5 touching eye. Lower la- bials usually 10-10 (N = 22), with other variants being 8-8 (1), 9-9 (1), 9-10 (4), and 10-11 (2), the first pair in contact be- hind the mental, 1-5 (occasionally 1-4) touching an anterior genial, 5-6 (occasion- ally 4-5) touching a posterior genial. An- terior genials shorter than posterior geni- als. Loreal present. Preocular single. Tem- porals 2-1-2 (rarely 1 or 3 anterior or posterior temporals). Body slightly higher than wide; ventro- lateral edge of body angulate. Head slight- ly wider than neck. Pupil round. Eye mod- erately large, its diameter slightly greater than the distance from eye to posterior edge of nostril (x = 1.21 ± 0.1; range = 1.06-1.44; N = 19). Scattered pits present on head scales, most consistently and densely on circum- orbital scales and on prefrontals and nasal; in some specimens, they are liberally sprin- kled over most of the head plates and su- pralabials except for the central parts of the parietals and frontal. Dentition. Maxillary teeth 20-25 + 2 (N = 28; X = 22.3 ± 1.38 prefang teeth). Diastema absent; gap <1 tooth width sep- arating tooth row from enlarged fangs. Ungrooved fangs not offset from tooth row, twice as large as the posteriormost max- illary teeth; having a rounded anterior sur- face (except for distal portion, which has a cutting edge) and a flattened knifelike posterior surface. The tips of the fangs are slightlv compressed. The skulls of two fe- males,'MCZ 180357 and 180370, have, re- spectively, 14-14 and 15-16 palatine teeth, 30-28 and 30-30 pterygoid teeth, and 27- 26 and 28-28 dentary teeth. Hemipenis (Fig. 36). Deeply bilobed, noncapitate, acalyculate (entirely spinose), with a deeply bifurcate centrolineal sulcus spermaticus. Distal tips of the lobes with a central "umbelliform" depression. Coloration in Life (see Glaiv and Vences, 1994:pl. 349 [L. "thieli"7, which is similar to many specimens from the RNP). MCZ 180355 (female): Dorsum Figure 28. Llopholldophls Infrasignatus (Gunther). Lateral view of head of lectotype (BMNH 1946.1 .7.57). olive brown, with indistinct indication of golden dorsolateral stripes anteriorly. Black postocular bar to corner of mouth, crossing middle of last two supralabials (see Do- mergue, 1973:fig. 6). Venter dull grayish yellow (tending to grayish white), with thin black longitudinal markings tending to form lines midventrally and ventrolater- ally. Black speckling on outer edges of ven- trals. A few dorsolateral black specks form- ing roughly two longitudinal rows just be- hind head (ca. 5-10 cm). Upper labials whitish, suffused with brown anteriorly. Lower labials whitish. MCZ 180354 (male): Similar to MCZ 180355, but with orange wash on venter, especially posteriorly. Ventral dark mark- ing forms midventral dark line on most of body and tail. The dorsal ground color in the RNP sample ranges from dull grayish to olive brown to rich golden brown. Some dorsal scales, especially medially on the anterior body, have white scale borders similar to those in stump ffi, epistibes, and lateralis; these do not appear as constant or as vivid in infrasignatus as in these other species. The postocular bar usually crosses the last two supralabials but sometimes ends on the penultimate one; often there is a separated extension on the ultimate supralabial (Figs. 23, 28) The dorsolateral light stripes may be evident primarily on the anterior part of the body, most of the body, or they may be rather indistinct. Most specimens have some indication of black spots dorsolater- ally on the anterior trunk (usually occu- 428 Bulletin Museum of Comparative Zoology, Vol. 154, No. 5 pying one dorsal scale or less); these usually fade by midbody but occasionally are pres- ent the length of the body. A ventral or- angish wash is characteristic of many spec- imens. Coloration in Preservative. Grayish brown to grayish olive dorsal ground color, usually with some indication of dorsolat- eral light stripes anteriorly. Dorsolateral stripes anteriorly on rows 5-6, not conflu- ent with light color of throat (Fig. 23), usually fading by midbody. A blackish postocular stripe from the eye to the corner of the mouth is universally present, and dark punctations are sometimes present on the dorsum, especially anteriorly. Supra- labials whitish except dorsally, where the dorsal ground color encroaches; infrala- bials and throat region whitish. Venter dull white with dark grayish black peppering, spotting, or streaking, usually forming a continuous dark midventral line (Figs. 27, 29; see also Domergue, 1973:fig. 7). Dark grayish pigment usually encroaches upon the venter from the flanks, occasionally reaching the midventer, and sometimes forming a broken dark line on the lateral edges of the venter; in some specimens, most of the venter is dark gray, but the midventral line is usually still evident in such specimens. Smaller specimens tend to have light venters, suggesting an ontoge- netic component to development of the ventral pigmentation. The stratum cor- neum is easily lost in preservative, giving a grayish cast to the dorsum. Natural History. Liopholidophis in- frasignatus is diurnal and terrestrial. This was the most frequently encountered di- urnal snake in forested areas of the RNP, usually active or sunning on trails from early morning to later afternoon; it was found in primary montane rainforest, 900- 1,050 m, and in one higher elevation (1,130 m) short-stature forest, but not in second- ary forests or open habitats. Other species, such as L. sexlineatus and L. lateralis, are possibly numerically more abundant in open habitats such as rice fields, marshes, and secondary forests. Liopholidophis infrasignatus bites in defense and also dorsoventrally flattens the anterior portion of the body. Domergue (1973:1409) reported neck flattening, as well as inflation of the body to reveal white borders of the dorsal scales. Four Liopholidophis infrasignatus con- tained one food item each, all swallowed head first: MCZ 180370 (SVL 552 mm) contained the hind limbs of a large Pleth- odontohyla inguinalis, a large terrestrial microhylid frog; MCZ 180373 (SVL 216 mm) and MCZ 180374 (SVL 330) each contained remains of Plethodontohyla al- luaudi, a small terrestrial microhylid; MCZ 180359-60 (food regurgitated into a com- mon collecting bag) (SVLs 461 and 432 mm, respectively) contained a Chamaeleo nasutus. Domergue (1973) reported frogs in the diet of Liopholidophis ^^thielf (=infrasignatus) . Four females in the RNP sample were gravid: MCZ 180370 (SVL 552 mm), col- lected 21 December contained small yolk- ing follicles; MCZ 180372 (SVL 530 mm), collected 20-23 November, and MCZ 180356 (SVL 555 mm), collected 10 De- cember, contained six and three, respec- tively, large, but nonoviductal yolking fol- licles; MCZ 180362 (SVL 609 mm), col- lected 19 December, contained nine shelled eggs, one of which contained an embryo in Zehr (1962) stage 18. Claw and Vences (1994) reported that gravid females of L. infrasignatus (as L. thieli; locality not stat- ed) collected in November laid six to seven eggs. Domergue (1973) reported six eggs in a female (SVL 546 mm) collected in November, a clutch of six laid by a female (SVL 593 mm) at the end of March, and a clutch of seven laid by a female (SVL 567 mm) in mid-November; all specimens were from Perinet, but details on captive maintenance were not given. Remarks. The type locality of Ptyas infrasignatus, Ankafana (=Ankafina), is a regional name for a forest just west of Tsar- afidy (Carleton and Schmidt, 1990) near the eastern edge of the high plateau. It lies at approximately 1,600 m elevation ac- LiOPHOLiDOPHis (Colubridae) FROM MADAGASCAR • Cadle 429 Figure 29. Liopholidophls infrasignatus (Gunther). Dorsal and ventral views of specimen from the RNP (MCZ 180359). cording to MacPhee (1987).'^ Other recent specimens of Liopholidophis infrasigna- tus have come from the same vicinity (Do- mergue, 1973:1405; two specimens of "thieli" in table II from Tsarafidy). Rax- worthy and Nussbaum (1994:8) cited MacPhee (1987:5) as the authority that the correct name for this locahty is "Ankafi- na," not "Ankafana," based on the desig- nation in descriptions of small mammals collected by Cowan at this locality. How- ever Carleton and Schmidt (1990) used the two names interchangeably. Boulenger (1893:247) corrected Gunther's (1882) er- roneous designation "Arkafina" to "An- kafana." Cowan (1883) himself was prob- ably responsible for the confusion: in the text he refers at least twice to the locality as "Ankafana" (e.g., p. 147), but on the accompanying map it is plotted as "An- kafina." '- The elevation is 1,300-1,540 m according to Rax- worthy and Nussbaum (1994). The FTM 1:1,000,000 map shows a peak at this locahty of 1,679 m. On Cowan's (1883) map, Tsarafidy is denoted as 'Ttsaf- idy. " Domergue (1973) reported occasional presence of two apical pits in Liopholi- dophis "thieli." I noted the presence of 0- 2 apical pits, the number highly variable within and between individuals. Parker (1925:391, footnote) observed much vari- ation in apical pit occurrence in species of the stumpffi group; he noted that one of the three syntypes of infrasignatus had apical pits (number not stated), whereas the other two lacked them. Domergue (1973) reported that the Mal- agasy name Menamaso ("orange eye") was used for Liopholidophis '^thieli" in the Perinet region, in reference to the often- orangish coloration of the iris. The name Mandodrano is used in the RNP area. Notes on Type Specimens of Junior Synonyms. Because I resurrect the name Ptyas infrasignatus from synonymy and place two names as new synonyms of it, I here provide notes on the relevant type specimens of junior synonyms. References are given in the synonymy. 1. Dromicus haroni Boulenger (holo- type, BMNH 1946.1.7.67 [old number 87.12.22.38], aduh female): Total length 734 mm; tail length 167 mm; tail as a per- 430 Bulletin Museum of Comparative Zoology, Vol. 154, No. 5 centage of total length, 23%. Ventrals 158 ( + 2 preventrals), subcaudals 69, one preo- cular, two postoculars, 2-2 temporals. Su- pralabials 8-8 (4-5 touching eye), right infralabials 10 (left side damaged). Dorsal scales P-19-17. Maxillary teeth 24 + 2. Dorsum dark grayish or greenish black, somewhat lighter anteriorly; tail not dif- ferentiated in color. Vague indication of some darker spots or markings when spec- imen under fluid, but this is subtle; ante- riorly, there are dark spots on the neck, forming indistinct reticulated pattern, but no light dorsolateral stripes are evident. Each dorsal scale very finely speckled with light yellowish spots, giving overall velvety appearance ("powdered with yellowish"; Boulenger, 1888:104). Black postocular bar extending diagonally down across last su- pralabial. Otherwise, supralabials white (some grayish suffusion on anterior one or two). Infralabials and gular region white. Dorsal pigment encroaches onto outer 20- 25% of each ventral edge; medial to this and not cleanly separated is a series of large irregular dark splotches (one pair per ventral); midventrally, a series of oblong dark spots forms a more or less continuous midventral line (see Boulenger, 1888:pl. V, fig. 5). Posteriorly on venter, dark pigment increases; underside of tail mostly dark (concentrated midventrally, lighter later- ally). The type of Dromicus baroni has an unusual coloration and pattern from other Liopholidophis, and its placement in the synonymy of Ptyas infrasignatus is pro- visional. Based on coloration, the specimen could be considered a rather unusual vari- ant of either infrasignatus or of stumpffi sensu lato, where baroni has previously been placed (e.g., Boulenger, 1898) (ep- istibes in this work). (Interestingly, the ho- lotype of L. thieli [=i7ifrasignatus] shows fine stippling of yellowish similar to, but less distinct than, that of baroni.) How- ever, unlike all other specimens of either infrasignatus or epistibes studied, the type of baroni has no indication of light dor- solateral stripes, and none was mentioned in the original description (Boulenger, 1888). The proportional tail length (23% of total) and subcaudal counts (69) of bar- oni are within the range of other infrasig- natus females and considerably outside the range of epistibes females (see Table 2). The position of the postocular dark stripe extending diagonally downward across the last supralabial, rather than across its up- per border, is also typical of infrasignatus rather than epistibes (cf. Figs. 8, 23, 28). Hence, the name baroni is synonymized with infrasignatus. Its status should be re- evaluated if additional specimens with precise locality data and having the un- usual coloration of baroni are discovered. 2. Liopholidophis thieli Domergue (ho- lotype, MNHN 1971-332, aduh male with everted hemipenes): Total length 695 mm; tail length 169 mm; tail as a percentage of total length, 24%. Ventrals 144 ( + 2 pre- ventrals), subcaudals 69, anal divided; one preocular, two postoculars, 2-2 temporals. Supralabials 8-8 (4-5 touching eye), in- fralabials 10-10. Dorsal scales 19-19-17; dorsal reduction by fusion of rows 3 + 4 at ventrals 75-72. Maxillary teeth 24 + 2; no diastema. All of these values are typical of infrasignatus (Table 2). The coloration of the type of thieli is identical to that already described for in- frasignatus, although it appears somewhat darkened, perhaps as a preservation arti- fact. The venter of MNHN 1971-332 is strongly patterned, with a median and a lateral series of irregular half-moon-shaped blotches, as well as other irregular spotting. This pattern is within the range of varia- tion observed in the RNP sample of in- frasignatus. Although Domergue (1973) properly resurrected Liopholidophis stumpffi (Boettger) from the synonymy of L. later- alis, failure to examine type material of previously described nominal taxa caused confusion of two species under the name stumpffi, discussed earlier, as well as re- sulting in the description of thieli for a previously described taxon. Direct com- parison of the types reveals that Liophol- idophis thieli Domergue, 1973, is identical with Ptyas infrasignatus Giinther, 1882, LioPHOLiDOPHis (Colubridae) FROM MADAGASCAR • Cadle 431 Figure 30. Hemipenis of Liopholidophis rhadinaea, new species. Fully everted organ of MCZ 180394 (from Talatakely in the RNP), shown in sulcate (left) and asulcate (right) views. Scale bar = 1 mm. a species variously subsumed under later- alis or stumpffi for more than a century (Gunther, 1890; Boulenger, 1893; Moc- quard, 1909; Parker, 1925; Guibe, 1954, 1958). Hence, thieli Domergue is a junior synonym of infrasignatus Gunther. HEMIPENIAL MORPHOLOGY IN LIOPHOLIDOPHIS Everted hemipenes of all currently rec- ognized nominal species of Liopholido- phis are described here. Brief comparisons to the corresponding inverted organs are given for some taxa as necessary. The sexlineatus Group Liopholidophis rhadinaea (Fully Everted Left Organ of MCZ 180394; Fig. 30). The organ is deeply bilobed, non- capitate, acalyculate (entirely spinose), with small cylindrical awns at the tips of the lobes (described later) and a deeply bifurcate centrolineal sulcus spermaticus. Total length of the everted organ approx- imately 6.5 mm, bilobed for the distal 2.5 mm. Sulcus spermaticus forked distally for 3 mm. No basal pockets or lobes. The sulcus spermaticus is a deep groove, bifurcate for about V2 its length, with the branches terminating on the same side of the organ at the base of the apical awns (centrolineal in orientation). The tip of each branch broadens slightly, resulting in fun- nel-shaped distal end of each branch. The stalk of the organ below the lobes is covered on all sides with small hooked spines. The stalk abruptly broadens slight- ly just below the sulcus division, the spines also coincidently increasing in size (spines here about twice as large as those on the base of the stalk). The lobes, including the crotch and inner and outer surfaces, are covered with hooked spines up to the distal tips of the branches of the sulcus. The spines 432 Bulletin Museum of Comparative Zoology, Vol. 154, No. 5 are arrayed more or less in longitudinal rows. Distally, beyond the tips of the branches of the sulcus spermaticus, each lobe has a nude, cylindrical projection (cylindrical awn), each somewhat <1 mm in length (i.e., considering only the nude portion); the distal tip of each awn is more or less flat but is slightly dimpled. These awns are not set off from the tips of the lobes except in lacking ornamentation (nude) and in projecting beyond the ends of the sulcus tips. Dowling (1959) and Dowling and Sav- age (1960) used the term awn for elongate, pointed projections from the apex of col- ubrid hemipenes. My use of the term cy- lindrical awn for the structures in Lio- pholidophis rhadinaea suggests a different shape but does not necessarily imply ho- mology with those as seen, for example, in Tropidoclonion (Dowling, 1959). Lio- pholidophis dolicocercus (see later) has ta- pered apical structures similar to, but less differentiated than, those of L. rhadinaea. The form of the apical structures in rhad- inaea are unique among known colubrid hemipenes. Although the awns on the hemipenes of rhadinaea might be construed as an arti- fact of overeversion, two other specimens with well-everted organs (MCZ 180392, 180396) had similar ornamentation, whereas a specimen with clearly unevert- ed tips to the lobes (MCZ 180402) does not show these structures. To more fully char- acterize these peculiar structures, the ven- tral lobe of an inverted hemipenis (MCZ 180389) was slit midventrally and exam- ined in situ. The hemipenis extends to the level of the suture between subcaudals 6- 7. The awn appears as a nude region (slightly > 1 subcaudal scale in length) be- yond the spinous portion of the lobe. The sulcus, in the dorsolateral wall of the lobe, ends in a slight expansion at the proximal end of the nude region. Liopholidophis dolicocercus (Fully Everted Right Organ of MCZ 180405; Fig. 31). The organ is deeply bilobed, non- capitate, and acalyculate (entirely spi- nose). Sulcus spermaticus deeply bifurcate, centrolineal. Total length of the everted organ 19 mm, bilobed for the distal 10.5 mm. Sulcus spermaticus forked distally for approximately 9.5-10 mm. No basal pock- ets or lobes. The sulcus spermaticus is a deep groove, forked for about half of its length, the branches passing distally on the same side of the organ (centrolineal). Distal tips of the forks not expanded, ending at edge of an apical nude area. Stalk of organ below the lobes on sulcate surface ornamented with tiny hooked spines; these are arrayed in a few rows paralleling the basal undivided part of the sulcus, and with spines generally covering the stalk to one side of the sulcus. "Lateral" surface of stalk between sulcate and asul- cate surfaces largely nude (a few scattered small spines). The asulcate surface of stalk has a me- dian patch of spines from near the base of the organ nearly to the point at which the organ divides. A highly unusual feature of this patch is that each spine appears to be recessed within a small pocket. Distal to division of the organ, the facing surfaces of the lobes are closely appressed and nude (as seen by prying the lobes apart), but distally the facing surfaces di- verge and are ornamented with spines on all sides. At the level of the division of the sulcus spermaticus, the body of the organ is abruptly expanded (from a width of ap- proximately 4.5 mm to approximately 8.5 mm). Concomitantly, the size of the spines abruptly increases, although toward the tips of the lobes spines again gradually become smaller. The narrow, distal portion of each lobe (especially on asulcate side and "crotch" side) is only sparsely covered with tiny spines. Tips of the lobes nude and with a median dimple. The narrow distal portions of the hem- ipenial lobes in Liopholidophis dolicocer- cus appear similar to the apical awns of L. rhadinaea, with two exceptions (cf . Figs. LioPHOLiDOPHis (Colubridae) FROM MADAGASCAR • Cadle 433 Figure 31. Hemipenis of Liopholidophis dolicocercus (Peracca). Fully everted organ of MCZ 180405 (from Talatakely in the RNP), shown in sulcate (left) and asulcate (right) views. Scale bar = 1 mm. 30-31): (1) in dolicocercus the narrow por- tion has a sparse covering of tiny spines (nude in rhadinaea) and (2) the unex- panded tips of the sulcus spermaticus ex- tend to the edge of the distal nude area in dolicocercus, whereas in rhadinaea the tips of the sulcus are expanded and end at the base of the apical awns. Liopholidophis grandidieri (Fully Everted Right Organ of MCZ 180297; Fig. 32). The organ was nearly completely everted upon preservation but subsequent- ly everted fully using the technique of Pe- santes (1994). The organ is deeply bilobed, noncapi- tate, acalyculate (entirely spinose), and with a deeply bifurcate centrolineal sulcus spermaticus. The organ is 11 mm total length, bilobed for the distal 5 mm. The sulcus spermaticus is bifurcate for the dis- tal 6 mm. No basal pockets or lobes are present. The sulcus spermaticus is a broad, deep groove, forked for about V2 of its length, with the branches passing distally on the same side of the organ (centrolineal). Dis- tal tips of the forks not expanded, ending at the distal tips of the lobes. Entire organ ornamented with hooked spines, smallest on the lobes, with an array of larger spines encircling the organ at the point where the lobes join (approximately 8-10 enlarged spines around base of each lobe from sulcus to middle of asulcate side). 434 Bulletin Museum of Comparative Zoology, Vol. 154, No. 5 Spines sparser on extreme distal tips of the lobes than the adjacent proximal portions. Spines sparser in a band around the middle of the organ (immediately proximal to the lobes) than on the base of the stalk or on the lobes. A small nude area is present on each "lateral" surface of stalk just proxi- mal to the union of the lobes. Base of stalk with small dense arrays of spines on all sides, somewhat larger and less dense on asulcate than on sulcate side. The stalk of the organ is of uniform width (i.e., no abrupt expansion, as seen in dol- icocercus and rhadinaea). After division of the organ, the lobes diverge gradually and are densely ornamented with spines on all sides. Liopholidophis sexlineatus (Fully Everted Right Organ of MCZ 180337; Fig. 33). The organ is deeply bilobed, non- capitate, acalyculate (entirely spinose), with a deeply bifurcate centrolineal sulcus spermaticus. Total length of the everted organ approximately 10 mm, bilobed for the distal 4 mm. Sulcus spermaticus forked distally for 4.5 mm. No basal pockets or lobes. The sulcus spermaticus is a deep groove, bordered by thickened, overhanging lips; bifurcate for about V2 its length, with the branches terminating abruptly about 1.5 mm short of the tips of the lobes on the same side of the organ (hence, centroli- neal). Basal undivided portion of sulcus spermaticus bordered on either side by dense array of small hooked spines. Sub- sequent to division of the sulcus, these spines become gradually larger to approx- imately the midpoint of the lobes, then decrease in size toward the tips of the lobes. The basal V2-% of the stalk on the asul- cate side bears a patch of small hooked spines; distal half of the stalk on the asul- cate side is sparsely ornamented with spines, with large more or less nude areas. "Lateral" surface of stalk between the asulcate spinous portion and the sulcus is mostly nude (a few scattered spines, mostly concentrated proximally). Proximal por- tion of the asulcate and "lateral" surfaces of each lobe with approximately 12 some- what enlarged hooked spines; more distal portion of lobes entirely spinose with smaller spines. The facing surfaces of the lobes are entirely spinose, but the crotch has a small nude area between the lobes; on the asulcate side, the nude area in the crotch separates the enlarged spines en- circling the base of each lobe from the corresponding spines of the other lobe. Beyond the distal tips of the branches of the sulcus spermaticus the lobes have a somewhat unusual ornamentation, which is restricted to the apex of the sulcate side (i.e., not encompassing the apex on the asulcate side, which is simply spinose as just described). The apexes bear 8-10 en- larged papillae or folds, each capped by a single spine that is approximately the same size as spines on the adjacent, nonpapillate portions of the lobes. Between the papillae, the organ appears nude. The overall effect of this ornamentation under low magni- fication is to give the apexes of the asulcate surface a somewhat rugose appearance. The papillae on the hemipenis of Lio- pholidophis sexlineatus are not similar to the "apical papillae" described by Dowl- ing (1959), which are merely pointed, awnlike structures at the tips of some col- ubrid hemipenes (one per lobe). However, they are somewhat similar to the spinulate papillae on the lobes of Psomophis hem- ipenes (Myers and Cadle, 1994:13). Unlike Psomophis, in which enlarged papillae are capped by minute spinules, the papillae of L. sexlineatus are capped by a spine ap- proximately the same size as other distal spines on the organ. Based on the minute- ness of the spinules and seemingly weak mineralization of some papillae when mi- cromanipulated, Myers and Cadle (1994: 13) hypothesized that the spinulate papil- lae on Psomophis hemipenes were derived from fully mineralized spines. Such a der- ivation seems less likely for the spinose papillae of L. sexlineatus, in which the spines on the papillae are not noticeably smaller than other distal spines. The di- versity of apical structures in the sexli- LioPHOunoPHis (Colubridae) from Madagascar • Cadle 435 Figure 32. Hemipenis of Liopholidophis grandidieri Mocquard. Fully everted organ of MCZ 180297 (from Mt. Maharira in the RNP), shown in sulcate (left) and asulcate (right) views. Scale bar = 1 mm. neatus group (cf. rhadinaea and dolico- cercus) makes the homology and origin of these structures difficult to discern with present knowledge. Liopholidophis pinguis (Everted Right Organ of MCZ 11701, Prepared from the Inverted Organ by the Method of Pe- santes [1994]). Before the organ was re- moved, it extended to approximately the middle of subcaudal 9 and bifurcated at the level of the suture between subcaudals 5 and 6. The major retractor muscle di- vides at about the base of subcaudal 11. The ventral lobe of the left hemipenis was examined in situ by making a midventral incision. The organ is deeply bilobed, noncapi- tate, acalyculate (entirely spinose), with a deeply bifurcate centrolineal sulcus sper- maticus. Total length of the everted and injected organ approximately 15 mm, bi- lobed for the distal 6.5 mm. Sulcus sper- maticus forked distally for 10 mm. No bas- al pockets or lobes. Stalk and lobes narrow, with no abrupt expansions. No especially enlarged spines anywhere on organ. The sulcus spermaticus is a broad, deep groove, bordered by thickened, overhang- ing lips; bifurcate for about % its length, with the branches terminating at the tips of the hemipenial lobes on the same side of the organ (centrolineal). There seems to be slight displacement of the branches to- ward the outer sides of the lobes, but this may be an artifact of the preparation method; the sulcus in the opened lobe of the inverted organ was on the lateral side of the lobe, as typical for centrolineal sulci. 436 Bulletin Museum of Comparative Zoology, Vol. 154, No. 5 Basal undivided portion of sulcus bordered on either side by dense array of short, thick, curved spines. Similar spines line the outer border of the sulcus from the base to the tip, and the mesial border of the forks of the sulcus beginning at the fork; the latter are continuous with the spinous portion of the lobes, and afford the only continuity between the spinous portions of the lobes. Spines are short spikes sitting atop a broad base. Subsequent to division of the sulcus, the spines become gradually larger to ap- proximately the midpoint of the lobes, then decrease in size toward the tips of the lobes. There are no abruptly enlarged spines. The inner side of the crotch of the organ nude for approximately 25% the length of the lobes, as is the mesial portion of the stalk of the organ between the division of the sulcus and the crotch of the organ. Spinous portions of lobes mesially entirely sepa- rated by nude area in crotch. Asulcate side of stalk with sparse cov- ering of short hooked spines; crotch of or- gan on asulcate side nude. Body of lobes on the asulcate side with dense array of spines, longest proximally, gradually de- creasing in size distally; a short, nearly nude section at base of each lobe has only a few scattered spines. Extreme distal tip of lobes more or less nude (scattered, very minute spines). Stalk of the organ between asul- cate spinous portion and spines bordering the sulcus (i.e., the "sides" of the organ) is nude. The apexes of the lobes of the everted organ were punctured during preparation, but configuration of distal structures was confirmed by examination of the inverted organ of the same specimen. The sulcus extends to the tip of each lobe, which is more or less nude. No peculiar apical struc- tures, as seen in dolicocercus and rhadi- naea, are apparent. The stumpffi Group Liopholidophis epistibes (Fully Evert- ed Left Organ of MCZ 180318; Fig. 34). The organ is deeply bilobed, noncapitate, acalyculate (entirely spinose), with a deep- ly bifurcate centrolineal sulcus spermati- cus. Sulcus spermaticus divides approxi- mately 3 mm from the base of the organ. The lobes diverge strongly from one an- other, essentially lying at right angles to the stalk. Thus, the distal face of the hem- ipenis is formed by the surfaces of the lobes that would normally face one another (i.e., the crotch) if the lobes were not so diver- gent. The tips of the lobes face away from one another at nearly right angles to the axis formed by the crotch and basal stalk. No basal pockets or lobes. The sulcus spermaticus is a deep groove, bifurcate for about % its length, the branches terminating just short of a central depression at the tip of the lobes. The ori- entation of the sulcus is therefore centro- lineal, even though the lobes themselves diverge at nearly 180° from one another. The stalk of the organ is very short and ornamented with scattered minute spines. The base of the lobes is encircled by 3-4 rows of enlarged spines on an expanded midsection of the stalk (>30 enlarged spines around base of each lobe); the mid- section is set off from the short basal por- tion of the stalk by a distinct nude shelf. On the sulcate side, the enlarged spines of the midsection approach the sulcus sper- maticus at its point of division. On the asulcate side, the spines follow the periph- ery of the lobes distally, becoming rather abruptly smaller as the lobes turn away from the stalk. The crotch of the organ, including most of the mesial surfaces of the lobes, is nude except for an array of tiny spinules encircling the distal rim of the lobes. The asulcate surface between the lobes is nude, as is a broad expanse of tissue between the spinous midsections. The distal tips of the lobes have a deep central "umbelliform" depression where the retractor muscle attaches internally (see later). These distal surfaces are ornament- ed with a sparse array of very tiny spinules, arranged in rather indistinct concentric rows. LlOPHOLlDOPHIS (COLI'BHIDAE) FROM MADAGASCAR • Cadle 437 Figure 33. Hemipenis of Liopholidophis sexlineatus (Gunther). Fully everted organ of MCZ 180333 (from Ambatolahy near the RNP), shown In sulcate (left) and asulcate (right) views. Scale bar = 1 mm. Liopholidophis stumpffi. The follow- ing description is based on the right organ of FMNH 18219, a topotype. The inverted organ was studied superficially in situ, be- fore removal and eversion using the meth- od of Pesantes (1994). Although the ever- sion was successful, the tissue probably is not as expanded as would be an organ everted from a fresh specimen. Thus, al- though details of ornamentation are easily discernible, the overall shape of the organ, which has rather narrow, unexpanded lobes, would probably be more similar to that described earlier for epistibes. Before removal, the organ extended to the level of the suture between subcaudals 8 and 9, bifurcating at the level of the suture between subcaudals 2 and 3 (hence, having a short stalk and long lobes). The everted organ is approximately 12 mm to- tal length, bilobed for the distal 9-10 mm (about % bilobed). The sulcus is centroli- neal, dividing about 3 mm from the base of the organ. The lobes diverge from one another but may to a greater extent in a naturally everted organ. No basal pockets or lobes. Overall, the organ is deeply bi- lobed, noncapitate, and acalyculate (en- tirely spinose), with a deeply bifurcate centrolineal sulcus spermaticus. The sulcus spermaticus is a deep groove, bifurcate for about ^A its length, the branches terminating at a central depres- sion at the tip of the lobes on the same side of the organ. The orientation of the sulcus is therefore centrolineal. The distal de- pression of the lobes would likely assume the "umbelliform" shape seen in other 438 Bulletin Museum of Comparative Zoology, Vol. 154, No. 5 members of the stumpffi group if the lobes attained full expansion. The umbelliform area appears to have scattered minute spines in an otherwise nude area. The stalk of the organ is very short, or- namented with scattered minute spines. At the base of the lobes, their outer surface has three to four enlarged hooked spines more or less in a curved line around the outer surface. Above the enlarged spines, the outer surface of each lobe is nude for a small area, above which the lobes are ornamented with spines for the distal % of their length. The enlarged spines are sep- arated from the stalk by a nude shelf and a shallow groove. Except in having many fewer enlarged spines, the spinous mid- section of the stumpffi hemipenis appears similar to that of epistibes, although not as expanded as it would probably in a fully inflated organ. On the sulcate surface, the crotch of the organ has a narrow array of spines bor- dering the sulcus above its division; oth- erwise, the crotch is nude on that surface, as well as on the asulcate surface at the base of the lobes. The facing surfaces of the lobes are basally nude (i.e., in the crotch) for about V4 of its length and spi- nose for the distal %. These spines are somewhat larger proximally, decreasing in size distally. Liopholidophis lateralis. Domergue (1962:101-102, fig. 13; 1973:1410, fig. 3) briefly described and illustrated hemi- penes referred to Liopholidophis lateralis but did not indicate the specimens upon which these were based. His two illustra- tions appear rather different: the earlier figure and description has more strongly divergent and less globose lobes than the later one. Whether this reflects variation or misidentified taxa is unclear (stumpffi, epistibes, and thieli were subsumed within lateralis when the 1962 paper was writ- ten); the strongly divergent lobes of the organ illustrated in the former paper sug- gest hemipenes of L. epistibes, as already described (but see later summary and com- parisons). The lateralis hemipenis illus- trated in 1973 is similar to organs of that species I have studied. The following description is based on the fully everted right organ of MCZ 180380 (Fig. 35). The organ is deeply bi- lobed, noncapitate, and acalyculate (or- namentation consists entirely of spines), with a deeply bifurcate centrolineal sulcus spermaticus. Sulcus spermaticus divides approximately 4 mm from the base of the organ. The lobes diverge strongly from one another, creating overall a Y-shaped or- gan. The tips of the lobes face away from one another at somewhat >45° angles to the axis formed by the crotch and basal stalk. No basal pockets or lobes. The lobes of this preparation are slightly asymmet- rical in size (Fig. 35), but this appears to be subject to some variation, as the left organ of the same specimen and several others examined do not show this asym- metry. The sulcus spermaticus is a deep groove, bifurcate for about % its length, the branches extending to the tip of the lobes and terminating in a central depression at their distal tips. The thickened lips of the sulcus become closely appressed to one an- other, especially distally, essentially mak- ing a closed channel of the groove below the surface. At the distal end of the sulcus adjacent to the umbelliform depression, the channel of the sulcus spermaticus is >1 mm deep. The stalk of the organ is short and or- namented with scattered minute spinules. Just proximal to the point of division of the sulcus, the stalk abruptly expands, forming a broad midsection from which the lobes extend. The midsection is set off by a distinct nude shelf from the proximal narrower portion of the stalk. The mid- section is arrayed with enlarged hooked spines arranged in clusters: viewed from the sulcate side, a large spine occupies the lower corners of the expanded portion of the stalk, and a cluster of 6-8 medium- sized spines (distally grading into the smaller spines of the lobes) is adjacent to the point of sulcus division; on the asulcate LioPHOLiDOPHis (Colubridae) FROM MADAGASCAR • Cadle 439 Figure 34. Hemipenis of Liopholidophis epistibes, new spe- cies. Fully everted organ of MCZ 180318 (from Talatakely in the RNP), shown in sulcate (top) and asulcate (bottom) views. The distal "umbelliform" tips to the lobes in the sulcate view appear to be normal features, rather than a result of incomplete eversion (see text: "Summary and Comparisons of Hemipenes of Liopholidophis"). Scale bar = 1 mm. side the midsections bear 10-12 enlarged hooked spines (larger proximally, smaller distally) that grade into the small spines on the lobes. Large areas of nude tissue occupy the midsections between the clus- ters of spines. A few small spines are present in the fork of the sulcus. Small spines and spi- nules ornament the lobes except the nude central depression at their distal tips; a small wedge of nude tissue is in the crotch of the organ and adjacent basal facing por- tions of the lobes, forming a continuous stretch of nude tissue in the crotch between the asulcate and sulcate sides (i.e., the spi- nous areas of the lobes are not continuous with one another across the crotch). As the lobes diverge from one another, their distal ends turn slightly toward the Figure 35. Hemipenis of Liopholidophis lateralis (Dumeril, Bi- bron, and Dumeril). Fully everted organ of MCZ 180380 (from near Midongy du Sud), shown in sulcate (top) and asulcate (bottom) views. The distal "umbelliform" tips to the lobes ap- pear to be normal features, rather than a result of incomplete eversion (see text: "Summary and Comparisons of Hemipenes of Liopholidophis"). Scale bar = 1 mm. asulcate surface, so that more of the tips of the lobes is visible from the asulcate than from the sulcate side. Lobes, except for the distal depression, are entirely ornamented with small spines, which are larger prox- imally. The distal tips of the lobes are nude and have a deep central depression (de- scribed as "umbiliform" by Domergue, 1962:101) where the retractor muscle at- taches to the inside of the lobes. 440 Bulletin Museum of Comparative Zoologij, Vol. 154, No. 5 Liopholidophis infrasignatus. Do- mergue (1973:fig. 8) illustrated an everted hemipenis of infrasignatus {^^thieli"). The following description is based on the fully everted right organ of MCZ 180368 (Fig. 36). The organ is deeply bilobed, noncap- itate, and acalyculate (entirely spinose), with a deeply bifurcate centrolineal sulcus spermaticus. Total length approximately 15 mm, bilobed for the distal 4 mm. Sulcus spermaticus divides approximately 6.5 mm from the base of the organ. The lobes di- verge slightly, and their distal tips face away from one another (see comments in later summary and comparisons). No basal pockets or lobes are present. The sulcus spermaticus is a deep groove, bifurcate for somewhat more than V2 its length, the branches extending to the tip of the lobes and terminating in a central umbelliform depression. The organ has a narrow stalk orna- mented with scattered small spines and an abuptly expanded midsection proximal to each lobe. The expanded midsections have a battery of enlarged, hooked spines (15- 20 on each midsection) more or less evenly distributed around the circumference of the organ. These spines are larger on the sulcate than the asulcate side, arranged roughly into two to three rows, and grade into the smaller spines of the lobes. On the asulcate side, the spinous midsections of either side are separated from one another by a nude gap in the crotch. The enlarged spines are separated by a shelf of nude tissue and a distinct groove (most promi- nent on the asulcate side) from the spinous stalk. The spines of the midsection grade into those of the lobes on the "lateral" sur- faces of the organ, with an abrupt size transition at the juncture of the lobes and midsections. The distal tips of the lobes have a deep central "umbelliform" depression. The lobes are ornamented with minute spines except distally, where a band of nude tis- sue encircles the umbelliform depression, and proximally in the crotch of the organ. Except for several minute spines within the fork of the sulcus spermaticus, the crotch of the organ is nude from the sulcus spermaticus to the spinous stalk on the asulcate side. The inner surfaces of the lobes (i.e., facing the crotch) are also nude except for a spinous band encircling the distal tips of the lobes (occupying the distal 15-25% of the facing surfaces of the lobes). Summary and Comparisons of Hemipenes of Liopholidophis Hemipenial morphology in the sexli- neatus group is more heterogeneous than in the stumpffi group. Relative to body size, three species (grandidieri, rhadinaea, sexlineatus) have rather small organs, whereas pinguis is intermediate in size, and dolicocercus is large. Liopholidophis dolicocercus, L. rhadinaea, and L. sexli- neatus have peculiar apical structures that are quite different from one another; the others have no such structures. The hem- ipenes of sexlineatus and grandidieri are the most similar pair in size and details of ornamentation (Figs. 32-33), but these are the most generalized organs of the series, lacking any especially distinctive features except for the spinose papillae on the lobes in sexlineatus. In comparison to the sexlineatus group, hemipenes of species of the stumpffi group are more homogeneous but quite different from those of the sexlineatus group. Hem- ipenes of the stumpffi group have a rela- tively short basal stalk (essentially none in epistihes and stumpffi) compared to those of the sexlineatus group. The organ is about Vs bilobed in infrasignatus, about V2 bi- lobed in lateralis, and much more than V2 bilobed in epistihes and stumpffi. Hemi- penes in the sexlineatus group are about 50% or less bilobed (greatest in dolicocer- cus), and all species in this group have a prominent stalk. The organs of the stumpffi group are also large relative to body size compared to all species of the sexlineatus group except dolicocercus. Within the stumpffi group, the hemi- penes of stumpffi and epistihes are more similar to one another in having a very LioPHOLiDOPiiis (Colubridae) from Madac;asc;ah • Cadle 441 Figure 36. Hemipenis of Liopholidophis infrasignatus (GiJnther). Fully everted organ of MCZ 180368 (from Talatakely in the RNP), shown in sulcate (left) and asulcate (right) views. The distal "umbelliform" tips to the lobes In the asulcate view appear to be normal features, rather than a result of incomplete eversion (see text: "Summary and Comparisons of Hemipenes of Liopholidophis"). Scale bar = 1 mm. reduced stalk, whereas lateralis and infra- signatus have organs of more typical pro- portions; the former condition is consid- ered derived (see following section). On the other hand, the organs of stump ffi and lateralis are similar in having few enlarged spines on the midsection, whereas infra- signatus and epistibes have many en- larged spines in this area (cf. Table 3). The overall form of the hemipenis of epistibes, with its lobes widely diverging so that the "crotch" of the organ actually forms its distal face, appears unusual but may be at least partly influenced by in- ternal attachment and constraint by the retractor muscles. Such an effect is sug- gested by comparison of the organ of in- frasignatus previously described with its partner, which was inflated with jelly while still attached to the specimen. The infra- signatus organ described earlier had lobes only slightly diverging, whereas its partner is similar to epistibes in having the lobes much more widely diverging. Hemipenes of the stumpffi, group are characterized by two unusual features that are discussed separately here. (1) Presence of an "umbelliform" de- pression at the tip of the lobes. At first the depression appears to result from incom- plete inflation of the hemipenis. However, it is a consistent feature of the hemipenes of all specimens of epistibes, lateralis, and infrasignatus I prepared in the field, de- spite a conscious effort to effect greater eversion. That the distal depression is not an artifact was proved by internal dissec- tion of an everted organ of L. lateralis (MCZ 180347). The dissection revealed that the "dimpled" appearance results from broad internal attachment of the M. re- tractor penis magnus to the somewhat pleated tissue at the tip of the lobes. No "uneverted" tissue appeared to remain in- side the organ, and I conclude that the umbelliform structure is a normal feature of these organs. The umbelliform lobes of the stumpffi 442 Bulletin Museum of Comparative Zoology, Vol. 154, No. 5 group are unusual among colubrid hemi- penes and here interpreted as a synapo- morphy of the group (see "Monophyly of the Species Groups of Liopholidophis") . However, similar structures appear in hemipenes of some species of Liophidium (personal observations of Liophidium rho- dogaster, and a description and illustration of the organ of Liophidium vaillanti ["L'apex . . . avec une depression centrale" and fig. 5B in Domergue, 1983]). As al- ready alluded to (description of rhadi- naea), and as will be revisited later, ge- neric limits of these and other Malagasy colubrids need reevaluation (see "Mono- phylv of Liopholidophis'' and the discus- sion of MNHN 1988-331). Proper phylo- genetic interpretation of the umbelliform lobes of the stumpffi group will only be possible with a broader survey of hemi- penial morphology of other Malagasy col- ubrids. Nevertheless, the unusual nature of the umbelliform lobes in hemipenes of the stumpffi group are reasonably inter- preted as a synapomorphy of the group at present. (2) An expanded midsection of the hem- ipenial stalk, set off by a nude shelf and/ or groove from the narrow basal portion of the stalk (less distinctly differentiated from the lobes). In the stumpffi group, the enlarged spines encircling the base of the lobes occupy this expanded midsection. Among other Malagasy colubrid hemi- penes examined (representatives of Geo- dipsas, Liophidium, Lycodryas, Mim,o- phis, Drom,icodryas, Madagascarophis, and Pseudoxyrhopus) , only Pseudoxy- rhopus tritaeniatus had a expanded mid- section somewhat similar to that in the stumpffi group. However, in P. tritaen- iatus, the midsection is not as distinctly set off as in members of the stumpffi group. Because of the seemingly restricted taxo- nomic distribution of a differentiated spi- nose hemipenial midsection, I interpret this feature as a synapomorphy of the stumpffi group. Within the stumpffi group, a dif- ferentiated midsection seems least devel- oped in stumpffi (although this interpre- tation is perhaps influenced by the ever- sion method of the organ studied; see ear- lier) and best developed in infrasignatus and lateralis. The midsection is not dis- crete in organs that are not well inflated and is easily overlooked, for example, in everted organs that are nonetheless flaccid. In hemipenes of the stumpffi group, the nude shelf and/or groove delimiting the midsection is reminiscent of the over- hanging shelf setting off the capitulum of those Neotropical colubrid hemipenes de- scribed as "capitate" ("xenodontines" sen- su lato; see Myers, 1973:30-31, 1974:31; Myers and Campbell, 1981:15-17; Myers and Cadle, 1994:13-14). However, the capitation observed in the latter organs does not appear homologous with the condition seen in the stumpffi group, ^^ as suggested by two features: (a) the midsection of hem- ipenes in the stumpffi group is set off by a less well-defined groove and shelf than is the capitulum in truly capitate organs, and (b) in truly capitate organs, the groove delimits a distinct, distal "capitulum" in the case of non-bilobed organs or, in bi- capitate or semicapitate organs, a capitu- lum on each hemipenial lobe (in which case the overhang delimiting the capitu- lum is considerably distal to the division of the sulcus spermaticus). The differen- tiated midsection of hemipenes in the stumpffi group appears more closely as- sociated with the stalk of the hemipenis, rather than with the lobes, and the delim- iting groove and shelf are proximal to the division of the sulcus spermaticus. One other feature of all hemipenes of Liopholidophis seems worthy of note. In comparison to a wide variety of other col- ubrids, the sulcus spermaticus of hemi- penes of Liopholidophis seems unusually broad and deep, although I have been un- successful in quantifying the variation. In many colubrids, the sulcus spermaticus has '^ Some forms of capitation in Neotropical colu- brids also have apparently been independently de- rived more than once. See Myers and Cadle (1994: 13-14 and references therein) for discussion. LioPHOLinoPHis (Coli'rridae) from Madagascar • Cadle 443 a narrow opening on the surface of the hemipenis, is bordered by a very narrow (sometimes indistinct) Hp, and appears as a Hne on the surface of the organ. In Lio- pholidophis (all species of both species groups), the sulcus has a broad surficial opening and is bordered by thickened lips; it appears as a deep, open trough except in some cases (as in lateralis, described earlier) in which the lips are appressed to one another and essentially form a closed canal. No similar structure was observed in hemipenes of other Malagasy colubrids, with the exception of several Geodipsas spp., which otherwise have quite different hemipenes from Liopholidophis. I offer neither a functional nor systematic inter- pretation of the unusual sulcus structure here but call attention to this apparently variable feature of colubrid hemipenes, which seems not to have been previously reported. OSTEOLOGICAL COMPARISONS I have examined one skull each of the species dolicocercus, grandidieri, pinguis, and rhadinaea, and two skulls each of ep- istibes, infrasignatus, lateralis, and sex- lineatus (Appendix). I have not seen a skull of stump ffi, and where I generalize to Lio- pholidophis sensu lato or to the stumpffi species group later, I am assuming that the characteristic under consideration is sim- ilar in stumpffi as in other species of its group (for species group characters) or for the genus as a whole (other characters). This should present no problem, as the skulls of the other species are rather ho- mogeneous for those characters at the ap- propriate level of comparison. A complete osteological description is not attempted here. I discuss only some salient characteristics of the genus; skull characters differentiating the species groups are presented in a later section (see "Monophyly of the Species Groups of Lio- pholidophis"). Polarization of character states as primitive and derived is, in most cases, impossible without reference to an explicit series of outgroups, a hierarchy unavailable with present knowledge of Malagasy colubrids. General Features of the Skulls of Lio- pholidophis. Skulls of all species of Lio- pholidophis are lightly built and of rather ordinary colubrid proportions (Figs. 38- 40). Prefang maxillary teeth moderate in number in the sexlineatus group (17-26; X = 20-24) and in infrasignatus {x = 22); higher in stumpffi, epistibes, and lateralis (22-31, averaging >25 in each species) (Tables 1-2). Orbital Region. The frontals and pa- rietal are considerably emarginated in all species, forming a large orbital foramen (Fig. 40); hence, the ventral borders of the frontals and parietal are widely separated below the orbital foramen. In the stumpffi group, the ventral and posteroventral edg- es of the frontals are emarginated to a greater extent than in the sexli7ieatus group. Consequently, in the stumpffi group the frontals rest on a high frontal crest of the sphenoid (Fig. 40; see later); the fron- tals are less emarginated ventrally and the frontal crest of the sphenoid is more poorly developed in the sexlineatus group (Fig. 40). Ventral emargination of the frontals similar to that of the stumpffi group and associated features such as a high frontal crest on the sphenoid were observed also in Dromicodryas, but not in the other Mal- agasy colubrid skulls examined. The tra- becular groove is open along its entire length and is not obscured laterally by an overlapping flange of the frontal. Basicranial and Posterior Cranial Regions. The Vidian canals are of mod- erate length (see Myers and Cadle [1994: footnote 9] for notes concerning terminol- ogy of Vidian canals). The anterior Vidian foramen is well inside the border of the sphenoid and lies immediately anterior to a bony ridge on the basisphenoid. Detailed morphology of the sphenoid differs be- tween the two species groups, and those differences are described later. Trigeminal foramina double on each side, separated by flange of prootic. A pair of sympathetic foramina on each side ventral to trigemi- nal foramen. 444 Bulletin Museum of Comparative Zoology, Vol. 154, No. 5 PHYLOGENETIC RELATIONSHIPS Monophyly of Liopholidophis There appears to be no unequivocal syn- apomorphy of Liopholidophis sensu lato, which was characterized by Mocquard (1904:303-304) as follows: Maxillary teeth in continuous series, about 20 to 25; mandibular teeth subequal, decreasing gradu- ally in length from front to back; head more or less distinct from the neck; eye moderately developed, with round pupil; body cylindrical; tail usually much longer in males than in females; scales smooth, without apical pits, in 17 or 19 longitudinal series; ventrals without lateral keel; anal and subcaudals divided; posterior trunk vertebrae bearing hypa- pophyses; hemipenis strongly bifurcate. Mocquard's (1904) erection of Liophol- idophis resulted from his discovery that these Malagasy colubrids had a deeply bi- furcate hemipenis, in contrast to most spe- cies of Tropidonotus, where these species had been placed by Boulenger (1893). Boulenger (1893, 1915) had maintained the then-recognized species in the large genus Tropidonotus Kuhl (section Amphiesma; Boulenger, 1893:197), apparently based on their possession of hypapophyses on the posterior trunk vertebrae, but otherwise of rather generalized colubrid morphology (i.e., lacking "derived" features of other hypapophysiate Madagascan genera, such as enlarged anterior mandibular teeth in Dromicodryas) . Parker (1925) and subse- quent authors (Werner, 1929; Guibe, 1954, 1958; Domergue, 1973; Glaw and Vences, 1992, 1994) used Mocquard's concept of Liopholidophis. Indeed, other than the elongate tails of males, which pertains to only a subset of species, Liopholidophis has been a repos- itory for generalized, diurnal, smooth- scaled Malagasy colubrids lacking char- acters such as grooved rear fangs (all Mal- agasy colubrid genera except Dromicod- ryas, Leioheterodon, Liophidium, Lio- pholidophis, and Micropisthodon), en- larged mandibular teeth (e.g., Dromicodryas, Micropisthodon, Pseudox- yrhopus), rostral modifications (e.g., Leioheterodon), vertical pupils (e.g., Madagascarophis), or mandibular and dental modifications (e.g., Liophidium). This situation, in conjunction with external and internal morphological characters dif- ferentiating the two species groups (see later), strongly suggests the possibility of paraphyly (or even polyphyly) of Lio- pholidophis. However, there seems little point in altering the composition of the genus until broader relationships among Malagasy colubrids are examined. Until such time, Mocquard's (1904) definition of Liopholidophis need only be modified to reflect the fact that the tail is unusually elongate in males of only a section of the genus (sexlineatus group) and that these snakes otherwise lack the distinguishing features (? putative synapomorphies) of other Malagasy genera, as just noted. Nev- ertheless, the uncertain monophyletic sta- tus of Liopholidophis requires indepen- dent treatment of the two species groups (which, as documented later, appear to be monophyletic) in comparative or phylo- genetic analyses involving Malagasy col- ubrids. The content of Liopholidophis should be reevaluated as the morphology and relationships of Madagascan colubrids becomes better understood. Monophyly of the Species Groups of Liopholidophis Parker (1925) recognized two species groups of Liopholidophis based on two "key" characters: the sexlineatus group characterized by 17 midbody scale rows and the elongate tail of males, including sexlineatus, dolicocercus, and grandidieri, and the stumpffi group characterized by 19 midbody scale rows and the tail in males of "normal" proportions, including stumpffi and lateralis. Parker (1925) left L. pinguis, which has 17 scale rows but "normal" tail proportions (but see later), unplaced as to species group. I retain Parker's (1925) species groups, but their composition is changed to reflect subsequent new species and revisions (Do- mergue, 1973, and herein). Furthermore, I consider each a monophyletic clade, not- withstanding lack of supporting evidence LioPHOLiDOPHis (Colubridae) FROM MADAGASCAR • Cadle 445 for monophyly of Liopholidophis sensu lato. Thus, the sexlineatus group includes dolicocercus, grandidieri, pinguis, rhadi- naea, and sexlineatus; the stumpffi group includes infrasignatus, lateralis, epistibes, and stumpffi. Species of the two groups are easily distinguished by multiple char- acters, including tail sexual dimorphism, dorsal scale row number, and skull and hemipenial morphology. The sexlineatus group is supported by several apparent synapomorphies, whereas synapomorphies supporting the stumpffi group are fewer in number and more equivocal. However, species in the stumpffi group are similar to one another in external and (especially) hemipenial morphology. I here document the distinguishing characteristics of these groups and include an amplified discussion of several characteristics (e.g., tail length differences). Hemipenial characters are discussed more fully in the previous sec- tion. Apparently derived characteristics are indicated by "D" and the rationale for considering them derived is given. Other characters will not be polarizable until ad- ditional Malagasy colubrids are more com- prehensively studied. In assessing taxo- nomic distribution of several characters, I draw on personal observations from a wide variety of colubrids (especially Neotropi- cal). In addition to skulls of Liopholidophis (see the Appendix for listing), I examined skulls of the following Malagasy colubrids: Dromicodryas bernieri (JEC 12595, 12632), Geodipsas infralineata (JEC 11815), Langaha nasuta (MCZ 18017), Leioheterodon naodestus (MCZ 177382), Liophidium, torquatum (MCZ 11572) and L. rhodogaster (JEC 11571), Lycodryas betsileanus (JEC 11839), Mimophis mah- falensis (MCZ 11715), and Pseudoxyrho- pus tritaeniatus (JEC 11716) (JEC speci- mens to be cataloged in the MCZ). A few skull characters were discernible for sev- eral species of Liophidium from figures or descriptions in Morgan (1973). Numbered characters correspond under the headings for each species group, and present con- trasting characteristics for the two groups. The sexlineatus Species Group. All species of the sexlineatus group share the following characters: (1) 17 scale rows at midbody. (2, D) Strong sexual dimorphism in relative tail length (tail length as a per- centage of total length). The difference between means for males versus females ranges from 7% in pinguis to 20% in gran- didieri (1-2% in species of the stumpffi group; Fig. 37). Expressed differently, the total ranges of relative tail length in males and females do not overlap in any species of the sexlineatus group, whereas, al- though males tend to have longer tails in species of the stumpffi group, the sexes broadly overlap in their ranges of tail pro- portions (the usual situation in colubrids). When the overlap is expressed as [mini- mum 6 value minus maximum 9 value], the difference ranges from 4 to 29% in the sexlineatus group and negative values 1 to 3% in the stumpffi group (Fig. 37). Ad- ditional sampling of pinguis will possibly reveal less distinction between males and females of this species, in which case the extreme sexual dimorphism in tail length would be a synapomorphy of only the sec- tion of the sexlineatus group including dolicocercus, grandidieri, rhadinaea, and sexlineatus. A hypothesis for relationships among species put forward below suggests this as a possibility. As is apparent from Figure 37, the dis- tinctiveness of the tail proportions in the sexlineatus group is attributable to the ex- traordinary lengths of tails in males of dol- icocercus, grandidieri, rhadinaea, and sexlineatus. Females of these species, as well as both sexes of other species of Lio- pholidophis, are rather ordinary in pro- portional tail length. Ironically, as the tails of males of the sexlineatus group are pro- duced to extraordinary lengths, the tails of females (except grandidieri) revert to rel- atively shorter lengths in comparison to those of the stumpffi group (Fig. 37). Comparable data on relative tail length differences for other colubrids are widely scattered, but I am aware of no other spe- cies in which the sex differences approach 446 Bulletin Museum of Comparative Zoology, Vol. 154, No. 5 those of the sexlineatus group. One com- pilation for 16 Neotropical colubrids re- vealed, for two species, maximum differ- ences of 6% between means of propor- tional tail length for each sex; the modal value for the 16 species was 1% difference (Guyer and Donnelly, 1990:table 3). Two colubrids in that study with tail lengths >40% total length {Oxybelis aeneus and Rhadinaea decorata) showed typical over- lap in ranges of proportional tail length between the sexes (means for each sex were identical in O. aeneus; cf . also Myers, 1974: 59, 70, for R. decorata). Klauber (1943) reported similarly narrow differences in relative tail lengths between males and fe- males of a wide variety of North American colubrids. Clearly, species of the Liophol- idophis sexlineatus group are unusual, perhaps unique among colubrids, in this regard. The sexual dimorphism in tail length and subcaudal number in the sexlineatus group is presumably apparent at hatching or birth, but I have seen few specimens of that size, and all those were females, as determined by examination of gonads or for hemipenes. These included three near- hatchlings of rhadinaea (MCZ 180387-88, 180398; SVL 122-170 mm) and one of sexlineatus (MCZ 180378; SVL 180 mm). The smallest males of rhadinaea (MCZ 180396, 180402; SVL 281 and 245 mm, respectively), sexlineatus (MCZ 11606; SVL 259 mm), and dolicocercus (MZUT 796; SVL 265 mm; data from Peracca, 1892) are either at or toward the lower end of the ranges of proportional tail length for males of those species (Fig. 37). On the other hand, their subcaudal counts are to- ward the higher ends of the ranges for their respective species. These observa- tions suggest only a weak association be- tween subcaudal count and relative tail length, as well as an increase in relative tail length with growth in these species. Although seemingly counterintuitive, cor- relations between tail length and subcau- dal counts are weak in several species of colubrids (Klauber, 1945; see also Arnold and Bennett, 1988). One might expect the longer tails of spe- cies in the sexlineatus group to incur great- er frequency of breaks than those of the stumpffi group or greater frequency of breakage of the long tails of males in the former group compared to females. Nei- ther expectation holds: species of the sex- lineatus group do not show greater fre- quency of tail breakage than those of the stumpffi group. Moreover, only in later- alis, in which males do not have inordi- nately long tails, and sexlineatus, were most specimens with tail breaks males. Per- centages of specimens with healed breaks were as follows (percentage followed by total sample size and proportion of speci- mens with breaks that were male): doli- cocercus {0%,9),grandidieri (25%, 4, 0/1), pinguis (25%, 12, 0/3), rhadinaea (5%, 19, 1/1), sexlineatus (10%, 30, 3/3); infrasig- natus (20%, 32, 1/5), lateralis (20%, 44, 7/ 10), epistibes (4%, 24, 0/1), and stumpffi (0%, 13). (3, D) Male superiority in body size (SVL) and ventral counts. With the ex- ception of dolicocercus, males of species in the sexlineatus group reach greater maximum SVLs than do females (Table 1). The absence of this trend in dolicocer- cus is probably due to the small sample of males (5) of that species, and I predict its occurrence in dolicocercus when sufficient samples are available. Males are nearly 40% greater in maximum SVL than females in rhadinaea and sexlineatus, the two species with reasonable samples of both sexes (Ta- ble 1). Perhaps associated with superior male size in species of the sexlineatus group, males of this group (including dolicocer- cus) also have higher ventral counts than females (Table 1). In all species, ranges for ventral counts show virtually no overlap between the sexes, and means for the sexes differ by 8-18 ventrals. As with the statis- tical correlation between tail length and subcaudal number, Klauber (1945) was unable to demonstrate significant corre- lation between body length and ventral number. Female superiority in body size and LioPHOLiDOPHis (Colubridae) FROM MADAGASCAR • Cadle 447 grandidien (29. 20) sexlineatus (11. 19) dolicocercus (8. 13) rhadinaea (9. 15) pinguis (4. 7) stumpffi (-5. 1) epistibes (-3. 1) lateralis (-2. 2) infrasignatus (-1. 2) + 9 20 30 40 50 60 Tail length as a percentage of total length Figure 37. Distribution of tail lengths as a percentage of total lengths in species of Liopholldophis (horizontal axis). Bars indicate the total range of percentages for each species, separated by sex. Numbers within parentheses under each species name indicate, respectively, (a) the difference (%) between the minimum male value and maximum female value; and (b) the difference between the mean values for males and females (%). ventral counts is the rule in the stumpffi group (Table 2), as it is in most colubrids (see, e.g., tabulation in Shine, 1991; notable exceptions occur among garter snakes [Thamnophis] and their relatives, as dis- cussed by Arnold, 1988, and Arnold and Bennett, 1988). Based on the infrequent occurrence of male size superiority in col- ubrids, this character is considered a syn- apomorphy of the sexlineatus group. Giv- en the nonsignificant correlation between ventral counts and body size in snakes (Klauber, 1945), the former could perhaps be considered as a separate, corroborating synapomorphy, although conservatively not treated so here. (4, D) Contact or virtual contact be- tween the postorbital and frontal (Fig. 38). The postorbital nearly contacts the frontal in all species of the sexlineatus group; occasionally, the three bones more or less form a three-way junction. Based on examination of a wide variety of other colubrids, this character seems to appear most often in species known or suspected to be at least partly semifossorial or in di- minutive leaf -litter snakes. Its occurrence in terrestrial snakes such as species of the 448 Bulletin Museum of Comparative Zoology, Vol. 154, No. 5 sexlineatus group is considered derived. Among other Malagasy colubrids, this state is observed in Pseudoxyrhopus tritaenia- tus and species of Liophidium (mayotten- sis, rhodogaster, and vaillanti fide Mor- gan, 1973:figs. 25-27; personal observa- tions of rhodogaster and torquatum) , but not in species of Mimophis, Dromicod- ryas, Langaha, Leioheterodon, Lycod- ryas, Geodipsas, or the Liopholidophis stumpffi group (postorbial and frontal widely separated in these snakes). (5, D) Anterior end of the sphenoid narrow, ending in a single point (Fig. 39). (6, D) Lateral margins of the cultri- form process of the sphenoid convergent, forming a narrow isosceles triangle ex- tending forward from the basal part of the hone (Fig. 39). Characters (5) and (6) are uniformly present in species of the sexlineatus group, giving the sphenoid of these snakes an un- usual form. Pseudoxyrhopus tritaeniatus and Liophidium spp. also have a triangular sphenoid with a single point anteriorly, but details of shape differ from those of the sexlineatus group. (Morgan [1973] report- ed the sphenoid as "notched" anteriorly in Liophidium rhodogaster, but it had a sin- gle point in the specimen I examined.) No other Malagasy colubrids examined had a similar configuration. States (5) and (6) in the sexlineatus group are similar to those in Neotropical snakes of the tribe Pseu- doboini and to burrowing snakes of many clades. However, they are unusual among fully terrestrial colubrids and, consequent- ly, considered apomorphic states of the sexlineatus group. (7) Ventral surface of the sphenoid an- terior to the anterior Vidian foramina bears a deep median groove. The anterior median portion of the basisphenoid, more or less between the anterior Vidian foram- ina, bears a bulbous protuberance. The groove referred to extends forward from this protuberance and is between a pair of parallel bony ridges extending along the cultriform process of the sphenoid. The groove is deepest posteriorly (next to the protuberance); the bony ridges and the groove itself become less prominent an- teriorly. Although species of the stumpffi group have a similar median protuberance on the sphenoid, no bony ridges or asso- ciated groove occur in species of this group; instead, the sphenoid is flat or even slightly convex in this region (a very shallow groove is present in the two skulls of lateralis ex- amined but was not bordered by bony ridges). Pseudoxyrhopus tritaeniatus and Geodipsas infralineata also have a broad groove on the anterior portion of the sphe- noid. (8) Ventral border of frontal usually contacting the dorsal margin of the tra- becular grooves for well more than half the length of the ventral edge of the fron- tal (Fig. 40). This character state is most extreme in dolicocercus, pinguis, and rhadinaea, in which the entire ventral edges of the frontals parallel the dorsal border of the trabecular grooves; in these species, the sphenoid bears only a slight indication of a frontal step. In sexlineatus and grandidieri, the posteroventral edge of the frontals is emarginated and sup- ported on a short frontal step of the sphe- noid; in these species, the posteroventral border of the frontals forms an angle <30° with the dorsal margin of the trabecular grooves (cf. stumpffi group). (9) Dorsal plate of frontals, viewed as a unit, about as wide at its narrowest point as its length (Fig. 38). Species of the sex- lineatus group have a more or less squarish shape to the paired frontals, contrasted with the more rectangular shape seen in the stumpffi group. Thus, the interorbital por- tion of the dorsal plate of the frontals is relatively wide (Fig. 38). (10) Dark stripe occupying at least the lower portion of the first dorsal scale row (usually also occupying the suture line with the ventral scutes and the outer por- tion of the ventrals). In rhadinaea, the stripe is brown and generally restricted to dorsal row 1 (general darkening of the flanks in the "dark" morph extends to out- er edges of the ventrals); in the other spe- LiOPHOLinoPUis (Colubridae) from Madagascar • Cadle 449 Figure 38. Dorsal views of skulls of Liopholidophis showing differences between the sexlineatus and sftympff/groups (represented by dolicocercus and lateralis, respectively). Top: L. dolicocercus (MCZ 180409). Bottom: L. lateralis (MCZ 180350). See text for discussion. cies, the stripe is black, usually involves the adjacent venter (often substantially so in sexlineatus), and sometimes involves other dorsal rows. In pinguis, the stripe is indistinct anteriorly, often restricted to the suture line between ventrals and dorsal row 1. The stripe is well developed in dolico- cercus and grandidieri. No such discrete 450 Bulletin Museum of Comparative Zoology, Vol. 154, No. 5 stripe is present in species of the stumpffi group, although lateral edges of the ven- trals may be spotted or stippled. (11) Dorsolateral light stripe. Absent in all species of the sexlineatus group ex- cept rhadinaea (universally present in stumpffi group). Some specimens of Lio- pholidophis sexlineatus appear to have light dorsolateral stripes (e.g., Glaw and Vences, 1994:pl. 348), but this results from the generally dark middorsum and flanks, rather than from presence of a discrete dorsolateral light stripe. (12) Apical pits absent. Apical pits are absent in all species of the sexlineatus group, whereas they are present in species of the stumpffi group. (13) Passive defense. Species of the sexlineatus group normally seem to use no special defenses such as biting or neck/ body flattening (see species accounts; cf. stumpffi group). (14) Apical ornamentation of hemi- penes various, hut never "umbelliform." See hemipenial descriptions and compare stumpffi group. (15) Undivided portion of hemipenis (stalk) well developed. The stalk is ap- proximately 40-50% or more the total length of the hemipenis in the sexlineatus group. This state is probably plesiomorph- ic given its wide distribution in taxonom- ically diverse colubrids, but equivocally so given the lack of explicit outgroups for these snakes. (16) Stalk of hemipenis not differen- tiated into a narrow proximal portion and an expanded midsection that is set off from the proximal portion by a nude shelf and/or delimiting groove. See hemipen- ial descriptions and compare stumpffi group. The stump£B Species Group. The stumpffi species group is characterized by the following characters (additional com- mentary on some characters in the section immediately preceding; numbered char- acters in the two sections correspond): (1) 19 midbody scale rows. (2) Relative tail length not strongly sexually dimorphic (Fig. 37). (3) Female superiority in body size (Table 2). Characters (2) and (3) are the common conditions among colubrids (see, e.g., Klauber, 1943; Guyer and Donnelly, 1990; Shine, 1991). (4) Postorbital and frontal widely sep- arated by a flange of the parietal (Fig. 38). This is the most common condition observed in a taxonomically and geo- graphically diverse sample of terrestrial colubrids and, with the exception of Pseu- doxyrhopus and Liophidium, the state in all Malagasy colubrids examined. It is therefore probably a plesiomorphic state for the stumpffi group. (5) Anterior end of the sphenoid broad and bifurcate (Fig. 39). (6) Lateral margins of the cultriform process of the sphenoid parallel or slightly diverging (Fig. 39). States (5) and (6) are uniformly present in species of the stumpf- fi, group. The sphenoid, including the form of the cultriform process and of its anterior end, varies greatly in shape among colu- brids. Both states are present in a taxo- nomically and geographically diverse ar- ray of colubrids, but both were uncommon states among the Malagasy colubrids ex- amined (state (6) is seen in Mimophis and Dromicodryas). Given their universal presence in species of the stumpffi group, they probably are plesiomorphic within the group, but whether or not they are syna- pomorphies for the group remains unclear. (7) Ventral surface of sphenoid ante- rior to the anterior Vidian foramina fiat or convex (no median groove or parallel bony ridges extending forward along the cultriform process from median protu- berance). See discussion under sexlinea- tus group. (8) Ventral and posteroventral edges of frontal emarginate, resting high above the margins of the trabecular grooves on a high frontal crest of the sphenoid (Fig. 40). A consistent feature of species of the stumpffi group, the posteroventral margin of the frontal forms an angle >30° with the dorsal margin of the trabecular grooves (cf. sexlineatus group). Liopholidophis la- LioPHOLiDOPHis (Colubridae) FROM MADAGASCAR • Cadle 451 Figure 39. Basicranial region {sphenoid + basioccipltal) of Liopholidophis showing differences between the sexlineatus and stumpffi groups (represented by dolicocercus and lateralis, respectively). Top: L. dolicocercus (MCZ 180409). Bottom: L. lateralis {MCZ 180350). See text for discussion. teralis seems to have the least emargina- tion, whereas epistibes and infrasignatus are more emarginate. This character is widespread taxonomically and geograph- ically within colubrids. (9) Dorsal plate of frontals, viewed as a unit, longer than the width at its nar- rowest point (Fig. 38). See discussion un- der sexlineatus group. (10) No dark stripe on dorsal scale row 1 (of. sexlineatus group). (11) Dorsolateral light stripe. The light stripe is present on rows 5-7 or 5-6 in epistibes and infrasignatus, rows 4-5 in stumpffi, and rows 3-5 in lateralis. A light stripe is present also in rhadinaea of the sexlineatus group (row 6 anteriorly, 5 pos- teriorly) but is otherwise absent in that group. (12) Apical pits present. The number of apical pits appears to be highly variable even within a specimen (0-2 pits present) in the stumpffi group. The pits are readily detectable in some specimens; in others, a careful search is required to detect scat- tered scales with pits. When only a single 452 Bulletin Museum of Comparative Zoology, Vol. 154, No. 5 pit is present, it is asymmetrically placed to one side of the scale tip. Apical pits appear to be more consistently present and evident in lateralis (generally 2 pits) than in other members of the stumpffi group. Scale pits in colubrids vary greatly in their obviousness (see, e.g., Conant, 1961); with the exception of lateralis (when they are present in that species), those in the stumpffi group are not as easily seen as those of many other colubrids. Neverthe- less, all species of the stumpffi group have apical pits (none detected in the sexlinea- tus group). (13, D) Presence of dorsoventral neck flattening as a defensive display. Three species of the stumpffi group flatten the neck as a defensive display (no observa- tions for stumpffi). In at least lateralis, the display can involve a greater portion of the body. In all three species, the behavior highlights the white edges of the dorsal scales and exposes white skin between the scales. This behavior was not observed in any species of the sexlineatus group. The extent of white skin between the scales appears to vary within and among species. Often only small patches adjacent to white scale borders are white; in other cases, more extensive patches of skin are involved. Neck flattening is found in diverse col- ubrids but seems to be rather taxonomi- cally restricted (Greene, 1988). Myers (1986) used the behavior as a synapomor- phy for a Neotropical clade (Xenodontini) comprising six genera. I have not observed the behavior in Malagasy colubrids outside members of the stumpffi group and, thus, consider it also as a synapomorphy of the group. In addition to the use of neck flattening as a defensive display, epistibes, lateralis, and infrasignatus also bite readily in de- fense. This contrasts with species of the sexlineatus group, which appear to seldom bite in defense (see species accounts). (14, D) Distal tip of hemipenial lobes umbelliform (see hemipenial descriptions and discussion; Figs. 34-36). Based on its apparently nearly unique occurrence in species of the stumpffi group, this char- acter is considered a synapomorphy for the group. (15) Undivided portion of hemipenis (stalk) reduced. The stalk is especially re- duced in epistibes and stumpffi, which es- sentially have none. (16, D) Expanded spinose midsection of hemipenis distinctly set off from nar- rower proximal portion of stalk by a nude shelf and /or delimiting groove. Because of the unusual and apparently taxonomi- cally restricted nature of this feature, the differentiated midsection of hemipenes in the stumpffi group is considered a syna- pomorphy. Relationships within the Species Groups Accepting the monophyly of each of the species groups of Liopholidophis, I here briefly explore hypothesized relationships within each group. These hypothesized re- lationships and supporting evidence are summarized in Figure 41. Sexlineatus Group. The following is a suggested synapomorphy scheme for spe- cies of the sexlineatus group and assumes the following plesiomorphic conditions for the group (characters invariant within the group not listed; see "Monophyly of the Species Groups of Liopholidophis^') . (1) Minimal sexual dimorphism in rel- ative tail length (<10% differences be- tween means for the sexes; cf. Fig. 37). In having the least dimorphic tail length pro- portions, and in lacking other clearly de- rived character states, pinguis is consid- ered the most plesiomorphic member of the sexlineatus group. (2) Presence of vivid white borders on dorsal scale rows (present in pinguis, sex- lineatus, and grandidieri in the sexlinea- tus group). Minimally involving dorsal row 3, but often other rows as well (see species accounts). Plesiomorphic condition in- ferred on the basis of presence of this char- acter state in the stumpffi group (stumpffi, epistibes, lateralis, and infrasignatus) . (3) Ventrolateral black stripe on dorsal LlOPHOLIDOPHIS (COLUBRIDAE) FROM MADAGASCAR • Codle 453 Figure 40. Orbital region of skulls of Liopholidophis showing differences between the sexlineatus and stumpffi groups (repre- sented by dolicocercus and lateralis, respectively). Top: L dolicocercus (MCZ 180409). Bottom: L. lateralis (MCZ 180350). See text for discussion. row 1 or suture line between ventrals and row 1: well developed posteriorly (weak or absent anteriorly). In all species except pinguis, the stripe is well developed an- teriorly as well. (4) Lateral black stripe involving dorsal row 3 (may involve adjacent rows as well): well developed posteriorly. In pinguis, the stripe may be well developed (Parker, 1925) or weak (personal observations) an- teriorly. (5) Middorsal pattern uniform the length of the body. The median dorsal scale rows are uniform in ground coloration (i.e., not involving discrete mottling or blotch- ing). Plesiomorphic condition inferred from the condition in the stumpffi group. The diversity of hemipenial structure among species of the sexlineatus group and questionable outgroup structure made use of hemipenial characters for compar- ative purposes here virtually impossible. 454 Bulletin Museum of Comparative Zoology, Vol. 154, No. 5 The following characters unite clades within the sexlineatus group (D = derived; cf. Fig. 41; for ease of interpreting Fig. 41, and for distinguishing them from char- acters in the previous section, these char- acters are given letter designations): {dolicocercus, grandidieri): (a, D) dorsal pattern consisting of complex mottling or reticulations of black and brown anteri- orly, black chevrons or blotches posteri- orly; (b, D) ventral pattern of complete and uniform blackening of ventral scutes (except laterally in dolicocercus). Both patterns, but especially the ventral one, are unusual not only in Liopholidophis, but among colubrids generally; (c) lateral black stripe anteriorly involving scale rows 3-4 (vestigial anteriorly, and completely ab- sent posteriorly in dolicocercus) (In sexli- neatus and pinguis, rows 2-3 are involved, and the stripe is absent in rhadinaea) . (rhadinaea {dolicocercus, grandidieri)): (d, D) Loss of vivid white borders to dorsal scale rows. Plesiomorphic condition of presence of white borders, as already in- ferred. Postulating loss of white borders as a synapomorphy of this clade requires reacquisition in grandidieri (i.e., loss then gain). However, based on the distribution of other postulated derived states (Fig. 41) two independent losses would otherwise be required (in rhadinaea and dolicocercus) . (sexlineatus (rhadinaea (dolicocercus, grandidieri))): (e, D) Development of ex- treme sexual dimorphism in relative tail length (>10% difference between means of relative tail lengths for males and fe- males; see Fig. 37). Liopholidophis pin- guis, in having the least dimorphic tail proportions, is thereby considered the most plesiomorphic species of the sexlineatus group. Nonetheless, the phylogeny hy- pothesized in Figure 41 suggests that sex- ual dimorphism in tail length has not pro- gressively increased during the evolution of the sexlineatus group: sexlineatus and grandidieri, the two species with greatest male tail lengths and greatest dimorphism in relative tail length (Fig. 37), are not sister taxa. If the degree of tail dimorphism has had a complex evolutionary history, then hypothesizing that pinguis is the sis- ter species to the rest of the sexlineatus group on the basis of having the least tail dimorphism may be overly simplistic. However, based on characters examined, pinguis seems to share no unequivocally derived features with other species in the group. Stumpffi. Group. I have been less suc- cessful postulating relationships among species of the stumpffi group. In part this is due to these snakes seemingly being more generalized than those of the sexlineatus group, and in part to the mosaic distri- bution of character states among them (Table 3). Given the questionable mono- phyly of Liopholidophis and lack of ex- plicit outgroups, I have been unable to un- ambiguously polarize the variable char- acters (Table 3). Liopholidophis stumpffi and epistibes are superficially more similar to one an- other (longer tails with more subcaudals, more gracile habitus than lateralis and in- frasignatus) and have an extremely bi- lobed hemipenis (essentially no basal stalk), which seems to be a more derived mor- phology than the less bilobed organs of the other two species (f, D). Hence, I postulate that stumpffi and epistibes are sister spe- cies on this basis (Fig. 41), but any hy- pothesis of relationships within this group seems poorly supported with present in- formation. NOTES ON MNHN 1988-331 (GENUS AND SPECIES INQUIRENDA) Figure 42 In several instances I alluded to prob- lems concerning the generic limits of both Liopholidophis and Liophidium (see Dis- cussion under the description of rhadi- naea; "Monophyly of Liopholidophis''). The problem is sharply focused by one specimen with a mosaic of characteristics of both genera. Domergue (1988:143, "Specimen 1") referred MNHN 1988-331 to Liophidium incertae sedis, but the spec- imen is similar to Liopholidophis rhadi- LiOPHOLiDOPHis (Colubridae) FROM MADAGASCAR • Cadle 455 13*, 14*, 16 1, 2, 3, 4-12, 15 ,cr -i^ J^ xncP rSy ^ ?y ysy vfV j!^ Figure 41 . Hypothesized phylogenetic relationships and character summary for species of Liopholidophis. As noted in the text, the monophyly of Liopholidophis sensu lato is equivocal. Numbered characters discussed in the section "Monophyly of the Species Groups of Liopholidophis"; lettered characters discussed under "Relationships within the Species Groups." Asterisked (*) characters are putative derived characters; underlined characters have plesiomorphic states at the levels indicated; characters with neither designation cannot be postulated as either derived or primitive at the level indicated. The numbered characters, which define the two species groups, have alternative states for each group (see text). naea in overall appearance, although more gracile and with a less distinct head. Com- parison of its everted hemipenes with those of rhadinaea reveal that the organs are nearly identical! Nevertheless, other char- acteristics make this specimen particularly enigmatic. (Domergue's [1988] Liophi- dium "Specimen 2" is a Liopholidophis rhadinaea, which is superficially similar to some species of Liophidium.) Hence, I am- plify Domergue's treatment of MNHN 1988-331 in order to put the problem of generic limits in a broader perspective. MNHN 1988-331 (Fig. 42) was collect- ed 10 December 1966 at Perinet [=An- dasibe] according to a tag attached to the specimen (Domergue [1988] gave the col- lection date as 19 December 1966 and the collector as E. R. Brygoo). It is a male, apparently adult, as indicated by miner- alized spines on the hemipenes (Domergue [1988] reported the specimen as a juvenile), with hemipenes everted. Total length 313 mm, tail length 92 mm (29% of total), head barely wider than neck; 15-15-15 smooth dorsal scale rows without apical pits; 149 ventrals, divided anal plate, 77 subcaudals; loreal present, 1-1 preoculars, 2-2 posto- culars, 1-1 anterior temporals, 1-1 poste- rior temporals; 8-8 supralabials (4-5 touching eye) and 9-9 infralabials. The specimen has 27 + 2 right maxillary teeth, the fangs ungrooved and about twice the size of the teeth immediately preceding (Domergue states "25-30 maxillary teeth"). Teeth curved, sharp, of normal propor- tions, and firmly ankylosed to the bone. The articulation between the dentary and 456 Bulletin Museum of Comparative Zoology, Vol. 154, No. 5 Table 3. Variation in qualitative characteristics among species of the Liopholidophis stumpffi GROUP. stumpffi epistibes lateralis infrasignatus Relatively long tail yes yes yes (males) no (high subcaudal counts) Dorsolateral stripe rows 4-5 rows 5-6 or 5-7 (4-5 posteriorly) rows 3-4-5 (occa- sionally 4 only) rows 5-6 Light stripe confluent yes no yes no with Hght throat Stripe complete the Body: sometimes Body: sometimes Body: yes Body: rarely length of body and Tail: sometimes Tail: sometimes Tail: yes Tail: no tail Venter^ more or less im- usually heavily pig- immaculate usually heavily pig- maculate mented mented Ventral counts rela- yes no (>160 both sex- no (155, males; yes tively low (mean es) 160, females) 25 >25 >25 <25 (mean) Enlarged spines en- single row; few multiple rows; sparse, spines clus- multiple rows; circling base of many tered; few many hemipenial lobes Hemipenial stalk very short very short long intermediate Sulcus spermaticus^ forked >75% its forked >75% its forked ca. 60% its forked ca. 70% its length length length length ^ All species of the stumpffi group may have encroachment of dorsal pigment and/or dark dots at lateral edges of ventrals. This comparison refers to additional ventral pigmentation. ^ The length of the basal unforked portion of the sulcus spermaticus is correlated with the length of the hemipenial stalk, so these two characters may not be independent. However, nothing would seem to preclude a snake having a long hemipenial stalk from having the sulcus divide basally and, thus, having a short unforked portion of the sulcus. the compound bone of the lower jaw ap- pears to be about half way along the length of the dentary. The presence of hypapo- physes on posterior trunk vertebrae was not verified. Coloration in Life (Domergue, 1988: 144, Paraphrased). "Dorsum reddish brown; the head equally reddish brown (but darker than the body) and having three light yellow spots; the upper labials white, spotted with brown; venter red ex- cept for the throat, which is white." Coloration and Pattern in Preservative. General dorsal coloration medium brown to yellowish brown. Dorsolateral light stripe on lower portion of dorsal row 5, bordered above and below by a slightly darkened series of dashes; the light stripe continuous from neck to near the end of the tail and tending to be a series of dashes on the tail. Rows 1-3 light brown; the 3 middorsal rows slightly darker brown than adjacent IV2 rows, which border the dorsolateral light stripe. Three small light nape spots, each surounded by a dark brown line. Top of head brown; head cap separated from white supralabials by a thin dark brown streak along dorsal border of supralabials. A few scattered brown spots on posterior supralabials. Venter and subcaudals, in- cluding throat, gular region, and infrala- bials, immaculate whitish. No dark pig- ment at edges of ventrals, or on scale row 1 except posteriorly, where a series of small brown spots is present; the latter continues onto the tail to become a continuous line at the lateral edges of the subcaudals. Hemipenis. Both hemipenes of MNHN 1988-331 were everted upon preservation, the left one completely, the right organ with the tips of the lobes remaining in- verted. The left organ has a total length LioPHOLiDOPHis (Colubridae) FROM MADAGASCAR • Cadle 457 Figure 42. MNHN 1 988-331 , genus and species Inquirenda. Approximately x 1 .2. of approximately 5 mm, the length of the lobes approximately 2 mm. The sulcus spermaticus divides about 2.5 mm from the base of the organ. The organ is deeply bilobed, noncapitate, acalyculate (entirely spinose), with small nude areas at the tips of the lobes, and a deeply bifurcate cen- trolineal sulcus spermaticus. No basal pockets or lobes. The sulcus spermaticus is a deep groove, bifurcate for half its length, the branches terminating on the same side of the organ at the tips of the lobes (centrolineal in ori- entation). The tips of the branches of the sulcus on the left organ are difficult to dis- cern, as if they simply peter out rather than having a discrete endpoint (distal tips nar- row and very shallow; proper lighting nec- essary to see the ends). The stalk (about 40% the length of the organ) is covered on all sides with small hooked spines. The stalk abruptly broad- ens below the sulcus division, the spines coincidently increasing in size (spines here about twice as large as those on the base of the stalk). The lobes, including the crotch and inner and outer surfaces, are covered with hooked spines except for the distal nude tips of the lobes. The spines are ar- rayed more or less in longitudinal rows. The tips of the lobes are nude but not appearing as cylindrical or as discrete as the awns on the hemipenis of Liopholi- dophis rhadinaea. Discussion. Domergue (1988) did not give reasons for referring MNHN 1988- 331 to Liophidium. Other than head pro- portions (head small and little distinct from neck), the specimen shares few features with other species of Liophidium. The dentition (sharp, curved teeth; enlarged rear maxillary teeth) and a "normal" ar- ticulation between the dentary and com- pound bone of the lower jaw (see Discus- sion under Liopholidophis rhadinaea) seem to preclude association of MNHN 1988-331 with Liophidium as usually de- fined (Boulenger, 1896; cf. also Savitzky, 1983; rear maxillary teeth in some Lio- phidium are somewhat enlarged, but not to the extent seen in MNHN 1988-331 in specimens I examined). Other features, such as the lack of scale row reductions, 15 dorsal rows, labial formulae, and rela- tive tail length, are variable among the nominal taxa presently in Liophidium (ap- proximately 8 species in Madagascar and the Comoro Islands; Domergue, 1983, and personal observations). An unusual feature of MNHN 1988-331 appears to be the presence of a single posterior temporal (2 in other species of Liophidium; Guibe, 1958, and personal observations). The most puzzling aspects of MNHN 1988-331 are the striking similarities to Liopholidophis rhadinaea in color pattern, dentition, and hemipenis, but notable dif- ferences in most other aspects of scalation 458 Bulletin Museum of Comparative Zoology, Vol. 154, No. 5 and body proportions. These similarities include (1) darkened middorsal 3 scale rows, although not so darkened in MNHN 1988-331 as in rhadinaea; (2) dorsolateral light brown stripe (row 5 in MNHN 1988- 331 vs. centered on row 6 in rhadinaea); (3) three light nape spots surrounded by narrow dark brown line (smaller in MNHN 1988-331 than in rhadinaea); (4) top of head plain brown and unpatterned; (5) dark brown line at upper edge of supra- labials separating head cap from white su- pralabials (line not so dark or broad in MNHN 1988-331 as in rhadinaea); (6) im- maculate venter, red in Viiefide Domergue (1988) (pink to vermilion in rhadinaea); (7) dentition similar in overall appearance, and maxillary tooth number for MNHN 1988-331 within the range of variation seen in rhadinaea; and (8) overall similarity in hemipenial morphology. The hemipenes of Liopholidophis rhad- inaea and MNHN 1988-331 are similar in form, differing mainly in two features: (1) the cylindrical awns at the tips of the lobes in rhadinaea are discrete structures, some- what set off from the body of the lobes (Fig. 30), whereas the nude tips of the lobes in MNHN 1988-331 are not so discretely set off; and (2) the sulcus spermaticus in rhadinaea terminates at the base of the awns with a discrete endpoint, whereas the branches of the sulcus in MNHN 1988-331 appear to extend to the tips of the lobes, where they peter out rather than having a discrete endpoint. The organ of rhadi- naea may have a somewhat more dense array of spines on the lobes than MNHN 1988-331, but the difference is subtle. Al- though the hemipenis of MNHN 1988-331 differs in these ways from that of rhadi- naea, the organs of the two are exceed- ingly similar for snakes that otherwise dif- fer in many ways (more similar, for ex- ample, than the hemipenis of rhadinaea is to any other species of Liopholidophis) . In addition to having unreduced 15 dor- sal scale rows, scale counts and tail pro- portions of MNHN 1988-331 are well out- side the ranges for males of Liopholidophis rhadinaea (cf. Table 1): ventrals 149 (vs. 170-179), subcaudals 77 (vs. 126-137), and tail relative to total length 29% (vs. 37- 43%). Its peculiar mosaic suite of character- istics do not allow unambiguous allocation of MNHN 1988-331 to any Malagasy col- ubrid genus as currently defined. That, along with questions already raised con- cerning the proper definition of Liophi- dium vis-a-vis similarities between Lio- pholidophis rhadinaea and Liophidium, differences among species of Liophidium (see Discussion after description of L. rhadinaea), and the questionable mono- phyly of Liopholidophis, suggest that fu- ture work may result in reallocation of some nominal taxa with improved understand- ing of phylogenetic relationships among Malagasy colubrids. The question with re- spect to Liophidium is the extent of in- terspecific variation in the "unique" den- titional and other skull characteristics at- tributed to that genus (Boulenger, 1896: 598-599; Savitzky, 1981, 1983), particu- larly in the new species recently described (Domergue, 1983). Such investigation re- mains to be done. ACKNOWLEDGMENTS I thank Patricia Wright, without whose generous support of logistics my work in Madagascar would have been impossible. The Chicago Zoological Society and the Douroucouli Foundation provided finan- cial assistance for the fieldwork. Several people contributed snakes that found their way into this study, among them especially the RNP guides, Ron Altig, Dan Turk, and Patricia Wright. Talata Pierre and Rajer- iariason Emile were almost constant com- panions for much of the effort; Emile, Rak- otonirina Georges, Rakoto Rafael, and the kind people of Miaranony helped save my skin as well as valuable collections and notes when a lightning strike threatened person- al and scientific loss. Ernest Williams and Rita Gavelis generously assisted with Ger- man translations; Marina ^Verbeloff pro- vided consultation with Greek. Charles Blanc clarified some old locality infor- mation. Franco Andreone supplied photos LlOPHOLIDOPHlS (COLUBKIDAE) FROM MADAGASCAR • Cadle 459 of the holotype of Liopholidophis dolico- cercus, provided information on its pro- venience, and clarified some locality in- formation. Laszlo Meszoly prepared the illustrations of hemipenes. Alan Savitzky shared his knowledge of Liophidium and discussed skull morphology of Liopholi- dophis with me; Chuck Myers advised me on some aspects of hemipenial morphol- ogv. For loan of specimens I am grateful toR. F. Inger and H. Voris (FMNH); C. W. Myers and L. Ford (AMNH); K. Klem- mer and M. Landau (SMF); and R. W. McDiarmid, R. P. Reynolds, and G. R. Zug (USNM). Special thanks to the staffs of The Natural History Museum, London (E. N. Arnold, C. McCarthy, and B. Clarke) and the Museum National d'Histoire Natu- relle, Paris (I. Ineich and A. Dubois) for cordial hospitality and assistance during my visits to their institutions. Chris Rax- worthy facilitated my examination of the type of grandidieri. Charles W. Myers and Harry W. Greene commented on the manuscript. For their thorough and gen- erous support I am grateful to M. Benjamin Andriamihaja, Mme. Berthe Rakotosami- manana (MINISUP), Mme. Celestine Ra- vaoarinoromanga (MPAEF), M. Philemon Randrianarijaona (Directeur des Eaux et Forets), and countless other Malagasy friends. Finally, but critically, a grant from the Ernst Mayr fund of the MCZ permitted examination of types and other material at the BMNH and the MNHN; it is a trib- ute to a scientist of Mayr's stature that, realizing the continued critical need for support of museum-based systematic work for understanding life on earth, he estab- lished the Mayr fund to support this most basic biological enterprise. Publication costs were covered, in part, by the Wetmore- Colles Fund. APPENDIX: SPECIMENS EXAMINED The following abbreviations of collec- tions are used in the text and in the list of specimens examined. As all specimens are from Madagascar, localities begin with the province. Coordinates are given for those localities that could be reliably localized. However, because Malagasy place names are highly redundant, coordinates were not readily apparent for some older specimens examined for which no provinces were given. Bracketed information in localities are inferred political units (province and, where possible, fivondronana), coordi- nates, or updated names for towns. Par- enthetical expressions within localities are part of the original locality data. A useful reference for names of smaller political units within provinces (fivondronanas) is Brygoo (197Lmap4, p. 36), although some must now be updated to reflect current name usage. Some localities are annotated with collector or other historical infor- mation that help localize older sites. Skel- etal preparations examined are indicated as sk (skull) or skel (complete skeleton, in- cluding skull). Specimens of rhadinaea and epistibes are listed in the descriptions of those species. AMNH American Museum of Natural History, reptile collection. New York BMNH British Museum (Natural His- tory), London FMNH Field Museum of Natural His- tory, Chicago MCZ Museum of Comparative Zool- ogy, reptile collection. Harvard University, Cambridge MNHN Museum National d'Histoire Naturelle, Paris MZUT Museo Zoologica dell'Universita di Torino [now incorporated as part of the Museo Regionale di Scienze Natural! di Torino], To- rino SMF Natur-Museum und Forschung- sinstitut Senckenberg, Frankfurt USNM National Museum of Natural History, Smithsonian Institu- tion, Washington, D.C. Liopholidophis dolicocercus (Peracca) FIANARANTSOA: Fivondronana Ifan- adiana: Talatakely, Ranomafana National Park, 970-1 ,050 m [21°16'S, 47°25'E], MCZ 460 Bulletin Museum of Comparative Zoology, Vol. 154, No. 5 180403-08. Mountain ridge N Miaranony, approximately 9.8 km WNW (airline) Tsaratanana, Faravory River, Ranoma- fana National Park, 800 m [2r09'30"S, 47°33'E], MCZ 180409 (skel). [TOAMA- SINA: Fivondronana Moramanga]: "Valle dell'Umbi (Andrangoloka)" [Valley of the Umbi River (Andrangoloka)] [=Andran- goloaka] [19°02'S, 47°55'E], MZUT 0796 (holotype; color slides only seen). Mora- manga, eastern Madagascar [18°56'S, 48°12'E], SMF 17575 (-7246.1a, as listed by Boettger, 1898:25, 1913:312). Liopholidophis grandidieri Mocquard [FIANARANTSOA: Fivondronana Am- bositra]: Ambohimitombo Forest, Mada- gascar [20°43'S, 47°26'E], BMNH 95.7.4.32 (specimen b of Boulenger, 1896:607 [as Tropidonotus dolicocercus]) . Fivondron- ana Ifanadiana: Mt. Maharira, Ranoma- fana National Park, approximately 1,375 m [21°19'59"S, 47°24'57"E], MCZ 180297 (sk). [? TOAMASINA]: Eastern Imerina, BMNH 95.10.29.52 (specimen a of Bou- lenger, 1893:247 [as Tropidonotus doli- cocercus]; see remarks under "Distribu- tion" in species account for this locality). [TOLIARA: Fivondronana Toliara]: L'embouchure du Saint-Augustin ["mouth of the Saint-Augustin River," here consid- ered to be in error] [23°33'S, 43°46'E], MNHN 02-103 (holotype) [the Saint-Au- gustin River is now referred to as the On- ilahy River, and the town of Saint-Augus- tin at its mouth as lanantsony or Anant- sony]. Liopholidophis infrasignatus (Gunther) NO SPECIFIC LOCALITIES: Imeri- na," BMNH 95.10.29.53-55 (specimens k- m of Boulenger [1893:248], as [Liopholi- dophis] stumpffii; see remarks under "Dis- tribution" in grandidieri species account for this locality). [FIANARANTSOA: Fi- vondronana Ambohimahasoa]: Arkafana, eastern Betsileo [=Ankafana fide Boulen- ger, 1893:247; 21°12'S, 47°12'E; 1,600 m fide Carleton and Schmidt, 1990; see "Re- marks" in account for infrasignatus], BMNH 1946.1.7.57 (lectotype, herein des- ignated); BMNH 1946.1.7.56, 1946.1.7.58 (paralectotypes). [Fivondronana Ambosi- tra]: Ivohimanita [approximately 20°50'S, 47°30'E], BMNH 96.10.9.16-17 [specimens collected by Major, who, discussing the lo- cality as "Ivohimanitra" (Major, 1896), gave the elevation as 1,000-1,100 m; MacPhee (1987) gives 700 m, whereas Carleton and Schmidt (1990) give 900 m]. Fivondronana Ifanadiana: Talatakely, Ranomafana National Park, 970 m [21°16'S, 47°25'E], MCZ 180354-70 (180357, skel; 180370, sk). Ivalohoaka, Ranomafana Na- tional Park, approximately 1,040 m [21°17'50"S, 47°26'20"E], MCZ 180371. Mountain ridge N Miaranony, approxi- mately 9.8 km WNW (airline) Tsaratan- ana, Faravory River, Ranomafana Nation- al Park, 850 m [21°09'30"S, 47°33'E), MCZ 180373. Approximately 2.2 km (airline) SE Sahavondrona along Andranoroa River, 1,170 m [21°17'10"S, 47°2r20"E], MCZ 180372. Fivondronana Midongy du Sud: Approximately 7 km SW (airline) Midon- gy du Sud [=Midongy Atsimo], near Rian- ambo ("high waterfall") on Alapo River, 670 m [23°35'S, 47°0rE], MCZ 180374. [TOAMASINA: Fivondronana Moraman- ga]: Moramanga [18°56'S, 48°12'E], SMF 17578. Perinet forestry station, 900 m [=Andasibe; 18°56'S, 48°25'E], MNHN 1971-332 (holotype of Liopholidophis thieli Domergue). 8 km E Perinet [=An- dasibe; 18°56'S, 48°25'E], USNM 149895. Liopholidophis lateralis (Dumeril, Bibron, and Dumeril) NO SPECIFIC LOCALITIES: "Mada- gascar," BMNH 71.6.28.17, 1946.1.15.19 (syntypes of Dromicus madagascariensis Gunther). [ANTANANARIVO: Fivondron- ana Manjakatompo]: Monjakatompo [=Manjakatompo], 10 km W Ambatolom- py [= Ambatolampy] [19°20'S, 47°26'E; 1940 m fide Angel, 1934], AMNH 60675-76, 60679-80 (60676, sk). FIANARANTSOA: LioPHOLiDOPHis (Colubridae) FROM MADAGASCAR • Cadle 461 [Fivondronana Fianarantsoa]: Fianarant- soa [2r26'S, 47°05'E], SMF 57037. Fivon- dronana Ijanadiana: 1-2 km W Rano- mafana (by trail on S side of Namorona River), approximately 700 m [21°15'S, 47°27'E], MCZ 180344-45. Trail between Tsaratanana and Ambohipo, approximate- ly 400-500 m [2rirS, 47°37'E], MCZ 180346-52 (180350, skel). Talatakely, Ranomafana National Park, 970 m [21°16'S, 47°25'E], MCZ 180353. Fivondronana Mi- dongy du Sud: Approximately 4 km SW (airline) Midongy du Sud [=Midongy At- simo], approximately 600 m [23°35'S, 47°0rEl, MCZ 180380. Approximately 7 km SW (airline) Midongy du Sud [=Mi- dongy Atsimo], near Rianambo ("high wa- terfall") on Alapo River, 670 m [23°35'S, 47°01'E], MCZ 180375. [MAHAJANGA: Fivondronana Mahajanga]: Majunga [=Mahajanga; 15°43'S, 46°19'E], SMF 17586, 57163. [TOAMASINA: Fivondron- ana Toamasina]: Tampina [18°30'S, 49°16'E; part of Bluntschili collection; see Mertens, 19331, AMNH 71498. 85 km N Mormunga [? = Moramanga; ?18°56'S, 48°12'E], USNM 149243. [TOLIARA: Fi- vondronana TolagnaroJ: Eminiminy [ap- proximately 24°40'S, 46°55'E], AMNH 71506 [part of Bluntschili collection; in the Ambolo (=Manampanihy) Valley and 400 m elevation fide Mertens, 1933:2611. Fi- vondronana Betroka: Betroka [23°16'S, 46°05'E], USNM 149374-75. [? Fivon- dronana Moramanga]: "Eastern forest" [about half way between Tamatave (=Toa- masina) and Tananarive (=Antananari- vo)], MCZ 11659-68, 11670-73, 11675-81 (see "Remarks" in pinguis species account for discussion of locality). Liopholidophis pinguis Parker NO SPECIFIC LOCALITIES: "Nord- Madagascar," SMF 61909. "Madagascar," AMNH 60692. [TOAMASINA: Fivondron- ana Ambatondrazaka]: Lake Alaotra [17°30'S, 48°30'E], BMNH 1936.3.3.94-97. [Fivondronana Moramanga]: Perinet [=Andasibe; 18°56'S, 48°25'E], USNM 149242. [? Fivondronana Moramanga]: "Eastern forest" [about half way between Tamatave and Tananarive], MCZ 11698- 11701 (11701, sk) (see "Remarks" in pin- guis species account for discussion of lo- cality). Liophoiidophis sexlineatus (Gunther) NO SPECIFIC LOCALITIES: "Eastern Betsileo," BMNH 1946.1.13.17-19 (old numbers 82.5.8.2-4) (syntypes of Dromi- cus sexlineatus Gunther); BMNH 1946.1.13.28-30 (old number 82.2.25) (syntypes of Dromicus macrocercus Gunther). See "Remarks" in species ac- count for locality comments. INDETER- MINATE LOCALITY: Mangerano, SMF 57028, collected by K. L. Koch (probably = Mangarano; the Defense Mapping Agency [19891 lists 12 localities with this name; the SMF has specimens collected by Koch from widely scattered localities in Madagascar, so the particular locality rep- resented by SMF 57028 is unclear). AN- TANANARIVO: Fivondronana Manjaka- tompo: Monjakatompo [=Manjakatompol, 10 km W Ambatolompy [=Ambatolampyl [19°20'S, 47°26'E; 1940 m fide Angel, 1934], AMNH 60678 (sk). FIANARANTSOA: Fi- vondronana Ifanadiana: Ambatolahy, ap- proximately 2.3 km NW (airline) Rano- mafana, approximately 850 m [21°14'55"S, 47°25'48"E], MCZ 180325-35 (180332, skel). Approximately 2.2 km (airline) SE Sahavondrona along Andranoroa River, 1,170 m [21°17T0"S, 47°21'20"E], MCZ 180336-37. Ambodirafia [21°19'S, 47°35'E], MCZ 180338. Fivondronana Midongy du Sud: Approximately 7 km SW (airline) Mi- dongy du Sud [Midongy Atsimo], near Rianambo ("high waterfall") on Alapo River, 670 m [23°35'S, 47°01'E], MCZ 180376-79. [TOAMASINA]: Eastern forest [about half way] between Tamatave (=Toamasina) and Tananarive (=Anta- nanarivo)], MCZ 11602-06 (see comment on this locality in species account remarks for L. pinguis). 462 Bulletin Museum of Comparative Zoology, Vol. 154, No. 5 Liopholidophis stumpffi (Boettger) [ANTSIRANANA: Fivondronana Antsi- ranana]: Route from Antsohihy NW to Di- ego Suarez, NE Madagascar [approxi- mately 12°20'S, 49°05'E], MCZ 54368. Montagne d'Ambre [=Ambohitra; 12°30'S, 49°10'E], MNHN 1893.211 (syntype of Liophidium gracile Mocquard), USNM 150595. [Fivondronana Nosy Be]: Nossi- Be [=Nosy Be; 13°20'S, 48°15'E], SMF 17576 (lectotype, herein designated), 17577, 17580-84; FMNH 18291; BMNH 1946.1.23.51; MNHN 84-595 (syntype of Liophidium gracile Mocquard). LITERATURE CITED Ahl, E. 1928. Neue Frosche der Gattung Rhaco- phorus aus Madagaskar. Zoologischer Anzeiger, 75(11/12): 311-318. Angel, F. 1934. Sur une collection de reptiles et de batraciens de Madagascar. Faune des Colonies Franyais 5(6): 311-320. . 1936. Materiaux herpetologiques recueillis a Madagascar par M. Roger Heim, Charge de Mission: description de deux formes nouvelles. Bulletin du Museum National d'Histoire Natu- relle, Paris, ser. 2, 8: 125-129. Arnold, S. J. 1988. Quantitative genetics and se- lection in natural populations: microevolution of vertebral numbers in the garter snake Tham- nophis elegans, pp. 619-636. In B. S. Weir, E. J. Eisen, M. J. Goodman, and G. Namkoong (eds.), Proceedings of the Second International Confer- ence on Quantitative Genetics. Sunderland, Mas- sachusetts: Sinauer Associates. Arnold, S. J., .\nd A. F. Bennett. 1988. Behav- ioural variation in natural populations. V. Mor- phological correlates of locomotion in the garter snake [Thamnophis radix). Biological Journal of the Linnaean Society, 34: 175-190. Barbour, T. 1918. Vertebrata from Madagascar. 2. Amphibia and Reptilia. Bulletin of the Mu- seum of Comparative Zoology, 61: 479-489. Bastard, E. J. 1898. Quelques mots sur une mission a Madagascar. Bulletin du Museum d'Histoire Naturelle, Paris, 4: 248-251. Blackburn, D. G. 1993. Standardized criteria for the recognition of reproductive modes in squa- mate reptiles. Herpetologica, 49(1): 118-132. . 1994. Review: discrepant usage of the term 'ovoviviparity' in the herpetologica! literature. Herpetological Journal, 4: 65-72. Blommers-Schlosser, R. M. A., and C. P. Blanc. 1991. Faune de Madagascar 75 (1): Amphibiens (premiere partie). Paris: Museum National d'His- ■ toire Naturelle. 379 pp. ■ Boettger, O. "1879" [1877]. Die Reptilien und Amphibien von Madagascar. Abhandlungen der . Senckenbergischen Naturforschenden Gesell- I schaft, 11: 1-56. ' . 1881a. Diagnoses reptilium et batrachior- um novorum ab ill. Antonio Stumpff in insula Nossi-Be Madagascariensi lectorum. Zoologisch- er Anzeiger, 4(87): 358-362, 1881b. Die Reptilien und Amphibien von Madagascar. Dritter Nachtrag. I. Studien iiber Reptilien und Amphibien von Madagascar. II. Aufzahlung der bis jetzt von Madagascar und seinen Kiisteninseln bekannt gevvordenen Rep- tilien und Amphibien. Abhandlungen der Senck- enbergischen Naturforschenden Gesellschaft, 12: 435-558. . 1898. Katalog der Reptilien-Sammlung im Museum der Senckenbergischen Naturforschen- den Gesellschaft in Frankfurt am Main. II (Schlangen). Frankfurt am Main: Druck von Ge- briider Knauer. 160 pp. 1913. Reptilien und Amphibien von Mad- agascar, den inseln und dem festland Ostafrikas (Sammlung Voelzkow 1889-1895 und 1903- 1905), pp. 269-375 [-Hpls. 23-30]. In A. Voeltz- kow (ed.), Reise in Ostafrika in den Jahren 1903- 1905 (Wissenschaftliche Ergebnisse. Band III. Systematische Arbeiten. Heft IV). Stuttgart: E. Schweizerbart'sche Verlagsbuchhandlung, Na- gele & Dr. Sproesser. Boulenger, G. a. 1888. Descriptions of new rep- tiles and batrachians from Madagascar. Annals and Magazine of Natural History, 1, ser. 6: 101- 107. . 1893. Catalogue of the Snakes in the British Museum (Natural History), Vol. I. London: Brit- ish Museum (Natural History). 448 pp. . 1896. Catalogue of the Snakes in the British Museum (Natural History), Vol. III. London: British Museum (Natural History). 727 pp. 1915. A list of snakes of Madagascar, Com- oro, Mascarenes and Seychelles. Proceedings of the Zoological Society of London, 1915: 369- 382. Brygoo, E. R. 1971. Reptiles: Sauriens: Chamae- leonidae: genre Chaniaeleo. Faune de Madagas- car, 33: 1-318. . 1983. Les ophidiens de Madagascar. Me- morias do Instituto Butantan, 46(1982): 19-58. . 1987. L'Endemisme des reptiles de Mad- agascar. Bulletin de la Societe Zoologique de France, 112: 5-38. Cadle, J. E. 1995. A new species of Boophis (An- ura: Rhacophoridae) with unusual skin glands from Madagascar, and a discussion of variation and sexual dimorphism in Boophis albilabris (Boulenger). Zoological Journal of the Linnean Society, 115, in press. Carleton, M. D., and D. F. Schmidt. 1990. Sys- tematic studies of Madagascar's endemic rodents LlOPHOLIDOPHIS (COLUBRIDAE) FROM MADAGASCAR • Cadle 463 (Muroidea: Nesomyinae): an annotated gazetteer of collecting localities of known forms. American Museum Novitates, 2987: 1-36. CONANT, R. 1961. A new water snake from Mexico, with notes on anal plates and apical pits in Natrix and Thamnophis. American Museum Novitates, 2060: 1-22 Cowan, W. D. 1883. Notes on the natural history of Madagascar. Proceedings of the Royal Phys- ical Society of Edinburgh, 7: 133-150. Defense Mapping Agency. 1989. Gazetteer of Madagascar, 2nd edition. Washington, D.C.: De- fense Mapping Agency, xvii + 826 pp. DOMERGUE, C. A. 1962. Observations sur les penis des Ophidiens (deuxieme note). Bulletin de la Societe des Sciences Naturelles et Phvsiques du Maroc, 42(1-2): 87-105. . "1973" [1972]. Etude de trois serpents mal- gaches: Liopholidophis lateralis (D. & B.), L. stumpfii (Boettger) et L. thieli n. sp. Bulletin du Museum National d'Histoire Naturelle, Paris, 3e ser., no. 103, Zool. 77: 1397-1412. . 1983. Notes sur les serpents de la region malgache. III. Description de trois especes nou- velles rapportees au genre Liophidium Boulen- ger, 1896. Bulletin du Museum National d'His- toire Naturelle, Paris, ser 4, 5(sect. A, no. 4); 1109-1122. . 1988. Notes sur les Serpents de la region Malgache VIII. Colubridae nouveaux. Bulletin du Museum National d'Histoire Naturelle, Paris, 4 ser., 10(sect. A, no. 1): 135-146. DowLiNG, H. G. 1959. Apical papillae on the hem- ipenes of two colubrid snakes. American Museum Novitates, 1948: 1-7. DowLiNG, H. G., AND J. M. Savage. 1960. A guide to the snake hemipenis: a survey of basic struc- ture and systematic characteristics. Zoologica, 45: 17-31. Dum£ril, a. M. C., G. Bibron, and A. Dumeril. 1854. Erpetologie Generale, Vol. 7. Paris; Mu- seum National d'Histoire Naturelle. Premiere Partie, vii -I- 4 -I- xvi -I- 780 pp; Deuxieme Partie, xii + 780-1536 pp. Elter, O. 1981. Cataloghi V. — La Collezione Er- petologica del Museo di Zoologia dell'Universita di Torino. Torino: Museo Regionale di Scienze Naturali, Torino. 116 pp. Fischer, J. G. 1884. Uber einige afrikanische Rep- tilien, Amphibien und Fische des Naturhisto- rischen Museums. II. Uber einige Reptilien von Nossi-Be und Madagaskar. Jahrbuch der Ham- burgischen Wissenschaftlichen Anstalten, 1: 33- 38. Gallieni, J. S. 1908. Neuf ans a Madagascar. Paris; Librairie Hachette et Cie. xv + 372 pp. Glaw, F., and M. Vences. 1992. A Fieldguide to the Amphibians and Reptiles of Madagascar. Koln, Germany: Privately Published. 331 pp. . 1994. A Fieldguide to the Amphibians and Reptiles of Madagascar, 2nd edition. Koln, Ger- many; Privately Published. 480 pp. Grandidier, a. 1893. Les voyageurs franqiais a Madagascar pendant les trente dernieres annees. Bulletin de la Societe de Geographic de Paris, ser. 7, 14: 289-300. Greene, H. W. 1988. Antipredator mechanisms in reptiles, pp. 1-152. In C. Gans and R. B. Huey (eds. ), Biology of the Reptilia, Vol. 16, Ecology B: Defense and Life History. New York: Alan R. Liss. GuiBE, J. 1954. Etude de Liopholidophis lateralis (D. et B. ) et description dune espece nouvelle [Reptiles]. Memoires de Institut Scientifique de Madagascar, ser. A, 9: 241-246. . 1958. Les serpents de Madagascar. Me- moires de Institut Scientifique de Madagascar, ser. A, 12: 189-260. GvNTHER, A. 1872. Seventh account of new species of snakes in the collection of the British Museum. Annals and Magazine of Natural History, 9, ser. 4(49); 13-37. . 1882. Ninth contribution to the knowledge of the fauna of Madagascar. Annals and Maga- zine of Natural History, 9: 262-266. 1890. Tenth contribution to the knowledge of the fauna of Madagascar. Annals and Maga- zine of Natural History, 5, ser. 6: 69-72. GuYER, C., AND M. A. Donnelly. 1990. Length- mass relationships among an assemblage of trop- ical snakes in Costa Rica. Journal of Tropical Ecology, 6: 65-76. International Commission on Zoological Nomenclature. 1985. International Code of Zoological Nomenclature, 3rd edition. London; International Trust for Zoological Nomenclature. XX -I- 338 pp. Jan, G. "1863 ' [1862]. Enumerazione sistematica degh ofidi appartenenti al gruppo Coronellidae. Archivio per la Zoologia I'anatomia e la Fisiol- ogia, 2: 213-330. Jan, G., and F. Sordelli. 1879. Iconographie Ge- nerale des Ophidiens, Vol. 5, livr. 49, pi. II. Mil- an. Jourdran, E. "1904" [1903]. Les Ophidiens de Madagascar. Paris; A. Michalon. 63 pp. + 30 pis. Kaudern, W. 1922. Sauropsiden aus Madagaskar. Zoologische Jahrbucher Abteilung fiir Systema- tik, 45: 395-458. KlaL'BER, L. M. 1943. Tail-length differences in snakes with notes on sexual dimorphism and the coefficient of divergence. Bulletins of the Zoo- logical Society of San Diego, 18: 1-60. . 1945. Herpetological correlations. I. Cor- relations in homogeneous populations. Bulletins of the Zoological Society of San Diego, 21: 1- 101. Liddell, H. G., and R. Scott. 1968. A Greek- English Lexicon. Oxford: Clarendon Press. MacPhee, R. D. E, 1987. The shrew tenrecs of Madagascar; systematic revision and Holocene distribution of Microgale (Tenrecidae, Insectiv- ora). American Museum Novitates, 2889: 1-45. Major, C. I. F. 1896. On the general results of a 464 Bulletin Museum of Comparative Zoology, Vol. 154, No. 5 zoological expedition to Madagascar in 1894-96. Proceedings of the Zoological Society of London, 1896: 971-981. Marx, H. 1958. Catalogue of type specimens of reptiles and amphibians in Chicago Natural His- tory Museum. Fieldiana, Zoology, 36(4): 409- 496. Mertens, R. 1933. Die Reptilien der Madagaskar- Expedition Prof. Dr. H. Bluntschili's. Sencken- bergiana Biologica, 15(3/4); 260-274. MoCQUARD, F. 1895a. Sur les reptiles recueillis a Madagascar de 1867 a 1885 par M. Grandidier. Bulletin de la Societe Philomathique de Paris, ser. 8, 7: 93-111. . 1895b. Sur une collection de reptiles re- cueillis a Madagascar par MM. Alluaud et Belly. Bulletin de la Societe Philomathique de Paris, ser. 8, 7: 112-135. . 1904. Description de quelques reptiles et d un batracien nouveaux de la collection du Mu- seum. Bulletin du Museum National d'Histoire Naturelle, Paris, 10: 301-309. . 1908. Description de quelques reptiles et dun batracien nouveaux de la collection du Mu- seum. Bulletin du Museum National d'Histoire Naturelle, Paris, 14: 259-262. -. 1909. Synopsis des families, genres et es- peces des Reptiles ecailleux et des Batraciens de Madagascar. Nouvelles Archives du Museum, Paris, 5 ser., I: 1-110. Morgan, E. C. 1973. Snakes of the subfamily Si- bynophiinae. Ph.D. dissertation. University of Southwestern Louisiana, Lafayette, xiv -I- 260 pp. Myers, C. W. 1973. A new genus for Andean snakes related to Lygophis boursieri and a new species. American Museum Novitates, 2522: 1-37. . 1974. The systematics of Rhadinaea (Co- lubridae), a genus of New World snakes. Bulletin of the American Museum of Natural History, 153(1): 1-262. -. 1986. An enigmatic new snake from the Peruvian Andes, with notes on the Xenodontini (Colubridae: Xenodontinae). American Museum Novitates, 2853: 1-12. Myers, C. W,, and J. E. Cadle. 1994. A new genus for South American snakes related to Rhadinaea obtusa Cope (Colubridae) and resurrection of Taeniophallus Cope for the "Rhadinaea" bre- virostris group. American Museum Novitates, 3102: 1-33 Myers, C. W., and J, A. Campbell. 1981. A new genus and species of colubrid snake from the Sierra Madre del Sur of Guerrero, Mexico. Amer- ican Museum Novitates, 2708: 1-20. NicOLL, M. E., AND O. Langrand. 1989. Mada- gascar: Revue de la conservation et des aires pro- teges. Gland, Switzerland: World Wildlife Fund, xvii + 374 pp. NOBILI, G. 1905. Descrizione di un nuovo pota- monide di Madagascar. Bollettino dei Musei di Zoologia ed Anatomia Comparata della Univ- ersita di Torino, 20(507): 1-4. Parker, H. W. 1925. New and rare reptiles and batrachians from Madagascar. Annals and Mag- azine of Natural History, ser. 9, 16: 390-394. Peracca, M. G. 1892. Descrizione di nuove specie di Rettili e Anfibi di Madagascar. Bollettino dei Musei di Zoologia ed Anatomia Comparata della Universita di Torino, 7(112): 1-5. . 1893. Descrizione di nuove specie di Rettili e Anfibi di Madagascar. Bollettino dei Musei di Zoologia ed Anatomia Comparata della Univ- ersita di Torino, 8(156): 1-16. Pesantes, O. S. 1994. A method for preparing the hemipenis of preserved snakes. Journal of Her- petology, 28(1): 93-95. Raxworthy, C. J., and R. a. Nussbaum. 1994a. A rainforest survey of amphibians, reptiles and small mammals at Montagne d'Ambre, Madagascar. Biological Conservation, 69: 65-73. . 1994b. A review of the Madagascan snake genera Psetidoxyrhopus, Pararhadinaea, and Heterohodon (Squamata: Colubridae). Miscel- laneous Publications of the Museum of Zoology, University of Michigan, 182: 1-37. Savitzky, a. H. 1981. Hinged teeth in snakes: an adaptation for swallowing hard-bodied prey. Sci- ence, 212: 346-349. . 1983. Coadapted character complexes among snakes: fossoriality, piscivory, and duro- phagy. American Zoologist, 23: 397-409. Shine, R. 1983. Reptilian reproductive modes: the oviparity-viviparity continuum. Herpetologica, 39: 1-8. . 1991. Intersexual dietary divergence and the evolution of sexual dimorphism in snakes. American Naturalist, 138(1): 103-122. UICN/PNUE/WWF (Union Internationale pour LA Conservation de la Nature/Programme des Nations Unies pour l'Environnement/ World Wildlife Fund). 1990. Madagascar: Profit de I'environnement. Gland, Switzerland, and Cambridge, England: UICN. xv + 439 pp. Werner, F. 1929. tbersicht der Gattungen und Arten der Schlangen der Familie Colubridae III. Teil (Colubrinae). Mit einem Nachtrag zu den iibrigen Familien. Zoologische Jahrbucher Ab- teilung fiir Systematik, 57: 1-196. Williams, K. L., and V. Wallach. 1989. Snakes of the World, Vol. 1, Synopsis of Snake Generic Names. Malabar, Florida: Krieger Publishing, viii + 234 pp. Zehr, D. R. 1962. Stages in the normal develop- ment of the common garter snake, Thamnophis sirtalis sirtalis. Copeia, 1962: 322-329. 0 22 o u Harvard MC2 LIbrar 3 2044 066 304 346 DATE DUE