OF THE UN IVERSITY or ILLINOIS 590.5 FI V.43 cop. 3 NATURAL F!S-^OKY SURVlY. Edited by Lillian A. Ross Systematics of the LAND AND FRESH-WATER MOLLUSCA OF THE New Hebrides ALAN SOLEM Curator, Division of Lower Invertebrates FIELDIANA: ZOOLOGY VOLUME 43, NUMBER 1 Published by CHICAGO NATURAL HISTORY MUSEUM OCTOBER 19, 1959 PRINTED WITH THE ASSISTANCE OF The Frederick R. and Abby K. Babcock Fund Library of Congress Catalog Card Number: 59-13765 PRINTED IN THE UNITED STATES OF AMERICA BY CHICAGO NATURAL HISTORY MUSEUM PRESS CONTENTS PAGE List of Illustrations 5 Introduction 7 Historical notes 11 Material studied 13 Geography of the New Hebrides 15 Physical factors of the environment 17 Climate 18 Geology 19 Systematic methods 20 Faunal Review 29 Class Gastropoda 30 Subclass Pulmonata 36 Superorder Systellommatophora 36 Order Oncidiacea 37 Family Oncidiidae 37 Order Soleolifera 40 Family Veronicellidae 40 Superorder Stylommatophora 42 Order Tracheopulmonata 43 Family Athoracophoridae 44 Order Heterurethra 52 Family Succineidae 52 Order Orthurethra 58 Family Pupillidae 58 Family Enidae 59 Family Tomatellinidae 62 Family Partulidae 63 Order Sigmurethra 75 Suborder Aulacopoda 75 Superfamily Arionacea 76 Family Endodontidae 77 3 4 CONTENTS PAGE Superfamily Limacacea 89 Family Helicarionidae 91 Family Zonitidae 106 Suborder Holopoda 116 Family Subulinidae 118 Family Bradybaenidae 119 Family Camaenidae 120 Family Bulimulidae 123 Family Paryphantidae 147 Superorder Basommatophora 161 Family Planorbidae 161 Subclass Prosobranchia 166 Order Archaeogastropoda 166 Superfamily Neritacea 166 Family Helicinidae 166 Order Mesogastropoda 179 Suborder Architaenioglossa 179 Superfamily Cyclophoracea 179 Family Poteriidae 180 Family Pupinidae 186 Family Diplommatinidae 190 Superfamily Rissoacea 194 Family Hydrobiidae 194 Family Truncatellidae 197 Family Assimineidae 198 Analysis of New Hebridean non-Marine Snails 204 Number of species 204 Inter-island distribution 204 Endemism 206 Distribution of New Hebridean land snail taxa 211 Appendices I. Kuntz collecting localities on Espiritu Santo 216 II. Mollusca known from Santa Cruz and New Hebridean islands 219 References 222 LIST OF ILLUSTRATIONS TEXT FIGURES PAOB 1. Pallial organs of stylommatophoran snails 34 2. Pedal grooves in the Aulacopoda and Holopoda 36 3. Reticharopa helva (Cox), 1870, and Helix ardua Cox, 1870 = Liardetia samoensis (Mousson) 88 4. Variation in Trochomorpha rubens Hartman 110 5. Phylogeny of the Pacific Bulimulidae 130 6. Umbilical area of Pleuropoma 175 7. Distribution of families of New Hebridean land snails 213 8. Distribution of genera of New Hebridean land snails 214 PLATES 1. Map of the New Hebrides. 2. Southeastern Espiritu Santo. 3. Anatomy of New Hebridean slugs. 4. Anatomy of Aneitea. 5. Anatomy of Succinea. 6. Aulacopod and prosobranch anatomy. 7. Anatomy of Dendrotrochus. 8. Type photographs of Placostylus, Draparnaudia, and Rhachistia. 9. Type photographs of Pariula, Gonatoraphe, and Pupina. 10. Type photographs of Pleuropoma, Rhytida, Dendrotrochus, and Macrocycloides. 11. Type photographs of Delos and Omphalotropis. 12. Type photographs of Partula, Trochomorpha, and Placostylus. 13. Aneitea, Succinea, and Rhachistia. 14. 15. Partula. 16. Diplomorpha, Partula, and Delos. 17. Placostylus salomonis. 18. Placostylus. 19. Placostylus and Diplomorpha. 20. Placostylus fuligineus. 21. Placostylus and Diplomorpha. 22. Placostylus bicolor. 6 LIST OF ILLUSTRATIONS 23. Diplomorpha. 24. Trochomorpha and Inozoniies. 25. Dendrotrochus and Physastra. 26. Gonatoraphe, Pleuropoma, and Omphalotropis. 27. Minute prosobranchs. 28. Phrixgnathus and Ouagapia. 29. 30. Reticharopa. 31. Mocella and Reticharopa. 32. Discocharopa and Diastole. 33. Limacoids and Gyraulus. 34. Limacoids. Systematics of the Land and Fresh 'Water MoUusca of the New Hebrides* INTRODUCTION This study summarizes our present knowledge of the New Hebri- dean land and fresh-water Mollusca and surveys the distribution patterns of Pacific land snails. Too little collecting has been done in the New Hebrides to allow much discussion of speciation. It is only possible to discuss the nomenclatural status of the named forms, summarize the limited distributional data, and suggest possibilities for field studies. More attention has been focused on trying to determine the relationship of the New Hebridean species to those found on other islands of the Pacific. Resulting from these com- parative studies are the tentative phylogenetic trees of the land Mollusca (figs. 10 and 11) and the discussion of Indo-Pacific land snail geography (pp. 245-331). The major impetus to this study was provided by the collections made on Florida Island in the Solomons, and on Espiritu Santo in the New Hebrides, by Robert E. Kuntz from 1943 to 1945, and presented to the University of Michigan Museum of Zoology in 1947. Check-lists of the New Hebridean marine mollusks and a study on the Florida Island land and fresh-water species are in press. Attempts to identify the Espiritu Santo land and fresh-water snails soon involved summarization of the New Hebridean fauna, based on material in American museums. The few species known from the Santa Cruz Islands (see Appendix II) are considered here, since their affinities are with New Hebridean, not Fijian or Solomon Island taxa. The only previous comprehensive study of the New Hebridean non-marine mollusks is a checklist (Kobelt, 1881, pp. 19-20) com- piled solely from descriptive literature. Of the fifty-four species listed, five are synonyms, eleven are from Lord Howe Island off 1 Modified from a dissertation submitted in partial fulfillment of the require- ments for the degree of Doctor of Philosophy in the University of Michigan, 1956. 8 FIELDIANA: ZOOLOGY, VOLUME 43 Australia, and eleven are from other localities not in the New Hebrides. Only twenty-seven represent valid records. Sykes (1903) reported thirty species from various New Hebridean localities and recorded sixteen estuarine mollusks. The other literature specifi- cally concerned with New Hebridean non-marine mollusks is the scattered pre-1870 descriptions by Louis Pfeiffer in the Proceedings of the Zoological Society of London and short papers by Cox (1870), E. A. Smith (1884), Thomson (1885), Hartman (1886, 1888, 1889, 1890, 1891), Ancey (1896, 1897, 1905, 1906), Mabille (1895), Grimpe and Hoffmann (1925a), and Hoffmann (1929b). All of these papers are purely descriptive in character and are based on incidental collections. Except for a few species found by John Brazier in 1865 and described by Cox (1870), no New Hebridean material has been collected by a malacologist or even a person primarily interested in mollusks. The specimens in museums were either a by-product of expeditions oriented toward the collecting of other phyla or were haphazard gatherings by missionaries, traders, or planters untrained in science. It is remarkable, and extremely unfortunate, that prior to 1955 no person working on the mollusks of the Bismarcks, Solo- mons, and New Hebrides had ever been to the islands, much less collected the material on which his publications were based. It is hoped that my own lack of field experience in the area can be remedied in the near future. The age of the literature pertaining to the New Hebrides only reflects the general status of land snail taxonomy during the twen- tieth century. In the latter part of the nineteenth century, several devoted collectors visited the Pacific islands and many papers were written on their findings. W. Harper Pease, Andrew Garrett, Eduard Graeffe, J. Kubary, Otto von Moellendorff, Xavier Mon- trouzier, Saint-Martin Souverbie, John Brazier, and Edgar Leopold Layard laid the foundations of our knowledge. The papers of Pease, Garrett, and Moellendorff were based on material that they had collected, while the field efforts of the others resulted in the many papers of Mousson, Gassies, Hartman, Crosse, Ancey, and Fischer. During the twentieth century the only important land snail collectors in the Pacific have been the late C. Montague Cooke and his prot^g^, Yoshio Kondo, both of the Bernice P. Bishop Museum, Honolulu. Cooke published comparatively little, but his collections are invaluable and it will be many years before they have been fully studied. SOLEM: MOLLUSCA OF THE NEW HEBRIDES 9 In the last fifty years the important publications on Pacific snails have been few in number. Excellent monographs of the Partulidae (Pilsbry, 1909), Tomatellinidae (Pilsbry and Cooke, 1915-16), and Limacacea (H. B. Baker, 1938b, 1940, 1941) overshadow lesser ones of the Athoracophoridae (Grimpe and Hoffmann, 1925a) and Heli- cinidae (Wagner, 1907-11). Faunal studies on Australia (Hedley and Iredale), New Zealand (Suter, Powell, and Dell), New Cale- donia (Dautzenberg and Franc), Hawaii (Pilsbry and Cooke), and Indonesia (van Benthem Jutting, Bemhard Rensch) are compre- hensive, but little has been published on the fauna of the "oceanic" islands. Checklists for the Solomons (Clapp, 1923) and Hawaii (Caum, 1928) ; monographs of some of the endemic Hawaiian fami- lies by Pilsbry and Cooke; shorter papers by Use Rensch, Clench, Kondo, Abbott, Thiele, and Hoffmann; and the variational studies of Crampton and Welch, all provide fragmentary data. The phy- logeny and zoogeography of Pacific land snails were discussed by Hedley (1892a, 1899), Pilsbry (1900b, 1916, 1921), C. M. Cooke (1926), Germain (1932, 1934), and H. B. Baker (1941). All these studies were based on the Polynesian fauna and their conclusions are greatly altered when the Austro-Melanesian fauna is considered as a unit. The brief survey of the literature given above shows that, except for the Limacacea (see H. B. Baker, 1938b, 1940, 1941), the study of Pacific land snails has not benefited from the use of the "New Systematics." The many older publications contain a wealth of factual data waiting for synthesis and interpretation in the light of modem systematic theory. In trying to determine the relationships of the New Hebridean species, I found it necessary to review the classification of a number of families. Since this was primarily a faunal survey, it was not possible to revise thoroughly any family or any widely distributed genus. In most cases the available data suggested probable phylogenetic sequences and distributional pat- terns, but the information needed to test the hypothetical solutions was not available. Study of the shells and the few references in the literature to the anatomy of the Pacific Succineidae, Endodontidae, Bulimulidae, and Paryphantidae resulted in new interpretations of distribution, generic units and phylogeny, which, if confirmed by subsequent research, will greatly change widely held concepts of the age, origin, and means of dispersal of the land snails. The system- atic review (pp. 29-204) thus contains data bearing on the classifi- cation of snails from all parts of the Pacific as well as the New Hebrides. 10 FIELDIANA: ZOOLOGY, VOLUME 43 I have recognized 79 species as occurring in the New Hebrides and Santa Cruz Islands, although several additional names are listed in the systematic review. In several cases the material needed to prove the specific identity of named morphological variants was not available. These "species" are retained as nomenclatural units, but their probable relationship to other species is clearly indicated in the text and they are not accepted as valid species records. Six- teen new species and subspecies (listed on p. 331) and six new supra- specific names (listed on p. 332) are proposed. The major portion of this study was done at the University of Michigan Museum of Zoology from June, 1954, through May, 1956. Revision of the systematic review and expansion of the zoogeographi- cal survey was completed at Chicago Natural History Museum in the following year. In preparing this report, I have been aided by many people. For permission to study collections under their charge and for loan of specimens I am indebted to Drs. Fritz Haas (Chicago Natural History Museum), the late Henry A. Pilsbry and R. Tucker Abbott (Academy of Natural Sciences of Philadelphia), Harald A, Rehder (United States National Museum), William J. Clench (Museum of Comparative Zoology), and Juan Jos^ Parodiz (Carnegie Museum, Pittsburgh), Commander Walter B. Miller (Falls Church, Virginia), and Messrs. W. K. Dell (Dominion Museum, Wellington, New Zealand) and William D. Clarke (American Museum of Natural History). Drs. Lothar Forcart (Mus^e d'Histoire Naturelle de Bale), Andr^ Franc (Museum d'Histoire Naturelle, Paris), the late Guy L. Wilkins (British Museum [Natural History]), Yoshio Kondo (Bernice P. Bishop Museum, Honolulu), and Donald F. McMichael (Australian Museum, Sydney) sent photographs and information about specimens I was unable to see personally. Data on specific taxa were generously provided by Drs. William J. Clench, Joseph P. E. Morrison, Yoshio Kondo, R. Tucker Abbott, and Henry van der Schalie. Dr. Juan Jos6 Parodiz made available notes on South American non-marine fossil moUusks; Dr. Harry S. Ladd provided data on Pacific geology and the fossil organisms found at Bikini Atoll. The late Emmett Reid Dunn read the section of general sys- tematics; and Drs. Yoshio Kondo, Theodor Just, and the late Karl P. Schmidt read the section on zoogeography. For help with the preparation of the manuscript I am indebted to Mr. Dwight L. Chapman, Mr. William L. Cristanelli (pis. 1 and 2), Mr. Harold J. Walter, Mrs. Phyllis Kannowski, the late J. Speed SOLEM: MOLLUSCA OF THE NEW HEBRIDES 11 Rogers, and the members of my doctoral committee at the Univer- sity of Michigan. The revision of the manuscript has been aided by my colleagues at Chicago Natural History Museum, particularly Dr. Fritz Haas and the late Karl P. Schmidt. They first interested me in zoology and through the years have devoted many hours of their time to furthering my biological career. Their help in the preparation of this paper has been invaluable. The beautiful illus- trations of minute species (pis. 29-34) are the work of E. John Pfiffner and most of the text figures were rendered by Marian Pahl, both of Chicago Natural History Museum, Division of Illustration. In particular I am indebted to Dr. Henry van der Schalie of the University of Michigan. His constant encouragement, constructive criticisms, and generous support have made possible what at times seemed an impossible task. Without the help he has given me this study would never have been finished. Historical Notes Many explorers visited the New Hebrides between 1595 and 1825 (see Markham, 1873; Bourge, 1906; and Harrisson, 1937), but there are no records of the collecting of mollusks until the voyage of the Astrolabe to Vanikoro and Ticopia in the Santa Cruz Islands in 1828 (see Quoy and Gaimard, 1832-35). The many marine shells and four species of land snails — Partula vanikorensis, Pleuropoma taeniata, Trochomorpha sp., and an unidentifiable Truncatella — repre- sent the largest collection made until less than twenty years ago. Some of the specimens are still preserved in the Museum d'Histoire Naturelle, Paris. In the late 1820's sandalwood was discovered in the islands and the ships of the China trade visited the southern New Hebrides regularly. George Bennett was at Dillon's Bay, Erromanga, on August 24, 1829, and collected the first living specimen of Nautilus pompilius (Linnaeus) seen by a European. After the trading voyage was over, the specimen was sent to London and formed the basis of the classic memoir on the "pearly nautilus" by Richard Owen (1832). No information on collectors in the 1840's and early 1850's was available to me. During this period missionaries were attempting to settle on the islands and a few of the larger land snails of Aneiteum, Tanna, and Erromanga reached London and were described by Louis Pfeiffer between 1852 and 1861. Some came from a mission- ary. Rev. Turner, and a specimen of Placostylus collected by the famous Nova Scotian missionary, John Geddie, is still preserved 12 FIELDIANA: ZOOLOGY, VOLUME 43 (USNM 23017). The bloody history of the settlement of the New Hebrides is well known. Several times gunboats were sent from New Caledonia to raid native villages in retaliation for the deaths of traders and missionaries, and the cruises of the Herald in 1854, with John MacGillivray as naturalist, and the Curagao in 1865, with John Brazier of Sydney, resulted in small shell collections. Mac- Gillivray revisited the southern New Hebrides in 1858, 1859, and 1860 (Iredale, 1937c, p. 60), but Brazier only made the one trip, touching briefly at Aneiteum, Tanna, Erromanga (where he collected shells while under attack by the natives!), Vate, and Vanua Lava (see Brenchley, 1873), Brazier found several minute species (Cox, 1870) as well as large, showy shells. In the 1850's the expanding sugar-cane plantations of Fiji and Queensland caused an acute labor shortage and New Hebridean natives were impressed as indentured field hands. The boat trips between the New Hebrides and Queensland probably resulted in some chance introductions. Aneitea brishanensis W. Pfeiffer (1900) may have been carried to the botanical gardens of Brisbane on plants from the New Hebrides. Since plants were carried as food on the voyages, Triboniophorus graeffei Humbert, a Queensland species, could have been carried to the New Hebrides if the locality record of Glamann (1903) is correct. Rhachistia histrio (Pfeiffer) is an arboreal snail found in New Caledonia, Vate, Tanna, and Queens- land. The order of introduction is unknown, but I suspect a New Caledonia-New Hebrides-Queensland sequence. The original home of R. histrio may be Madagascar. In the 1870's and 1880's many plantations were established in the New Hebrides. Edgar Leopold Layard, the British Consul at Noumea, New Caledonia, was an ardent shell collector and exploited the potentialities of his position. Layard sent old European news- papers to New Hebridean planters in exchange for shells found by the planters while clearing land. George de Lautour on Espiritu Santo and W. Glisson on Vate are the only ones whose names were recorded, but there were undoubtedly others. Layard sent speci- mens to C. F. Ancey in France, who described several species be- tween 1884 and 1906, and to William D. Hartman in Philadelphia, who published papers between 1886 and 1891. Most of Hartman's types and the correspondence from Layard are preserved at the Carnegie Museum, Pittsburgh, and were made available by Dr. Juan Jos^ Parodiz. Prior to Layard's fatal illness, he and Hartman planned to summarize the New Hebridean land SOLEM: MOLLUSCA OF THE NEW HEBRIDES 13 snail fauna. Numerous preliminary notes in the Layard correspond- ence have been incorporated into this study. Layard 's collection was purchased by the late E. R. Sykes and is now in the British Museum (Natural History). Material collected primarily (if not entirely) on Espiritu Santo by Dr. Ph. Francois was studied by Jules Mabille (1895). Mabille's identifications were inaccurate and all his unconfirmed records have been rejected as unreliable (see Ancey, 1905). Dr. Andr^ Franc kindly provided photographs of the types of several of Mabille's unfigured species in the Paris Museum (see pi. 12). J. J. Walker, a ship's engineer, collected specimens during the cruise of the Ringarooma between June and October, 1900 (see Sykes, 1903), and found three new species. Felix Speiser, an anthropologist, who was working on Malo and Espiritu Santo in 1911 (see Grimpe and Hoff- mann, 1925a), found slugs which are now in Basel, Switzerland. Dr. Lothar Forcart was unable to locate any other mollusks col- lected by Speiser. Professor John R. Baker of Oxford visited Espi- ritu Santo and Gaua in 1922 and 1927 and spent a year (1933-34) at Hog Harbour, Espiritu Santo. He made general biological col- lections and the mollusks were turned over to Guy Robson of the British Museum (Natural History). The slugs were studied by Hoffmann (1929b) and a few notes on the fresh-water snails were published by J. R. Baker (1929). The material in the British Museum was not located until 1958 and reached me while this monograph was in press. A detailed report will be issued later, but a few important notes have been added below. E. Aubert de la Rue was in the New Hebrides from 1934 to 1936. The marine shells (Fischer and Fischer-Piette, 1938, 1939) and the fossils he collected (Abrard, 1946) are in the Paris Museum. I was unable to determine if he collected any non-marine shells. L. Mac- millan, of the Whitney South Sea Expedition, picked up a few shells on Tanna, Aneiteum, and Erromanga, in 1937. They were deposited in the American Museum of Natural History and lent to me for study. Several members of the United States Armed Forces made collections during World War II. Those that could be located are mentioned below. Undoubtedly many more are still in private hands. Material Studied About 5,100 specimens (581 lots) of New Hebridean land and fresh-water shells were examined. The largest single collection, 225 14 FIELDIANA: ZOOLOGY, VOLUME 43 lots, was made by Robert E. Kuntz on southeastern Espiritu Santo (see pi. 2 and Appendix I). Kuntz not only established geographic collecting stations but often divided each site into ecologic niches. My observations on the status of Trochomorpha ruhens vs. T. con- vexa, Pleuropoma albescens vs. P. sublaevigata, Dendrotrochus layardi, and the Diplomorpha complex all derive directly from the Kuntz ecological notes. Of particular value in the Kuntz material were the many specimens preserved in alcohol, the shells from an alluvial deposit (ML 33), and a stream drift sample (ML 95). The Kuntz collection is now integrated into the mollusk collection of the Uni- versity of Michigan Museum of Zoology, with a duplicate set in Chicago Natural History Museum and paratypes of new species distributed to several other institutions. Besides the non-marine shells, Kuntz collected 331 lots of marine and about 150 lots of estuarine Mollusca. During the summer of 1954, six museums in the eastern United States were visited and their mollusk collections searched for New Hebridean material. The relevant specimens were borrowed and studied over the period of the next two years. The institutions visited and the number of lots from each are: Carnegie Museum, Pittsburgh (43 lots) ; American Museum of Natural History (29 lots) ; Museum of Comparative Zoology, Harvard (53 lots); Academy of Natural Sciences of Philadelphia (51 lots); United States National Museum (29 lots); and Chicago Natural History Museum (12 lots). The Bryant Walker collection at the University of Michigan added 104 lots of New Hebridean shells. Most of the lots were from famous shell collections and many were types or from the type lots. Some previously unstudied material was located, however. Throughout the text, the following abbreviations indicate the depository of the specimens referred to : AMNH American Museum of Natural History ANSP Academy of Natural Sciences of Philadelphia BM British Museum (Natural History) BPBM Bernice P. Bishop Museum, Honolulu CM Carnegie Museum, Pittsburgh CNHM Chicago Natural History Museum DMNZ Dominion Museum, Wellington MCZ Museum of Comparative Zoology, Harvard Miller Walter B. Miller, Falls Church, Virginia ML, NH Kuntz station numbers (see Appendix I) UMMZ University of Michigan Museum of Zoology USNM United States National Museum SOLEM: MOLLUSCA OF THE NEW HEBRIDES 15 Geography of the New Hebrides The New Hebrides (pi. 1) comprise approximately eighty islands lying between 12° and 20° S. Lat. and 165° and 170° E. Long. They extend about 550 miles northwest to southeast, from the northern- most island, Hiu in the Torres Group (165° 40' E., 13° 10' S.), to the southernmost, Aneiteum (169° 51' E., 20° 15' S.). They are about 1,100 miles east of, and roughly parallel to, the coast of northern Queensland. The Santa Cruz (Queen Charlotte) Islands lie 100 miles north of the Torres, and the main island, Santa Cruz (Ndeni, Nitendi), is about 240 miles east of San Cristoval in the Solomons; the Loyalty Islands are about 120 miles southwest of Tanna; New Caledonia is 70 miles beyond the Loyalties; the Fijis are about 400 miles east of the New Hebrides; and the southern Solomon Islands are about 300 miles northwest of the Torres. The depth of water separating the various archipelagoes is as important as geographic proximity. Detailed information on the contour of the Pacific Ocean is difficult to find, but a few general conclusions can be drawn. Sarasin (1925, pp. 5, 6) gave charts showing the 2000-meter and 3000-meter contour lines for the Melanesian-New Zealand region. The 2000-meter contour joins the New Hebrides and Santa Cruz Islands but leaves them isolated from New Caledonia and the Solomon Islands. The 3000-meter contour connects the New Hebrides to the Solomons and New Guinea but probably not to New Caledonia. New Caledonia is joined to the Louisiades and New Guinea by a separate submarine ridge through the Huon group. The possible importance of the undersea topographic features is discussed in the section on geology and referred to frequently in the zoogeographic summary (see also Riech, 1937, pp. 138-141). The New Hebrides thus form a natural geographic unit isolated by both distance and oceanic deeps from the neighboring archipel- agoes. The islands are arranged in a "Y," the tail extending from Vate to Aneiteum, one arm through the Sheppard group north to Maewo, and the other arm through Malekula to Espiritu Santo. The small Torres and Banks groups lie north of Espiritu Santo. Information about the different islands is scattered through many books and periodicals. Good general accounts are given by Speiser (1913) and Aubert de la Rue (1945). Mawson (1905), Allen (1922), and Robson (1946) describe the individual islands; bibliographies are contained in Allen (1922, pp. 540-542) and Harrisson (1937). 16 FIELDIANA: ZOOLOGY, VOLUME 43 The following list of the principal islands gives alternate names, prominent geographic features, and the names of people who col- lected shells from each. The sequence is from south to north. Aneiteum (Anatom, Annaton, Annatom, Annatam, Aneitym) lacks the raised coral reefs characteristic of most of the other islands. Geddie, Brazier, MacGil- livray, Turner, and Macmillan collected, usually near Anelgauhat Bay. Tanna (Tana) is well wooded and mountainous, with an active volcano, Yasowa (Yasur), whose sulphurous fumes produce a central "desert." Turner, MacGillivray, J. J. Walker, Robertson, and Macmillan collected near Port Reso- lution and White Sands. Aniwa (Immer, Nina) is a small coral island located 15 miles east-northeast of Tanna. Native tradition says it was formerly connected to Tanna by a land bridge (see Mawson, 1905, p. 408). Boettger (1916) lists a few mollusks collected by the Hamburg Siid-See Expedition. Futuna (Erronan, Foutouna, Table-Top) is a volcanic island, 1,930 feet high, lying 44 miles east of Tanna. Macmillan found a few juvenile Placostylus there in 1937. Erromanga (Eromanga) has an extensive savannah area and a few raised coral platforms up to 1,000 feet elevation. Dillon's Bay is the most important locality, and Turner, MacGillivray, Brazier, and Macmillan collected on the island. Vate (Efate, Esafate, Sandwich) has mountains in the northern and western portions. No mollusks have been collected on the central and southern plain, but Havannah Harbour, Undine Bay, and Vila (Fila) have been visited by Brazier, Glisson, J. J. Walker, Froggatt, and Miller. Nguna Island off the coast of Vate has a few marine shells reported from it. Epi (Api, Tasiko) is mountainous and was briefly visited by the Challenger Expedition (see E. A. Smith, 1884). Ambrym (Ambrim) has no known land mollusks, although slugs have been observed on the island (see Grimpe and Hoffmann, 1925a). Pentecost (Pentecote, Whitsuntide, Aragh Aragh) is a long, narrow volcanic island with a central ridge 2,500 feet high. No land snails have been collected. Maewo (Aurora) is similar in structure to Pentecost. Mollusks have never been collected. Omba (Oba, Aoba, Isle of Lepers) lacks raised coral reefs. F. P. Drowne collected a few Diplomorpha there in 1927. Malekula (Malicolo, Mallicolo) is less densely forested in the lowlands than is Espiritu Santo. A few shells have been picked up at Port Sandwich and the satel- lite islands, Vao and Rano. Bougainville Strait between Malekula and Espiritu Santo is less than 300 feet deep, and the two islands probably were connected during the Pleistocene glaciation. Espiritu Santo (Santo, Marina) is the largest island (1,500 square miles). There are several permanent rivers (Jordan, Sarakata, Renee) and a chain of mountains which reach 5,566 feet (Santo Peak or laiiriiri) and 6,195 feet (Mount Tabwemasana). The eastern part of the island is a coral platform elevated 300 to 600 feet. Most of the collecting has been done on the extreme southeastern por- tion of the island near Segond Channel, the Malo pass, and the satellite islands: SOLEM: MOLLUSCA OF THE NEW HEBRIDES 17 Aore (Aura), Malo (St. Bartholomew), Tutuba (Sitova), Araki, and Tangoa (see pi. 2). A few shells have been picked up at Hog Harbour, Terebiu Mission (Te- rebu) on St. Phillips and St. James Bay (Big Bay), and Cape Lisburn. J. R. Baker found slugs in the interior of the island (see HofTmann, 1929b). The most important collectors were George de Lautour (1880's), Ph. Francois (1890's), Felix Speiser (1911), and Robert E. Kuntz (1943-44). Gaua (Santa Maria, Lacona) has a large fresh-water lake (four miles in cir- cumference) in the center of the island. J. R. Baker found some mollusks in this lake and J. J. Walker collected snails from other parts of the island in 1900. Valua (Saddle, Mota-Lava) was visited by J. J. Walker in 1900. Mawson (1905, p. 424) thought that Valua, Mota, and Vanua Lava were probably once part of a single volcanic crater. Vanua Lava is a volcanic island reaching 3,120 feet in elevation. Brazier and J. J. Walker collected there. In the Torres group, Hiu (Hiw, North) and Lo (Loh, Saddle) are coral lime- stone islands with a maximum height of 1,230 feet. J. J. Walker collected on both islands. Buka-Buka (see Ancey, 1905, p. 44) could not be recognized as an alter- nate name for any of the Torres Islands. The island to which it refers is unknown to me. During the voyage of the Astrolabe, mollusks were collected from Tucopia (Ticopia) and Vanikoro in the Santa Cruz Islands (Queen Charlotte) (see Quoy and Gaimard, 1832-35). An Australian Museum expedition may have collected subsequently on Vanikoro, but no published record of non-marine material is available. Land snails from Santa Cruz (Ndeni) Island were collected by A. F. Basset-Hull, Troughton and Livingston (see Iredale, 1927), W. M. Mann, and possibly the Rev. C. E. Fox of Malaita in the Solomons. Physical Factors of the Environment Boycott (1934) considered that the local distribution of land snails was primarily determined by moisture, shelter, and the avail- ability of lime. The extreme importance of moisture is well sum- marized by Graham (1957, p. 135) who stated: "No mollusc can ever be said to have become truely terrestrial in the sense that an arthropod or vertebrate has . . . they are terrestrial, as woodlice and Peripatus are terrestrial, by avoiding the truly terrestrial conditions and living in restricted habitats of high humidity." Availability of lime is undoubtedly critical and the number of obligatory calcicole species is large (Boycott, 1934, pp. 9-12). Unfortunately we do not know which forms of lime are biologically available to snails. Triibsbach (1943, 1947) suggested that only organic lime salts, usually citrate limes, can be utilized by the animal, but his hypothesis needs confirmation. Boycott (1934, p. 4) also believed that food, per se, "has no influence either by its quality or quantity on the recurrence of our 18 FIELDIANA: ZOOLOGY, VOLUME 43 land Mollusca." Most snails feed on decaying vegetation and fungi, but some are carnivorous, and a few feed directly on living plants. The tjrpe of plant cover supplying the decaying matter is immaterial, except for the effects it has on the physical nature of the habitat, i.e., acidity, shelter, moisture retention, and percentage of organic matter. Lundgren (1954, pp. 472-473) and Burch (1956, 1957) correlated snail distribution with vegetation, but I suspect that physical rather than biotic factors are involved. In analyzing snail distributions, the micro-environment is more important than the gross conditions. Little is known of the New Hebridean macro-environment and nothing of the micro-environ- ments, but some features of the New Hebridean climate and geology correlate with land snail distribution and variation. CLIMATE The New Hebrides extend for more than 500 miles and there are considerable differences in both temperature and rainfall between the Banks and Aneiteum. Precise data are almost unavailable and only a few general comments can be offered. The micro-climates, as determined by local topography, probably dominate most "pheno- typic variation," but some New Hebridean species show variations correlated with the over-all climatic conditions. In the southern islands the climate is distinctly seasonal, with a hot, wet period from November to April and a cool, dry season from May to October. Aneiteum, Tanna, and Erromanga have a climate much like that of New Caledonia, with an annual rainfall of about fifty inches. On Vate Island the average annual temperature is about 72° F., and there is a 9.8° F. difference between the average for the coolest month and the average for the warmest. The climate at Hog Harbour, Espiritu Santo, was studied by Baker and Harrisson (1936), who concluded that it "is hot and wet from June to October, and slightly hotter and considerably wetter from November to May." Over a period of years, the annual rain- fall averaged about 120 inches, the driest month (August) averaging 5.6 inches, and the wettest (January) 11.1 inches. The average annual temperature was 79.1° F., with the hottest month (January) averaging 81.1° F., and the coolest (July) 76.8° F. The western side of Espiritu Santo, in the rain shadow of Santo Peak and Mount Tabwemasana, has less rainfall and a dryer "dry" season, but no exact data are available. The Banks Islands have over 200 inches of rain a year, but no exact data could be located. SOLEM: MOLLUSCA OF THE NEW HEBRIDES 19 This gradient in rainfall from 50 inches annually on Aneiteum to over 200 inches annually on the Banks, with the southern islands having a true "dry season" and the northern islands a continual "wet season," is reflected in morphological gradients in the shells of several New Hebridean land snail taxa (see pp. 22-23). In discussing the origin of the fauna, it is necessary to consider the direction of prevailing winds and currents. Both winds and cur- rents come from Polynesia, and even the hurricanes, about two per year (see Visher, 1925), develop near Fiji and move toward the New Hebrides and New Caledonia. GEOLOGY Mawson (1905) and Abrard (1946) provide the only important sources on New Hebridean geology. The islands are a complex mix- ture of metamorphic, andesitic volcanic, and coralline rocks. Only surface features have been observed, sometimes from the deck of a ship, and the geologic history of the region is unknown. Mawson (1905, p. 474) suggested that the arc of the New Hebrides was formed during the Hercynian revolution (Pennsylvanian to Permian), but the earliest known fossiliferous rocks are only Early Miocene (Abrard, 1946, p. 105), probably equivalent to the Aquitanien of Europe. The New Hebrides are west of the "Andesite Line" and are a region of great seismic activity. Volcanoes are found on most islands, and earthquakes are extremely frequent, causing many local changes in topography (Mawson, 1905, pp. 432-433, and Speiser, 1913). An interesting example is seen on Tanna, where a sea-level rock on which Captain Cook stood in 1774 has been subsequently elevated sixty feet (Mawson, 1905, p. 430, and Allen, 1922, p. 157). Apparently there have been substantial changes in the size of the islands. Malekula has been relatively stable since Late Miocene or Early Pliocene times, but Vate and Espiritu Santo show Miocene and Pliocene outcrops at 1,000 and 2,000 feet, respectively. The raised Miocene deposits on the New Hebridean Islands present an intriguing correlation with the sunken Miocene deposits at Bikini and substantiate the theory that Micronesia has subsided and Mela- nesia been elevated during the Tertiary (Emery, Tracey and Ladd, 1954, pp. 152-154). The extent of subsidence in Micronesia is un- known, but the Eniwetok core drillings went through 4,222 feet of coral limestone before reaching base rock. Dobrin and Perkins (1954, p. 503) suggested that a minimum of 3,000 and a maximum of 13,000 20 FIELDIANA: ZOOLOGY, VOLUME 43 feet of limestone, which was deposited in shallow water, underlie Bikini Atoll. Without more geological study, the age of the New Hebridean area and the present islands cannot be determined. New Zealand and New Caledonia have long histories of elevation followed by par- tial subsidence, and Saipan in the Marianas appeared as volcanoes in the Eocene (Cloud, Schmidt and Burke, 1956, pp. 1, 20, 98). The biological data presented below strongly suggest that dry land has persisted in the New Hebridean region for long geological periods. Systematic Methods Early in this study it became evident that there are several im- portant differences between the systematic concepts and procedures of vertebrate zoologists and those of many malacologists. Since my zoogeographic conclusions are based primarily on molluscan evidence and this depends on the systematics, a brief statement of taxonomic theory seems necessary. Some of the material is a repetition of gen- eral systematic papers of the last few years, but it serves as a back- ground for what may seem unorthodox. IDENTIFICATION AND NOMENCLATURE Involved in systematics are several often quite independent dis- ciplines: identification, nomenclature, the study of speciation, and classification. Identification need concern itself only with the iden- tity of an organism with a reference specimen, figure, or description. Nomenclature is the attempt to apply a static legal system in nam- ing dynamically changing entities. When over-zealously applied, absurdities can occur such as dating the name of a species from the time and author of a "valid emendation" of a name (see p. 143). Use of priority and definite rules for generic and specific names are necessities, but placement of family and ordinal names under pri- ority and type concepts results only in confusion and is not accepted here. H. B. Baker (1956a, b) "hastily reviewed" the retroactive effects of this ruling on the land snail family names and I follow his lead in refusing to accept the immediately indicated changes, much less those which could be dragged from the back shelves of libraries. In phyla where family level taxonomy is stabilized, application of priority might be advisable, but in the fluid state of land snail tax- onomy, it is unacceptable. SOLEM: MOLLUSCA OF THE NEW HEBRIDES 21 SPECIATION Types of species. — No fully adequate definition of a species is available, but most systematists would agree that species seem "to be definable as distinct self-perpetuating units with an objective ex- istence in nature, and therefore on a different theoretical footing from genera or families or other higher categories, which are not definable in this concrete way." (Huxley, 1940, p. 4.) In the study of species (speciation) perhaps three arbitrary levels can be recognized: the typological or nomenclatural species, the morphologic species, and the biologic species. The nomenclatural or typological species is based on the study of individuals and on the belief that species are "kinds." The logical acme of the typologists in malacology came in the last half of the nineteenth century. Reeve, the Sowerbys, Bourguignat, Locard, and others named every variation as a species. The shattering of the two species of Anodonta (a fresh-water clam) found in France into 251 "species" may serve as an extreme example (see Mayr, Linsley and Usinger, 1953, p. 86). Most Pacific Island land snail species are "nomenclatural," and prior to this study nearly all the New Hebri- dean "species" had been named by nomenclatural systematists. The "New Systematics" focused attention on populations rather than individuals and the early attempts of Sarasin and Kleinschmidt to understand geographic variation have been brought to fruition in the studies of Rensch, Huxley, Mayr, and many others. Utilizing the conclusions of geneticists in regard to populations, they have tried to delimit species through the study of variation in morphologic characters. In birds and mammals, which have few species per stu- dent, the process of descriptive morphological systematics is well advanced; in mollusks and insects, with many species per student, it has barely begun. The criteria for species remain morphological, but a biological basis is assumed, since species are groups of popula- tions that are actually or potentially capable of interbreeding under natural conditions. Careful study of many "morphologic species" has revealed the presence of "biologic races" which are reproductively isolated under natural conditions and are distinguished by behavioral or genetic rather than morphologic characters. The taxonomic treatment of such "races" is one of the great problems facing biologists. Botanists partially resolve the problem by recognizing different types of spe- cies — morphologic "Linneons" and ecologic "elementary species." Biosystematics or experimental taxonomy has been primarily de- 22 FIELDIANA: ZOOLOGY, VOLUME 43 veloped by botanists, but many of the concepts are now being utilized by zoologists. Subspecific variation. — In taxa where speciation has been thor- oughly studied, modern systematic effort is directed toward inter- pretation of variation within species. Nomenclaturally, groups of populations within species are termed subspecies and are usually based on minor morphological differences. Edwards (1954, pp. 12- 13) listed several types of subspecies, but common usage implies only geographic variation. Each subspecies occupies a definite geographic area and most specimens can be distinguished from most specimens of the adjacent subspecies. Occasionally, as in the pocket gopher, Thomomys hottae, with more than 150 "subspecies," each local pop- ulation is nomenclaturally recognized. Subspecies are often based on artificially isolated segments of uniform clines (see Wilson and Brown, 1953) and then serve no useful purpose. For emphasizing the breaks in "step-clines," or striking "insular" variations, the subspe- cies, used cautiously, can be a helpful tool. For gradual morphologic trends in species, the nomenclaturally neutral "cline" is perhaps more suitable. Study of the variation of Pacific Ocean land snails is practically synonymous with the work of Crampton (1916, 1925, 1932) on Partula. Besides his monumental efforts, only the papers of Neal (1934) on Hawaiian Helicinidae and the many papers of Powell (see References) on New Zealand Paryphanta are important. Research on snails from other areas of the world has provided certain "princi- ples of variation" which are referred to throughout this text. In the New Hebrides different aspects of land snail variation can be directly correlated with the rainfall differential, altitude, frequency of topo- graphic changes caused by earthquakes, and ecological stratification. Effects of moisture and climate: Many land snails, particularly the prosobranchs, are very sensitive to reduced moisture supply, and the occurrence of populations of dwarfed individuals in local areas with reduced rainfall is very well documented. Shells of Gonatoraphe fornicata (Pfeiffer) from Espiritu Santo are very large, while those from Aneiteum are much smaller. Shells from the Port Vila, Vate, populations of G. fornicata and Pleuropoma articulata (Pfeiffer) are smaller than those from Aneiteum, but Vila is in the rain shadow of Mount MacDonald and probably has less rainfall than Aneiteum. The difference between the shells from the three islands is only in size, not in sculpture, shape, or other characters. Adopting the "75 per cent rule," the populations, on size alone, would be sub- SOLEM: MOLLUSCA OF THE NEW HEBRIDES 23 specifically, if not specifically, "distinct." Yet the variation only reflects local physical conditions of the years the specimens were col- lected and, by itself, has no evolutionary significance. In the case of Trochomorpha bakeri Solem from Aneiteum, the xeric adaptations of heavier, thicker shell, lighter color, and more prominent sculpture are joined by the taxonomically important character of color band position. Specific separation from the Espiritu Santo T. rubens Hartman is therefore justified. The Partula and Placostylus of the southern islands show similar xeric adaptations, sometimes with sys- tematically important differences, sometimes without. Snails from the bottom of a valley will often be quite different in color, shape, and size from those found near the top of the hills. The correlation is with altitude, but the causative factor probably is moisture conditions in the micro-habitats. The slopes of a valley dry out much more quickly than the bottom, and snails near the top can be active for shorter periods of time than those at the bottom. The latter are usually bigger, more colorful, and with thinner shells. Often the height of the spire varies with altitude (see Solem, 1955). In the New Hebrides, "high altitude" specimens of Partula and Diplomorpha had higher spires than "low altitude" examples. In Placostylus, however, the reverse was true. The differences were large enough to be "subspecific" and many of the variations have been called "species" by typological systematists. Such variations reflect local conditions and I do not consider them worth nomen- clatural recognition. Effects of topography : The influence of local topography on vari- ation of land snails is extremely important. Minor geographic obsta- cles are major dispersal barriers to snails, and local morphologically distinct populations, sometimes occupying a few square yards or a single tree, can develop. Examples of this are seen in the Hawaiian Achatinella, the Society Island Partula, the Cuban Urocoptis, and the Florida Ldguus. For speciation to occur, however, relatively long term maintenance of the isolating mechanisms is necessary. In many land snail taxa, geographic isolation has been followed by spe- ciation because of the snail's inability to cross filter zones. In some families, local colonies are effectively isolated from their neighbors, but relatively frequently the individual snails are mechanically dis- persed by storms, winds, etc. There is, at irregular intervals, sudden mixing of morphologically distinct breeding populations, a period of amalgamation and formation of a new local "subspecies," chance dispersal resulting in another mixing of races, and so on, indefinitely. 24 FIELDIANA: ZOOLOGY, VOLUME 43 This is apparently the normal course of events for fresh-water snails and the land snails which live near the strand line of the ocean. Passage of time results in so many gene exchanges between popula- tions, essentially in a random manner, that the classical concepts of geographic speciation are inapplicable (for example, see Mayr and Rosen, 1956, on the West Indian Cerion). In the New Hebrides, Pleuropoma articulata (Pfeiffer) and pos- sibly Omphalotropis (Oriella) setocinda (Ancey) closely correspond to Cerion in habitat and variability, while the fresh-water planorbid, Physastra layardi (Ancey), shows a range of variation which encom- passes ten of the New Caledonian "species" (see Franc, 1957, pi. 9, figs. 111-118). Placostylus shows similar variability. In New Zea- land (Powell, 1947, 1951) and New Caledonia (Pain, 1955), Placo- stylus forms local morphologic races ("subspecies"), but in the New Hebrides there seems to be only bewildering variation (see pp. 132- 135). The frequent earthquakes which change local topography provide a mechanism whereby morphologically divergent races can intermingle, hybridize, and produce extreme intra-populational vari- ability. If time permits, a stabilized morphotype can develop before the united populations are mixed with yet a third, or the variation can be increased by addition of yet another set of characters to the gene pool. I suspect the New Hebridean Placostylus will be found to have local populations with stabilized morphotypes interspersed be- tween populations showing every conceivable degree of variability. The process of separation and reunion of populations has been discussed recently by Hubbell (1956). For the union of previously isolated populations he has coined the term "phylosynapsis," and for the situation of highly variable populations in a zone between two stabilized morphotypes, the term "mosaic-discordant intergra- dation." If we strictly follow Hubbell's latter definition, the New Hebridean Placostylus does not seem to represent a zone of variabil- ity between two stable populations, but rather a single zone of vari- ation. Possibly the difference is in the factor of area size, since Hubbell's example was based on a North American grasshopper which varies only in central Florida, while the New Hebrides have no center of stable populations. The type of variation is the same, and the causative factors (earthquakes in the New Hebrides, Pleisto- cene sea-land fluctuations in Florida) are equivalent. Effects of ecological stratification : Ecological stratification is com- mon in many groups of animals, but it has not been emphasized in molluscan systematics. There are no true arboreal snails in the tern- SOLEM: MOLLUSCA OF THE NEW HEBRIDES 25 perate zones. A few species will ascend the trunks of trees, but their presence is only temporary. Only in the tropics has a rich and varied fauna of tree snails developed. The change from terrestrial to arbo- real habitat has occurred in many families and the basic modifica- tions accompanying the change were discussed by Pilsbry (1894, p. xiv). The phenomenon of sympatric ecological speciation (or sub- speciation) is virtually unknown in vertebrates, but it is quite com- mon in insects and probably also in land snails, although it has been rarely reported. Indications that this has happened or is happening in the New Hebridean land snails were seen in the Bulimulidae, Zonitidae, and Helicinidae. On Espiritu Santo, the terrestrial species Placosiylus salomonis (Pfeiffer), Trochomorpha rubens ruhens Hart- man, and Pleuropoma suhlaevigata (Pfeiffer) have their arboreal counterparts in Placostylus hicolor (Hartman), Trochomorpha ruhens convexa Hartman, and Pleuropoma albescens (Hartman). On Erro- manga there are two terrestrial and one arboreal Placostylus. Criteria for specific recognition. — From the above discussion, it follows that a species of land snail can show many shell variations. Undoubtedly many of the variations are genetic rather than pheno- typic, but the difference cannot be recognized in museum specimens. In the present study of New Hebridean snails, only the morphologic species concept could be utilized. Ecological notes with the Kuntz collection aided interpretation of the morphologic data but are most useful in indicating possibilities for field studies. In delimiting spe- cies, morphologic intergradation has been accepted as proof of specific identity, and variations in size, color, height of spire, thickness of shell, by themselves, are considered to be of little or no importance in separating species. A few "typological species" in Aneitea, Partula, and Diplomorpha are retained because insufficient material was avail- able to prove intergradation between the "species." I have recognized very few subspecies in this study. This is par- tially caused by lack of material from most of the islands, but mainly by my inclination not to use "subspecies" as a means of indicating variation obviously correlated with physical conditions of local areas. Populations on different islands are recognized as subspecies in the cases of Orpiella retardata (Cox) and Dendrotrochus eva (Pfeiffer), but only in the case of Aneitea robsoni Hoffmann are subspecies recognized on the same island. The line between subspecific and specific difference is nebulous. Lamprocystis guttula (Pfeiffer) and L. mendanae Solem are species because they are slightly more sharply differentiated than are Orpiella retardata retardata (Cox) and 0. r. 26 FIELDIANA: ZOOLOGY, VOLUME 43 depressa Solem. The variations and geographic ranges are parallel, but the taxonomic treatment is arbitrarily different. CLASSIFICATION Although there can be little doubt that certain species are more closely related than others, whether a taxon is a section, subgenus, genus, subfamily, or family is a matter of personal opinion on the part of any taxonomist. The degree of relationship of species is provable, but the relative value of a taxon in a taxonomic hierarchy is arbitrary. Meaning of genera. — Every species has two names; the second indicates its uniqueness; the first is a collective term designating relationship. As emphasized above, the species has a "real" exist- ence, but the genus is an artificial device for pigeon-holing relation- ships. Obviously no uniform agreement as to the use of the genus can be expected. Within any group of animals the genus can have the same importance and meaning, but it can have quite different meanings in different groups. Perhaps the only direct com- parison which can be made is by means of the number of species per genus and this information is presented for several taxa in Table I. If Iredale is temporarily ignored, the fact that genera of mollusks are much larger than genera of vertebrates is self-evident. In any given geographic area, the number of niches available to land snails is much greater than the number available to vertebrates. This un- doubtedly has influenced both specific and supra-specific evolution. It cannot be stated that genera of land snails are larger because of more "minor" evolution, since within the land snail genera many species groups are recognized as "subgenera" and "sections." Even with the addition of these subgeneric units. North American land snails would have 4.5 species per taxon, instead of 7.7 species per full genus. Thus, in relation to numbers of species per unit, a sec- tion of land mollusks is still larger than a genus of vertebrates, and a genus of land mollusks is much, much larger. Part of the difference reflects the state of taxonomic knowledge of land mollusks. Many more people have studied vertebrates than have studied land snails. Specific relationships are better known and usually have been nomenclaturally recognized. The ultimate result of this is seen in ornithology, where 10,000 generic names are avail- able for 8,500 species of birds (Mayr, 1942, p. 286). In vertebrate systematics this trend toward monotypic genera has been largely reversed, but in malacology it is in full swing. In the marine cowry SOLEM: MOLLUSCA OF THE NEW HEBRIDES 27 Table I. — The Genus in Modern Systematic Usage Based on the Fauna of North America North of Mexico Number of Number of Species/ Group species genera genus Authority Birds 800 400 2.0 Mayr (1942, p. 287) South Africa 865 620 1.4 Mayr (1942, p. 286) World, est 8500 2600 3.27 Mayr (1942, p. 287) Mammals 377 141 2.7 Burt and Grossen- heider (1952) Fishes» 4137 1490 2.8 Jordan, Evermann, and Clark (1930) Great Lakes 172 94 1.8 Hubbs and Lagler (1947) Mexican FW 179 102 1.8 De Buen (1947) Reptiles* 236 89 2.6 Schmidt (1953) Amphibians* 142 38 3.7 Schmidt (1953) Butterflies" 432 140 3.1 Klots (1951) Land snails 714 93 7.7 Pilsbry (1948) Japan 669 105 6.4 Kuroda (1953) South Africa 615 69 8.9 Connolly (1939) New Zealand 262 43 6.1 Powell (1946a) Australia 644 175 3.7 Iredale (1937a, b, 1938) ' Includes marine fishes from Panama and north; also Central American fresh- water fishes. * Includes Lower California. * East of Great Plains only. snails (family Cypraeidae), Hidalgo (1906-7) recognized 222 species in the Linnean genus Cypraea. After twenty years of studying shells, Schilder and Schilder (1938-39) recognized 165 species with 361 sub- species. These were placed in three subfamilies with 26 genera and 21 additional subgenera. Allan (1956) elevated subgenera and sub- jective synonyms to generic rank and placed 170 species with 420 subspecies in 13 subfamilies with 61 genera. Allan's classification, with 2.79 species per genus, compares favorably in size to vertebrate genera, but it has been unanimously criticized in reviews by con- chologists for its extreme splitting. Even the Schilders' classification, with 6.35 species per genus, is regarded by many conchologists as being based on too fine a series of generic distinctions. Generic classification in land snails. — For the past thirty years the leading students of land mollusks — Pilsbry, H. B. Baker, Watson, Connolly, Thiele, Boettger, Odhner, and CM. Cooke, to name a few — have deliberately retained large genera and indicated smaller divi- 28 FIELDIANA: ZOOLOGY, VOLUME 43 sions by subgeneric and sectional categories. The subgenera and sections provide a detailed analysis of affinities which can also be interpreted by the non-specialist. Elevation of the sections and sub- genera to genera "isolates" the species from their relatives and im- measurably increases the difficulties of the non-specialist who tries to use a systematic monograph to provide data for zoogeographic or biologic studies. The use of subgenera and sections is counter to the procedures of vertebrate zoologists, but, I believe, serves the func- tion of demonstrating relationship much better than the multiplica- tion of genera. An important exception to this view of classification is seen in the many papers of Tom Iredale (1933-45). Iredale has published valu- able checklists on the land mollusks of Australia, Papua, Lord Howe, and Norfolk Island, in which several hundred new generic names are proposed. Type species are always designated, but generic descriptions and comparisons range from inadequate to non-existent. One genus is described as "more like the former than the latter, and it is conchologically neither." (Iredale, 1933, p. 57.) Iredale con- siders that anything Australian is generically distinct and that even most families are endemic to Australia. In order to include Iredale's regions in my zoogeographic survey, drastic alterations of his classi- fications have been necessary. Iredale emphasizes differences; I em- phasize relationships. Revision of his genera is based on the study of specimens in the University of Michigan Museum of Zoology and Chicago Natural History Museum. In the time available for this project, it has been impossible to do the careful comparative studies that Iredale neglected. The conclusions as to Iredale's genera are thus presented here in as arbitrary a fashion as Iredale originally described them. Many of Iredale's genera delineate groups of spe- cies, but they greatly overemphasize the differences in extra-limital relatives. Criteria for generic recognition. — My criteria for generic recogni- tion have been primarily morphological. An ecological basis (see Inger, 1954, pp. 194-199) is not possible with island snails because of ecological speciation. Wherever possible, ecological and anatom- ical data have been utilized, but very little information on Pacific land snails is available. Thus most of the genera are based on con- chological criteria. The classical "generic" characters of spire height, shape, color, and size are often of little value in classification, but H. B. Baker (1938b, p. 6) emphasized that "positive shell-characters (e.g., definite shell sculpture, columellar folds or teeth) are very use- SOLEM: MOLLUSCA OF THE NEW HEBRIDES 29 ful in classification, but that negative ones (i.e., the convergence exhibited by smooth, featureless, or even vestigial shells) have been repeatedly produced in widely divergent groups and may mean prac- tically nothing." In the absence of definitive anatomical studies, extra-limital relationships of the New Hebridean species have been de- termined by close similarities or by obvious geographic trends in these positive shell features. The detailed classifications will undoubtedly be changed when comprehensive anatomical and ecological studies are available, but it is hoped that they closely approximate reality. Criteria for family recognition. — Supra-generic categories have been based on the criteria used in generic recognition. Families and orders are primarily evolutionary patterns rather than units with a few key characters, and minor deviations are too often allowed to over-ride major similarities. Discontinuous southern distributions are extremely common in land snails, although they are uncommon in birds and mammals. Iredale has created endemic families for the Australian elements, but his actions are rejected in this study. FAUNAL REVIEW The classification followed below is adopted almost completely from the studies of Pilsbry, H. B. Baker, Watson, and Boettger, and the ordinal categories are quite different from those in Thiele (1929, 1931), Iredale (1937a et seq.), and Franc (1957). Since my classifi- cation has been taken from scattered studies, I have summarized briefly the bases utilized in recognizing orders of land snails, together with references to the pertinent literature. As background for the zoogeographic discussion, the distribution of several family groups not found in the New Hebrides is briefly summarized. While they are not directly related to the New Hebridean fauna, information about their distribution and history has bearing on the time and mode of origin of the New Hebridean land snails. Under each family I have attempted to give a brief summary of its major diagnostic characters, distribution, and probable relation- ships, and the important references in the literature. Keys have been provided to aid in the identifications below family level. No keys have been provided to introduced species, or to the Partulidae and Athoracophoridae. Complete synonymies are given for all species native to the New Hebrides, but only one or two key references for introduced or very wide-ranging forms. Following the literature citations are (1) local 30 FIELDIANA: ZOOLOGY, VOLUME 43 and extra-limital distribution, (2) a list of material examined, and (3) remarks on the ecology, anatomy, variation, and typology of the species. Detailed descriptions of species have been omitted, since this is not primarily a study in speciation. Data concerning the variation in many species have been pre- sented in tabular form. All measurements are in millimeters; whorl counts were made to the nearest eighth. Shells more than 5 mm. in maximum size were measured with a vernier caliper to the nearest 0.1 mm.; specimens under 5 mm. were measured by a calibrated Whipple counting disc under 28 X magnification. Class GASTROPODA Cuvier, 1798 Of the five molluscan classes, only the Gastropoda have invaded the terrestrial habitat. Several different branches have made the transition from water to land, but generally the Prosobranchia are marine (with some land and fresh-water forms), the Opisthobranchia exclusively marine (except for the very questionable Lake Baikal Ancylodoris) , and the Pulmonata land and fresh-water (with a few marine taxa). The basic classification of the gastropods was estab- lished in the last century, but in the past fifteen years a series of papers have appeared which provide much new information on which a reclassification can be based. No synthesis is as yet possible, but many important ideas have been offered by the studies of C. M. Yonge on functional anatomy, Fretter (1943) on Oncidiella, Huben- dick (1945) on the Basommatophora, Pilsbry (1948) on the land mollusks, Graham (1949) on the molluscan stomach, Knight (1952) on primitive fossil gastropods, Boettger (1954) on the "euthyneura," Morton (1955) on the Ellobiidae, and H. B. Baker (1955) on the Pulmonata The land prosobranchs represent only a small fraction of the spe- cies found in that subclass, and comments in this study are restricted to subfamilial classification. The Pulmonata, however, are consid- ered in detail, since all of the major groups are found in the New Hebrides or adjacent areas. The division of the Pulmonata into Stylommatophora (land snails) and Basommatophora (fresh-water snails) recognizes basic adaptive radiations and is supported by morphological and physiological characters. Surprisingly, there is no single adequate discussion of the adaptive differences, although much information is contained in Boycott (1934, 1936) and Hopwood (1944, 1945). SOLEM: MOLLUSCA OF THE NEW HEBRIDES 31 Pilsbry (1948) established a third superorder, the Systellommato- phora/ for three famiHes usually included with the Stylommatophora: the Oncidiidae, Veronicellidae, and Rathouisiidae. These families dif- fer from the other pulmonates in lacking any shell and mantle cavity, and in having the lung, breathing pore, excretory opening, and anus at the posterior end. The eyes are at the tips of the peduncles, as in the Stylommatophora, but there are separate gonopores, as in the Basommatophora. Ecologically, the three families have barely left the sea, since the Oncidiidae are primarily inter-tidal and the Veronicellidae and Rathouisiidae are restricted to very humid trop- ical areas. The Systellommatophora seem to be a natural group, but there is controversy as to its place in the Gastropoda. Some authorities — Plate (1893), Colosi (1921), and Fretter (1943)— believe that they are opisthobranchs, while others — Hoffmann (1925, 1928, 1929a), Watson (1925), and Pilsbry (1948) — consider them pulmonates. Homologies of structure are still uncertain (see Odhner, 1917), and comparative studies on the anatomy, embryology, and life histories are needed. The Oncidiidae are the most primitive and show the closest affinities to the opisthobranchs. Fretter (1943) summarized the reasons for removing them from the Pulmonata, but neglected their similarity to the Veronicellidae and Rathouisiidae, which show distinctly pulmonate characteristics (see Hoffmann, 1925, pp. 326- 338). Probably the three families are more closely related to each other than to either the main pulmonate or opisthobranch lines. Division of the Systellommatophora into two orders, the Oncidi- acea (Oncidiidae) and the Soleolifera (Veronicellidae and Rathouisi- idae), recognizes the important habitat difference (inter-tidal vs. land) and several structural adaptations probably correlated with the environmental conditions. The Basommatophora were surveyed by Hubendick (1945), and his classification has been retained. It is generally believed that the Basommatophora are more primitive than the Stylommatophora and that the latter were derived from an "Urtyp" corresponding rather closely to some of the more primitive Ellobiidae. The Ellobiidae are the only land Basommatophora (see Morton, 1955) and they present many characters from which the basic stylommatophoran structures can be derived. Since the only New Hebridean Basommatophora • The name Systellommatophora is equivalent to the Ditremata, Teletremata, Clasthurethra, Digonopora, and Gymnophila of authors. Strict priority would require use of Fischer's name, Ditremata, but since Pilsbry's name is in etymologi- cal conformity with the names of the other two superorders, it is here retained. 32 FIELDIANA: ZOOLOGY, VOLUME 43 considered here are planorbid snails, no comments have been made on the supra-familial classification of the Basommatophora proposed by Hubendick (1945) and Morton (1955). The Stylommatophora contain about 88 per cent of the New Hebridean Pulmonata. No system for classifying the Stylomma- tophora has gained universal acceptance. Those of Pilsbry (1900a, 1948), Pelseneer (1906), and Thiele (1931) are most widely used today; the older systems of Morch, Semper, Fischer, and others have only historical interest. Pelseneer (1906) established four divisions: the Ditremata (=Sys- tellommatophora), Elasmognatha (=Heterurethra and Tracheopul- monata), Agnatha (the carnivorous snails), and Holognatha (remaining families). This system was based on the number of gonopores and the structure of the jaw. It failed to provide any concept of relation- ships between the many families of the Holognatha and overempha- sized the phylogenetic importance of the carnivorous habit. Watson (1915) showed that carnivorous snails have developed in several dif- ferent families and that the characters of Pelseneer's Agnatha are the result of convergent evolution. Thiele (1931) divided the pulmonates into a number of "Stirpes" or superfamilies, but his classification gave no indication of any rela- tionships between them. As part of the only general classification of the Mollusca to appear since Pelseneer (1906), undoubtedly Thiele's system will be widely used in regional studies. It has already been partially adopted by Iredale (1937a, 1938, 1941) for Australia and Papua, by van Benthem Jutting (1948, 1950, 1952, 1953a, 1956) for Java, and by Franc (1957) for New Caledonia. Despite many excel- lent innovations, much of Thiele's classification of the Stylomma- tophora is faulty and it has been criticized by several malacologists (Pilsbry and Cooke, 1934a; H. B. Baker, 1938b, 1941; and Pilsbry, 1948). Several of Thiele's "Stirpes," i.e., Zonitacea, Ariophantacea, are clearly heterogenous. Much of Thiele's classification below fam- ily level has been utilized here, but his higher categories have not been accepted. The most comprehensive classification of the Stylommatophora has been developed by Pilsbry (1900a, 1919, 1948) and H. B. Baker (1955, 1956a, b). Based on the structure of the pallial region, five orders — Heterurethra, Tracheopulmonata, Orthurethra, Mesurethra, and Sigmurethra — have been recognized. Originally Pilsbry sepa- rated the Tracheopulmonata from the typical Stylommatophora because of the peculiar structure of the lung in this group. Baker SOLEM: MOLLUSCA OF THE NEW HEBRIDES 33 (1955) suggested that the Tracheopulmonata are closely related to the Heterurethra, but the subject needs more investigation. Use of the pallial organs as primary criteria for ordinal recognition was criti- cized by Simroth and Hoffmann (1908-28, pp. 408, 1223-24), Wachtler (1934), and Thiele (1935, p. 1064). Their objections were based on variations in the position of the ureter, which seemed to form a series of transitions between the ordinal categories; in one case these variations occurred in a single family (see Wachtler, 1934, for a bibliography). Without more study, it is impossible to deter- mine the importance of the pallial regions in pulmonate classification. I consider that the cited variations will probably be found to be sec- ondary modifications of the basic plan, and thus will not affect the usefulness of the criteria in recognizing the ordinal categories. Most American authors, and three European malacologists — Watson (1915, 1920), Steenberg (1925), and Boettger (1952, 1954)— have adopted Pilsbry's classification. The "key character" used in separating the various orders is the relationship of the ureter to the kidney and other pallial organs. Briefly, the five orders are characterized as follows: Heterurethra: Ureter along front margin of kidney, then following rectum to pneumostomal opening (fig. 1, a). Tracheopulmonata: Ureter multi-looped, opening into respiratory pore or directly to exterior (fig. 1,6). Orthurethra: Ureter remote from hindgut, running from tapered anterior end of kidney to anterior border of lung (fig. l,c). Mesurethra: Ureter represented mainly by lateral opening of the relatively short kidney (fig. l,d). Sigmurethra: Ureter abruptly refiexed, passing to posterior end of lung cavity, then turning across to the last fold of the gut and following it forward to the mantle edge (fig. l,e). During the evolutionary change from a shell-bearing snail to a "Halbnacktschnecke" or a slug, the pallial region may become greatly altered in position, size, and organization. Thus, the many slug taxa do not seem to follow the above criteria, but must be related to ordinal categories through their affinities with shell-bearing snails, rather than on their own characteristics. When the nearest generic relatives are unknown, great confusion can result. For example, the genus Aillya is known from the Cameroons and Fernando Po (fide Watson). Odhner (1927) described the anatomy of Aillya and placed it in the Amphibuliminae (family Bulimulidae) on the basis of the radula, jaw, anal pouch, and stomach. He noted, however, that the structures of the nephridium, genitalia, intestine, and spermathecal Fig. 1. Pallial organs of stylommatophoran snails: a, Heterurethra; h, Tracheo- pulmonata; c, Orthurethra; d, Mesurethra; e, Sigmurethra (after various sources). Kidney (K), ureter (U), renal orifice (KO), hindgut (HG), principal pulmonary vein (HV), heart and pericardium (H). 34 SOLEM: MOLLUSCA OF THE NEW HEBRIDES 35 stalk are similar to those of the Succineidae, while the ureter is par- tially transitional between the types found in the Succineidae and Athoracophoridae. Pilsbry (1946b) thought that Aillya might be an aberrant zonitid snail (Sigmurethra, Aulacopoda), and H. B. Baker (1955, p. 110) placed it in the Heterurethra. Similarly the genus Ceciliodes was of uncertain position before Watson's study (1928). Such examples of the difficulty in classifying particular genera have been used by many conchologists (students of shells) as an excuse for ignoring the importance of the soft parts in systematics and for rejecting the conclusions of malacologists (students of shells and soft parts). Little can be said about the relationships between the ordinal categories. The pallial region of the Basommatophora is most similar to that of the Orthurethra (fig. 1, c), and Pilsbry (1900a, et seq.) considered the Orthurethra to be the most primitive order of the Stylommatophora. The Heterurethra (Succineidae or amber- snails) and Tracheopulmonata (the slugs of the family Athoraco- phoridae) have quite different pallial regions (fig. 1, a, b) but the importance of this difference is unknown. The multi-looped kidney and "tracheal" lung of the Athoracophoridae are distinctive, but somewhat similar variations have been produced in other slug taxa. H. B. Baker (1955, p. 110) pointed out parallelisms in both the Veronicellidae and Ancylidae and placed both the Aillyidae and Athoracophoridae in the Heterurethra. Without more study of the taxa involved, including the aberrant succineids Omalonyx and Hya- limax, I prefer to retain the Heterurethra and Tracheopulmonata as separate categories. If their union proves justified, the name Elas- mognatha Morch should be applied to the combined category, since it was originally proposed to cover both the Succineidae and Athora- cophoridae. Baker (1955) suggested that the Sigmurethra were derived from a heterurethrous and aulacopod ancestor, rather than from an orthur- ethran, but further study is needed. The Mesurethra represent a side branch from the main stream of pulmonate evolution and prob- ably had no part in the ancestry of the dominant Sigmurethra. No members of the Mesurethra are found in the New Hebrides, but the order is zoogeographically important and a brief synopsis of its dis- tribution and content is given (pp. 42-43). Within the Sigmurethra two main lines of development have been recognized by establishing the suborders Aulacopoda and Holopoda. Proposed by Pilsbry (1896), these suborders are based on the rela- 36 FIELDIANA: ZOOLOGY, VOLUME 43 tive position and prominence of the pedal grooves (see fig. 2). In the Aulacopoda they are conspicuously impressed and situated well above the margin of the foot; thus a part of the sole of the foot forms the lower portion of the side of the animal (fig. 2, a). In the Holo- FlG. 2. Pedal grooves in the Aulacopoda and Holopoda: a, aulacopod type (from Hemiplecta neptunus Pfeiffer; CNHM 33041); b, holopod type (from Zachry- sia guanensis Poey; CNHM 53481). poda the grooves are rather inconspicuous and are at or close to the angle of the foot (fig. 2, 6). Wachtler (1935) reviewed the morphol- ogy of the gastropod foot and concluded that the position of the pedal grooves was characteristic, at least for families. He was uncertain as to the value of the Aulacopoda and Holopoda as primary divisions but perhaps was confused by the variation in other orders, since he had previously (1934) rejected the Heterurethra-Orthurethra-Sig- murethra concept of Pilsbry (1900a). Pilsbry (1946b, pp. 231-232) admitted that in a few cases the distinction between the Aulacopoda and Holopoda becomes blurred, but he considered the separation valid. The Aulacopoda are the more primitive and are more impor- tant in the faunas of the southern hemisphere than in those of the northern. By far the largest number of species of land snails belong to the Holopoda. Subclass PULMONATA Cuvier, 1817 Superorder SYSTELLOMMATOPHORA Pilsbry, 1948 Pulmonata with eyes on contractile (Soleolifera) or inversible (Oncidiacea) stalks. Body oval or lengthened, slug-like. Convex or keeled dorsal integument without mantle cavity or shell, extending down over head on all sides. Lung pos- terior, breathing pore, anus, and nephridial pore behind foot. Male genital opening at right side of head ; female opening midway on right side of hyponotum or near anus. (Modified from Pilsbry, 1948, p. 1062.) SOLEM: MOLLUSCA OF THE NEW HEBRIDES 37 Order ONCIDIACEA Thiele, 1931 Shell-less animals with one pair of tentacles bearing terminal eyes. Mouth surrounded by large sensory lobes. Radular teeth unicuspid. Respiration by lung and branchial gills. Male gonopore at right side of head, either inside or outside right tentacle; female gonopore at posterior end of body near anus. Family ONCIDIIDAE The studies of Semper (1885), Plate (1893), Watson (1925), Hoffmann (1928, 1929a), H. B. Baker (1938a), Fretter (1943), and Awati and Karandikar (1948) summarize present knowledge of the family and provide a guide to the vast literature. Oncidiids are found in all oceans from 60° N. to 55° S. but are most numerous in the oceans of the triangle formed by the Andamans, the Philippines, and New Caledonia. Most Oncidiidae live in the tidal zone. A few are terrestrial (Fretter, 1943, p. 716, and Hoffmann, 1932, p. 134), while others inhabit brackish water (Plate, 1893). Oncidiids have pelagic larvae and are distributed like marine organisms, but are included here because of their pulmonate affinities. The only New Hebridean Oncidiidae available to me were found in a marine tidal zone on Espiritu Santo by G. S. Banner in 1943. Except for a record by Bretnall (1919), they are the only oncidiids known from the New Hebrides. Of the six genera of Oncidiidae, Oncis and Oncidium are known from the islands, while Oncidina probably will be found there. The classification of the Oncidiidae is based on the internal anatomy, but the New Hebridean species can be identified from external characters (see pi. 3, fig. 2). The following key includes Oncidina australis Semper, as well as the species actually examined. Key to the New Hebridean Oncidiidae 1. Dorsal eyes present, either singly or in groups; hyponota present 2 Dorsal eyes absent; no hyponota; no branchial plumes; sole and head reddish- gray Oncidina australis Semper 2. Hyponota narrower than sole of foot; dorsal eyes arranged in groups; branchial plumes present; size large 3 Hyponota wider than sole of foot; dorsal eyes single; branchial plumes absent; size small Oncis martensi Plate ' If type concepts and priority are applied to ordinal categories, earlier names can be utilized. Since Thiele was the first to use a name of more than family rank for the Oncidiidae only, his name has here been accepted. 38 FIELDIANA: ZOOLOGY, VOLUME 43 3. Branchial plumes distributed over entire notum; respiratory pore one-fourth to one-third width of hyponotum from anus, off median line; hyponota equal to greatest width of foot sole Oncidium peronii Cuvier Branchial plumes on posterior fourth of notum only; respiratory pore one-half to one-third width of hyponotum from anus, on median line; hyponota nar- rower than greatest width of sole Oncidium verruculatum Cuvier Genus ONCIDINA Semper, 1885 Type species.— Onchidina australis Semper. Oncidina australis Semper Onchidina australis Semper, 1885, Reisen im Philippinen, 3, (7), p. 287, pi. 19, figs. 11, 14, 15, pi. 21, fig. 27, pi. 23, fig. 10; Bretnall, 1919, Rec. Australian Mus., 12: 325. Oncidina australis (Semper) Plate, 1893, Zool. Jahr., Anat., 17: 208, pi. 7, fig. 13, pi. 8, figs. 24, 35, pi. 10, fig. 51; Hoffmann, 1928, Zool. Jahr., Syst., 55: 102. Range. — Brisbane, Australia, New Caledonia, and Fiji. Material. — No material available. Remarks. — The range of this monotypic genus is such that it may be found in the New Hebrides. The absence of hyponota and dorsal eyes at once separate Oncidina from the other New Hebridean On- cidiidae. Genus ONCIS Plate, 1893 Type species. — Oncidina coriacea Semper. Remarks. — The thirteen species of Oncis range throughout much of the Indo-Australian region (Hoffmann, 1929a, p. 254). Although previously known from New Britain and Queensland, this is the first record from east of the 160th meridian. Oncis martensi Plate Oncis martensi Plate, 1893, Zool. Jahr., Anat., 7: 196-197, pi. 7, fig. 7 — Singa- pore (error); Martens, 1897, in Weber's Zool. Ergeb. Ost-Indien, 4: 128 — Petshaburi, Gulf of Siam (correction of error); Hoffmann, 1928, Zool. Jahr., Syst., 55: 89. Range. — Espiritu Santo, Gulf of Siam, India. Material— Espiritu Santo (AMNH, G. S. Banner!). Remarks. — The single juvenile specimen is most similar to 0. mar- tensi Plate and, pending study of adult individuals, is referred to that species. The specimen is only 7 mm. long and has six dorsal eyes, in comparison with the type, which is 66 mm. and has ninety dorsal SOLEM: MOLLUSCA OF THE NEW HEBRIDES 39 eyes. The unspecified Indian record is taken from Awati and Karan- dikar (1948). Genus ONCIDIUM Buchanan, 1800 Type species. — Oncidium typhae Buchanan. Oncidium peronii Cuvier Onchidium peronii Cuvier, 1805, Ann. Mus. Paris, 5: 37; Bretnall, 1919, Rec, Australian Mus., 12: 311 — Lord Howe, Santa Cruz Islands. Onchidium tonganum Quoy and Gaimard, 1832, Voy. Astrolabe, 2: 210, pi. 15, figs. 17-18— Tonga Islands. Ovcidium peroni (Cuvier) Hoffmann, 1928, Zool. Jahr., Syst., 55: 71-72. Range. — Lord Howe, Santa Cruz Islands. From Red Sea and Mauritius to New Caledonia, Samoa, and the Marshall Islands. Material. — No material available. Remarks. — Oncidium peronii is the common species of Polynesia and since it has been found on all sides of the New Hebrides it un- doubtedly lives there. 0. peronii and 0. verruculatum have often been confused. Numerous differences exist between the two (see Bretnall, 1919, and Hoffmann, 1928), the most obvious of which are given in the key. Oncidium verruculatum Cuvier Onchidium verruculatum Cuvier, 1830, Reg. Anim., 2nd ed., 3: 46; Bretnall, 1919, Rec. Australian Mus., 12: 309-310. Oncidium verruculatum (Cuvier) Hoffmann, 1928, Zool. Jahr., Syst., 55: 72-74; Awati and Karandikar, 1948, Univ. Bombay, Zool. Mem., no. 1, 52 pp., 62 figs. Range. — Espiritu Santo, Africa to Japan, New Guinea, Australia, and New Caledonia. Also common in Hawaii, but not known from Polynesia. Afo^maZ.— Espiritu Santo (AMNH, G. S. Banner). Remarks. — Branchial plumes are developed in the one adult and smallest juvenile but are lacking in the other five specimens. All have groups of dorsal eyes, fourteen groups of two to six eyes in the adult and three to eight groups of two to four eyes each in the juve- nile specimens. Awati and Karandikar (1948) give an excellent sum- mary of our knowledge of this species. Their findings contradict the observations of Fretter (1943) on Oncidiella and suggest pulmonate rather than opisthobranchiate affinities for the Oncidiidae. Further study is needed to resolve the differences. 40 FIELDIANA: ZOOLOGY, VOLUME 43 Order SOLEOLIFERA Simroth, 1890 Slug-like mollusks without external or internal shell. Head bearing two pairs of tentacles, the upper ommatophores, the lower tactile. Radular teeth unicuspid. Male genital opening at right side of head, female opening in pedal groove or hyponotum about midway along right side of body. The Veronicellidae are herbivorous, the Rathouisiidae carnivo- rous, and the characters separating the two famihes which comprise the Soleolifera primarily reflect this dietary distinction. In the Rathouisiidae the jaw is lacking; the female gonopore is in the pedal groove; the genital organs lack accessory glands; the radular teeth are slender and dagger-shaped; and, in the Austro-Melanesian spe- cies, the notum is sharply rounded or keeled (see Laidlaw, 1940) . Rathouisiids (fig. 17) are known from southeast China, Tonkin, Siam, Burma, Malaya, Indonesia, the Philippines, New Guinea, the Bismarcks, and northern Queensland (see Hoffmann, 1925). Prob- ably both rathouisiids and veronicellids will be found in the Solo- mon Islands, although no slugs have been reported from the islands up to the present time. Family VERONICELLIDAE (=Vaginulidae) Terrestrial slugs without internal or external shell. Rounded dorsal surface covered by mantle (or notum) and separated from ventral surface by a sharp keel, the perinotum. Ventral surface with central sole, separated from the lateral hy- ponota (which are ventral continuations of the notum) by a pedal groove. Head with four tentacles, the upper pair having terminal eyes. Head and tentacles of contracted specimens hidden beneath notum. Female gonopore about midway in right hyponotum. Male gonopore at right side in groove between mouth and foot. Jaw low, arcuate, formed of narrow vertical plates. Radula with narrow central and broad lateral teeth, all unicuspid. (Adapted from Pilsbry, 1948, p. 1062 and van Benthem Jutting, 1952, pp. 327-328.) Hoffmann (1925) discussed the systematics, anatomy, zoogeog- raphy, and phylogeny of the Veronicellidae. Unfortunately his nomenclature is incorrect, and H. B. Baker (1925a, 1931) should be consulted in conjunction with this otherwise excellent contribution. The Veronicellidae have a circumtropical distribution which closely approximates that of the palm trees except in North Africa and the Middle East, where no veronicellids are found. The slugs live under stones, grass tufts, or decaying wood and feed on living or decaying plants. Occasionally they assume economic importance through the destruction of crops. Collinge (1900), Grimpe and Hoffmann (1925a, b) and Hoffmann (1929b) studied New Hebridean veronicellids. Two genera, each SOLEM: MOLLUSCA OF THE NEW HEBRIDES 41 represented by a single species, have been reported from tiie islands. Both the South American Angustipes and the Indonesian Eleuthero- caulus were probably introduced. I have seen no specimens and the information presented below has been abstracted from the literature. Material from Vate Island is particularly desirable to determine if V. brunnea Collinge is a synonym of A. pleheius (Fischer) as reported by Grimpe and Hoffmann (1925a). Genus ANGUSTIPES Colosi, 1921 (= Belocaulus and Sarasinula of Grimpe and Hoffmann, 1925a) The South American Angustipes differs from the African Imerinia in lacking a doubled spermatheca (H. B. Baker, 1931, p. 136). An- gustipes s.s. (= Belocaulus) usually has the greatest width of the verge below the middle and the verge is without an "S"-shaped curve. Section SARASINULA Grimpe and Hoffmann, 1925 Type species. — Vaginula grandidieri Crosse and Fischer. Angustipes (Sarasinula) pleheius (Fischer). Plate 3, fig. 4. Vaginula pleheius Fischer, 1868, Jour, de Conch., 16: 145-146 — Noumea, New Caledonia. Vaginula brunnea Collinge, 1900, Willey's Zool. Results, 4: 435, pi. 41, figs. 18-23— Vate, New Hebrides. Vaginula hedleyi (Simroth) Collinge, 1900, op. cit., p. 435 — Vate. Sarasinula plebeja (Fischer) Grimpe and Hoflfmann, 1925, Nova Caledonia, 3, (10), pp. 357-362, 365-366, pi. 6, figs. 1-3— Malo, New Hebrides. Imerinia plebeja (Fischer) HofTmann, 1929, Zool. Anz., 84, (5-6), p. 117 — Espiritu Santo. Angustipes (Sarasinula) plebeius (Fischer) H. B. Baker, 1931, Nautilus, 44, (4), pp. 134-136. Range. — Vate, Malo, and Espiritu Santo. Also known from New Caledonia, Loyalty Islands, Queensland, Fiji, Samoa, Tahiti, Mas- carene Islands, Trinidad, Antigua, St. Thomas, Rio de Janeiro and Bahia, Brazil. Material. — No material available. Remarks. — Hoffmann (1927, p. 35) united A. plebeius with the Brazilian- An tillean A. dubia (Semper). Both Hoffmann and Baker attest to the pan-tropical distribution of this species. Baker consid- ered that it was introduced into Polynesia during historical times. 42 FIELDIANA: ZOOLOGY, VOLUME 43 Genus ELEUTHEROGAULUS Simroth, 1913 (=Laevicaulus Simroth, 1913, and Meisenheimeria Grimpe and Hoffmann, 1925) Type species. — Vaginula comorensis Fischer (=V. alte Ferussac). Eleutherocaulus alte (Ferussac). Plate 3, fig. 3. Vaginula alte Ferussac, 1823, Hist. nat. Moll. terr. fluv., 2: 96 — Pondicherry. Veronicella leydigi (Simroth) Collinge, 1900, Willey's Zool. Results, 4: 435— Vate, New Hebrides. Meisenheimeria alte (Ferussac) Grimpe and Hoffmann, 1925, Nova Caledonia, 3, (10), pp. 362-367, pi. 4, figs. 4-6. Laevicaulus alte (Ferussac) Hoffmann, 1929, Zool. Anz., 84, (5-6), p. 117; van Benthem Jutting, 1952, Treubia, 21, (2), p. 330. Range. — Vate(?). East Africa to the Philippines and Tenimber Islands, Queensland, New Caledonia, and the Loyalty Islands. Material. — No material available. Remarks. — Collinge's record from Vate Island needs confirmation, but it is probably correct. Superorder STYLOMMATOPHORA Schmidt, 1855 Snails with eyes on tips of a pair of superior invaginable tentacles and with a smaller inferior pair of feelers which are rarely (Athoracophoridae and some Pupil- lidae) lacking. Animal hermaphroditic, with a common external gonopore (rarely narrowly separated) opening near the right tentacle. Shell external, varjdng from large (helices) to vestigial (Testacella) , or internal and consisting of a calcareous plate or series of fragments (except in the Philomycidae where the shell is com- pletely absent). Pallial cavity without ctenidia and respiration primarily through a more or less heavily veined lung. All are terrestrial. Of the five orders of the Stylommatophora, only the Mesurethra is not found in the New Hebrides. The Heterurethra and Tracheo- pulmonata are represented by only one or two species and the Orthu- rethra by probably not more than ten; the great majority of the species belong to the Sigmurethra. The Heterurethra and Tracheo- pulmonata never form a large element in a fauna, and only on Rapa and the Hawaiian Islands do the Orthurethra become dominant. In all other regions of the world, various sigmurethrous taxa have the greatest numbers of species. The order Mesurethra was created by H. B. Baker (1955) for the relict land snail families Cerionidae, Corillidae, and Acavacea. The Cerionidae are West Indian and live only on the portion of the sea shore that is above high tide but is reached by salt spray. The SOLEM: MOLLUSCA OF THE NEW HEBRIDES 43 Corillidae are found in India, China, southeast Asia, and Ceylon, with a few representatives in South Africa. The Acavacea include genera from South America, South Africa, Madagascar, the Seychelles, Ceylon, Tasmania, New South Wales, and Queensland. The shells are very different in appearance, but anatomical structures indicate their close relationship. The conchological dissimilarity led Iredale (1937b, pp. 14-19) to create four families and seven genera for the thirteen Australian species. Watson (in Connolly, 1915), Germain (1925) and Boettger (1936) reviewed the phylogeny of the Acavacea. They recognized four subfamilies (or families) : the Strophocheilinae, Acavinae, Dorcasiinae, and Caryodinae. The Strophocheilinae of South America are the most primitive, the Acavinae of Ceylon, the Seychelles, and Madagascar the most advanced. The Dorcasiinae from South Africa and the Caryodinae from Australia and Tasmania are intermediate in the possession of primitive characters. In each area the primitive genera are more southern in distribution, and the more advanced are northern. The distribution has been interpreted by Watson and Germain to indicate an origin for the Acavacea in the "Gondwana" area. Order TRACHEOPULMONATA Simroth, 1890 Animal without external shell. Integument with furrows outlining a triangu- lar head shield and a usually triangular mantle on right side of animal. Head with two eye-bearing invaginable tentacles. Lung with trachea-like air spaces. Kid- ney with highly convoluted ureter. Jaw elasmognathous. Radula with central tooth well developed, vestigial, or lacking. One or more calcareous bodies present beneath mantle shield. (Adapted from Thiele, 1931, p. 494.) The highly unusual pallial system of the Tracheopulmonata has been carefully studied by Plate (1897), W. Pfeiffer (1900), and Gla- mann (1903). Although they worked with different genera, their conclusions agree except in minor details. Plate (1897) compared the lung structure of Athoracophorus to the tracheal system of insects, but most probably it is only a highly specialized "Gefasslunge" (see Simroth, 1918). Equally interesting is the multi-folded ureter (see fig. 1, b) which forms one of the primary reasons for regarding the Tracheopulmonata as a separate order. The single family, the Athoracophoridae (fig. 18), is restricted to New Guinea, Queensland, New South Wales, the Bismarcks, the Admiralty Islands, the New Hebrides, New Caledonia, and New Zealand. Probably species will be found in the Solomon Islands. 44 FIELDIANA: ZOOLOGY, VOLUME 43 Family ATHORACOPHORIDAE (= Janellidae) Many generic names have been proposed for athoracophorids, but I am recognizing only four genera. Simroth (1918) described two genera, Ottonita Strand, 1932 (=Ottonia Simroth, 1920, non Gistl, 1848), and Neomecklenburgia, from the Bismarcks, but their status is uncertain. Grimpe and Hoffmann (1925a) were unable to determine their validity, and until new material becomes available they are better left as dubious taxa. Athoracophorus Gould (fig. 21, B) is restricted to New Zealand and the neighboring islands (see Suter, 1913). The rachidian tooth is large, with four to seven denticles, and the lateral teeth normally have six to seven cusps. The notum is deeply furrowed and there is much variation in the shape and relative positions of the mantle and cephalic shields. All species of Athoracophorus have the notum sep- arated from the foot sole by a narrow hyponotum. Powell (1946a) placed the New Zealand species in four genera. The differences be- tween these "genera" are much smaller than those between the New Zealand species and the other genera. A better understanding of the importance of the radiation in New Zealand is obtained by reducing Powell's genera to subgenera and sections of Athoracophorus. Aneitea Gray (fig. 21, C) is found in New Caledonia and the New Hebrides but has not been found in the Loyalty Islands. Specimens have evidently been introduced into the botanical gardens at Bris- bane, Australia (Aneitea hrishanensis W. Pfeiffer, 1900). The cen- tral tooth is greatly reduced or absent and the lateral teeth have four cusps. The notum is deeply furrowed and the mantle and cephalic shields are clearly outlined and in contact with one another. Tri- boniophorus is similar in external appearance, but there are radular and anatomical differences. Triboniophorus Humbert (fig. 21, A) is reported from Queens- land, New South Wales, New Guinea, New Caledonia, and the New Hebrides, but the last two records are questionable. Aneitella Cockerell (fig. 21, D) has been found on New Britain and the Admiralty Islands. The rachidian tooth is absent and the first two lateral teeth are partially fused. The notum is unfurrowed and the mantle and cephalic shields are only weakly outlined. For the Athoracophoridae Grimpe and Hoffmann (1925a) postu- lated an Antarctic origin followed by two radiations. The first was from New Zealand to New Caledonia and the New Hebrides, the second from Australia to the Bismarck and Admiralty Islands. SOLEM: MOLLUSCA OF THE NEW HEBRIDES 45 This theory assumed that athoracophorids are absent from New Guinea and the Solomons and that Aneitella, without a rachidian tooth, is more primitive than Athoracophorus, which has a central tooth. Van Benthem Jutting (1933) reported Triboniophorus from the interior of New Guinea, and probably some species of Athoraco- phoridae will be found in the Solomon Islands. Possession of a ra- chidian tooth is a basic pulmonate character. Other factors being equal, a genus with a well-developed central tooth is more primitive than a genus which is losing or has lost the central tooth. Aneitella, without a rachidian tooth, is thus less primitive than Athoracophorus. The loss of the skin furrows in Aneitella is another indication of its advanced status. Much less difficulty is encountered if a northern origin is postu- lated for the Athoracophoridae, with New Guinea as the dispersal center from which the present genera were derived. Athoracophorus is the most primitive, Triboniophorus slightly less primitive, Aneitea specialized and Aneitella highly specialized. The most probable path of immigration would have been from New Guinea through the Solo- mons and New Hebrides(?) to New Caledonia and New Zealand. Triboniophorus was probably preserved in the highlands of New Guinea and only comparatively recently spread into eastern Aus- tralia. Whether the Athoracophoridae originated in New Guinea or spread there from southeastern Asia is unknown. Possibly ad- vanced athoracophorids will be found in the Solomons and in the lowlands of New Guinea. Two genera of athoracophorids have been reported from the New Hebrides. Glamann (1903) recorded Triboniophorus from an unknown locality, and Aneitea is an important part of the fauna. Genus TRIBONIOPHORUS Humbert, 1863 Rachidian tooth large, with three to five cusps; lateral teeth usually with four cusps. Oviducal accessory gland near the albumen gland. Kidney crescent- shaped, ureter with four or five loops (Thiele, 1931, p. 495). Type species. — Triboniophorus graeffei Humbert. Remarks. — Hedley (1889) recognized only one species, although several names have been applied to the Australian specimens. Gla- mann (1903) made a histological examination of the pallial region of Triboniophorus and at the end of his paper mentioned that the specimen was from the New Hebrides. The validity of this record is doubtful and needs confirmation. Grimpe and Hoffmann (1925a) described T. sarasini from Mount Humboldt, New Caledonia, on 46 FIELDIANA: ZOOLOGY, VOLUME 43 the basis of four specimens about 16 mm. long. I suspect that juvenile specimens of Aneitea rather than adults of Trihoniophorus were involved. Van Benthem Jutting (1933, p. 90) reported speci- mens from Dutch New Guinea collected between 4,600 feet and 7,800 feet elevation. They were at most subspecifically distinct from the east Australian T. graeffei. The great amount of variation she found in the shape of the rachidian tooth of Trihoniophorus supports my conclusions on Aneitea (see pp. 47, 50). Triboniophorus graeflfei Humbert Trihoniophorus graeffei Humbert, 1863, Mem. Soc. Phys. Nat. Geneve, 17, (1), p. 116 — Wollongong, New South Wales, Australia. Aneitea graeffei (Humbert) Glamann, 1903, Zool. Jahr., Anat., 17: 684 — New Hebrides (doubtful). Aneitea (Trihoniophorus) graeffei Humbert var. insularis Grimpe and Hoff- mann, 1925, Nova Caledonia, 3, (10), pp. 442, 448 — based entirely on the paper of Glamann (1903). Aneitea (Trihoniophorus) graeffei var. papuensis van Benthem Jutting, 1933, Nova Guinea, Zool., 17: 90-92 — Dutch New Guinea. Range. — New Guinea to southern New South Wales. Possibly introduced into the New Hebrides. Material. — No material available. Remarks. — There are no characters given in Glamann's study by which the "New Hebridean" specimen could be separated from the east Australian forms studied by others. Grimpe and Hoff- mann's var. insularis was based solely on the locality and need not be carried in the literature. If the New Hebridean locality record is correct, specimens probably were accidentally introduced from Queensland. Genus ANEITEA Gray, 1860 Mantle clearly outlined by deep furrows. Shell composed of a single piece. Radula with rudimentary or no central tooth, lateral teeth with four cusps. Oviducal accessory gland far removed from hermaphroditic duct. Kidney with four- to seven-looped ureter. (Modified from Grimpe and Hoffmann, 1925a.) Type species. — Aneitea macdonaldi Gray. Remarks. — Grimpe and Hoffmann (1925a) recognized sixteen species and varieties of Aneitea: eleven from New Caledonia, four from the New Hebrides, and one from Brisbane, Australia. Hoff- mann (1929b) described a fifth species from the New Hebrides. The Brisbane A. hrisbanensis W. Pfeiffer (1900) was probably intro- duced into the botanical gardens. The source of the introduced SOLEM: MOLLUSCA OF THE NEW HEBRIDES 47 specimens is unknown. Franc (1957, pp. 97-101) added no new data, but provided a convenient summarization of Grimpe and Hoffmann's studies on the New Caledonian Aneitea. The infor- mation on morphologic variation presented below raises doubts as to the validity of the many New Caledonian "species," and field studies are needed to clarify the status of the named forms. The primary character utilized for recognition of species in Anei- tea has been the shape of the central tooth of the radula. Each specimen examined in this study had a differently constructed cen- tral tooth and van Benthem Jutting (1933, p. 90) found similar variability in the New Guinea Trihoniophorus. Within a species, the central tooth is roughly comparable, but the individual varia- tions exceed the limits of "specific" difference established by Grimpe and Hoffmann (1925a). The central tooth is in the process of being lost and it is a basic tenet of evolutionary theory that vestigial struc- tures are not subject to selection and thus are apt to vary widely. The internal shell of Aneitea is deeply grooved and pitted and varies widely in shape. The drawings in Grimpe and Hoffmann (1925a) suggest that the shell provides a guide to recognition of species through its shape, but this is not borne out by the material exam- ined. Color variation in slugs is equally unreliable as a specific char- acter, particularly in preserved specimens. Whether the color pattern in slugs is essentially cryptic (Scharff) or is correlated with tempera- ture (Simroth) is unknown. The New Hebridean slugs show quite different color patterns and there are excellent opportunities for field studies on this factor. Specimens of the genotype, A. macdonaldi, and a series from Espi- ritu Santo were available for this study. The most constant charac- teristics separating the two populations were the shape of the jaw, the number of lobes in the hermaphroditic gland, the position of the oviducal accessory gland, the presence or absence of an epiphallic caecum, and the shape of the spermatheca. Without re-examining their original material, it is impossible to evaluate the many species proposed by Grimpe and Hoffmann (1925a). The characters listed above may provide a basis for specific recognition, but more material must be examined before their importance can be evaluated. For convenience the "species" of Grimpe and Hoffmann are briefly dis- cussed below and the original figures of the anatomy reproduced (pi. 3, figs. 5-13). Slugs collected on Aneiteum, Tanna, and Erromanga are referred to A. macdonaldi. The material from Espiritu Santo is most similar to A. robsoni Hoffmann but differs enough to be given nomenclatural 48 FIELDIANA: ZOOLOGY, VOLUME 43 recognition. A. rohsoni was collected on Mount Tabwemasana (6,000 feet), while A. r. santoensis came from near the seacoast. The new "subspecies" may only represent a local population, but its true status can only be evaluated when more adequate collections have been made on the island. Grimpe and Hoffmann (1925a) mention that slugs were seen on Ambrym by Speiser but none were collected. Aneitea maloensis Grimpe and Hoffmann. Plate 3, figs. 5, 9, 11, a, 12, a. Aneitea macdonaldi maloensis Grimpe and Hoffmann, 1925, Nova Caledonia, 3, (10), pp. 436-438, figs. 15h, 16to, 111 (not m), 181, 19a, 20; pi. 5, fig. 8a,b — Malo Island, New Hebrides. Range. — Malo Island. Material. — No material available. Remarks. — Basal plate of rachidian tooth not bifid; tooth itself indistinctly bicuspid. Both upper and basal plate of jaw very wide. Jaw without median projection. Hermaphroditic gland bilobed, epi- phallic flagellum lacking. Animal light-colored with two longitudi- nal black stripes. Shell short and broad. The differences found in the anatomy of A. macdonaldi justify raising this "subspecies" to "specific" rank. Aneitea speiseri Grimpe and Hoffmann. Plate 3, figs. 6, 10, 11, b, 12, b. Aneitea speiseri Grimpe and Hoffmann, 1925, Nova Caledonia, 3, (10), pp. 433- 436, figs. 15/, 16A;, 17m (not I), 18k, 20, pi. 5, figs. 10a, b, 11, 12— Espiritu Santo, Malo; Hoffmann, 1929, Zool. Anz., 84, (5-6), pp. 108-109— east central Espiritu Santo. Range. — Espiritu Santo, Malo. Material. — No material available. Remarks. — Rachidian tooth without cusps, only a small basal plate. Basal plate of jaw as in maloensis, but upper plate narrower and with low median cutting edge. Hermaphroditic gland bilobed, epiphallic flagellum lacking. Animal darker than maloensis, with two or three longitudinal black stripes, or the surface densely spotted with black. Shell elongate, narrow. Aneitea elisabethae Grimpe and Hoffmann. Plate 3, figs. 7, 8, 11, c, 12, c. Aneitea elisabethae Grimpe and Hoffmann, 1925, Nova Caledonia, 3, (10), pp. 431-433, figs. 15g, 161, Ilk, 18i, 20, pi. 5, fig. 13— Malo; Hoffmann, 1929, Zool. Anz., 84, (5-6), pp. 109-111— Espiritu Santo(?). SOLEM: MOLLUSCA OF THE NEW HEBRIDES 49 Range. — Malo and Espiritu Santo (?). Material. — No material available. Remarks. — Central tooth weakly or not bifid, cusps little or not developed. Basal plate of jaw more elongate than in first two spe- cies, upper plate narrow, long, with low median projection. Her- maphroditic gland bilobed, epiphallic flagellum lacking. Animal dark in color with rounded posterior. Shell elongate, slightly notched at one end. The specimens reported by Hoffmann (1929b) differed slightly from the original description, but probably are this "species." Aneitea robsoni robsoni Hoffmann. Plate 3, fig. 13. Aneitea robsoni Hoffmann, 1929, Zool. Anz., 84, (5-6), pp. 111-114, figs. 3-6— Mount Tabwemasana (ca. 6,000 feet), central Espiritu Santo. Range. — Mount Tabwemasana, Espiritu Santo. Material. — No material available. Remarks. — Basal plate of rachidian tooth deeply bifid, cusp struc- ture well developed, distinctly bicuspid. Upper plate of jaw narrow, with prominent median projection, lower plate wide, tapering poste- riorly. Hermaphroditic gland trilobed in drawing, bilobed in text. Epiphallus with small caecum (?). Penis strongly coiled. Animal almost black, mantle grooves colorless. Shell elongate, strongly curved. The dark color of A. r. robsoni probably is correlated with its mon- tane habitat and contrasts with the lighter coloration of A. r. santo- ensis. The two subspecies(?) have quite different radulae, jaws, and genitalia, but santoensis had best be considered a subspecies until spatially intermediate populations have been sampled. Aneitea robsoni santoensis, new subspecies. Plate 4, figs. 3, 4, 6, 8, 10, 11; plate 13, fig. 1. An Aneitea similar to robsoni in rachidian tooth structure, tri- lobed hermaphroditic gland, highly convoluted penis, shell, and basic color pattern. It differs from robsoni in jaw structure, lack of an epiphallic flagellum, in having the oviducal accessory gland much nearer the albumen gland, and in having only partially pigmented mantle grooves. Animal (pi. 13, fig. 1) elongate, tapering posteriorly, 22-43 mm. in length (preserved). Color reddish olive green, mottled with black spots on dorsal area. Sole of foot same color as dorsal region, but with fewer black spots. Cephalic 50 FIELDIANA: ZOOLOGY, VOLUME 43 shield triangular, outlined with dark streaks, inner edge of mantle grooves black. Respiratory, anal, and excretory pores nearest lower corner of mantle, distinctly separated and connected only by a groove in the epithelium. Upper plate of jaw (pi. 4, fig. 4) widest at center, tapering distally, without median projection. Basal plate short, broad, not tapering posteriorly. Internal shell (pi. 4, figs. 6, 10) irreg- ular, more deeply grooved and less solid than in A. macdonaldi. Radula with 127-135 rows and 115-121 teeth per half row. First few lateral teeth (pi. 4, fig. 11, right) large with well-developed cusps; after nineteenth lateral, the teeth diminish greatly in size. Central tooth varying both in depth and width of bifid basal plate; cusps may be absent (figs. 8, b-c) or claw-shaped (fig. 11, left). Genitalia (pi. 4, fig. 3) with trilobed hermaphroditic gland. Ovoid albumen gland with ducts of oviduct and vas deferens issuing near insertion of hermaphroditic duct. Prostate and vas deferens loosely bound to oviduct by connective tissue. End of prostate near oviducal diverticulum. Vas deferens passing along convoluted oviduct to genital atrium, then alongside penis to insertion of penial retractor. Penial re- tractor arising from body wall on left side, slightly above sole and 5 mm. behind posterior edge of mantle shield. Penis greatly convoluted. No epiphallic caecum. Epiphallus (above lowest convolution) internally sculptured with pustulose nod- ules; penis proper (below convolutions) with smooth tubercles only. Right ocular retractor passing between male and female genitalia. Spermatheca spherical even when filled with sperm. Type. — American Museum of Natural History. Collected on the leaves of plants and trees near Segond Channel, Espiritu Santo, by G. S. Banner on September 9, 1943. Paratypes. — Collected with the holotype and in the collection of the American Museum. Remarks. — The availability of 36 specimens from a single popu- lation made possible a study of variation in characters previously used for specific delineation. The shell and rachidian tooth varied extensively and are quite unreliable for identification. In contrast the shape of the jaw seemed constant, as also were the structures and relative proportions of the genitalia. The presence or absence of an epiphallic caecum, the position of the oviducal accessory gland and the shape of the albumen gland seem to be useful in classification. The other "species" from Espiritu Santo need to be re-examined before these criteria can be employed, however. The penis showed a peculiar dimorphism which is, at present, inexplicable. Most penes were of normal shape, with the penial re- tractor inserting at the juncture of the "thin" (epiphallic) and the "thick" (penis) sections of the penial complex. In about one-fourth of the animals dissected the upper half of the "thick" section was no larger than the vas deferens. All the parts of the genitalia were normal in proportion so that there was no question of misplacement of the muscle insertion. When first observed it was assumed that SOLEM: MOLLUSCA OF THE NEW HEBRIDES 51 the specimen was immature, but when this condition was found in a mating specimen that possibiHty was discarded. Both members of the other mating pairs were normal so that the probability of protogyny is lessened. The bionomics of Aneitea are unknown, although it is assumed that the slugs feed on plants. Among the 36 specimens, three pairs were in copulation and two others seemed to have been separated in the preservative. One of the copulating pairs was dissected completely; in the others the apical portions of the penes were examined. Copu- lation (pi. 13, fig. 1) is head-on and reciprocal. Both penes were in- serted into the vaginae and extended about two millimeters beyond the openings to the spermathecae. As is usual in the Stylommato- phora, the penes were everted. In both cases the spermathecae were tightly packed with sperm, a conclusive indication that insemination was mutual. The tubercles of the penes were prominent and evi- dently served as stimulators. The epiphallus and penial retractor were uncoiled and extended. No traces of a solid spermatophore could be found. The date of collection established that mating occurred in September. The relationship of santoensis to typical robsoni is uncertain. The trilobed hermaphroditic gland and the bifid basal plate of the rachidian tooth indicate affinities, but the structure of the jaws is quite different and robsoni seems to have a small epiphallic caecum which is lacking in santoensis. The latter may prove to be a distinct species, but until the former can be restudied and spatially inter- mediate populations examined, it is perhaps best to treat it as a subspecies. Aneitea macdonaldi Gray. Plate 4, figs. 1, 2, 5, 7, 9, 12; plate 13, fig. 2. "Bitentaculate slug," MacDonald, 1856, Ann. Mag. Nat. Hist., (2), 18: 38-42, pi. 3 — Aneiteum. Aneitea macdonaldi Gray, 1860, Ann. Mag. Nat. Hist., (3), 6: 195-196. Range. — Aneiteum, Tanna, Erromanga. Material— Aneiteum (AMNH, Macmillan, August, 1937) ; Erro- manga (AMNH, Macmillan, March-April, 1937); White Sands, Tanna (AMNH, Macmillan, December, 1936). Remarks. — Aneitea macdonaldi has not been reported in the liter- ature since its original collection and description. Hoffmann (1929b) mentioned deficiencies in the original study but they were only the result of the lack of knowledge of molluscan anatomy at that time. 52 FIELDIANA: ZOOLOGY, VOLUME 43 The genitalia, radula, and jaw have been refigured at this time for comparison with A. robsoni santoensis. The animals (pi. 13, fig. 2) range from 22 to 65 mm. in length, with only slight traces of dark mottling and a few dark streaks out- lining the cephalic shield. Externally they do not differ from A. r. santoensis, and the internal morphology does not differ greatly from that of A. brisbanensis (W. Pfeiffer, 1900). The jaw is quite narrow centrally, with a small but distinct median projection. Distally it widens, reaching a maximum width just before the "flaps" bend under. The basal plate of the jaw is only slightly broader than wide. The radula has 118-134 rows with 150+ teeth per row. The lateral teeth of the radula (pi. 4, fig. 5, b-f) show the same trends seen in A. r. santoensis and the central tooth (pi. 4, fig. 5, a) is even more vari- able than in the latter species. The cusps of macdonaldi are more blunted than those of A. r. santoensis, a condition possibly caused by feeding on harder substances than its northern relatives do. The shell is narrower than that of most Aneitea, varying in outline from elongate-ovate to a rough parallelogram; it is also thicker and more solid than that of A. r. santoensis. The genitalia fpl. 4, fig. 1) have the hermaphroditic gland weakly trilobed, with the separate ducts very short and the lobes closely bound; a casual inspection might suggest it is unilobed. The presence of an epiphallic caecum, an elon- gate spermatheca and the relative position of the oviducal accessory gland are the main characters separating macdonaldi from santoensis. Most specimens of macdonaldi were juvenile, but four adults were dissected. No significant differences were found between specimens from Tanna, Aneiteum, and Erromanga. Order HETERURETHRA Pilsbry, 1900 Jaw elasmognathous. Kidney broader than long, extending from pericardium to hindgut, ureter lying along front margin, then following rectum to pneumo- stome. Lung short, but with normal venation. Shell oval, thin, with spire of few whorls or uncoiled. (After Pilsbry, 1948, p. 771.) As utilized in this study, the Heterurethra contains only the fam- ily Succineidae. Possibly the African Aillyidae and the Tracheopul- monata may prove to be closely related to the Succineidae (see p. 33). Family SUCCINEIDAE The amber snails are worldwide in distribution and the rudimen- tary shells provide few characters of value in classification. In recent SOLEM: MOLLUSCA OF THE NEW HEBRIDES 53 years Quick (1933, 1934, 1936, 1939a, b), Boettger (1939), Pilsbry (1948), and Odhner (1950) have attempted to base succineid classi- fication on the structures of the soft parts. Following Odhner (1950), I have placed species without a penial sheath in a subfamily Catel- linae, species with a sheath in the Succineinae. Lee (1951) erected a third subfamily in an unpublished thesis, but his action does not affect the status of any of the species considered below. Hawaiian and some Polynesian succineids belong to the Catellinae, but all the Australian, Indonesian, and Melanesian species, as far as now known, belong to the Succineinae. Subfamily SUCCINEINAE A large number of nomenclatural genera are available for the members of this subfamily, but too little is known about individual anatomical variation and too many species have not yet been dis- sected to enable recognition of many biological groups. Separation of the species with a penial appendix as Oxyloma seems justifiable, but it is not possible to go further at the present time. Genus SUCCINEA Draparnaud, 1801 Type species. — Helix putris Linnaeus. Remarks. — Many sections of Succinea, such as Novisuccinea Pils- bry (1948, p. 801) and Desmosuccinea Webb (1954, p. 10), have been based on the degree to which the epiphallus and penis are free from the sheath. The variation observed in the specimens studied raises doubt as to the validity of this character for specific, much less ge- neric determinations. Webb (1953, p. 216) has shown that during copulation the "free loop" of the penis and /or epiphallus disappears into the sheath upon extrusion of the penis. This fact provides a ready explanation for variation in the size and prominence of the loop. The differences may be caused by incomplete retraction of the penis after copulation or spasmodic muscular contractions in the killing solution. Apparently a more reliable specific character is the exact point of insertion of the penial retractor muscle on the sheath, penis and /or epiphallus (Pilsbry, 1948, p. 801). Unfortunately this information is known for only a few species. Preserved material of two species, S. simplex and S. kuntziana, was available to me, and the anatomy of several other Indo-Pacific succineids has been par- tially figured in the literature. 54 FIELD! ANA: ZOOLOGY, VOLUME 43 A survey of published records and two collections of shells (UMMZ and CNHM) indicated that the Austromalayan succineid fauna con- tains two groups, distinguished by the presence or absence of micro- scopic criss-cross sculpture on the shell and the relative lengths of radular cusps and basal plates, A group with microscopic criss-cross sculpture and the cusps on the rachidian tooth as long as the basal plate was defined by Quick (1939a, b, 1951). He listed several spe- cies probably belonging to it. A later paper by van Benthem Jut- ting (1952) enabled me to add the Javanese forms. A species de- scribed by Abbott (1950), although aberrant in several aspects, probably can be included in the same taxon. The following seem to be related: obesa von Martens (Java), minuta von Martens (Java), keelingensis Abbott (Cocos-Keeling Atoll), solitaria Smith (Christ- mas Island), australis Ferussac (Victoria), norfolkensis Sykes (Norfolk Island), caduca Mighels (Hawaii), texta Odhner (Masatierra), ande- cola Crawford (Peru) and striata Krauss (South Africa). The sec- tional name Austrosuccinea Iredale (1937a, p. 307, type S. australis Ferussac) can be applied to this group. The more recent "genus" Spirancinea Iredale (1945, p. 53, type S. norfolkensis Sykes) is a probable synonym, although the sculpture of S. norfolkensis is much heavier than in the other species examined. The nearly world-wide distribution of succineid genera and sections is well established, al- though its cause is unknown. Transport of Succinea by birds has been reported (see Nautilus, 27, p. 71; 50, p. 31), but the importance of such factors in molluscan distribution can easily be over-emphasized. A second group of species without any trace of the microscopic sculpture, even under 100 X magnification, and with the rachidian cusps much shorter than the basal plates contains the following spe- cies: strubelli Strubell (Papua), papuana Strubell (Papua), brittaniae Mousson (Bismarck and Admiralty Islands), simplex Pfeiffer (Solo- mon Islands), kuntziana, new sp. (New Hebrides), modesta Gould (Samoa and Tonga), archeyi Powell (New Zealand) and probably the New Caledonian species. The name Papusuccinea Iredale (1941, p. 63) is available and has been adopted as a sectional name for these species. Quick (1951) studied the anatomy of archeyi and called it an Austrosuccinea. The radula with short-cusped rachidian tooth (Powell, 1950) and the lack of criss-cross sculpture on the shell (Quick, 1951) indicate that it is a Papusuccinea, a group of species evidently not examined by Quick. Apparently there are no readily discernible differences in the genitalia of Austrosuccinea and Papu- succinea. SOLEM: MOLLUSC A OF THE NEW HEBRIDES 55 Section PAPUSUCCINEA Iredale, 1941 Shell smooth, without microscopic criss-cross striations. Radula with mesocone of central tooth extending only slightly past mid-point of basal plate; in Austro- succinea it is at least two-thirds the length of and often as long as the basal plate. Genitalia similar to those of Succinea avara of eastern North America, but shell and radula quite different. Penis more slender and vagina longer than in Calci- succinea. Type species. — Succinea strubelli Strubell. Remarks. — Members of Papusuccinea replace each other geo- graphically in a long arc from New Guinea to New Caledonia (see Franc, 1957, pp. 94-96, for New Caledonian Succinea). There is a species in Samoa and Tonga, but no succineids have been reported from the Fijis. Conchologically the species are very similar and records in the literature are quite untrustworthy for specific designa- tions. I. Rensch (1937, p. 593) suggested they may represent a classical Rassenkreis. Pending anatomical studies of the entire com- plex, however, it is perhaps best to consider them distinct species. Previously published figures of the genitalia of species of Papusuc- cinea do not illustrate diagnostic characters, nor do accounts of the radula offer sufficient information to enable proper allocation of the nomenclatural units listed above. It is not possible to determine the affinities of Austrosuccinea and Papusuccinea to "genera" found in other areas. The New Hebridean succineid has been recorded in the literature, but not named. I take great pleasure in naming it after Commander Robert E. Kuntz, whose donation of so much carefully preserved and localized material made this study possible. Succinea (Papusuccinea) kuntziana, new species. Plate 5, figs. 1, 3-8; plate 13, fig. 3. Succinea sp. Sykes, 1903, Proc. Malac. Soc. London, 5: 198 — Valua, Banks Islands and Vila, Vate. Succinea simplex Iredale (not PfeifTer, 1855), 1941, Australian Zool., 10, (1), p. 63 — New Hebrides. A Papusuccinea with flatter whorls and fewer lateral teeth than are found in S. simplex Pfeiffer from the Solomon Islands, and with smaller nuclear whorls and less elongate spire than S. montrouzieri Crosse from New Caledonia. Shell thin, elongate-ovate, translucent, light horn-colored, apex often reddish. Whorls 2H to 3, gently rounded. Spire-aperture ratio 0.27 to 0.40 in adults; young shells with a lower ratio and whorls more sharply rounded than in adult. Surface uneven and slightly pitted. Growth striae absent to prominent. Radula 56 FIELDIANA: ZOOLOGY, VOLUME 43 with 78 to 112 rows, formula (13-20)-(7-9)-l-(7-9)-(13-20). Mesocone of central tooth (pi. 5, fig. 3, a) slightly less than two-thirds the length of basal plate. Lateral and marginal teeth (pi. 5, fig. 3, b-d) with small endocone, occasionally lacking. Jaw (pi. 5, fig. 6) without accessory ribs and having only a small median pro- jection. Talon (pi. 5, figs. 1, 5) unequally bilobed in many specimens. "Free loop" of epiphallus (pi. 5, fig. 8) completely contained within the penial sheath in a few specimens (pi. 5, figs. 1, 7), partially protruding in the majority. Penis slender. Penial retractor inserting at junction of epiphallus and vas deferens, some fibers attaching to penial sheath. Penis and epiphallus undifferentiated externally except for a slight constriction at their junction, concealed by the penial sheath. Epi- phallus sculptured internally by a series of longitudinal beaded ridges. Penis with four to six unbeaded longitudinal ridges. The same pattern is found in the North American Succinea avara (Pilsbry, 1948, p. 818, fig. b). Height of shell 8.2 to 12.8 mm., diameter 5.4 to 8.0 mm., height of aperture 5.4 to 8.0 mm. Type. — University of Michigan Museum of Zoology no. 183566. Collected at Mile. Havet Plantation on Espiritu Santo (ML 96) by Robert E. Kuntz in July, 1944. Paratypes. — On Espiritu Santo, Kuntz collected specimens at the following stations: ML 13, ML 22, ML 26c, ML 31f, ML 39, ML 43, ML 46, ML 63, ML 64, ML 69, ML 70, ML 74, ML 76a, ML 96 (type locality) . Most of these are in the University of Michigan Museum of Zoology. Some have been distributed to other institutions (CNHM, MCZ, ANSP, BPBM, USNM, and Australian Museum, Sydney). Other paratypic material includes : USNM 432456 (Espiritu Santo) ; USNM 515359 (Tanna); and USNM 598358 (Vila, Vate, ex Miller 452). Range. — Espiritu Santo, Valua (Banks Group), Vate, and Tanna. It probably will be found on most of the other islands. Remarks. — Through the kindness of Dr. William J. Clench, one adult and one juvenile specimen of Succinea simplex Pfeiffer from Paivi, Ugi, Solomon Islands, were made available for dissection. Un- fortunately the specimens had previously been preserved in formal- dehyde. The radulae and jaws were extracted and the general facies of the male genitalia could be observed, but the hardness of the tis- sues made critical studies impossible. The radular teeth are quite similar to those of kuntziana, although the mesoconal cusp of the central tooth is slightly shorter in simplex, and the endocone of the lateral teeth is more developed in the latter species. There may be a clinal gradient in the endoconal development of the lateral teeth in Papusuccinea. The endocone is very much reduced in S. kuntzi- ana, stronger in S. simplex and quite prominent in S. brittaniae Mousson (I. Rensch, 1937, p. 593, fig. 38). S. kuntziana and S. sim- plex differ in radular formula. The adult of the latter has the formula 15-13-1-13-15, and the juvenile (4.2 mm. high) has the formula 15- SOLEM: MOLLUSCA OF THE NEW HEBRIDES 57 9-1-9-15. Quick (1934) correlated changes in radular formula with increase in shell size in two British species. His data show that even in specimens three and four millimeters in height, differences in the number of lateral teeth are significant. A 4 mm. specimen of kuntzi- ana had only five lateral teeth. In view of the results of Quick's study the difference between S. simplex and S. kuntziana is consid- ered to be of specific value. S. struhelli (see Hedley, 1892b, p. 691) has the formula 16-13-1-13-16; I. Rensch (1937, p. 593) has reported 51-53 teeth "in einer halben Reihe" of S. brittaniae without giving the actual formula, but no other Papusuccinea have been examined. No significant differences were found in the genitalia of S. kuntziana (pi. 5, fig. 1) and S. simplex (pi. 5, fig. 9). The anatomical figures of succineids given by B. Rensch (1932, p. 127, figs. 52, 55) and I. Rensch (1937, p. 593, fig. 37) do not treat diagnostic characters. Seemingly, as in all succineids, the jaws of the sections Austrosuc- cinea and Papusuccinea show no constant differences. Specimens of S. modesta from Tutuila, Samoa, have a more obese shell. No authentic material of succineids from the Bismarcks or New Guinea was available. Adult simplex and kuntziana are illustrated (pi. 13, figs. 3^). The juvenile shells of S. kuntziana are nearly as rounded as those of S. simplex. The characteristic flatness of the whorls of S. kuntziana is apparent only in specimens with 2% or more whorls. Material from Tanna and Vate (USNM) is juvenile, with only 2^ whorls. One adult shell from Miller (number 542) showed no differences from the Espiritu Santo adults. S. kuntziana was collected on Espiritu Santo from October, 1943, to June, 1944, but no definite reproductive pattern was discernible. Young shells (less than 5 mm. in height) were found in October and May, half-grown specimens (5-7 mm.) in October, January and June, and adults (more than 7 mm.) in December, January, April and June. Powell (1950) reported an annual life cycle for the New Zealand succineid; a similar cycle has been established for several North American species, but the data on S. kuntziana are insufficient to establish its life span. Most living specimens of S. kuntziana were collected under piles of decaying leaves, sticks, or palm fronds in heavy shade. Many were found in areas near streams, but one lot (ML 39) was found on a hillside and a few adults were collected on the trunks of trees and shrubs during a rain (ML 31f). Except for S. archeyi (Powell) (1950), all Papusuccinea have similar ecology. Both S. simplex Pfeiffer (see E. A. Smith, 1885) and S. strubelli (see Hedley, 1892b) live in native 58 FIELDIANA: ZOOLOGY, VOLUME 43 gardens near water. The general habitat of Papusuccinea is similar to that of the North American avara complex, which seems to be a group convergent, if not related, to Papusuccinea. Order ORTHURETHRA Pilsbry, 1900 Holopod snails with the kidney tapering anteriorly into the ureter, which runs directly forward, remote from the hindgut and opens within the forward border of the lung. (After Pilsbry, 1948, p. 848.) Division of the Orthurethra into family categories is still pro- visional. Watson (1920) recognized four families, Steenberg (1925) sixteen, and H. B. Baker (1956a) ten. Pilsbry (1948, p. 848) con- sidered that families are primarily recognized on the basis of expedi- ency and that shell characters are often more useful in classification of the Orthurethra than the soft parts. This apparent similarity in anatomy may be more the result of convergent evolution than any basic conservatism, but much work needs to be done before a defini- tive classification can be attempted. Of the ten families recognized by H. B. Baker (1956a), four — Pupillidae, Enidae, Tornatellinidae, and Partulidae— are found in the New Hebrides. The single enid is very probably introduced. The single pupillid and two tornatellinids are found over much of the Pacific and probably have been dispersed by natives. Only the partulids form an important endemic part of the fauna. Family PUPILLIDAE (=Vertiginidae) Genera such as Nesopupa, Pupisoma, and Cylindrovertilla may subsequently be found in the New Hebrides, but only the very wide- ranging Gastrocopta pediculus (Shuttleworth) is known from the islands at the present time. Gastrocopta (Sinalbinula) pediculus (Shuttleworth) Pupa pediculus Shuttleworth, 1852, Mitth. Naturf. Gesell. Bern, 1852: 296— Marquesas Islands. Gastrocopta pediculus (Shuttleworth) Pilsbry, 1917, Man. Conch., (2), 24: 145- 152, pi. 25, figs. 1-3, 5-8, 12-15. Range. — Espiritu Santo. From Java to Australia, New Cale- donia, all of Polynesia and Micronesia and Hawaii. Material.~ML 35, ML 39, ML 40, ML 69, ML 78, ML 95. Remarks. — The range of Gastrocopta pediculus and its occurrence only at lower elevations on high islands (Pilsbry, 1917, p. 148) lends SOLEM: MOLLUSCA OF THE NEW HEBRIDES 59 credence to the possibility that it may have been introduced over much of its range. Several weakly characterized varieties have been named. Although a few New Hebridean specimens may qualify as these forms, the great majority were typical pediculus. Family ENIDAE ( = Buliminidae) Enids are common in the temperate and tropical regions of Africa, Europe, and Asia. The distribution and taxonomy of the African and European species have been extensively studied, although seri- ous gaps in our knowledge still remain. The exact limits of distribu- tion in Asia are very uncertain. Enids are common in India and China, and members of the typical subfamily, Eninae, are reported from Java (van Benthem Jutting, 1952, pp. 366-373) and Lombok (B. Rensch, 1932, pp. 118-119). Another series of species close to, if not congeneric with, the South African Rhachistia, is discussed below. The status of several fringe groups is less certain. Pseudonapaeus aperttis (Martens), found on Timor (B. Rensch, 1935, p. 320), has typically enid radular dentition but the genital anatomy is unknown. Apoecus colonus (Mllf.) from Constantinhaven, New Guinea (see Kobelt, 1902a), although unknown anatomically, is similar enough to Pseudonapaeus apertus to be tentatively referred to the Enidae (see Solem, in press-B). Aminopina Iredale (1933, p. 42; 1938, p. 93; 1941, p. 64) from New Guinea and northern Queensland prob- ably is an enid (see Solem, op. cit.). Thiis, enids are known definitely from Java, Lombok and Timor, and possibly from Australia and New Guinea (fig. 16). No species referable to the Enidae have been reported from the Bismarcks, the Solomon Islands, New Zealand or Polynesia. Many enids are con- chologically similar to the Partulidae, which possibly will be found to be ancestral to the Enidae. In the Partulidae there are no acces- sory organs on the penis; accessory organs are always present in the Enidae. In this respect the enid-partulid may parallel the helicid- camaenid relationship in the holopod Sigmurethra. Subfamily relationships within the Enidae are obscure. The most recent classification is that of Thiele, who places the New Hebridean species in the Pachnodinae. I have accepted his classification but have made no attempt to characterize the subfamily. The Pachnodi- nae are found chiefly in southern Africa and Asia, although a few species live on the Mediterranean islands. There is a large group of 60 FIELDIANA: ZOOLOGY, VOLUME 43 very similar species found in South Africa, India, southeast Asia, the PhiUppines, Celebes, Timor, New Hebrides, New Caledonia and Queensland. Thiele (1931) and Connolly (1939) place them in a single genus, Rhachistia; Tomlin and Peile (1930) put the Asiatic species in a new genus Eorrhachis; and Iredale (1933) created a new genus, Rachispeculum, for the Queensland "species." There are only minor conchological differences between African and New Hebridean specimens and radular structure is identical (see Tomlin and Peile, 1930) . Study of the genitalia may indicate that Eorrhachis is a valid genus, but in the light of our present knowledge only one genus is admissible. Genus RHACHISTIA Connolly, 1925 ( = Eorrhachis Tomlin and Peile, 1930, and Rachispeculum Ire- dale, 1933) Shell comparatively large, short turriform, solid, smooth, fairly glassy, with dark spots and brown or purple bands on ground of rose, buff, flesh, or cream. Aperture ovate, peristome acute, not expanded. Teeth of radula large, lobe-shaped, central single, blunt or round (Connolly, 1939, pp. 418-419). Type species. — Buliminus rhodotaenius Martens. Remarks. — The Asiatic species of Rhachistia have been charac- terized mainly on the basis of color pattern. R. sulphur eus (Tomlin and Peile) has only a few maculations, and R. zonulatus (Pfeiffer) has a series of spiral bands and spots. R. zonulatus has been reported from Timor, the Celebes, and the Philippines. The spotty distri- bution of both zonulatus and histrio suggests that they may be intro- duced species, since R. punctatus Anton is an Indian species widely introduced into Africa (Pilsbry, 1919, pp. 304-315). R. histrio is known from Queensland, New Caledonia, and the New Hebrides. A lot labeled Tamatave, Madagascar (CNHM 48032), appears to be typical histrio and I suspect that R. histrio may prove to be an African species imported to New Caledonia and then to the New Hebrides and Queensland. Rhachistia histrio (Pfeiffer). Plate 8, fig. 9; plate 13, fig. 5. Bulimus histrio Pfeiffer, 1854, Proc. Zool. Soc. London, 1854: 124 — Tanna. Bulimus magenii Gassies, 1856, Jour, de Conch., 5: 181, pi. 6, fig. 5 — New Caledonia. Bulimus hidwilli Cox, 1868, Monog. Australian Land Shells, p. 72, pi. 13, fig. 11 — Burnett River, Queensland ("in the tops of trees"). Buliminus (Rhachis) histrio (Pfeiffer) Kobelt, 1902, Conch. Cab., I, 13, (1), p. 749, pi. 110, figs. 9, 10. SOLEM: MOLLUSCA OF THE NEW HEBRIDES 61 Buliminus {Rhachis) bidmlli (Cox) Kobelt, 1902, op. cit., pp. 752-753, pi. 110, figs. 18, 19. Rhachis magenii (Gassies) Sykes, 1903, Proc. Malac. Soc. London, 5: 198 — Vila, Vate. Rhachistia (Eorrhachis) histrio (Pfeiffer) Cockerell, 1937, Nautilus, 51, (1), p. 35. Implies that histrio might be introduced. Rachistia (sic) {Eorrhachis) histrio (Pfeiffer) Franc, 1957, Mem. Mus. Nat. d'Hist. Nat., N.S., (A), Zoologie, 13: 108, pi. 11, fig. 141. Range. — Tanna, Vate, Loyalty Islands, New Caledonia, Queens- land, and Madagascar (?), Material— Vi\2i, Vate (USNM 598634, Miller); many lots (UMMZ and MCZ) from New Caledonia, Loyalty Islands, and Queensland. Photograph of holotype of Bulimus histrio Pfeiffer (courtesy of BM). Remarks. — The holotype of Bulimus histrio Pfeiffer (pi. 8, fig. 9) is juvenile and has a color pattern quite dissimilar to the specimens figured by Kobelt (1902a) as this species. The color pattern in the twenty lots seen varied widely and encompassed the descriptions of histrio, magenii, and bidwilli, as well as the color forms named by Gassies (1871) and Crosse (1894). The color varies from albino to a pattern of four dark color bands and a row of spots. The Vate specimens (pi. 13, fig. 5) lack the two upper color bands and closely resemble the type figure of magenii. Numerous New Caledonian specimens are as fully colored as Kobelt's histrio; others are albino or had the magenii pattern. The size, shape, and sculpture of the holotype and the New Caledonian and New Hebridean specimens are identical. The name histrio is retained, despite the wide diver- gence of the color pattern of the holotype from that shown in Kobelt's figures and the Vate specimens. The figures, description, and specimens of bidwilli leave little doubt that it is the same species as R. histrio. R. bidwilli has been found in the Port Curtis district, Port Denison, Bundaburg, Mount Dryander, and near Maryborough, Queensland. All are near the active centers for the blackbirding trade between the New Hebrides and Queensland during the late 1800's. There was thus ample op- portunity for accidental importation either to or from Queensland. In the majority of cases the occurrence of a snail away from human habitation is an indication that it is "native." From the published records, it is probable that bidwilli has been collected "in the scrub" away from the cities, although the recorded localities are all from coastal regions. This is not significant, as R. punctatus has become similarly adapted in Africa (Pilsbry, 1919, pp. 304-305). Iredale's 62 FIELDIANA: ZOOLOGY, VOLUME 43 belief that bidwilli is a derivative from a Papuina-\ike shell indicates lack of familiarity with non-Australian Mollusca. Family TORNATELLINIDAE A monograph of the Tornatellinidae by the late C. Montague Cooke and Dr. Yoshio Kondo is in press at the Bishop Museum, Honolulu. For this reason, diagnoses of higher taxa and discussions of distribution, phylogeny, and systematics are not given below. The Tornatellinidae are found on most islands of the Indian and Pacific Oceans (fig. 13) . A few genera reach the Galapagos and Juan Fernandez Islands but no tornatellinids have been reported from continental areas. The previous major studies on the Tornatellini- dae are those of Pilsbry and Cooke (1915-16) and Odhner (1922). I tentatively identified the two New Hebridean species, and then forwarded them to Dr. Kondo at the Bishop Museum. I am in- debted to Dr. Kondo for his kindness in checking my identifications and for providing me with the information of the distribution of the Tornatellinidae included in my zoogeographic survey. Genus ELASMIAS Pilsbry, 1910 Elasmias apertum (Pease) Range. — Espiritu Santo. Widely distributed in Polynesia. Material— ML 95. Remarks. — The New Hebridean specimens of E. apertum are most similar to the Huaheine, Society Island, shell figured by Pilsbry and Cooke (1915-16, pi. 30, fig. 3). E. apertum is extremely widely dis- tributed in the Pacific and is an inhabitant of many coral atolls. Genus LAMELLIDEA Pilsbry, 1910 Lamellidea pusilla (Gould) Range. — Espiritu Santo, most of Polynesia, Micronesia, and Melanesia. Material— Espiritu Santo (USNM 432457; Harrington). Remarks. — The single adult specimen is referred to L. pusilla on the authority of Dr. Kondo. Younger specimens of the same spe- cies were seen from the Solomon Islands (UMMZ) during this study. L. pusilla is easily separated from Elasmias apertum by the globose SOLEM: MOLLUSCA OF THE NEW HEBRIDES 63 shape and apertural folds present in the adult of the latter. Young L. pusilla have prominent teeth, which are lost in the adult, and are elongate conic in shape. Family PARTULIDAE Pilsbry (1909), Pilsbry and Cooke (1934a, p. 21), and Kondo (1948) emphasize that the Partulidae are related to the Orthurethra rather than the sigmurethrous Bulimulidae as suggested by Thiele (1931, p. 658). Placement of the Partulidae in the Orthurethra is relatively certain, but as yet no pulmonate family can be definitely selected as either its ancestor or its descendant. A few Tertiary fos- sils have been erroneously referred to Partula, but there is no fossil record (Pilsbry, 1909, p. 164). The Partulidae (fig. 12) are confined to the high islands of the southern and western Pacific. A few species are found in the Palau, Caroline, and Mariana Islands, but the main development occurred in Polynesia and Melanesia(?). The Society Islands, with the larg- est number of species, have been studied by Crampton (1916, 1932). The partulid fauna of the Marquesas, Australs, Samoan, Fijian, and Tonga Islands is less well known, and no speciation studies have been made on the Partulidae of Melanesia or Micronesia (except Guam, by Crampton, 1925). No partulids are known from New Caledonia, Hawaii, the Marshall Islands, or any part of the Indo-Australian archipelago, except eastern New Guinea and the Louisiades. Pilsbry (1900b, p. 568, and 1909, pp. 166-167) and Crampton (1932, pp. 194-197) cited Partula as evidence of a former Pacific continent, but the involved phylogeny and leaf-clinging habit sug- gest that passive dispersal by winds may have played an important part in their distribution. Early students of the Partulidae were W. Harper Pease, Andrew Garrett and W. H. Hartman. Their contributions have been sum- marized by Pilsbry (1909). Subsequent studies on variation by Crampton (1916, 1925, 1932), while not concerned with New He- bridean species, give valuable information as to the extent and trends of variation within the family. For many years the family Partulidae was considered monogeneric, but Pilsbry and Cooke (1934a) recog- nized several genera and subgenera based on anatomical differences. Dr. Yoshio Kondo recently completed an extensive anatomical survey of the Partulidae. His conclusions as to generic and anatom- ical relationships were generously made available to me and are in- 64 FIELDIANA: ZOOLOGY, VOLUME 43 corporated here with his permission. Time and distance have not allowed co-operation on specific relationships and Dr. Kondo is in no way responsible for any errors and omissions in regard to individ- ual species. Of the three genera Kondo recognizes, only one, Partula, is found in Melanesia. The Melanesian partulids are all true Partula and have one or the other of the two most generalized types of penial structure. These types are found throughout the range of the genus and are of no value in recognition of subgeneric or sectional cate- gories. The Melanesian species are thus placed in Partula without subgeneric designations. Kondo (personal communication) informs me that the Melanesian partulids probably represent a backwater of partulid evolution and are neither primitive nor advanced. The present survey of New Hebridean partulids is based on a re- examination of Hartman's collection (now in CM) and study of speci- mens in several other museums. In only one case were more than two or three specimens of a species available and no anatomical material was seen. A critical revision could not be attempted, but types have been refigured, errors in localities corrected and the "spe- cies" grouped into "superspecies." It is very probable that these "superspecies" are biological species and that the "species" will be recognized as synonyms or local races when adequate series are available for study. After exclusion of spurious records and species of uncertain status, there are five recognizable groups of partulids in the New Hebrides. P. vanikorensis, P. minor Hartman and P. milleri, new species, are related to Solomon Island species; and there are three "superspecies" seemingly restricted to the New Hebrides. Genus PARTULA Ferussac, 1821 On conchological criteria the Melanesian species of Partula can be divided into two series, one restricted to the New Hebrides, the other ranging from Espiritu Santo to New Guinea. They differ in size, prominence of spiral sculpture and development of the parietal callus. A complete discussion of their differences is given in Solem (in press-A). P. minor, P. vanikorensis, and P. milleri belong to the second series. The first series contains three "superspecies," grouped around P. macgillivrayi Pfeiffer, P. pyramis Hartman and P. au- raniana Hartman. The third group is quite distinct, but the inter- relationship of the other two is uncertain. Typically they are easily separable, but a few puzzling lots were seen. Without further mate- SOLEM: MOLLUSCA OF THE NEW HEBRIDES 65 rial it cannot be decided whether the first two "superspecies" repre- sent two taxonomic entities or are a series of convergent intraspecific variations. PYom notes accompanying the Hartman collection it is evident that Hartman came to believe that the macgillivrayi and pyramis series were actually only one species. The probability of this is admitted, but without more material than was available the two series should tentatively be considered distinct on the basis of the differences noted below. The morphometry of the New Hebri- dean partulids is summarized in Table II. SUPERSPECIES OF Partula macgillivrayi Shell slightly smaller than in the pyramis group (usually less than 23 mm. high), sculpture lighter, fading out on the last whorl except at the base, sutures less im- pressed and whorls less rounded. Little or no parietal callus and columellar callus reduced. Color pattern of longitudinal bands and a strong to subobsolete spiral band near the periphery of the body whorl. The species are quite similar, differing only by minor variations in color, shape and sculpture. There has been much confusion as to relationships, partly because of lack of material and partly because the exact morphotype of each taxonomic unit was uncertain. Figur- ing of type material fixed the morphotypes, but the question of rela- tionships can only be settled by field studies. Until adequate series from Aneiteum, Erromanga, Tanna and Vate are available, it is probably best to consider the "species" distinct entities. Partula macgillivrayi Pfeiffer. Plate 9, fig. 1. Partula macgillivrayi Pfeiffer, 1855, Proc. Zool. Soc. London, 1855: 97 — New Hebrides (MacGillivray); Pfeiffer, 1857, Nov. Conch., 1 : 61, pi. 17, figs. 14, 15; Pilsbry, 1909, Man. Conch., (2), 20: 278, pi. 33, figs. 15, 16. Range. — Aneiteum. Material.— New Hebrides (UMMZ 145746) . Photograph of holo- type (BM). Remarks. — The photograph of the holotype (pi. 9, fig. 1) shows that macgillivrayi has a wider, more reflexed columella, more trun- cate base, more prominent sculpture, and slightly more rounded whorls than do other members of the macgillivrayi superspecies. There is no suprasutural cord. Cox (1868) reported P. macgillivrayi from Aneiteum and the type locality is restricted to that island. From information in Layard's notes, it probably will be found on the southwest part of Aneiteum. ::^ d ^^^. _. M OOOOOOg^OOi-lOOOOO o OOOOOOOOOOi-HOOOO PL, 0) ■l-> "—I fc(M'-Hi-H^t-T-H05 (M03eO«DO«£iOO«5Tj CO Q i-ioooO(MT-iui_;«>t-coo(MUi(MiO t-t-ixioo-^coeciososusoscoo'^eo 1-1 t, o "S O Oi O -* U5 t- 2 (M(MO 00(MC O O lO i-H C<1 S coT-HC^ieoeoi-ileococgiMC^c^icoc^ (M(Mi-irHT-iT-ii-ioiMi-i(M0500(jicg CSi-HtHi— It— (T— (1— l'~'i— (rHi-Hi— li— li— IrHT— I T— It— (tH>— It—It— ll— ll— ll— (i— ll— I T— I 5 d ^. u:)(MUit-t~t-??W5c^o-*iot-(M t-oot-Ti>« J3 ,.- o o s JS S 0,0, 0, 0,0, 66 SOLEM: MOLLUSCA OF THE NEW HEBRIDES 67 Partula radiosa (Pfeiffer). Plate 9, fig. 3. Bulimus radiosus Pfeiflfer, 1854, Proc. Zool. Soc. London, 1854: 58 — New Hebrides. BuUminus {Rhachis) radiosus (Pfeiffer) Kobelt, 1902, Conch. Cab., I, 13, (2), p. 987. Range. — New Hebrides. Material. — Photograph of holotype (BM). Remarks. — The juvenile holotype was the first New Hebridean partulid seen by Pfeiffer. The similarity in sculpture and coloration to many buliminids makes this error in describing it as an enid excus- able. It is evident from the photographs that Partula radiosa and macgillivrayi (pi. 9, fig. 1) are very closely related. The only differ- ence is the presence of a supra-sutural cord in P. radiosa. The type locality of radiosa is unknown, but it is probably from one of the southern islands. Partula turneri Pfeiffer. Plate 9, fig. 9; plate 16, figs. 2, 3. Partula turneri Pfeiffer, 1860, Proc. Zool. Soc. London, 1860: 40 — Erromanga (Turner); Pilsbry, 1909, Man. Conch., (2), 20: 277, pi. 33, figs. 5, 6. Partula turneri perstrigata Pilsbry, 1909, Man. Conch., (2), 20: 277, pi. 33, fig. 4 — Tanna. Partula (Melanesica) turneri perstrigata (Pilsbry) Pilsbry and Cooke, 1934, Occ. Pap. B. P. Bishop Mus., 10, (14), p. 15, figs. 5c, d (anatomy) — Tanna (Robertson). Range. — Erromanga, Tanna. Maierial.—New Hebrides (UMMZ 14035, MCZ 24836, CM 62.4303 ex Geale, CM 62.4304 ex Brot) ; Tanna (CM 62.4302 ex Geale, ANSP 69174, paratype of perstrigata); Erromanga (AMNH, Macmillan). Photograph of syntype (BM). Remarks. — Typical turneri is elongate and perstrigata is short and obese. Both variations were found in specimens from Tanna and Erromanga, but each lot was assignable to one variety or the other. No mixed lots were seen. The syntype of P. turneri (pi. 9, fig. 9), however, is intermediate between the elongate and the obese series and the separation based on material in American museums cannot be maintained. P. turneri has no supra-sutural cord and the mark- ings are much lighter than those of P. macgillivrayi. The type local- ity is uncertain. The original specimens of P. turneri were obtained from the Rev. George Turner. It could not be ascertained if Turner actually visited Erromanga, but the specimens might have been col- lected by Samoan missionary teachers. 68 FIELDIANA: ZOOLOGY, VOLUME 43 Partula caledonica Pfeiffer. Plate 9, fig. 2; plate 14, figs. 1-4. Partula caledonica Pfeiffer, 1861, Proc. Zool. Soc. London, 1861: 389 — New Caledonia (error); Brazier, 1871, Proc. Zool. Soc. London, 1871: 585 — correction of type locality to Havannah Harbour, Vate, and report of it from Vanua Lava, Banks Group; Pilsbry, 1909, Man. Conch., (2), 20: 278-279, pi. 33, figs. 12-14. Partula pfeifferi Crosse, 1871, Jour, de Conchy., 19: 184 — substitute name. Range. — Vate. Material. — Vate (CM 62.4287, ex Brazier), Havannah Harbor, Vate (CM 62.4286, CM 62.4288), Vila, Vate (ANSP 133297, ex Frog- gatt; USNM 598361 and Miller 542, coll. by W. B. Miller). Remarks. — P. caledonica is the only New Hebridean partulid of which series of shells were available. Two lots were from near Vila; Froggatt's shells were of uncertain locality and Miller's came from the banks of a small stream about two miles north of Vila. Two lots were from the type locality, Havannah Harbor (CM 62.4286, CM 62.4288, collected by Brazier in 1865). All the specimens cor- responded well to the syntype (pi. 9, fig. 2). Four of the ten speci- mens collected by Miller are illustrated (pi. 14, figs. 1-4) to show the infra-populational variation in shell contour and apertural shape. One specimen of P. caledonica from near Vila (pi. 14, fig. 4) is similar to P. pyramis, but the sutures are less deeply impressed. The only character separating P. caledonica from P. turneri is the development of a subsutural cord and darker coloration in P. caledonica. Brazier's (1871, p. 585) report of P. caledonica from Vanua Lava needs confir- mation. Hartman (1886, p. 34) described P. proxima from specimens collected on Vanua Lava by Brazier, and the record of P. caledonica is probably based on the same specimens. SuPERSPEClES OF Partula pyramis Shell larger than in the macgillivrayi series (usually more than 23 mm. high). Sculpture more prominent, continuing over the last whorl. Suture well impressed and whorls strongly rounded. Parietal and columellar calluses well developed. The coloration is unknown, as most individuals seen have been quite worn. Sykes (1903, p. 198) first postulated the unity of these species and study of type specimens and photographs confirms his conclu- sion. The pyramis complex has a slightly more elongate shell with more impressed sutures than the macgillivrayi series. The differences are more of degree than sharp separation and the two groups may be ecological variations, rather than distinct taxonomic entities. The worn condition of the specimens of the pyramis "superspecies" SOLEM: MOLLUSCA OF THE NEW HEBRIDES 69 may indicate that they have been washed down from higher eleva- tions and collected from stream drift. The elongate spire and impressed sutures could then be correlated with the same type of variation found in Diplomorpha (see pp. 139-141). Partula pyramis Hartman. Plate 14, fig. 9. Partula pyramis Hartman, 1886, Proc. Acad. Nat. Sci. Philadelphia, 1886: 34, pi. 2, fig. 12— Vate; Sykes, 1903, Proc. Malac. Soc. London, 5: 198— Terebu and Renee Rivers, Espiritu Santo(?); Pilsbry, 1909, Man. Conch., (2), 20:281-282, pi. 33, fig. 7. Range. — Vate (the Espiritu Santo record is not accepted). Material. — Vate (CM 62.4305, holotype and paratype). Remarks. — The aperture and body whorl of P. pyramis are more swollen and the whorls more rounded than in P. eximia. Otherwise the two species are very similar. The Espiritu Santo records for P. pyramis are probably based on specimens of P. albescens Hart- man, if, indeed, the two should prove to be distinct. Partula proxima Hartman Partula proxima Hartman, 1886, Proc. Acad. Nat. Sci. Philadelphia, 1886: 34, pi. 2, fig. 11 — Vanua Lava, Banks Group; Pilsbry, 1909, Man. Conch., (2), 20: 286-287, pi. 34, figs. 12, 13. Range. — Vanua Lava, Banks Group. Material. — Type photograph (CM). Remarks. — The type specimens are in the Australian Museum, Sydney. A photograph of the holotype in the Carnegie Museum was too faded to be reproduced. P. proxima is quite similar to P. caledonica but has sufficiently impressed sutures and elongate spire to be placed in the pyramis series. As mentioned in the dis- cussion of P. caledonica, the record of that species from Vanua Lava may have been based on the specimens Hartman later described as proxima. Partula eburnea Hartman Partula eburnea Hartman, 1886, Proc. Acad. Nat. Sci. Philadelphia, 1886: 33, pi. 2, fig. 10— no locality; Pilsbry, 1909, Man. Conch., (2), 20: 281, pi. 33, fig. 8. Range. — Locality unknown. Material. — No material available. Remarks. — The type specimens are in the Australian Museum, Sydney. A note on a photograph (ANSP) indicates that Hartman 70 FIELDIANA: ZOOLOGY, VOLUME 43 considered P. ehurnea to be identical with P. macgilUvrayi. P. ehur- nea has been placed in the pyramis series on the basis of its slender, produced spire and rounded whorls. Partula eximia Hartman. Plate 14, figs. 5, 6. Partula eximia Hartman, 1886, Proc. Acad. Nat. Sci. Philadelphia, 1886: 35, pi. 2, fig. 14— Aneiteum (ex Thomson, Layard); Pilsbry, 1909, Man. Conch., (2), 20: 280-281, pi. 33, fig. 11. Range. — Aneiteum. Material. — Aneiteum (CM 62.4292, holotype and paratype). Remarks. — P. eximia differs from P. proxima in being less obese and in having the body whorl more swollen. Partula albescens Hartman. Plate 14, figs. 7, 8. Partula albescens Hartman, 1888, Proc. Acad. Nat. Sci. Philadelphia, 1888: pi. 13, fig. 4 — Aura (=Aore) and Sitova (=Tutuba) Islands (Layard); Pilsbry, 1909, Man. Conch., (2), 20: 282-283, pi. 33, figs. 9-10. Range. — Aore and Tutuba Islands off Espiritu Santo. Material. — Aore (CM. 62.4290, holotype and paratypes) ; Tutuba (ANSP 60534); Aolia (=Aore?, Aoba?) (ANSP 144263); New Heb- rides (USNM 608987 ex Quadras). Remarks. — P. albescens comes close to forming a connecting link between the macgilUvrayi and pyramis series. The shells from Tutuba (pi. 14, fig. 7) are shorter and proportionately broader, and have a narrower lip than those from Aore. Population studies may show that Tutuba has a local race, but taxonomic recognition on the basis of two shells is not warranted. SUPERSPECIES OF Partula auraniana Shell thin, smaller than the macgilUvrayi and pyramis superspecies. A broad white band is below the suture; apex and spire brownish, fading to greenish-yellow on body whorl. Spiral sculpture prominent. The coloration and strong spiral sculpture serve to separate the auraniana superspecies from the other two. The auraniana complex has not yet been found in the central and southern New Hebrides. Partula auraniana Hartman. Plate 15, figs. 1, 2, 6. Partula auraniana Hartman, 1888, Proc. Acad. Nat. Sci. Philadelphia, 1888: 250, pi. 13, fig. 1— Aore Island; Sykes, 1903, Proc. Malac. Soc. London, 5: 198 — Lo and Hiu Islands, Torres Group (J. J. Walker); Ancey, 1905, Nautilus, 19, (4), p. 44— Buka-Buka, Torres Group; Pilsbry, 1909, Man. Conch., (2), 20: 284-285, pi. 34, figs. 7-10. SOLEM: MOLLUSCA OF THE NEW HEBRIDES 71 Range. — Aore near Espiritu Santo and Lo, Hiu, and Buka-Buka Islands in the Torres Group. Material.— Aore (CM 62.4295, holotype and paratype; UMMZ 145681, ex B. Walker, Tomlin); Hiu Island (UMMZ 145682, ex B. Walker, Tomlin, Sykes). Remarks. — P. auraniana is smaller and has fewer whorls than P. fraterna Hartman and P. carnicolor Hartman. The holotypes of the three species seem quite different, but enough intergradation could be found among the paratypes to make it probable that the "species" represent infraspecific variations rather than distinct spe- cies. Sykes (1903, p. 198) considered that the specimens from Lo Island, Torres Group, represented a local race. I saw no material from this island. Partula fraterna Hartman. Plate 15, figs. 3, 4. Partula fraterna Hartman, 1888, Proc. Acad. Nat. Sci. Philadelphia, 1888: 250, pi. 13, fig. 2— Aore Island; Sykes, 1903, Proc. Malac. Soc. London, 5: 198 — Ravenga, Vanua Lava and Lakona, Gaua (Banks Group) ; Pilsbry, 1909, Man. Conch., (2), 20: 285-286, pi. 34, fig. 4. Range. — Aore, Vanua Lava, and Gaua. Material. — Aore (CM 62.4294, holotype and paratype). Remarks. — P. fraterna has more whorls, and a more capacious um- bilical fissure, and is larger than P. auraniana. P. carnicolor is larger and more obese, and has a columellar sinuation which is lacking in P. fraterna. No specimens of P. fraterna from the Banks Islands were seen. Partula carnicolor Hartman. Plate 15, figs. 5, 9. Partula carnicolor Hartman, 1888, Proc. Acad. Nat. Sci. Philadelphia, 1888: 250, pi. 13, fig. 3— Aore Island; Pilsbry, 1909, Man. Conch., (2), 20: 286, pi. 34, figs. 1-3. Range. — Aore Island. Material— Aore (CM 62.4289, holotype and paratype). Remarks. — The size, elongate spire, and columellar sinuation sep- arate P. carnicolor from P. auraniana and P. fraterna. The paratype (see Table II) is much smaller than the holotype and forms a con- necting link with other members of the auraniana superspecies. The second series of Partula is found in northern Melanesia. Two species (P. minor Hartman and P. milleri, new sp.) have been re- 72 FIELDIANA: ZOOLOGY, VOLUME 43 ported from the New Hebrides, P. minor may be a mislabeled Solo- mon Island specimen, but P. milleri represents a valid record from Espiritu Santo. P. vanikorensis from the Santa Cruz Islands also belongs here. Partula minor Hartman. Plate 15, fig. 7. Partula minor Hartman, 1886, Proc. Acad. Nat. Sci. Philadelphia, 1886: 31, pi. 2, fig. 5— Erromanga (doubtful); Pilsbry, 1909, Man. Conch., (2), 20: 287-288, pi. 34, figs. 5, 6, 14. Range. — Erromanga (?) . Material. — Erromanga (CM 62.4243, holotype and paratype). Remarks. — P. minor is very similar to P. pellucida Pease and P. coxi Pease from the Solomon Islands. The sculpture of the last two "species" is more prominent than that of P. minor, but other- wise they are similar. Hartman (1886) described P. hastula and P. minor from Erromanga. Later, the type lot of P. hastula was found to have been collected on Eddystone Island in the Solo- mons. It is possible that P. minor has an erroneous locality, but until Erromanga has been explored for partulids, it would be better to consider P. minor a dubious member of the New Hebridean fauna, instead of rejecting the locality completely. Partula milleri, new species. Plate 15, fig. 8; plate 16, fig. 4. A species of Partula separated from P. regularis Hartman and P. minor Hartman by its very strong spiral sculpture, prominent suprasutural cord, and obese shape. Shell ovate-conic, thin, translucent, nearly devoid of color. Whorls 4^, rounded, without trace of a carina. Sutures lightly impressed. Body whorl lat- erally compressed behind the aperture. Sculpture of numerous dense, wavy, spiral lines distinctly narrower than the interstices. A very prominent suprasutural cord on all but embryonic whorls. Aperture ovate, parietal callus thick, slightly opaque. Columellar wall less angulate than in regularis. Lip expanded, thickened internally. Umbilicus open. Height 15.8 mm., diameter 9.1 mm., aperture 8.6 by 5.4 mm. Type. — United States National Museum no. 619738. Collected on the south side of Pallikula Bay, Espiritu Santo, New Hebrides, by Walter B. Miller. Remarks. — P. milleri is most closely related to P. regularis Hart- man from Guadalcanal in the Solomons. P. regularis is less obese, has a less prominent suprasutural cord and weaker spiral sculpture in which the striae are as wide as their interstices. P. milleri is the first species of the northern Melanesian type of Partula to be defi- nitely reported from the New Hebrides. SOLEM: MOLLUSCA OF THE NEW HEBRIDES 73 It is named after the collector, Commander Walter B. Miller of Falls Church, Virginia. Partula vanikorensis (Quoy and Gaimard). Plate 12, figs. 1, 2. Helix vanikorensis Quoy and Gaimard, 1832, Voy. I'Astrolabe, Zool., 2: 116, pi. 9, figs. 12-17 — Vanikoro, Santa Cruz Islands (found under the leaves of trees at the abandoned village of Ocili). Partula vanikorensis (Quoy and Gaimard), Pilsbry, 1909, Man. Conch., (2), 20: 289-290, pi. 35, figs. 15-17. Range. — Vanikoro, Santa Cruz Islands. Material— Yanikoro (MCZ 24844, MCZ 192352, ex Stevenson, Fox). Remarks. — On the basis of the figures in Quoy and Gaimard (1832-35), I had considered that vanikorensis might belong to the genus Samoana Pilsbry (1909). Of the two available specimens, one (MCZ 24844 ex British Museum) is probably a mislabeled Partula otaheitana sinistrorsa Pease from Tahiti, and the other (MCZ 192357) was very similar to Samoana alahastrina (Pfeiffer) from Fiji (CM 62.4293, cotypes of Partula nematoraphe) . The Vanikoro specimen had a less widely reflexed columellar whorl, much less prominent suprasutural cord, stronger spiral sculpture, and a more thickened lip. The apical sculpture is like that shown for Partula attenuata Pease (see Pilsbry, 1909, pi. 24, fig. 3). Photographs of the type and a cotype furnished by Dr. Andr^ Franc of the Paris Museum (pi. 12, figs. 1, 2) showed specimens com- pletely different from the figures in Quoy and Gaimard, and very similar to Partula milleri in general appearance. The specimens are so unlike the original figures that it is difficult to believe that there has not been an error, either in the illustrations or in the selection of type specimens. Without new material from Vanikoro and personal examination of all the Paris Museum Partulidae, it is impossible to clarify the identity of Helix vanikorensis or the status of the Vanikoro popula- tion. As a temporary expedient, they are grouped under the name Partula vanikorensis (Quoy and Gaimard). Iredale (1927, p. 74) reported that a Partula was collected on Santa Cruz Island, but it has never been described or figured. Several species of partulids have been erroneously reported as coming from the New Hebrides. Others were so inadequately de- 74 FIELDIANA: ZOOLOGY, VOLUME 43 scribed and figured that they cannot be recognized. References are included for completeness. Partula concinna Pease Partulaconcinna Tease, 1871, Amer. Jour. Conch,, 7:196 — Tanna (Cox); Hart- man, 1886, Proc. Acad. Nat. Sci. Philadelphia, 1886: 35, pi. 2, fig. 16; Pilsbry, 1909, Man. Conch., (2), 20: 288, pi. 36, figs. 9, 12. Hartman's single specimen (CM 62.4244, ex Hartman, Taylor) from "Tanna" probably is a mislabeled Partula taeniata nucleola from Moorea, Society Islands (see Pilsbry, 1909, p. 288, and H. H. Smith, 1902, p. 455). Partula glaber Hartman Partula glaber Hartman, 1885, Proc. Acad. Nat. Sci. Philadelphia, 1885: 205, 1 fig. — no locality except "New Hebrides(?)." This is the Peruvian Drymaeus strigatus, var. purus (Pilsbry, 1909, p. 319). Partula hastula Hartman Partula hastula Hartman, 1886, Proc. Acad. Nat. Sci. Philadelphia, 1886: 33, pi. 2, fig. 9 — Erromanga, Solomon Islands (sic) (Brazier); Pilsbry, 1909, Man. Conch., (2), 20: 291, pi. 35, figs. 6-8— Simbo, Eddystone Island, Solomon Islands. The correction of type locality is based on a note in Hartman's handwriting found with the original label. Partula hollandiana Pilsbry Partula hollandiana Pilsbry, 1909, Man. Conch., (2), 20: 293-294, pi. 37, figs. 8-10 — locality unknown. This is very similar to Partula regularis Hartman from Guadal- canal, Solomon Islands, and may be a local race of that species. Partula nematoraphe Pilsbry Partula nematoraphe Pilsbry, 1909, Man. Conch., (2), 20: 279-280, pi. 35, figs. 1-3 — unknown locality, "but the shell has wholly the appearance of the New Hebridean Partulae." Pilsbry and Cooke (1934a, p. 17) synonymize this species with P. alabastrina Pfeiffer from Moala, Fiji Islands. Partula repanda Pfeiffer. Plate 9, fig. 7. Partula repanda Pfeiffer, 1855, Proc. Zool. Soc. London, 1855: 98 — New Heb- rides (Cuming collection); Pilsbry, 1909, Man. Conch., (2), 20: 288-289, pi. 34, fig. 11. SOLEM: MOLLUSCA OF THE NEW HEBRIDES 75 Hartman considered that Partula repanda might be related to the pale varieties of P. recta Pease from the Marquesas, but com- parison of recta with the type of repanda in the British Museum indi- cated that they were distinct (see H. H. Smith, 1902, p. 459). The photograph of the holotype (pi. 9, fig. 7) shows a shell that resembles the Society Island partulid more than any Melanesian species. Partula turricula Pease Partula turricula Pease, 1871, Amer. Jour. Conch., 7: 196 — New Hebrides(?) (received from Cox); Pilsbry, 1909, Man. Conch., (2), 20: 283-284. The lack of transverse striae and the sinistral coiling separate P. turricula from macgillivrayi and caledonica. The location of the type of P. turricula is unknown and the species has never been figured. Order SIGMURETHRA Pilsbry, 1900 Snails in which the ureter of the kidney is abruptly reflexed, passing to the posterior end of the lung cavity. An open groove or closed tube (secondary ureter) continues across to the last fold of the gut, then follows it forward to the mantle edge. (After Pilsbry, 1900a, pp. 562-563.) Family relationships within the Sigmurethra are complex and little understood. The dominant land snails of continental areas belong to the Sigmurethra and adaptive radiations have been numer- ous and extensive. Much of the confusion as to distribution and relationships has resulted from the refusal of certain conchologists to recognize that species whose shells differ because of ecological adap- tations may actually be very closely related (see Pilsbry, 1894, pp. viii, xiv). Conversely, strikingly similar shells may belong to differ- ent families and resemble each other only because they are adapted for similar ecologic niches. The shells of land snails are very plastic, and form alone is unreliable in classification. Use of the divisions Aulacopoda and Holopoda (see p. 36) recog- nizes two major overlapping distributional patterns which probably reflect a basic evolutionary divergence. Suborder AULACOPODA Pilsbry, 1896 Sigmurethrous snails in which the pedal groove is conspicuously impressed and situated well above the lateral angle of the foot, which is bordered by a band of the sole (Pilsbry, 1946b, p. 231). Pilsbry (1896, p. 110) recognized two family groups in the Aula- copoda. Division I, containing the Zonitidae, Helicarionidae, and 76 FIELDIANA: ZOOLOGY, VOLUME 43 Limacidae, was named the superfamily Limacacea by H, B. Baker (1941, p. 206) ; division II, containing the Endodontidae, Arionidae, and Philomycidae, was called the superfamily Arionoidea by H. B. Baker (1955, p. 109). The name Arionoidea is here emended to Arionacea to conform with the usual superfamilial endings. Shell-bearing limacoids and arionoids differ in the general appear- ance of the shell, but anatomical differences between the superfami- lies are small because of convergent evolution in slug taxa. The most constant difference is in the shape of the marginal teeth. The Limacacea have marginal teeth with narrow, elongate, basal plates and the teeth are unicuspid or bicuspid as a result of the elevation of the outer cusp on the middle cusp. The Arionacea have marginal teeth with short, squarish, basal plates and with either one or sev- eral cusps. Radular teeth are usually of little value in delineating categories at the family level, but the Limacacea and Arionacea have slug taxa which are convergent in most other characters. The persistent prim- itiveness of the marginal teeth in the Aulacopoda, despite the great divergence in centrals and laterals, is in keeping with the law of meso- metamorphosis: "All modifications in the teeth proceed from the median line of the radula outwards toward the edges, the outer mar- ginal teeth being the last to be modified." (Pilsbry, 1894, p. xiii.) The Limacacea and Arionacea have shells which are easily distin- guished, while the slugs can be separated by anatomical details. Despite the convergent slug taxa, recognition of the two superfam- ilies seems both desirable and valid. Both the Limacacea and Arionacea are represented in the New Hebrides. With very few exceptions, the Limacacea are at their southern limit of distribution, but the Arionacea are important con- stituents of the fauna of New Caledonia and New Zealand. On the basis of distribution and unspecialized structure, it seems probable that the Arionacea are the more primitive taxon and stand nearer the sigmurethran ancestor than either the Limacacea or the Holopoda (Pilsbry, 1894, p. xxxix). On continental areas the shell-bearing limacoids are much more numerous, and they appear to be replacing the shell-bearing arionids. The latter still seem to be dominant in Tasmania, New Zealand, New Caledonia, and many oceanic islands. Superfamily ARIONACEA Marginal teeth with short, wide and squarish basal-plates with one or several cusps, the outer cusp never elevated on middle cusp (Pilsbry, 1896, p. 110). SOLEM: MOLLUSCA OF THE NEW HEBRIDES 77 Two of the three famih'es, the Arionidae and Philomycidae, are of limited distribution and obviously derived from the third family, the Endodontidae. The Arionidae are Holarctic, with genera in Europe, Asia {Arion in Siberia and Anadenus in the Himalayas and China), western North America, northern Africa, and South Africa (Oopelta) (Pilsbry, 1948, p. 661). They are restricted to areas of high humidity and seem far less adaptable than either the limacoid or the helicoid slugs. A few species of Arion have been disseminated by commerce. In the more humid, temperate regions of America, South Africa, New Zealand, and Australia they have become well established. In the Philomycidae the mantle covers the entire body, and the shell has been completely lost. Philomycus is restricted to eastern North America from Canada to Florida and northeastern Texas. Pallifera is found in the northeastern United States as far south as North Carolina and in Central America from Mexico to Colombia (Pilsbry, 1948, pp. 748-770). Meghimatium, which is very closely related to Pallifera, is found in Japan, China, Tonkin, India, and Indonesia east to Borneo and the Celebes (van Benthem Jutting, 1952, p. 424). The ecological relationship between Meghimatium and the veronicellid slugs is unknown. It is not improbable that they compete, since both are vegetarians and restricted to similar humid situations. The Endodontidae probably are the most primitive living sig- murethrans. They certainly represent the most widely distributed family of land mollusks. Unfortunately, the minute size of the ani- mals and their distribution in the Southern Hemisphere has pre- vented any extensive survey of the entire family as to anatomical structure. Recent studies (Iredale, 1913, 1915, 1933, 1937a; Suter, 1913; and Dell, 1952a) have shown that many taxa are well charac- terized by sculptural differences of the shells. I found it necessary to make a comprehensive survey of the Pacific Endodontidae in order to place the New Hebridean species in genera. Family ENDODONTIDAE Shell umbilicate, depressed-heliciform to discoidal, many species flammulate or banded. Aperture rounded-lunate or compressed, lamellate in many species; peristome sharp, expanded only near the columellar margjin. Sculpture of various combinations of radial, spiral, or oblique ribs and/or striae. Foot with pedal grooves meeting above the tail; sole not divided longitudinally. Lung without noticeable venation other than the pulmonary vein. Kidney triangular, squarish, or U-shaped. Genitalia simple; spermatheca on a long duct. Jaw arcuate, entire, 78 FIELDIANA: ZOOLOGY, VOLUME 43 or formed of loosely connected squarish plates. Rachidian tooth tricuspid, nar- rower than or nearly equal to the laterals. Laterals bi- or tri-cuspid; marginals with wide, short basal plates and one to several cusps. (Modified from Pilsbry, 1948, pp. 565-566.) Intergeneric relationships of the Endodontidae are very poorly understood. Iredale (1937a) recognized by implication at least six families, while Thiele (1931) divided the single family into eight sub- families. As emphasized by Pilsbry (1948, p. 566), the anatomy of only a very few genera is known. Although two markedly distinct subfamilies can be recognized, most of the genera must still be placed in a large subfamily, the Endodontinae. Species with the jaw formed by loosely connected plates belong to the subfamily Punctinae Morse, 1864 (=Laominae Suter, 1913), which Iredale (1937a) raised to family rank. The Punctinae (fig. 13) include the Holarctic and South African Punctum, Neotropical Radio- discus, Australian "Laomidae" (Iredale, 1937a, 1939), New Zealand "Laomidae" (Powell, 1946a), Kermadec Island Paralaoma (Iredale, 1913), and probably the New Hebridean Phrixgnathus reported below. The North American Helicodiscus and Mexican Chanomphalus include discoidal species in which the whorls do not substantially increase in size. Possibly Polygyriscus Pilsbry (1948, p. 1097) also belongs to the Helicodiscinae. Stenopylis (fig. 24) from the Indo- Melanesian region (see Solem, 1957) has very similar shell structure, but also a prominent rachidian tooth, less prominent endo- and ecto- cones on the laterals, and a jaw composed of separate (?) plates. The difference between the jaws of Stenopylis and Helicodiscus is partially bridged by the stegognath jaw of Hebetodiscus, a subgenus of Helico- discus (see Pilsbry, 1948, p. 635). Further study may indicate that Stenopylis should be placed in the Helicodiscinae, rather than in a separate subfamily (Thiele, 1931, p. 569) or family (Iredale, 1937b, pp. 1-2). For the other genera, there seem to be no criteria by which easily recognizable subfamilies can be delineated, Suter (1913) attempted to divide the New Zealand species into families based on the presence (Flammulinidae) or absence (Endodontidae) of a caudal foss. H. B. Baker (1941, p. 205) and Pilsbry (1896, p. 109, and 1946b, p. 232) have pointed out the slight utility of the caudal foss in taxonomy, and Suter's divisions are not accepted here. In regard to the generic affinities of the Pacific "Endodontinae," numerous changes seem necessary. Essentially there are four types of apical sculpture found in the Pacific species: radial ribs only; radial SOLEM: MOLLUSCA OF THE NEW HEBRIDES 79 and spiral ribs; spiral ribs; and smooth apical whorls. It is probable that the species with smooth apices were derived from one of the other three types, and other characters will have to be used to deter- mine their exact affinities. When used in conjunction with the sub- sidiary characters of apertural armature and shape, a reasonably logical classification results. Full elucidation of the classification is not possible at this time, but the broad outlines are an important preamble to the zoogeographic survey. The sculpture of spiral striae is found throughout most of the Pacific. In the Indonesian region there is the minute toothed Beila- nia (see Solem, 1957) which is probably ancestral to all the Microne- sian species. Beilania (fig. 25) possibly reaches New Guinea and the Bismarcks but not Australia or Polynesia. In Polynesia, New Zea- land, western Australia, the Solomon Islands, and the New Hebrides there is a larger toothless series of species which have relatively few spiral ribs on the apical whorls. The earliest name for this group seems to be Mocella Iredale, 1915 (fig. 25). In New Caledonia and eastern Australia (including Tasmania) species with many more spiral ribs are found (fig. 25). On the basis of evidence from the Microne- sian derivatives of Beilania, it seems probable that the pattern of few spiral lines is primitive, that of many spiral lines advanced (see p. 298). Consideration of the interrelationships of the New Cale- donian and East Australian species is well beyond the scope of this study. The radially ribbed apical sculpture is represented by two main groups, Discocharopa and Ptychodon. Discocharopa (fig. 24) is minute, greasy white in color, usually very widely umbilicate, and either toothed or toothless (see Solem, 1957). Ptychodon is larger, horn- colored or flammulated, has many apertural lamellae and a narrower umbilicus. It is found in Polynesia and New Zealand (fig. 26). A few Australian species with radially ribbed apices have been placed in "genera" by Iredale and are not considered at this time. The species with both radial and spiral ribs on the apical whorls form two series. In the Indonesian-New Guinean complex (Solem, 1958b) the ribs are much modified; in the New Hebridean endemic, Retickaropa, the radial and spiral ribs are nearly equal in size. Pos- sibly the New Guinean endemic, Paryphantopsis, is derived from the Indonesian forms with radial and spiral ribs (see Solem, 1958b). Of the taxa with smooth apical whorls the New Zealand Charopa (fig. 24) may be derived from Mocella, and Endodonta, Nesophila, and Nesodisctis from Ptychodon (see figs. 24, 26). 80 FIELDIANA: ZOOLOGY, VOLUME 43 The above sketch of generic affinities gives no idea of relation- ships between the taxa with different types of apical sculpture and only a brief outline of the distribution and classification. It is hoped that study of this family can be pursued at a later date with more material than is available at the present time. The New Hebridean Endodontidae belong to four genera: the endemic Reticharopa; the punctid genus Phrixgnathus, found else- where only in New Zealand; Mocella, which ranges throughout most of Polynesia, Melanesia, New Zealand, and western Australia; and Discocharopa, which has a scattered distribution from the Philippines to the Kermadec Islands. Of the nine species which I've examined, only three had been pre- viously described. Five of the nine were found only in stream drift deposit (ML 95) on Espiritu Santo. A key, based on apical sculp- ture and number of ribs, is given below. There are published records for one additional species, Charopa perryi Smith, which has never been figured. Key to the New Hebridean Endodontidae 1. Body whorl keeled; no prominent radial ribs; apex smooth 2 Body whorl rounded or angulated; with prominent radial ribs; apex with dis- tinct sculpture (except in worn specimens) 3 2. Body whorl sharply carinated; diameter 6 mm. Phrixgnathus tenuiscripta (Ancey) Body whorl bluntly keeled; diameter less than 5 mm. Phrixgnathtis glissoni (Ancey) 3. Spire elevated or in plane of body whorl 4 Spire distinctly depressed Reticharopa sp. 4. Apical sculpture of radial ribs or reticulated 5 Apical sculpture of spiral ribs only Mocella euryomphala, new sp. 5. Apex reticulated; umbilicus contained more than three times in diameter. . .6 Apex radially ribbed; umbilicus contained twice in diameter. Discocharopa planulata, new sp. 6. Ribs on body whorl 35-40 7 Ribs on body whorl more than 65 8 7. Body whorl angulated; diameter 3 mm.; Aneiteum . Reticharopa geddiei, new sp. Body whorl rounded; diameter 2 mm.; Espiritu Santo. Reticharopa latecosta, new sp. 8. Ribs on body whorl 70-75; Espiritu Santo. .Reticharopa stenopleura, new sp. Ribs on body whorl about 110; Aneiteum Reticharopa helva (Cox) Subfamily PUNCTINAE Genus PHRIXGNATHUS Hutton, 1883 Shell imperforate to widely umbilicate, trochoidal to planulate, body whorl rounded to sharply keeled. Sculpture smooth to strongly radially ribbed. Aper- SOLEM: MOLLUSCA OF THE NEW HEBRIDES 81 ture toothless. Animal without caudal mucous pore. Jaw composed of papillose, imbricating plates. Radula with narrow uni- or tri-cusped central tooth, bicuspid laterals, and low, wide, bicuspid marginals. (Adapted from Suter, 1913, p. 733.) Type species. — Phrixgnathns celia Hutton. Remarks. — Detailed study of the New Zealand Phrixgnathus may indicate that the genus, as now constituted, is polyphyletic. How- ever, inspection of the material in the University of Michigan Museum of Zoology suggests that a complete series of transitional species exists between the smooth, trochoidal type, P. celia; the radially ribbed, helicoid P. allochroida Suter; and the smooth, planulate, widely umbilicate P. sciadium (Pfeiffer). The soft parts of very few species have been examined and divisions based on anatomical features are not yet possible. Two New Hebridean "Endodonta" glissoni Ancey and tenui- scripta Ancey, are very similar to Phrixgnathus sciadium. Although the animals have not been studied, the conchological similarity is so great that I feel little hesitation in placing them in Phrixgnathus. Phrixgnathus glissoni is easily separated from P. tenuiscriptus by the larger size and sharp, not rounded, carina of the latter species. Phrixgnathus glissoni (Ancey). Plate 28, figs. 3-5. Patula glissoni Ancey, 1889, Le Naturaliste, 11: 50 — Sea View Estate, Vate Island (Glisson!). Endodonta glissoni Ancey, 1896, Nautilus, 10, (8), p. 90. Range. — Vate, Espiritu Santo. Material.— ML 68, ML 95; Vate (UMMZ 136690, ex Walker, Ponsonby, Layard, from the type lot). Remarks. — The few specimens examined show a considerable range of variation in whorl convexity, degree of carination, and H/D ratio. Ancey 's variety "B" represents one of the more tro- choidal, less sharply carinate specimens, and is probably taxonomi- cally insignificant. No constant differences could be found between the Espiritu Santo and Vate shells. The larger specimens have A}4 to 43^ whorls and are 4.5 to 4.8 mm. in diameter. Phrixgnathus tenuiscriptus (Ancey). Plate 28, figs. 1, 2. Endodonta tenuiscripta Ancey, 1896, Nautilus, 10, (8), p. 90 — Malekula (ex Layard); Ancey, 1905, op. cit., 19, (4), p. 42. Range. — Malekula, Espiritu Santo. Material.— ML 95. 82 FIELDIANA: ZOOLOGY, VOLUME 43 Remarks. — A single, broken, juvenile specimen (33/^ whorls) seems referable to tenuiscriptus. The knife-edge keel and close-set radial ribs of the shell easily separate tenuiscriptus from glissoni and were the characters used by Ancey in his original diagnosis. In P. glis- soni the keel is rounded; there is a supracarinal groove; the sculpture consists of a few indistinct, broad ribs at irregular intervals; and the shell is about 4.7 mm. in diameter. In P. tenuiscriptus the keel is produced into a knife-edge carina; the sculpture consists of promi- nent, close-set radial ribs at regular intervals; and the shell is about 6.5 mm. in diameter with five whorls (Ancey, 1896, p. 90). Subfamily ENDODONTINAE Genus DISCOCHAROPA Iredale, 1913 Shell minute, widely umbilicated, discoidal, with numerous very close-set radial ribs. Apical sculpture continued on rest of shell. No spiral sculpture. Aperture toothed or toothless. Type species. — Discocharopa exquisita Iredale. Remarks. — Several species from various parts of the Pacific are grouped under Discocharopa. They agree in sculpture, discoidal shape, and usually wide umbilicus. The aperture is toothed in two species, D. exquisita Iredale and D. werneri Solem (1957), and tooth- less in the others. The anatomy is unknown. Several specimens of a Discocharopa were found in stream drift collected on Espiritu Santo. Comparisons with paratypes of most of the other species indicated that they belong to a new species which is described below. Discocharopa planulata, new species. Plate 32, figs. 1-3. A species of Discocharopa characterized by its slightly elevated spire, very close-set radial ribs, relatively gradual whorl increase, and toothless aperture. Shell minute, discoidal, spire only slightly raised above plane of body whorl. Whorls 3-4, gradually increasing in size. Aperture ovate, slightly flattened above. Umbilicus widely open, shallow, contained 1.96 times in the diameter. Sculpture of numerous close-set, slightly retractive radial ribs; about 145 on body whorl. Interstices with several very fine radial riblets. No spiral sculpture. Diameter of holotype 1.45 mm., lesser diameter 1.29 mm., height 0.55 mm., with 3^ whorls. Type. — University of Michigan Museum of Zoology no. 186037. Collected from stream drift in the Sarakata River Valley, Espiritu Santo (ML 95), by Robert E. Kuntz in May, 1944. Paratypes. — From ML 95, specimens distributed as follows: UMMZ 186036, CNHM 54904, BPBM 212378, MCZ 186826. SOLEM: MOLLUSCA OF THE NEW HEBRIDES 83 Remarks. — Discocharopa planulata differs from the Kermadec Island D. exquisita by its toothless aperture, more closely set ribs, and slightly elevated spire; it differs from the Indonesian D. micro- discus van Benthem Jutting by its less rapidly increasing whorls and less elevated spire. Genus MOCELLA Iredale, 1915 Small, tightly coiled shells with slightly elevated spire. Umbilicus moderately open, deep. Apical sculpture of 8 to 12 spiral striae; spire and body whorl with slightly retractive radial ribs. Interstices with fine radial riblets crossed by spiral striae (pi. 31, fig. 4). Aperture toothless, ovate; apertural edge slightly flattened above. Type species. — Helix corniculum Reeve (=Mocella cogitata Ire- dale, 1941, p. 91). Remarks. — Mocella is equivalent to Charopa as used in previous publications on the endodontids of Melanesia and Polynesia. Cha- ropa must be restricted to New Zealand species that have smooth apical whorls. The Polynesian and Melanesian species have spirally striated apices, as do the New Zealand species that Iredale placed in Mocella. In the absence of conflicting evidence, they are considered congeneric. It is probable that the west Australian species that Ire- dale (1937a) placed in Luinodiscus are not more than subgenerically separable. Species belonging to Mocella have been found from the Bismarcks to the Society Islands, but are not known from the Tuamotus, Mar- quesas, Australs, or New Caledonia. Several shells from Espiritu Santo represent the following new species: Mocella euryomphala, new species. Plate 6, figs. 7, 8; plate 31, figs. 1-5. A species of Mocella most similar to M. solomonensis (Clapp) but easily separated from that and all other species by the comparatively large umbilicus. Shell small, depressed-helicoid, spire only slightly elevated. Whorls 3 ^ to 4 ^, tightly coiled, gradually increasing in size, sutures deeply channeled. Aperture ovate, flattened dorso-laterally. Umbilicus open, deep, contained about 3.2 to 3.5 times in the diameter. Apical whorls IJ^, with about eleven spiral ribs (pi. 31, fig. 5). Spire and body whorl with strong radial ribs (ca. 90 on body whorl). Numerous fine riblets crossed by spiral lines between primary ribs (pi. 31, fig. 4). Color light reddish-horn. Diameter 2.5 to 2.8 mm., height 1.3 to 1.4 mm. Type. — University of Michigan Museum of Zoology no. 186042. Collected at Brigstock Point, Espiritu Santo, New Hebrides (ML 63), under logs, by Robert E. Kuntz on April 25, 1944. 84 FIELDIANA: ZOOLOGY, VOLUME 43 Paratypes. — On Espiritu Santo, Kuntz collected specimens at the following stations: ML 40, ML 63 (type locality), ML 78, and ML 95. Paratypic specimens are CNHM 54906, MCZ 186828, BPBM 212379, and ANSP. An additional shell from Espiritu Santo (USNM 432458, Harrington !) is also a paratype. Remarks. — Many "species" of Mocella will probably be synony- mized when the genus has been critically reviewed. M. euryomphala was compared with all the New Zealand and most of the Polynesian species. On the basis of sculpture it is closest to M. solomonensis, but in shape and general appearance it more nearly resembles some of the Polynesian species. The relatively wide umbilicus is diagnostic and at once separates M. euryomphala from the other species of Mocella. Lot ML 63 was collected alive under bark on a rotting log in deep shade. Fully satisfactory dissections of the animal were not obtained, but I extracted and studied the terminal portions of the male genitalia, jaw and radula. No other "Charopa" except the aberrant New Zealand species Egestula egesta (Gray) has been dis- sected (see Suter, 1913). It has an epiphallic flagellum, different insertion of the penial retractor, and less swollen penis. In Mocella euryomphala the foot is not divided longitudinally but has pedal grooves with a distinct pit where they meet over the tail. Iredale (1913, pp. 375-376) probably mistook the caudal pit for a true "mucous gland" in some Kermadec Islands endodontids. The jaw of M. euryomphala is membranous and solid, but without radial striae. There are slight indications of concentric growth rings and a low median projection. The radula (pi. 6, fig. 7) has the formula 12-6-1-6-12 in the single specimen examined. The central and lat- eral teeth are poorly differentiated, both with three cusps; the meso- cone, by far the largest, is as long as the basal plate. The marginals are subequally tricuspid, thus agreeing well with the teeth of Mocella corniculum (Reeve) (see Suter, 1913, p. 726). The male genitalia of M. euryomphala (pi. 6, fig. 8) are remark- able mainly for their swollen, sharply differentiated epiphallus. In the other species of endodontids from the Pacific (both in Egestula and Endodonta), the epiphallus is not sharply demarcated from the penis and the penial retractor inserts at the junction of the vas defer- ens and epiphallus rather than at the penial-epiphallic angle. Com- parison should be made with the anatomy of Endodonta (Cooke, 1928) and the North American Anguispira (Pilsbry, 1948, p. 568). SOLEM: MOLLUSCA OF THE NEW HEBRIDES 85 Details of the female and internal structure of the male genitalia were not worked out. RETICHAROPA, new genus Small, toothless, deeply umbilicated shells with slightly elevated spire. Whorls 3-4, sutures moderately impressed, last whorl descending slightly in front. Apical sculpture reticulate (pi. 31, figs. 6, 7), formed by radial and spiral ribs. Same sculp- ture continued on body whorl with decrease in size (or actual loss) of spiral ribs. Aperture ovate, with thin parietal callus, lip simple, not reflexed or thickened. Animal unknown. Type species. — Reticharopa latecosta, new species. Remarks. — Reticharopa is proposed for five species of New He- bridean "Charopa" that have reticulate apical sculpture (pi. 31, figs. 6, 7) . In most species, the spiral ribs of the apex are continued over the radial ribs on the spire and body whorl, although they are greatly reduced on the post-apical whorls of R. stenopleura. Reti- charopa is closely allied to an Indonesian-New Guinean genus {Parvi- charopa), which differs from Reticharopa in having the spiral and radial ribs fused (see Solem, 1958a), more loosely coiled whorls, and a less elevated spire. It is possible that collections from the Bismarcks, the Solomons, and New Guinea will contain species which are intermediate between the Indonesian and New Hebridean groups. Three of the five species of Reticharopa were recovered from a single drift sample (ML 95) ; the fourth species, R. geddiei, was found on Aneiteum in a specimen of Placostylus fuligineus; and the fifth species, R. helva (Cox), is also from Aneiteum. Reticharopa geddiei, new species. Plate 29, figs. 4-6. A species of Reticharopa characterized by its large size (3 mm.), widely spaced radial ribs (35-40 on body whorl), prominent spiral ribs, and angulated body whorl. Shell small, solid, deeply umbilicate, base convex. Whorls 4, rounded, increas- ing regularly in size; body whorl angulated, slightly descending in front. Sutures well marked, but only slightly impressed. Apical sculpture partially eroded. Spire and body whorl with widely spaced, retractive, radial ribs crossed by numerous (ca. 23 on third whorl) spiral ribs. Several riblets between the primary radial ribs. Aperture subquadrangulate, lip simple, not reflexed or thickened. Parietal callus thin. Umbilicus 0.75 mm. wide, open to the apex, contained about 4 times in the diameter. Diameter 3.1 mm., height 1.8 mm. Type. — Chicago Natural History Museum no. 72335. Collected on Aneiteum Island. 86 FIELDIANA: ZOOLOGY, VOLUME 43 Remarks. — The unique holotype was stuck in some dirt in the aperture of a specimen of Placostylus fuligineus. R. geddiei most closely resembles R. latecosta, but it is easily separable by its larger size and angulated body whorl. The other endodontid known from Aneiteum, Reticharopa helva (Cox) (see p. 87), has a rounded body whorl and much more crowded ribbing. R. geddiei is named after the pioneer missionary on Aneiteum, the Rev. John Geddie of Nova Scotia. Reticharopa latecosta, new species. Plate 29, figs. 1-3. A species of Reticharopa characterized by its widely spaced pri- mary ribs (35 on last whorl), prominent spiral ribs, rounded body whorl, and minute size (2 mm.). Shell small, thin, deeply but narrowly umbilicate, spire slightly elevated. Whorls 4 to 4J^ (holotype), gradually increasing in size, last whorl descending in front. Sutures deeply impressed. Apical whorls 13^, sculpture composed of equally prominent radial and spiral ribs (pi. 31, fig. 6). Primary ribs slightly retractive and more widely spaced on body whorl and spire than on the apex. Spiral ribs continued over primary radials. Numerous secondary riblets between primary radials. Aperture ovate, lip simple, not thickened or refiexed. Parietal callus very thin. Umbilicus 0.5 mm. wide, contained four times in the diameter. Diameter of holotype 2.0 mm., height 1.1 mm. Type. — University of Michigan Museum of Zoology no. 186040, Collected from stream drift in the Sarakata River Valley, Espiritu Santo, New Hebrides (ML 95), by Robert E. Kuntz, May- June, 1944. Para«2/pes.— Paratopotypes are UMMZ 186041, BPBM 212376, and MCZ 186829. Remarks. — R. latecosta has the same sculpture as R. geddiei, but the latter is easily recognized by its larger size, only slightly im- pressed sutures, and angulated body whorl. Reticharopa stenopleura, new species. Plate 30, figs. 1-3. A species of Reticharopa in which the spiral ribs have been re- duced on the body whorl and spire and the primary radial ribs are very close-set (73 on the body whorl) . Shell minute, solid, closely coiled, deeply and narrowly umbilicated, spire slightly elevated. Whorls 4 to 43^, sharply rounded, body whorl descending in front. Sutures prominent, but not deeply channeled. Apical whorls 1 }4, sculp- ture reticulated, spiral ribs less prominent than the radial (pi. 31, fig. 7). Body whorl and spire without prominent spiral ribs. Radial ribs prominent, close-set, slightly retractive. Secondary riblets present on and between the primaries. Aperture ovate, lip simple. Parietal callus thin, translucent. Umbilicus 0.4 mm. SOLEM: MOLLUSCA OF THE NEW HEBRIDES 87 wide, contained five times in the diameter. Diameter of holotype 2 mm., height 1.3 mm. Type. — University of Michigan Museum of Zoology no. 186038. Collected from stream drift in the Sarakata River Valley, Espiritu Santo, New Hebrides (ML 95), by Robert E. Kuntz, May- June, 1944. Paratyves.—\]MUZ 186039 and BPBM 212377. Remarks. — The reduced spiral ribbing of the post-apical whorls separates R. stenopleura from the other Reticharopa. Reticharopa sp. Plate 30, figs. 4-6. Range. — Espiritu Santo. Material.— ML 95. Remarks. — A single specimen with depressed spire, broken apical whorls, and reticulated body sculpture probably represents another species of Reticharopa. The shell is flammulated with reddish-brown and white and is similar to the New Zealand species grouped in Cavellia (sens, lat.), Dell (1952a) showed that Cavellia, characterized by a depressed spire, is polyphyletic. Similarly, Iredale (1937a) placed the Austro-Tasmanian species with depressed spires in genera having the same apical sculpture rather than in a separate genus. Although the apical whorls of the New Hebridean specimen are miss- ing, the spiral ribbing suggests that it is a Reticharopa. Reticharopa helva (Cox). Figure 3, a-c. Helix helva Cox, 1870, Proc. Zool. Soc. London, 1870: 82 — Aneiteum. Range. — Aneiteum. Material—At 200 feet elevation, Aneiteum (DMNZ, W. H. Daw- bin!). Photographs of holotype, courtesy of Australian Museum, Sydney). Remarks. — The original description of R. helva failed to mention many of the systematically important characters, and even exami- nation of the figures of the holotype furnished by Dr. Donald F. McMichael (fig. 3) did not enable me to classify this species satis- factorily. Fortunately, two partially broken specimens lent by the Dominion Museum identified Helix helva as a Reticharopa. The holo- type (Australian Museum no. C. 62202) has the apical sculpture worn off and there is no spiral sculpture apparent on the post-apical whorls. It is 2.5 mm. in diameter and 1.3 mm. high, and has 3% whorls. Both of the Aneiteum specimens from the Dominion Museum had worn apical whorls, but in one the umbilicus was free of obstruc- 88 FIELDIANA: ZOOLOGY, VOLUME 43 I 1 Fig. 3. a-c, Reticharopa helva (Cox) (photographs of holotype) ; d-f, Helix ardua Cox {=Liardetia samoensis (Mousson); photographs of holotype). Photographs courtesy of D. F. McMichael, Australian Museum, Sydney. Scale line=2 mm. tions and examination of the under side of the embryonic whorls clearly revealed the reticulated pattern of Reticharopa. The one adult specimen has 4^ whorls and is 2.46 mm. in diameter and 1.36 mm. high. It has 106 major ribs on the last whorl. One New Hebridean "Charopa" could not be identified. It has never been figured. The original description is reprinted below: *'Charopa" perryi Smith Charopa perryi Smith, 1897, Ann. Mag. Nat. Hist., (6), 20: 520 — Mota Island, New Hebrides, and Rotuma Island off the Fijis. "Testa orbicularis, latissime umbilicata, alba, epidermide olivacea induta, lineis incrementi elevatis curvatis tenuibus instructa; spira depressa, supra anfr. ultimum paulo elata; anfractus quatuor, convexi, primus laevis, pellucidus, ulti- mus supra leviter declivis, antice subdescendens; apertura obliqua, irregulariter rotundata, intus alba; peristoma tenue, marginibus conniventibus, callo tenui albo junctis, columellari ad insertionem vix expanso. Diam. maj. 3J^ millim., min. 3; alt. 2." Charopa perryi was named after Mr. W. W. Perry, who collected it on Mota Island. Mota is a small "hat" island off Valua and Vanua Lava in the Banks Group. It is possible that perryi is not an endo- SOLEM: MOLLUSCA OF THE NEW HEBRIDES 89 dontid. It may be related to Ouagapia radicalis (Mousson) from Fiji. No material of perryi was available. UMMZ 136770 (ex Walker, Ponsonby, Garrett) contains a single specimen that had been labeled C. perryi. This specimen is 6.3 mm. in diameter and has five whorls with a distinct shoulder on the body whorl. It is obviously misidentified and probably represents an undescribed species. Unfortunately, no locality data were given. Superfamily LIMACACEA Marginal teeth with narrow, lengthened basal-plates, either unicuspid and thorn-like, or bicuspid by elevation of outer on middle cusp (Pilsbry, 1896, p. 110). The three families, the Limacidae, Zonitidae, and Helicarionidae, are very closely related. The limacid slugs are easily characterized, but the differences between the Zonitidae and Helicarionidae are almost entirely anatomical. Shells of the two families cannot be separated unless the anatomy of the soft parts is known. Occasion- ally the zonitid subfamily, Vitrininae, is considered to be a fourth family intermediate between the Limacidae and Zonitidae (see Pils- bry, 1946b, p. 233). I have followed H. B. Baker (1941, p. 205) in recognizing the Zonitidae and Helicarionidae as distinct families, rather than Pilsbry, who included at least one helicarionid subfam- ily, the Euconulinae, in the Zonitidae, or Iredale (1939, et seq.), who divided the zonitoid snails into over a dozen families. The Limacidae are Palearctic slugs. One genus, Deroceras, has spread into northern Asia and North America, and several others have been widely disseminated by commerce. It is extremely difficult to determine the distribution of the other two families, since the anatomy of the Eurasian, African, and Neo- tropical "zonitoid" genera is imperfectly known. The Pacific Island species were very ably monographed by H. B. Baker (1938b, 1940, 1941). Melanesia was not included in the area covered by his study, but nearly all of the New Hebridean species belong to taxa repre- sented in Polynesia. Because the characteristics and probable rela- tionships of the supraspecific taxa have been summarized by Baker (loc. cit.), it was considered unnecessary to repeat them here. For convenience, however, diagnoses of the families Helicarionidae and Zonitidae have been included. I saw fourteen species and subspecies of New Hebridean limacoid mollusks. One additional species, "Helix" vannae-lavae Cox (1870), 90 FIELDIANA: ZOOLOGY, VOLUME 43 has never been figured and no material of it could be located. Speci- mens of two more species probably represent mislabeled lots. Of the fourteen species, eleven are endemic, two are widely distributed in the Pacific, and one is of uncertain status. I have described four species and one subspecies as new and have provided a single key for both families, since there are no obvious conchological characters separating them. All the New Hebridean limacoids belong to the most primitive subfamilies. The zonitid subfamily Trochomorphinae has "the most endodontid shell in the entire Limacea" (Baker, 1941, pp. 269-270), and it shows affinities to the most primitive helicarionid subfamily, the Euconulinae, which, in turn, is closely related to the Micro- cystinae (Baker, 1938b, p. 11). The other helicarionid subfamily, the Sesarinae, is also considered primitive (Baker, 1941, pp. 238, 352). None of the advanced taxa of "zonitid" mollusks is represented in the New Hebrides. In the genera for which anatomical material was available, Den- drotrochus and Trochomorpha, the New Hebridean species were found to belong to new sections or subgenera. Study of the animals of the endemic Lamprocystis, Diastole, and Orpiella will probably produce similar results. Without dissection of the soft parts, however, the infrageneric relationships of the endemic species cannot be deter- mined with accuracy. Pending such studies, I have tentatively assigned the species to previously named subgeneric taxa. Unlike most families of land Mollusca, the "zonitid" snails do not grow to a certain size and then cease addition of shell material. Growth is more or less continuous throughout the life of the individ- ual and it is therefore impossible to assign an "adult" size to any species. For this reason, measurements comparing species are only valid when shells with the same numbers of whorls are used. Com- parative measurements for the minute species are given in Table III. The measurements are not as detailed as those given by H. B. Baker (1938b, 1940, 1941), since most of the New Hebridean species are characterized by positive sculptural features or are representatives of widely distributed taxa. Both Dendrotrochus and Trochomorpha depart from the pattern shown by the minute species, since there is a partial thickening of the lip and columella and cessation of shell increment at about the same number of whorls. This may be a sea- sonal phenomenon. The morphometry of the New Hebridean Den- drotrochus and Trochomorpha is summarized in Tables IV- VI. SOLEM: MOLLUSCA OF THE NEW HEBRIDES 91 Key to New Hebridean Limacacea 1. Shell large, more than 10 mm. in diameter 2 Shell small, less than 7 mm. in diameter 7 2. Shell imperforate, or umbilicus contained more than eight times in the diameter 3 Umbilicus open, contained about five times in the diameter 6 3. Umbilicus completely closed; lip thickened, slightly reflexed; shell without a supraperipheral ridge 4 Umbilicus narrowly open; lip thin, not reflexed; shell with a prominent supra- peripheral ridge Inozonites bicarinata (Semper) 4. Shell with less than six whorls; less than 15 mm. in height 5 Shell with 6H or more whorls; more than 15 mm. in height. Dendrotrochus layardi (Hartman) 5. Shell solid; keeled (not carinate); 5}4 whorls; Vate Island. Dendrotrochiis eva eva (Pfeiflfer) Shell thin; sharply carinate; 5^ whorls; Banks, Torres or Santa Cruz Islands. Dendrotrochus eva stramineus Sykes 6. Shell solid; color band, when present, not bordering the suture. Trochomorpha bakeri, new sp. Shell thin; color band, when present, bordering the suture. Trochomorpha rubens Hartman 7. Shell perforate or openly umbilicate 9 Shell with umbilicus completely closed 8 8. Diameter at 4% whorls 5.7 mm.; H/D ratio 0.54-0.61; Espiritu Santo. Lamprocystis mendanae, new sp. Diameter at 4^ whorls 4.3 mm.; H/D ratio 0.66-0.75; Vate and Aneiteum. Lamprocystis guttula (Pfeiffer) 9. Shell with spiral rib(s) on apex and/or spire 10 Shell smooth or with close-set radial ribs on spire 12 10. Several spiral ribs on apex and spire 11 One spiral rib on spire, another on periphery of body whorl. Coneuplecta (Conibycus) bicarinata, new sp. 11. Shell minute (2.5 mm.); ribs continued on body whorl; base flat. Coneuplecta (Sitalina) microconus (Mousson) Shell large (4.8 mm.); ribs fading out on body whorl; base convex. Diastole subcarinata, new sp. 12. Shell smooth or with microscopic spiral or radial lines 13 Shell with fine radial ribs on spire, spiral lines on base. Liardetia samoensis (Mousson) 13. Shell large (4.5-6.0 mm.); no spiral sculpture 14 Shell minute (2 mm.); microscopic spiral lines Wilhelminaia sp. 14. Shell relatively depressed; Espiritu Santo. Orpiella {IHalozonites) retardata depressa, new subsp. Shell more elevated; Vate, Aneiteum . . .Orpiella (?Halozonites) retardata (Cox) Family HELICARIONIDAE Primitively with tripartite sole, well-developed caudal foss overhung by "horn" and multicuspid radular marginals, all of which are usually accentuated or reverted to when shell is reduced; advanced groups with well-developed amatorial organs or dart-apparatus on female side of genitalia; shell rarely umbilicate and with per- foration or rimation often closed by internal callus (H. B. Baker, 1941, p. 205). 92 FIELDIANA: ZOOLOGY, VOLUME 43 The subfamily Microcystinae is endemic to the Pacific islands, but most of the other subfamilies are continental groups. The Euco- nulinae is the most primitive, but it is very closely related to the Microcystinae. The third subfamily found in the New Hebrides, the Sesarinae, has a discontinuous distribution probably resulting from replacement by more advanced helicarionids on continental areas. The characteristics allying and separating the various sub- families of the Helicarionidae are given by Baker (1941, pp. 208-209). Subfamily EUCONULINAE Of the three New Hebridean species of Euconulinae, one, Coneu- plecta (Conibycus) hicarinata, is endemic, another, Coneuplecta {Sita- lina) microconus (Mousson), is widely dispersed in the Pacific, and the third, Wilhelminaia sp., has uncertain status. Discussions of the systematic position and content of the Euconulinae are given by Baker (1941, pp. 212-213) and Pilsbry (1946b, pp. 233-234). Genus WILHELMINAIA Preston, 1913 Shell corneous, turbinate, perforate, sculptured throughout with very fine, spiral, punctate striae (Preston, 1913, p. 434). Type species. — Wilhelminaia mathildae Preston. Remarks. — The status of the species included in Wilhelminaia is very uncertain. Usually they are placed in "Microcystis" or "Micro- cystina," catch-all taxa for small to minute Indo-Pacific zonitids. H. B. Baker (1938b, p. 57) restricted Microcystis to some species from the Austral and Cook Islands. Microcystina, described from the Nicobar Islands and subsequently enlarged to include many In- donesian species (see B. Rensch, 1932, and van Benthem Jutting, 1950), has a much larger shell and different radular and genital structure than the only Wilhelminaia whose anatomy has been fig- ured (see H. B. Baker, 1941, p. 227, pi. 44, figs. 1, 2). W. minuscula Preston, a manuscript species of Preston's from Batjan Island, "Mi- crocystina" gratilla van Benthem Jutting (1950, pp. 448^50, figs. 68-69), the Caroline Island specimens reported by H. B. Baker (1941, pp. 226-227), the genotype, and the Melanesian specimens I have seen during this study I have temporarily grouped as a generic unit. Discoconulus sinapidium (Reinhardt) from Harima, Japan (UMMZ 147956, ex Walker, Ponsonby, Gude, Hirase), is similar but has weaker apical sculpture and differently shaped radular teeth. W. mathildae has never been figured, and I could not locate any specimens. Shells of W. minuscula Preston and the Batjan species SOLEM: MOLLUSCA OF THE NEW HEBRIDES 93 Table III. — Measurements of the New Hebridean Limacacea Species Wilhelminaia sp. Florida Island, Solomons (FLW 40) (FLW 38) New Hebrides (ML 39).. (ML 35) (ML 95) Batjan Island, Moluccas (UMMZ 148388) Liardetia samoensis (ML 95) Coneuplecta microconus . . . . (ML 95) C. bicarinata, type Lamprocystis guttula (Miller 598) UMMZ 147868 Aneiteum (AMNH) L. mendafiae, holotype paratypes CNHM 55197, paratype. Diastole subcarinata holotype Orpiella retardata 0. r. depressa, type Greater Lesser diameter diameter Height H/D Whorl 1.62 1.49 0.92 0.57 S'A 1.49 1.39 0.92 0.58 3^ 1.52 1.42 0.99 0.65 sy2 1.88 1.72 1.19 0.63 ^Vs 1.72 1.55 1.06 0.62 3H 1.91 1.73 1.19 0.62 ^H 1.65 1.52 0.99 0.60 SVs 1.88 1.75 1.19 0.63 4^ 1.65 1.52 1.06 0.64 SVs 1.72 1.62 1.12 0.65 3H 1.85 1.72 1.16 0.63 4 1.78 1.65 1.12 0.63 3% 1.58 1.48 0.99 0.63 3M 1.58 1.45 0.99 0.63 3H 3.14 2.98 2.38 0.76 4% 2.68 2.54 2.24 0.84 5H 2.61 2.44 2.48 0.95 5H 2.44 2.31 2.14 0.88 4K 4.4 4.1 3.3 0.75 5 4.3 4.0 2.9 0.68 4^ 3.4 3.1 2.2 0.65 3K 5.0 4.6 3.7 0.74 5 3.3 3.0 2.2 0.67 ^Vs 5.8 5.4 3.4 0.59 m 4.7 4.4 2.8 0.60 SVs 4.8 4.4 2.9 0.60 SVs 5.7 5.3 3.1 0.54 4K 4.8 4.4 3.2 0.67 4^ 4.8 4.4 3.2 0.67 4M 4.8 4.4 3.3 0.69 4Ji 4.5 4.0 2.5 0.56 43^ 5.0 4.6 3.2 0.64 ^H 5.1 4.5 2.7 0.53 4^ 5.9 4.6 3.3 0.54 5 (UMMZ) were very similar to the Melanesian shells. H. B. Baker (1941, pp. 226-228) referred some Caroline Island specimens to Discoconulns and compared them to sinapidium. He had not seen any specimens of Wilhelminaia. His description and figure of one specimen (op. cit., p. 227, pi. 60, fig. 14) match the specimens of Wilhelminaia better than those of Discoconulus sinapidium. Dr. Y. Kondo compared the New Hebridean and Caroline Island shells without discovering any differences. 94 FIELDIANA : ZOOLOGY, VOLUME 43 Wilhelminaia sp. (?mathildae Preston). Plate 33, figs. 1-3. Range. — Espiritu Santo, Florida Island, Solomon Islands, Caro- line Islands, and Indonesia. Material— ML 35, ML 39, ML 44, ML 74, ML 95; New Heb- rides (UMMZ, ex Walker, Ponsonby); Florida Island, Solomon Is- lands (UMMZ, ex Kuntz!); Batjan Island, Moluccas (UMMZ 148386, UMMZ 148387, and UMMZ 148388, ex Walker, Preston). Remarks. — The Indonesian specimens cited above were under a manuscript name of Preston's and their identification with W. ma- thildae is uncertain. Certainly there were no important differences between the Indonesian and Melanesian specimens that I examined during this study. I suspect that only one wide-ranging species of Wilhelminaia will eventually be recognized. Genus CONEUPLECTA Moellendorff, 1893 Coneuplecta microconus (Mousson) and C. calculosa (Gould) have been widely distributed in the Pacific by human agency (H. B. Baker, 1941, p. 235). Coneuplecta is common in the Philippines and Indo- nesia, but only the two species mentioned above reach Polynesia. H. B. Baker (op. cit., pp. 232-233), with some hesitation, included a number of genera described by Thiele and Iredale in the synonymy of Coneuplecta. One New Hebridean species, C. bicarinata, new spe- cies, seems referable to Conibycus (Thiele, 1928a, p. 135). Pending study of the soft parts, I have retained Conibycus as a subgenus of Coneuplecta, although it may prove to be a separate genus. Coneu- plecta microconus is widely distributed in the New Hebrides, but C. calculosa has not yet been found there. Coneuplecta (Sitalina) microconus (Mousson). Plate 34, fig. 10. Nanina microconus Mousson, 1865, Jour, de Conch., 13: 192 — Lomma-Lomma, Fiji Islands. Helix sansitus Cox, 1870, Proc. Zool. Soc. London, 1870: 83 — Viti Levu, Fiji, and Vanua Lava, Banks Islands. Trochonanina microconus (Mousson), Garrett, 1887, op. cit., 1887: 172 — Tonga and Samoa. Sitala sansitus (Cox), Gude, 1913, Proc. Malac. Soc. London, 10: 328, pi. 14, fig. 7a-c — Suva Harbor, Fiji Islands. Coneuplecta (Sitalina) microconus (Mousson), Baker, 1941, Bull. B. P. Bishop Mus., 166: 236-237, pi. 55, fig. 6. Range. — Espiritu Santo, Aneiteum, Vate, and Vanua Lava. Fiji, Samoa, and Tonga. SOLEM: MOLLUSC A OF THE NEW HEBRIDES 95 Material— ML 26a, ML 31a, ML 50, ML 63, ML 66, ML 70, ML 76a, ML 95; New Hebrides (UMMZ 161077, ex Walker, Pon- sonby); Vila, Vate (Australian Museum, A, R. MeCulloch!); Anei- teum (DMNZ, W. H. Dawbin!); numerous lots from Polynesia. Remarks. — The New Hebridean specimens are not separable from the Fijian "sansitus" studied by Gude (loc. cit.) (UMMZ 148445 and CNHM 43411). I have therefore synonymized "Helix" sansitus despite the fact that I have not re-examined the types. Coneuplecta (Conibycus) bicarinata, new species. Plate 33, figs. 7-9. A species of Conibycus easily separated from C. dahli Thiele and C aruensis Thiele by having a supra-medial thread-like carina on the spire and a second carina on the periphery of the body whorl. Shell trochoidal, apex slightly depressed, whorls 4>2 to 4^, regularly and gradually increasing in size. Sutures impressed, whorls rounded. Aperture obliquely ovate. Umbilicus barely perforate, basal lip reflected over and par- tially covering perforation. Sculpture of many fine spiral threads, crossing low, irregular growth wrinkles. One thread, slightly above the middle of the whorl, greatly enlarged, forming a distinct carina on the spire. A second carina on periphery of body whorl. Diameter of holotype 2.44 mm., height 2.14 mm., with A}/2 whorls. Type. — University of Michigan Museum of Zoology no. 186108. Collected from stream drift in the Sarakata River Valley, Espiritu Santo (ML 95), by Robert E. Kuntz, May-June, 1944. Paratype. — A single paratopotype is BPBM 212375. Remarks. — Only two specimens, one broken, of C. bicarinata were available. The basic sculpture is typical of Coneuplecta, but the thread-like carinae at once separate bicarinata from any Pacific spe- cies. An extremely similar shell, "Sitala" subbilirata Godwin-Austen, is found in the Andaman Islands. Specimens of subbilirata (UMMZ 148446, ex Walker, Godwin-Austen) have the identical sculpture of threads and carinae found in bicarinata. C. bicarinata differs from subbilirata, however, in having a much narrower umbilicus and a more elevated spire. The relationship of the Indian and Indonesian Sitala to Coneuplecta remains to be determined. Without study of the soft parts it will be impossible to decide whether bicarinata and subbilirata represent convergent species in unrelated genera or paral- lel developments from the same ancestral stock {?Conibycus from Papua) . On the basis of other taxa, I consider that the latter pos- sibility is probably correct. 96 FIELDIANA: ZOOLOGY, VOLUME 43 The nearest relatives of C. bicarinata in the Pacific seem to be C. dahli from New Britain and C. aruensis from the Aru Islands (Thiele, 1928a, p. 137, pi. 5, figs. 26-27). Both species have the apex of the spire slightly, though distinctly depressed. Typical Coneu- plecta has the apical whorls at the same angle as the rest of the spire. Subfamily MICROCYSTINAE If broad interpretations of subfamilial categories are utilized, the Microcystinae can be merged with the Euconulinae into one sub- family. Following H. B. Baker (1941, pp. 208-209), I have adopted the narrower subfamily units. Three of the four microcystine spe- cies found in the New Hebrides, Lamprocystis guttula (Pfeiffer), L. mendanae, new sp., and Diastole subcarinata, new sp., are endemic; the fourth, Liardetia samoensis (Mousson) {=striolata Pease), is widely distributed in the Pacific. Genus LIARDETIA Gude, 1913 The only endemic Polynesian species of Liardetia are found in the Society Islands. Two species, L. discordiae (Garrett) and L. samoensis (Mousson), appear to have been widely disseminated. It seems probable that the center of distribution lies in Indonesia, since by far the greatest number of species are found there (van Benthem Jutting, 1950, pp. 393-412). The New Hebridean species, L. samo- ensis, has the widest range of any species. Liardetia (Liardetia) samoensis (Mousson). Figure 3, d-f; Plate 33, figs. 4-6. Nanina samoensis Mousson, 1865, Jour, de Conch., 13: 165 — Upolu, Samoa. Helix ardua Cox, 1870, Proc. Zool. Soc. London, 1870: 82 — Erromanga, New Hebrides (Brazier!). Helix antelata Cox, 1870, op. cit., pp. 83-84 — Aneiteum, New Hebrides (Brazier!). Guppya papuana Thiele, 1928, Zool. Jahr., Syst., 55: 113, pi. 5, fig. 20— Ralum, New Britain. Liardetia (Liardetia) striolata (Pease), Baker, 1938, Bull. B. P. Bishop Mus., 158: 22-24, pi. 9, figs. 5, 6; pi. 14, fig. 12. Liardetia (Liardetia) samoensis (Mousson), Baker, 1940, op. cit., 165: 190 — correction of name, since striolata is preoccupied. Range. — Espiritu Santo, Aneiteum, and Erromanga. New Cale- donia, Solomons, Bismarcks, Marshalls, Ellices, Fiji, Samoa, Cook, Society, and Marquesas. SOLEM: MOLLUSCA OF THE NEW HEBRIDES 97 Material— ML 50, ML 63, ML 69, ML 95; Aneiteum (UMMZ, ex Australian Museum, paratypes of Helix antelata Cox) ; Erromanga (UMMZ, ex Australian Museum) ; New Hebrides (UMMZ 162016, ex Walker, Ponsonby); Bourail, New Caledonia (UMMZ); many Polynesian specimens. Remarks. — Ldardetia samoensis varies greatly in size and shape. Juvenile shells have a low spire and correspond to the form called antelata Cox; gerontic individuals have a very high spire and repre- sent the form ardua Cox (fig. 3) . Specimens matching both extremes, as well as numerous intermediate individuals, were found on Espiritu Santo. Franc (1957, p. 147) reports this species from New Caledonia as Kaliella subfulva (Gassies). Genus DIASTOLE Gude, 1913 The nomenclatural confusion necessitating the replacement of Trochonanina of authors by Diastole Gude was summarized by H. B. Baker (1938b, pp. 45^6). Diastole is primarily a Polynesian genus, with species found in the New Hebrides and Fijis. Study of the Papuan and Indonesian "zonitids" may result in a great westward extension of its range. The New Hebridean Diastole subcarinata, new sp., is somewhat intermediate between Diastole (sens, str.) and Trochonanita Baker, 1941. It has the scarcely beaded spiral ridges and no columellar fold of the former, but the number of major spiral ridges of the latter. D. subcarinata is characterized by its small size, weak carina, and lack of a columellar fold. Diastole (Diastole) subcarinata, new species. Plate 32, figs. 4-6. A species of Diastole most closely related to the Society Island D. necrodes Baker but easily separated by virtue of its smaller size, less acute carina, and lack of a columellar lamella. Shell depressed trochoidal, body whorl angulated with a thin epidermal carina, whorls slowly and gradually increasing in size. Base of shell slightly flattened laterally, sutures impressed, whorls of spire gently and evenly rounded. Holo- type very light horn-colored, paratypes bleached. Whorls 4^, embryonic ones (pi. 32, fig. 6) with eleven spiral ridges, faintly beaded by radial lines. Ridges becoming less prominent on spire, almost completely lacking on body whorl. Growth wrinkles becoming more prominent on later whorls. Base of shell with only faint traces of sculpture. Aperture compressed, basal margin paralleling con- tour of penultimate whorl, distinctly angulate at periphery. Columella slightly thickened, but without trace of a lamella. Diameter of holotype 4.8 mm., height 3.3 mm. Type. — University of Michigan Museum of Zoology no. 186103. Collected from stream drift in the Sarakata River Valley, Espiritu 98 FIELDIANA: ZOOLOGY, VOLUME 43 Santo, New Hebrides (ML 95), by Robert E. Kuntz, May- June, 1944. Paratypes.—Faratopotypes are UMMZ 186102, BPBM 212372. Remarks. — With the exception of D. lamellaxis Baker, D. sub- carinata is the smallest member of the genus known. D. subcarinata differs from D. lamellaxis in being much less sharply carinate and more depressed. Only three specimens of D. subcarinata were avail- able. Genus LAMPROCYSTIS Pfeiffer, 1883 H. B. Baker (1938b, pp. 68-92) monographed the Polynesian and Micronesian species of Lamprocystis. Some Indonesian species have been referred to the genus (see van Benthem Jutting, 1950), but the western limits of distribution are unknown. No preserved specimens of the New Hebridean species were available. On the basis of con- chological characters, both Lamprocystis guttula (Pfeiffer) and L. mendanae, new sp., belong to Lamprocystis (sens. str.). They differ from the Fijian and Samoan species in having the columellar lamel- lae reduced and more elevated; from the Mariana L. denticulata (Quadras and Moellendorff) they are distinguished by having the columellar lamellae higher and less prominent. Lamprocystis (Lamprocystis) guttula (Pfeiffer). Plate 34, figs. 11, 12. Helix guttula Pfeiffer, 1853 (March), Zeits. f. Malak., 10, (4), p. 53— New Zealand (error); Reeve, 1853 (May), Conch. Icon., Helix, pi. 158, fig. 1040 — New Zealand (error); Pfeiffer, 1854, Proc. Zool. Soc. London, 1853: 58 — New Zealand (error); Pfeiffer, 1859, Monog. HeUc. viv., 4: 15. Hyalina (Conulu^) layardi Thomson, 1885, Proc. Zool. Soc. London, 1885: 26-27, figs.— Vate Island, New Hebrides. Microcystis guttula (Gould) (sic), Sykes, 1903, Proc. Malac. Soc. London, 5: 196— Port Vila, Vate Island. Lamprocystis layardi (Thomson), Ancey, 1905, Nautilus, 19, (4), p. 42 — Vate (Glisson!). Lamprocystis (Lamprocystis) excrescens (Mousson), Baker, 1938, Bull. B. P. Bishop Mus., 158: 85— (part). Range. — Vate and Aneiteum. Material— Yi\a, Vate (Miller 598); Vate (UMMZ 147868, ex Walker, Ponsonby, Layard, paratypes of layardi Thomson; CNHM 43041, ex Webb, Gude, paratype of layardi Thomson); Aneiteum (AMNH, Macmillan!, 1937); New Hebrides (UMMZ 147867, ex Walker, Ponsonby, Garrett). SOLEM: MOLLUSCA OF THE NEW HEBRIDES 99 Remarks. — The identity of Helix guttula Pfeiffer has long been uncertain. Sykes (1903) united layardi and guttula. Specimens from the British Museum (CM 62.29194), which are the excrescens of H. B. Baker (1938b, pi. 13, figs. 5, 6) are labeled guttula. The speci- men figured in the Conchologia Iconica clearly shows the high spire and single columellar "tooth" of the Vate shells, rather than the characters of the true excrescens, as emphasized by Ancey (1905, p. 42). Evidence that Helix guttula is a mislabeled New Hebridean shell is provided by Delos (Hebridelos) rapida (Pfeiffer), which was described from "New Zealand" at the same time and is now known to be New Hebridean. L. guttula differs from L. excrescens in lacking basal teeth in young shells, in having a shorter columellar lamella, and in being much more trochoidal. L. mendanae has the same columellar structure as L. guttula, but is more depressed and slightly larger. Lamprocystis (Lamprocystis) mendanae, new species. Plate 34, figs. 7-9. Nanina (Microcystis) excrescens Mabille (not Mousson), 1895, Bull. Soc. d'hist. nat. d'Autun, 8: 109 — New Hebrides (Frangois!). A species of Lamprocystis characterized by its large size (5.8 mm. at 4% whorls), low spire, angulated body whorl, and small columel- lar lamella. Shell broadly conoid, depressed, imperforate; body whorl slightly angulated on adults (33^ whorls), more sharply angulated on juveniles. Surface smooth, shining, with only slight suggestion of spiral striae. Suture broadly overriding, only slightly impressed. Aperture with rounded periphery, small, not flaring. Lip thin, slightly thickened and reflexed basally. Columella with a small, rounded spiral lamella, more prominent and sharply delimited in young shells. Color very light horn. Diameter of holotype 5.8 mm., height 3.4 mm., with 4^ whorls. Type. — University of Michigan Museum of Zoology no. 181741. Collected from river drift in the Sarakata River Valley, Espiritu Santo, New Hebrides (ML 95), by Robert E. Kuntz, May-June, 1944. Paratypes. — On Espiritu Santo, specimens were collected at ML 69 and ML 95 (type locality). Paratypes are UMMZ 181740, UMMZ 186100, CNHM 55197, BPBM 212373, ANSP, MCZ. Remarks. — In columellar structure L. mendanae is almost iden- tical with L. guttula, but the smaller size and relatively higher spire of the latter easily separate the two species. As in most zonitids, there is considerable difference in shape between old and young speci- mens. L. mendanae is named after the Spanish explorer, Alvaro de Mendafia. 100 FIELDIANA: ZOOLOGY, VOLUME 43 Subfamily SESARINAE The Sesarinae are rather primitive heUcarionids which have a discontinuous distribution. Although diverse in shell structure, ana- tomical similarities indicate that the Sesarinae are a natural assem- blage. Two genera, Orpiella and Dendrotrochus, yielded interesting information on the affinities of the New Hebridean species. Genus ORPIELLA Gray, 1855 Only the Fijian species of Orpiella have been adequately studied for anatomical characters and conchological variation. Numerous "species" are known from the Solomons, Bismarcks, and New Guinea, but their relationship to the Fijian taxa cannot be determined with- out more study (H. B. Baker, 1941, p. 352). The two New Hebri- dean morphs appear to be most similar to the Fijian group that H. B. Baker (1941, pp. 240-243) placed in Halozonites Pilsbry and Cooke, 1941. Pending study of the animals, I have tentatively referred 0. retardata and 0. retardata depressa to Halozonites. The only dif- ference between retardata and depressa is that the latter is less ele- vated. I consider them to be only subspecifically distinct. Orpiella (PHalozonites) retardata retardata (Cox). Plate 34, figs. 1-3. Helix retardata Cox, 1870, Proc. Zool. Soc. London, 1870: 84 — Aneiteum (Brazier). Macrochlamys (?) annatonensis Ancey (not Pfeiffer), 1905, Nautilus, 19, (4), p. 42— Vate (Glisson). Range. — Vate, Aneiteum. Material.— Yi\a, Vate (Miller 598; USNM 598360 ex Miller; UMMZ 147918 ex Walker, Ponsonby, Ancey, Layard) ; Vate (UMMZ 162021 ex Walker, Ponsonby, Layard) ; Aneiteum (UMMZ ex Aus- tralian Museum, paratypes of Helix retardata Cox). Remarks. — The paratypes of Helix retardata are very young indi- viduals and quite different in appearance from typical adults. Kobelt placed retardata in the rhytidid genus Macrocy chides, but examina- tion of the paratypes does not support this. 0. retardata differs from Orpiella of similar size by its narrow umbilicus and angulated body whorl. Orpiella (PHalozonites) retardata depressa, new subspecies. Plate 34, figs. 4-6. A subspecies of Orpiella retardata which differs from the nominate form only by having a more depressed spire. SOLEM: MOLLUSCA OF THE NEW HEBRIDES 101 Sculpture, umbilicus, aperture, color, and whorl increase as in typical retardata. Spire more depressed and shells larger in diameter than specimens of the nominate form with the same number of whorls. Diameter of holotype 5.1 mm., height 3.0 mm., with 4^ whorls. Type. — University of Michigan Museum of Zoology no. 186112. Collected in the jungle between the Renee and Sarakata Rivers, Espiritu Santo, New Hebrides (ML 26b), by Robert E. Kuntz on January 20, 1944. Paratypes. — On Espiritu Santo, Kuntz collected this subspecies from ML 26b (type locality), ML 33, ML 39, ML 74, ML 95. Para- types are UMMZ 186109, UMMZ 186110, UMMZ 186113, BPBM 212371, CNHM, MCZ, ANSP, and USNM. Remarks. — Comparisons of the 57 specimens of 0. r. depressa with the 19 specimens of 0. r. retardata failed to reveal any charac- ters, other than height and diameter, by which the two can be sepa- rated. Most individuals could be easily placed in one subspecies or the other, but a few intergrading shells were seen. Despite the geo- graphic isolation, retardata and depressa are probably only subspe- cifically separable. Genus DENDROTROCHUS Pilsbry, 1894 (=Trochonanina Rensch, 1930 et seq., not Mousson, 1869) For many years both the systematic and the nomenclatural status of Dendrotrochus were uncertain. On the basis of conchological simi- larity, Pilsbry established Dendrotrochus as a section of the camaenid genus Papuina. Hedley (1895a) dissected the animals and placed Dendrotrochus near Trochomorpha. Convergence in shell structure led L & B. Rensch to combine Dendrotrochus with the microcystine genus Diastole (= Trochonanina Garrett, Rensch, etc., not Mousson, 1869). H. B. Baker (1941, p. 256) placed Dendrotrochus in the Sesari- nae after dissecting several species. His classification is followed here. The exact range of Dendrotrochus is uncertain. It is definitely known from the New Hebrides, Santa Cruz, Solomons, Bismarcks, Admiralty and Caroline Islands (fig. 19), but anatomical investiga- tions may prove that some of the New Guinean and Australian "Papuina" are congeneric. I. Rensch (1934b, 1937) reviewed the species of Dendrotrochus, but the taxonomic status of the numerous morphs remains uncertain. The shells vary greatly in size, color, elevation and degree of carina- tion. Apparently there are terrestrial and arboreal species, since both types of radulae are found (I. Rensch, 1934b, pp. 3, 23). Par- 102 FIELDIANA: ZOOLOGY, VOLUME 43 ticularly in respect to the taxa of northern Melanesia, no satisfactory conclusions can be drawn as to the type and extent of speciation. Only a few species have been dissected. They agree in the gen- eral facies of the genitalia, jaw, mantle lappets, and caudal areas, but show a remarkable series of divergences in the internal structure of the penis. The Caroline Island species was placed in a new sub- genus, Ponapea, by Baker (1941, p. 256), because it lacked a utricle surrounding the verge. Typical Dendrotrochus is found in the Solo- mons, and I dissected one species, D. cleryi cleryi (Recluz), for com- parison with the New Hebridean D. layardi (Hartman) . The external features of the genitalia are quite similar, but the internal structures of the penis (pi. 23, figs. 1-4) are quite different. The New Hebri- dean species has two large pilasters flanking the verge and a large ovoid stimulator in a sac to one side of the atrium, but all of these are lacking in D. cleryi. On the basis of the penial structure, D. lay- ardi is herein made the type of a new subgenus, Santotrochus. Speci- mens of D. eva from Vate (BM loan in 1958) have the same structures as D. layardi. SANTOTROCHUS, new subgenus Male genitalia with a verge flanked by two pilasters and a large, flat stimulator at one side of the atrium. Female genitalia and shell as in the typical subgenus. Type species. — Oxychona layardi Hartman. Remarks. — Two species of Dendrotrochus are known from the New Hebrides. D. layardi is restricted to Espiritu Santo and D. eva (Pfeiffer) apparently is found on Vate, Epi, the Banks, Torres, and Santa Cruz Islands. D. eva eva lives on Vate and Epi; the northern islands have a weakly differentiated race, stramineus Sykes. A third species, D. cyrene (Crosse) (see Jour, de Conch., 17: 183-184; 18: 102-103, pi. 2, fig. 2), from an unknown locality, may be from the New Hebrides. It is equally possible, however, that it may have come from the Solomons. The unique holotype of D. cyrene is in the collection of the Journal de Conchyliologique in Paris. "Helix" jenynsi Pfeiffer, described without locality, was reported from Tanna by Reeve (Conch. Icon., Helix, pi. 150, fig. 979). The figured shell may be a young Dendrotrochus, but the correctness of the locality is questionable. Dendrotrochus (Santotrochus) eva eva (Pfeiffer). Plate 10, fig. 13; plate 25, figs. 1,2. Helix eva Pfeiffer in Reeve, 1853, Conch. Icon., Helix, pi. 150, fig. 977— Solo- mon Islands (error); Pfeiffer, 1854, Proc. Zool. Soc. London, 1852: 84— SOLEM: MOLLUSCA OF THE NEW HEBRIDES 103 New Hebrides; Pfeiffer, 1854, Conch. Cab., I, 12, (3), p. 490, pi. 160, figs. 22-23— New Hebrides; Brazier, 1880, Jour, de Conchy., 28: 313- 314— Vate. Helix (Papuina) eva Pfeiffer, Pilsbry, 1891, Man. Conch., (2), 7: 78-79, pi. 15, figs. 76-78, 84-85, pi. 16, figs. 1-4— Vate. Dendrotrochus eva (Pfeiffer), Pilsbry, 1894, Man. Conch., (2), 9: 143; Sykes, 1903, Proc. Malac. Soc. London, 5: 196— Port Vila, Vate (J. J. Walker); Ancey, 1905, Nautilus, 19, (4), p. 43— Vate. Range. — Vate Island and Epi Island. Material.— Vila, Vate (USNM 598363 ex Miller; Miller 538; ANSP 133298 ex Froggatt); Vate (UMMZ 76488 ex Walker, Ponsonby, Willey; UMMZ 76485 ex Walker, Ponsonby; ANSP 48922 ex Depuy; MCZ 8656 ex Winkley; MCZ ex Beddome; CM 62.4491 ex Hartman; CM 62.13644 ex Clapp, Hartman; CNHM ex Webb, Gude, Layard); Nivenue, Epi, at 900 feet elevation (DMNZ, A. G. Horwell!); New Hebrides (ANSP 1912 ex Brown, Cuming, Cox; UMMZ 11931 ex Stearns; UMMZ 76484 ex Walker, Ponsonby, Wetherby, Garrett; AMNH 61730; AMNH 38397 ex Crooke; MCZ ex Grand Rapids Museum; MCZ ex Pease; MCZ 133876 ex Putzeys; MCZ ex Cox; MCZ 102842 ex Bequaert). Remarks. — At least two recognizably different races of D. eva are found on Vate, only one of which can be accurately localized. Shells from near Vila correspond perfectly to the holotype (pi. 10, fig. 13). The type locality is here restricted to the vicinity of Port Vila. The Port Vila shells are keeled, not carinate, appreciably higher, and with more whorls than are found in the northern race, stramineus (see Table IV). Vate Island specimens distributed by Layard, however, are intermediate between eva and stramineus in respect to whorl count (5% to 53^ whorls), coloration, and degree of carination. It is probable that the shells were collected at Seaview and Rathmor plantations, the locations of which I was unable to discover. Pos- sibly other morphological forms will be found when Vate has been more thoroughly explored. The Epi Island specimens show a range of variation encompassing both the Vate Island morphotypes. Dendrotrochus (Santotrochus) eva stramineus Sykes. Plate 10, fig. 14; plate 25, figs. 3-5. Dendrotrochus stramineus Sykes, 1903, Proc. Malac. Soc. London, 5: 196-197, fig. 1 — Lo, Torres Islands and Vanua Lava, Banks Islands (type locality) (J. J. Walker). Range. — Torres, Banks, and Santa Cruz Islands. 104 FIELDIANA: ZOOLOGY, VOLUME 43 Table IV, — Size Variation in the New Hebridean Dendrotrochtis Diameter Height H/D Whorls D.evaeva Mean 13.8 10.4 0.74 5^ Vila Vate (17 specimens) Range 13.1-15.0 9.3-11.2 0.63-0.82 SJ^-B S.D. 0.66 0.53 0.048 D.evaeva Mean 14.0 10.0 0.72 5% Nivenue, Epi (7 specimens) Range 13.1-14.8 9.1-10.9 0.64-0.79 S^-SJ^ S.D. 0.85 0.65 0.057 D. e. stramineus Mean 13.8 9.2 0.67 5K Santa Cruz Island (20 specimens) Range 12.6-15.0 8.5-10.1 0.61-0.71 5-5^ S.D. 0.73 0.48 0.037 D.layardi Mean 16.6 14.4 0.87 6^ Espiritu Santo (37 specimens) Range 15.3-18.0 12.7-16.6 0.76-0.98 eVg-lH S.D. 0.76 0.61 0.062 Material. — Graciosa Bay, Santa Cruz Island (MCZ 192354 ex A. G. Stevenson, C. E. Fox); Santa Cruz Island (AMNH 73353, W. M. Mann; UMMZ 182858 ex MCZ, W. M. Mann; MCZ 32476, MCZ 32477, MCZ 32478, MCZ 32479, all W. M. Mann; UMMZ 76487 ex Walker, Ponsonby); Banks Islands (MCZ 108715 ex P. Dautzenberg) ; photograph of holotype (courtesy of BM). Remarks. — The Port Vila and Santa Cruz Island specimens are usually separable on the basis of color and degree of carination but not morphometrically, since the coefficient of difference is less than 1,28 for every measurable character. If the other Vate shells are considered, the differences between the Vate and Santa Cruz popu- lations practically disappear. Pending further study I have placed the Vate shells in one subspecies, D. eva eva, and the Torres, Banks, and Santa Cruz specimens in another, D. e. stramineus. Further study is needed on this complex before the classification is settled, however. Collection of adequate series from the Banks and Torres may necessitate the addition of a name for the Santa Cruz population. Only two adult shells were seen from the Banks (type locality of stramineus). Both lacked the color band and were slightly smaller than the Santa Cruz shells. If these differences are constant, the Santa Cruz population could possibly be subspecifically distinct. SOLEM: MOLLUSCA OF THE NEW HEBRIDES 105 Dendrotrochus (Santotrochus) layardi (Hartman). Plate 6, figs. 5, 6; plate 7, figs. 3, 4; plate 25, figs. 6-9. Oxychona layardi Hartman, 1889, Proc. Acad. Nat. Sci. Philadelphia, 1889: 91, pi. 5, fig. 2 — Aore Island (Delautour). Helix (Papuina) layardi (Hartman), Pilsbry, 1891, Man. Conch., (2), 7: 79- 80, pi. 15, figs. 94-97. Dendrotrochus layardi (Hartman), Pilsbry, 1894, Man. Conch., (2), 9: 144; Sykes, 1903, Proc. Malac. Soc. London, 5: 196 — Renee River, Espiritu Santo (J. J. Walker); Ancey, 1905, Nautilus, 19, (4), p. 43. Trochomorpha (Oxychona) layardi (Hartman), Mabille, 1895, Bull. Soc. d'hist. Nat. d'Autun, 8: 409 — Espiritu Santo (Frangois). Range. — Aore and Espiritu Santo. Material— ML 17, ML 31e, ML 32, ML 33, ML 36, ML 74, ML 76, ML 84, ML 95; Aore Island (ANSP 60063 ex Hartman, Layard, Delautour, holotype; ANSP 194310 ex Hartman, Layard, Delautour, paratypes; UMMZ 76501 ex Walker, Ponsonby, Fulton, Layard, type lot; MCZ 182264 ex Fiske- Warren; CM 62.9811 ex Clapp, Hartman, paratype; CM 62.2910 ex Hartman, Layard, para- types) ; Hog Harbour, Espiritu Santo (MCZ 147386 M. J. McMil- lan) ; Espiritu Santo (UMMZ 162118 ex Walker, Ponsonby) ; New Hebrides (ANSP 109278 ex Sowerby and Fulton; UMMZ 76502 ex Walker; MCZ 102077 ex Bequaert). Remarks. — The holotype (pi. 25, fig. 8) of D. layardi is quite slender, but numerous intergrades connect it with the more obese specimens. Three color variations are found: ashy white (pi. 25, fig. 6) ; with a dark brown line on the keel and suture (pi. 25, fig. 8) ; and with an additional brown band between the keel and sutures (pi. 25, fig. 7). All three variations are found in most lots and prob- ably are of no taxonomic significance. Fresh material of D. layardi often shows small translucent spots and fine black lines on white background. From the esthetic viewpoint, D. layardi is undoubtedly the most "beautiful" snail found in the New Hebrides. The genital anatomy of D. layardi (pi. 7, figs. 3, 4) is typical of Dendrotrochus, except for the remarkable internal structure of the penial complex. The "biscuit" at the side of the atrium (pi. 7, fig. 4) is probably stimulatory. It is, however, much too large to fit through the undistended gonopore and the mechanism by which it is everted is unknown. The material available was not suitable for histological study. It could not be ascertained, therefore, whether the "biscuit" contained spaces that, when filled with fluid, resulted in great turgor, or whether it is a solid organ evertable only because of great disten- tion of the gonopore. 106 FIELDIANA: ZOOLOGY, VOLUME 43 The brilliant coloration, trochoidal shell, and radula (pi. 6, fig. 5) suggest that D. layardi is an arboreal species. The ecological infor- mation with the Kuntz specimens confirms this, since most living specimens were collected on leaves or trunks of trees between 5 and 15 feet from the ground. The radula of D. layardi (pi. 6, fig. 5) has developed lateral teeth with all the characteristics of arboreal rad- ulae, that is, with broad, gouge-like cusps (Pilsbry, 1894, p. xiv), while the marginals have the slender, pointed teeth that are charac- teristic of terrestrial genera. Two radulae of D. layardi were mounted. The larger had 118 rows with the formula 151-13-1-13-151; the smaller had 100 rows with the formula 134-12-1-12-134. One dis- sected specimen had a spermatophore (pi. 6, fig. 6) in the sperma- theca. Apparently it is essentially the same as the spermatophores of D. ponapensis H. B. Baker (1941, pi. 46, fig. 6). Unhke many zoni- tids, for example, Ryssota (op. cit., pi. 46, fig. 12), the spermatophore of D. layardi is without any prominent external sculpture. Family ZONITIDAE Primitively with undivided sole, becoming tripartite when shell is reduced; caudal foss not overhung and often obsolescent or lacking (rarely accentuated when shell is reduced); principal radular marginals not more than bicuspid and more usually unicuspid; amatorial organs or dart-apparatus (when rarely devel- oped) penial; shell usually umbilicate or with open perforation or rimation (H. B. Baker, 1941, p. 205). The subfamilial relationships within the Zonitidae have been ably discussed by H. B. Baker (1941, p. 270) and Pilsbry (1946b, p. 233). Of the six "restricted" subfamilies recognized by Baker (loc. cit.) only the Trochomorphinae are widely distributed in the Pacific. The Hawaiian Islands have a few endemic species of Zonitinae and Gastrodontinae; otherwise, only introduced species are found on Pa- cific islands. The Trochomorphinae are an important constituent of the island fauna over most of the Pacific. They probably repre- sent the most advanced land snail taxon native to Polynesia, although they are the most primitive subfamily of the Zonitidae. Subfamily TROCHOMORPHINAE Trochomorpha-like shells have been reported from India, Anda- man and Nicobar Islands, southeast Asia, China, Japan, the Ryuk- yus, Formosa, the Philippines, Indonesia, Melanesia (except New Caledonia), Polynesia east to the Society Islands, and the Palau, Caroline, and Marshall islands of Micronesia (fig. 19). None are SOLEM: MOLLUSCA OF THE NEW HEBRIDES 107 known from the Hawaiian, Cook, Austral, Marquesan, or Tuamotu Islands, New Caledonia, Australia, or New Zealand. The Polynesian and Micronesian Trochomorphinae were mono- graphed by H. B. Baker (1941, pp. 270-321). He recognized five genera, Kondoa, Hogolua, Brazieria, Videna, and Trochomorpha. The first three are restricted to the Caroline Islands; Videna ranges from the Palau and Philippine Islands westward (?); and Trocho- morpha includes the species from Polynesia and Melanesia. It re- mains for future studies to append the Asian, Indonesian, and Pap- uan species to Baker's classification. Videna and Trochomorpha apparently will receive many of the species, but new generic units may be necessary for the Asian shells. Genera extralimital to Baker's study which he included in the Trochomorphinae are Bertia, Coxia, and Calostropha, all of Ancey, 1887. The name Rosselidena Iredale, 1941, is a nomenclatural unit only, whose status cannot be decided without anatomical investigation. Several species of Trochomorpha (sens, lat.) have been reported from the New Hebridean region. All belong to Trochomorpha, ex- cept pulcherrima Hartman, which seems to be a mislabeled specimen of the Philippine Island Inozonites bicarinata (Semper). Genus TROCHOMORPHA Albers, 1850 H. B. Baker (1941, pp. 285-286) divided the Polynesian and Micronesian species of Trochomorpha into two subgenera, Nigritella for those species without a high, ovoid, penial stimulator and Tro- chomorpha for those species with a stimulator. Nigritella is found in the Caroline Islands, Samoa, Fiji, and westward to the Moluccas. Trochomorpha (sens, str.) was previously known only from the Tonga and Society Islands. The only Melanesian species examined by Baker was Trochomorpha sanctae-annae Smith. On the basis of its unicuspid marginal teeth. Baker established Lentitrochus as a new subgenus of Trochomorpha. Transitional forms between the radulae of the subgenera Lentitrochus and Nigritella exist (Solem, in press-A). Pending detailed study of the Solomon Island species I have retained Lentitrochus as a subgenus, although it may prove to be a section of Nigritella. The only New Hebridean species of which anatomical material was available, T. rubens Hartman, has a well-developed stimulator and thus belongs in Trochomorpha (sens. str.). The other New He- bridean species, T. bakeri, new sp., is herein tentatively placed in the same section. 108 FIELDIANA: ZOOLOGY, VOLUME 43 Subgenus TROCHOMORPHA Albers, 1850 The shell (pi. 24, figs. 1-17) and genital anatomy (pi. 6, figs. 1, 2) of T. rubens show several differences from those of the previously described sections of the subgenus Trochomorpha. In T. rubens the penial stimulator has a short free tip, like the Tongan Cotitrochus, but the epiphallus is less sharply differentiated from the vas deferens and the spiral striae of the protoconch present in Cotitrochus are ab- sent in rubens. The section Trochomorpha (sens, str.) differs from T. rubens in having a longer free tip of the stimulator, different apical sculpture and more poorly differentiated vas deferens and epiphallus. Probably the penial stimulator in T. rubens is not strictly homolo- gous, since it is located near the middle of the penis (see pi. 6, fig. 1). In Trochomorpha (sens, str.) and Cotitrochus its basal tip lies at or near the junction of the penis and atrium. Because of these differences, a new section is herein established for the New Hebridean species. HARTMANITROCHUS, new section Penis with several well-developed pilasters and a medial, fusiform stimulator. Vas deferens and epiphallus with longitudinal pilasters; epiphallus entering penis apically but not sharply demarcated from it. No penial sphincter. Shell lenticu- lar to trochiform, apical whorls with weak growth striae and very faint wrinkles similar to those found in Trochomorpha (sens. str.). Type species. — Trochomorpha rubens Hartman. Remarks. — Retention of Hartmanitrochus in the subgenus Tro- chomorpha is for reasons of practicality rather than phylogeny. Apparently development of a penial stimulator has occurred several times in the Trochomorphinae and is an advanced character. The subgenus Trochomorpha, with sections in Tonga, Society Islands, and the New Hebrides, is probably composed of parallel termini of sepa- rate phylogenetic lines. Until the Indonesian Trochomorphinae have been thoroughly studied, the present classification can be utilized. Two species, Trochomorpha rubens Hartman and T. bakeri, new sp., are recognized from the New Hebrides. Quoy and Gaimard (1832-35, p. 127, pi. 10, figs. 22-25) figured a Trochomorpha from Vani- koro Island as "Helix" exclusa Ferussac, a Moluccan species. Photo- graphs (pi. 12, figs. 3-5) of the Vanikoro shell (courtesy of Andr6 Franc, Paris Museum) indicate that the figures in Quoy and Gaimard are very inaccurate in portraying the shape of the shell. The photo- graphs show a shell very similar to T. bakeri from Aneiteum, but with a narrower, more medial color band and a more convex base. Prob- ably the Vanikoro population represents an undescribed species, but SOLEM: MOLLUSCA OF THE NEW HEBRIDES 109 without personal examination of the original specimen or study of topotypic material, nomenclatural recognition would be unwarranted. In T. bakeri the color band is well removed from the carinal sulcus, while in T. rubens it borders the carinal sulcus. Trochomorpha (Hartmanitrochus) rubens Hartman, Plate 6, figs. 1, 2, 4; plate 24, figs. 1-11, 14. Trochomorpha rubens Hartman, 1888, Proc. Acad. Nat. Sci. Philadelphia, 1888: 251, pi. 13, figs. 5, 5a, 56— Aore Island; Pilsbry, 1892, Man. Conch., (2), 8: 129, pi. 57, figs. 1-3; Mabille, 1895, Bull. Soc. d'hist. Nat. d'Autun, 8: 409 — Espiritu Santo; Sykes, 1903, Proc. Malac. Soc. London, 5: 196 — Terebu and Renee Rivers, Espiritu Santo, Havannah Harbour, Vate (J. J. Walker). Trochomorpha convexa Hartman, 1889, Proc. Acad. Nat. Sci. Philadelphia, 1889: 93, pi. 5, fig. 7— Aore Island; Pilsbry, 1892, Man. Conch., (2), 8: 131-132, pi. 30, figs. 20-25; Mabille, 1895, Bull. Soc. d'hist. Nat. d'Autun, 8: 409 — Espiritu Santo (Frangois); Sykes, 1903, Proc. Malac. Soc. Lon- don, 5: 196 — Terebu and Renee Rivers, Espiritu Santo (J. J. Walker). ?Trochomorpha sp. Sykes, 1903, Proc. Malac. Soc. London, 5: 196 — Renee River, Espiritu Santo and Ravenga, Vanua Lava (J. J. Walker). /?awge.— Erromanga, Vate, Malekula, Espiritu Santo and Vanua Lava(?). Material.— ML 9, ML 12 (convexa), ML 26 (all vars.), ML 31b (rubens to convexa), ML 31e, ML 32 (rubens and "dome"), ML 33 (all vars.), ML 36, ML 37 (convexa), ML 39, ML 44, ML 59, ML 63 (rubens), ML 70, ML 74, ML 76, ML 86 (all vars.), ML 95; Aore Island (ANSP 49007, holotype of rubens, ex Hartman; ANSP 60062, holotype of convexa, ex Hartman; CM 62.4408, paratypes of convexa, ex Hartman; CM 62.13887, paratypes of rubens, ex Clapp, Hartman; CM 62.4410, paratypes of rubens, ex Hartman; CNHM 43675, ex Webb, Gude, Layard, type lot of rubens; UMMZ 138890, paratypes of rubens, ex Walker, J. H. Thompson; UMMZ 139618, type lot of rubens, ex Walker, Ponsonby, Layard; MCZ, topotypes of convexa, W. L. Nutting coll. 1944); Segond Channel, Espiritu Santo (ANSP 132673, convexa, ex Brazier; CNHM 43629, ex Webb, Gude, Layard; UMMZ 139730, ex Walker, Ponsonby, Layard) ; Hog Harbour, Espi- ritu Santo (MCZ, rubens to convexa, ex J. H. Daniels) ; Espiritu Santo (UMMZ 161023, rubens to convexa, ex Walker, Ponsonby; AMNH, rubens to "dome," G. S. Banner, Sept. 9, 1943) ; Malekula (UMMZ 139617 and 184324, ex Walker, Ponsonby, Melville; Marsh 8604, ex UMMZ, Walker, Ponsonby, Melville) ; Erromanga (MCZ 12113, ex Layard); New Hebrides (UMMZ 139627, ex Walker, Tomlin). Remarks. — Typologically, rubens (pi. 24, fig. 1) differs from con- vexa (pi. 24, fig. 2) in having a flatter spire, a larger diameter, fewer ^ / <"'' \ . ■"^^-^ \ o m o 1- o to o o CM — susiupads p -ON suaoipads p '0|\| 110 SOLEM: MOLLUSCA OF THE NEW HEBRIDES 111 Table V. — Size Variation in Trochomorpha rubens Hartman Total ML 33 (rubens) ML 33 (convexa) Live Holotype Holotype (rubens) (convexa) No. of specimens 186 41 74 71 Height. . Mean Range S.D. 6.8 5.0-8.9 0.63 6.7 5.6-7.5 0.56 7.2 6.0-8.9 0.54 6.6 5.0-7.5 0.57 6.3 6.9 Diameter.. Mean Range S.D. 13.5 10.6-17.1 1.44 15.4 13.2-17.1 0.76 12.4 10.6-14.5 0.62 13.5 10.6-16. 1.21 14.7 5 13.1 H/D Mean Range S.D. 51 37-74 7.18 44 37-48 5.24 57 49-74 4.21 49 41-64 5.10 42 53 Whorls. .Mean Range 5^-6^ 5^ 5^-6 6 5^-6^ 5H-6 5H 6 whorls and a more prominent red band. Actually rubens is one end of a clinal series, convexa is somewhere in the middle, and an un- named variety, "dome" (pi. 24, figs. 3, 4), is the other extreme. Museum specimens can be fairly easily divided into "rubens" and "conveza-dome," but the Kuntz material contained numerous inter- grades. Including juveniles and museum specimens I saw over 300 individuals of this complex. For statistical analysis, all adult shells collected by Kuntz, that is, those with the lip internally thickened, were measured and the data studied. One lot, ML 33, consisted of a thanatocoenosis of 115 adult shells from a road cut in the Sarakata River Valley. Before this study was started, the lot had been divided into "rubens" and "convexa-dome." The series were analyzed sep- arately and significant morphometric differences found between them (see Table V and fig. 4). Examination of "live" material (71 adults) showed mainly intermediate specimens, one lot, ML 31b (pi. 24, figs. 5-11, 14), containing a complete series while other lots, ML 26, ML 32, ML 39 and ML 86, had less complete series of intergrades. Intergrades were most common between the nomenclatural types of rubens and convexa and less frequent between convexa and "dome." Animals of all forms were available, and two of each variety were dissected. No taxonomically significant variations were found in the radula (pi. 6, fig. 4) or the genitalia (pi, 6, figs. 1, 2). The radular formula was (29-33)-(10-13)-l-(10-13)-(28-33) X 115-130. The teeth are quite typical, with the bicuspid marginals of Trochomorpha (sens, str.) and Nigritella. The unusual features of the genitalia are 112 FIELDIANA: ZOOLOGY, VOLUME 43 discussed under the sectional description of Hartmanitrochus. No major anatomical differences exist between "rubens" and "convexa- dome." There is, however, an indication of slight differences in the proportions of the mantle organs and parts of the genitalia. The differences are small, and not enough material was available to prove whether or not they are statistically significant. Despite the lack of specific anatomical characters and the inter- mediate nature of the shells of living specimens, the fact remains that museum specimens and the subfossil material from lot ML 33 can be divided into two series which are significantly different mor- phologically. According to Mayr, Linsley and Usinger (1953, p. 146), the two series mentioned above are well above the conventional limit of subspecific discrimination in respect to diameter (coefficient of difference 2.18) and H/D ratio (CD. 1.38). There is no possibility of geographic isolation, since the forms are sympatric (ML 26, ML 31b, etc.). Apparently, however, there is an ecologic differ- ence, as most specimens of "ruhens" were collected on the ground and most specimens of "convexa-dome" on trees. A similar case of habitat subspeciation is found in Trochomorpha nigritella (Pfeiffer) from Ponape in the Caroline islands (H. B. Baker, 1941, pp. 292- 294), and actual ecological speciation occurs in the Raiatean Trocho- morpha typus H. B. Baker and T. swainsoni (Pfeiffer) (H. B. Baker, 1941, p. 318). In the case of the New Hebridean T. rubens complex, field studies are needed to establish the exact degree of ecological separation. All the museum specimens of T. rubens were collected prior to the wide-scale establishment of plantations; indeed, most of them were found by new settlers while they were clearing land for plant- ing. The subfossil specimens from ML 33 almost certainly were de- posited long before extensive cultivation was practiced. These specimens can be easily divided into rubens and convexa-'dovae" and intergrading individuals are few. The living material collected by Kuntz had a much higher percentage of intergrading individuals and was mainly obtained at or near the boundary between jungle and plantation. There is also a possibility that "rubens" and "convexa- dome" are well differentiated and isolated in natural jungle areas, but that in cultivated or ecotonal areas the complete or partial isola- tion breaks down. Without field studies the exact relationship of "rubens" and "con- vexa-doTcie" must remain uncertain. On the basis of variation found in the Kuntz material, they are here united in a single species. How- SOLEM: MOLLUSCA OF THE NEW HEBRIDES 113 ever, they may be ecologically isolated subspecies or even species which hybridize under "artificial" plantation conditions. The distributional limits of the rubens complex is uncertain. Specimens from Malekula are conchologically identical with typical rubens as is the single specimen seen from Erromanga. No material from Vate was available, nor could I determine the status of the form reported by Sykes from Vanua Lava and Espiritu Santo as "approach- ing approximata." The presence of rubens on Malekula is not sur- prising, but the Erromanga record needs confirmation, particularly since Layard, in a letter to Hartman (March 17, 1888) refers to bakeri as living on both Aneiteum and Erromanga. Trochomorpha (Hartmanitrochus) bakeri, new species. Plate 24, figs. 13, 15-17. Helix apia Pfeiffer (not Hombron and Jacquinot, 1852), 1854, Conch. Icon., Helix, pi. 199, fig. 1402— Aneiteum (MacGillivray) ; Pfeiffer, 1859, Mon. Helic. viv., 4: 183— New Hebrides; Tryon, 1887, Man. Conch., (2), 3: 88, pi. 17, figs. 45-46 — New Hebrides. Helix (Discus) apia Pfeiffer (not Hombron and Jacquinot, 1852), Cox, 1868, Ex- change list, p. 38, no. 7 — Aneiteum. A species of Trochomorpha distinguished from the rubens complex by its heavier shell, coarser surface sculpture, sharper, wider keel, and lighter ground color, and by having the spiral color band well removed from the carinal sulcus. The anatomy is unknown. Shell lenticular to trochiform, thick, solid, base slightly convex; carinate, with a broad, prominent sulcus above the periphery, greatly reduced below. Straw- colored, with (or without) supracarinal and basal red spiral bands well removed from the straw-colored carina. Sculpture of equally vague protractive and retrac- tive striae, crossed by prominent growth wrinkles on the later whorls. Embryonic whorls with fainter sculpture than the spire, eroded in most adults. Aperture ovately rostrate, upper lip thin, slightly indented, lower lip strongly thickened internally. Parietal callus thin. Umbilicus variable in size, 1.8-2.3 mm. in diam- eter, contained 6.3-9 times in the base. Diameter of holotype 14.5 mm., height 7.8 mm., whorls 5^; umbilicus 2.2 mm., contained 6.5 times in the diameter. Type. — University of Michigan Museum of Zoology no. 184722. Collected on Aneiteum Island, New Hebrides. Specimen from the Walker collection. Paratypes. — The following specimens examined during this study may be considered paratypes: Aneiteum (UMMZ 139690, ex Walker, Ponsonby, Layard, the "type lot"; AMNH, Macmillan!, 1937; CM 62.4406, ex Hartman, Layard; CM 62.13885, ex Clapp, Mazyck, Layard; CNHM 43696, ex Webb, Gude, Layard); New Hebrides (MCZ, ex Thaanum, Shackleford; MCZ, from unknown source; 114 FIELDIANA: ZOOLOGY, VOLUME 43 UMMZ 139689, ex Walker, Tomlin; UMMZ 139691, ex Walker, Ponsonby; ANSP 1932, ex A. D. Brown; Marsh 8544, ex UMMZ, Walker, Ponsonby). Range. — Aneiteum, but may be found on Tanna and Erromanga. Remarks. — H. B. Baker (1941, pp. 311-313) restricted the name apia Hombron and Jacquinot, 1852, to a Samoan shell and pointed out the error in using apia for the New Hebridean species. Since the New Hebridean species were not included in Baker's monograph, he did not propose a new name, based on Reeve's illustration, for the Aneiteum species. Great pleasure is taken in naming this species after Dr. Baker in appreciation of his invaluable assistance. T. hakeri is superficially similar to the Samoan T. apia, but the latter has on the embryonic whorls a strong spiral sculpture which is lacking in T. baker i. Although the anatomy of T. hakeri is unknown, there can be little doubt that it is closely related to the rubens complex. The sculpture of the two species is identical and the varia- tions in contour and size are quite similar. T. hakeri has been given specific status on the basis of the differences noted in the diagnosis. With the exception of the position of the color band (pi. 24, fig. 13) all of the differences can be considered as adaptations to the drier climate of Aneiteum; increase in shell thickness, prominence of growth wrinkles and lightening of coloration are all well-documented adaptations of snails to xeric conditions. The position of the color band has proven to be indicative of specific differences in other island groups, so that, despite lack of anatomical material, specific status for T. hakeri is warranted. A series of specimens showing variation in contour and color is illustrated (pi. 24). The morphometry of the twenty-seven available specimens is summarized in Table VI. Table VI. — Size Variation in Trochomorpha hakeri Height Diameter H/D Whorls Mean 7.4 15.1 49 bV^ ^ Range 6.5-9.2 13.6-16.3 42-62 5^-53^ S.D 0.55 0.73 5.07 Through the courtesy of Dr. Donald F. McMichael of the Aus- tralian Museum (Sydney), it was possible to determine the identity] of most of the unfigured species described by Cox (1870) . One "zoni- tid," Helix vannaelavae Cox, 1870, could not be located in the Aus-| tralian Museum collection and its identity remains unknown. The' original description is reprinted below. i SOLEM: MOLLUSCA OF THE NEW HEBRIDES 115 Helix vannaelavae Cox. Helix vannaelavae Cox, 1870, Proc. Zool. Soc. London, 1870: 82-83 — Aneiteum (Brazier!). Shell imperforate, conoidly depressed, dark yellowish horn colour, smooth, shining; spire broadly conoid; whorls 5^, only slightly convex, last not descend- ing, base convex; suture shallow; aperture lunate; peristome thin, simple, columel- lar margin a little dilated above. Diam., greatest 0.14, least 0.12; height 0.11 of an inch. No shell seen during this study even slightly approximated the brief description. The four small zonitoid snails known from Anei- teum, Coneuplecta (Sitalina) microconus, Ldardetia samoensis, Lam- procystis guttula, and Orpiella retardata, all have conspicuous umbilici and a different shape than Helix vannaelavae. Specimens of the following two species were in museum collec- tions with a label stating they were from the New Hebrides. They almost certainly represent mislabeled specimens rather than valid records. Inozonites bicarinata (Semper). Plate 24, fig. 12. Euplecta bicarinata Semper, 1870, Reisen im Philippinem, (2), 3: 16-17, pi. 2, fig. 8a-c — Bery Arayat, Pampanga, Luzon, Philippine Islands; Faustino, 1930, Philippine Jour. Sci., 42, (1), p. 95— Luzon. Trochomorpha pulcherrima Hartman, 1890, Proc. Acad. Nat. Sci. Philadelphia, 1890: 288, pi. 3, fig. 13— Aore Island, New Hebrides. Range. — Luzon, Philippine Islands. Material. — Aore, New Hebrides (CM 62.5825, ex Hartman, Lay- ard, holotype of pulcherrima Hartman); Morong, Luzon (UMMZ 79824, ex Walker, Ponsonby, Moellendorff ; ANSP 63988, ex Moel- lendorff); Manila, Luzon (ANSP 148327, ex Walker, Fulton). Remarks. — The unique holotype of pulcherrima cannot be dis- tinguished from Philippine specimens of Inozonites bicarinata. Prob- ably either Layard or Hartman mixed the Philippine shell with New Hebridean material in packing specimens for shipment. The holo- type of pulcherrima measures 6.2 mm. in height and 11.8 mm. in diameter, and has six whorls. Mendana (Tahuatoa) garrettiana (Garrett) A single specimen of this Marquesan species was in a lot of Orpi- elh retardata retardata (Cox) from Vate Island (UMMZ 147918, ex 116 FIELDIANA: ZOOLOGY, VOLUME 43 Walker, Ponsonby, Ancey, Layard). Ancey had material of M. gar- rettiana (see H. B. Baker, 1938b, pp. 37-38, pi. 15, figs. 3, 4) at the same time (1883-89) that he was receiving New Hebridean shells. It is quite possible that specimens became mixed in preparing ex- change collections. Suborder HOLOPODA Pilsbry, 1896 Sigmurethrous snails with the pedal grooves usually rather inconspicuous and at or close to the angle of the foot. (After Pilsbry, 1946b, p. 231.) Several groups of families can be recognized within the Holo- poda, but because relationships are still uncertain, no taxa above the family have been formally recognized below. The content and dis- tribution of the broad categories are briefly summarized, and discus- sions of their characteristics can easily be found in Pilsbry (1894, 1900a) and Thiele (1931) or the cited references. Watson (1915) demonstrated the polyphyletic nature of Pelse- neer's Agnatha, but there are still several families which seem to be more closely related to each other than to any of the herbivorous taxa. Further study may reveal their vegetarian ancestors, but until then the Agnathomorpha remains a useful concept. Pilsbry (1900a, 1908), Watson (1915), and H. B. Baker (1930, p. 408) pointed out that the Aperidae, Haplotrematidae, Streptaxidae, and Paryphantidae possess many characteristics in common and may be phylogenetically related. Watson (1915) derived both the Aperi- dae and Streptaxidae from primitive paryphantids. The relation- ship of the Haplotrematidae is less certain, but it has been included here on the authority of H. B. Baker. The great majority of the holopod snails have a vegetarian diet. Our knowledge of their classification is still essentially that of Pils- bry (1894, 1900a) as subsequently modified by Pilsbry himself. The herbivorous holopod Sigmurethra consist of five groups: the Achatinacea, Bulimulacea, Polygyracea, Helicacea, and Clausiliacea. Only the Bulimulacea are native to the New Hebrides, ^ but a brief survey of the distribution of the others is included for subsequent reference. The Achatinacea are predominately African, but several genera seem to be native to the West Indian-Brazilian region (Pilsbry, 1906, p. vi), and the relict family Megaspiridae is found in Brazil, Obi in 1 The genus Draparnaudia (p. 121) has been left in the helicacean family Camaenidae, but its position is doubtful. SOLEM: MOLLUSCA OF THE NEW HEBRIDES 117 the Moluccas, New Guinea, and Queensland. A few small achatina- ceans have been widely spread by commerce and are now practically ubiquitous. The Bulimulacea are primarily South and Middle American, with two taxa, Bothriembryon and Placostylus, in the Pacific and possibly one genus, Prestonella (see p. 123), in South Africa. The Bulimulidae are found throughout Middle and South America with Bothriembryon and Placostylus in the Pacific and Prestonella in Africa. Iredale (1937a, 1944) created families for the two Pacific taxa, but I am ignoring his action. The other family, the Urocoptidae, is found in the West Indies, Central America, and northern South America, and a few species reach the southwestern United States. The Clausiliacea are found in Europe, northern Africa, Asia, South America, and the West Indies. A few species (fig. 14) reach the Indo-Australian archipelago (see Loosjes, 1953), and one species has been found in New Guinea (Loosjes, 1956). The Clausiliacea are of uncertain affinities, but H. B. Baker (1956a, p. 131) suggests that they may be related to the Bulimulacea. The Polygyracea (Pilsbry, 1948, p. vi) probably are the most primitive existing stock of the true helices. They are restricted to North and Middle America and are the only holopod mollusks found in eastern North America. The western part of North Amer- ica served as a causeway for the helices which populated the West Indies, Central and South America, but none were able to cross the Great Plains region. The Helicacea represent the most advanced land snails. Within this broad taxon, many families can be recognized. The more prim- itive families are found in southern areas; the more specialized fam- ilies are northern in distribution. The over-all pattern of distribution has been interpreted to indicate a southern Asian origin for the Helicacea. Of all the holopod superfamilies, only the Agnathomorpha and Bulimulacea are native to the New Hebrides. The Helicacea may possibly be represented by Draparnaudia but until its anatomy has been fully studied, it is uncertain whether Draparnaudia is a bulim- ulid, a camaenid, or an orthurethrous tornatellinid. No attempt has been made to define the higher taxa of the sev- eral introduced holopods, or to characterize the superfamilial cate- gories outlined above. 118 FIELDIANA: ZOOLOGY, VOLUME 43 Family SUBULINIDAE Many subulinids have been accidentally imported on plants into tropical countries or greenhouses of the temperate zone. Thaanum (1927) illustrated the frequency with which they are found during customs inspection of plants and produce entering Hawaii. The sub- ulinids found in the New Hebrides are obvious importations; for a full discussion of their classification see Pilsbry (1946b) . Subulina octona is the largest of the three species. It is imper- forate at all stages of growth and has much larger nuclear whorls than Lamellaxis gracilis, which is perforate in young shells and has rather small nuclear whorls. Opeas pumilum is characterized by its sinuous lip and arcuate sculpture. Sykes (1903, p. 198) reported Lamellaxis oparanum (Pfeiffer) from Vate, Vanua Lava, and Valua. Probably the record was based on the ubiquitous L. gracilis. Other wide-ranging subulinids such as Lamellaxis micra (Orbigny) and L. mauritianum (Pfeiffer) will eventually be found in the New Hebrides. Subulina octona (Bruguiere) Bulimus octonus Bruguiere, 1792, Encycl. Meth., I, p. 325. Subulina octona (Chemnitz), Sykes, 1903, Proc. Malac. Soc. London, 5: 198 — Vila, Vate, New Hebrides. Subulina octona (Bruguiere), Pilsbry, 1906, Man. Conch., (2), 18: 72, pi. 12, figs. 8, 9; van Benthem Jutting, 1952, Treubia, 21, (2), pp. 376-378, figs. 52-53. Range. — Vate, Epi, and Espiritu Santo. Circum tropical. Material.— ML 8, ML 14, ML 21, ML 23, ML 26, ML 31, ML 34, ML 39, ML 49, ML 69, ML 70, ML 95; Vila, Vate (USNM 598357, Miller!); Vate (DMNZ, W. H. Dawbin!); Nivenue, Epi (DMNZ, A. G. Horwell!). Remarks. — Subulina octona probably was native to the Western Hemisphere and introduced into the Pacific in comparatively recent times (Pilsbry, 1906, p. 74). In the New Hebrides it is very com- mon in cultivated areas and near houses. Although it is found under decaying logs, sticks, and leaves, its favorite habitat seems to be in decaying coconuts. Opeas pumilum (Pfeiffer) Helix goodalli Miller, 1822, Ann. Phil., n.s., 3: 381 (not Ferussac, 1821). Bulimus pumilus Pfeiffer, 1840, Arch. Naturg., p. 252. Opeas pumilum (Pfeiffer), Pilsbry, 1946, Monog. N. Amer. Land Shells, 2, (1), pp. 181-182. SOLEM: MOLLUSCA OF THE NEW HEBRIDES 119 Range. — Espiritu Santo and Vate. Circum tropical. Material.~ML 95; Vate (DMNZ, W. H. Dawbin!). Remarks. — The small nuclear whorls and obese shape separate 0. pumilum from Suhulina octona. Lamellaxis gracilis is much slenderer and has stronger sculpture than the other two species. Lamellaxis (AUopeas) gracilis (Hutton) Bulimus gracilis Hutton, 1834, Jour. Asiat. Soc. Bengal, 3: 93. Stenogyra juncea Mabille (not Gould), 1895, Bull. Soc. d'Hist. Nat. d'Autun, 8: 411— New Hebrides. Opeas gracile (Hutton), Pilsbry, 1906, Man. Conch., (2), 18: 125-132, pi. 18, figs. 3-6; van Benthem Jutting, 1952, Treubia, 21, (2), pp. 378-380, figs. 55, 56a. Range. — Aneiteum and Espiritu Santo. Circumtropical. Material— ML 8, ML 14, ML 23, ML 31c, ML 33, ML 35, ML 39, ML 69, ML 74, ML 76, ML 95; Espiritu Santo (USNM 432453, Harrington!); Aneiteum (AMNH; DMNZ, W. H. Dawbin!). Remarks. — Although L. gracilis and Suhulina octona were found together at four localities on Espiritu Santo (excluding the drift sam- ple, ML 95), most of the specimens were collected at different sites. Possibly the two species are partially ecologically isolated, but no definite data are available. Family BRADYBAENIDAE (=Eulotidae) Key references to the classification of the Bradybaenidae are Connolly (1939) and Pilsbry (1939, p. 15). The only species found in the New Hebrides, Bradyhaena similaris (Ferussac), has been widely distributed by commerce. Bradyhaena similaris (Ferussac) Helix similaris Ferussac, 1821, Tabl. Syst. Limacons, 3: 47. Bradyhaena similaris Ferussac, van Benthem Jutting, 1950, Treubia, 20, (3), pp. 501-503, figs. 104-105. Range. — Vate and Espiritu Santo. Circumtropical. Material— ML 18, ML 20, ML 26, ML 31c, ML 31e, ML 34, ML 39, ML 44, ML 50, ML 51, ML 63, ML 64, ML 66, ML 70, ML 94; Vila, Vate (Miller, USNM 598356). Remarks. — Bradyhaena similaris is native in eastern Asia from southern China to Java and Celebes. There it lives in the "wild" as well as in "cultivated" areas. Opinions differ as to the means of 120 FIELDIANA: ZOOLOGY, VOLUME 43 dispersal. Many consider that it has traveled with the coffee tree, but Pilsbry (1894, pp. 203-204) suggested that sugar cane serves as a more likely carrier. Both banded and unhanded shells were found, occasionally in equal numbers, sometimes with one predominating. Van Benthem Jutting (loc. cit.) gave a good summary of the distribution, varia- tion, and ecology of this species, and Komai and Emura (1955) ana- lyzed the genetical basis for its color variation. Family CAMAENIDAE (= Pleurodontidae) Helices without dart apparatus; penis continued in an epiphallus and a flagel- lum (the latter sometimes vestigial or wanting) ; spermathecal duct not branched (Pilsbry, 1939, p. 411). Camaenids are widely distributed in the Indo-Australian and American tropics, and in the Palaeo-Oriental land snail region they form the "dominant" land snail taxon. Both anatomy and distri- bution place the camaenids as more primitive than and ancestral to the dart-bearing helicids of Eurasia and western North America (see Pilsbry, 1894) . The camaenids have occupied a number of habitats, producing such terrestrial genera as Chloritis (sens, lat.), Ther sites, and Rhagada, arboreal genera such as Papuina and Amphidromus, and desert snails such as Glyptorhagada and Xanthomelon. Iredale (1937b, 1938, 1941) grouped the Austro-Melanesian spe- cies into several "families." Few Austro-Melanesian species have been dissected, but the published data show the same types and ex- tent of variations that are found in the West Indian camaenids (see Wurtz, 1955). Iredale's creation of families for Chloritis, Papuina, Xanthomelon, Hadra, and Rhagada ignores important and obvious affinities to such Malayan taxa as Amphidromus, Landouria, and Ganesella. The West Indian genera show an equal amount of concho- logical variation, and careful study of their anatomy has shown their essentially similar basic structure. Iredale is consistently unable to recognize that diversity in con- chology is often produced by ecological adaptations and that arboreal and terrestrial genera such as Papuina and Chloritis, despite strik- ingly different shells, are actually closely related. Many "Papuina" and "Chloritis" are found in the Solomons, Bismarcks, and New Guinea, but none have been found on any New Hebridean island. Jaeckel and Schlesch (1952, pp. 157-158, pi. 6, figs. 5a-c; see also Watson, 1953, pp. 99-100) described Papuina SOLEM: MOLLUSCA OF THE NEW HEBRIDES 121 charlottae from a single shell supposedly found on Santa Cruz Island. P. charlottae is closely related to the Guadalcanal P. caerulescens Angas and an unnamed form from Florida Island, Solomon Islands (Solem, in press-A). Iredale (1927) did not mention any Santa Cruz Papuina, nor were any collected by W. M. Mann. It is suspected that Papuina charlottae was actually collected in the Solomons rather than on Santa Cruz Island because there are other errors in Jaeckel and Schlesch (1952), only a single specimen is known, and the source of the specimen has been responsible for other locality errors. The presence of a Papuina on Santa Cruz Island would be very significant zoogeographically, but the record of Jaeckel and Schlesch cannot be accepted as valid. A genus of small terrestrial snails found in the New Hebrides, Loyalties, and New Caledonia (fig. 17), Draparnaudia, usually has been placed in the Camaenidae. Its resemblance to a miniature Amphidromus or Pseudopartula perhaps influenced its inclusion in the Camaenidae rather than the Bulimulidae. The shells which most resemble Draparnaudia belong to the Hawaiian tornatellinid genus Auriculella. Genus Draparnaudia Montrouzier, 1859 Shell small, sinistral, perforate, turbinate-conic, covered with a yellow or brown cuticle. Whorls 5J^ to 7, convex, very obliquely striated. Aperture very oblique, truncate-oval, the peristome expanded or simple, columellar margin broadly dilated and built forward, columella simply reflexed (Pilsbry, 1901, p. 13). Type species. — Draparnaudia michaudi Montrouzier (= Helix sin- istrorsa Deshayes, 1840). Moss and Webb (1897) figured the jaw, radula, and terminal genitalia of Draparnaudia lifuana Pilsbry. Pilsbry (1899), placed Draparnaudia in the Camaenidae on the basis of those illustrations but stated that "until the pallial region is investigated, we cannot be certain that it is not a member of the Bulimulidae." On the basis of the published drawings of the anatomy, I consider it probable that Draparnaudia represents a tornatellinid which has lost the penial appendage rather than a camaenid which has lost the characteristic flagellum. Without study of the pallial region, however, speculation on affinities is useless, and I have left the species, with some hesita- tion, in the Camaenidae. Six species and a number of varieties of Draparnaudia have been recognized (Pilsbry, 1901, pp. 12-18, 283, Ixxi; Sykes, 1903; and 122 FIELDIANA: ZOOLOGY, VOLUME 43 Ancey, 1905). The species from New Caledonia and the Loyalty Islands have been reviewed by Franc (1957, pp. 162-164). The New Hebridean species are reviewed below. Key to the New Hebridean Draparnaudia 1. Shell less than 7 mm. high; six whorls; obtusely or not keeled at the periphery. . 2 Shell 9-10 mm. high; 6^ whorls; acutely keeled at periphery. D. s. singularis (Pfeiffer) 2. Shell rounded at periphery of body whorl; Espiritu Santo. . .D. walkeri Sykes Shell obtusely keeled at periphery; last whorl not deflected in front; Aneiteum. D. s. diminuta Ancey Draparnaudia singularis singularis (Pfeiffer). Plate 8, fig. 7. Helix singularis Pfeiffer, 1854, Proc. Zool. Soc. London, 1854: 290 — Aneiteum (MacGillivray!). Helix (Geotrochus) singularis Pfeiffer, Cox, 1868, Exchange List, p. 44, no. 103 — Aneiteum. Bulimus sinistrorsus var. carinatus Gassies, 1876, Faune Conchy. N. Cale- donie, 2: 92 — Isle of Pines. Draparnaudia singularis (Pfeiffer), Pilsbry, 1901, Man. Conch., (2), 14: 14-15, pi. 3, fig. 1; Dautzenberg, 1923, Nova Caledonia, 3, (1), p. 143— Uvea, Loyalty Islands; Franc, 1957, Mem. Mus. d'Hist. Nat., (n.s.), Zool., 13: 162-163, pi. 21, fig. 216. Range. — Aneiteum. New Caledonia and Loyalty Islands. Material— 'New Hebrides (ANSP 31512) ; New Caledonia (MCZ 74686, UMMZ 145974) ; Isle of Pines (UMMZ 146973) . Photograph of holotype (BM). Remarks. — Draparnaudia singularis is the only species of land snail found in both New Caledonia and the New Hebrides which has not been obviously introduced by man. I have based the spe- cific identity of the New Hebridean and New Caledonian specimens on a comparison of the photograph of the holotype with New Cale- donian shells, since no New Hebridean material with exact locality data was available to me. The very prominent thread-like keel on the body whorl easily separates D. s. singularis from the other New Hebridean morphs. Draparnaudia singularis diminuta Ancey Draparnaudia singularis Ancey (not Pfeiffer), 1897, Nautilus, 11: 27 — Aneiteiun. Draparnaudia singularis diminuta Ancey, 1905, Nautilus, 19: 42-43 — Aneiteum. Range. — Aneiteum. Material. — No material available. SOLEM: MOLLUSCA OF THE NEW HEBRIDES 123 Remarks. — According to the original description, this variety is smaller and less acutely keeled than typical singularis. It has not been figured, and I have examined no specimens. Draparnaudia walkeri Sykes. Plate 8, fig. 8. Draparnaudia walkeri Sykes, 1903, Proc. Malac. Soc. London, 5: 197, fig. 2 — Renee River, Espiritu Santo. Range. — Espiritu Santo. Material. — ML 95. Photograph of holotype (BM). Remarks. — The small size and slightly keeled periphery distin- guish D. walkeri from D. s. singularis. According to Ancey (1905, pp. 42-43), D. walkeri has a higher body whorl and more rounded periphery than D. s. diminuta. A juvenile specimen collected by Kuntz has been referred here, although it is too young to have devel- oped the characters used in specific determination. Family BULIMULIDAE Shell ovate to cylindric. Body without visible pedal grooves. Kidney short, triangular, cardiac edge equal to pericardium in length, secondary ureter complete. Penis entering epiphallus without external distinction; epiphallus often continued in a flagelliform appendage. Penial retractor attached at or near end of fiagellum or epiphallus. Jaw composed of vertical or medially converging imbricating plates. Radula of helicid type; central tooth as wide as adjacent laterals or somewhat narrower. Right ocular retractor muscle arising from columellar muscle; left ocular united shortly or for some distance with pharyngeal retractor. (Adapted from Pilsbry, 1946b, p. 1.) The Bulimulidae are primarily South American, reaching maxi- mum development in Brazil, where seventeen genera are found (Pils- bry, 1946b, p. 2). A few genera extend north into Mexico, the West Indies, and the southern United States; others are found in Austra- lia, New Zealand, and Melanesia (fig. 15). Two African genera, Aillya (see pp. 33, 35) and Prestonella (see Connolly, 1939), have been referred to the Bulimulidae, but their taxonomic position is uncer- tain. H. B. Baker (1955) considers Aillya to be a relative of the Succineidae, and (personal communication) thinks that Prestonella may also belong to the Heterurethra. No study on the anatomy of Prestonella has been published, and determination of its affinities must await dissection of the animal. The exact phylogeny of the Bulimulidae is uncertain, but the present distribution, comparative anatomy and fossil record suggest a South American adaptive radiation in the Mesozoic and Tertiary. 124 FIELDIANA: ZOOLOGY, VOLUME 43 Fossil bulimulids, in many cases referable to modern genera, are found in the Eocene and Paleocene of Patagonia. In Brazil, bulimu- lids do not appear until the middle Tertiary (Parodiz, 1946, 1949). Outside of South America, the only known fossil bulimulids are from the Miocene of Florida. The genus occurring there, Hyperaulax, still lives on Fernando Noronha off Brazil and is a primitive member of the Odontostominae. During the Miocene, Florida land snails were Antillean in affinities and origin. The living Hyperaulax may be the last remnant of an early northward radiation which reached the in- sular fringes in the Miocene but has later been replaced by more advanced bulimulids. Apparently the bulimulids are entering North America in the wake of the warming climate. The affinities of the Austro-Melanesian bulimulids have long been doubtful. Iredale (1937a, 1944) created separate families for the two Austro-Melanesian taxa, Bothriembryon and Placostylus. Recent studies by Pilsbry (1946a) and Kondo (1948) reaffirmed the bulimu- lid relationship of both genera. Their positions within the family differ, however. Bothriembryon (fig. 20) is found in southwest Australia and Tas- mania. The northern limit seems to be Shark's Bay, the eastern limit is Port Lincoln, and there is a single species in Tasmania. Ire- dale (1933, 1937a, 1939) studied Bothriembryon and in the process proposed at least eight new generic or subgeneric names and twenty- three new specific ones. The embryonic shell sculpture varies widely, and exactly parallels the variations found in South American Bulimulus (Pilsbry, 1900c, pp. 1-2, pi. 4). The central Australian Bothriembryon spenceri Tate and the Tasmanian B. gunnii (Sowerby) (=tasmanicus Pfeiffer) have the most primitive embryonic sculpture, and the species of King George's Sound have the most advanced type. Probably this indi- cates a secondary radiation from a southwest Australian "refuge zone" rather than an original radiation from south to north. The area in which Bothriembryon lives has a very limited molluscan fauna and only recently has been re-united to the rest of Australia. Previ- ously the large central sea and desert provided effective isolating mechanisms. Pilsbry (1901, p. 1; 1946a) examined the anatomy of Bothriem- bryon and found it essentially the same as the more primitive South American bulimulids. On the basis of conchology, ecology, and anat- SOLEM: MOLLUSCA OF THE NEW HEBRIDES 125 omy there is absolutely no basis for more than generic separation of Bothriembryon and Bulimulus. Both Pilsbry (1901) and Kondo (1948) dissected Placostylus. In most characters it is the same as the South American bulimuline genera, but the shortened spermathecal duct, very capacious penis, and rather enlarged pericardial vein of the species studied led Pils- bry (1946a) to propose a subfamily, the Placostylinae, for Placostylus and Diplomorpha. Iredale (1944, p. 309) listed a name, Placostyli- nae, but gave no diagnosis or identifying characters. The Inter- national Rules of Zoological Nomenclature, at the present time (Hemming, 1953, p. 35, paragraph 52), accept such nomina nuda as being nomenclaturally published. If priority for family names is accepted, Iredale must be credited with the family name, although Pilsbry is certainly the taxonomic author. Actually the anatomy of Diplomorpha (see Kondo, 1948) and the New Hebridean Placostylus (see p. 128) is transitional between the generalized South American Bulimulinae and Bothriembryon and the specialized Placostylus on which Pilsbry based his Placostylinae. The subfamily Placostylinae is weakly differentiated, but it can be re- tained in the classification pending more comprehensive investigations. If Dryptus is added and Prestonella, Aillya, Partula, and Peltella are removed, the classification of the Bulimulidae in Thiele (1931) adequately summarizes the New World Bulimulidae. The Placo- stylinae are considered in detail below. Subfamily PLACOSTYLINAE Bulimulidae with a short, capacious penis, enlarged first branch of the peri- cardial vein, and shortened spermatheca (in some species). (Modified from Pils- bry, 1946a, p. 3.) In this study only two genera are recognized in the Placostylinae. Placostylus will probably be split into genera along the lines indicated below, but further study is needed before additional units should be recognized. Diplomorpha is restricted to the New Hebrides. Key to the New Hebridean Placostylinae 1. Shell large, height more than 29 mm Placostylits 2 Shell small, height less than 29 mm Diplomorpha 4 2. Umbilical chink completely closed fuligineiis (Pfeiffer) Umbilical chink partially to widely open 3 126 FIELDIANA: ZOOLOGY, VOLUME 43 3. Shell thin, without prominent surface sculpture; habitat arboreal. (a) Espiritu Santo bicolor (Hartman) (b) Erromanga turneri (Pfeifler) Shell thick, with prominent surface sculpture; habitat terrestrial. (a) Santa Cruz Island; apical sculpture with 10 ridges/mm. hullianus (Iredale) (b) Aneiteum to Espiritu Santo; apical sculpture with 18 ridges /mm. salomonis (Pfeiffer) 4. Shell ovate, 33^-4^ whorls Diplomorpha (sens, str.) 5 Shell elongate, 4J^-55^ whorls, surface with longitudinal sculpture. Diplomorpha {Quires) bernieri (Hartman) 5. Shell with a distinct parietal tooth layardi Ancey Shell without a parietal tooth 6 6. From southern New Hebrides (Aneiteum?) coxi (Pease) From Espiritu Santo or outlying islands 7 7. Reflexed portion of lip orange, lip only slightly expanded . . . brazieri Hartman Reflexed portion of lip white, lip broadly expanded 8 8. Spire short, last whorl broadly expanded, umbilicus widely open . .peasei (Cox) Spire more elongate, umbilicus less widely open delautouri Hartman Genus Placostylus Beck, 1837 Type species. — Placostylus hootis Menke (emended), which has been subsequently restricted to the New Caledonian Ldmax fibratus Martyn, 1789 (see Pilsbry, 1900c, p. 19). Remarks. — Although Martyn's work has been rejected for use in nomenclature (Opinion 456 of the International Commission on Zoo- logical Nomenclature), an effort is being made to conserve many of his names. The nomenclatural status of a New Caledonian species is outside the scope of this study and Martyn's name is quoted above, although it may be nomenclaturally replaced at a later date. Species assigned to Placostylus (sens, lat.) are known from the Solomon Islands south of Bougainville, the Santa Cruz group, the New Hebrides, the Fijis exclusive of the Lau Archipelago, New Cale- donia and the Loyalty Islands, Lord Howe Island, and the tip of North Island in New Zealand (figs. 15, 20). Hedley (1898, p. 97) described a "Placostylus'' from New Guinea, but the relationships of this species are very uncertain (see Iredale, 1941, p. 64). No Placo- stylus, living or fossil, has ever been reported from Australia, Indo- nesia, or the Pacific Islands west of the Fijis. Because of their large size and striking coloration, specimens of Placostylus are great favorites with shell collectors. Individual shells vary greatly and, like stamp collectors, shell collectors want "new" kinds. As a result, nearly every morphological variation in Placo- stylus has been called a species, and the literature abounds with names SOLEM: MOLLUSCA OF THE NEW HEBRIDES 127 of little or no taxonomic significance. Crosse, Kobelt, Hedley, and Pilsbry attempted to untangle the nomenclatural chaos, but much work needs to be done. In this study several unsorted series of New Hebridean Placostylus were available. The range of variation found in these series is so great that I have synonymized most of the New Hebridean "species." Union of several obviously "distinct" species will seem erroneous to some, so the variation is very fully illustrated (pis. 17-22). In the past, too many species have been described on the basis of one or two individuals, and no consideration has been given to the possibility of variation. It is possible that field studies will demonstrate the presence of local, morphologically stable populations which correspond to some of the "species" here suppressed. The available evidence suggests that the variation is infrapopulational, rather than between popula- tions, but more data are needed. In New Zealand (Powell, 1947, 1951) and New Caledonia (Pain, 1955) local "subspecies" of Placo- stylus have been described, but recognition of similar situations in Melanesia must await detailed field studies. The theoretical implica- tions of the variation in Placostylus are discussed above (p. 24). Not only the variation between specimens but the contrast be- tween arboreal and terrestrial species in one archipelago and the differences between species from different archipelagos have resulted in the proposal of several subgeneric and sectional names. The earlier ones are listed in Pilsbry (1900c, p. 19) ; others have been proposed by Iredale (1927), Haas (1935), and Clench (1941). All the names are based on characters of the shell and often indicate the adaptive differences between terrestrial and arboreal species. While they are useful as a means of recognizing species groups within island chains, they fail to show the phyletic relationships of species found in dif- ferent archipelagos. Characters of possible value in determining phylogenies do exist, but more study is needed before a definitive classification can be established. Haas (1935) was the first to recognize the importance of embryonic sculpture in classification, and it is apparent that, as in the South American Bulimulidae, there are several kinds of sculp- ture. Diplomorpha (sens, str.) has "wrinkle-striate" sculpture (Pils- bry, 1900c, pi. 72, fig. 17) as do the Santa Cruz and New Hebridean Placostylus and the New Zealand and Lord Howe Island species that Haas (1935, p. 188, fig. 3) placed in Maoristylus. The New Hebri- dean Diplomorpha (Quiros) (see Haas, 1935, p. 190, fig. 1) and New Zealand Basileostylus (loc. cit., fig. 4) have longitudinal striae. New 128 FIELDIANA: ZOOLOGY, VOLUME 43 Caledonian species (loc. cit., fig. 1) are finely, densely, and irregularly punctate, while Solomon Island (Pilsbry, 1900c, pi. 72, fig. 18) and Fijian species have regularly spaced, prominent punctations. Pils- bry (1900c, pp. 1-2) considered the striate pattern primitive and the "pitted" sculpture the most advanced. Therefore the Santa Cruz Islands and the New Hebrides, Lord Howe Island, and New Zea- land have primitive species; the Solomon and Fiji Islands have ad- vanced species; and the New Caledonian Placostylus are only slightly less advanced than the Solomon and Fijian taxa. The New Hebri- dean and New Caledonian species are totally unrelated. Pilsbry (1900c), I. and B. Rensch (1935), and Kondo (1948) illustrated the genital anatomy of several species of Placostylus, but their results have not been correlated. Only a few species have been dissected, but the trends in the organs of the genitalia parallel those found in the apical sculpture of the shell. Diplomorpha has a long spermatheca and a long, slender penis with one large pilaster (see Kondo, 1948), which is the condition found in the primitive South American Bulimulinae and Australian Bothriemhryon. Both the Santa Cruz Island (see p. 132) and New Zealand (Pilsbry, 1900c) Placostylus have similar penial structure but possess shortened sper- mathecae. P. Fischer (1871) published crude figures of the genitalia of two New Caledonian Placostylus, but the internal structure of the penis is not shown. Apparently the New Caledonian species have a very short penis and a medium length spermatheca. In the Solo- mon Island (I. and B. Rensch, 1935; Kondo, 1948) and Fijian (Kondo, 1948) Placostylus, both penes and spermathecae are greatly shortened and the penis is greatly modified internally. A long spermatheca, a long unspecialized penis, and wrinkled api- cal shell sculpture seem to be the "primitive" condition, while a short spermatheca, a short specialized penis, and pitted apical shell sculp- ture might be considered "advanced" characters. Before a reclassification of the Placostylinae can be presented, the Lord Howe, New Caledonian, and New Zealand Placostylus will have to be dissected. Even with the limited information available, however, certain tentative conclusions can be reached. In the New Hebrides, New Caledonian, Solomon Island and Fijian Placostylus, arboreal species are more closely related to terrestrial species from the same area than to arboreal species from another island group. The terrestrial and arboreal species of any one island area have come from the same ancestral stock, that is, there has probably been eco- SOLEM: MOLLUSCA OF THE NEW HEBRIDES 129 logical speciation. Similar situations are found in the zonitid genus Trochomorpha (p. 112). Diplomorpha seems to be the most primitive, and the New Hebri- dean and Santa Cruz Placostylus can be easily derived from a Diplo- morpha-\ike ancestor. The New Hebrides have the most primitive Placostylinae, and, in Quiros, the type which is probably nearest the South American Bulimulinae. The anatomical differences between Diplomorpha and the Santa Cruz Placostylus are not very significant, and without the obvious conchological differences generic separation would not be advisable. The New Zealand and Lord Howe Island Placostylus are quite similar to the New Hebridean species. Until the anatomy has been studied, the exact degree of relationship will remain unknown, but on the basis of apical sculpture not even sec- tional separation is justified. The New Caledonian species of Placo- stylus have a completely different apical sculpture. Although the anatomical structures are still unstudied, probably subgeneric or generic separation of the New Caledonian from the New Hebridean species will be justified. The Solomon Island and Fijian species have shell sculpture and anatomical features very different from those of the New Hebridean Placostylus, but they appear to be closely related to each other. These tentative relationships are shown (fig. 5) . A formal classification based on the relationships suggested here is not presented at this time, but the phylogenetic ideas are used in discussing the zoogeography of the Pacific land snails. Hedley (1892a, 1898, 1899) postulated a former Melanesian continent embracing the area from New Guinea to the Fijis and New Zealand. This area of land connections, even if it did not form a continuous mass, was sub- ject to sufficient fluctuations to allow essentially dry land passage of the placostyline shells. Hedley (1892a) suggested an Indonesian de- rivation, but later (1899), recognizing the affinities with the South American Bulimulidae, he proposed an Antarctic migration in Meso- zoic or Paleozoic times. This view was adopted by Pilsbry (1900c, 1911) and several other authors. In the zoogeographical survey (p. 327) it is suggested that less difficulties are encountered if a north- em origin is postulated for the Placostylinae. The New Hebridean species of Placostylus have been placed in two sections, one containing the arboreal, the other the terrestrial species. The terrestrial species, P. hullianus (Iredale), P. salomonis (Pfeiffer), and P.fuligineus (Pfeiffer), are in the section Santacharis; the arboreal species, P. bicolor (Hartman) and P. turneri (Pfeiffer), are in the section Poecilocharis. The apical sculpture of the species is SOLOMON FIJIAN PLACOSTYLUS NEW CALEDONIAN PLACOSTYLUS NEW HEBRIDEAN NEWZEALAND LORD HOWE ISLAND PLACOSTYLUS DIPLOMORPHA QUIROSIELLA BULIMUL1n£' STOCK Fig. 5. Phylogeny of the Pacific Bulimulidae. 130 SOLEM: MOLLUSCA OF THE NEW HEBRIDES 131 identical and the sectional names will probably be dropped when the anatomy has been studied. They are convenient indicators of the habitat and have been retained, but not characterized, pending fur- ther study. There are literature records for several Placostylus which are not actually found in the New Hebrides. Placostylus cuniculinsulae and P. hivaricosus are found only on Lord Howe Island off Australia (Iredale, 1944) and not in the New Hebrides (see Kobelt, 1881). Placostylus stutchburyi and P. palmarum are common in the Solomon Islands (see Clench, 1941) but the specimens in museums labeled "New Hebrides" (UMMZ 816, UMMZ 13891) are from the Cuming collection, which is notorious for its erroneous data. The "Bulimus janus" of Kobelt (1881) is an Amphidromus from Burma and Mergui (see Pilsbry, 1900c, p. 156). Section SANTACHARIS Iredale, 1927 Type species. — Santacharis hullianus Iredale. Remarks. — Variation within the terrestrial species, P. hullianus, P. Suligineus, and P. salomonis, is large and only study of series of specimens clearly reveals the patterns. All three species show clinal variations from slender to obese, smooth to rugose, unicolored to streaked, flaring to indented lip, white to orange-colored aperture, strong to weak columellar sinus, and dark to light ground color. The named varieties are combinations selected from these independ- ent clinal gradients. Population studies may demonstrate the exist- ence of local morphologic races, but the available material suggests that there has been extensive phylosynapsis between local popula- tions (see Hubbell, 1956). P. fuligineus can be easily separated from the other two species by its completely closed umbilical chink, although in a few specimens of P. salomonis (pi. 19, figs. 5-8) the umbilicus is rimate rather than open. The aflfinities of P. salomonis and P. hullianus are less clear, but the difference in apical sculpture probably is indicative of spe- cific separation. Placostylus (Santacharis) hullianus (Iredale). Plate 18, fig. 8. Santacharis hullianus Iredale, 1927, Rec. Australian Mus., 16, (1), p. 77, pi. 5, fig. 1 — Santa Cruz Island. Santacharis hullianus expeditionis Iredale, 1927, Rec. Australian Mus., 16, (1), p. 77, pi. 5, fig. 2— Carlisle Bay, Santa Cruz Island. 182 FIELDIANA: ZOOLOGY, VOLUME 43 Placostylus (Santacharis) salomonis odhneri Jaeckel and Schlesch, 1952, Jour, de Conchy., 92, (4), p. 156, pi. 6, fig. 3— Santa Cruz Island; Watson, 1953, Jour, de Conchy., 93, (3), p. 98. Range. — Santa Cruz Island. Material— Santa, Cruz Island (AMNH). Remarks. — The original descriptions and illustrations of P. hulli- anus, expeditionis, and odhneri show the kinds of variations found in specimens of P. salomonis. No adults from Santa Cruz Island were available, but R. H. Beck collected two lots of juveniles in February, 1927, one from 300 feet elevation, the other from 1000-2000 feet. The specimens from the lower altitude have a more elongate spire, thus paralleling the postulated situation in P. bicolor (see p. 137). The spire of P. hullianus is more swollen than that of P. salomonis and the apical sculpture more prominent (10 ridges/mm. in hullianus; 18/mm. in salomonis). Probably expeditionis and odhneri only rep- resent local populations. Several juvenile animals were dissected. The terminal portions of the genitalia are shown on Plate 6. Although the specimens are not fully mature, a few comparisons with Diplomorpha (Kondo, 1948, pi. 8, and specimens from Omba) can be made. P. hullianus has a shorter, broader penis, longer vagina, and shorter spermatheca, but the internal structure of the penis is the same as in Diplomorpha. Both Diplomorpha and P. hullianus have a single long, rugose pilaster and no trace of the complicated epiphallic structure found in the Solomon Island and Fijian Placostylus. Iredale (1927, p. 77) reported that the field notes of Troughton and Livingston state that the variety expeditionis was found "on leaves." The shell has all the characteristics of a terrestrial species. Since terrestrial shells are occasionally found on the leaves of bushes and succulents, as were the AMNH shells, Iredale's conclusion that they were "undoubtedly collected on trees" cannot be accepted. Placostylus (Santacharis) salomonis (Pfeiffer). Plate 8, figs. 4, 5; plate 17; plate 18, figs. 1-7; plate 19, figs. 5-13. Partula salomonis Pfeiffer, 1852, Proc. Zool. Soc. London, 1852: 137 — Solo- mon Islands (error); Pfeiffer, 1855, Conch. Cab., I, (13), p. 276, pi. 66, figs. 10, 11; Dohrn, 1862, Malak. Blatt., 9: 213— identity of pyrostomus and salomonis. Bulimus pyrostomus Pfeiffer, 1860, Proc. Zool. Soc. London, 1860: 137 — Erro- manga; Brazier, 1890, Jour, of Conch., 6: 79. Bulimus salomonis Pfeiffer, Crosse, 1864, Jour, de Conchy., 12: 131-133, pi. 7, fig. 8. SOLEM: MOLLUSCA OF THE NEW HEBRIDES 133 Placostylus pyrostomus Pfeiifer, Kobelt, 1891, Conch. Cab., I, (13a), p. 68, pi. 16, figs. 5-8. Placostylus salomonis Pfeiffer, Pilsbry, Man. Conch., (2), 13: 69-70, pi. 6, figs. 6-9. Range. — Erromanga, Aneiteum, Tanna, Futuna(?), Espiritu Santo(?). Material. — Erromanga (AMNH) ; Aneiteum (AMNH) ; Tanna (USNM 99692, ex LeGrand; photograph from BM collected by MacGillivray in 1854) ; ?Futuna (AMNH, juvenile) ; New Hebrides (UMMZ 146861, ex Walker, Tomlin; CM 62.2855, ex Hartman); photograph of holotype of pyrostomus (BM). Remarks. — The identity of salomonis and pyrostomus was estab- lished as far back as 1862. Many have preferred to use the later name pyrostomus because of the "misleading" nature of the term salomonis for a New Hebridean species. Brazier (loc, cit.) reported both P. fuligineus and P. salomonis from Aneiteum and this is con- firmed by the material (AMNH). P. salomonis is also known from Terebu, Espiritu Santo (Sykes, 1903, p. 197). The holotype of pyrostomus (pi. 8, fig. 5) is more slender than most of the specimens seen, but it is not too atypical. The holotype of salomonis could not be located in the British Museum. Both "species" were probably collected on Erromanga. Eighteen of the thirty specimens seen were found on Aneiteum and Erromanga by L. Macmillan in 1937 (AMNH collection). The wide range of variation in obesity, spire-aperture ratio, degree of closure of umbilical chink, extent of columellar sinus, shape of aper- ture, prominence of parietal and columellar callus, and extent of surface sculpture is illustrated (pis. 17, 18). All of the above char- acters have been utilized in determining "specific" differences. Specimens from Tanna (pi. 8, fig. 4) are of both the elongate and the globose varieties. Juvenile shells from Futuna Island (AMNH) have the apical sculpture of P. salomonis but are too young for posi- tive specific identification. P. salomonis differs from P. fuligineus in being generally more obese and lighter in color and in always having an open umbilical chink. Color patterns and sculpture overlap, although P. fuligineus is usually darker and more heavily malleated. The orange lip color mentioned by Jaeckel and Schlesch (1952, p. 156) is more often absent than present. 184 FIELDIANA: ZOOLOGY, VOLUME 43 P. salomonis lives "under tufts of grass ... in the mountain ranges of Aneiteum." (Brazier, 1890, p. 79.) No other ecological informa- tion is available. Placostylus (Santacharis) fuligineus (Pfeiffer). Plate 8, figs. 3, 6; plate 19, figs. 1-4; plate 20; plate 21, figs. 1-6. BulirmLS fuligineus Pfeiffer, 1852, Proc. Zool. Soc. London, 1852: 85 — New Hebrides; Pfeiffer, 1854, Conch. Cab., I, (13), p. 157, pi. 48, figs. 5-6; Crosse, 1864, Jour, de Conchy., 12: 129, pi. 7, fig. 4; Brazier, 1890, Jour, of Conch., 6: 80. Bulimu^ fuligineus var. beta Pfeiffer, 1859, Monog. Helic. viv., 4: 363 — Anei- teum (MacGillivray). Placostylus fuligineus Pfeiffer, Kobelt, 1891, Conch. Cab., I, (13a), pp. 44-46, pi. 10, figs. 5-8, pp. 74-75, pi. 17, fig. 8(?); Pilsbry, 1900, Man. Conch., (2), 13: 70-71, pi. 11, figs. 2-4, pi. 14, figs. 9-12. Placostylus alienus Pilsbry, 1893, Nautilus, 6: 116-117 — New Hebrides; Pils- bry, 1900, Man. Conch., (2), 13: 72-73, pi. 8, figs. 18-19. Placostylus heterostylus Pilsbry, 1900, Man. Conch., (2), 13: 72, pi. 7, figs. 11-12 — New Hebrides. Range. — Aneiteum and Erromanga. Mo^erioZ.— Aneiteum (AMNH; USNM 23017, Geddie; ANSP 4883, ex A. Brown, Cox; ANSP 8365, ex A. Brown; CM 62.2856, ex Hart- man) ; Erromanga (AMNH); New Hebrides (UMMZ 13890; UMMZ 146811, ex Walker, Ford; UMMZ 146812, ex Walker, Wetherby; ANSP 8364, holotype of heterostylus, ex Swift, Bland; ANSP 62431, holotype of alienus, ex Cox). Remarks. — P. fuligineus has the same variations in obesity, spire- aperture ratio, columellar sinus and callus, color, sculpture and aper- tural shape that are found in the other species. In some specimens there is a strong columellar cord (see pi. 20, figs. 6-8). Pilsbry's species, alienus and heterostylus, were based on single specimens and represented accentuated development of several clinal gradients. Without the series collected by Macmillan they would have been considered distinct. P. alienus (ANSP 62431, pi. 21, fig. 5) is small and has a very strong columellar cord and callus, the outer lip is slightly indented and there is no columellar sinus. P. heterostylus (pi. 20, fig. 3) is smooth, light-colored, and obese, with a wide, flar- ing aperture, a thin parietal callus and a slightly sinuate columella. The type locality of these two forms is here restricted to Aneiteum. The typical variety of Pfeiffer (pi. 8, fig. 6, courtesy of BM) is a dark, rather squat and obese, heavily malleated shell. The variety beta (pi. 8, fig. 3, courtesy of BM), figured by Kobelt (loc. cit., pi. 17, SOLEM: MOLLUSCA OF THE NEW HEBRIDES 135 fig. 8) is the slender, elongate shell commonly seen in collections. It must be emphasized that these named varieties are not constant — that each specimen presents a different combination of characters selected from the several clinal gradients. On the basis of concholog- ical characteristics, each specimen could be called a different variety. Previously P. fuligineus was reported only from Aneiteum, but one specimen of fuligineus was found in a lot of salomonis from Erro- manga (AMNH). A single specimen of salomonis was in the lot of fuligineus from Aneiteum. Brazier (1890, pp. 79-80) reported find- ing both P. fuligineus and P. salomonis together under clumps of grass on Aneiteum. Field studies are needed to verify the occurrence of both species on Aneiteum and Erromanga and to establish their ecologic relationship. Table VII. — Size Variation in Placostylus salomonis and Placostylus fuligineus P. salomonis P. fuligineus (16 specimens) (17 specimens) Range Mean S.D. Range Mean S.D. Height 36.8-43.2 39.3 2.07 33.7-40.9 37.1 1.99 Diameter 20.6-24.0 22.5 0.98 17.6-20.3 19.2 0.68 H/D 1.60-1.97 1.74 0.12 1.77-2.16 1.94 0.07 Spire 12.6-21.7 15.7 2.22 13.3-18.1 15.9 1.50 Aperture 21.3-26.0 23.6 1.37 19.1-23.0 21.1 1.03 S/A 0.51-0.99 0.67 0.12 0.60-0.88 0.76 0.07 A morphometric table (Table VII) comparing P. fuligineus and P. salomonis is based solely on material collected in 1937 (AMNH). Inclusion of other lots affected mainly the range and lessened the value of this table as a comparison of two simultaneously sampled populations. The major difference between P. salomonis and P. fuli- gineus lies in diameter; 94 per cent of all observed specimens of P. fuligineus were less than 21 mm. and 94 per cent of all observed specimens of P. salomonis were more than 21 mm. in diameter. Section POECILOCHARIS Kobelt, 1891 Type species. — Placostylus hartmanni Kobelt (=hicolor Hartman). Placostylus bicolor (Hartman) and P. turneri (Pfeiffer) are arboreal species closely allied to Santacharis. The apical sculpture is the same in Santacharis and Poecilocharis, while the thinner shell, brighter color, and larger aperture of Poecilocharis are correlated with the arboreal habitat. P. turneri differs from P. bicolor in having a much larger aperture, a less reflexed and thicker lip, and an imperforate umbilicus. 136 FIELDIANA: ZOOLOGY, VOLUME 43 Placostylus (Poecilocharis) bicolor (Hartman), Plate 12, figs. 6-9; plate 22. Charis bicolor Hartman, 1889, Proc. Acad. Nat. Sci. Philadelphia, 1889: 91, pi. 5, figs. 1, la — Aore Island. Charis rossiteri Hartman, 1889, Proc. Acad. Nat. Sci. Philadelphia, 1889: 91, pi. 5, figs. 2, 2a (not Bulimus rossiteri Brazier, 1881). Placostylus (Poecilocharis) hartmanni (sic) Kobelt, 1891, Conch. Cab., I, (13a), p. 78, pi. 18, figs. 7-9 — new name for rossiteri Hartman. Placostylus (Charis) bicolor Hartman, Kobelt, 1891, Conch. Cab., I, (13a), p. 81, pi. 19, figs. 5, 6. Placostylus francoisi Mabille, 1895, Bull. Soc. d'Hist. Nat. d'Autun, 8: 410 — Espiritu Santo; Pilsbry, 1900, Man. Conch., (2), 13: 74-75; Ancey, 1905, Nautilus, 19: 44. Placostylus hebridarum Mabille, 1895, Bull. Soc. d'Hist. Nat. d'Autun, 8: 410 — Espiritu Santo; Pilsbry, 1900, Man. Conch., (2), 13: 75-76. Placostylus (Poecilocharis) hartmani Pilsbry, 1900, Man. Conch., (2), 13: 73- 74, pi. 45, figs. 8-12 — emendation of Kobelt's misspelling. Placostylus (Poecilocharis) bicolor (Hartman), Pilsbry, 1900, Man. Conch., (2), 13: 74, frontispiece, figs. 1-4. Range. — Espiritu Santo and Aore Islands. Material— ML 33, ML 76, NH 2; Hog Harbour (MCZ 132314); holotypes of rossiteri Hartman (= hartmanni Kobelt) (ANSP 60067) and bicolor Hartman (CM 62.4683), Segond Channel, Espiritu Santo (CM 62.2885, ex Hartman, Layard) ; New Hebrides (MCZ 102017, ex Bequaert). Remarks. — Specimens which are intergrades between hartmanni and bicolor are illustrated (pi. 22). The greater obesity (lower spire) of bicolor is probably correlated with altitude and is not taxonomi- cally significant. P. francoisi Mabille (pi. 12, figs. 8, 9) and P. hebri- darum Mabille (pi. 12, figs. 6, 7) are within the range of variation shown by the Kuntz material and are here synonymized. P. bicolor has a dimorphic color pattern, typified by the "vari- eties" frangoisi and hebridarum. In the former, longitudinal streaks predominate, in the latter, spiral bands. The epidermal color pat- tern is formed by the deposition of waste products as pigments (Com- fort, 1951). Spurts of shell growth might conceivably result in a streaked pattern, while slow continuous growth would emphasize the spiral pattern. Underneath the epidermal color pattern is a ground color which consists of a saffron spire fading to white on the body whorl (see Pilsbry, 1900c). The type localities cited by Hartman are questionable. An orig- inal label with bicolor reads "Santo Is., near Aure (=Aore) Island" SOLEM: MOLLUSCA OF THE NEW HEBRIDES 137 and a partly decipherable note with the holotype of rossiteri indicates that it was collected on "Segon or Saigon on Santo Is. near Aura." Probably Segond Channel on Espiritu Santo is the type locality. All available specimens of typical bicolor are worn and without epidermis. If they were collected on Espiritu Santo, they undoubtedly were found near the coast and probably near a river or stream.' It would then be quite possible that the shells had been washed down from the interior of the island. This would not mean that bicolor and hart- manni are distinct, since altitudinally correlated variation is well known in land mollusks (see A. P. Brown, 1911; Pelseneer, 1920; van der Schalie, 1948; Solem, 1955) and is found in the New Hebridean Diplo- morpha (see pp. 140, 141, 143). Placostylus (Poecilocharis) turner! (Pfeiffer). Plate 8, figs. 1, 2. Bulimus turneri Pfeiffer, 1860, Proc. Zool. Soc. London, 1860: 138, pi. 51, fig. 10 — Erromanga (Turner). Placostylus (Poecilocharis) turneri Pfeiffer, Pilsbry, 1900, Man. Conch., (2), 13: 75, pi. 45, fig. 5. Range. — Erromanga. Material. — Photograph of holotype (BM). Remarks. — Placostylus turneri was listed for exchange by Cox (1868), but I could locate no specimens. It is closely allied to bi- color, but has an imperforate umbilicus and a larger aperture. The color pattern is similar to P. bicolor var. hebridarum. Genus DIPLOMORPHA Ancey, 1884 Shell small, elongate conic to ovate, umbilicate or rimate, solid, whorls 33^ to 5J^. Apical whorls IJ-^, with longitudinal striae or wrinkle-striate when not eroded. Cuticle thin, brown. Aperture ovate, peristome thickened or reflexed, usually stained with orange or red. (Adapted from Pilsbry, 1900c, p. 114.) Type species. — Diplomorpha layardi Ancey. Remarks. — Early authors such as Kobelt and Ancey considered that Diplomorpha represented an intermediate stage between Placo- stylus and Partula. The wide anatomical divergence between the latter two became known in the 1890's, but the affinities of Diplo- morpha remained uncertain until Kondo (1948) demonstrated its close relationship to Placostylus. In several respects Diplomorpha is more primitive. The long spermatheca inserted near the atrium ' Except for J. R. Baker (see p. 13), no person has collected shells from the interior of Espiritu Santo. All known collections were made near the sea coast or from the banks of a stream not more than four miles inland. 138 FIELD! ANA: ZOOLOGY, VOLUME 43 and the long slender penis are much nearer the corresponding struc- tures of the South American Bulimulidae than those of the specialized Placostylus of the Solomons and Fiji (see Kondo, 1948) . The apical sculpture of Diplomorpha, consisting of longitudinal striae in Quiros, new subgenus, and the wrinkle-striate type in Diplomorpha (sens, str.) are more primitive than the pitted sculpture of the New Caledonian, Fijian, and Solomon Island Placostylus. Apparently Diplomorpha is restricted to the New Hebrides. Sev- eral species, erroneously described from the Solomon Islands, have later been found in the New Hebrides. No Diplomorpha have been found on other island groups. Twelve names are available in Diplomorpha, of which perhaps four are entitled to specific rank. One of these, Diplomorpha hernieri, is here placed in a new subgenus, Quiros. The other three are the toothed D. layardi from Vate, D. coxi from Aneiteum(?), and the closely related D. peasei-delautouri-brazieri from Espiritu Santo and Omba Islands. The latter three names probably refer to variants of the same species. There is a prior generic name of nearly identical spelling, Diplo- morphus Giraud, 1871 (Ann. Soc. Ent. France, (5), 1: 409). Under present rules of zoological nomenclature, Ancey's genus is not affected. If future changes in the rules should invalidate Ancey's name, Hehri- daria Pilsbry (1900c, p. 115) has been made available as a replacement. Subgenus DIPLOMORPHA Shell ovate, whorls 33^ to 4^, spire-aperture ratio less than 0.85, surface without longitudinal folds and plications, suture never crenulate. Apical sculpture wrinkle striate, as in Santacharis. Type species. — Diplomorpha layardi Ancey. Diplomorpha (Diplomorpha) layardi Ancey. Plate 23, figs. 7-12. Diplomorpha layardi "Brazier" Ancey, 1884, II Naturaliste Siciliano, 3: 344 — generic description. Partula layardi "Brazier" Hartman, 1885, Proc. Acad. Nat. Sci. Philadelphia, 1885: 223 — observations, but not description, of the species. Partula {Diplomorpha) layardi "Brazier" Hartman, 1886, op. cit., p. 35, pi. 2, fig. 18 — figure, but no description. Diplomorpha layardi "Brazier," var. alticola Ancey, 1889, Le Naturaliste, 1889, p. 266 — ". . . une haute montagne, Vate" (Mt. MacDonald?). Partula layardi Brazier, 1890, Jour, of Conch., 6: 80— Tuker Tuker, Vate or Sandwich Island (no description or figure). SOLEM: MOLLUSCA OF THE NEW HEBRIDES 139 Partula layardi var. alba Brazier, 1890, op. cit., p. 80 — described without locality. Placostylus (Diplomorpha) layardi ("Brazier," Hartman), Pilsbry, 1900, Man. Conch., (2), 13: 115-116, pi. 72, figs. 1-5, 17— described and figured. Range. — Vate Island. Material.~Yi\a, Vate (CM 62.5291, J. Jetschin, Dec. 1904; Miller 548; UMMZ 146731, ex Walker, Tomlin; ANSP 133293, ex Frog- gatt); 2,000 feet up, Vate (UMMZ 183209, ex Walker, Ponsonby); Vate (MCZ 180554, ex Shackleford, Tomlin; ANSP 8338, ex Hart- man; USNM 603091, var. alba ex Higgins, ?Brazier; USNM 603092, ex Higgins, ?Brazier; ANSP 8336, ex Dupuy; MCZ 10209, ex Wink- ley; CM 62.4676, ex Hartman, Layard); "Salisboe Is." (USNM 56322, error by Hartman in locality) ; New Hebrides (UMMZ 14045; ANSP 8337, ex Garrett; MCZ 102018, ex Boston Soc. Nat. Hist., Roper; MCZ, ex Bequaert; MCZ 152750, ex Putzeys; MCZ 194761, ex Archer, Fulton). Remarks. — Brazier never published the name layardi, which must date from Ancey's description of Diplomorpha in 1884. Under Opin- ion 43 of the International Commission of Zoological Nomenclature, Ancey's usage is equivalent to the term "new genus, new species" of modem systematics. Since Ancey never considered that he described D. layardi, it is certain that no specimens labeled "types" are extant. The presence of named varieties, alticola Ancey and alba Brazier, necessitated selection of a neotype for the nominate variety. For this purpose, ANSP 8336 (pi. 23, fig. 10) is designated neotype. The shell has been previously figured by Pilsbry (1900c, pi. 72, fig. 2) and, although darker in coloration and less elongate than many specimens, is nevertheless representative of the original specimens sent to Layard by several collectors. Information on the ecology and type locality of D. I. layardi is contained in letters with the Hartman correspondence (CM). In a letter to Brazier, Layard indicated that D. layardi was not collected at "Tuker-Tuker (Ford's place), it came first from 'Rathmor' and then the fine large spms. from 'Seaview,' both properties of W. Glis- son who discovered it on Vate." The exact localities of Rathmor and Seaview are unknown to me, but probably they are near Havannah Harbour. The type locality is restricted to Rathmor, where it was first collected, and the site of the smaller specimens of which the neotype is representative. A letter from Layard to Garrett is here quoted at length : "As far as I know it [D. layardi] is only found on the elevated land at a distance from the coast. They were got at 140 FIELDIANA: ZOOLOGY, VOLUME 43 Seaview (his place) 3 miles from the coast at an elevation of about 1500 feet. It does not seem to move about in the day time, or in dry weather, but remains under dead leaves on which, I think, it feeds. It has not been seen to travel up bushes or trees or attach itself to their leaves. After rain, or when the ground is very moist, it moves about vigorously. As yet I but know it from Vate or Sandwich Is- land, and they don't seem to have a Bulimus [^ Placostylus] there, though they have another Partula . . . which is also found on SW of Aneiteum." Apparently this is a quotation from a letter sent to Layard by W. Glisson. Brazier (1890, p. 80) reported that D. layardi is found under sticks and leaves. Very few of the forty-seven available specimens had precise local- ity data. ANSP 133293, from near Vila, represented a population of large, thin, obese shells (pi. 23, figs. 11, 12) which are bright orange in color and have a very small parietal tooth. Miller 548 came from the banks of a small stream about two miles north of Vila. The speci- mens are smaller, slightly less obese, and darker in color, with a larger parietal tooth. In comparison, the neotype and Hartman's shells (pi. 23, figs. 7, 10) are very solid, somewhat stunted, and dark in color. UMMZ 146731 represented a mixed lot. One label read "Port Vila" (ex Tomlin); the other "2,000 feet up, Vate" (ex John Ponsonby) . The mixed lot can be divided into two series correspond- ing to typical layardi and variety alticola. The latter were catalogued as UMMZ 183209 with the locality "2,000 feet up, Vate." I did not see var. alha Brazier. The only specimen labeled "alba" (USNM 603091, ex Higgins) was an albino alticola. The type of alba is somewhere in England (Brazier, 1890, p. 66). Variety alticola Ancey (pi. 23, figs. 8, 9) has an elongate spire, and specimens are usually much lighter in color than typical D. layardi. The type locality of alticola is here restricted to Mount MacDonald, Table VIII. — Size Variation in Diplomorpha layardi D. layardi D. layardi var. alticola Range Mean S.D. Range Mean Height 18.9-24.4 21.0 1.29 23.2-25.4 24.2 Diameter 11.9-16.5 14.3 1.06 14.5-16.0 1.63 H/D 1.33-1.59 1.48 0.05 1.48-1.75 1.63 Aperture 12.3-16.4 14.1 1.06 14.4-16.1 15.2 Spire 5.5-8.0 6.9 0.52 8.5-9.5 9.0 S/A 0.38-0.61 0.50 0.05 0.56-0.61 0.59 Whorls 3>^-4 3^ .... 4 4 SOLEM: MOLLUSCA OF THE NEW HEBRIDES 141 Vate Island. D. I. var. alticola would be considered a geographic subspecies by many, but the higher spire in mountain forms is such a common adaptation that nomenclatural recognition only serves to burden the literature. The variation in the columellar tooth (pi. 23, figs. 7-12) is potentially of much more evolutionary significance than the variation in height of spire, and alticola, although a well-marked form, is not here accepted as a "subspecies." Several specimens of D. layardi collected by Miller have all of the internal partitions of the shell missing. Destruction of the apical whorls and spire by shrew and rodent predation is well known in continental areas, but I know of no predator on snails that will de- stroy only the internal partitions and not harm the spire. The variation in forty specimens of D. layardi and six of variety alticola is summarized in Table VIII. Diplomorpha (Diplomorpha) peasei (Cox). Plate 19, fig. 15. Partula peasei Cox, 1871, Proc. Zool. Soc. London, 1871: 644, pi. 52, fig. 2— Solomon Islands (Rainbird!) (error). Placostylus {Diplomorpha) peasei (Cox), Pilsbry, 1900, Man. Conch., (2), 13: 117, pi. 72, figs. 7-9. Range. — Aore Island (?). Material— "AolisL Is." (ANSP 144262). Table IX. — Size Variation in Diplomorpha D. delautouri Hog Harbour Sarakata River Mean Range S.D. Mean Range S.D. Height 22.4 20.9-24.6 1.08 24.4 20.6-26.9 1.15 Diameter 16.3 14.1-17.9 0.95 17.0 15.2-19.2 0.71 H/D 1.39 1.23-1.48 0.07 1.44 1.26-1.64 0.08 Spire 9.00 8.0-10.1 0.68 9.8 8.3-12.3 0.94 Aperture 13.3 12.2-14.7 0.72 14.4 12.3-15.7 0.72 S/A 0.69 0.57-0.78 0.05 0.69 0.54-0.84 0.05 Whorls A% AH-m 4^ 4-4^ D. peasei D. coxi D. brazieri Mean Range Height 21.8 20.0-23.7 24.5 19.0 Diameter 16.3 13.2-18.0 16.4 13.6 H/D 1.34 1.21-1.57 1.49 1.40 Spire 8.3 7.2-8.7 10.3 7.6 Aperture 13.5 11.5-15.1 13.2 11.4 S/A 0.61 0.53-0.74 0.78 0.67 Whorls 4M 4-4H ^H 4 142 FIELDIANA: ZOOLOGY, VOLUME 43 Remarks. — Some years after publishing his review of Diplomorpha, Pilsbry received eleven specimens of D. peasei from I. S. Oldroyd. The locality is "Aolia Island," which can be taken to be a misspelling of either Aoba (=Omba) or Aore Island. Several specimens of Par- tula albescens were received from the same locality so it is possible that Aore Island was intended. Pilsbry (1900c, p. 117) separated D. peasei from D. delautouri only by the more open umbilicus (pi. 19, fig. 15) and more expanded body whorl of peasei. It is doubtful whether field studies will maintain the separation of the three "spe- cies," delautouri, peasei, and brazieri, although a complete series of intergrades could not be found in the available material. A sum- mary of the morphometry of the lot of D. peasei (ANSP) is given in Table IX. Diplomorpha (Diplomorpha) brazieri Hartman. Plate 23, fig. 4. Diplomorpha brazieri Hartman, 1889, Proc. Acad. Nat. Sci. Philadelphia, 1889: 92-93, pi. 5, fig. 6— Aore Island; Ancey, 1905, Nautilus, 19, (4), p. 43 — Espiritu Santo. Placostylus (Diplomorpha) brazieri Hartman, Pilsbry, 1900, Man. Conch., (2), 13: 116-117, pi. 72, fig. 6. Range. — Espiritu Santo and Aore. Material— Aore Island (CM 62.4677 holotype). Remarks. — Hartman stated that all known specimens are worn and without epidermis, but I found only the holotype. The chief characteristics of D. brazieri are the narrow orange lip, small size, slender, flat whorls, and non-sinuate peristome. All of these char- acters vary greatly in Placostylus, but not enough is known about Diplomorpha to be positive that brazieri and delautouri are conspecific. Diplomorpha (Diplomiorpha) delautouri (Hartman) (emended). Plate 19, fig. 14; plate 23, figs. 1-6. Partula (Diplomorpha) delatouri Hartman, 1886, Proc. Acad. Nat. Sci. Phila- delphia, 1886: 35, pi. 2, fig. 19 — Aore Island, New Hebrides. Placostylus (Diplomorpha) delatouri Hartman, Pilsbry, 1900, Man. Conch., (2), 13: 117-118, pi. 72, figs. 10-12. Diplomorpha delautouri (sic) Hartman, Sykes, 1903, Proc. Malac. Soc. Lon- don, 5: 197 — Renee River, Espiritu Santo. Diplomorpha delautouri (sic) var. major Ancey, 1905, Nautilus, 19, (4), pp. 43- 44 — Espiritu Santo. Range. — Aore, Omba, and Espiritu Santo. Material— ML 10, ML 31f, ML 31g, ML 31h, ML 32, ML 33, ML 37, ML 37, ML 39, ML 42, ML 76, ML 90, NH 2; Aore Island (CM 62.4680, ex Hartman, lectotype); Omba (AMNH, Browne!); SOLEM: MOLLUSCA OF THE NEW HEBRIDES 143 Espiritu Santo (CM 62.4679, ex Hartman); Hog Harbour, Espiritu Santo (MCZ 109445); Segond Channel, Espiritu Santo (UMMZ 146792, ex Walker, Ponsonby, Layard); New Hebrides (MCZ 102019, ex Bequaert; MCZ 10208, ex Winkley). Remarks. — The collector of this species, George de Lautour, was a middle-aged planter later killed by natives and not "an enthusiastic young scientist" (see Rannie, 1912, pp. 151-152). Hartman's orig- inal spelling, delatouri, is invalid. Under the present rules of zoolog- ical nomenclature (Hemming, 1953, p. 45, paragraph 73) the name, delautouri, as a "valid emendation" must date from this author and paper. Such a course is ridiculous, since the species was described by Hartman and the addition of a single letter to the name does not change the species in any way. The species is quoted hereafter as Diplomorpha delautouri Hartman. The type locality is herein restricted to the portion of Aore Island facing Espiritu Santo (see pi. 2), and CM 62.4680 is selected as the type specimen. Specimens from Aore Island are smaller (19.3-19.7 mm. high) than those from Espiritu Santo (20.6-26.9), and Ancey (1905, p. 44) named the Santo shells var. major, based on material in the Paris Museum collected by Francois (see Mabille, 1895, p. 411). Dwarfed shells on small islands are common in land snails (Pelseneer, 1920, p. 560) and should not be nomenclaturally recognized. The variation in a lot of seventeen specimens from Hog Harbour (MCZ 109445), in a lot of sixty-five specimens from a road cut in the Sarakata River Valley (ML 33), and in D. peasei and D. hrazieri is summarized in Table IX. The differences between the Hog Harbour and Sarakata River specimens are of the same type as those found between D. layardi and D. layardi var. alticola. The material from ML 33 is a thanatocoenosis of shells deposited by flood waters. Dead land shells with a bubble of air trapped inside the spire are often picked up by flood waters and carried for considerable distances be- fore being deposited and covered by debris. Many of the shells in ML 33 were probably washed down from the mountains and it is quite possible that a high-spired form of delautouri inhabits the inte- rior of Espiritu Santo. The direction of the variational trend is oppo- site in Placostylus (higher spire in lowlands) and Diplomorpha (higher spire in uplands). Study of the ecology of the two genera may help to explain the mechanism controlling clinal variation in shell height, which is correlated with altitudinal changes. The variations found in height, peristomal swelling, and degree of lip expansion are illustrated (pi. 23). There was no significant 144 FIELDIANA: ZOOLOGY, VOLUME 43 color variation, all specimens having the streaked shell, orange aper- ture, and white reflected lip. Little is known of the ecology of D. delautouri. The appearance is that of a terrestrial species, and most of Kuntz's specimens were found under decaying logs, in leaf mould, or under tufts of grass. A few specimens, however, were found on the trunks and limbs of trees from one to seven feet above the ground. Young shells were found in January and early February, and nearly adult specimens in late February and early March. Only adult specimens were collected in October and June. Drowne picked up both adult and very young specimens on Omba in June, 1927. Kondo (1948, pp. 119-122, pi. 8) studied the anatomy of a single adult specimen. Three juveniles preserved by Kuntz were too young for study, but I dissected several adults obtained by Drowne and they completely confirmed the observations made by Kondo. The anatomy has not been refigured. Diplomorpha (Diplomorpha) coxi (Pease). Plate 16, fig. 1. Bulimus {?Borus) coxi Pease, 1871, Amer. Jour. Conch., 7: 197 — Solomon Islands (Rainbird) (error). Diplomorpha coxi Hartman, 1891, Proc. Linn. Soc. New South Wales, (2), 6: 571, pi. 21, figs. 1, 3, 6— Aneiteum(?). Placostylus (?) coxi Pease, Pilsbry, 1900, Man. Conch., (2), 13: 90. Placostylus (Diplomorpha) coxiana Pilsbry, 1900, Man. Conch., (2), 20: 118, pi. 72, figs. 13-14 — new name for P. coxi Hartman (not Pease). Diplomorpha coxi Pease, Clench, 1932, Nautilus, 46, (2), pp. 68-69, pi. 2, fig. 8. Range. — Aneiteum(?) . Material. — Solomon Islands (MCZ 86495, holotype of B. coxi Pease) . Remarks.~The holotype of D. coxi (Pease) (MCZ 86495) had been badly broken and then repaired at several different times dur- ing the life of the animal. The apical whorls are cracked and have several chips missing, the spire shows signs of at least two minor in- juries, the body whorl is badly broken near the columellar wall, and there is another bad break behind the lip. Thus much of the appear- ance of the holotype can be attributed to former injuries. The spire of coxi is wider than in D. delautouri, but otherwise the species are quite similar. The relationship to D. coxi Hartman is quite un- certain (see Clench, 1932). The illustrations of that species do not allow a close comparison. The spire of Hartman's species is much more produced and narrower than that of Cox's species. The rela- SOLEM: MOLLUSCA OF THE NEW HEBRIDES 145 tionship of the two can only be determined by re-examination of the type of coxi Hartman in the Australian Museum. Until then, noth- ing can be gained by separating the "species." In a letter to Hart- man, written in Sydney, Layard indicated that he doubted that Cox's specimen, the holotype of coxi Hartman, came from Aneiteum. Both forms may prove to be long-spired forms of delautouri. OUIROS, new subgenus Shell elongate, whorls 4K to 5%, spire-aperture ratio 0.87 or greater, surface with longitudinal folds or plications, sutures crenulate in many specimens. Apical sculpture of widely spaced, parallel longitudinal striae. Anatomy as in Diplo- morpha (sens. str.). Type species. — Bulimus hernieri Hartman. Remarks. — The apical sculpture, elongate form, greater number of whorls and surface sculpture separate Quiros from Diplomorpha (sens. str.). The apertural shape and color, umbilicus, genital anat- omy, and columellar structure are quite similar to Diplomorpha (sens, str.) and suggest that Quiros be considered a subgenus rather than a genus. Diplomorpha (Quiros) bernieri (Hartman). Plate 21, figs. 7, 8. Bulimus ruga Hartman, 1890, Proc. Acad. Nat. Sci. Philadelphia, 1890: 284, pi. 3, fig. 1 — Segou Island (=Segond Channel, Espiritu Santo). Bulimus bernieri Hartman, 1890, Proc. Acad. Nat. Sci. Philadelphia, 1890: 284, pi. 3, fig. 2 — Segou Island (=Segond Channel, Espiritu Santo). DipUrmorpha ruga (Hartman), Ancey, 1897, Nautilus, 11, (3), pp. 26-27. Diplomorpha bernieri (Hartman), Ancey, 1897, Nautilus, 11, (3), pp. 26-27. Placostylus {Diplomorpha) ruga (Hartman), Pilsbry, 1900, Man. Conch., (2), 13: 119, pi. 72, fig. 15. Placostylus {Diplomorpha) bernieri (Hartman), Pilsbry, 1900, Man. Conch., (2), 13: 119-120, pi. 72, fig. 16. Placostylus (?) bernieri (Hartman), Sykes, 1903, Proc. Malac. Soc. London, 5: 197 — Renee River, Espiritu Santo. Range. — Espiritu Santo. Material— ML 11, ML 33, ML 42, ML 46, ML 76; Segond Chan- nel (ANSP 132670). Remarks. — Hartman's type specimens could not be located. They are not in the Carnegie Museum, and, since Ancey, Kobelt, Sykes, and Haas have examined specimens, the types may be in a European museum. For a long time it was uncertain whether D. hernieri was a bulimulid or a partulid, but the roseate aperture and embryonic 5 :^ I ^07 W rn' t- 3 2 00 a s t-: W "S, j X w Eh J§ Ph w 146 SOLEM: MOLLUSCA OF THE NEW HEBRIDES 147 sculpture (see Haas, 1935, p. 190, fig. 1) are unknown in the Partuli- dae but are common to both Placostylus and Diplomorpha. On the basis of sutural sculpture (pi. 21, figs. 7, 8), two species were originally recognized. Both types of sculpture are often found on the same specimen; thus only one species can be recognized. The name ruga has page and line priority, but the name hernieri is here selected, as it represents the common variety, and ruga, with its "ruffled" suture, is a rather rare variant. Nothing is known of the ecology of D. hernieri. The specimens in lot ML 11 were found on the decaying trunk of a coconut palm, but Commander Kuntz's field notes fail to indicate if they were alive or dead. ML 33 contained many dead shells and the morphometry of 62 adults is summarized in Table X. Family PARYPHANTIDAE' (= Rhy tididae) Carnivorous agnathomorphous sigmurethra. Shell depressed-globose, planu- late, or auriform, umbilicated. Paryphanta and Schizoglossa have shells with few calcareous elements. Color usually brown or green, shell more or less prominently variegated with reddish-brown. Periphery rounded or keeled. Sculpture of weak spiral lines and/or radial ribs, or smooth. Peristome simple, not reflected or thick- ened. Animal without peripodial grooves or caudal mucous pore. Mantle not re- flected over shell. Jaw absent. Radula with less than 100 rows of teeth; central tooth absent in many species; lateral teeth unicuspid and large. Reproductive system without talon, epiphallic fiagellum, dart-sac, or glands. Duct of sper- matheca unbranched. Penis with papillae or large stimulatory organ. The Paryphantidae are found in South Africa, the Seychelles, Indonesia, Melanesia, the Caroline Islands, Samoa, Tonga, New Zea- land, and eastern Australia (fig. 18). The only monographic study is by Moellendorff (1903-1904). No general account has appeared since then but the papers of Watson (1934), Connolly (1939), Kondo (1943), and Powell (1930-49) contain much valuable information. The South African species (see Watson, 1934, and Connolly, 1939) are placed in two genera, Natalina and Nata. Natalina appears to be an analogue of the Austro-Zelandic Rhytida and was considered a subgenus of Rhytida by Thiele (1931, p. 726). Very probably, Nata- lina and Rhytida are only convergent (see Watson, 1934, p. 153), and they are here maintained as distinct genera. The shell of Nata is very similar to some Australian Delos, but until the anatomy of the latter genus has been fully studied, the exact nature of their relation- ' Paryphantinae Godwin-Austen, 1893, antedates Rhytididae Pilsbry, 1895 (see H. B. Baker, 1956a, b), and since both have been used with equal frequency, the law of priority has been followed here. 148 FIELDIANA: ZOOLOGY, VOLUME 43 ship will remain obscure. The South African paryphantids are quite restricted in range, none having been found north of 25° S. Lat. Genera from tropical Africa formerly included in the Paryphantidae are now known to belong elsewhere (Watson, 1934, p. 152). The relationships of the Pacific paryphantids are very poorly understood. Thiele (1931, pp. 724-726) utilized nine genera and subgenera for all the Pacific species, while Iredale (1933, 1938) cre- ated eleven new genera for the Australian species alone. In arranging the paryphantids for zoogeographic considerations, I have demoted Iredale's genera to sections or synonyms of previously described taxa. This reclassification indicates the broad lines within the family in regard to shell structure. The anatomy of the paryphantids is too incompletely known to be used in creating a phylogenetic classifica- tion, although previous studies (see Kondo, 1943, for references) have shown that numerous characters of value in taxonomy exist in the anatomy. A clearer concept of distribution and conchological relationships can probably be obtained by retention of broad taxo- nomic categories than by a great proliferation of generic names. For this reason, the Indo-Pacific Paryphantidae have been grouped as follows: GROUP A. Genus Paryphanta Albers, 1850 (type species, Helix busbyi Gray). Section Powelliphanta O'Connor, 1945. Section Victaphanta Iredale, 1933. Distribution: New Zealand, Victoria. Genus Wainuia Powell, 1930 (type species, Helix urnula Pfeiflfer). Melavitrina Iredale, 1933, is doubtfully separable. Distribution: New Zealand, Tasmania. Genus Schizoglossa Hedley, 1892 (type species, Daudebardia novoseeland- ica Pfr.). Distribution: New Zealand. GROUP B. Genus Rhytida Albers, 1860 (type species. Helix greenwoodi Gray), Distribution: New Zealand, Tasmania, Australia, New Caledonia. Genus Delos Hutton, 1904 (type species, Zonites coresia Gray). Distribution: Rotuma, New Hebrides, New Caledonia, New Zealand, Australia, Caroline Islands. Genus Macrocycloides Martens, 1869 (type species, Helix arthurii Pfeififer). Distribution: Buru, Borneo, Ambon, Saparua, Ceram, Solomon Islands, New Hebrides. SOLEM: MOLLUSCA OF THE NEW HEBRIDES 149 Genus Ouagapta Crosse, 1894 (type species, Helix raynali Gassies). Distribution: New Caledonia, Solomon Islands, New Guinea, Bismarcks, New Hebrides, Tonga, Samoa, Fijis, GROUP C. Genus Diplomphalus Crosse and Fischer, 1873 (type species, Helix cabriti Gassies). Distribution: New Caledonia. The use of groups "A," "B," and "C" show the major types of conchological variation and do not necessarily reflect phylogeny. The three genera of group "A" have a globose or auriform shell com- posed mainly of a thick layer of conchin with very few calcareous elements. The animals are too large to retreat into the shell, and the species are restricted to very humid situations. In New Zealand, Paryphanta are found in isolated colonies, usually at more than 2,000 feet elevation, and have evolved numerous local forms (Powell, 1930, 1946a, b, 1949). Group "C" contains a few specialized New Cale- donian species. Their multi-whorled, flat shell cannot be mistaken for any other paryphantid. The genera in group "B" may be prim- itive and "A" and "C" specialized offshoots. Several genera which are sometimes placed in the Paryphantidae need further discussion. Occirhenea Iredale, 1933, was proposed for Helix georgiana Quoy and Gaimard, 1832, a species from King George Sound, Southwest Australia. Later, Iredale (1939, p. 73) created a new family, the Occirheneidae, for this species. The original illus- trations show a helicid type of shell, but determination of its system- atic position must await study of the shell and its anatomy. The New Guinean "paryphantid" genera Gallodema and Illonesta Ire- dale (1941, p. 94) are endodontids synonymous with Paryphantopsis (Thiele, 1928a, pp. 125-126) (see Solem, in press-A). In his masterly essay on the carnivorous pulmonates, Hugh Wat- son commented on the past history of the Paryphantidae. The little information appearing since 1915 does not affect his conclusions and the following account has been condensed, with minor alterations, from Watson (1915, pp. 231-256). As to both distribution and anatomy, the Paryphantidae seem to be the oldest carnivorous family.' After an early Mesozoic or late Paleozoic origin, the paryphantids radiated, eventually reaching South Africa, Australia, Melanesia, and New Zealand. It is not • The carnivorous Rathouisiidae belong to a more primitive stock of the Pul- monata, but their distribution suggests that they evolved at a later date than the paryphantids. While they are more primitive, they are also more recent. 150 FIELDIANA: ZOOLOGY, VOLUME 43 known that they ever inhabited South America, but a few genera such as Guestieria and Macrocyclis could conceivably be aberrant paryphantids. The present distribution of the Paryphantidae can best be understood in relation to the evolution and distribution of some other carnivorous land mollusks. The Streptaxidae are the common carnivorous mollusks of Africa, South America, and the mainland of Asia. It is quite probable that they evolved from primitive paryphantids. Within the Streptaxidae there have been three lines of evolution: the helicoid, the pupaeform, and the cylindrelloid types of shell. In a few areas, for example, East Africa and the Mascarene Islands, large pupaeform streptaxids have developed, but generally they are small. The pupaeform and cylindrelloid streptaxids are adapted to the pursuit and capture of prey different from that of the predominately helicoid paryphantids and thus probably do not compete with the latter. Pupaeform strep- taxids and helicoid paryphantids are both common in South Africa. In areas where the helicoid streptaxids are present, however, as in central and northern Africa, Central America, South America, and the mainland of Asia, the paryphantids are absent. This may indi- cate that the helicoid streptaxids have replaced the paryphantids whenever the two groups have come into competition. It is not known if there is a barrier preventing the spread of the helicoid strep- taxids into South Africa, or whether the group is slowly extending its range and will eventually "eliminate" the paryphantids and occupy the entire African continent. Helicoid streptaxids are found in southeastern Asia, but they are all small and do not equal the size of many African and American species. Possibly the carnivorous rathouisiid slugs of southeastern Asia utilize the food sources that the large helicoid streptaxids prey upon in other areas. A radiation of large helicoid streptaxids from another region could possibly compete successfully, but it is difficult to understand how early stages in the evolution of large streptaxids from small species could compete with the slugs. Possibly of sig- nificance is the fact that in areas where both paryphantids and rathouisiids are found (Lombok to the Bismarcks), no large pary- phantids have been discovered. Northern Queensland may be an ex- ception, and the ecological position of Atopos australis (Heynemann) and the large paryphantids needs investigation. Except for the introduced Gulella (Huttonella) bicolor (Hutton), no streptaxids are known from the Indonesian area. Although com- mon on the mainland, they do not seem to have crossed from Malaya SOLEM: MOLLUSCA OF THE NEW HEBRIDES 151 to Java and there is a puzzling absence of small carnivorous snails in the area from Java to Lombok. The cause of this distributional gap is unknown. Pupaeform and helicoid strep taxids have been found in the Celebes and Borneo (van Benthem Jutting, 1954), but these probably represent a recent invasion from the Philippines where streptaxids are common. All the New Hebridean Paryphantidae belong to group "B." Examination of most of the named species resulted in my recogni- tion of four broad categories based on shell sculpture and type of whorl increment. These groups are here considered genera. The type species are easily separable, but many species found near the geo- graphical limits of distribution could be placed in more than one genus. Pending anatomical studies, it is probably better to maintain four large genera with several sections than to make each small spe- cies group a genus as Iredale did (1933, 1938). In the Malayan and Melanesian area, there are several conchological groups of species which have not been given generic names. One of these is so dis- tinct that I have created a new subgenus, Hebridelos, for it. A brief diagnosis of conchological characters is given for each of the genera and sections. All four genera of group "B" have been reported from the New Hebrides. The opportunity is taken to indi- cate the probable relationships of all the Pacific species and to clas- sify the New Hebridean species. Key to the New Hebridean Paryphantidae 1. Shell with definite surface sculpture 2 Shell smooth, without definite surface sculpture. Macrocycloides annatonensis (Pfeiffer) 2. Whorls rapidly increasing in width; body whorl at aperture as wide as spire. (Delos) 3 Whorls slowly increasing in width; body whorl at aperture not as wide as spire. . 5 3. Body whorl not sharply keeled 4 Body whorl sharply keeled Delos gassiest (Pfeiffer) 4. Umbilicus shallow, wide at apex; whorls not laterally compressed. Delos rapida (Pfeiffer) Umbilicus deep, narrow at apex; whorls laterally compressed. Delos haasi, new sp. 5. Umbilicus less than a fourth of the diameter in width 6 Umbilicus more than half of the diameter in width. Ouagapia santoensis, new sp. 6. Shell 10 mm. in diameter, with apertural tooth, weak transverse sculpture. Rhytida aulacospira (Pfeiffer) Shell more than 15 mm. in diameter, no apertural tooth, strong transverse sculpture Rhytida inaequalis (Pfeiffer) 152 FIELDIANA: ZOOLOGY, VOLUME 43 Genus DELOS Hutton, 1904 (=Elaea Hutton, 1883, not Ziegler, 1833, and Rhenea Hutton, 1893, not Saalmuller, 1884) Paryphantids with whorls rapidly increasing in size, spire flat, depressed or only slightly elevated. Color pattern of reddish streaks or blotches. Surface sculpture of weak to strong spiral striae, growth lines and longitudinal striae never prominent. Type species. — Zonites coresia Gray. Remarks. — On the basis of the several types of shell sculpture observed in Delos, three subgenera and several sections are here ten- tatively recognized : Subgenus Delos (sens. str.). Sculpture of very few short wavy spiral lines; much of the surface smooth. Section Delos (sens. str.). New Zealand species. Umbilicus wide, coloration light. Delouagapia Powell, 1952, is probably a synonym. Section Prolesophanta Iredale, 1933. Tasmanian species with a narrow umbilicus. Tasmadelos Iredale, 1933, is a synonym. Subgenus Saladelos Iredale, 1933. East Australian species with narrow or wide umbilicus. Subgenus Hebridelos, new subgenus. Species from the New Heb- rides, Rotuma, and the Caroline Islands with prominent anastomos- ing spiral striae and dark red zigzag markings. Subgenus HEBRIDELOS, new subgenus Shell discoidal, spire flat or depressed. Whorls 3^ to 4^, rapidly increasing in size. Umbilicus widely open, either shallow or deep, contained four to five times in the diameter. Body whorl rounded or keeled. Shell with thick epidermal layer. Color dark or light greenish-horn with red blotches or zigzag streaks. Sculpture of anastomosing spiral striae crossed by a few irregular growth lines. Aperture ovate to compressedly lunate, rarely deflected. Lip thin, simple. Diameter 5-12 mm., height 2-4 mm. Type species. — Helix rapida Pfeiffer. Remarks. — The color pattern and distribution (fig. 28) overlap Ouagapia, but the rapidly increasing whorls, narrow umbilicus, and spiral striae relate Hebridelos to Delos. Kondo (1943) investigated the anatomy of D. (H.) oualanensis (Pease) and two species of Oua- gapia. The former differed in having a true verge. The drawings of Murdoch (1900) are incomplete, but it is probable that the genital anatomy of Delos coresia (Gray) is very similar to that of D. oualanensis. The three New Hebridean species, D. rapida (Pfeiffer), D. gassiesi (Pfeiffer), and D. haasi, new sp.; D. gardineri Smith, 1897, from SOLEM: MOLLUSCA OF THE NEW HEBRIDES 153 Rotuma; and D. oualanensis (Pease) from the Caroline Islands are the only species in the subgenus Hebridelos. Previous accounts of D. rapida and D. gassiest have been unsatisfactory, and for conven- ience they, as well as D. haasi, new sp., are here described. Deles (Hebridelos) rapida (Pfeiflfer). Plate 11, figs. 1-3. Helix rapida Pfeiffer, 1853, Zeitschr. f. Malak., 1853: 54-55— New Zealand (error) (March); Pfeiflfer, 1853, Conch. Icon., Helix, pi. 158, fig. 1038— New Zealand (error) (May); Pfeiflfer, 1853, Proc. Zool. Soc. London, 1853: 58-59— New Zealand and Solomon Islands (error) (July 25); Cox, 1868, Monog. Austr. Land Shells, pp. 19-20, pi. 3, fig. 9, a, b — Cape York, Queensland and New Hebrides. Helix rapida var. beta Pfeiflfer, 1853, Zeitschr. f. Malak., 1853: 55 — Solomon Islands. ?Helix {Discus) rapida Pfeiflfer, Cox, 1868, Exchange List, p. 43, no. 92 — Erromanga. Elaea rapida (Pfeiflfer), Tryon, 1885, Man. Conch., (2), 1: 129, pi. 26, fig. 13— New Zealand, Solomon Islands (errors); Hedley, 1890, Proc. Roy. Soc. Queensland, 5, (2), p. 100 — New Hebrides. Rhenea rapida (Pfeiflfer), Sykes, 1903, Proc. Malac. Soc. London, 5: 196 — Port Vila, Vate (J. J. Walker). Range. — Vate, Erromanga (?). Material— ^e-w Hebrides (AMNH 55455, AMNH 65958, Marsh 6648, UMMZ 135627, ex Walker, Ponsonby); Vate Island (UMMZ 135626, ex Walker, Eaton; CM 62.5806, ex Hartman, Layard; MCZ, ex Putzeys); Vila, Vate (Miller 593 and USNM 598359, ex Miller); unknown locality (UMMZ 127597, ex Walker, Ponsonby). Photo- graph of syntype (BM). Shell small, subdiscoidal, spire flat or slightly depressed. Whorls 3^ to 3% (average S'Vu). rapidly increasing in size. Umbilicus shallow, wide at apex, con- tained about A\i times in diameter. Body whorl obtusely rounded, not keeled. Shell with a thin calcareous and thick epidermal layer. Color greenish-horn with numerous reddish blotches. Occasionally blotches form zigzag stripes on lower surface. Surface sculpture of anastomosing spiral striae crossed by growth lines, stronger above, weaker below periphery. Below the epidermal layer, a sculpture of minute, transverse, anastomosing striae. Aperture ovate, rarely deflected. Lip thin, simple. Diameter 7.5-8.7 mm. (average 8.1); height 2.6-3.5 mm. (average 3.2). Remarks. — Photographs of a sjmtype (pi. 11, figs. 1-3) enabled satisfactory identification of Helix rapida Pfeiffer. Moellendorff (1903, p. 18) was unable to identify the species and Hedley and Suter had previously stated that rapida did not live in either Australia or New Zealand. The syntype is identical with specimens from near Vila, Vate, and the type locality is restricted to the southern New Heb- 154 FIELDIANA: ZOOLOGY, VOLUME 43 rides. Further restriction is impossible until more specimens of Delos (Hehridelos) haasi, new sp., are known. The possibility exists that the Erromangan specimens of Delos reported by Cox (1868) belong to D. (H.) haasi. If this is true the type locality of D. (H.) rapida could be restricted to Vate; if D. (H.) rapida is found on Erromanga as well as Vate, then Erromanga should be selected as type locality. There is some variation in color pattern. Many specimens have the reddish blotches scattered, as in the syntype; others have the blotches organized into transverse stripes. Kondo (1943, p. 245) mentioned a dimorphic condition in D. oualanensis, where both a lightly sculptured and colored form and a heavily sculptured and darkly colored form were found together. The presence of both lightly colored and darkly colored specimens of D. (H.) rapida may indicate that a similar dimorphism exists in that species. Delos (H.) rapida has a wider, shallower umbilicus and fewer whorls than D. (H.) haasi, and it lacks the sharp keel which is characteristic of D. (H.) gassiesi. Delos (Hebridelos) gassiesi (Pfeiffer). Plate 11, figs. 4, 6. Helix gassiesi Pfeiffer, 1861, Proc. Zool. Soc. London, 1861: 21 — Erromanga (Turner); Pfeiffer, 1861, Malak. Blatt., 7: 234— Erromanga (Turner). Helix (Videna) gassiesi Pfeiffer, Cox, 1868, Exchange List, p. 40, no. 45. Trochomorpha (Videna) gassiesi (Pfeiffer), Clessin, 1881, Nomen. Helic. viv., p. 85, no. 1302 — Erromanga. Range. — Erromanga. Material. — Erromanga (USNM, ex Calvert, Chamberlain) ; New Hebrides (AMNH 59127, ex Haines); photograph of holotype (BM). Shell small, subdiscoidal, spire distinctly depressed below the body whorl. Whorls 4-43/2. rapidly increasing in size. Umbilicus shallow, contained four times in diameter. Body whorl keeled, periphery at center of whorl. Keel slightly ele- vated, carina-like. Color pale horn, with transverse streaks of red. Sculpture of anastomosing spiral striae, stronger above the keel, weaker below. Subepidermal sculpture as in D. (H.) rapida. Aperture heart-shaped, laterally compressed. Lip thin, simple. Diameter 7.5-10.1 mm., height 2.7-3.8 mm. Remarks. — The shallow umbilicus and size ally Delos (Hehridelos) gassiesi to D. (H.) rapida. The prominent keel probably misled Cox and Clessin, who placed D. gassiesi in the zonitid genus Trocho- morpha. The specimens from the American Museum had the color pattern faded, but otherwise corresponded well to the type photo- graphs. The single example from the U. S. National Museum was in a lot of Gonatoraphe fornicata (Pfeiffer). It is 10.1 mm. in diameter and 3.8 mm. high, and has 43^ whorls. SOLEM: MOLLUSCA OF THE NEW HEBRIDES 155 Delos (Hebridelos) haasi, new species. Plate 16, figs. 5-7. A species of the subgenus Hebridelos distinguished from D. (//.) rapida (Pfeiffer) by its deeper, narrower umbilicus, weaker spiral sculpture, and laterally compressed whorls; from D. (H.) gassiesi (F*feifTer) it is separated by its deeper, narrower umbilicus, weaker sculpture, and rounded (not keeled) body whorl. Shell small, subdiscoidal, spire flat, not depressed. Whorls 3% (paratypes 4 J^ to 45^, average 45^), rapidly increasing in size. Umbilicus deep, very narrow at apex, contained about five times in the diameter. Body whorl rounded, slightly compressed laterally. Color greenish-horn with transverse zigzag red markings. Surface sculpture of a few weak anastomosing spiral striae. Subepidermal sculp- ture as in D. (//.) rapida. Aperture ovate, compressed laterally. Lip thin, simple. Diameter 7.5 mm. (paratypes 9.3-10.1 mm., average 9.6 mm.), height 3.0 mm. (paratypes 3.8-4.4 mm., average 4.1 mm.). Type. — Dominion Museum, Wellington, New Zealand, no. M.F, 11263. Collected at 200 feet in forest on the island of Aneiteum, New Hebrides, by W. H. Dawbin, on October 22, 1954. Paratypes.— VMMZ 127596 and CNHM 72440 with the locality "New Hebrides" from the Bryant Walker, John Ponsonby, and An- drew Garrett collections. Remarks. — The Aneiteum holotype is much smaller than the para- types, but the differences probably only reflect age or local environ- mental conditions. The specimens that Cox (1868) listed as D. rapida may actually have been D. haasi. Great pleasure is taken in dedicating this species to my friend and mentor, Dr. Fritz Haas of Chicago Natural History Museum. Genus MACROCYCLOIDES von Martens, 1869 Paryphantids with whorls gradually increasing in size, rounded, and with im- pressed sutures. Spire slightly elevated. Surface smooth with a few wide, low, growth lines, or widely spaced striae as in Retinella. Never with prominent spiral sculpture. Color hyaline or light horn with only traces of reddish flammulations. Aperture oblique, but not deflected. Type species. — Helix arthurii Pfeiffer, Remarks. — Species referable to Macrocydoides are found in Borneo {arthurii Pfeiffer), Amboina {microcystis Boettger), Saparua {saparu- ana Boettger), Haruku {sericina Boettger), Ceram {quadrispira von Martens), New Guinea {kapaurensis Smith), and the New Hebrides {annatonensis Pfeiffer) (fig. 27). Other species with flat spires, from Buru {lutea von Martens, euglypta von Martens) and the Solomon Islands {veronica Pfeiffer), do not seem separable. Zonites hameliamis Crosse and Helix subnitens Gassies, which Franc (1957, pp. 174-175) 156 FIELDIANA: ZOOLOGY, VOLUME 43 placed in Macrocycloides are much smaller than any of the other species and are brown in color. It is very probable that they are ariophantids, not paryphantids. The "finement stride spiralement" of subnitens definitely removes it from the Paryphantidae, but with- out examination of specimens no suggestions can be made as to its systematic position. The general appearance of the listed species is the same, but Macrocycloides may be polyphyletic. No species have been dissected and very few specimens are known. Without anatomical studies and abundant conchological material, recognition of more supraspecific categories is unwarranted. Only a single species, M. annatonensis (Pfeiffer), is known from the New Hebrides. Macrocycloides annatonensis (Pfeiffer). Plate 10, figs. 15-17. Helix annatonensis Pfeiffer, 1854, Conch. Icon., Helix, pi. 200, fig. 1409 — Aneiteum (MacGillivray) ; Pfeiffer, 1855, Proc. Zool. Soc. London, 1854: 288; Kobelt, 1897, Conch. Cab., I, 12, (4), pp. 552-553, pi. 166, figs. 14-15. Helix (Hyalinia) annatonensis Pfeiffer, Cox, 1868, Exchange List, p. 38, no. 4 — Aneiteum. Hyalinia (Polita) annatonensis (Pfeiffer), Tryon, 1886, Man. Conch., (2), 2: 169, pi. 52, fig. 100. Rhytida (Macrocycloides) annatonensis (Pfeiffer), Moellendorff, 1903, Conch. Cab., I, 12, (B), p. 48, pi. 9, fig. 7. Range. — Aneiteum. Material— Aneiteum (CNHM 37019, ex Webb, Gude, Stevens); photographs of syntype (BM). Remarks. — The size (9-10 mm. in diameter), pale coloration, very deep and wide umbilicus, and lack of spiral sculpture suffice to sep- arate this species from the other New Hebridean paryphantids. The shell of M. annatonensis is superficially similar to some Asiatic and Malayan Helicarionidae, but the smooth surface and "greasy-white" color is characteristic of the carnivorous taxa. The only New He- bridean shell which is at all similar is the minute Orpiella retardata (Cox), which is 5 mm. in diameter and has a much smaller umbilicus. The single specimen of M. annatonensis that I saw, CNHM 37019, is 9.3 mm. in diameter, 4.6 mm. high, has 5 whorls, and the umbilicus is contained 3.1 times in the diameter. r> Genus RHYTIDA Albers, 1860 ^ Medium to large paryphantids with regularly coiled whorls, impressed sutures, and a deep, widely open umbilicus. Body whorl rounded or keeled. Sculpture of SOLEM: MOLLUSCA OF THE NEW HEBRIDES 157 prominent transverse ribs, reduced or absent in a few species. Spiral striae are prominent in some species, absent in others. Spire flat or elevated. Aperture large, lunate, not as oblique as in Macrocy chides . Type species. — Helix greenwoodi Gray. Remarks. — The species included in Rhytida can be divided into several conchological sections. Apparently there is a north-south gradation in sculpture which may prove of evolutionary importance when the genus has been thoroughly studied. New Zealand and Tasmanian species have very prominent transverse sculpture which usually is continued on the base of the shell; Australian and New Caledonian species have the sculpture weakened or absent on the base; and the species from the northern parts of Queensland have almost completely lost the transverse sculpture. On the basis of sculpture and distribution it is possible to recognize two subgenera and several sections. The following classification is suggested : Subgenus Rhytida (sens, str.) . Transverse ribs strongly developed on spire and body whorl or secondarily lost. Section Rhytida (sens. str.). New Zealand species with transverse ribs on spire, sculpture becoming rugose to pitted on body whorl and base. Section Tasmaphena Iredale, 1933. Tasmanian shells with trans- verse ribs on spire, body whorl and base. Section Strangesta Iredale, 1933. East Australian species with transverse ribs on spire and body whorl above periphery. Base and body whorl below periphery smooth or with a few weak spiral striae. Probably Montidelos Iredale (1943a, p. 68) is synonymous. Section Murphitella Iredale, 1933. Species from northern Queens- land in which the transverse sculpture is lost and a very fine series of spiral lines has been developed. Namoitena Iredale, 1933, is a synonym. Subgenus Ptychorhytida Moellendorff , 1903. A sculpture of widely spaced spiral lines has been developed in addition to the transverse ribs. Section Ptychorhytida (sens. str.). New Caledonian shells with the transverse ribs well developed. Section Echotrida Iredale, 1933. Australian species with the trans- verse sculpture greatly reduced. Two species belonging to the subgenus Ptychorhytida, aulacospira Pfeiffer and inaequalis Pfeiffer, have been reported from the New Hebrides. Both are common in New Caledonia and probably the New Hebridean records are erroneous, since they originated from the Cuming collection. The two species are listed below but are not here accepted as being part of the New Hebridean fauna. 158 FIELDIANA: ZOOLOGY, VOLUME 43 Rhytida (Ptychorhytida) aulacospira (Pfeiffer). Plate 10, figs. 10-12. Helix aulacospira Pfeiffer, 1846, Proc. Zool. Soc. London, 1846: 37 — no local- ity; Pfeiffer, 1853, Conch. Icon., Helix, pi. 150, fig. 975— New Hebrides (Cuming collection) ; Pfeiffer, 1854, Conch. Cab., I, 12, (3), p. 491, pi. 160, figs. 24-26. Helix multisulcata Gassies, 1857, Jour, de Conchy., 6: 272-273, pi. 9, fig. 3 — New Caledonia; Gassies, 1863, Faune Conchy. Nouv. Caledonie, I, p. 118, pi. 1, figs. 5-6. Rhytida multisulcata (Gassies), Tryon, 1885, Man. Conch., (2), 1: 118, pi. 23, figs. 40-42. Patula (Punctum) aulacospira (Pfeiffer), Tryon, 1887, Man. Conch., (2), 3: 35, pi. 7, fig. 86. Rhytida (Ptychorhytida) multisulcata (Gassies), Moellendorff, 1903, Conch. Cab., I, 12, (B), pp. 71-72, pi. 12, figs. 6-8. Rhytida (Ouagapia) aulacospira (Pfeiffer), Moellendorff, 1904, Conch. Cab., I, 12, (B), pp. 81-82, pi. 13, figs. 6-8. Ouagapia (Ptychorhytida) aulacospira (Pfeiffer), Franc, 1957, Mem. Mus. Nat. Hist. Nat., n.s., Zool., 13: 173, pi. 22, fig. 230. Range. — New Caledonia and Loyalty Islands. Material.— 't^ew Hebrides (ANSP 23685, ex Wilslach) ; New Cale- donia (several lots in UMMZ, ANSP, CNHM). Photograph of syntype (BM). Remarks. — Sykes (1895, p. 72) united aulacospira and multisul- cata after examining the types in the British Museum (Natural His- tory). R. aulacospira is a common New Caledonian shell and the New Hebridean record from the Cuming collection is probably in- correct. The ANSP lot was identified as "Delos rapida," a New Hebridean shell (see p. 153), and may have been mislabeled. Photo- graphs of the syntype of aulacospira are reproduced here (pi. 10, figs. 10-12) for convenient reference. Rhytida (Ptychorhytida) inaequalis (Pfeiffer) Helix inaequalis Pfeiffer, 1854, Conch. Icon., Helix, pi. 198, fig. 1394 — Isle of Pines (MacGillivray); Pfeiffer, 1854, Proc. Zool. Soc. London, 1854: 286; Gassies, 1863, Faune Conchy. Nouv. Caledonie, I, p. 32, pi. 1, fig. 17. Helix fischeri Gassies, 1857, Jour, de Conchy., 6: 271, pi. 9, figs. 1, 2. Rhytida (Ptychorhytida) inaequalis (Pfeiffer), Moellendorff, 1904, Conch. Cab., I, 12, (B), pp. 75-76, pi. 12, figs. 17-19. Ouagapia (Ouagapia) inaequalis (Pfeiffer), Franc, 1957, Mem. Mus. Nat. Hist. Nat., n.s., Zool., 13: 169, pi. 21, fig. 225. Range. — New Caledonia. SOLEM: MOLLUSCA OF THE NEW HEBRIDES 159 Material— New Hebrides (ANSP 1909, ex A. D. Brown; MCZ 10990, ex Anthony, Cuming; CM 62.14564, ex Hartman); New Cale- donia (UMMZ.11040, ex Steams; UMMZ 135598, ex Walker, Weth- erby; UMMZ 135599, ex Walker, Ponsonby, Garrett, Layard; UMMZ 135600, ex Walker, Ponsonby, Layard). Remarks. — Layard, in his letters to Hartman, did not mention R. inaequalis as a New Hebridean shell, although the Hartman col- lection contains many New Caledonian specimens received from Layard. The source of the New Hebridean locality probably was the Cuming collection, the localities of which are notoriously inac- curate. R. inaequalis is characterized by its size, strong spiral and transverse sculpture, and lack of an apertural tooth. Genus OUAGAPIA Crosse, 1894 Small to large paryphantids with a deep, wide umbilicus. Whorls increasing regularly in size, sutures impressed and channeled. Color horn with reddish macu- lations. Sculpture varies. Type species. — Helix raynali Gassies. Remarks. — There are three species groups included in Ouagapia. Diverse in size and sculpture, their common color pattern and island distribution (fig. 27) have resulted in their being united. Rather than add more names to the overburdened taxonomy, the species are here retained in the same genus. The genotype, 0. raynali, is very large, with a shell sculpture of prominent but very fine, close-set spiral lines. The same sculpture is found on the base of Rhytida (Murphitella) franklandensis (Forbes) (UMMZ 135582, ex Walker, Thompson, Brazier, MacGillivray from Frankland Island, Queensland). The color patterns of raynali and franklandensis are quite different, but since neither has been dissected their relationship is uncertain. 0. villandrei (Gassies) from the Solomon Islands (see Clapp, 1923) is very similar to raynali in size, color pattern, and size of umbilicus. 0. villandrei differs in having the base of the shell with only a few weak, irregular spiral striae and the upper portion with strong trans- verse ribs as in the third group. The third group consists of 0. gradata (Gould) from Samoa and Tonga (see Kondo, 1948); 0. ratusukuni (Cooke) from the Fijis; 0. opaoana (Gassies), 0. lamberti (Gassies), and 0. rufotincta (Gas- sies) from New Caledonia; 0. santoensis, new sp., from the New Hebrides; and, less certainly, 0. radicalis (Mousson) and 0. vicaria (Mousson) from Tonga. All have a deep, very wide umbilicus; trans- 160 FIELDIANA: ZOOLOGY, VOLUME 43 verse sculpture above the periphery; and spiral lines below the periph- ery of the body whorl. "Rhytida" hednalli Ponsonby, 1907, from New Guinea, "Endodonta" acuticarinata Thiele, 1928, from New Britain, and "Champa" delectans Smith, 1898, from New Guinea also seem to belong to this series of Ouagapia (see Solem, in press-B). Only 0. gradata and 0. ratusukuni have been dissected (Kondo, 1943). Until the large New Caledonian and Solomon Island Oua- gapia have been dissected, taxonomic recognition of the three groups outlined above would be unwise. Probably the genus Ouagapia, as now constituted, is polyphyletic. 0. santoensis, new species, was collected from ground detritus in the Sarakata River Valley on Espi- ritu Santo. Only dead shells were found. Ouagapia santoensis, new species. Plate 28, figs. 6-8. ?Nanina (Endodonta) acetabulum Mabille (not Pease), 1895, Bull. Soc. d'Hist. Nat. d'Autun, 8: 409 — New Hebrides (Francois). A species of Ouagapia related to 0. opaoana (Gassies) and 0. gra- data (Gould) but distinguished from them by having a wider, shal- lower umbilicus, the base of the shell more sharply keeled, the spire and body whorl narrower, and the coloration lighter. Shell small, flat, discoidal. Spire slightly elevated or flat, sutures impressed and channeled. Whorls 3^ to 4, gradually increasing in size. Umbilicus widely open, more than one half of the diameter in size. Base of shell distinctly keeled around the umbilicus. Sculpture of prominent, close-set, transverse ribs on the spire and above the periphery of the body whorl. Below the periphery and in the umbilical region a sculpture of a few growth lines and weak, widely spaced spiral striae. A few fine, close-set spiral striae on the spire. Color light horn, with a few irregular transverse reddish maculations. Aperture large, lunate, sharply angled below. Lip thin, simple. Diameter 5.5-5.8 mm., height 2.4-2.8 mm. Type. — University of Michigan Museum of Zoology no. 188435. Collected from stream drift in the Sarakata River Valley, Espiritu Santo, New Hebrides (ML 95), by Robert E. Kuntz in May and June, 1944. Paratypes. — On Espiritu Santo, specimens were collected from ML 33 and ML 95 (type locality). Paratopotypes are UMMZ 188436 and CNHM 73061. One paratype is UMMZ 188437. Remarks. — Mabille's record for the Polynesian endodontid Neso- discus acetabulum was probably based on 0. santoensis. In size, shape, and color the two species are very similar, despite their very dif- ferent taxonomic positions. Similar convergence probably caused L Rensch (1937) to place the Bismarck Ouagapia acuticarinata (Thiele) in Nesodiscus. SOLEM: MOLLUSCA OF THE NEW HEBRIDES 161 0. santoensis has a wider umbilicus than that of any other species except 0. ratusukuni Cooke (UMMZ 164673), which is larger and flatter and lacks the color flammulations of the New Hebridean shell. Superorder BASOMMATOPHORA Schmidt, 1855 Snails bearing eyes at base of a pair of contractile, but not invaginable ten- tacles. Animals hermaphroditic, with separate genital openings, the male more anterior. Shell external, variously modified. Respiration through lung, gills, and/or epithelium. Except for the tidal, estuarine, and terrestrial Ellobiidae, marine Patelliformia, and estuarine Amphibolidae, the Basommatophora are fresh- water inhabitants. Three families, the Siphonariidae, Ellobiidae, and Planorbidae, are found in the New Hebrides. The Siphonariidae belong to the littoral fauna and will be considered with the other marine taxa. The Ellobiidae contain tidal, supratidal, and estuarine species which tend to form local colonies showing a wide range of morphologic vari- ation. Similar situations complicate the study of the Thiaridae and Neritina (sens, lat.) and the Ellobiidae will be considered with the latter. Both the Siphonariidae and Ellobiidae belong to the more primitive "Archaeopulmonata" (Morton, 1955, p. 163), perhaps more often called, respectively, parts of the "Thalassophila" and "Acto- phila" (Hubendick, 1945, p. 164). The Planorbidae, together with the Lymnaeidae, Ancylidae, and Physidae, form the Hygrophila (Thiele, 1931), "higher limnic Basommatophora" (Hubendick, 1945), or "Branchiopulmonata" (Morton, 1955). Comprehensive surveys of the relationships and phylogeny of the Basommatophora can be found in Hubendick (1945), Graham (1949), and Morton (1955). Family PLANORBIDAE Shell basically sinistral, flat, discoidal, lens-shaped, ovate, or turreted, but pseudodextral in many taxa. Foot small, broad; tentacles slender. Most species with a secondary gill (pseudobranch) below pulmonary orifice on the left side. Radula with normally bicuspid central tooth, tricuspid laterals, and serrate mar- ginals (Mandahl-Barth, 1954, p. 72). Because certain planorbid snails are vectors of an important human parasitic disease, schistosomiasis or bilharzia, much attention has been focused on the Planorbidae in recent years. Most reports have been restricted to the few species of immediate medical importance, but the studies of Hubendick (1948, 1955) and Mandahl-Barth (1954) supplement the systematic conclusions presented in the uncompleted 162 FIELDIANA: ZOOLOGY, VOLUME 43 monograph by F. C. Baker (1945), which was published posthumously. The classification below follows Hubendick (1955). Two genera, Gyraulus and Physastra, are found in the New Heb- rides. Gyraulus is probably world-wide in distribution; Physastra is restricted to the Indo-Australian region. The single species of each genus is also found in New Caledonia. Subfamily PLANORBINAE Genus PHYSASTRA Tapparone-Canefri, 1883 Type species. — Physastra vestita Tapparone-Canefri. Remarks. — Prior to 1948, the Indo-Melanesian physoid snails were grouped with the African carriers of bilharzia as Bulinus (sens. str.). Hubendick (1948) demonstrated that there are two types of "Physa"- like planorbs. One, the Afro-Australian Bulinus, has the male copu- latory organ with a pseudopenis (Hubendick, 1948, p. 25, fig. 93). It is placed in a subfamily, the Bulininae. The Planorbinae contain species with a true penis and seem to be more primitive than the Bulininae. Three genera of physoid planorbinine shells are recognized ; Phy- sastra, in which the penis has a homy terminal stylet and a lateral penis pore; Amerianna, which has no stylet and a terminal penis pore; and Miratesta, which has a lateral pore but no stylet. The distribu- tion and probable synonyms of the genera are given by Hubendick (1948, pp. 57-59). Franc (1957, pp. 83-88) recognized twelve "species" of Physastra from New Caledonia. Material in the University of Michigan Museum of Zoology suggests that most of the "species" should be united into one taxon. Layard expressed a similar opinion in a letter to Hart- man (March, 1890), since he had often observed different "species" of Physastra in copulation. The earliest name available for the New Caledonian Physastra is nasuta Morelet, 1856. Union of the named forms into one species is premature, but probably will be eventually accomplished. Several Fijian specimens {sinuata Gould, 1852) in the material studied do not differ from the New Hebridean and New Caledonian shells. If the possible relationship of the Tongan popu- lation (tongana Quoy and Gaimard, 1832) is also considered, obvi- ously no decision on the specific status or names of the Melanesian Physastra can be reached at this time. Until the majority of the named entities can be critically examined, both as to anatomy and SOLEM: MOLLUSCA OF THE NEW HEBRIDES 163 shell structure, any name applied to the complex must be only pro- visional. Hubendick (1948, p. 16, figs. 50-53, 58, pi. 1, fig. 4) partially fig- ured the anatomy of Physastra doliolum (Gassies) from Kone, New Caledonia. I dissected several specimens from Espiritu Santo. No differences of any significance were found. Proportional size and details of the penial region were identical. Without any doubt the New Hebridean populations and New Caledonian doliolum belong to the same species. Unfortunately, no other Melanesian Physastra has been dissected. Little would be accomplished by uniting the New Hebridean pop- ulation to P. doliolum only, and it is premature to synonymize all of the New Caledonian species. For convenience, Ancey's name, Physa layardi, is used for the New Hebridean Physastra until such time as the remaining Melanesian Physastra can be studied. Physastra layardi (Ancey). Plate 25, figs. 10-14. Physa sinuata Mabille (not Gould, 1852), 1895, Bull. Soc. d'Hist. Nat. d'Autun 8: 409 — New Hebrides (Frangois). Physa layardi Ancey, 1905, Nautilus, 19, (4), p. 44 — Vate (Glisson). Isidora caledonica Boettger (not Morelet, 1856), 1916, Abhl. Senck. Naturf. Gesell., 36, (3), p. 296— Aniwa. Physa cf. obtusa J. R. Baker, 1929, Man and animals in the New Hebrides, p. 148 — Steaming Hill Lake, Gaua, Banks Group. Range. — Espiritu Santo, Gaua, Vate, Aniwa, Tanna, Aneiteum, Aore and Futuna. Material— ML 60, ML 61a, ML 61b, ML 62, ML 65, ML 68, ML 78, ML 80, ML 82a, ML 82b, ML 85a, ML 85b, ML 91, ML 93 Vate (CM, ex Hartman, Layard, type lot material; DMNZ, W. H Dawbin!, at 800 feet elevation) ; Aniwa (CNHM 35109, ex Boettger) Tanna (MCZ, ex BPBM 81642); Anelgauhat, Aneiteum (DMNZ M. Laird!); Aneiteum (DMNZ, M. Laird!); Aore (DMNZ, M. Laird!) Imatangi Village, Futuna (DMNZ, M. Laird!). Remarks. — The juvenile specimens (type lot; CM) compare well with the original description of P. layardi (Ancey). On Espiritu Santo, adult shells were collected May 5, but only very young shells (5-6 mm.) in late May and June. Slightly larger specimens (7-10 mm.) were found in late July. Most of the material came from a semi-permanent pond near Brigstock Point. Kuntz recorded in his field notes that the animals were infected with small stylet cercariae. Locality ML 68 was in a flowing creek; otherwise all specimens were from still water. 164 FIELDIANA: ZOOLOGY, VOLUME 43 The conchological variation found within a single population of P. layardi is exceptionally large (pi. 25). Such plasticity of form is well documented in physoid Planorbidae and appears to be the norm rather than the exception (see Hedley, 1917; Mandahl-Barth, 1954, pp. 101-116). The largest specimen of P. layardi was 19.4 mm. high and 10 mm. in diameter, with 6}4 whorls. Most individuals collected in July (ML 85) were the size (10 mm. high) and whorl count (5 whorls) of the cotypes from Vate (CM). Specimens from the other islands were well within the range of variation found in the Espiritu Santo population. I have assumed that the Gaua popula- tion does not differ significantly. Several individuals from ML 65 were dissected. The jaw, radula, and distal male genitalia were compared to the figures in Hubendick (1948). No differences were found between the Espiritu Santo ani- mals and Physastra doliolum (Gassies), which Hubendick studied. The anatomy is not refigured. Genus GYRAULUS Charpentier, 1837 Type species. — Planorbis hispidus Draparnaud (=alhus Miiller). Remarks. — Gyraulus-\ike snails are found in fresh water nearly everywhere. Little is known of their anatomy, and the generic posi- tion of most of the Indo-Pacific species is very poorly understood. The New Hebridean G. montrouzieri (Gassies) has not been dissected. The shell is very similar to that of the Indonesian G. convexiusculus (Hutton), which is a typical Gyraulus (F. C. Baker, 1945, pp. 67-70). Gyraulus (Gyraulus) montrouzieri (Gassies). Plate 33, figs. 10-12. Planorbis montrouzieri Gassies, 1863, Faune Conchy. Nouv. Caledonie, I, p. 283, pi. 7, fig. 17— Kanala, New Caledonia; Bavay, 1908, Nova Guinea, ZooL, 5: 289 — Lac Sentani, Dutch New Guinea. Planorbis (Spirodiscus) montrouzieri Gassies, Crosse, 1894, Jour, de Conchy., 42: 334-335. Planorbis (Gyraulus) montrouzieri Gassies, Germain, 1923, Rec. Indian Mus., 21, (3), pp. 147-148, pi. 2, figs. 7-9. Gyraulus montrouzieri (Gassies), F. C. Baker, 1945, MoUuscan family Plan- orbidae, p. 71; Franc, 1957, Mem. Mus. hist, nat., n.s., Zool., 13: 90, pi. 9, fig. 122. Range. — Espiritu Santo, New Caledonia, possibly New Guinea. Material.— ML 62, ML 85a, ML 85b, ML 82a; Kanala, New Cale- donia (ANSP 71178, ex Vanatta, topotypes) ; New Caledonia (ANSP SOLEM: MOLLUSCA OF THE NEW HEBRIDES 165 21694, ex Marie; UMMZ 16116, ex Steams; UMMZ 86793, ex Walker, MacAndrews) . Remarks. — Topotypes of G. montrouzieri (ANSP 71178) were more sharply carinated than most of the New Hebridean shells. The fig- ures in Germain (loc. cit.) and the other New Caledonian specimens, however, all closely approximated the Kuntz series. There are sev- eral species of New Caledonian "Planorhis" (see PYanc, 1957, pp. 89- 91, figs. 121-123). Related extralimital species include all the Glyptanisus of Iredale (1943b, pp. 225-227) and Cotton (1943) and G. convexiusculus (Hut- ton). The Australian Glyptanisus appear to be badly oversplit and the genus itself is probably only a section of Gyraulus (sens. str.). Australian material (UMMZ) is intermediate between montrouzieri and convexiusculus in that the umbilicus is moderately excavated and the apical whorls are deeply sunken; in montrouzieri the base is rather deeply excavated and the apical whorls are only slightly sunken; and in convexiusculus the base is nearly flat and the apical whorls are sunken well below the level of the spire. G. montrouzieri differs from both Glyptanisus and convexiusculus in having a submedian rather than a median keel. G. convexiusculus has an exceedingly wide range of variation, and long synonymy. Information on this species is summarized in the papers of van Benthem Jutting (1931), B. Rensch (1934), and Be- quaert and Clench (1939). Definitely known from as far east as the Celebes and Buru, probably Gyraulus elberti Haas, 1912, from Lom- bok, and Planorhis turbinellus Tapparone-Canefri, 1883, from the Aru Islands, will be added to the synonymy of G. convexiusculus. The report of montrouzieri from New Guinea needs confirmation. It may only be a new record for convexiusculus. Three of the New Hebridean series, ML 82a, ML 85a, and ML 85b, were collected in June and July from the quiet waters of a dammed Table XI. — Measurements of Gyraulus montrouzieri (Gassies) Greater Lesser diameter diameter Height H/D Umbilicus Whorls ANSP 71178... 4.47 3.70 1.25 0.35 1.55 S^^ (topotypes).. 4.17 3.52 1.21 0.29 1.49 35^ 4.00 3.22 1.12 0.28 1.29 3^ ML 62 5.93 4.78 2.05 0.35 2.17 AH 5.30 4.44 1.99 0.38 1.99 4 4.73 3.82 1.54 0.34 1.59 SH 4.16 3.54 1.31 0.31 1.37 3}^ ML 85 4.05 3.31 1.48 0.34 1.37 3^ 3.76 3.08 1.42 0.38 1.14 3 166 FIELDIANA: ZOOLOGY, VOLUME 43 creek and a fresh-water pool connected with the creek. The shells were appreciably smaller than those collected in May from a semi- permanent pond (ML 62) on the east side of Brigstock Lagoon. The differences were in whorl count and diameter, large specimens from the first three series having the same whorl count and diameter as small specimens from the fourth. Generally the larger specimens were more sharply keeled. Diameter is directly dependent on whorl count, and measurements of several shells are given in Table XI for comparative purposes. As in all Pacific Gyraulus there is much vari- ation in the degree of carination. Subclass PROSOBRANCHIA Milne-Edwards, 1848 The recent prosobranch classification of Thiele (1929) has been accepted here without major changes, and diagnoses of the higher categories are not usually given below. An expanded classification, including the fossil mollusks, can be found in Wenz (1938-44). Several prosobranch families — the Neritidae, Viviparidae, Ampul- lariidae, Valvatidae, Hydrobiidae, Micromelanidae, and Thiaridae — colonized the fresh-water habitat, while others — the Hydrocenidae, Helicinidae, Cyclophoracea, Truncatellidae, Pomatiasidae, Acmidae, and Assimineidae — have become land dwellers. The only truly fresh- water prosobranch known from the New Hebrides is the hydrobiid, Fluviopupa brevior (Ancey). Many "primary" fresh-water Neritidae and Thiaridae are found in other areas of the world, but none are known from the New Hebrides. All the Melanesian Thiaridae and Neritidae seem capable of passing salt water filter zones with relative ease and are very widely distributed. Of the land prosobranch fam- ilies, the Helicinidae, Cyclophoracea, Truncatellidae, and Assiminei- dae are native to the New Hebrides. Order ARCHAEOGASTROPODA Superfamily NERITACEA Family HELICINIDAE Shell imperforate, strongly depressed to conic, with few whorls. Umbilical region covered by a prominent callus. Operculum with thin to moderately heavy calcareous layer, either concentric or paucispiral. Radula rhipidoglossate. Foot broad, undivided or tripartite. (Modified from Pilsbry, 1948, p. 1078.) The most recent monograph of the Helicinidae (Wagner, 1907- 11) is very inadequate. The illustrations are useless, no attention SOLEM: MOLLUSCA OF THE NEW HEBRIDES 167 has been paid to the work of previous authors, nomenclatural proce- dure is not followed, and errors in citation, fact, and localities are many. Fulton (1915) listed 137 species of Helicinidae which Wag- ner did not even mention. In recognizing the distinctness of the Antillean and Indo-Pacific helicinids, Wagner made a valuable con- tribution, but the numerous errors and omissions mar the otherwise favorable impression that his work merits. Six new genera, and many subgenera, based entirely on shell and opercular characters, were established by Wagner. Unfortunately he did not designate type species, and over a span of two or three years he indiscriminately shifted species from genus to genus. H. B. Baker (1922), Thiele (1929), Pilsbry and Cooke (1934b), Iredale (1937a) , and Wenz (1938-44, pp. 435-444) all tried to settle the nomen- clatural status of Wagner's genera. Iredale (1937a, pp. 291-293) complicated the issue by accepting Wagner's "Formenkreise" as formally proposed sectional names. They were not intended as such, being simply "species groups" often designated by the trivial name of one of the included species (see Zilch, 1948). A list of the available generic units of Pacific helicinids follows, giving type designations and geographic range: Waldemaria A. J. Wagner, 1905 Type species. — Helicina japonica A. Adams, by subsequent designation of H. B. Baker (1922, p. 41). Range. — Japan. This is a section of the North American Hendersonia. Sulfurina MoellendorflF, 1893 Type species. — Helicina citrina Grateloup. Range. — Indonesia, the Philippines, Nicobar, Palau, and the Caroline Islands (see fig. 23). Sturanyella Pilsbry and Cooke, 1934 Type species. — Helicina plicatilis Mousson. Range. — Carolines, Solomons, Fiji, Tonga, Wallis, Samoa, Cook, and (?)Galap- agos (see fig. 22). Remarks. — Sturanyella is a new name for the group usually called Sturanya Wagner. The latter name was transferred to Orobophana (see below) by the type designation of H. B. Baker (1922). Geophorus Fischer, 1885 (see Bartsch, 1918) Type species. — Helicina agglutinans Sowerby. Range. — Philippines and Indonesia (see fig. 23). Pleuropoma Moellendorfif, 1893 Type species. — Helicina dichroa Moellendorfif. Range. — Indo-Pacific Islands except Tuamotus, Marianas, and Marshalls (see fig. 22). 168 FIELDIANA: ZOOLOGY, VOLUME 43 Synonyms of Pleuropoma are: Aphanoconia Wagner, 1905 Type species. — Helicina verecunda Gould, by subsequent designation of Gude (1921, p. 366). Sphaeroconia Wagner, 1905 Type species. — Helicina verecunda Gould, by subsequent designation of H. B. Baker (1922, p. 43). Orobophana Wagner, 1905 Type species. — Helicina uberta Gould, by subsequent designation of H. B. Baker (1922, p. 43). Range. — New Caledonia(?), Fiji, Tonga, Cook, Niue, Society, Tuamotus, Mar- quesas, and Hawaii (see fig. 23). Sturanya Wagner, 1905 Type species. — Helicina laciniosa Mighels by subsequent designation of H. B. Baker (1922, p. 43). Remarks. — Johnson (1949, p. 226) clearly showed that the Helicina laciniosa of authors (for example, Neal, 1934) is not the laciniosa of Mighels. The latter is equivalent to Orobophana berniceia Pilsbry and Cooke, which means that Sturanya becomes a subjective synonym of Orobophana, and no name of generic rank is available for the "zonata" group of Pleuropoma (see p. 177). No substitute name is proposed, since the generic value of the latter group is doubtful. Palaeohelicina Wagner, 1905 Type species. — Helicina moquiniana Recluz, by subsequent designation of Thiele (1929, p. 87). Range. — The Solomons, Bismarcks, New Guinea, and Palau Islands (see fig. 23). Ceratopotna Moellendorff, 1893 Type species. — Helicina caroli Kobelt. Range. — The Philippines, Indonesia, and Australia (see fig. 22). Remarks. — Wagner considered Ceratopoma to be a subgenus of his later genus Palaeohelicina. Ceratopoma has priority, and if the two prove to be related, Cera- topoma must be used as the generic name. The above list is primarily nomenclatural, since little is known about the Pacific Helicinidae. Bourne (1911), H. B. Baker (1922, 1925b, 1926), and van Ben them Jutting (1941) published papers on the anatomy of the Helicinidae. Very few Pacific species have been examined. Bourne (1911) dissected four Pleuropoma, one Orobo- phana, and a Ceratopoma. H. B. Baker (1922) studied the radulae of Waldemaria and Pleuropoma verecunda (Gould), and van Ben them Jutting (1941, p. 6) the radula of Sulfurina cerinella van Benthem Jutting. These papers and the brief note on Hawaiian helicinids by Pilsbry and Cooke (1909) form the total published knowledge of the Pacific Helicinidae. SOLEM: MOLLUSCA OF THE NEW HEBRIDES 169 Orobophana has a distinctive radula, but there are only minor differences between Pleuropoma, Sturanya, Aphanoconia and Cerato- poma. Animals of three New Hebridean species were available. I saw no extralimital specimens, and, other than radular mounts, did not study the anatomy. On the basis of radular structure, the New Hebridean species are here placed in Pleuropoma. The radula of the Helicinidae is extremely complex and difficult to interpret correctly (see H. B. Baker, 1922, p. 33). Since no comparative material was available and I have not previously worked with helicinid anatomy, the radulae are not figured at this time. The radulae of Pleuropoma varians (Sykes), P. albescens (Hartman), and P. sublaevigata (Pfeiffer) are extremely similar and apparently there are no major differences between them. The land operculates are all dioecious and in many genera there is considerable sexual dimorphism. In the Neritidae, Hydrobiidae, Cyclophoracea, and Assimineidae the females are much larger than the males and the males have a prominent cephalic penis. In the Helicinidae there are no external sexual differences, and H. B. Baker's (1925b, pp. 274-275) attempt to determine sexual dimorphism yielded mainly negative results. It was not possible to undertake a study of sexual dimorphism in the present investigation. Wagner (1907-11) and Bourne (1911, pp. 760-762) discussed the geographic distribution of the Helicinidae. The family is absent from Europe, Africa, most of Asia, New Zealand, Norfolk and the Lord Howe Islands, and much of Australia. Most helicinids are found in the Greater Antilles and the Indo-Polynesian region. There are only a few continental relatives, and very few species live outside the tropics. Twelve species of Helicinidae have been described from the New Hebrides. Only six are here accepted as valid ; three, modesta Pfeiffer, zebriolata Pfeiffer, and subreticulata Wagner, are probably based on erroneous locality data; and three, bairdii Reeve, layardi Hartman, and novella Mabille, are reduced to synonymy. Of the six recognized species, one, Pleuropoma articulata CPfeiffer), is related to a widely distributed group, while the other five, P. varians (Sykes), P. taeniata (Quoy and Gaimard), P. albescens (Hartman), P. sublaevigata (Pfeif- fer), and P. rotella (Sowerby), form a homogeneous series without discernible extralimital relatives. No preserved material of P. articu- lata was available, but subsequent study may place this species in a different genus from the others. There is no apparent relationship between the Solomon Island and New Hebridean Helicinidae, and 170 FIELDIANA: ZOOLOGY, VOLUME 43 only a weak one between the New Hebridean and New Caledonian species. Key to the New Hebridean Helicinidae 1. Body whorl with knife-edge carina; diameter more than 9 mm 2 Body whorl rounded or angulated; if carinate, then diameter less than 6.5 mm.. .3 2. Columellar margin with basal tooth; Vanikoro. Pleuropoma taeniata (Quoy and Gaimard) Columellar margin without basal tooth; Santa Cruz Island ... P. varians (Sykes) 3. Color pattern of spiral bands or unicolored 4 Color pattern of alternating red and white radial fiammulations. P. articulata (PfeiflFer) 4. Body whorl angulated; lip strongly reflected; weak columellar sinus 5 Body whorl rounded; lip weakly reflected; strong columellar sinus. P. sublaevigata (Pfeiffer) 5. Shell larger, 9-11 mm.; Tanna P. rotella (Sowerby) Shell smaller, 6-8 mm.; Espiritu Santo P. albescens (Hartman) Genus PLEUROPOMA Moellendorff, 1893 In accordance with the determinations of H. B. Baker (1922), Pilsbry and Cooke (1934b), and Wenz (1938-44), Aphanoconia and its synonym Sphaeroconia are here considered congeneric. P. articu- lata would be placed in the zonata group, if the latter proves to be separable, while the other New Hebridean species form an isolated group of Pleuropoma. Anatomical studies may reveal that the New Hebridean species belong in a special section, but nomenclatural rec- ognition is not warranted at this time. Pleuropoma varians (Sykes). Plate 26, figs. 3, 4. Helicina varians Sykes, 1903, Jour, of Malac, 10, (2), p. 67 — Santa Cruz Islands; Sykes, 1903, op. cit., 10, (13), p. 78, pi. 6, figs. 5, 6. Range. — Santa Cruz, Santa Cruz Islands. Ma^erm^.— Santa Cruz Island (UMMZ 72060, ex Walker, Pon- sonby, Sykes, paratypes; MCZ 32609, ex W. M. Mann!). Remarks. — P. varians superficially resembles the northern Mela- nesian Palaeohelicina, but it lacks the spiral striae of that genus and has the same embryonic spiral ridges, deciduous periostracal bands, and type of radula found in the New Hebridean Pleuropoma. Most of the 89 individuals that I examined were yellow with red supra- peripheral color band, but others were uniform yellow and a few had radiating reddish-brown color streaks. The variation in size and o JS rG \00 ■^ Tj< -^ ■^ 7T CO CO TT ^ai ^^ '^-' *— ' Q t- 00 o C is ■^ IM 00 00 00 CO I I I OJ 00 -l t- J> CO CD CO 00 la 00 to CO S J." lO to CO to CO to X Xi tf o o o CO • »— I t~ t- rH I i I eo to i-H CO CO OJ K III § CO t> O '-( o ^'^ 1— ( >— t I— I • o> 00 o> Q i-H (M (M CZJ o o o to CO CO I I I CO 00 CO ^ y-l 1-i la 2 IM eg >-H 00 CO ^^ to to CO CO CO ^ » CS C5 S ^'^Is IM l-H o o to -^ I I 00 o> 5 to ""• OS Q 2 s O to o d O to CO U3 I I o 00 to T]i O 00 to eg to ■«S' to U3 to Oi o d d d t> ^ OS 00 eo eo 00 op ( 1 (M CO oq d eo oi eo" CO to eg •>* 00 00 to 00 IM 00 eo 1— I I— I O rt IM .-H ^ OS 5 171 172 FIELDIANA: ZOOLOGY, VOLUME 43 shape was the most found in any New Hebridean hehcinid (see Table XII). Nothing is known about the ecology of P. varians. The bright color and trochoidal shape suggest that it may be an arboreal species. Pleuropoma taenia ta (Quoy and Gaimard) Helicina taeniata Quoy and Gaimard, 1832, Voy. Astrolabe, Zool., 2: 194, pi. 12, figs. 6-10— Vanikoro, Santa Cruz Islands; Sowerby, 1847, Thes. Conchy., I, p. 14, pi. 2, figs. 49-51; Pfeiffer, 1846, Syst. Conch. Cab., I, 18, (1), p. 63, pi. A, fig. 14; Sowerby, 1866, Thes. Conchy., 3: 287, pi. 274, figs. 322-323; Sowerby, 1873, Conch. Icon., Helicina, pi. 10, figs. 85-86. Range. — Vanikoro Island. Material. — No material available. Remarks. — All of the records in the literature for this species are based on the original description and figures. The figured spec- imens could not be found in Paris by Andr^ Franc, and the only specimens in the Paris Museum labeled taeniata did not match the description or figures. Without topotypic material or rediscovery of the type specimens, taeniata must remain a dubiously valid species. It is most similar to P. varians (Sykes) but differs by having a dis- tinct basal tooth within the columellar lip. Pleuropoma albescens (Hartman). Plate 26, figs. 5-7, 11. Helicina albescens Hartman, 1890, Proc. Acad. Nat. Sci. Philadelphia, 1890: 285, pi. 3, fig. 5 — Segond Channel, Espiritu Santo. Range. — Espiritu Santo. Material.~ML 1, ML 2, ML 31c, ML 31d, ML 31e, ML 31g, ML 32, ML 37a, ML 39, ML 51, ML 66, ML 76a, ML 86, ML 95; Segond Channel, Espiritu Santo (CM 62.15313, ex Hartman, Layard, holotype); New Hebrides (UMMZ 72208, ex Walker). Remarks. — The holotype of P. albescens (pi. 26, fig. 5) is a large individual, 7.5 mm. in diameter, 5.7 mm. high, with 43/^ whorls. Most specimens (see Table XII) are smaller, differing from the type only in size. P. albescens superficially resembles P. sublaevigata. On Espiritu Santo, the two species were found at the same collecting site, but apparently they are ecologically isolated, sublaevigata living on trees and albescens usually under ground debris. Individuals of sublaevigata and albescens can be separated by several characters. For convenience they are listed in the following table. SOLEM: MOLLUSCA OF THE NEW HEBRIDES 173 P. sublaevigata 1. A strong columellar sinus 1. (see fig. 6). 2. Periphery rounded. 2. 3. Apex flattened. 3. 4. Lip weakly or not reflected. 4. 5. Usually with spiral color bands. 5. 6. Juvenile shells without any hairs. 6. 7. Spire usually convex. 7. 8. Operculum unicolored. 8. P. albescens A weak or no columellar sinus (see fig. 6). Periphery keeled. Apex elevated. Lip strongly reflected. No spiral color bands. Juvenile shells with two or more spiral rows of hairs. Spire usually flattened. Operculum with a red border. Numbers 1, 2, 4, and 8 are the most useful for identifying adult shells, numbers 1 and 6 for juveniles. All specimens of albescens have a light spiral band on the periph- ery of the body whorl. The color of the rest of the shell varies from yellow or white to orange and wine-red. The color is not arranged in bands but is distributed over the entire shell in a uniform hue. The spiral ridges characteristic of sublaevigata are also present in albescens, although to a much lesser degree. It is suspected that albescens and sublaevigata may have the same ecological and system- atic relationship as the Trochomorpha rubens (Hartman) complex. Unlike the latter, however, there is no indication of hybridization between albescens and sublaevigata in the plantation areas of Espi- ritu Santo. Pleuropoma sublaevigata (Pfeiffer). Plate 10, figs. 1-3; plate 26, figs. 12, 13. Helicina sublaevigata Pfeiffer, 1852, Monog. Pneumon. viv., I, p. 384 — New Hebrides; Pfeiffer, 1854, Proc. Zool. Soc. London, 1852: 87; Sowerby, 1866, Thes. Conchy., Ill, p. 290, pi. 274, figs. 339-340; Sowerby, 1873, Conch. Icon., Helicina, pi. 29, fig. 265a, b; E. A. Smith, 1884, Proc. Zool. Soc. London, 1884: 268— Api Island; Sykes, 1903, Proc. Malac. Soc. London, 5: 200 — Vate, Espiritu Santo, Valua, Vanua Lava, Gaua, Hiu, Lo. Helicina layardi Hartman, 1888, Proc. Acad. Nat. Sci. Philadelphia, 1888: 251, pi. 13, fig. 6— Aore Island; Ancey, 1905, Nautilus, 19, (4), pp. 45-46. Helicina novella Mabille, 1895, Bull. Soc. d'Hist. Nat. d'Autun, 8: 400-401— New Hebrides (Francois). Helicina bairdi Reeve, Mabille, 1895, Bull. Soc. d'Hist. Nat. d'Autun, 8: 401 — New Hebrides. Orobophana sublaevigata (Pfeiffer), Wagner, 1905, Denk. Akad. Wien, 77: 426, pi. 7, fig. 17a-c. Orobophana sublaevigata layardi (Hartman), Wagoner, 1905, Denk. Akad. Wien, 77: 427, pi. 7, fig. 18. Aphanoconia (Sphaeroconia) sublaevigata (Pfeiffer), Wagner, 1909, Conch. Cab., I, 18, (2), p. 208, pi. 41, figs. 23-27. 174 FIELDIANA: ZOOLOGY, VOLUME 43 Palaeohelicina (Ceraiopoma) layardi (Hartman), Wagner, 1909, Conch. Cab., I, 18, (2), p. 261, pi. 51, figs. 18-22. Range. — Tanna, Vate, Epi, Aore and Espiritu Santo; Valua, Vanua Lava, and Gaua in the Banks; and Hiu and Lo in the Torres. Material— ML 1, ML 2, ML 26f, ML 31e, ML 32, ML 37a, ML 39, ML 51, ML 66, ML 69, ML 70, ML 76a, ML 86, ML 95; Port Reso- lution, Tanna (UMMZ 72207, ex Walker) ; Vate (MCZ 140592, ex Bishop Museum; ANSP 14498, ex Dupuy; CNHM 28627, ex Webb, Ritchie; CM 62.15391, ex Hartman, Layard; ANSP 179794, ex Wolf; UMMZ 72209, ex Walker, Ponsonby, Layard) ; Vila, Vate (ANSP 133300, ex Froggatt; ANSP 133292, ex Froggatt; Miller 68); Hog Harbour, Espiritu Santo (MCZ 141035, ex J. McMillan) ; Espiritu Santo (MCZ 113620, ex Bequaert; USNM 432449, 432450, 432451, 432452, ex Harrington; AMNH, ex Banner); Aore (CM 62.15656, ex Hartman, Layard, holotype and paratypes; MCZ, ex W. L. Nut- ting); New Hebrides (UMMZ 72304, ex Walker, Fulton; ANSP 60914, ex Hartman, Cox). Remarks. — The only extralimital species closely related to Pleuro- poma suhlaevigata (Pfeiff er) is P. primeana (Gassies) from the Loyalty Islands and New Caledonia. The two species are very similar in shape, color variations, and size but are easily separated by their different columellar lip margin (see fig. 6). No Solomon Island or Polynesian helicinid is closely related to suhlaevigata or primeana. The Espiritu Santo endemic, P. albescens (Hartman), resembles suh- laevigata but is distinguished by several characters (see above) . A. Wagner (1907-11, pp. 208 and 261) divided suhlaevigata into two species belonging to two genera. The species were established on the presence (layardi) or absence (suhlaevigata) of spiral ridges on the embryonic whorls. The type specimens of both layardi and suh- laevigata have the spiral ridges. In only 5 per cent of the 575 speci- mens examined are the ridges completely absent; in many others, however, they are greatly reduced in prominence. Apparently the ridges have no taxonomic value. The background color of P. suhlaevigata varies from white to pale yellow, with the latter more common. Most specimens have red supra- and subperipheral color bands. The bands are usually nar- rower than those found in the syntype (pi. 10, figs. 1, 2). In a few specimens they cover the entire whorl; in others they are split into two fine red lines. Either or both color bands can be absent. In about 15 per cent of the specimens the color bands are unpigmented and are represented by clear, translucent areas in the shell. All vari- SOLEM: MOLLUSCA OF THE NEW HEBRIDES 175 a Fig. 6. Umbilical area of Pleuropoma. a, Pleuropoma primeana (Gassies); Lifu, Loyalty Islands (UMMZ 72122, ex Walker, Ponsonby, Layard). b, Pleuro- poma albescens (Hartman); Espiritu Santo (ML 32). c, Pleuropoma sublaevigata (Pfeiffer); Espiritu Santo (ML 66). Scale line=l mm. ations were found in the larger samples. It is suspected that this is the same type of genetically determined polymorphism found in Brady- baena similaris (Ferussac) by Komai and Emura (1955). Most sam- ples did not contain enough individuals to make tabulation of the different percentages of color forms worth while. In the one large sample (ML 1, with 109 specimens), 35 specimens lacked all bands, 38 had both, 15 had only translucent areas, 7 had no supraperipheral band, 6 had no subperipheral band, 5 had both bands very wide, and 3 had only a very wide subperipheral band. A few large specimens of sublaevigata were more than 8 mm. in diameter, but the majority were much smaller (see Table XII). "Helicind" novella Mabille was based on specimens 8-10 mm. in diameter. Although Andr^ Franc was unable to locate the types in the Paris Museum, it is suspected that novella was based on large specimens of sublaevigata. The record for P. sublaevigata from Lifu, Loyalty Islands (see Franc, 1957, pp. 34-35, fig. 40) is incorrect, since the figured specimen is quite different from the holotype and the New Hebridean specimens. Most specimens of sublaevigata were collected from the leaves of trees or shrubs, from two to six feet above the ground; a few were collected under debris and logs, and the ecological position of sub- laevigata needs to be clarified. 176 FIELDIANA: ZOOLOGY, VOLUME 43 Pleuropoma articulata (Pfeiffer). Plate 10, figs. 7-9. Helicina articulata Pfeiffer, 1854, Malak. Blatt., 1 : 103 — Tanna, New Hebrides (April); Pfeiffer, 1854, Proc. Zool. Soc. London, 1853: 53— Tanna, New Hebrides (July); Pfeiffer, 1858, Monog. Pneum. viv., suppl. 1, pp. 191-192; Sowerby, 1866, Thes. Conchy., 3: 294, pi. 12, figs. 409-410; Sowerby, 1873, Conch. Icon., Helicina, pi. 26, fig. 236; Sykes, 1903, Proc. Malac. Soc. London, 5: 200— Port Vila, Vate (J. J. Walker). Aphanoconia articulata (Pfeiffer), Wagner, 1908, Conch. Cab., 1, 18, (2), p. 162, pi. 32, figs. 1-3. Range. — Tanna, Vate, Aneiteum. M aterial— Aneiteum (ANSP 132647, ex Brazier! 1865; DMNZ, W. H. Dawbin!); Vila, Vate (Miller 69, USNM 598362, ex Miller; UMMZ 72115, ex Walker, Ponsonby, Sykes); Vate (UMMZ 72116, ex Walker, Layard; UMMZ 72199, ex Walker, Layard, Garrett; CM 62.15338, ex Hartman, Layard; DMNZ, W. H. Dawbin!); New Heb- rides (UMMZ 72114, ex Walker; AMNH; MCZ; CNHM 28729, ex Webb, Sowerby; CNHM 28820, ex Webb, Sowerby). Remarks. — Pleuropoma articulata is quite variable in color pattern, degree of carination, and sculpture. There are two different color patterns. Most specimens (pi. 10, figs. 7-9) have the middle of each whorl clear and slightly translucent with the alternating red and white flammulations confined to the periphery and areas bordering the sutures. In some the zigzag pattern is continuous over the entire whorl, as in fulgora Gould, zehriolata Pfeiffer, and suturalis von Mar- tens. In most shells the stripes are absent on the base of the shell, but in a few specimens they continue to the umbilical callus. Many specimens of intermediate aspect were seen. Acutely carinated, ob- tusely keeled (pi. 10, fig. 8) and intermediate shells were also found. Narrow spiral ridges, arising from a deciduous periostracum, are present in some shells but are totally absent or partially eroded in the majority. The Vate and Aneiteum populations differ in size and degree of carination (see Table XII). The Aneiteum population closely approx- imates the type figures of P. zehriolata (Pfeiffer) (see p. 179), and perhaps the name should be applied to the Aneiteum shells rather than the supposed Samar (Philippine Islands) population. In view of the uncertainty of the reference of the name zehriolata, the uncer- tainty of the importance of the variation between the Vate and Aneiteum populations, and the complexity of the entire "zonata" group (see below), I have not given taxonomic recognition to the Aneiteum population. SOLEM: MOLLUSCA OF THE NEW HEBRIDES 177 It is interesting to note the differences between the populations collected in different years on Vate and Aneiteum (see Table XII). There is, of course, little probability that the identical local popula- tions were sampled at the two times, but the size differences between the two sets illustrates the unreliability of size as a valid character in separating species of land prosobranchs. The characters mentioned in the preceding paragraph have been used to separate the many small helicinids related to articulata. The range of variation shown by the New Hebridean shells encompasses most of the named taxa. E. A. Smith (1897, p. 522) previously sug- gested that several of the "species" are probably synonymous. In none of the "species" have the soft parts been examined and only in Hawaii have local colonies of these helicinids been sampled and com- pared (see Neal, 1934, pp. 39-78). It would be very unusual if only one species could be recognized over the entire Pacific, but certainly many of the names in current use will be discarded when the entire complex is studied in detail. A probably incomplete list of "species" closely related to articu- lata includes the following "Helicina": zonata Lesson, 1830, zigzag Pease, 1867, and ponapensis Wagner, 1911, from the Caroline Islands; suturalis von Martens, 1864, from the Moluccas; possibly zebriolata Pfeiffer, 1855, from the Philippines; gallina Gassies, 1870, and mariei Crosse, 1870, from New Caledonia; vicina Wagner, 1911, from Fiji; fulgora Gould, 1847, musiva Gould, 1847, hrenchleyi Baird, 1873, and altivaga Ancey, 1889, from Samoa; ^amwea Quoy and Gaimard, 1832, and diminuta Mousson, 1871, from Tonga; oceanica Pease, 1867, from the Gilbert Islands; and laciniosa Neal, 1934 (not Mighels, 1845), oahuensis Pilsbry and Cooke, 1908, nonouensis Neal, 1934, and sub- sculpta Neal, 1934, from the Hawaiian Islands. Many specimens of the above "species" are in the University of Michigan Museum of Zoology. The Hawaiian, Moluccan, and Gilbert Island shells are easily separated from the New Hebridean, but the great majority of the others are very similar. Pleuropoma rotella (Sowerby), Plate 10, figs. 4-6. Helicina rotella Sowerby, 1847, Thes. Conchy., I, p. 12, pi. 3, figs. 135-136— locality unknown; Pfeiffer, 1852, Monog. Pneum. viv., I, p. 387; Pfeiffer, 1858, op. cit., II, p. 109— Tanna, New Hebrides; Sowerby, 1866, Thes. Conchy., 3: 291, pi. 275, fig. 358— New Hebrides; Sowerby, 1873, Conch. Icon., Helicina, pi. 12, fig. 104. Range. — Tanna. Material. — Photograph of holotype (BM). 178 FIELDIANA: ZOOLOGY, VOLUME 43 Remarks. — Identification of the Tanna shells (probably collected by MacGillivray) with Helicina rotella needs confirmation. The only helicinids from Tanna examined during this study (UMMZ 72207) were slightly atypical suhlaevigata. P. rotella differs from suhlaevi- gata in having a more sharply angulated periphery and an expanded lip. In this respect, as well as in color variation, rotella resembles the Espiritu Santo albescens. The latter is substantially smaller than rotella (which is 9-11 mm. in diameter) and has a less indented colu- mellar sinus. The following three species, described from the New Hebrides, probably were from other areas. P. subreticulata (Wagner) and Stur- anyella modesta (Pfeiffer) almost certainly are extralimital, but Pleuro- poma zebriolata (Pfeiffer) may be a synonym of P. articulata (Pfeiffer) . Sturanyella modesta (Pfeiffer) Helicina modesta Pfeiffer, 1854, Proc. Zool. Soc. London, 1853: 52 — Tanna, New Hebrides; E. A. Smith, 1885, op. cit., 1885: 598-599— Shortland, Treasury, and Guadalcanal Islands in the Solomons. Sturanya modesta (Pfeiffer), A. J. Wagner, 1907, Conch. Cab., I, 18, (2), p. 43, pi. 6, figs. 25-29; I. and B. Rensch, 1936, Rev. Suisse Zool., 43, (32), p. 683— New Georgia; Dell, 1955, Pacific Science, 9, (3), p. 325— Nissan Island, northern Solomons. Range. — Solomon Islands. Material. — Many lots in the collection of the University of Mich- igan Museum of Zoology. Remarks. — The identity of Cuming's types with Solomon Island specimens was established by E. A. Smith. Helicina modesta, type locahty Tanna, New Hebrides, is "No. 19" in Pfeiffer's list of new species. "No. 18" in the same list is Pupina cumingiana Pfeiffer, type locality Solomon Islands, which has only been found on Tanna in the New Hebrides (see p. 189). It is very probable that the local- ities were accidentally transposed. No shell resembling S. modesta is known from the New Hebrides, and I have not accepted the species as part of the fauna. Pleuropoma subreticulata (Wagner) Aphanoconia (Sphaeroconia) subreticulata A. J. Wagner, 1909, Conch. Cab., I, 18, (2), pp. 196-197, pi. 40, figs. 5-9— New Hebrides. Range. — Queensland . MafermZ.— Queensland (UMMZ 72212, ex Walker, Rolle, cotype). SOLEM: MOLLUSCA OF THE NEW HEBRIDES 179 Remarks. — Wagner compared P. subreticulata to the Australian P. gouldiana (Forbes), noting only a very few minor differences. The presence of a cotype in the collection (UMMZ) from Queensland, where P. gouldiana lives, makes it very probable that the New Hebridean locality for subreticulata is an error in Wagner's mono- graph. Pleuropoma zebriolata (Pfeiffer) Helicina zebriolata Pfeiffer, 1855, Proc. Zool. Soc. London, 1855: 101-102— Lord Howe Island (error); Pfeiffer, 1858, Monog. Pneum. viv., suppl. 1, p. 191; Sowerby, 1866, Thes. Conchy., 3: 294, pi. 12, figs. 407, 408— Samar, Philippine Islands; Sowerby, 1873, Conch. Icon., Helicina, pi. 26, fig. 229. Geophorus (?) zebriolata (Pfeiffer), Faustino, 1930, Philippine Jour. Sci., 42, (1). p. 174. Range. — Samar, Philippine Islands. Material— Samar (UMMZ 72062, ex Walker, Ponsonby, Melville). Remarks. — No helicinids live on Lord Howe Island (Iredale, 1944) and the original locality is thus an obvious error. Sowerby's identifi- cation of zebriolata with the Samar shells is here accepted with some reservations. The coloration and size of zebriolata are the same as the large, sharply keeled variety of Pleuropoma articidaia (Pfeiffer). Study of the types of zebriolata may result in its being added to the synon- ymy of articulata instead of being referred to the Samar shells. Moellendorff (1897, pp. 179-182) did not include zebriolata among the Philippine Island Helicinidae, although the record for Samar had been published more than thirty years prior to his study. It is thus possible that Samar may be another erroneous locality. Order MESOGASTROPODA Suborder ARCHITAENIOGLOSSA Superfamily CYCLOPHORACEA Aside from the systematic compilation of Kobelt (1902b), no family-wide studies of the Cyclophoracea have appeared since those of L. Pfeiffer in the middle 1800's. In recent years systems of classification have been proposed by Thiele (1929), Wenz (1938-44), Tielecke (1940), Torre, Bartsch and Morrison (1942), and Morrison (1955). Thiele's and Wenz's systems were only slightly changed from Kobelt's, but the others proposed important modifications. In establishing higher categories Tielecke relied on the structures of the 180 FIELDIANA: ZOOLOGY, VOLUME 43 central nervous system, male and female genitalia, and pallial region, Torre, Bartsch and Morrison on the shell and operculum, and Mor- rison on the external male reproductive organs. Differences between the classifications are numerous and fundamental. None is fully satisfactory, but that of Tielecke probably has the soundest basis, in that several organ systems are used as diagnostic characters. For this reason, Tielecke's family divisions have been accepted here (see also van Benthem Jutting, 1948). Three of Tielecke's five families, the Poteriidae (fig. 16), Pupini- dae (fig. 14), and Diplommatinidae (= Cochlostomatidae) (fig. 17), are found in the New Hebrides, The Cyclophoridae are common in Indonesia and reach the Solomons. The fifth family, Maizaniidae, is endemic in Africa. Because of the unsettled classification, no attempt has here been made to characterize suprageneric taxa. Family POTERIIDAE Tielecke's family Poteriidae is equivalent to the subfamily Poteri- inae of Thiele (1929, pp. 102-103) with the addition of Ostodes and related Pacific genera (see Clench, 1949, pp. 4-29). Despite concho- logical differences, the genital tracts and nervous systems show that these Oceanic and Neotropical genera are more closely related to each other than to any other cyclophoraceans (see Tielecke, 1940, pp. 361- 363) . Further study may warrant subfamily recognition for the Pa- cific genera, and it is also possible that the African Maizaniidae may prove to be a closely related group. Morrison (1955) does not consider the Pacific cyclophoraceans, but it is worth while to contrast his classification of the American poteriine genera with Tielecke's single family: Family Cyclophoridae Subfamily Cyclophorinae Genus Buckleyia Subfamily Neopupininae Genus Aperostoma {=Cyrtotoma) Subfamily Neocyclotinae Genus Poteria Genus Neocyclotus Family Amphicyclotidae Genus Amphicyclotus Genus Crocidopoma SOLEM: MOLLUSCA OF THE NEW HEBRIDES 181 Morrison's families and subfamilies are based on the structure of the male verge, an organ which is quite variable in land Mollusca. The verge is very useful in systematics, but too much importance can be attached to it, as well as to the shell and the operculum. Clench (1949, pp. 4-29) revised the Pacific Poteriidae, basing the seven genera (five new) on opercular characters. The only "cyclo- phorid" not mentioned in his study is Cyclostoma ohligatum Gould, 1852, from Tahiti. Examination of a specimen (UMMZ 85675) showed that it belongs to the assimineid genus Cyclomorpha. The relation of the New Zealand endemics, Liarea and Murdochia, to the other Pacific cyclophorids is uncertain. The reproductive anat- omy of Liarea is unknown, while the only study on Murdochia (Morton, 1952) does not offer sufficient information for a detailed comparison with the other Pacific genera. The radulae of Murdochia (Morton, 1952, pi. 25, fig. 4) and Ldarea (Powell, 1954, pi. 48, figs. 38-40) are quite different from those of the Polynesian cyclophora- ceans figured by Clench (1949, p. 48, fig. 28), but so few cyclophorid radulae have been studied that the importance of the differences is unknown. According to Morton (1952, pp. 76-77) the oviduct of Murdochia enters the receptaculum seminalis, which opens into the uterus (pal- lial oviduct). No mention is made of the bursa copulatrix. It is impossible to compare the female organs of Murdochia with the spe- cies studied by Tielecke (1940) without knowing the position of the bursa; the point where the oviduct enters the uterus; and the rela- tion between the bursa and receptaculum. Both males and females of Ostodes strigatus (Gould) were dissected by Tielecke (1940, pp. 331-332). Females of Ostodes adjunctu^ (Mous- son) and both sexes of Gonatoraphe fornicata (Pfeiffer) were available for this study. The female anatomy of the three species differed only in minor details. The external male organ of Ostodes strigatus is lo- cated mid-dorsally and well behind the tentacles. It consists of a large, rugose, fleshy verge with narrow terminal appendage. There is a distinct sperm furrow running from the end of the vas deferens to the tip of the penial appendage. The shape and location of the verge in Gonatoraphe fornicata (pi. 6, fig. 10) are the same, but G. for- nicata differs in having the sperm furrow closed to form an internal, tubular "vas" running from the pallial genital pore to the tip of the cephalic penis. In Murdochia (Morton, 1952, pp. 75-76, fig. 2) the penis is lateral and attached behind the right tentacle. There is a 182 FIELDIANA: ZOOLOGY, VOLUME 43 tubular vas deferens, and the tip has developed into a retractile intro- mittent organ. According to the criteria of Morrison (1955), Ostodes would belong to the subfamily Neocyclotinae of the Cyclophoridae, Gonatoraphe to the Amphicyclotidae, and Murdochia to the cyclophorid subfamily Neopupininae. The transition from a seminal groove to a tubular vas deferens is probably only of generic value and not indicative of family affinities in the Cyclophoracea. Until the anatomy of the New Caledonian Gassiesia and Fijian Gonatoraphe is known, the relationship between Gassiesia, Gona- toraphe, and Ostodes will be uncertain. The shell of Gonatoraphe is transitional in shape and sculpture between Gassiesia and Ostodes. The latter two have horny operculi but the operculum of Gona- toraphe is heavily calcified. The importance of the opercular and vergic differences and the affinities of Fijipoma and the Caroline Island Paramia and Kondoraphe cannot be determined at present. I suspect that the Pacific Poteriidae are relicts which have been re- placed in Indonesia and northern Melanesia by the cyclophorine genera Lagochilus and Leptopoma, and that Gassiesia and Ostodes will be found to be primitive relatives of Gonatoraphe. The anatomical information given by Tielecke (1940) and the results of this study enable modification of the generic descriptions in Clench (1949). Expanded generic definitions are given below, not only for the Fijian-New Hebridean Gonatoraphe, but also for the Samoan Ostodes, and the New Caledonian Gassiesia. n Genus OSTODES Gould, 1862 Shell globose to depressed-turbinate, solid, and opaque. Surface with spiral lirae (except 0. cookei Clench) and, in many species, broad, flattened axial plicae. Umbilicus wide, body whorl rounded or acutely keeled. Operculum horny, multi- spiral, with depressed nucleus. Male verge external, mid-dorsal, posterior, with open seminal groove and terminal appendage. Oviduct, seminal receptacle, and bursa copulatrix opening into uterus through a single pore. (Modified from Tielecke, 1940, pp. 331-332, and Clench, 1949, p. 9.) Type species. — Cyclostoma strigatum Gould. Range. — Savaii, Upolu, and Tutuila, Samoa. Genus GONATORAPHE Moellendorff, 1898 Shell depressed turbinate, solid, widely umbilicate. Sculpture of spiral lirae, most species with broad axial plicae. Operculum flat, calcareous, multispiral, with depressed nucleus. Male verge external, mid-dorsal, posterior, with internal tubu- lar ejaculatory duct. Oviduct, seminal receptacle, and bursa copulatrix opening into uterus through a single pore. (Modified from Clench, 1949, p. 18.) SOLEM: MOLLUSC A OF THE NEW HEBRIDES 183 Type species. — Cyclostoma recluzianum Pfeiffer (=fornicatum PfeifTer). Remarks.— Thiele (1928a, p. 128, and 1929, p. 173) placed Gona- toraphe in the Assimineidae, but the radula, operculum, and anatomy prove that it is a cyclophoracean. Clench (1949, pp. 18-23) described and figured the three Fijian species and summarized previously pub- lished data on the New Hebridean representatives. My study of 250 New Hebridean specimens indicated that only one species is found on the islands. Gonatoraphe fornicata (Pfeiffer). Plate 6, figs. 9, 10; plate 9, figs. 4-6; plate 26, figs. 1, 2. Cyclostoma (Cyclophorus) fornicatum Pfeiffer, 1854, Proc. Zool. Soc. London, 1852: 146— New Hebrides (Cuming). Cyclostoma (Cyclotus) recluzianum Pfeiffer, 1854, Proc. Zool. Soc. London, 1853: 51 — Solomon Islands (Cuming) (corrected to New Hebrides by Brazier, 1871, p. 587). Cyclostoma fornicatum Pfeiffer, 1854, Conch. Cab., 1, 19, p. 376, pi. 49, figs. 6-8. CyclostoTna (Cyclotus) ma^gillivrayi Pfeiffer, 1855, Proc. Zool. Soc. London, 1855: 103-104— Aneiteum (MacGillivray). Cyclophorus fornicatus Pfeiffer, 1861, Conch. Icon., Cyclophorus, pi. 18, fig. 86a, b; Cox, 1868, Exchange List, p. 49, no. 196— New Hebrides. Cyclotus recluzianus Pfeiffer, 1863, Conch. Icon., Cyclotus, pi. 9, fig. 53; Cox, 1868, Exchange List, p. 49, no. 193; Brazier, 1871, Proc. Zool. Soc. Lon- don, 1871 : 587 — restriction of type locality to Dillon's Bay, Erromanga. Cyclotus macgillivrayi PfeifTer, Cox, 1868, Exchange List, p. 49, no. 192 — Aneiteum; Brazier, 1871, Proc. Zool. Soc. London, 1871: 587 — Aneiteum. Cyclotus charmian Hutton, 1883, Trans. New Zealand Inst., 16: 209 — Horo- kiwi, Wellington, New Zealand (error); Suter, 1913, Man. New Zealand Moll., p. 684 — synonjmiizes with m,acgillivrayi PfeifTer. Cyclophorus vieillardi Mabille (not Gassies), 1895, Bull. Soc. d'Hist. Nat. d'Autun, 8: 401 — New Hebrides (Francois). Osiodes fornicata fornicata (PfeifTer), Sykes, 1903, Proc. Malac. Soc. London, 5: 199— Vila, Vate. Ostodes fornicata var. mncgillivrayi (PfeifTer), Sykes, 1903, Proc. Malac. Soc. London, 5: 199 — Vila, Vate and Renee River, Espiritu Santo. Gonatoraphe fornicata fornicata (PfeifTer), Clench, 1949, Bull. B. P. Bishop Mus., 196: 19-20. Gonatoraphe fornicata macgillivrayi (Pfeiffer), Clench, 1949, Bull. B. P. Bishop Mus., 196: 20. Range. — Aneiteum, Erromanga, Vate, Espiritu Santo. Material.— ML 59, ML 70, ML 72, ML 76a; Erromanga (USNM 515694, ex Chamberlain, Calvert; ANSP 13398, ex Wilstach); Anei- 184 FIELDIANA: ZOOLOGY, VOLUME 43 teum (ANSP 132659, Brazier!); Vila, Vate (ANSP 133299, Froggatt!; USNM 598355, ex W. B. Miller; Miller 108); Vate (UMMZ 74393, ex Walker, Ponsonby) ; Vate and Aneiteum (UMMZ 74391, ex Walker, Ponsonby, Layard) ; Renee River, Espiritu Santo (UMMZ 74390, ex Walker, Ponsonby, Sykes, J. J. Walker!); New Hebrides (UMMZ 74389, ex Walker, Ponsonby; CNHM 32431, ex Webb, Sowerby and Fulton). Remarks. — The conclusion of Sykes (1903, p. 199) that the three named "species" are inseparable is confirmed by the type photo- graphs (pi. 9, figs. 4-6). As was suspected from the original measure- ments, fornicata (pi. 9, fig. 6) is based on very juvenile specimens, which have a much sharper keel than adult material, and variety recluziana is also a juvenile (pi. 9, fig. 4), with only a slightly less prominent keel than fornicata. Most adult specimens are close to macgillivrayi (pi. 9, fig. 5), except that gerontic individuals (pi. 26, figs. 1, 2) have a much higher spire and a somewhat deflected aper- ture. Typologically, then, fornicata, recluziana and macgillivrayi rep- resent growth stages, rather than different taxonomic entities. The populations from Aneiteum and Erromanga are not separable, although Brazier (1871, p. 587) considered that they were different. The specimens differed only in size and whorl count, since the sculpture, whorl increase, height of spire, umbilicus and degree of carination showed the same variations in each population. Speci- mens from Vate were very small (diameter 7.1 mm., 43^8 whorls), those from Aneiteum and Erromanga were intermediate (diameter 9.5 mm., 4}^ whorls), and those from Espiritu Santo (Lot ML 72, collected alive) were large (diameter 11.1 mm., 4J4 whorls). The size and whorl count differences are probably not taxonomically impor- tant, but are the result of differing ecological conditions. Land operculates are quite sensitive to lack of moisture, and examples of stunted populations in ecologically unfavorable areas are numerous (see van der Schalie, 1948, pp. 26-30). The sculpture varied in the prominence, spacing, and number of lirae as well as in the number of longitudinal plications. In the latter respect, gerontic shells from Espiritu Santo (pi. 26, figs. 1, 2) were quite similar to Gonatoraphe stricta (Garrett) from Fiji. The sculp- ture was not constant within a lot, nor could height of spire and umbilical variations be used to separate populations. Several animals of G. fornicata were dissected. Male and female genitalia are figured (pi. 6, figs. 9, 10). In preserved specimens, the SOLEM: MOLLUSCA OF THE NEW HEBRIDES 185 reflexed portion of the female oviduct was bound to the seminal re- ceptacle by connective tissues and the bursa was appressed against the kidney. The organs are shown in unnatural positions to indicate their relative sizes and the points at which they connect with each other. The diagram of the external male organ shows the position of the verge with the mantle slit. The dotted line indicates the path of the ejaculatory duct. Several radulae were examined and found to be the same as that of G. stricta (Garrett) figured by Clench (1949, p. 48, fig. 28c). The teeth have not been figured here. Genus GASSIESIA Clench, 1949 Shell depressed-turbinate, solid, widely umbilicate. Sculpture of weak spiral threads; some species with an axial sculpture of growth lines. Operculum horny, multispiral, with depressed nucleus. Male with medially placed dorsal verge. (Modified from Clench, 1949, p. 5.) Type species. — Cyclostoma artense Montrouzier. Remarks. — The information on the position of the verge is taken from Fischer and Crosse (1880-1902, p. 137) and refers to G. mon- trouzieri (Souverbie). One species, Gassiesia forhesianus, was de- scribed from "Lord Howe's Island, New Hebrides." The species probably is based on a mislabeled New Caledonian shell. Gassiesia forhesianus (Pfeiflfer) Cyclostoma (Cyclophorus) forhesianus Pfeiffer, 1856, Proc. Zool. Soc. London, 1855: 104 — Lord Howe Island, New Hebrides (MacGillivray!) (error). Cyclophorus forhesianus Pfeiffer, 1858, Monog. Pneum. viv., II, p. 63; Pfeiffer, 1861, Conch. Icon., Cyclophorus, sp. 72. Ostodes forhesianus (Pfeiffer), Kobelt, 1902, Das Tierreich, 16: 155 — Santa Cruz Islands. Gassiesia forhesianus (Pfeiffer), Clench, 1949, Bull. B. P. Bishop Mus., 196: 6-7. Range. — New Caledonia or Isle of Pines(?). Material— New Hebrides (UMMZ 85670, ex Walker, Ponsonby; ANSP 12860, ex Swift, Cuming, cotypes?). Remarks. — Clench (1949, p. 6) tentatively equated forhesianus with the New Caledonian hocageanus Gassies. Probably this is cor- rect, but until types can be compared and adequate collections made in New Caledonia and the Santa Cruz Islands, the type locality and identity oi forhesianus will remain uncertain. The operculum of the ANSP cotypes(?) is like the New Caledonian Gassiesia rather than the opercular type of Gonatoraphe. 186 FIELDIANA: ZOOLOGY, VOLUME 43 Family PUPINIDAE The two subfamilies, Pupininae and Pupinellinae, are easily sep- arated by differences in anatomy (see Tielecke, 1940, p. 366) and the dull, radially sculptured pupinellids are quite different from the smooth, glossy shells of the pupinids. Pupinellids range from Asia to the Louisiades and Australia, but do not reach the Solomons. The sub- family Pupininae reaches its eastward limit in the New Hebrides and Fiji. Subfamily PUPININAE Species assignable to the Pupininae are found in India, China, Java, and the Philippines, and throughout most of Micronesia and Melanesia. Apparently they do not reach the Marianas, Polynesia, New Caledonia, New Zealand, ^ or Australia south of northern New South Wales (fig. 14). Numerous generic and subgeneric names, based on the size, shape, and number of apertural notches, have been proposed (see Kobelt, 1902b, pp. 301-336; Iredale, 1937a, pp. 295- 299; and Clench, 1949, pp. 30^8). Similar apertural configurations are found in specimens from widely separated regions, but it is un- known whether these variations are phyletic or convergent. Although lack of both material and time has prevented more than cursory review of the classification, I have restudied the Melanesian and Micronesian species and have tentatively subordinated Iredale's gen- era to previously known taxa. Dr. William J. Clench has been most helpful in permitting me to study the extensive series in the Museum of Comparative Zoology and in making suggestions about this dif- ficult group. After separating a few specialized taxa — Hargravesia, Moulinsia, Porocallia, Callianella, and Pupinoa — most workers divide the re- maining Pupininae into two large and widely distributed genera, Pupina and Tylotoechus. As Clench (1949, p. 44) stated, many spe- cies are easily placed, but others are intermediate. Without an exten- sive revision, their true relationship is uncertain. For convenience, I have here maintained the historical status quo. The New Hebri- dean species belong to Pupina. Acting on his usual assumption that any Australian species is, ipso facto, generically different from any non-Australian species. Ire- dale (1937a, pp. 295-299) created several new "genera" for the Queens- land and New South Wales pupinids. All seem closely related and ' Liarea (or Realia) from New Zealand (see Powell, 1954) has been placed in the Pupininae, but its affinities are uncertain. SOLEM: MOLLUSCA OF THE NEW HEBRIDES 187 are probable derivatives of Pupina (sens. str.). The Australian spe- cies are considered to be a subgenus of Pupina, called Signepupina Iredale, 1937. Of Iredale's other "genera," Estopupina and Necopu- pina are probably sections of some pupinelline genus, while Lopupina, Diplopupina, Parpupina and Dolopupina represent evolutionary de- velopments from Signepupina of not more than sectional value. Genus PUPINA Vignard, 1829 Species from the Philippines (gracilis Moellendorff) and Australia (pfeifferi Dohrn) are similar enough to the genotype, P. keradreni Vignard, to be included in the same genus. The Australian species belong to a subgenus, Signepupina, while typical Pupina includes the island species from the Philippines to the Fijis. Subgenus PUPINA (sens, str.) Although a few Philippine Island species probably belong to Pupina (sens, str.), this discussion is limited to the Micronesian and Melane- sian species. The species of New Guinea and northern Melanesia have both a well-developed parietal tooth and a columellar notch. In the island chains fanning out from this core, it is possible to trace three parallel lines of development. In the Palau Islands, P. dijffi- cilis Semper is quite close to P. keradreni, but P. hrenchleyi E. A. Smith from Lukunor and P. complanata (Pease) from Kusaie in the Carolines show a gradual reduction in the size of the parietal notch. The Fijian species (Pupinesia Clench, 1949, p. 39) have lost the parietal notch, and the columellar notch has been considerably mod- ified. The New Hebridean species (Kanapa Clench, 1949, p. 41) lack a prominent parietal notch, although retaining the columellar notch. Pupinesia and Kanapa represent obvious offshoots of Pupina only slightly more differentiated than the Palau-Caroline Island series. Their relationships are much more clearly shown by making Kanapa and Pupinesia sections of Pupina rather than distinct genera. Section KANAPA Clench, 1949 Similar in general shell outline to Pupina, s. s., but differs in lack of a parietal notch and parietal tooth. Columellar notch is small but enlarged somewhat be- neath lip margin (Clench, 1949, p. 41). Type species. — Registoma brazieri Crosse. Remarks. — Of the three species Clench placed in Kanapa, P. bra- zieri and P. cumingiana are from the New Hebrides; the third, 188 FIELDIANA: ZOOLOGY, VOLUME 43 P. manni (Clench), is from Three Sisters Island in the Solomons. Probably it is only a local development from Pupina keradreni Vign- ard, since the columellar structure of manni is like that of keradreni rather than that of the New Hebridean Kanapa. Table XIII. — Size Variation in the New Hebridean Pupina Height Diameter H/D Mean Range Mean Range Mean Range P. cumingiana 9.1 8.0-9.7 3.7 3.3-4.1 2.4 2.2-2.7 (MCZ 140552) P. cumingiana 8.3 8.0-8.6 3.6 3.3-3.9 2.3 2.2-2.5 (ANSP 13167) P.brazieri 6.0 5.8-6.3 3.0 2.9-3.2 2.0 1.9-2.1 P. brazieri has no trace of a groove at the parietal angle, while P. cumingiana has either a weak groove or a prominent notch. Generally, brazieri is much smaller (5.8-6.3 mm.) than cumingiana (8.0-9.7 mm.), but a few dwarf specimens of cumingiana (6.5-9.7 mm.) were seen. Pupina (Kanapa) brazieri (Crosse) Registoma complanatum Cox (not Pease), 1868, Exchange List, p. 50, no. 210 — Erromanga. Registoma brazieri Crosse, 1870, Jour, de Conchy., 18: 250-251 — Erromanga; Crosse, 1871, Jour, de Conchy., 19: 321-322, pi. 13, fig. 6. Pupina (Pupina) brazieri (Crosse), Kobelt, 1902, Das Tierreich, 16: 302. Kanapa brazieri (Crosse), Clench, 1949, Bull. B. P. Bishop Mus., 196: 42-43, figs. 23fe, 24a. Range. — Erromanga, Espiritu Santo. Material.-~ML 95; Erromanga (MCZ 14109, ex Beddome, Bra- zier; MCZ 141008, ex Pease, Brazier; UMMZ 87408, ex Walker, Ponsonby, Beddome, Brazier) ; New Hebrides (ANSP 13177, ex Wil- stach; UMMZ, ex Walker, Ponsonby; MCZ 141010, ex Dohrn; MCZ 141011). Remarks. — The small (5.8-6.3 mm.) size and compressed spire of P. brazieri at once separate it from P. cumingiana. Unlike the latter species, P. brazieri had no trace of an apertural notch in any of the specimens I saw. The Espiritu Santo shells are inseparable from the syntypes in the Museum of Comparative Zoology. The holotype and four paratypes are preserved in the collection of the Journal de Conchyliologique in Paris (see Jour, de Conchy., 90: 77). SOLEM: MOLLUSCA OF THE NEW HEBRIDES 189 Pupina (Kanapa) cumingiana Pfeiflfer. Plate 9, fig. 8. Pupina (Pregistoma) (sic) cuminffiana Pfeiffer, 1853, Proc. Zool. Soc. London, 1853: 52 — Solomon Islands (error). Registoma cumingiana (Pfeiffer), 1858, Monog. Pneum. viv., II, p. 97 — Lord Howe's Island (MacGillivray) (error); Cox, 1868, Exchange List, p. 50, no. 211. Registoma cumingiana minor Pfeiffer, 1865, Monog. Pneum. viv., suppl. 3, p. 97 — New Caledonia (error). Pupina (Registoma) cumingianum Pfeiffer, 1876, Monog. Pneum. viv., suppl. 4, p. 152— Tanna. Pupina {Pupina) cumingianum Pfeiffer, Kobelt, 1902, Das Tierreich, 16: 303. Pupina cumingiana Pfeiffer, Sykes, 1903, Proc. Malac. Soc. London, 5: 200 — Resolution Bay, Tanna (J. J. Walker). Kanapa cumingiana (Pfeiffer), Clench, 1949, Bull. B. P. Bishop Mus., 196: 43-44, fig. 23c. Kanapa cumingiana minor (Pfeiffer), Clench, 1949, op. cit., p. 44. Range. — Tanna. Mafma/.— Tanna (UMMZ 87824, ex Walker, Ponsonby; MCZ 140551, ex Bishop Museum, Brazier, MacGillivray, cotypes; MCZ 140552, ex Bishop Museum, E. Robertson! July, 1925; MCZ 165719, ex Boston Soc. Nat. Hist., Geale; ANSP 13167, ex Brazier); New Hebrides (UMMZ 161465, ex Walker; UMMZ 87507, ex Walker, Ponsonby) ; erroneous localities (UMMZ 13155, ANSP 13165, ANSP 13166, ANSP 13164, ANSP 13163). Remarks. — Most specimens of cumingiana are the size and shape of the holotype (pi. 9, fig. 8), but those collected by Robertson in 1925 (MCZ 140552) are much more elongate. Comparative measurements of cumingiana and brazieri are given in Table XIII. All specimens of cumingiana have an internal parietal groove but show a great variation in the extent of development of the parietal notch. In the holotype and about one-third of the specimens, the notch is absent. In others it is partially developed, and in one speci- men (UMMZ 87824) the notch is as fully developed as in P. compla- nata Pease (UMMZ 87501, 87502, 87503 and 87504; see also Clench, 1949, p. 34, fig. 20c). The "notched" ponsonbyi and the "notchless" cumingiana intergrade fully and cannot be separated. The presence of a variety of cumingiana with a parietal notch weakens the validity of Kanapa, but it can still be retained as a useful indicator that the New Hebridean species are different from, although closely related to, the Solomon Island Pupina. A lot (UMMZ 87507) of dwarf P. cumingiana contained one small specimen 6.5 mm. in height and five others ranging from 7.1-7.4 mm. 190 FIELDIANA: ZOOLOGY, VOLUME 43 Family DIPLOMMATINIDAE The family group name Diplommatinidae must replace the name Cochlostomatidae (Tielecke, 1940, p. 364), since the former dates from L. Pfeiffer in 1856 (Malak. Blatt., 3: 118). Two subfamilies are usually recognized, the European Cochlostomatinae and the In- do-Pacific Diplommatininae. Possibly the Central American Adelo- poma should be placed with the Diplommatininae. Of the many genera in the Diplommatininae (fig. 17), only Pal- aina, Diplommatina, Diancta, Pseudopalaina, and the monotjrpic Hun- gerfordia reach the islands of the Pacific, excluding the Philippines and Indonesia. Palaina occurs in the Palau, Caroline, and Mariana Islands of Micronesia, all of Melanesia, Lord Howe and Norfolk Is- lands, and east to Tonga; Diplommatina in Fiji, Samoa, and the Solomons; Diancta in Fiji; and Pseudopalaina and Hungerfordia in the Palau Islands. All genera are based on shell characters only. Palaina includes species without a parietal tubercle and Diplommatina species with a parietal tubercle. Some species can be placed in either genus (Solem, in press-A), but the New Hebridean species are unquestionably Palaina. Excellent figures showing the variation found in the Dip- lommatininae are given by Zilch (1953). ( \ Genus PALAINA 0. Semper, 1865 Type species. — Diplommatina macgillivrayi Pfeiffer by subsequent designation of Iredale (1944, p. 303). Remarks. — Several subgeneric names, based on shell contour, have been proposed. Probably these names identify convergent evo- lution, since similarly shaped shells are found in all parts of the range of the genus. A list of these nomenclatural units is included for future reference. All type species are by original designation. Eupalaina Kobelt and Moellendorflf, 1898 An absolute synonym of Palaina (sens. str.). Cylindropalaina Moellendorflf, 1897 Type species. — Palaina chrysalis Moellendorflf. Macropalaina Moellendorflf, 1897 Type species. — Palaina pomatiaeformis Mousson. Eclogarinia Iredale, 1933 Type species. — Diplommatina gowllandi Brazier. Velepalaina Iredale, 1937 Type species.— Diplommatina oreadis Hedley. SOLEM: MOLLUSCA OF THE NEW HEBRIDES 191 Palmatina Iredale, 1944 Type species. — Diplommatina coxi H. Adams. Fermepalaina Iredale, 1945 Type species. — Palaina nancena Iredale. Many papers have appeared on the Austro-Melanesian Palaina. The important sources for information are: Australia: Iredale, 1937a, pp. 300-301. Bismarcks: I. and B. Rensch, 1929, pp. 84-86; I. Rensch, 1937, pp. 602-605. Solomon Islands: Dell, 1955, p. 425; Solem, in press-A. Fijis: Mousson, 1870, pp. 180-189. New Caledonia: Crosse, 1894, pp. 381-383; Hedley, 1898, pp. 102-104; Cockerell, 1930; Franc, 1957, pp. 40-i2. Lord Howe Island: Iredale, 1944, pp. 303-307. Norfolk Island: Iredale, 1945, pp. 49-50. Specimens of most of the species were examined (UMMZ and ANSP). It is not possible to recognize phyletic groups, but certain species are very similar in shape. Thus, the following have a constricted body whorl and flaring aperture: P. martensi H. Adams, 1866: Fiji (UMMZ 87918, ex Walker, Ponsonby, Bednall) ; Tongatabu, Tonga (UMMZ 87919, ex Walker, Ponsonby, Schliiter), a new locality record. P. distorta Mousson, 1869: Viti Levu, Fiji (ANSP 13235, ex Godef- froy); Suva Bay, Viti Levu (UMMZ 88697, ex Walker, Ponsonby). Palaina, new species: Florida, Solomon Islands (UMMZ 181755, ex Kuntz!). P. schneideri I. and B. Rensch, 1929: New Britain. Three species have a V-shaped spire with shallow sutures and the body whorl not constricted (Macropalaina) : P. pomatiaeformis Mousson, 1870: Fijis.' P. gardneri Dell, 1955: Treasury Island, Solomon Islands. P. perroquiniana Crosse, 1871 : New Caledonia (ANSP 64179, ex Fulton; UMMZ 87934, ex Walker, Ponsonby, Beddome). Three species are cylindric with shallow sutures (Cylindropalaina) : P. subregularis Mousson, 1870: Fiji. P. mariei Crosse, 1867: New Caledonia (UMMZ 89912, ex Walker, Tomlin). 192 FIELDIANA: ZOOLOGY, VOLUME 43 P. montrouzieri Crosse, 1874: New Caledonia (UMMZ 87922, ex Walker, Wetherby; UMMZ 87923, ex Walker, Ponsonby, Marie); Noumea, New Caledonia (ANSP 13233, ex Dupuy); Bourail, New Caledonia (ANSP 150987, ex Cockerell!). P. m. humilior Cockerell, 1930: Bourail, New Caledonia (ANSP 150986, ex Cockerell, paratypes). The remaining species from the Bismarcks, Fiji, the Solomons, Lord Howe and the New Hebrides have swollen spires, deep sutures, and slightly constricted body whorls. The Norfolk Island Palmatina (Iredale, 1945, pp. 49-50) are characterized by a solute body whorl. The resemblances noted above may be caused by convergence, but the recognition of these "species groups" is useful if for no other reason than to facilitate identifications. Specimens of New Hebridean Palaina were available only from Espiritu Santo. Two species are here recognized : one with the radial ribs widely spaced on the upper spire and closely spaced on the lower spire; the other with the ribs only slightly less widely spaced on the lower spire and body whorl. The larger, more evenly ribbed species is described as P. sykesi; the smaller species with crowded ribs on the body whorl is identified as P. frariQoisi Ancey. n Table XIV. — Size Variation in New Hebridean Palaina P. sykesi P. franQoisi Males Females Males Females No. of specimens 3 10 48 21 Height Mean 3.58 4.22 2.35 2.89 Range 3.30-3.86 4.07-4.45 2.10-2.61 2.70-3.17 S.D. .... 0.12 0.13 0.10 Diameter Mean 1.83 2.14 1.17 1.49 Range 1.71-1.93 2.01-2.27 1.06-1.29 1.37-1.59 S.D. .... 0.089 0.068 0.049 H/D Mean 1.95 1.97 1.99 1.94 Range 1.93-2.00 1.83-2.22 1.82-2.15 1.81-2.06 S.D. .... 0.04 0.083 0.043 Whorls 6M-6% 6^-7 6-6^ 6-6i^ Within each sculptural type, there were two size classes (see Table XIV). No preserved specimens were available, but probably the differences reflect sexual dimorphism. In most land prosobranchs the females are considerably larger than the males (H. B. Baker, 1925b, p. 274). Pending study of the soft parts, I have assumed that the larger shells of each species came from females, the smaller from males. SOLEM: MOLLUSCA OF THE NEW HEBRIDES 193 Palaina frangoisi Ancey. Plate 27, figs. 3, 4. Palaina franQoisi Ancey, 1905, Nautilus, 19, (4), pp. 44-45 — Espiritu Santo, New Hebrides. Range. — Espiritu Santo. Material— ML 39 (1 male, 1 female), ML 51 (1 male), ML 78 (2 females), ML 95. Remarks. — The unique example of Ancey's species may be in Brussels, Belgium, but it was not seen or located during this study. All of Ancey's Espiritu Santo shells came from the region in which Kuntz collected. Since the original measurements (height 2.75 mm., diameter 1.5 mm., whorls 63^) fall into the range of variation shown by "female" specimens of the smaller species, /rawfom has been iden- tified with that population. Of the extralimital species, P. frangoisi is most similar to P. gode- froyana (Mousson) and P. g. latecosta (Mousson) from Fiji. The two latter species differ in having deeper sutures, more evenly spaced rib- bing, and more rounded whorls. Palaina sykesi, new species. Plate 27, figs. 1, 2. A species of Palaina characterized by its subcylindric shape, shal- low sutures, evenly spaced radial ribs, and only slightly constricted body whorl. The only other New Hebridean species, P. frangoisi Ancey, is smaller and has more closely spaced ribs on the body whorl. Shell sinistral, subcylindric, sutures well marked but not deeply impressed. Spire increasing regularly in size, distinctly asymmetrical, skewed to the right. Aperture ovate, lip moderately expanded, closely appressed to the penultimate whorl. Umbilicus closed. Apical whorls smooth, spire with widely spaced, slightly protractive ribs, gradually becoming closely spaced on body whorl. Fresh speci- mens with microscopic spiral lines. Operculum not seen. Holotype, 2.14 mm. in diameter, 4.28 mm. high, with 6H whorls. Allotype, diameter 1.84 mm., height 3.66 mm., with 6% whorls. Type. — University of Michigan Museum of Zoology no. 186158. Collected from stream drift in the Sarakata River Valley, Espiritu Santo, New Hebrides (ML 95), by Robert E. Kuntz in May and June, 1944. Allotype. — UMMZ 186159. From the same locality as the type. Paratypes. — Specimens from the type locality are UMMZ 186160, BPBM, MCZ, and CNHM 62236. Remarks. — Only drift specimens were available. In size and shape P. sykesi resembles some of the Bismarck species (I. and B. Rensch, 1929, pp. 84-86) but differs from them by having a circular aperture, 194 FIELDIANA: ZOOLOGY, VOLUME 43 shallower sutures, a less constricted body whorl, and different micro- sculpture. The Lord Howe Island Palaina (Iredale, 1944, pp. SOS- SOT) are more obese and have a more rhomboidal aperture and more crowded ribbing. Palaina sykesi is named for the late Ernest Ruthven Sykes, who published papers on New Hebridean non-marine Mollusca in 190S and 1904. Superfamily RISSOACEA Family HYDROBIIDAE American authors have used the family name Amnicolidae and European authors the name Hydrobiidae. The latter has probable priority (teste Morrison) and has been adopted here.^ Thiele (1928b and 1929, pp. 1S6-156) and Wenz (1938-44, pp. 555-581) pubhshed comprehensive classifications based primarily on radular structure. Recent studies by Berry (194S) and Morrison (1949) have suggested that the verge is an important aid to understanding hydrobiid rela- tionships, but in only a few species have the vergic structures been described. Following Pilsbry (1948, p. 1065) and Clench and Turner (1948), I have listed the Truncatellidae as a separate family. One hydrobiid, Fluviopupa brevior (Ancey), is found in the New Hebrides. Fluviopupa was established by Pilsbry (1911) for a Fijian species, and the genus has been reviewed by Hubendick (1952). In accordance with the classification of Thiele, Fluviopupa is here as- signed to the subfamily Hydrobiinae, tribe Littoridineae. Genus FLUVIOPUPA Pilsbry, 1911 Shell minute, pupiform, with obtuse apex and only slightly rounded whorls. Aperture ovate, vertical or sloping forward below, peristome completely or almost completely free from upper portion of body whorl. Operculum thin, horny, pauci- spiral, with subcentral nucleus. Radula typically amnicolid, central (4-5)-l- (4-5) / (2-4)-(2-4), lateral with 7-12 cusps, marginals with about 30 minute cusps. Penis with single duct, tip simple or bilobed, some species with a median bulb. (Modified from Pilsbry, 1911, p. 549, and Hubendick, 1952.) Type species. — Fluviopupa pupoidea Pilsbry. ' Opinion 475 of the International Commission on Zoological Nomenclature (issued July 31, 1957) validates the name Bithynidae of Gray, 1857. Whether or not Bithynia {=Bulimus Scopoli, 1777) and Hydrobia belong to the same family is uncertain. In any case, I prefer to use the more familiar name Hydrobiidae, regard- less of the exact relationship of the two genera. SOLEM: MOLLUSCA OF THE NEW HEBRIDES 195 Remarks. — Hubendick (1952) summarized previous information on Fluviopupa, described two species from Rapa, and indicated a number of probable relatives from the Pacific Ocean. My study of specimens (UMMZ) has enabled me to make several additions to Hubendick's survey. Iredale (1944, p. 332) proposed names for many Fluviopupa from Lord Howe Island and divided them into two genera, Fluviorissoina and Pupidrobia, without giving illustrations or adequate descriptions. The original study on Lord Howe and Norfolk Island mollusks was done by H. B. Preston, who carefully selected "types" and distrib- uted many new species under manuscript names. Iredale (1944, 1945), working from a series of "cotypes" and the general collection in the Australian Museum (Sydney), published reviews of the Lord Howe and Norfolk fauna. He used most of Preston's manuscript names as his own, but he changed a few names and reported on addi- tional material. Preston's collection was purchased by Bryant Walker and is now in the University of Michigan Museum of Zoology. All of Iredale's species were represented by shells and operculi. No soft parts were available, but Fluviorissoina and Pupidrobia show no conchological criteria justifying generic separation from Fluviopupa. The New Hebridean Fluviopupa brevior is most similar in shell structure to the Fijian F. pupoidea, but the latter is larger and more cylindric, and has a proportionately smaller aperture. Fluviopupa brevior (Ancey). Plate 6, fig. 11; plate 27, figs. 5, 6. Potamopyrgus brevior Ancey, 1905, Nautilus, 19, (4), p. 46 — Vate Island, New Hebrides. Potamopyrgm sp. J. R. Baker, 1929, Man and Animals in the New Hebrides, p. 148 — Steaming Hill Lake, Gaua, Banks Group. Range. — Vate, Espiritu Santo, Gaua. Material.— ML 69, ML 78, ML 79, ML 85b, ML 87, ML 95. Remarks. — The location of the unique specimen is unknown to me. Layard mentioned a "Hydrobia" from Vate in an invoice of specimens sent to Hartman, but the specimens could not be found in the Carnegie Museum collection. Little would be gained by describ- ing the Espiritu Santo population as a new species, simply because they are from another island, and I have applied the name brevior to the entire complex. I examined about 280 specimens, half of them preserved in alco- hol. The variation in a small sample is summarized in Table XV. The males were generally smaller than the females, but the sizes over- 196 FIELDIANA: ZOOLOGY, VOLUME 43 lapped and size alone cannot separate male and female specimens. Destruction of the shell was necessary in order to get at the retracted animal. Fifteen individuals were dissected. The males were much more heavily pigmented than the females. Mr. Harold J. Walter of the University of Michigan Museum of Zoology informs me (personal communication) that in the North American hydrobiid, Pomatiopsis, the female is very dark and the male very light in color. Table XV. — Size Variation in Fluviopupa and Omphalotropis F. brevior O. setocincta O. poecila 0. conella Lot no ML 69 ML 40 ML 31d CM & UMMZ No. of specimens 27 29 79 6 Height Mean 2.08 4.81 6.94 4.29 Range 1.71-2.40 4.47-5.23 5.56-8.55 3.87^.97 S.D. 0.14 0.20 0.81 0.35 Diameter Mean 1.23 3.22 5.72 3.22 Range 1.03-1.41 2.88-3.65 4.70-6.81 2.86-3.38 S.D. 0.08 0.16 0.64 0.22 H/D Mean 1.69 1.50 1.21 1.37 Range 1.56-1.79 1.37-1.59 1.10-1.33 1.32-1.47 S.D. 0.06 0.053 0.048 0.058 Whorls 5-5^ 5M-6M 53^-6 Without destroying the shells, it was impossible to establish sex ratios or the degree of sexual dimorphism. In the field notes, Kuntz records that dicrocoeliid cercariae were recovered from the animals. Fluviopupa brevior was collected alive in quiet fresh-water pools of the Sarakata River on rocks, twigs, and leaves. Both Physastra layardi and Gyraulus montrouzieri were collected at the same stations. There can be no question but that Fluviopupa brevior is a true fresh- water snail and not an estuarine species. The radula of F. brevior did not differ from those figured by Huben- dick (1952). The penis, however, showed an important modification (pi. 6, fig. 11). As in the other species, there is a single duct in the verge, but the tip in brevior is distinctly bilobate. In this respect, brevior more closely resembles the Rapan shells than the Fijian. I saw no significant variation in the shape of the verge in the eight males dissected. In the time available, it was not possible for me to undertake a study of the internal genital anatomy. SOLEM: MOLLUSCA OF THE NEW HEBRIDES 197 Family TRUNCATELLIDAE Clench and Turner (1948) reviewed current classification and pre- sented a nomenclatural catalogue of the Truncatellidae. Pilsbry (1948, pp. 1065-68) summarized what little is known of their ecology and method of locomotion. Truncatellids live in scattered colonies among the debris just above high-water mark. They are thus quite well situated for me- chanical dispersal by storms and currents. Most of the numerous named forms will probably be found to be variants of a few widely distributed species. Quoy and Gaimard (1832-35, p. 186, pi. 12, figs. 27-30) described Cyclostoma striata from Vanikoro Island. The specimens were sub- sequently lost (Clench and Turner, 1948, p. 207), and the original figures and description are unrecognizable. Probably C. striata was a truncatellid, but since the types are destroyed and the name unrec- ognizable, striata must be considered a nomen dubium. The only species definitely known from the New Hebrides is the extremely common, wide-ranging Truncatella guerinii A. and J. B. Villa. Cox (1868, p. 48, no. 181) reported a Japanese species, T. pfeifferi von Martens, from Erromanga, but the record needs confirmation, Truncatella (Truncatella) guerinii A. and J. B. Villa. Plate 27, fig. 7. Truncatella guerinii A. and J. B. Villa, 1841, Conchy, terr. fluv. Mediolani, p. 59 — Reunion. Truncatella valida PfeiflFer, 1856, Zeitschr. f. Malak., 3: 182; Cox, 1868, Ex- change List, p. 48, no. 103 — Aneiteum; Sykes, 1903, Proc. Malac. Soc. London, 5: 199 — Resolution Bay, Tanna (J. J. Walker). Truncatella pacifica Pease, Mabille, 1895, Bull. Soc. d'Hist. Nat. d'Autun, 8: 401 — New Hebrides (Francois). Truncatella (Truncatella) guerinii A. and J. B. Villa, Clench and Turner, 1948, Occ. Pap. Mollusks, 1, (13), pp. 167-169, pi. 23, figs. 12-13. Range. — Aore, Vate, Aneiteum, Tanna, Erromanga(?). Extralim- itally from Africa to Japan, New Caledonia, and the Society Islands. Material— kore (MCZ 192072, W. L. Nutting) ; "Woody Island," New Hebrides (ANSP 12617, ex Cox); Vate (DMNZ, W. H. Dawbin!). Remarks. — Colonies of T. guerinii will probably be discovered on the shore line of every New Hebridean Island. 198 FIELDIANA: ZOOLOGY, VOLUME 43 Family ASSIMINEIDAE (= Realiidae, Synceridae of authors) There is some confusion as to the correct name to be used for this family. Iredale (1941, p. 59) believed that Realia must replace Omphalotropis, rather than being used for the New Zealand cyclo- phorid Ldarea (see Thiele, 1929, p. 104). If Iredale is correct, then the family name must be Realiidae Pfeiffer, 1858. Abbott (1949, p. 263) stated that there is no evidence for Iredale's conclusion that Gray's Figures of Molluscous Animals, volume 4, appeared before June, 1850. The introduction is dated February 12, 1850, but un- doubtedly it appeared several months later. Kobelt (1906, pp. 49, 121, 138) named the subfamilies Garrettinae and Adelomorphinae, both of which have priority over the Assimineidae and Omphalotro- pidinae of Thiele (1927, pp. 125-127), The systematic position of Kobelt's subfamilies is uncertain and I prefer to utilize Thiele's name. Several authors substituted Syncera Gray, 1821, for Assiminea Flem- ing, 1828. Iredale (1922) showed that Syncera is a nomen nudum and thus neither Syncera nor the Synceridae can replace Assiminea or the Assimineidae. Kobelt (1906), Thiele (1927, 1929, pp. 168-173) and Wenz (1938- 44, pp. 631-640) presented synoptic classifications of the Assimineidae. Recently, I. Rensch (1937, pp. 614-625), Cooke and Clench (1943), Kondo (1944), Clench (1946, 1948, 1955), and Abbott (1949, 1958) published papers on some of the Pacific Island species. The New Hebridean species were reviewed by Ancey (1906). I could come to no conclusions about the distribution and phy- logeny within the family. Radular studies show that very similar shells belong to widely divergent genera, few genotypes have had their soft parts examined, and most species have not been adequately described or figured. Most of the species are referable to Omphalo- tropis, but a single shell from Vate Island (DMNZ) is here tentatively referred to Assiminea nitida (Pease). A. nitida has been widely dis- persed throughout most of Polynesia and has even been reported from Java (see van Benthem Jutting, 1956) . Its presence in the New Hebrides, if confirmed, may have resulted from accidental introduc- tion by human agency. Key to the New Hebridean Assimineidae 1. Shell with prominent spiral ridges; operculum with nearly central nucleus and calcareous granules on outside 2 Shell macroscopically smooth; operculum with acentric nucleus, horny 4 2. Diameter more than 4 mm.; umbilicus without keel 3 Diameter less than 4 mm.; umbilicus with a distinct keel. Omphalotropis (Lyrotropis) conella Sykes SOLEM: MOLLUSCA OF THE NEW HEBRIDES 199 3. Height at six whorls more than 9 mm O. (L.) annatonensis (Pfeiffer) Height at six whorls less than 8.5 mm O. (L.) poecila Ancey 4. Umbilicus with a keel 0. (Oriella) setocincta Ancey Umbilicus without a keel Assiminea (A.) nitida (Pease) Genus ASSIMINEA Fleming, 1828 Type species. — Assiminea grayana Fleming, 1828. Remarks. — Syncera Gray, 1821, is a nomen nudum (see Iredale, 1922, and Thiele, 1929) although Abbott (1949) and van Benthem Jutting (1956) considered it a validly described genus. Subsequently Abbott (1958, pp. 232-233) reconsidered his position and has accepted Assiminea. The shells of Assiminea and PaludineUa are practically inseparable, and only radular studies can determine the generic position of any particular morph. A single shell from Vate Island seems to belong to the complex usually called Assiminea nitida (Pease). The exact relationships of the forms grouped under this name remain to be de- termined (see Abbott, 1949) and the New Hebridean shell is only tentatively referred to the species nitida. Assiminea (Assiminea) nitida (Pease) Hydrocena nitida Pease, 1865, Proc. Zool. Soc. London, 1864: 674 — Huahine, Society Islands. Syncera nitida (Pease), Abbott, 1949, Occ. Pap. B. P. Bishop Mus., 19, (15), p. 272, fig. 7. Syncera nitida (Pease), van Benthem Jutting, 1956, Treubia, 23, (2), pp. 355- 356, fig. 68. Range. — Vate. Reported from Java, Mauritius, Ceylon, China and the Philippines east to the Society Islands by various authors. Material— At 800 feet elevation in forest, Vate Island (DMNZ, W. H. Dawbin!). Many Polynesian lots (UMMZ and CNHM). Remarks. — As emphasized above, any assimineid identifications based on single shells are apt to be questionable and the exact refer- ence of the Vate population must await an extensive revision of the Pacific "Assiminea" and '*Paludinella." Genus OMPHALOTROPIS Pfeiffer, 1851 Type species. — Cyclostoma aurantiaca Deshayes, by subsequent designation (Gude, 1921, p. 355). Remarks.— Iredale (1941, p. 59) and Abbott (1949, p. 262) cite Omphalotropis hieroglyphica Ferussac as genotype, but the first type 200 FIELDIANA: ZOOLOGY, VOLUME 43 designation is that of Gude, cited above. The typical section of the genus is found on the Mascarene Islands, but none of these species have had their soft parts examined. Thiele (1927, p. 127) recognized several conchological sections from Pacific Islands. Later, Cooke and Clench (1943), Iredale (1944, p. 301) and Clench (1946, 1948, 1955) described new generic units from the Pacific Islands. After examining specimens or illustrations of all the type species, I placed Omphalotropis setocincta Ancey in the section Oriella Thiele, 1927, and described below a new section, Lyrotropis, for the other species. Only dried specimens of Omphalotropis setocincta were available, but many preserved animals of 0. poecila were in the Kuntz collec- tions. Both species have the pectinate marginal teeth of the Omphalo- tropidinae. The verge (pi. 6, fig. 12) of 0. poecila is very similar to that of 0. cookei Abbott from the Marianas (Abbott, 1949, p. 265, fig. le) . I did not study further details of the soft parts, since com- parative material was lacking. Section ORIELLA Thiele, 1927 Shell very small, ovate-conic. Whorls rounded, with deep sutures, body whorl obtusely angled. Umbilicus narrow, with a strong keel. Aperture vertical, oval, angulated above. Lip thickened, columellar and basal margins slightly expanded. Surface macroscopically smooth with a microsculpture of fine pits and spiral lines. (Modified from Wenz, 1938-44, p. 636.) Type species. — Omphalotropis submaritima Quadras and Moel- lendorff. Remarks. — Study of specimens in the University of Michigan Museum of Zoology revealed several very similar "species" which belong to Oriella. Not enough material was available to determine whether they are specifically distinct or only one very wide-ranging species. Besides the genotype, 0. submaritima from the Marianas, 0. granum Pfeiffer (=maritima Montrouzier) from New Caledonia, 0. moussoni Pease from Fiji, and probably some of the other Steno- tropis (see Kobelt, 1906, pp. 85-93) are closely related to 0. seto- cincta Ancey. Omphalotropis (Oriella) setocincta Ancey. Plate 27, figs. 8, 9. Omphalotropis setocincta Ancey, 1890, Le Naturaliste, 1890: 26 — Vate (Glisson!). Omphalotropis (Eurytropis) setocincta Ancey, Kobelt, 1906, Jahrb. Nass. Ver. Naturk., 59: 82, 143. Range. — Vate, Espiritu Santo. Material.~ML 24, ML 40, ML 95; Vate Island (UMMZ 74254, ex Walker, Ponsonby, Boettger, Layard, Glisson, type lot specimens). SOLEM: MOLLUSCA OF THE NEW HEBRIDES 201 Remarks. — The living specimens of 0. setocincta were found under debris or in decaying coconuts relatively near the seacoast, but defi- nitely not under "strand line" conditions. The conchologically sim- ilar genus Assiminea (=Syncera) is an inhabitant of brackish water or intertidal areas (see van Benthem Jutting, 1956, p. 349) and Palu- dinella lives in fresh water (op. cit., p. 362). The terrestrial habitat of Omphalotropis is a definite contrast. Of the 150 specimens seen, 135 came from ML 40 and were found in decaying coconuts. The animals had been dried, but radulae were extracted and portions of verges examined in softened material. The radula is almost identical to that of 0. poecila (pi. 6, fig. 13) and the verge is slender near the base and situated mid-dorsally. In the study of softened specimens, I noticed that the shells of males were slenderer and had a proportionately narrower aperture than did those of females (see pi. 27, figs. 8, 9). No attempt was made to sex the vast majority of the dried animals in ML 40; the measurements of a sample of males and females are summarized in Table XV. As with Pleuropoma articulata (Pfeiffer), it is impossible to deter- mine the specific relationships of Omphalotropis setocincta. Both 0. setocincta and Pleuropoma articulata have extremely similar spe- cies on many Pacific islands. Insufficient material is available, how- ever, to decide whether one wide-ranging or several closely related species are involved. LYROTROPIS, new section Shell small, solid, trochiform. Whorls slightly flattened with moderately deep sutures, body whorl rounded or obtusely angulated. Umbilicus narrow, with or without basal keel. Aperture slightly protractive, circular, with columellar margin flattened. Lip thin, margin slightly expanded in adults. Surface with several to many prominent spiral lirae crossed by retractive radial riblets. Operculum horny, with only slightly acentric nucleus and varying amounts of calcareous granules deposited on the outer surface. Radula with pectinate marginals. Type species. — Omphalotropis poecila Ancey. Remarks. — The operculum of Lyrotropis is most similar to that of Rapanella (Cooke and Clench, 1943, p. 254, fig. 3) but differs in hav- ing a surface deposit of calcareous granules as in Omphalotropis cookei (Abbott, 1949, p. 265, fig. le). The shell form and sculpture of Lyro- tropis is most similar to Pseudocyclotus from the Solomons and Bis- marcks, but Pseudocyclotus has non-pectinate marginal teeth and a heavily calcified operculum. None of the Solomon Island (Solem, in press-A), Bismarck (I. Rensch, 1937, pp. 614-625), or New Caledonian (Franc, 1957, pp. 202 FIELDIANA: ZOOLOGY, VOLUME 43 50-52) Omphalotropis appear to be closely related to the New Heb- ridean species. Indeed, Lyrotropis is a very "isolated" group and, if I had followed most recent workers on Pacific assimineids, it would have been called a distinct genus. The three species, 0. annatonensis (Pfeiffer), 0. poecila Ancey, and 0. conella Sykes, were reviewed by Ancey (1906). 0. conella is quite distinctive, but the relationship between poecila and annato- nensis is uncertain. They differ primarily in size, and the situation may parallel that found in Gonatoraphe fornicata, where populations from different islands are quite different in size (see p. 184). Vate Island specimens of Omphalotropis poecila are much smaller than Espiritu Santo shells, but juveniles from Espiritu Santo matched the Vate adults. The one specimen of 0. annatonensis was much larger than any specimen of poecila from Espiritu Santo. The status of poecila and annatonensis can only be settled by population studies and comparisons of their anatomy. Omphalotropis (Lyrotropis) poecila Ancey. Plate 6, figs. 12, 13; plate 26, figs. 9, 10. Omphalotropis poecila Ancey, 1890, Le Naturaliste, 1890: 12 — Vate (Glisson); Ancey, 1904, Jour, de Conchy., 52, (4), p. 308; Ancey, 1906, Jour, de Conchy., 53, (3), p. 300. Omphalotropis varians Moellendorff, 1897, Nachr. d. Malak. Gesell., 1897: 166- 167— Vate (ex Layard); Sykes, 1903, Proc. Malac. Soc. London, 5: 199— Port Vila, Vate (J. J. Walker). Omphalotropis annatonensis santoensis Ancey, 1906, Jour, de Conchy., 53, (3), pp. 300-301, fig. 1 — Espiritu Santo and Aurora. Omphalotropis (Eurytropis) poecila Ancey, Kobelt, Jahrb. Nass. Ver. Naturk., 59: 79, 143. Omphalotropis (Eurytropis) varians Moellendorff, Kobelt, 1906, Jahrb. Nass. Ver. Naturk., 59: 83. Range. — Vate, Maewo, and Espiritu Santo. Material~ML 26h, ML 31d, ML 44, ML 46, ML 50, ML 51, ML 63, ML 66, ML 86, ML 95; Vate Island (UMMZ 76621, ex Walker, Ponsonby, Boettger, Layard); New Hebrides (CM). Remarks. — The specimens from Vate Island in the University of Michigan Museum of Zoology may be part of the type lot of 0. poecila. They are much smaller (average diameter 4.35 mm., average height 3.74 mm.) than the adult shells from Espiritu Santo (see Table XV), but juvenile specimens from Espiritu Santo with 4% whorls are iden- tical in size, shape, and sculpture. Ancey' s variety santoensis referred to examples of 0. poecila in which the white flammulations are miss- ing. Although it could be applied to the larger shells found on Espiritu I SOLEM: MOLLUSCA OF THE NEW HEBRIDES 203 Santo, I consider the size difference to be of little taxonomic impor- tance and I have not utilized the name santoensis. Most of the Espiritu Santo shells have whitish color streaks, but in about 30 per cent of the specimens the color markings were absent. The size variation found in 79 adult specimens from ML 31d is sum- marized in Table XV. The comparatively large standard deviation is probably the result of sexual dimorphism. Lot ML Bid contained only dead shells, but in other lots the shells of males were smaller than those of females. The radula (pi. 6, fig. 13) and external male genitalia (pi. 6, fig. 12) are typical of the Omphalotropidinae. The teeth are shown as iso- lated objects rather than in their natural positions (see Kondo, 1944, p. 314, fig. 1). The verge of 0. poecila is quite similar to that of 0. cookei Abbott (1949, p. 265, fig. le) from Saipan in the Marianas, but it is more slender and has less highly developed basal glands. The shell and operculum of 0. cookei are quite different from those of 0. poecila. Omphalotropis (Lyrotropis) conella Sykes. Plate 11, fig. 5. ?Omphalotropis sp. Sykes, 1903, Proc. Malac. Soc. London, 5: 200 — Renee River, Espiritu Santo. Omphalotropis conella Sykes, 1903, loc. cit., fig. 3 — Port Vila, Vate Island (J. J. Walker!). Omphalotropis conella Sykes var., Ancey, 1905, Nautilus, 19, (4), p. 45 — Espiritu Santo. Omphalotropis (Eurytropis) conella Sykes, Kobelt, 1906, Jahrb. Nass. Ver. Naturk., 59: 62. Range. — Vate and Espiritu Santo. Material. — Vate Island (UMMZ 75540, ex Walker, Ponsonby, Boettger, Layard; CM, ex Hartman, Layard); Renee River, Espiritu Santo (UMMZ 76539, ex Walker, Ponsonby, Sykes). Photograph of the holotype (BM). Remarks. — 0. conella is easily separated from the other New Heb- ridean Omphalotropis by its small size, prominent umbilical keel, and trochoidal shape. No operculi or preserved specimens were available. The Vate Island specimens from Layard are probably part of the type lot. The Renee River individual mentioned by Sykes (1903, p. 200) as Omphalotropis sp. is now in the University of Michigan Museum of Zoology. It falls within the range of variation shown by the six Vate Island specimens (see Table XV). No specimens of 0. conella were found by Kuntz. 204 FIELDIANA: ZOOLOGY, VOLUME 43 Omphalotropis (Lyrotropis) annatonensis (Pfeiflfer). Plate 26, fig. 8. Cyclostoma annatonensis Pfeiffer, 1856, Proc. Zool. Soc. London, 1855: 105 — Aneiteum, New Hebrides (MacGillivray!). Omphalotropis (Eurytropis) annatonensis (Pfeiffer), Kobelt, 1906, Jahrb. Nass. Ver. Naturk., 59: 65. Range. — Aneiteum. Ma^ma^.— Aneiteum (UMMZ 76520, ex Walker, Ponsonby, Boett- ger, Layard). Remarks. — The single specimen of 0. annatonensis examined is 10.0 mm. high, 8.3 mm. in diameter, and has 63^8 whorls. The only difference from 0. poecila is the much larger size. Pending study of further material of both species, I have retained 0. annatonensis as a distinct species. ANALYSIS OF NEW HEBRIDEAN NON-MARINE SNAILS Number of Species The literature contained approximately 135 specific names which had been applied to shells supposedly found in the New Hebrides, Some were misidentifications, others were based on erroneous local- ity data, and many have been placed in synonymy. In the zoogeo- graphical discussion, I am recognizing 79 species of land snails from the New Hebrides and Santa Cruz Islands (see below, p. 207) : 56 are land pulmonates, 17 are land prosobranchs, 3 are fresh-water inhabit- ants, and 3 are inter-tidal pulmonates. Several names listed in the systematic review (see Aneitea, Partula, and Diplomorpha) are omitted below since they probably do not represent valid species. Their in- clusion would give a distorted picture of the relative prominence of the three genera and unduly increase the marked endemism reported below. No specimens of Helix vannaelavae Cox, 1870, and Champa perryi Smith, 1897, were seen, and the identity of these named forms could not be determined. Of the 79 recognized species, 22 are reported for the first time. The 135 specific names are thus reduced to fifty-seven valid species records. Inter-Island Distribution A list of the species known from each island is given in Appen- dix IL For the 79 recognized species, there are only 152 island records (1.92 records per species). SOLEM: MOLLUSCA OF THE NEW HEBRIDES 205 This lack of locality data prevented any detailed analysis of speci- ation and distribution within the New Hebrides. Only twenty-five islands have any land snails reported from them; ten islands have more than 3 species; five islands more than 10; and only one island, Espiritu Santo, more than 25. Five islands, Aneiteum, Erromanga, Tanna, Vate, and Espiritu Santo account for 77 per cent (117) of the species records; Aneiteum, Vate, and Espiritu Santo for 63 per cent (95) ; Vate and Espiritu Santo for 49 per cent (75) ; and Espiritu Santo alone for 32 per cent (48). Forty-four of the 79 species have been reported from only a single island, but at least 20 of these should be rather widely distributed. Robert E. Kuntz obtained 41 of the 48 species known from Espi- ritu Santo; 24 were new records. Three of these were introduced, and the shells of the other 21 were all less than 7 mm. in diameter. Kuntz's most prolific sample was river drift (ML 95). It contained 38 species, 8 of which were found only in this sample. Kuntz's col- lections were limited to southeastern Espiritu Santo (see pis. 1 and 2) and undoubtedly many more species remain to be discovered in the vastly greater unexplored sections of the island. Even Espiritu Santo cannot be considered adequately explored for mollusks. Practically no mollusks are known from the large islands between Espiritu Santo and Vate (see pi. 1), and, except for Espiritu Santo, no island has more than a few minute snails reported from it. The easiest way for a non-malacologist to collect minute snails is in river drift. Kuntz's one drift sample contained 90 per cent of all the spe- cies he found. Drift has the disadvantage of containing only dead shells, but it provides specimens from a variety of habitats and is the only feasible method for an untrained collector to find minute species. In view of the limited number of locality records, little can be said about distribution patterns within the New Hebrides. Many species range throughout the archipelago. Others are known from only one island but probably are equally widely distributed. There are no clear faunal dissimilarities between the northern and southern islands, and no faunal areas can be recognized. Extensive collecting may change the picture, but I suspect the islands will prove to have a singularly homogeneous fauna. Probably this can be traced to the unstable topography. Relatively long-term maintenance of barriers to dispersal is necessary for successful speciation. A continuing alter- nation of separation and reunion of populations through geological time would tend more toward variation within species than to sep- aration into species. 206 FIELDIANA: ZOOLOGY, VOLUME 43 Only two of the Santa Cruz Islands, Vanikoro and Santa Cruz, have land snails reported from them. With one exception (Papuina charlottae Jaeckel and Schlesch) all the species are closely and obvi- ously related to New Hebridean shells and unrelated to Solomon Is- land or Fijian taxa. Papuina charlottae probably is a mislabeled Solomon Island shell. Only the holotype is known, and the source of the specimen has been responsible for other erroneous localities. In regard to land snails, the Santa Cruz Islands form a single unit with the New Hebrides. A similar situation apparently exists in in- sects (Gressitt, 1956, p. 15), although the reptile fauna (p. 312) shows considerable differences. In discussing the land snails, the Santa Cruz and New Hebrides are hereafter referred to jointly as the New Hebrides. Perhaps the only definite indication of an inter-island distribu- tional pattern can be seen in a parallelism between the snails of the southern New Hebrides and the Santa Cruz Islands. Trochomorpha hakeri from Aneiteum is much more similar to the Vanikoro Island Trochomorpha sp. than to the central New Hebridean T. rubens; the Vate Island Dendrotrochus e. eva is represented in the Santa Cruz and Banks Islands by D. e. stramineus, with the quite different D. layardi on Espiritu Santo; and the southern New Hebridean Placostylus salo- monis is more closely related to the Santa Cruz P. hullianus than to the Espiritu Santo P. bicolor (Sykes, 1903, reported P. salomonis from Espiritu Santo, but the record needs confirmation). The data listed above are insufficient to draw anything more than very tentative conclusions, but the immediately suggested idea is that Espiritu Santo has served as a dispersal center from which both the southern New Hebrides and Santa Cruz Islands have been populated. Until the islands have been much more thoroughly explored for land snails, however, this can only be regarded as a tentative suggestion. Endemism On most Pacific Islands it is possible to recognize species having one of five relatively distinct types of distribution: (a) The "tropical tramps" which have been spread throughout the world by commerce of the last 200 years; (b) species widely dispersed over the Pacific by primitive man; (c) species found in only two archipelagos; (d) species endemic to an island or archipelago with congeneric relatives in other regions; and (e) species belonging to a genus found only in one archi- pelago. Most species can be easily placed in one of the five categories, but a few are of uncertain relationship. SOLEM: MOLLUSC A OF THE NEW HEBRIDES 207 In the following list of the 79 recognized species, the symbols indi- cate whether the species are endemic (E), found on another island group (N), very widely distributed in the Pacific (W), or a "tropical tramp" (T). SYSTEMATIC LIST OF NEW HEBRIDEAN AND SANTA CRUZ LAND AND FRESH-WATER MOLLUSCA PULMONATA SYSTELLOMMATOPHORA Oncidiacea Oncidiidae Oncidium peronii Cuvier — W Oncidium verruculaium Cuvier — W Oncis aff . mariensi Plate — W Soleolifera Veronicellidae Angustipes plebeius (Fischer) — T Eleutherocaulus alte (Ferussac) — W PULMONATA STYLOMMATOPHORA Tracheopulmonata Athoracophoridae Aneitea macdonaldi Gray — E Aneitea speiseri Grimpe and Hoffmann — E Heterurethra Succineidae Succinea (Papusuccinea) kuntziana Solem — E Orthurethra Pupillidae Gastrocopta pediculus (Shuttleworth) — W Tornatellinidae Elasmias apertum (Pease) — W Lamellidea piisilla (Gould) — W Enidae Rhachistia histrio (Pfeiffer)— T(?) Partulidae Partula macgillivrayi Pfeiffer — E Partula pyramis Hartman — E Partula auraniana Hartman — E Partula milleri Solem — E Partula vanikorensis (Quoy and Gaimard) — E Sigmurethra Aulacopoda Arionacea Endodontidae Phrixgnathtis glissoni (Ancey) — E Phrixgnathus tenuiscrtptus (Ancey) — E Discocharopa planulata Solem — E Mocella euryomphala Solem — E Reticharopa geddiei Solem — E Reticharopa stenopleura Solem — E 208 FIELDIANA: ZOOLOGY, VOLUME 43 Reticharopa latecosta Solem — E Reticharopa helva (Cox) — E Reticharopa sp. — E Limacacea Helicarionidae Euconulinae Wilhelminaia afF. mathildae Preston — W(?) Coneuplecta (Sitalina) microconus (Mousson) — W Coneuplecta (Conibycus) bicarinata Solem — E Microcystinae Liardetia (Liardetia) samoensis (Mousson) — W Diastole subcarinata Solem — E Lamprocystis guttula (Pfeiffer) — E Lamprocystis mendanae Solem — E Sesarinae Orpiella (fHalozonites) retardata (Cox) — E Dendrotrochus (Santotrochus) eva (Pfeiflfer) — E Dendrotrochus {Santotrochus) layardi (Hartman) — E Zonitidae Trochomorphinae Trochomorpha (Hartmanitrochus) rubens Hartman — E Trochomorpha (Hartmanitrochus) bakeri Solem — E Trochomorpha sp. — E Holopoda Subulinidae Subulina octona (Bruguiere) — T Opeas pumilum (Pfeiffer) — T Lamellaxis (Allopeas) gracilis (Hutton) — T Bradybaenidae Bradybaena similaris (Ferussac) — T Camaenidae(?) Draparnaudia singularis (Pfeiffer) — ^N Draparnaudia walkeri Sykes — E Bulimulidae Placostylus hullianus (Iredale) — E Placostylus fuligineus (Pfeiffer) — E Placostylus salomonis (Pfeiffer) — E Placostylus turneri (Pfeiffer) — E Placostylus bicolor (Hartman) — E Diplomorpha (Diplomorpha) layardi Ancey — E Diplomorpha {Diplomorpha) coxi (Pease) — E Diplomorpha {Diplomorpha) delautouri (Hartman) — E Diplomorpha {Quiros) bernieri (Hartman) — 3 Paryphantidae Delos {Hebridelos) rapida (Pfeiffer) — E Delos {Hebridelos) gassiesi (Pfeiffer) — E Delos {Hebridelos) haasi Solem — E Macrocycloides annatonensis (Pfeiffer) — E Ouagapia santoensis Solem — E PULMONATA BASOMMATOPHORA Planorbidae Physastra layardi (Ancey) — W Gyraulus montrouzieri (Gassies) — N SOLEM: MOLLUSCA OF THE NEW HEBRIDES 209 PROSOBRANCHIA ARCHAEOGASTROPODA Helicinidae Pleuropoma varians (Sykes) — E Pleuropoma taeniata (Quoy and Gaimard) — E Pleuropoma sublaevigata (PfeiflFer) — E Pleuropoma albescens (Hartman) — E Pleuropoma rotella (Sower by) — E Pleuropoma articulata (Pfeiflfer) — E or W PROSOBRANCHIA MESOGASTROPODA Cyclophoracea Poteriidae Gonatoraphe fornicata (Pfeiflfer) — E Pupinidae Pupina (Kanapa) brazieri (Crosse) — E Pupina (Kanapa) cumingiana PfeiflFer — E Diplommatinidae Palaina franqoisi Ancey — E Palaina sykesi Solem — E Rissoacea Hydrobiidae Fluviopupa brevior (Ancey) — E Truncatellidae Truncatella Truncatella guerinii A. and J. Villa — W Assimineidae Omphalotropis (Oriella) setocincta Ancey — E or W Omphalotropis (Lyrotropis) annatonensis (PfeiflFer) — E Omphalotropis (Lyrotropis) poecila Ancey — E Omphalotropis {Lyrotropis) conella Sykes — E Assiminea (Assiminea) nitida (Pease) — W "Tropical tramps." — During the past few centuries commerce has spread many land snails throughout the tropical and warm-temperate regions of the world. These "tropical tramps" or "culture snails" are of uncertain geographic origin but now live in nearly every tropical country. Five New Hebridean pulmonates belong in this category: the veronicellid, Angustipes plebeius (Fischer); the three subulinids; and the helicid, Bradyhaena similaris (Ferussac). Other "tropical tramps" found in Melanesia and Australia, but not yet reported from the New Hebrides, include the subulinids, Lamellaxis micra (Orbigny) and L. mauritianum (PfeiflFer); the streptaxid, Gulella (Huttonella) hicolor (Hutton) ; the edible snails. Helix aspersa and Helix pomatia; and species of the slug genera, Deroceras, Ldmax, and Arion. Rhachis- tia histrio (Pfeiflfer) may belong to this category. R. histrio lives in the New Hebrides, New Caledonia and Queensland, and it may be an African species introduced into the Pacific. Without further study, however, its status cannot be precisely determined. Atoll fauna. — The low coral atolls of the Pacific Ocean have a characteristic but limited snail fauna. The species are minute, widely 210 FIELDIANA: ZOOLOGY, VOLUME 43 distributed, and, at least in some cases, carried from island to island by natives (see Cooke, 1926, pp. 2278-2279). Eight New Hebridean species range throughout most of Polynesia, Micronesia and Indo- nesia: the veronicellid, Eleutherocaulus alte (Ferussac); the pupillid, Gastrocopta pediculus (Shuttleworth) ; the tornatellinids, Elasmias apertum (Pease) and Lamellidea pusilla (Gould); the helicarionids, Coneuplecta microconus (Mousson) and Liardetia samoensis (Mous- son); and the operculates Truncatella guerinii (A. and J. Villa) and Assiminea nitida (Pease) . One other species, the helicarionid, Wilhel- minaia aff . mathildae Preston, probably lives in Indonesia, northern Melanesia, and the Caroline Islands. It is not known from Polyne- sia and New Caledonia. The three Oncidiidae are very widely dis- persed in the Indo-Pacific. In their mode of distribution (by pelagic larvae) the Oncidiidae are marine rather than terrestrial, and they are not discussed further in this study. Thirteen of the 56 land pulmonates and two of the 17 land proso- branchs are obvious introductions or are so widely distributed that they are of no value in a zoogeographic analysis. Of the remaining 43 pulmonates and 15 prosobranchs, only one (Draparnaudia singu- laris Pfeiffer) is found elsewhere (New Caledonia). The other 57 are endemic. The three fresh-water species will be considered later. Endemic species. — Some of the endemic species belong to very widely distributed genera — Succinea, Discocharopa, Mocella, Macro- cycloides, Ouagapia, two species of Palaina, Pleuropoma articulata (Pfeiffer), the five partulids, and Omphalotropis setocincta Ancey — which, at the present time, do not seem to be divisible into subgen- eric categories. Diastole, Conihycus, Orpiella, the two species of Lam- procystis, three Dehs, and the other five Pleuropoma probably belong to endemic sections or subgenera. Until their anatomy has been studied, however, description of subgeneric units is unwarranted. The two species of Dendrotrochus, three Trochomorpha, five Placo- stylus, three Omphalotropis (Lyrotropis) , and two Pupina definitely belong to endemic sections or subgenera. A few species — the endemic Draparnaudia, Gonatoraphe, two Aneitea, and two Phrixgnathus — belong to genera found in only one other archipelago. The remain- ing nine species — five Reticharopa and four Diplomorpha — belong to endemic genera. Fresh-water snails. — Two pulmonates, Physastra layardi (Ancey) and Gyraulus montrouzieri (Gassies),and one prosobranch, Fluviopupa brevior (Ancey), comprise the primary fresh-water snail fauna. Neri- SOLEM: MOLLUSCA OF THE NEW HEBRIDES 211 tina (sens, lat.) and the Thiaridae are at least partially salt-water tolerant and are not included in this study as being fresh-water organ- isms. Physastra layardi is found in New Caledonia and possibly Fiji and Tonga. Gyraulus montrouzieri also lives in New Caledonia, but a different Gyraulus is found in the Solomons and Fiji (Solem, in press-A). Fluviopupa brevior is endemic, with close relatives in the Fijis and Rapa. The reasons for the decreased endemism of the fresh-water snails are discussed below (pp. 258-259) . Summary. — Considering just the seventy-three land dwellers, six species (8.2 per cent) are "tropical tramps," nine (12.3 per cent) are part of the coral atoll fauna, one (1.4 per cent) is found on one other archipelago, and fifty-seven (79.1 per cent) are endemic. Of the endemic species, fourteen (24.6 per cent) belong to genera found on several archipelagoes, thirteen (22.8 per cent) probably belong to endemic subgeneric taxa, fifteen (26.4 per cent) definitely be- long to endemic subgeneric units, six (10.5 per cent) belong to genera found on only one other archipelago, and nine (15.7 per cent) belong to endemic genera. I have repeatedly emphasized that the classification of Pacific land snails is poorly understood and that New Hebridean species have been placed in endemic supraspecific taxa only when sufficient evi- dence of extralimital affinities allowed possible recognition of phyletic lines. Genera of mollusks are not equivalent to genera of vertebrates (see pp. 26-27), and if the smaller genera of Iredale had been uti- lized here, quite different results would have been obtained. Perhaps thirty-four genera (compared to twenty-five) of which twenty-three were endemic (compared to two) would have been recognized. Distribution of New Hebridean Land Snail Taxa If we exclude the five or six "tropical tramps," nine atoll species, three oncidiids, Pleuropoma articulata,^ and Omphalotropis setocincta,^ there are fifty-nine species (three living in fresh water) with limited enough distributions to be useful in a zoogeographic analysis. The fifty-six terrestrial species are placed in twenty-four genera, of which only two, Reticharopa and Diplomorpha, are endemic. The genera belong to fourteen different families. The distribution of the families and genera with New Hebridean endemic species is shown in Tables XVI and XVII and figures 7 and 8. ' These two species may be widely distributed "strand-line" dwellers (see pp. 177, 200). Table XVI. — Distribution of New Hebridean Families Family NG^ SI NH NC NZ F T A C Athoracophoridae x ? x x x x Succineidae' x x x x x x x x Partulidae' x x x x x x Endodontidae' xxxxxxxxx Helicarionidae' x x x x x x x Zonitidae^ x x x x x x x Draparnaudia x x Bulimulidae x x x x x x'^ Paryphantidae^ x x x x x x x x x Camaenidae x x x Helicinidae' x x x x x x x x Cyclophoridae x x ? x^ Poteriidaei x x ? x x Pupinidae x x x ? x x x Diplommatinidae^ x x x x x x x x x Assimineidae^ x x x x x x x x x 1 Also in Samoa. ^ A different subfamily. ^ Only in northern Queensland. * NG (New Guinea), SI (Solomon Islands), NH (New Hebrides), NC (New Cale- donia), NZ (New Zealand), F (Fiji), T (Tonga), A (Australia), C (Caroline Islands). Table XVII. — Distribution of New Hebridean Genera Genus NG^ SI NH NC NZ F T Aneitea R* ? x x R Papusuccinea^ x x x x x x Partula^ x^ x x x x Phrixgnathus x x Discocharopa ? ? x x Mocella^ x x x x x Reticharopa R ? x Conibycus ? ? x Diastole' ? ? x (x) (x) Lamprocystis' (x)^ ? x (x) (x) Orpiella (x) (x) x (x) (x) Dendrotrochus ? (x) x Trochomorpha' (x) (x) x (x) (x) Draparnaudia x x Placostylus (x) x (x) x (x) Diplomorpha x Delos X (x) x Macrocycloides ? x x Ouagapia' x x x x x x Pleuropoma' (x) ? x x(?) (x) (x) Gonatoraphe x R x Pupina (x) (x) x (x) Palaina x x x x x x Omphalotropis' (x) (x) x (x) (x) (x) ' Also Samoa. ^ One record. ^ In southwest Australia only. * See footnote 4 on Table XVI. * A related genus. * A different subgenus. R R R(?) X x X3 R (X) (X) (X) (X) (X) (X) (X) R (X) (X) X X (X) (R) 212 ISLANDS « Truk Ponape i3; Kusaie DISTRIBUTION OF NEW HEBRIDEAN FAMILIES ADDITIONAL FAMILIES OF LAND SNAILS I SAME FAMILIES AS I IN NEW HEBRIDES ii2= SOLOMOir^k SLANDS* • . NEW HEBRIDEsi\ «^,// 14 . '» 9 10 Lord Howe Island NEW CALEDONIA FIJI TONGA Norfolk Island NEW ZEALAND Fig. 7. Distribution of families of New Hebridean land snails. 213 ^Ponape Kusaie , ^„ SOLOMON ^^^ V ISLANDS ISLANDS DISTRIBUTION OF NEW HEBRIDEAN GENERA LIMITED TO TWO OR THREE AREAS RELATED GENERA OR SUB- GENERA WIDELY DISTRIBUTED f.'.'.'J GENERA endemic g ener a T XI NEW 4> HEBRIDES y^' FIJI CALEDONIA Norfolk Island Lord Howe Island NEW ZEALAND Fig. 8. Distribution of genera of New Hebridean land snails. 214 SOLEM: MOLLUSCA OF THE NEW HEBRIDES 215 The raw data are summarized in figures 7 and 8. New Zealand and Tonga have the fewest New Hebridean famiUes and the Solo- mons and Fiji the most. New Guinea, the Solomons, and Australia (but none of the other areas) have families which are not found in the New Hebrides. Obviously there are many important faunistic changes, since from two to six of the New Hebridean families are absent from each area. The New Hebridean genera (fig. 8) show three types of distri- bution: taxa found in the New Hebrides and only one or two other areas; taxa represented by distinct subgeneric units in other areas; and taxa widely distributed over the Pacific. It proved impossible to include the genera present in these other areas but not in the New Hebrides, and figure 8 shows only the distribution of genera found in the New Hebrides. Because of inadequate collections of small shells, the records for the Solomons and New Guinea are based on a number of probable but unconfirmed records. New Caledonia and New Zealand have the fewest New Hebridean genera, the Solomons, Fiji, and New Guinea the most. The great difference between the New Hebridean and New Caledonian fauna is one of the most strik- ing and most unexpected results of this study. The difference cannot be considered an artifact of collecting, since New Caledonia has been the object of more intensive malacological collecting than any area of the Pacific except Hawaii and the Society Islands. Another im- portant factor is the distribution of genera found on only one or two archipelagos other than the New Hebrides. New Caledonia and Fiji are the nearest islands, but the New Zealand and Caroline Island affinities are surprising, as is the lack of any close affinities with the Solomon Islands. The absolute presence or absence of a taxon is useful, but a more realistic comparison would encompass the relative importance of the taxa in each area as well as the general types of distribution shown by the families and genera. Our knowledge about the extent of speci- ation in land snails over most of the Pacific region is so pitifully inade- quate that no comparison of the relative abundance of species in different areas can be attempted. For many areas we have a fair idea of qualitative, but very little of quantitative content. Appendix I: ROBERT E. KUNTZ COLLECTING LOCALITIES ON ESPIRITU SANTO / . ML 1. On dagger plants near Surundo River. Oct., 1943. ML 2. Near Turtle River, early morning. Oct. 16, 1943. ML 3. Plants and rocks within 20 feet of water. Oct., 1943. ML 5. On rocks along shore. Oct. 3, 1943. ML 6. Underneath large clam shell, west of Renee River. Oct. 21, 1943. ML 7. Turtle Bay. On vegetation, 20-40 feet from stream in deep shade. Oct. 18, 1943. ML 8. Decaying coconut. Sarakata River peninsula. Nov., 1943. ML 9. On ground in dense jungle, near Sarakata River. Oct., 1943. ML 10. On ground in jungle, near Sarakata River. Oct., 1943. ML 11. On tree trunk on rainy day. Sarakata River peninsula. Nov. 17, 1943. ML 12. On decaying cacao leaves near Malaria Control Headquarters. Oct. 22, 1943. ML 13. On decaying cacao leaves near Malaria Control Headquarters. ML 14. In decaying coconut, Sarakata River peninsula. Oct., 1943. ML 15. Under side of rocks in Renee River. Dec. 30, 1943. ML 16. From green leaves and dead wood in damp shaded area. Dec. 28, 1943. ML 17. Under side of green leaves of banana-like tree. Dec. 28, 1943. ML 18. Dead and decaying sticks and limbs, some from ground. Jan. 6, 1944. ML 19. Dead and decaying sticks and limbs; some from ground. ML 20. Dead and decaying sticks and limbs; some from ground. ML 21. Dead and decaying sticks and limbs; some from ground. ML 22-ML 24. Dead limbs on ground near edge of woods. Jan. 13, 1944. ML 25. Rocks along shore of Renee River. Occupied by crabs. Dec. 30, 1943. ML 26. From vegetation, boards and decaying wood in deep shade in woods. Jan. 20, 1944. ML 27. Curtiss Creek, in mud on bottom. Jan. 22, 1944. ML 28. Roots of dead trees along bay at mouth of Sarakata River, low tide. Jan. 31, 1944. ML 29. Renee River, 1^^ miles from mouth, in shallow running water attached to rocks. These eaten by Hawaiians and Filipinos. Dec. 30, 1943. ML 30. Same as ML 29, except 3^ to IJ^ miles upstream. ML 31. At edge of jungle during rain. Jan. 22, 1944. a. Under side of decaying logs. 6. Tree trunks and fence posts; 1-15 feet up. c. Rotting logs, on trees and ground. d. Trunks of trees and shrubs. e. Leaves of trees and shrubs. f-h. Trunks of trees and shrubs. i. Under sides of boards, rotting logs, coconuts. ML 32. Trunks and limbs of trees and shrubs; 1-7 feet up. Feb. 6, 1944. ML 33. From alluvial deposits in road cuts, etc. Feb. 6, 1944. Most common within 150 yards of river. ML 34. Under logs in Sarakata River valley. Jan. 30, 1944. ML 35. Under side of logs in short grass meadow. Jan., 1944. 216 SOLEM: MOLLUSC A OF THE NEW HEBRIDES 217 ML 36. Along road. Feb. 24, 1944. ML 37. From vegetation, trees and shrubs. Feb. 24, 1944. ML 38. From lava rocks and coral, 13^ miles up Renee River; water fresh. Dec. 30, 1943. ML 39. Leaves and decaying logs in deep shade. Small snails more abundant under sticks in open places on hillside. Feb. 27, 1944. ML 40. In rotting coconuts, Brigstock Point. March 11, 1944. a. Under coconut fronds on ground. ML 41. Bought on Vao, Malekula, from natives; ready to string. Dec. 5, 1943. ML 42. From rocks along and in Renee River, 1 3^ miles upstream. Water fresh. Mar. 24, 1944. ML 43. Base of papaya tree, Cabar Plantation. Apr. 4, 1944. ML 44. From logs and under coconuts. Apr. 13, 1944. Upper Renee River. ML 45. Water cress in Cabar Creek. Apr. 15, 1944. ML 46. Same as ML 44. On logs. ML 47. Under rocks in stream below dam, Brigstock Point. Water swift. April 10, 1944. ML 48. Same as ML 47. From stream bottom, some on large rocks. ML 49. Venus Island. March 19, 1944. Eaten by natives and Tonkinese. ML 52. From water cress, Cabar Creek. Water swift. ML 53. From coral rocks in swift-flowing water. Apr., 1944. ML 54-57. From coral rocks in swift-flowing water. ML 58. From coral rocks in swift-flowing water. Under side of rocks; a few on blades of grass. ML 59. Under side of logs. Apr., 1944. ML 60. Small semi-permanent pool near Brigstock Point. Chara, lilies and grass in pool. May 5, 1944. ML 61-62. Small semi-permanent pool near Brigstock Point. Chara, lilies and grass in pool. ML 63. On damp under side of logs in shade near pond. Apr. 25, 1944. ML 64. On damp under side of logs in shade near pond. ML 65. Semi-permanent pool near Brigstock Point. May 7, 1944. Infected with small stylet cercariae. ML 66. From pandanus and other low trees. Apr. 19, 1944. ML 67. In swift water below dam, Brigstock Point. Apr. 28, 1944. ML 68. Water cress in Cabar Creek. May 25, 1944. ML 69. Under side of damp, decaying logs. May 27, 1944. ML 70. Shrubs and decaying, fallen trees. Apr., 1944. ML 71. From muddy bottom of swift-flowing stream, Naranga Village, 250 feet altitude. May 28, 1944. ML 72. On damp ground under trees, buried in decaying logs and under leaves. ML 73. From rocks in running water in shade, 2 miles up. June 10, 1944. ML 74. On ground among decaying leaves, 300 yards from Renee River. June 8, 1944. ML 75. Small stream running into south end of Brigstock Lagoon. From edge of running water in deep shade of mangroves. June 3, 1944. ML 76. From ground and dead leaves in shade, 300 yards from Renee River. June 2, 1944. ML 77. Cabar Creek. Half buried in muck and decaying vegetation in still places along stream. May 14, 1944. ML 78. From rocks in quiet side pool of Sarakata River. Some algae. June 17, 1944. ML 79. From rocks and leaves in small pool in Sarakata River bed. Pool shaded, water running slowly. Opisthorcid cercaria. End of logging trail. June, 1944. ML 80. In small semi-permanent ground pool in coconut grove, Ratard Planta- tion. Water muddy. May, 1944. ML 81. From rocks on bottom of stream, Renee River. June 19, 1944. 218 FIELDIANA: ZOOLOGY, VOLUME 43 ML 82. From sticks, dead leaves and duckweed in quiet waters of creek running into Brigstock Lagoon. June 21, 1944. ML 83. From decaying leaves and branches, 100 yards from shore line of bay near Renee River. June, 1944. ML 84. From stalk and under side of leaves of banana-like plant along Renee River. Found in deep shade where plants are moist all the time. June, 1944. ML 85. From quiet pool connected with dammed creek at Brigstock Lagoon. In masses of duckweed, water lilies, and chara. July, 1944. ML 86. From leaves of shrubs (a), decaying logs (6) and ground (c) in shaded jungle adjacent to coconut plantation. June 26, 1944. ML 87. In small side pool on south bank of Sarakata River. June, 1944. Small snails with dicroclid cercaria. ML 88. From shaded large coral rock 1 3^ miles up Renee River, in swift water. May, 1944. ML 89. From rocks in partly shaded, swift-flowing stream running into Brigstock Lagoon. June, 1944. ML 90. From tree trunks in shaded jungle. Aug., 1944. ML 91. From shallow trailside pool in well-shaded water protected by overhang- ing grass near Havet Plantation. July 15, 1944. ML 92. Under side of logs and planks off beach near Sarakata River. Aug. 23, 1944. ML 93. From small ground pool in coconut grove, Aore Island. Aug. 15, 1944. ML 94. Under side of fallen tree trunks on high ground above Brigstock Lagoon. Aug. 24, 1944. ML 95. From pile of debris left at base of tree during high water in Sarakata River. May-June, 1944. ML 96. Dead shells from banks of stream. July, 1944. Marine Collections NH 1. Palikulo Bay. Along shore. Oct., 1943. NH 2. Woods near shore. Oct., 1943. NH 3. Roots of mangrove trees. Oct., 1943. NH 4. Near shore at low tide, night. Oct., 1943. NH 5. Shore line on Segond Channel. Oct. 20, 1943. NH 6. Roots of mangrove trees in Pallikulo Bay. Nov. 22, 1944. NH 7. Shores of Segond Channel. Oct. 24, 1943. NH 8. Shores of Malo Island. Dec. 14, 1943. NH 9. Shores of Palikulo Bay. Nov. 24, 1943. NH 10. Segond Channel. Nov. 24, 1943. NH 11. Shore of Araki Island, facing sea; steep rough volcanic banks. NH 12. Dead tree trunks and rocks along shore at Brigstock Point. Feb. 20, 1944. NH 13. Palikulo Bay at low tide. NH 14. At night near Brigstock Point. March 20, 1944. NH 15. From coral pile. March, 1944. Appendix II. MOLLUSCA KNOWN FROM SANTA CRUZ AND NEW HEBRIDEAN ISLANDS "?"=a doubtful record; *=literature reference only. Names of questionable validity included. Island sequence from south to north. ANEITEUM Aneitea macdonaldi Gray Partula macgillivrayi Pfeiffer* Partula eximia Hartman Reticharopa geddiei Solem Reticharopa helva (Cox) Lamprocystis guttula (Pfeiffer) Coneuplecta microconus (Mousson) Liardeiia samoensis (Mousson) Orpiella retardata (Cox) Trochomorpha bakeri Solem "Helix" vannaelavae (Cox)* Lamellaxis gracilis (Hutton) Draparnaudia s. singularis (PfeifiFer)* Draparnaudia s. diminuta Ancey* Placostylus salomonis (Pfeiffer) Placostylus fuliginetis (Pfeiffer) ?Diplomorpha coxi (Pease) Delos haasi Solem Macrocycloides annatonensis (Pfeiffer) Physastra layardi (Ancey) Pleuropoma articulata (Pfeiffer) Gonatoraphe fornicata (Pfeiffer) Truncatella guerinii A. and J. Villa* Omphalotropis annatonensis (PfeifiFer) TANNA Aneitea macdonaldi Gray Sticcinea kuntziana Solem Rhachistia histrio (Pfeiffer)* Partula turneri PfeifiFer Placostylus salomonis (PfeifiFer) Physastra layardi (Ancey) Pleuropoma sublaevigata (PfeifiFer) Pleuropoma rotella (Sowerby)* Pleuropoma articulata (PfeifiFer)* Pupina cumingiana PfeifiFer Truncatella guerinii A. and J. Villa* Placostylus salomonis (PfeifiFer) FUTUNA Physastra layardi (Ancey) ANIWA Physastra layardi (Ancey) ERROMANGA Aneitea macdonaldi Gray Partula turneri PfeifiFer fPartula minor Hartman Trochomorpha rubens Hartman Liardetia samoensis (Mousson) Placostylus salomonis (PfeifiFer) Placostylus fuligineus (PfeifiFer) Placostylus turneri (PfeifiFer)* ? Delos rapida (PfeifiFer)* Delos gassiesi (PfeifiFer) Pupina brazieri (Crosse) Gonatoraphe fornicata (PfeifiFer) Truncatella sp.* 219 220 FIELDIANA: ZOOLOGY, VOLUME 43 VATE Angustipes plebeius (Fischer)* Eleutherocaulus alte (Ferussac)* Succinea kuntziana Solem Rhachistia histrio (Pfeiffer) Partula caledonica (Pfeiffer) Partula pyramis Hartman Phrixgnathus glissoni (Ancey) Coneuplecta microconus (Mousson) Lamprocystis guttula (Pfeiffer) Orpiella retardata (Cox) Dendrotrochus e. eva (Pfeiffer) Trochomorpha rubens Hartman* Opeas pumilum (Pfeiffer) Subulina octona (Bruguiere) Dendrotrochus e. eva (Pfeiffer) Subulina octona (Bruguiere) Bradybaena similaris (Ferussac) Diplomorpha layardi Ancey Delos rapida (Pfeiffer) Physastra layardi (Ancey) Pleuropoma sublaevigata (Pfeiffer) Pleuropoma articulata (Pfeiffer) Gonatoraphe fornicata (Pfeiffer) Fluviopupa brevior (Ancey)* Truncatella guerinii A. and J. Villa Omphalotropis setocincta Ancey Omphalotropis conella Sykes Omphalotropis poecila Ancey Assiminea nitida (Pease) EPI Pleuropoma sublaevigata (Pfeiffer)* MALEKULA Phrixgnathus tenuiscripta (Ancey)* Trochomorpha rubens Hartman ?Diplomx)rpha peasei (Cox) OMBA Diplomorpha delautouri Hartman MAEWO fOmphalotropis poecila Ancey* ESPIRITU SANTO Oncis aff. martensi Plate Oncidium verruculatum Cuvier Angustipes plebeius (Fischer)* Aneitea maloensis Grimpe and Hoffmann* Aneitea speiseri Grimpe and Hoffmann* Aneitea elisabethae Grimpe and Hoffmann* Aneitea r. robsoni Hoffmann* Aneitea r. santoensis Solem Succinea kuntziana Solem Gastrocopta pediculus (Shuttleworth) Partula albescens Hartman Partula auraniana Hartman Partula fraterna Hartman Partula carnicolor Hartman Partula milleri Solem Phrixgnathus glissoni (Ancey) Phrixgnathus tenuiscriptus (Ancey) Discocharopa planulata Solem Mocella euryomphala Solem Reticharopa latecosta Solem Reticharopa stenopleura Solem Reticharopa sp. Wilhelminaia aff. mathildae Preston Coneuplecta microconus (Mousson) Coneuplecta bicarinata Solem Liardetia samoensis (Mousson) Diastole subcarinata Solem Lamprocystis mendanae Solem Orpiella retardata depressa Solem Dendrotrochus layardi (Hartman) Trochomorpha rubens Hartman Subulina octona (Bruguiere) Lamellaxis gracilis (Hutton) Opeas pumilum (Pfeiffer) Bradybaena similaris (Ferussac) Draparnaudia walkeri Sykes Placostylus bicolor (Hartman) Placostylus salomonis (Pfeiffer)* Diplomorpha brazieri Hartman Diplomorpha delautouri Hartman SOLEM: MOLLUSCA OF THE NEW HEBRIDES 221 ESPIRITU SANTO— continued Diplomorpha bernieri (Hartman) Palaina franqoisi Ancey Ouagapia santoensis Solem Palaina sykesi Solem Physastra layardi (Ancey) Fluviopupa brevior (Ancey) Gyraulus montrouzieri (Gassies) Truncaiella guerinii A. and J. Villa* Pleuropoma albescens (Hartman) Omphalotropis setocincta Ancey Pleuropoma sublaevigata (Pfeiffer) Omphalotropis poecila Ancey Gonatoraphe fornicata (Pfeiffer) Omphalotropis conella Sykes Pupina brazieri Crosse GAUA, BANKS Partula fraterna Hartman* Physastra layardi (Ancey)* Pleuropoma sublaevigata (Pfeiffer)* Fluviopupa brevior (Ancey) VANUA LAVA, BANKS Partula proximo Hartman* Dendrotrochus eva stramineu^ Sykes* Partula fraterna Hartman* ?Trochomorpha rubens Hartman* Coneuplecta microconus (Mousson)* Pleuropoma sublaevigata (Pfeiffer)* VALUA, BANKS Succinea kuntziana Solem* Pleuropoma sublaevigata (Pfeiffer)* MOTA ISLAND, BANKS "Charopa" perryi Smith* LO ISLAND, TORRES Partula auraniana Hartman* Pleuropoma sublaevigata (Pfeiffer)* Dendrotrochus eva stramineus Sykes* BUKA-BUKA ISLAND, TORRES Partula auraniana Hartman* HIU ISLAND, TORRES Partula auraniana Hartman Pleuropoma sublaevigata (Pfeiffer)* "NEW HEBRIDES" Triboniophoru^ graeffei Humbert* Delos haasi Solem Partula radiosa (Pfeiffer)* VANIKORO, SANTA CRUZ Partula vanikorensis (Quoy and Gaimard) Pleuropoma taeniata Trochomorpha sp.* (Quoy and Gaimard)* Truncaiella striata (Quoy and Gaimard)* SANTA CRUZ ISLAND, SANTA CRUZ Oncidium peronii Cuvier* Placostylus hullianus (Iredale) Dendrotrochus eva stramine