Vfi ;S) HARVARD UNIVERSITY Library of the Museum of Comparative Zoology SuL Latin OF TH seum o m f f;^--,-.;.;: vv JJ«--i^A„ m sl^f if«--i':i Volume 141 1971 ^/^r HARVARD UNIVERSITY CAMBRIDGE, MASSACHUSETTS 02138 U.S.A. CONTENTS No. 1. West Indian Xenodontine Colubrid Snakes: Their Probable Origin, Phylogeny, and Zoogeography. By Vincent J. MagHo. December, 1970 1 No. 2. The Milliped Family Conotylidae in North America, With a Description of the New Family Adritylidae ( Diplopoda : Chorde- umida). By William A. Shear. February, 1971 55 No. 3. Monograph of the Cuban Genera Emoda and Ghjptemoda ( Mol- lusca: Ai-chaeogasti-opoda: Helicinidae). By William J. Clench and Morris K. Jacobson. February, 1971 99 No. 4. The Diadematus Group of the Orb-Weaver Genus Araneus North of Mexico (Araneae: Araneidae). By Herbert W. Levi. February, 1971 131 No. 5. Evolutionary Relationships of Some South American Ground Tyrants. By W. John Smith and Frangois Vuilleumier. March, 1971 : 181 No. 6. Systematics and Natural History of the Mygalomorph Spider Genus Antrodioetiis and Related Genera (Araneae: Antrodiaeti- dae). By Frederick A. Coyle. July, 1971 269 No. 7. A Monograph of the Genera Calidviono, Ustronia, Troschelviana, and Semitrochatella (Mollusca: Archaeogasti-opoda: Helicinidae). By William J. Clench and Morris K. Jacobson. August, 1971 403 No. 8. The Orb Weaver Genus Neoscona in North America (Araneae: Araneidae). By Jonathan D. Berman and Herbert W. Levi. September, 1971 465 BuLLetln OF THE Museum of Comparativ e oology -.i.,T:|-./..^i>,'ti\ -- West Indian Xenodontine Colubrid Snakes: Their Probable Origin, Phylogeny, and Zoogeography VINCENT J. MAGLIO HARVARD UNIVERSITY VOLUME 141, NUMBER 1 CAMBRIDGE, MASSACHUSETTS, U.S.A. 17 DECEMBER 1970 PUBLICATIONS ISSUED OR DISTRIBUTED BY THE MUSEUM OF COMPARATIVE ZOOLOGY HARVARD UNIVERSITY Bulletin 1863- Breviora 1952- Memoirs 1864-1938 JoHNSONiA, Department of Mollusks, 1941- OccAsioNAL Papers on Mollusks, 1945- Other Publications. Bigelow, H. B., and W. C. Schroeder, 1953. Fishes of the Gulf of Maine. Reprint, $6.50 cloth. Brues, C. T., A. L. Melander, and F. M. Carpenter, 1954. Classification of Insects. $9.00 cloth. Creighton, W. S., 1950. The Ants of North America. Reprint, $10.00 cloth. Lyman, C. P., and A. R. Dawe (eds.), 1960. Symposimn on Natural Mam- malian Hibernation. $3.00 paper, $4.50 cloth. Peters' Check-list of Birds of the World, vols. 2-7, 9, 10, 12-15. (Price list on request. ) Turner, R. D., 1966. A Survey and Illustrated Catalogue of the Teredinidae ( Mollusca: Bivalvia ) . $8.00 cloth. Whittington, H. B., and W. D. I. Rolfe (eds.), 1963. Phylogeny and Evolution of Crustacea. $6.75 cloth. Proceedings of the New England Zoological Club 1899-1948. ( Complete sets only. ) Publications of the Boston Society of Natural History. Publications Office Museum of Comparative Zoology Harvard University Cambridge, Massachusetts 02138, U. S. A. © The President and Fellows of Harvord College 1970. WEST INDIAN XENODONTINE COLUBRID SNAKES: THEIR PROBABLE ORIGIN, PHYLOGENY, AND ZOOGEOGRAPHY VINCENT J. MAGLIO CONTENTS Al)stract 1 Introduction 1 A^e versus Haliitus 9 The Species Assemblages 9 cautJierigerus species assemblage 10 melanotus species assemblage 27 (ludrcae species assemblage 32 fiiiicreus species assemblage 36 A Problematical Cenus (laltiis) 48 Conclusions 48 Acknowledgments 50 Literature Cited 50 Appendix „_ 52 ABSTRACT The relationships between the thirty-three spe- cies of xenodontine snakes in the West Indies are reviewed primarily on the basis ot osteological and hemipenial morphology. Foiu" species assem- blages are recognized, distinguislied by the shape of the frontal and prefrontal bones and b\' the structure of the hemipenis. Within the caniheri- gcnis species assemblage three genera are recog- nized-A/.so/j/u's, HypsirJujnchus, and Uromacer. It is suggested that this group entered the West Indies from South or Central America, deri\ed from a primitive form of Alsophis. The South American species Alsophis chamissonis appears to be a relict of that primitive stock. The mainland genera Philoditjas and Conophis appear to be later specialized descendants from that same early stock. The three Galapagos species formerly referred to the genus Dromicits {hiserialis, dorsalis, and sk'vini), are placed in the genus AlsopJiis and regarded as close to the primitive mainland forms. The relationships of the genus laltris remain uncertain, but descent from West Indian Alsojiliis is reasonable. The mekinotits species assemblage has not progressed into the West Indies beyond the northern Lesser Antilles, and has almost certainly been derived from the mainland Leimadophis- Liophis-Lygophis complex. The generic name Dioniiciis is applied to these West Indian forms with the name Lcimadopliis as a junior synonym. Two species, andreae and parvifrons from Cuba and Hispaniola, respectively, share a number of osteological characters with Alsophis, but are like Dromicus externally. The hemipenis is of the Alsophis type and unlike that of Dromicus. Be- cause of their peculiar combination of characters these two species cannot readily be accommodated in any existing genus. The name AntiUopliis nov. gen. is proposed for them. It is suggested that they may be closely related to the mainland form Lijgophis hoursieri while the type species of that genus, L. lincatiis, appears to be closer to Dromi- cus. Eight species formerly placed in the genera Arrhytoii, Dromicus, and Darlingtonia, are con- sidered to form the funcrcus species assemblage. Except for the retention of Darlingtonia for hactiana, the species of this group are referred to the genus Arrhyton. A close relationsliip to the mainland genus Rhadinaca is postulated, and it is suggested that the two genera may have been derived from a connnon ancestor. The osteological similarities between these two groups are dis- cussed in terms of general seniiburrowing adapta- tions and are compared with other semi])urr()wiiig to burrowing New World colubrid snakes. It is concluded that these similarities represent a plnlo- genetic relationship rather than morpliological convergence. Four oversea colonizations from the mainland and numerous inter-island dispersals are recjuired to explain the recent West Indian fauna and its present distribution. INTRODUCTION Th(' \Vest Indies today contain an en- demic snake fauna of modest size. In the absence of an adequate fossil record, any Bull. Mus. Comp. Zool., 141(1): 1-54, December, 1970 1 BuUctiii Miisciiiu of Conipcnalive Zoology, Vol. 141, No. 1 Cuba 0 I— SCALE 100 200 ' ■ MILES '° B] Great Inagua Anguilla \ Antigua S^Cro.x\ \Guadeloupe Puerto Rico %°"^\ \G '7 jAj ., .. , f^Dominica Nevisl V(. Montserrat/^j/ Marie Galante Martinique^'^St. OLucia Borbd^os Grenada, "^-0 Tobago Trinidad^ Fig. 1. Map of the West Indies in Mercator s projection. discussion of relationships within this group and of its history must ultimately be based on the inferred relationships of livinci; species. It is the purpose of the present study to examine the West Indian species of the subfamily Xenodontinae (sensii Dunn, 1928) of the family Colubridae, with reference to their origin, phylogeny, and zoogeography, so far as these can be de- duced from their anatomy and distribution. The only previous attempt to consider a large segment of this group was that by Dunn, 1932, but his work concerned only the Greater Antilles. Dimn relied heavily on the number of sensory pits present on each of the dorsal body scales and conse- quently recognized two basic generic groups in the West Indies; Ahophis was distinguished as having two pits per scale, and Dromicus only one pit. From these two groups Dunn derived all of the other endemic genera of the Greater Antilles. He also examined the dentition and hemipenis. concluding that these did not contradict his proposed relationships. However, ex- amination of Lesser Antillean and mainland species in the present stud}', as well as a re-evaluation of all West Indian xeno- dontines, does contradict these conclusions. Dimn weighted his characters in such a way that several well-defined groups of species remaincxl imrecognized. The xenodontine fauna of the \Vest Indies consists of the thirty-three species and their subspecies listed in Table 1. Tretanorhinus tjariahilis ssp. occurs on Guba, and appears to be a recent immi- grant from Gentral Anu^rica where se\'eral closely related species occur; it will not be considered further here. The remaining thirty-two species — except for ''Leima- doplm' mclanotus which occurs both on Trinidad and on th(> mainland (see Fig. 1 for map) — are endemic to the West Indies and form the basis of this work. All except I a It lis parislii ha\'e been examined. They West Indian Xenodontine Colubrid Snakes Maglio Table 1 Checklist and distribution of West Indian and some mainland^ and Galapagos^ xenodontine colubrid snakes. generic assignments recognized previous to this paper and recognized in the PRESENT STUDY ARE GIVEN FOR COMPARISON. SpECIES ARE ARRANGED IX ALPHABETICAL ORDER. Generic assignment Species Previous Present Distribution {almadensis) andreae andreae andrcac Diclopliyna andreae nehidatus andreae orientalis andreae ))eniu.stdae (an^iistilineatu-s) anomalus ater antillen.sis antillensis antillensis antiquae antillensis nianselli antillen.si.s sanetuntni antillensis sibonius ( bi.seriali.s) callilaennis cantherigerus cantherigerus canthcrigerus adspersus canth erigenis h rooksi cantherigerus caijmamts cantherigerus fiiscicauda cantherigerus pepei cantherigerus ruttiji cantherigerus sell wartzi catesbyi ( cluiniissiDiis) cursor dolichurum dorsulis dorsalis (dorsalis) exiguus cxiguus exiguus stahli exiguus subspadix ferox fremitus funereus haetiana haetiana liaetiana perfector Jul id e jidiae juliae co])eae jidiae mariae melanichnus melanotus ornatus (ixyrhynchus parishi parvifrons parvifrons Leimadophis Dromicus Drotiiicus Alsophis Alsopliis Alsophis Dromicus Dromicus Alsophis Uroniacer Dromicus Dromicus Arrhyton lultris Uromacer Dromictis Dromicus Ilypsirhynchus Uromacer Dromicus Darliugionia Dromicus Alsophis Leiniado])his Dromicus Uronuicer laltris Dromictis Dromicus Antillophis nov. gen. Alsophis Alsophis Alsophis Alsophis Alsophis Arrhyton Alsophis Uromacer Alsoiihis Dromicus Arrhyton laltris Uromacer Alsophis Arrhyton Ilyiisirhynchus Uromacer Arrhyton Darlingtonia Dromicus Also 1)1 1 is Dromicus Dromicus Uromacer laltris Antillophis nov. gen. Brazil Cuba Cuba Isle of Pines Cuba Cu])a Peru Hispaniola Jamaica Guadeloupe Antigua Montserrat Les Saintes Dominica Galapagos Jamaica Cuba Cuba Swan Island Grand Cayman Cayman Brae Cuba Little Cayman Cuba Hi.spaniola, Tortue Island, Vache Island, Gonave Island Chili, Argentina Martinique Cuba Hispaniola Gonave Island Galapagos American Virgin Islands Puerto Rico Puerto Rico Hispaniola Hispaniola, Beata Island Jamaica Hispaniola Hispaniola Dominica Guadeloupe Marie Galante Hispaniola Trinidad, Tobago, Soutli America St. Lucia Hispaniola Hispaniola Hispaniola ^ Non-West Indian species are enclosed in parentheses. Bulletin Miisciiiti of Coinparative Zoology, Vol. 141, No. 1 Table 1 ( Continued ) Generic assignment Species Previous Present Distribution parvifrans ciUeni parvifrou.s- lincolni IKirvifroii.s /i/^'cr l)(inifroiis jxiKiiii^cr parv if runs prat en us parvifron.s- rosanionde panifrons sii/giiis pdnifions tortugensis perfuscus po]\llepi.s ])()itoriccn.sis poitoiiccnsis- poiioriccnsis ancgadac puifoiircnsis a})luinl\is lioitoiiccii.si.s niclmtsi pcrtoriccn.si-s pn/nntii.s purtonccnsis richardi portoriccnms varicgatiis rijger.sntdi nifiroitiis sancticntcis (slevini) (tachymenoidcs) tacniatiun variahilis vHtdfitni fUtatinu vittdttnii Idiidoi vudii I' lid a I lid a allciiiiiiis vudii picticep.^ vudii rainetji vudii utawanae Lciiiiddopliis Droniicu.s Also)}his Ahophis Alsophis Also])hi.s Droiiiicus Drondcus Arrhijton Trctdnorhinus Arrhijton Alsophis Droviicus Arrhijton Alsophis Alsophis Also I )h is Alsophis Also I )h is Ahophis Arrhyton (not considered) Arrhijton Alsophis Gonave Island Hispanio]a Hispaniola Hispaniola Hispaniola Vachc Island Hispaniola Tortue Island Barbados Jamaica Puerto Rico Anegadae Vieques Buck Island Puerto Rico St. Thomas Mona Island Anguilla Bank Islands St. Kitts, Saba, St. Eustatius, Nevis St. Croix Galapagos Peru Cuba Cuba Cuba Cu1>a Gri'at Baliama Bank Islands Grand Bahama Island Bimini Island Crooked Island Great Inagua Island arc divided into four species assemblages on the basis of skull, heniipenial, and external characters, and will be treated within thes(> groups in the following pages. All described subspecies except "Dioiiicus' amireae pen/n.sfi/ae and "D." o. meloplujrrlia from Cuba, "Dromicus" jiiliae copeoe from Guadeloupe, and "D." exis until probable relationships have been assessed and assemblages of probable generic \'alue can be recognized. Changes in nomenclature are made only where West Indian Xenodontine Colubrid Snakes • Maglio 5 s soc Fig. 2. Skull of Alsophis conf/iengerus (MCZ 44874) showing general relationships of bones in xenodontine colubrid snakes. A, dorsal view; B, lateral view; C, ventral view. Left tooth-bearing elements removed. Abbreviations: bo, basioc- cipitcl; bs, basisphenoid; d, dentary; ec, ectopterygoid; eo, exoccipital; I, frontal; m, maxilla; n, nasal; p, parietal; pi, pala- tine; pm, premoxiila; po, postorbital; pr, prootic; pri, prefrontal; ps, parasphenoid; pt, pterygoid; q, quadrate; s, stapes; sa, surangular; sm, septomaxilla; soc, supraoccipital; st, supratemporal; tc, trabecular canal; v, vomer. Approx. X 7. 6 Bulletin Museum of Comparative Zoology, Vol. 141, No. 1 necessary for consistency and for the logical implementation of generic concepts as developed here. In a final section I use the postulated relationships to suggest a possible interpretation of the zoogeographic history of what I consider to be the valid groups in the West Indies. It is my feeling that snake genera are oversplit and not comparable to genera in other reptilian groups. This is a subjective judgment, however, and certainly not con- sistent with the philosophy and usage of the majority of herpetologists working on snakes. Since it is certainly desirable that taxonomic usage within the West Indies conform to that customary elsewhere, I have conserx'atively retained generic names (e.g., Darliniifonia) though I believe them to be of limited usefulness. In one case I have raised to generic rank a species group which, on current taxonomic usage, cannot be accommodated within any other existing genus. As Darlington (1938) and Simpson ( 1956 ) have discussed, the islands of the Caribbean do not appear to haxe been connected with the mainland during the later Cenozoic and overseas migration best explains the available faunal evidence. This hypothesis seems best also to explain the present data. Characters utilized The choice of specific skull characters was made only after more than 200 skulls of West Indian and mainland species had been examined to determine which char- acters were least variable within a species and to discover which ones could therefore be used to infer possible relationships with other species. The characters so chosen include^ the following: 1. The number of teeth which, in several of the species assemblages recognized here, show trends of reduction or increase from species to species (see Figs. 6, 20, and 29 below and the appendix). 2. The frontal bones (see Fig. 2 for labeled skull) also show considerable variation in shape within the West Indies, but prove to have distinctive proportions in certain assemblages, long and narrow on one group and nearly s(|uare in another. Figure 3 shows the distribution of length/ width ratios for the frontal bones in 29 West Indian species. The four groups of species labeled A, B, C, and D represent the species assemblages recognized here on the basis of all characters combined. Nevertheless, even with respect to frontal shape alone, it is clear that assemblage C may be distinguished from assemblages A and D in having a frontal pair that is roughly as wide as it is long. In assem- blage A, most of the species have a frontal pair that is one and one-half to two times as long as wide. 3. The shape of the prefrontal bone was found to be very reliable in separating assemblages in most cases, but (juanti- fication of this character was difficult. In Figure 4 the same 29 species as in Figure 3 are compared with respect to the length/ width index for the prefrontal bone. Al- though assemblages B and C overlap) completely, both exhibit a considerably longer and narrower prefrontal than in assemblage A (Fig. 5). Although there is a certain degree of variation with respect to shape of the prefrontal and other bones, they remain sufficiently distinctiv^e in each group to be taxonomically useful. Other characters such as the parasphenoid width and skull proportions are discussed under each species assemblage. Fig. 3. Length/width indices for the frontal bone pair of 29 species of West Indian colubrid snakes. A, conther/gerus species assemblage; B, me/anofus species assemblage; C, lunereus species assemblage; D, andreae species assemblage. Hori- zontal line, observed range; vertical line, mean; open rectangle, ± one standard deviation from the mean; solid rectangle, 95 per cent confidence interval of the mean. West Indian Xenodontine Colubrid Snakes • Maglio o ooo oooooooooooo aD0>O — cviro'tio <£>r^oo o>p — cm J L o o o o to ^ in to cvJ cvi cvj cvj cvi cj oj J \ I I \ \ L canfherigerus N=I2 vudi! N = 6 ater n=3 anomalus N=2 • melanichnus n=I -ig^„^ — portoricensis n=9 -1- sancticrucis N=2 -i- rufivenfris N=2 1 — rijgersmai N=2 1 00 I - antillensis N=6 -^ ferox N = 4 catesbyi n=7 ^i^^m^^- oxyrhynchus N=2 ' dor sails N=4 — I frenatus N=6 -i _L melanotus N=3 -L perfuscus n=2 -L- ornatus N = 2 -L cursor N=2 —1— yu/Zoe N = 4 i- funereus n = 2 . polylepis N=2 callilaemus N=2 — L hoetiana N=3 — L. exiguus N=4 vittatum N=7 — I/, landoi N = 3 T D 1 taeniatum n=2 I andreae N=2 parvifrons N=i3 O "1 — r "1 — r T ^ ^ 1 1 1 ^ \ \ \ \ OOOoOOOOOOOOOOOOOOO oooo— cvjro^iocot^oooo — cji*J'*;^ are based on dissections in situ of the une verted organ. Terminology is after Dowhng and Savage (1960). It must be emphasized that in this study most measured parameters show various degrees of overlap in range for various species. This does not, however, lessen their value in attempting to recognize phylogenetic relationships through an analysis of moiphological similarities. If an insular series of closely related species has been successively derived by island- hopping in a linear sequence, we might expect any two adjacent forms to show a greater similarity, barring extreme special- ization, than the two geographically terminal species of the series. The characters used here are those that combine relatively little intraspecific vari- ability and enough variation between spe- cies to be useful in the study of intrageneric relationships. Certain characters such as an apical awn on the hemipenis, the shape of the prefrontal bone, or certain skull pro- portions appear to be constant within groups of species that show a close relation- ship in totality of characters combined, and are therefore considered to be of maximum value in indicating true affinities. Others, such as the number of pits on each dorsal body scale, have sometimes proved incon- sistent \\'ith the majority of other traits and have therefore been considered un- reliable at the le\'el of major groups. ^ AGE VERSUS HABITUS When attempting to establish relation- ships between extant forms based entirely on morphological considerations, it must be reasonably certain that differences are not the result of allometric phenomena. Likewise, it must be reasonably clear that apparent similarities are not the result of habitus rather than a close phylogenetic relationship. Juvenile specimens, when available for comparison with the corresponding adult forms, show a number of consistent differ- ences in the structure of the skull which appear to be related solely to size. The major differences may be summarized as follows : Juvenile skull Skull relatively wide Quadrate thin and triangular Supratemporal relatively short Crests low and rounded Postorbital small, non-projecting Orbital foramen very large Pterygoids short, not projecting beyond foramen magnum Maxilla relatively lightly built Bones of brain case thin Adult skull Skull relatively narrow Quadrate with rodlike shaft Supratemporal relati\'ely long Crests high and sharp Postorbital large, projecting Orbital foramen small Pterygoids long, projecting far beyond foramen magnum Maxilla relatively massive Bones of brain case thick It is clear that these characters should not ordinarily be given high taxonomic weight unless the comparison is between two species of comparable adult size. Similarities resulting from habitus adap- tations are more difficult to establish be- cause the mode of life of these species is so poorly understood, and also because the adaptive significance of certain characters, such as scale pits, is at present unknown. Character convergence resulting from habitus similarity can, however, be inferred if a large number of characters are studied together. This point will be discussed in greater detail under the funereus species assemblage below. THE SPECIES ASSEMBLAGES Four species assemblages have been dis- tinguished among the 32 species of West Indian xenodontine snakes here under study. These are characterized on the basis of a number of traits as follows: 10 Build ill Museum of Cotu para five Zoology, Vol. 141, No. 1 Prefrontal Frontals Hcmipenis Size long anteroposteriorly long and narrow no disk large short anteroposteriorly short and narrow apical disk medium short anteroposteriorly long and narrow no disk medium short anteroposteriorly square no disk small cantherigerus assem])lage inelanotus assemblage undrcae assemblage funereus assemblage CANTHERIGERUS SPECIES ASSEMBLAGE Included West Indian species: anomalus Peters, Hispaniola; antillensis ( = leu- comdasY Schlegel, Antigua, Montserrat, Guadeloupe, Les Saintes, Dominica; ater GossE, Jamaica; cantherigerus- Bibhon, Cuba, Isle of Pines, Swan Island, Grand Cayman, Little Cayman, Cayman Brae; cateshiji Schlegel, Hispaniola, Tortue Is- land, Vache Island, Gonave; dor.salis Dunn, Gonave; ferox Gunther, Hispaniola; fren- atus GiJNTHER, Hispaniola; mehnichnus Cope, Hispaniola; oxyrliynchus Dumeril AND BiBRON, Hispaniola; portoricensis Rein- HARDT AND LuTKEN, PucrtO RicO, Moua Island, Virgin Islands; rijgersmm Cope, Anguilla Bank Islands; rufiventris Dumeril AND BiBRON, Saba, St. Eustatius, St. Kitts, Nevis; .sancticrucis Cope, St. Croix; viidii Cope, Great Bahama Bank Islands, Great Inagua. Osteology. The present group of species may be distinguished from other West Indian xenodontincs by a number of skull features, the most characteristic of which is the shape of the prefrontal bone (see Fig. 5). Here this element is wide antero- posteriorly with a broad and strongly con- vex anterior edge. The lower margin of this anterior edge cun^es medially aboxe the lacrimal foramen, so that the latter opens anteroventrally. ^As discussed by Schwartz (1966: 178), Brongersnia's (1937) analysis of Schlegel's co- types and his choice of the Guadaloupe-like speci- mens as the lectotype of Ahophis antillensis have reduced the name leucomclas to the junior synonymy of antillensis. - Senior synonym of angulifer; see Smith and Grant, 1958. Within the assemblage, the species cantherigerus (Cuba) has the lowest num- ber of teeth, with an average dental formula of about 12 -t- 2 maxillary, 10 palatine, 26 pterygoid, and 19 dentary teeth (see Fig. 6 and the Appendix for vari- ation). The skull (Fig. 7) is long and the cranium is moderately deep dorsoventrally. The frontals are widest anteriorly where they make contact with the prefrontals and are strongly emarginated above the orbits. A short, stout postorbital bone is articulated in a deep notch on the parietal bone in such a way that a prominent flange or lateral extension of the parietal intervenes between the postorbital and the frontal bones (see Fig. 2). A weak, but clearly visible groove marks the dorsal midsagittal line of the parietal bone. The parasphe- noid, forming the midventral surface of the skull, is narrow and has a deep trabec- ular canal on each lateral surface. This groove extends from the orbital foramen to the nasal capsule. A dorsal extension of the parasphenoid bone above the trabecular canals separatees the two orbits, forming a thin intcrorbital partition.' The supratemporal is strong and curved, and extends some distance beyond the occiput. The cjuadrate is long and straight. The species vudii on the islands of the Great Bahama Bank does not significantly differ ostcologically from cantherigerus. The dental formula is about 12 + 2 maxil- lary, 10 palatine, 24 pterygoid, and 21 dentary teeth for vudii vudii and is roughly comparable to that of cantherigerus. A peculiar feature of vudii is the melanic ^ Equals frontal crests of Underwood, 1967. West Indian Xenodontine Colubrid Snakes • Maglio 11 Fig. 5. Comparison of the three prefrontal bone types found in the four species assemblages of West Indian colubrid snakes as discussed in the text. A, conther/gerus assemblage type [Ahophis confher/gerus, MCZ 11200); B, me/onofus and andreae assemblage type (Antillophis parvi- frons, nov. gen., MCZ 77227); C, funereus assemblage type {Arrhyton polylepls, MCZ 81020). For each: left, lateral view; right, anterior view. Not to scale. Approx. X 10. tissue lining the cranial cavity of every specimen examined. This tissue imparts a bluish gray color to the skull. A similar condition occurs occasionally in cantheri- genis, and also in cateslnji, dorsalis, frenatus, and oxijrhijnchus where it is the usual condition. Although only two specimens including the type were available for study, vudii utowanoe from Great Inagua differs from all the subspecies of vudii on the Great Bahama Bank in several characters. The nasal bone is distinctive in shape, but this character appears to be somewhat more variable than most skull characters and is therefore of uncertain significance. In its dentition, however, vudii utowanae is quite distinct from the other subspecies. The dental formula is about 15 + 2 maxillary, 13 palatine, 30 pterygoid, and 23 dentaiy teeth, and is greater for every tooth-bearing element. When additional specimens of utowanae become available, a greater de- gree of overlap with the Bahama Bank forms may become evident. However, utowanae still will largely lie outside the neatly clustered range for the other sub- species of vudii. As in vudii vudii, the skull of utouanae appears bluish gray due to the melanic tissue lining the cranial cavity. On Jamaica the species ater has a dental formula higher than that of eantJierigerus. The skull is generally flatter (Fig. 8) and, as a result of this flattening, the nasal bones are closer to the septomaxilla, and the frontal bones touch the trabecular canals ventrally so that the interorbital partition is very small, consisting only of that portion of the parasphenoid bearing the trabecular canals. The frontal bones are relatively shorter and wider (Fig. 9) in contrast to the long, narrow frontals of the Cuban species. In all other skull characters the two forms are very similar. The septo- maxillae are expanded anteriorly and widen immediatelv behind the premaxilla (Fig. lOB). The name capistrata, introduced by Gosse (1851: 373) for a patterned form from Jamaica, was synonymized with ater by Boulenger (1894) without comment. Two specimens of this form from the British Museum were made available to me for comparative puiposes. Both are smaller than typical ater and differ from it in a number of osteological characters which in other species are related to 12 Bulletin Miisciiiii of Coiujniiafivc Zoolofnj, Vol. 141, No. 1 MAXILLARY TEETH — cvjrO'j-mtflr-oooo— c^ff' — — — fsj OJtMCVJ 1 I I I I I I \ I L canfherigerus N = I2 — — vudii N = 4 • vudii utowanae N=2 • ater N = 4 • anomalus N-2 • melanichnus N-l ^—^—. portoricensis N=IO ^— sancticrucis N=2 — rufiventris N=2 — rijgersmai N-2 antillensis N=6 — ferox N=3 catesbyi N = 6 oxyrhynchus N-7 — dorsalis N=2 frenatus N = 4 1 — I — \ — rT~i — r~\ — n i i r~ — cjro^intON-oooO — cMto CVJOJOJM PTERYGOID TEETH f^ooo^O — cjro^iou)h-00CT)O — cvj cJcJcvjc\Jc\JcMc\Jc\Jc\JOJroroio I I I I I I \ I I I I I I I I I vudii canfherigerus vudii utowanae — ater • anomalus • melanichnus portoricensis sancticrucis rufiventris rijgersmai antillensis ferox -— catesbyi oxyrhynchus • dorsalis — frenatus I I I I I I I I I I I I h-oo0— cvjrO'^-ioUJt^ajaiO— cj cvJOJCvjoJCvJCNJOJoJcocvjfOtOro TT PALATINE TEETH N CO CT) O — cviro^ioUJh- _l I \ I I I I I canfherigerus — vudii — vudii utowanae ater — anomalus • melanichnus portoricensis — sancticrucis rufiventris rijgersmai antillensis • ferox — catesbyi oxyrhynchus . dorsalis . frenatus 1 — I — rn — rn — i i i i — i — r~cocn2z^[2i:2^I^ DENTARY TEETH cDf^ooo^O— cvJro^incor-ooo^O — ojro^io cvjc\JCVJcJojcocvJCMoJcvjror<^rOrOrotO I I I \ \ l_J I 1 I I I 1_J \ \ \ \ 1 L_ canfherigerus — vudii — vudii utowanae ater — anomalus melanichnus portoricensis — sancticrucis rufiventris — rijgersmai antillensis — ferox catesbyi oxyrhynchus •^— dorsalis • frenatus ~\ I I I I \ — \ — \ — \ — \ — \ — \ — \ — \ — \ — \ — \ — I — I — r lOf^oooiO — cviro'^u^tDf^coo o — cjrO'j-io oJOJcvJCMCJOJOJcdOJCO rorOfOtOfOro Fig. 6. Observed ranges of variation in numbers of teeth on each tooth-bearing element for the 15 West Indian species of the contherigerus species assemblage. ontogenetic changes. These inehide rela- capistroto represents a juvenile stage ofj tively narrower frontals, a broad rounded ater, and we may follow Boulenger in cranium, low crests and ridges, and thin rc\garding the two as synonymous, cranial bones. It thus seems likely that On Hispaniola there are eight species West Indian Xenodontine Colubrid Snakes • Maglio 13 5 mm Fig. 7. Skull structure of Alsophis cantherigerus, MCZ 56429. A, dorsal view; B, lateral view; C, ventral view. that may be placed in the present species 28 pteiygoid, and 24 dentary teeth. The assemblage. In the rare form melanichmis\ posterior processes of the vomer are later- the numbers of teeth are higher than in ally expanded into flat plates, oval in shape either cantherigerus or ater- the dental when viewed from below. This character formula is 18 + 2 maxillary, 16 palatine, is not seen in any other West Indian 14 Bulletin Museum of Conipamtive Zoology, Vol. 141, No. 1 oooooooooooo r^oooo — coroT^intDN-QO J III' I I I I \ L^ cantherigerus N = I2 vudii N = 6 ^ ater N = 3 — I — anomalus N = 2 . melanichnus N= I -i^^^a^ — porforicensis N=7 — ' — sancticrucis N = 2 — I — rufiventris N=2 — I— rijgersmoi N=2 -• — fez-ox N = 3 cafesbyi N = 7 oxyrhynchus N=6 -L- dorsalis N = 2 J frenatus N = 4 anfillensis N=6 1 o ■f- O ■~r- o 1 o ■■I — o 1 o 1 o 1 O 1 o 1 o 1 o 1 o N- CO en o -oocDO — c\Jro^tnaturc> alone. However, it is clear from other characters that the use of scale pits to define major groups may result in over- splitting of otherwise closely related as- semblages. With the recognition that scale pits by themselves are useful as taxonomic characters within this group only at the species or species-group level, the West Indian species of the cantherigerus as- semblage (excluding Ilypsirhynchtis and Uromacer) may be considered congeneric with chamissonis, angiistilineatus, and tachymenoides from South America, and with dorsalis, hiserialis, and slevini on the Galapagos archipelago. Smith and Grant (1958) have shown that Bibron's (1843) type of Dromicus was cursor from Martinique. This is a form unrelated to South American "Dromicus," as I will show below. WUh the name Dromicus thus unavailable, the present as- semblage of species is referred to Adsophis with the type antillensis ( = leucomelas) Fitzinger. (See Brongersma, already cited above, for the identity of the name antillen- sis.) The suggested phyletic relationships be- tween the genera and species of the can- therigerus assemblage are shown in Figure 18. Zoogeography. Two lines of evidence indicate a western origin for thc^ cantheri- gerus species assemblage into the West Indies. First, Alsophis cantherigerus from Cuba is the Antillean species most similar to A. chamissonis of the mainland; this similarity is most notable in osteological characters and especially in the dental formula, which in these two species is the lowest of the whole assemblage. The Gala- pagos forms, likewise, have relatively few teeth, as do the suggested mainland deriv- atives PJiilodryas and Conopliis. Within the West Indies a general trend toward increased number of teeth is evident, espe- cially in the specialized arboreal species of Uromacer, and in the portoricensis species group. From the geographic distribution of dental formulae, it would appear that a low number of teeth is primitixe for the mainland ancestor of this assemblage. In other characters also, the more easterly distributed species show a greater diver- gence from the mainland forms, thus tend- ing to support the \'iew that the group entered from the ^\'est. The second line of evidence indicating a western origin is that no member of this West Indian Xenodontine Colubrid Snakes • Maglio 25 CO .CO ^ ^ 05 "co "^ -Q CO O O ■"::: c: to 5q C o ^ -XT ^ C O -S CD to tu S2 . o — cvjro cvj rofororo J I L LJjll I I 4V — melanotus — perfuscus . ornatus — cursor 1 — r—\ — r I I im ih juliae _____ cvjrorOrOfO DENTARY TEETH (Dr-oocnO— cjfO^iouJh-oocno— cvi ro^ cvioj cJCvicvJCJCJcvJCNJCNJrorOfororo I I I I L_L I I I I _LJ L_L melanotus perfuscus .ornatus • cursor — juliae -1 — \ — [—] — I — I — m — I — n — I I I I \ n r ____c>jcvjrvjcvj(\jojcvjc\jcvjcvjrotorof0f0 Fig. 20. Observed ranges of variation in numbers of teetfi on each tooth-bearing element for the five West Indian species of the melanotus species assemblage. often results in the well-known phenom- enon of charaeter divergenee whereby the competmg forms become adapted ( first ecologically, then morphologically) to somewhat different aspects of the environ- ment, thus reducing competition. It is interesting to note that, even in the rela- tively less specialized Hispaniolan species anomalus and melonichnus, character di- vergence has progressed to a remarkable degree, so that these forms lie near the t\\ o extremes of variation for the entire assem- blage in many of their characters (see Figs. 6, 8, 9, and 11). Another waif dispersal from the main- land to the Galapagos almost certainly resulted in the differentiation of Imerialis, (lorsalis, and sJcvini on these islands. Later, rear-fanged specialization within the an- cestral mainland genus pbssibly led to Philochyas and Conophis which displaced their antecedent from much of its former range. The three closely related species, chamissonis, taclujmenoides, and ongusti- Uneofus, have remained as relatively un- modified relicts of the original mainland stock, except for reduction in the number of scale pits. MELANOTUS SPECIES ASSEMBLAGE Included West Indian species: cursor Lacepede, Martinique; julioe (including marlae) Cope, Guadeloupe, Marie Galante, Dominica; melanotus Shaw, South America, Trinidad, ? Grenada; ornatus Gaeman, St. Lucia; po-juscus Cope, Barbados. 28 Bulletin Museum of Comparative Zoology, Vol. 141, No. 1 o cvl o O o IT) O CD o o 00 o (7) o o o — CM 1 CJ CM CM 1 CJ CM 1 CM 1 ro ro 1 1 1 mPlnnniu^ m = o N = 2 _ 1 perfuscus 1 _1_ ornatus N=2 N = 2 1 jullae ^ 1 = 4 — I— o 1 O ro , .J ... O 1- o IT) 1 o CC — r- o r- — I— o CO — |— o 1 1 o o o — Fig. 21. Skull length width incJices for the five West Indiarn species of the melanotus species assemblage. Widths are taken at the otic region of the skull. Osteologxj. This assemblage of species is distributed from the island of Guadeloupe south to Trinidad. It is distinguished from the Alsophis cantherigerus species assem- blage and other Antillean xenodontines primarily by the shape of the prefrontal bone and b\' the structure of the hemipenis (discussed below). The prefrontal is long dorso^'entrally and narrow anteroposteriorly with a sharply pointed anterior projection at about midlength (see Fig. 5). The species for which this group is named, melanotus, occurs on Trinidad and Tobago and has been questionably re- ported from Grenada. It also occurs in Venezuela and Colombia. Like all mem- bers of this group melanotus is about half the size of A. cantherigerus and contrasts with it in the following skull characters: the postorbital region is proportionately longer; the frontal bone is relatively shorter anteroposteriorly with very little emargin- ation above the orbits; the skull is con- siderably more flattened dorsoventrally as compared with its width; as a result of the latter character, the interorbital partition formed by the parasphenoid bone is shorter in its dorsoventral extension; long, thin lateral processes extend back from the premaxilla in contrast to the short, ^^ide based processes of A. cantherigerus. The dental formula for melanotus is about 15 + 2 maxillary, 10 palatine, 24 pterygoid, and 19 dentary teeth (Fig. 20 for variation). The skull of the endemic species per- fuscus on the island of Barbados is pro- portionately longer and narrower than that of melanotus (Fig. 21). The supratemporal and quadrate bones are more elongated and comparatively narrower. The orbit is small. The postorbital bone lies far for- ward on the parietal and is nearly in con- tact with the frontal. In this last character perfuscus is distinct from other members of the present group; in the latter, a wide expansion of the parietal bone separates the postorbital from the frontal. In general skull proportions and in its low dental formula, perfuscus is closer to melanotus and other mainland representatixes of this assemblage than it is to other Antillean species on adjacent islands. The species ornatus from St. Lucia is slightly larger than melanotus (Trinidad) and has a somewhat broader skull. The prefrontal bone is of the melanotus type, but is \\'ider than in that species. In most other skull characters the two forms are very similar except for the dental formula, which is higher in ornatus. Clearly related to the latter is the species cursor from Martinique (Fig. 22), which differs only in the smaller size of the nasal bone. On Guadeloupe, Dominica, and Marie Galante the species juliae has a higher dental formula than cursor, and the supratemporal, quadrate, and premaxilla are greatly re- duced in relative size. A similar reduction is seen in some related mainland species such as bimaculatus, but this probably represents parallelism rather than an inde- pendent derivation from one of these main- land forms. It is suiprising that no member of this group has been reported from St. Vincent. This island forms an important intermediate stepping stone between Grenada and St. Lucia. The mongoose is widespread on this island, and it is possible that a formerly present species, endemic or not, has be- come extinct. External morphology. Externally the melanotus species assemblage is a homo- West Indian Xenodontine Colubrid Snakes • Maglio 29 5 m m J I 1 — \ — I Fig. 22. Skull structure of Dromi'cus cursor, MCZ 6011a reversed). A, dorsal view; B, lateral vievv; C, ventral view. 30 Bulletin Museum of Comparative Zoology, Vol. 141, No. 1 Fig. 23. Hemipenis morphology in Dromi'cus cursor (MCZ 6011) showing the typical apical awn of the melanofus species assemblage; semidiagrammatic. The organ is un- everted and dissected in situ. Approx. X 5. geneous one and in many respects is similar to the cantherigerus assemblage. As in the latter group there are eight supralabials, but here only the fourth and fifth enter the orbit. The number of scale rows may be 17 or 19, and the ventral scales number appro-ximately between 150 and 200. In contrast to Alsophis, the number of sub- caudals is rarely over 100. The anal plate is divided. All of these snakes are moder- ate in size and are roughly between 600 and 1000 millimeters in total body length. A single apical pit is usually present on the dorsal body scales, but may be absent as in jiiliae mariae. Hemipenis. The hemipenis of cursor ( Fig. 23 ) is shorter proportionately than in Alsophis cantherigerus, and the sulcus spermaticus is less deeply divided. As in that species, several rows of stout spines extend along the sides of the organ but, in contrast to it, small spines are also present between the diverging branches of the sulcus. The organ is generally weakly bifurcated and the apical ornament is dis- tinct from all other West Indian xenodon- tines. Here papillae are never present; instead, a series of membranous folds radi- ate from the apex and terminate in a transverse fold of tissue which encircles the tip of each lobe. This fold forms a well-defined apical disk on each lobe when the organ is everted. The sulcus forks and proceeds onto the disk and to the tip of each lobe. The structure of the hemipenis in the other species in this assemblage is essen- tially like that of cursor. Origin and Zoogeography. The melanotus group offers no problem of origin. This well-defined and closely related assem- blage is moiphologically continuous with the widespread series of species currently referred to the genus Leirnadophis, common on the mainland.^ In both its osteology and hemipenial morphology the type species of Leimadophis — L. almadensis — is clearly al- lied to the present group, and there are no external characters which would pre- clude such a relationship. It seems probable that a northward colonization of the Lesser ^ Dromicus anuizouicus is known only from the type specimen, MCZ 2820, and may also be re- ferred to the present group. Its prefrontal bone is essentially of the melanotus type and the hemi- penis has a weak apical disk. In all its osteological characters this specimen may be distinguished from other members of the present group only in ha\ing those features, such as broad skull, tliin cranial bones, low, rounded crests, etc., which normally characterize juvenile specimens. Even the relatively weak disk on the hemipenis probably reflects the youthful condition of the specimen. It is thus quite possible that this form represents an immatmc specimen of a previously described South American species of the present assemblage. West Indian Xenodontine Colubrid Snakes • Maglio 31 SCALE 0 100 200 MILES D. perfuscus Fig. 24. Proposed routes of colonization of the West Indies by the melanotus species assemblage. Distribution of extant species is as indicated. Antilles by a mainland species similar to melanotus occurred in relatively recent times. From this species or its ancestor, ornatus (St. Lucia) was almost certainly derived, possibly via a now extinct inter- mediate species on St. Vincent. Successive overseas migration (Fig. 24) probably re- sulted in cursor on Martinique and jtilioe on Guadeloupe, Dominica, and Marie Galante. The Barbados species, perfuscus, appears to have been derived from one of the islands to the west, perhaps from St. Vincent or Grenada. Taxonomy. As mentioned above. Smith and Grant (1958) have sho\\'n that Bibron's (1843) type of Dromicus was cursor. With cursor and ahnadensis here considered as congeneric, Dromicus Bibron 1843 and Leimadophis Fitzinger 1843 become syno- nymic names for this assemblage. Although the actual dates of publication of these two names remain in question, December 31, 1843, is now to be taken as the official publication date of Fitzinger's Sijstema Reptilium (Smith and Grant, 1958), and Bibron's Dromicus thus becomes the senior synonym for the present assemblage with Dromicus cursor as the type species.^ The genus Dromicus is vei-v similar to the South American genus Liophis Wagler 1830. In all of the characters studied, Wagler's type of Liophis — L. cobella — is close to the present assemblage. The pre- frontal is like that of D. melanotus, the frontal is short with very little emargination above the orbits, and the interorbital par- tition is veiy small. The hemipenis of Liophis has a pair of well-developed apical disks as in D. melanotus, but differs in the presence of basal hooks (Roze, 1964). Body ^ After the present paper was in manuscript form, it was brouglit to my attention that Drs. Albert Sehwartz and Richard Thomas reached the conchision of "Dromicus" cf. cursor - Leimadophis synonymy independently of the present autlior. Their conclusion was reported in a letter to Dr. E. E. Williams. 32 Bulletin Museum of Comparative Zoology, Vol. 141, No. 1 .O Ci CO CO ^ :3 o to :3 CO o CD C3 s^ o CO .D qS E C: :3 ^ Q. o o B Barbados D Dominico L Lesser Antilles M Martinique s St Lucia SA South America — oversea colonization Fig. 25. Suggested phyletic relationships between species of the me/onotus species assemblage and the genus Liophis. Short horizontal lines indicate proposed oversea coloniza- tions. size and scale counts are within the range of Dromicus (present concept). The question arises as to the vaUdity of the generic distinctions between ^'Leima- dophis" (i.e., Dromicus) and Liophis made by Roze (1964: 535). As we have seen, the presence or absence of scale pits may not be as important a distinction as for- merly belie\'ed. Thus, the only major difference between these two genera is in the maxillaiy dentition in which "Leim- adophis" has a diastema with much en- larged posterior teeth, whereas Liophis lacks a diastema and the posterior teeth are little enlarged ( Roze, 1964 ) . I suspect that these two groups will be considered as congeneric when better known, but on present evidence I here treat them as valid genera. The inferred phyletic relationships be- tween Liophis and the West Indian species of Dromicus are presented in Figure 25. ANDREAE SPECIES ASSEMBLAGE Included West Indian species: amlreae Reinhaedt and LxJTKEN, Cuba; parvifrons Cope, Hispaniola. OsteolofS,y. Only two species of this assemblage are extant in the West Indies, andreae on Cuba and parvifrons on Hispan- iola and nearby islands. They are generally considerably smaller in size than Alsophis and are about the size of Dromicus mela- nofus. In cranial osteology amlreae and parvifrons show features characteristic of both Alsophis and Dromicus (Fig. 26). The frontal bones are very long and narrow with a deep emargination above the orbits as in Alsophis, but unlike the pro- portionately shorter and \\ ider frontals of Dromicus (see Fig. 3). On the contrary, in the structure of the prefrontal bone, the present assemblage is close to Dromicus with a long and narrow prefrontal bearing a shaiply rounded anterior process. This is (juite distinct from the relati\'ely wider prefrontal of Alsophis in \vhich the an- terior surface is broadly rounded and the bone is proportionately wider (see Fig. 5). In general skull proportions andreae and parvifrons are closer to Alsophis than to Dromicus. The parasphenoid bone is very narrow as in A. cantheriiierus and other westerly distributed species of Alsophis, as well as mainland forms, but unlike the rather broad shape of that bone in the A. portoricensis species group or in Dromi- cus. The parasphenoid partition between the orbits extends high abo\e the trabecular canals, as in most species of Alsophis except West Indian Xenodontine Colubrid Snakes • Maglio 33 5mm I I I I L Fig. 76. Skull structure of Antillophis parvilrons nov. gen., MCZ 77228. A, dorsal view; B, lateral view; C, ventral view. 34 Bulletin Museum of Comparaiwe Zoology, Vol. 141, No. 1 Fig. 27. Hemipenis morphology in Aniillophis parvitrons nov. gen., MCZ 60064; semidiagrammatic. The organ is uneverted and dissected in situ. Approx. X 5. portoricensis and related species. The j)re- niaxillary bones in andreae and pawifrons lack the long lateral process as in Dromicus. The dental formnlae in these two species is comparable to that of both Alsopliis and Dromicus; that is, about 16 + 2 niaxillaiy, 12 palatine, 26 pterygoid, and 21 dentary teeth in parvifrons, and 21 + 2 maxillary, 16 palatine, 35 pterygoid, and 26 dentaiy teeth in andreae. Nonosteological characters. In external characters andreae and pawifrons are similar to Dromicus. The number of ven- tral scales is generally lower for these species than in Alsopliis and about the same as in Dromicus. The subcaudal scale number in parvifrons is higher than that of andreae and of species of Dromicus, while it is within the normal range for species of Aho})]}is. In the present assemblage, each dorsal body scale bears a single sensory pit in contrast to the two pits in all West Indian species of Alsophis, and in this respect is like most species of Dromicus. It is in the structure of the hemipenis, however, that andreae and parvifrons de- part radically from Dromicus. Here the organ has a deeply dix'ided sulcus sperm- aticus, bordered by a thick fold of spinose tissue. Several rows of stout spines parallel the sulcus, each branch of which terminates on a diskless apex bearing long filiform calyces (Fig. 27). There are no basal hooks, neither are there spines between the branches of the sulcus spermaticus as in Dromicus. This structure is basically like that of Alsopliis and is quite unlike that of Dromicus in \\'hich strong apical disks are present. Taxonomy and Origin. As noted above, the two species of the andreae group ex- liibit osteological features characteristic of l)oth Alsophis and Dromicus, though closer to the former. In most external characters they seem closer to Dromicus, but in the structure of the hemipenis they differ markedly from that group, being extremely close to Alsophis. Taken as a whole, this group cannot easily be referred to either genus. With regard to both skull and hemi- penial characters, these two species re- semble certain members of the South American genus Lygophis. The hemipenis in members of the latter genus, as currently recognized, is very heterogeneous. The organ may possess well-differentiat(>d apical disks as in L. lineatus (Fig. 2SA), the type species, and L. flavifrenatus, or it may be essentially of the Alsophis type, as in L. hoursieri (Fig. 2SB), in which the apical calyces are more spinulate along their margins than in the Alsophis type. Of these three species, Lygophis hoursieri (Ecuador West Indian Xenodontine Colubrid Snakes • Magjio 35 Fig. 28. Hemipenis morphology in the South American genus Lygophis; semldiagrammatic. The organ Is uneverted and dissected in situ. A, Lygophis lineatus, MCZ 80994; B, Lygophis boursieri, MCZ 36948. Approx. X 5. and Colombia) is very close to andreae and parvifrons in both skull and hemipenial characters, as well as in external scale pat- tern. From this incomplete study of Lygophis., it seems possible that we may be dealing with a compound genus of distantly re- lated forms. The evidence seems to sug- gest that L. boursieri might be placed in a separate genus with andreae and parvi- frons as a specialized radiation, perhaps derived from mainland Alsophis; and that the type of Lygophis, L. lineatus, may need to be synonymized with Dromicus from which it differs primarily in th(> lack of scale pits and the presence of basal hooks on the hemipenis (Roze, 1964). However, we must await more detailed taxonomic work to clarify these issues.^ Nevertheless, it seems clear that generic distinctions be- tween Dromicus (present concept) and ^ In a reappraisal of South American snakes related to LijgopJiis boursieri, Myers ( 1969 ) established the boursieri species group including Rhadinaea antioquien,sis, R. tiistriata, and L. boursieri. The group is placed in Lijgophis "... simply because boursieri already resides there." But the artificiality of the resulting genus is clearly pointed out, and the necessity of generic reassignment is affirmed. 36 Bulletin Museum of Comparative Zoology, Vol. 141, No. 1 Lygopliis (see Roze, 1964) rest on slim evidence. Since on present generic concepts it is not possible to accommodate the species of the ondreae group in any existing genus, it seems best to propose a new generic name. Though close similarities exist be- tween andreae and parvifrons and at least part of what is now called Lijr form a square plate above the orbits, unlike the condition in other West Indian groups. Ventrally the parasphenoid is proportion- ately wider along its entire length, especially posteriorly. The skull is pro- portionately flatter, and thus the inter- orbital partition formed by the dorsal extension of the parasphenoid does not extend above the trabecular grooves. Rather, the frontal bone on each side covers the entire lateral aspect of the para- sphenoid. The supratemporal and cjuad- rate are reduced in size, and the latter is flat and triangular in shape. On Jamaica there are three species of this group, funereus, polylcpis, and calli- lacmus. The specific distinction between funercus and polylcpis has recently been demonstrated by Buden (1966), but osteo- logically they are very similar. They have the highest dental formulae of the group (Fig. 29) with about 19 + 2 maxillary, 11 palatine, 19 pter\'goid, and 24 dentaiy teeth in funercus, and 17 + 2 maxillaiy, 11 palatine, 24 pterygoid, and 27 dentary teeth in polylepis. The parasphc>noid bone beneath the orbits is broad throughout its West Indian Xenodontine Colubrid Snakes • Maglio 37 MAXILLARY TEETH ,^oocnO-cM;o^|ncDf-ooo^o-cvj I I I I I 1 I I I I I • funereus N=2 • polylepis N - 2 — colli laemus N-Z — haetiana N-3 exiguus N=4 vittafum N=7 — V. landoi N-3 • dolichurum N = l taenia turn N = 2 1 I r 1 — \ — \ — I — I — r (-ooo^2z£d:25:!e^^?^g^^ PTERYGOID TEETH -r.mOOJ'a-CDOOOCVJ'd-lOCD yJCDii OJCJOJCMCJ I I I I I I I I I I I I • funereus — polylepis . callilaemus haetiana —^^—— exiguus — vittatum • / landoi • dolichurum — taenia turn T1 — \ — I — r I I I I U)Q0r:!:^:z.__c\jCJOJcM(\J PALATINE TEETH iou)t^condently in many groups vmder the in- fluence of similar habitus selection. But the exact way in which parallel characters are achieved, even under identical selection West Indian Xenodontine Colubrid Snakes • Maglio 45 pressures, depends upon raw materials in the form of existing moqohological struc- tures, and upon genetic variability. The more distantly related any two forms are, the more likely it is that they will achieve functionally similar adaptations in a di- vergent way. An examination of semiburrowing adap- tations in xenodontine snakes of the New World shows similar osteological trends common to all of them, but, as expected, they differ from each other in details. In most osteological characters studied, Rha- dinaea and the funereus group exhibit a similarity of form which suggests more than morphological parallelism with re- spect to semiburrowing adaptations. The osteological modifications which generally appear to be associated with semiburrowing adaptations are: small body size; reduction in relative orbit size; short- ening of the parietal region of the skull; enlargement and consolidation of the pre- orbital bones to form firm contacts with each other; broadening of the parasphenoid bone associated with the reduced orbits; relative broadening of the otic region so that it becomes the widest part of the skull (probably associated with general stream- lining ) ; reduction of the supratcmporal and quadrate bones; and, a relatively low num- ber of teeth. With respect to all of these characters, as well as others not obviously correlated with burrowing, the funereus group and Rhodinoea show a close correspondence, differing only in the structure of the pre- frontal bone. Tlie frontal bones form a nearly square plate above the orbits and contact the parietals in a broad, nearly straight suture. In contrast, the parietals of Geophis miititorqids extend lateral to the posterior half of the frontals, occupying a deep groove in the latter. A similar con- dition exists in ApostoJepis am])imii,ra, where the frontals are trapezoidal in shape, being wider anteriorly. In Atractus Jati- frons the frontals are oval in shape, with their long axis in the transverse direction. In the funereus group and Rhadinaea the parietal retains the shape normal for nonburrowing xenodontines. In Geophis mutitorquis a posterior process extends into the deeply divided interparietal, and in Atractus lotifrons and Rhinostoma guia- nense the parietals are so shortened that they are broader than long. The reduced quadrate in the funereus group and Rhadinaea is thin and tri- angular in shape. In Rhinostoma two forms appear to be relatively West Indian Xenodontine Colubrid Snakes • Maglio 47 C3 O Q: to to 03 CD W) =3 E S rx CJ CD o .Cj _2 —J c: — *^ *•*«..» ^ ■ -«^ Qj 03 O o i» Ci >< O CJ T3 ii :^ cb -c: c Cuba CA Central America H Hispaniola J Jamaica p Puerto Rico V Virgin Islands — oversea colonization Fig. 36. Suggested phylefic relationships between the species of the funereus species assemblage and the genus Rhad- inaea. Short horizontal lines indicate proposed oversea colonizations. recent products of speciation on that island. The most primitive member of the genus is A. funereus which forms the base of both the A. funereus-A. toeniatiim series and the A. funereus-A. exiguus series. It is the only Antillcan species (except for A. pohjiepis) with a nonbifurcated hemipenis and in this respect is closest to species of Rliadinaea. Within the A. funereus-A. taeniatum series it is the least specialized in terms of the reduction of skull bones and other semi- burrowing adaptations. From this early A. funereus stock two basic lines appear to have been established; A. coIUkiemus (Jamaica) began a trend toward reduction of the posterior skull bones and in the dentition, and a widening of the para- sphenoid bone. A colonization of Cuba from this early A. calliloemus stock led to the Cuban forms which, through repeated speciation, continued the trend, with A. taeniatum as the most highly specialized species of the group. The second line re- sulted from a migration to Hispaniola and tended toward a narrower parasphenoid, larger size, and in many ways a general convergence toward other W'cst Indian xenodontine groups, especially Dromicus. This may have been related to a general tendency away from semiburrowing adap- tations. In order to derive A. exiguus from this early Hispaniolan form we must postu- late a widespread distribution for this latter form before (or concurrent with) its differ- 48 Bulletin Museum of Comparative Zoology, Vol. 141, No. 1 SCALE 0 100 200 I I 1 MILES D. haetiana A . funereus ^ ■ exiguus ^^ A . poly le pis 0 0 Fig. 37. Proposed routes of colonization of the West Indies by the funereus species assemblage. The arrows are not in- tended to represent exact paths. The distribution of extant species is as indicated. entiation into the morphologically some- what specialized and ecologically restricted species Darlingtonia haetiana. Today D. haetiana occurs only in the montane massifs of the southwestern and Barahona penin- sulas at altitudes ranging from 1000 to 5600 feet. One possible explanation for the peculiar distribution of this species is an ecological replacement of the former wide- spread species (possibly by pawifrons?) with D. haetiana remaining as a montane relict. This zoogeographic pattern is sum- marized in Figure 37. A PROBLEMATICAL GENUS Two species remain to be discussed: laltris (lor.salis and 7. parlshi from Hispan- iola. laltris dorsalis (I have not seen I. pari.shi) Ls a large species and is most similar to Alsophis in many skull char- acters, but distinct in many external and hemipenial characters. The skull is nar- rower, especially in the otic region. In its dentition this species is unique among West Indian xenodontines in having prominent grooves on the enlarged posterior maxillary teeth. The bilobed hemipenis is very long, ridged with numerous folds, and has an apical ornament of weakly developed flounces. Externally there are seven up- per labials as in Arrhijton generally, but the \'entral and subcaudal scale counts are similar to those of Alsophis. laltris dorsalis (and presumably I. parishi also) is not very close to any other Antillean species as far as can be determined from its present morphological specializations and therefore certainly should be retained in a distinct genus. In most characters laltris shows its greatest similarity to Alsophis, and it may have been derived from that genus on Hispaniola. CONCLUSIONS The use of skull and hemipenial char- acters, in addition to those of external West Indian Xenodontine Colubrid Snakes • Maglio 49 Table 2 Distinguishing charactebistics of the eight genera of xenodontine colubrid snakes in the West Indies as discussed in the text. Genus prefrontal frontal hemipenis size supra- labials grooved max. teeth I anal scale no. sensor>- pits loreal scale Ahophis cantherigeriis type long and narrow no disk- large 8 no divided 2(1) present Dromicus mehnotiis type short and narrow apical disk medium 8 no divided 1(0) present AiitiJloiyJii.s- mdanotiis type long and narrow no disk medium 8 no divided 1 present Arrlnjton f tine reus t>pe sqnare no disk small 7-8 no divided 0-1 present (absent in A. taeniatum) Durlingionia fiiuereus square no disk small 7-8 no complete 0 absent type Hijpsirhijnchus cantherigeriis long and no disk type narrow Uromacer cantherigerus long and no disk type narrow large no divided 1 present lalt ris cantherigerus long and no disk type narrow large, 8 no divided 0 present arboreal large 7 yes dixided 0 present to be of significant suggesting relationships between morphology, appears aid in species of West Indian xenodontine eohi- brid snakes. They not only provide data for a proposed redefinition of generic con- cepts, but suggest certain phylogenetic relationships with mainland groups. Such relationships are of considerable interest, since they allow a tentative reconstruction of the possible origin and history of these snakes in the Antilles. The generic groups of xenodontine snakes here recognized in the West Indies and listed in Table 1 may be distinguished as in Table 2. The present xenodontine fauna of the West Indies was possibly wholly derived from Central and South American stocks through at least four oversea colonizations. Based on present evidence, a summaiy of the postulated historical events follows: 1) From the formerly widespread South American genus Alsophis, a waif coloni- zation established this group on Cuba. Sub- sequent radiation into a number of species and endemic genera led to its present distribution throughout the Greater An- tilles and the northern Lesser Antilles. A minimum of three separate inter-island migrations of this group is required to explain the peculiar faunal assemblage of Hispaniola. 2) The specialized genus laJtris possibly emerged from Alsophis on Hispan- iola. 3) Using Jamaica as a port of entry and center of dispersion, a single stock, possibly derived from the Central Ameri- can genus RhacUnaea, successfully spread through the Greater Antilles in two distinct but closely related lines and established the genera Arrliyton and Dorlingtonia. 4) Pos- sibly derived from part of what is now called Lijg,oph\s in Soutli America, the species andreae and parvifrons may have reached Cuba and Hispaniola by a direct oversea colonization. 5) A relatively recent invasion of the Lesser Antilles by a species of Dromicus ( = Leimadophis) almost cer- 50 Bulletin Museum of Comparative Zoology, Vol. 141, No. 1 tainh' (Mitered vhx Trinidad, but has not yet progrt^ssed be>'ond (iiiadcloupe. The chronological sequence of coloni- zations cannot definitely be established on present evidence. However, a sequence roughh' similar to that above is not un- reasonable. It, of course, cannot be as- sumed that the West Indies were devoid of a xenodontine ophi fauna before the series of colonizations that established the present fauna, but our knowledge of earlier xenodontine colonizers is nonexistent be- cause of the lack of a significant fossil record. From the patterns of dispersion discussed in this paper it would appear that numerous combinations of inter-island migrations have occurred. The main sequences have progressed from one island to the next adjacent island and in this sense were for the most part linear. The following series have been proposed: mainland-Cuba- Hispaniola-Puerto Rico-Lesser Antilles; mainland-Cuba-Jamaica-Hispaniola; main- land-Cuba-Bahamas; mainland-Jamaica- Cuba; mainland-Jamaica-Hispaniola-Puerto Rico; mainland-Trinidad-Lesser Antilles. Inter-island migration, especially to cen- trally located Hispaniola, seems to have been more frec^uent than mainland-island migrations. This was certainly the result of the greater cross-water distance between the mainland and any island than between the various islands themselves, as Simpson (1956) and Darhngton (1957) have sug- gested. The greatest diversity in species and genera occurs on Hispaniola; this is to be expected in view of its central position and consequently greater number of coloni- zations. Its large size, varied habitats, complex physiography and history have prox'ided an excellent opportunity for im- migrants to differentiate into noncompet- ing forms. The zoogeographical patterns here pro- posed are based on limited evidence and are in large measure speculative. It is hoped that they offer a workable contri- bution toward the continued study of this group. However, only when adequate in- formation about the comparative anatomy, karyotypes, ecology, physiology, and bio- chemistry of all Antillean snakes and their mainland relatives is available will we be able to draw firmer conclusions concerning the origin and zoogeography of xeno- dontine snakes in the West Indies. ACKNOWLEDGMENTS I wish to express my deep gratitude to Professor Ernest E. Williams for suggesting the problem which began this work, for innumerable discussions and suggestions, and for his constant encouragement. With- out his interest and assistance this study could never have been completed. For their reading of various versions of the manuscript and for many useful com- ments and criticisms, I extend my thanks to Drs. Richard Estes, George Gorman, Max Hecht, Edmond Malnate, Charles Myers, James Peters, Albert Schwartz, Richard Thomas, Paulo Vanzolini, and Ernest Wil- liams. The conclusions reached in this paper and any errors which remain are, of course, solely my responsibility. I am grateful to the following persons who kindly loaned specimens from their collections and ga\e permission to extract skulls: Dr. Richard Zweifel, Miss A. G. C. Grandison, Drs. Edmond Malnate, Albert Schwartz, and Ernest Williams. I wish thankfully to acknowledge Mr. Laszlo Meszoly \\'ho prepared the maps, graphs, and skull drawings, and Miss Catherine McGeary and Mrs. B. Gail Browne who typed various versions of the manuscript. The research was supported in part by National Science Foundation grants nos. GB-6944 and NSF B 019801X. LITERATURE CITED Barbour, T., and B. Shrevk. 1938. Concerning some Bahamian reptiles with some notes on' the fauna. Proc. Boston Soc. Nat. Hist., 40 (5): 347-366. BiBROK, G. 1843. Reptiles, in Ramon De La Sagra, Histoire Physique PoHtique et Natur- elle de I'ile de Cuba, pp. 1-239. West Indian Xenodontine Colubrid Snakes • Maglio 51 BouLENGER, G. A. 1894. Catalogue of the snakes in tlie British Museum (Natural History). London, vol. 2: 1-382. . 1896. Catalogue of the snakes in the British Nhiseum (Natural History). London, vol. 3: 1-727. Broxgersma, L. D. 1937. The types of Psam- moi)lii.s antillensis Schlegel, 1837. Heipeto- logical Note XIV, Zoologische Mededeelingen, 20: 1-10. BuDEN, D. W. 1966. An evaluation of Jamaican Dromicus with the description of a new species. Breviora, No. 238: 1-10. Cochran, D. M. 1941. The Herpetology of Hispaniola. Bull. Smithson. Inst., U. S. Nat. Mus., No. 177: 1-398. Cope, E. D. 1862. Synopsis of the species of Holcosis and Ameiva, with diagnoses of new W. Indian and S. American Colnbridae. Proc. Acad. Nat. Sci., Philadelphia, 1863: 60-82. Darlington, P. J., Jr. 1938. The origin of the fauna of the Greater Antilles, with discussion of dispersal of animals over water and through the air. Quart. Rev. Biol., 13: 274-300. . 1957. Zoogeography. New York: John Wiley and Sons, 675 pp. Dowling, G. H., and J. M. Savage. 1960. A guide to the snake hemipenis: a survey of basic structure and systematic characters. Zoologica, 45 (1): 17-27. Dunn, E. R. 1928. A tentative key and arrange- ment of the American genera of Colnbridae. Bull. Antivenin Inst. Amer., 2(1): 18-24. . 1932. The colubrid snakes of the Greater Antilles. Copeia, 1932 (2): 92-98. Gorman, G. C, and L. Atkins. 1969. The zoo- geography of Lesser Antillean Atioli.s lizards- an analysis based upon chromosomes and lactic dehydrogenases. Bull. Mus. Comp. Zool., 138 (3): 53-80. Gosse, p. H. 1851. Naturalists' sojourn in Jamaica. London: Longman, Brown, Green, and Long- mans, pp. 1-508. Horn, H. 1969. Polymorphism and evolution of the Hispaniolan snake genus Uromaccr (Colnbridae). Breviora, No. 324: 1-23 Lanix), R. v., and E. E. Williams. 1970. Notes on the herpetology of the U. S. Naval Base at Guantanamo Bay, Cuba. Studies of the fauna of Curacao and other Caribbean islands. No. 116: 159-201. Mertens, E. 1939. Herpetologische Ergebnisse eine Reise nach der Insel Hispaniola, Westin- dien. Abh. Senckenberg. Naturforsch. Ges., No. 449: 1-84. Myers, C. W. 1967. The pine woods snake, Rhadinaca flacilata (Cope). Bull. Florida State Mus., 11 (2): 47-97. . 1969. South American snakes related to Lygophis hoiiisieii: a reappraisal of Rhadinaca antioquicnsis, Rhadinaca tristriata, Coronclla wJiymperi, and Liophis ataliiiallpae. Amer. Mus. Novitates, 2385: 1-27. RozE, J. A. 1958. A new species of the genus Urothcca (Serpentes: Colnbridae) from Vene- zuela. Breviora, No. 88: 1-5. . 1964. The snakes of the Leimadophis- Urotheca-Liophis complex from Parque Na- cional Henri Pittier ( Rancho Grande ) , Vene- zuela, with a description of a new genus and species ( Reptilia, Colnbridae ) . Senckenbergi- ana Biol., 45 (3/5): 533-542. Schwartz, A. 1965. Snakes of the genus Alsophis in Puerto Rico and the Virgin Islands. Studies of the fauna of Curacao and other Caribbean islands. No. 90: 177-227. . 1966. Review of the colubrid snake genus Arrhijton with a description of a new sub- species from southern Oriente Province, Cuba. Proc. Biol. Soc. Washington, 78: 99-114. Simpson, G. G. 1956. Zoogeography of West Indian land mammals. Amer. Mus. No\'itates, 1759: 1-28. Smith, H. M., and C. Grant. 1958. The proper names for some Cuban snakes: an analysis of dates of publication of Ramon De La Sagra's Historia Natural de Cuba, and of Fitzinger's Systema Reptilium. Herpetologica, 14: 215- 222. Underwood, G. 1962. Reptiles of the eastern Caribbean. Carib. Affairs (n.s.), Dept. Extra- mural studies, Univ. West Indies, No. 1 : 1-92. . 1967. A contribution to the classification of snakes. British Mus. (Nat. Hist.), Pub. No. 653: 1-179. 52 BuUciin Museum of Comparative Zoology, Vol. 141, No. 1 Appendix Dental pormxjlae and variation for West Indian and some related mainland and Galapagos xenodontine colubrid snakes. variation includes ranges of subspecies.'^ N2 maxillary palatine pterygoid dentar>' Also phis angustilincaius 1 11 + 2 11 17 18 (iiunimlus 1 12+2 8-9 20 18-19 autillcn.'ii'i 6 16-19 + 2 11-13 28-30 24-26 ater 4 16+2 i;3-16 26-27 22-25 hi.scridlus 1 12 + 2 7 16 17 canihcri^erus 12 11-15 + 2 9-12 23-30 17-21 chammoni.1 2 9 + 2 7-8 10-14 15-16 (lor.saIi.s- 1 12+2 7 16 17 melauiclinus 1 18+2 16 28 24 partoricensis 10 14-18 + 2 9-12 26-32 23-35 rij^cr.smai 2 16-17+2 11-13 26-30 23-24 rufivcntris 3 16-18+2 11-15 25-28 21-26 sancticnicis 2 17-19+2 12-13 30-32 22-23 slevini 1 11 + 2 8 18 19 tachijmenoides 1 11 + 2 10 16 21 vudii vudii 4 11-13 + 2 9-10 20-27 19-23 viulii uiowanae 2 15 + 2 12-13 29-31 22r-23 Aniillophis gen. nov. andrcae 4 18-21 + 2 15-16 34-35 25-28 panifwns 13 13-16+2 10-13 25-27 19-24 Arrhyton callikiem i/.v 2 11-12 + 2 7 16 18 dolicliuium 1 10+2 7 16 12 exiguua 4 13-16+2 7-11 17-27 19-23 funcreus 2 19+2 11 19 24 ])ohi\cpi.s 2 17+2 11 2.3-24 26-27 tiiciiiatum 2 6+2 5 7-8 10 vittatum vittatum 7 12-15+2 10-14 9-10 15-17 vitiattim Icmdoi 3 10-11+2 9 9 14-15 Conophis lineatus 1 10+2 7 17-18 18 Darlingtonia haetiana 3 16-17 + 2 11 22^25 20-24 Dromicus (dmadensi.s 2 18-19 + 2 15-16 27-28 26-29 amazonicus 1 18+2 12 22 23 hiniacidaiiis 1 21 + 2 14 28 26 cursor 2 20-21 + 2 14-15 27-28 26 epinc'phalus 2 22^24 + 2 16-18 27-32 32 jidiac 4 24-26 + 2 29-33 30-34 33-34 niclanohts 3 15-16 + 2 10-11 2.3-26 16-22 onuitus 2 19-20 13 26 24-27 })crfiiscus 2 15-16+2 12-13 20-22 18-19 pscudocobeUa 1 18+2 13 23 19 rcginae 1 16+2 12 27 20 taeniurus 1 19 + 2 11 25 25 Htjpsirhynchus ferox 3 11-12+2 7 17-19 19-20 West Indian Xenodontine Colubrid Snakes • Maglio 53 AprEXDix { Continued ) N2 maxillarv palatine pterygoid dentary laliris dorsalis- Lio))Jiis- aiioDuila cohella jacgcri mcrrcmi Lijguphis hoursieri flavifrenutus liucatits Philodryas acsiivus Jntrmeisteri olfersii Uromaccr catcshtji dorsalis fremitus oxyrJiyucIiu.s Rhadinaca Incvirostrls d ecu rat (I flavdata ■serperaster 16 + 2 21 20 1 12 + 2 10 19 17 3 19-20+2 13-14 26-29 24-30 1 21 + 2 18 27 1 18+2 14 26 22 1 22 + 2 14 28 25 1 26+2 23 35 35 2 19-20+2 14-15 29-32 27-29 1 14+2 11 20 19 1 12+2 9 16 19 1 10+2 9 17 13-14 7 15-18 + 2 10-11 24-25 22-28 2 13-14+2 9 18 20-22 4 14-16+2 9 18-19 24 7 15-17 + 2 8-10 17-21 22-28 1 14+2 9 21 18 1 22+2 14 34- 24 1 24+2 11 24 20 1 17 + 2 9 14-16 14 1 Problematical subspecies which may be full species are listed separately. - Number of specimens examined. .vf-t OF THE Museum of Comparative Zoology The Milliped Family Conotylidae in North America, with A Description of the New Family Adritylidae (Diplopoda: Chordeumida) WILLIAM A. SHEAR HARVARD UNIVERSITY CAMBRIDGE, MASSACHUSETTS, U.S.A. VOLUME 141, NUMBER 2 4 FEBRUARY 1971 PUBLICATIONS ISSUED OR DISTRIBUTED BY THE MUSEUM OF COMPARATIVE ZOOLOGY HARVARD UNIVERSITY Bulletin 1863- Breviora 1952- Memoirs 1864-1938 JoHNSONiA, Department of Mollusks, 1941- OccAsiONAL Papers on Mollusks, 1945- Other Publications. Bigelow, H. B., and W. C. Schroeder, 1953. Fishes of the Gulf of Maine. Reprint, $6.50 cloth. Brues, C. T., A. L. Melander, and F. M. Carpenter, 1954. Classification of Insects. $9.00 cloth. Creighton, W. S., 1950. The Ants of North America. Reprint, $10.00 cloth. Lyman, C. P., and A. R. Dawe (eds.), 1960. Symposium on Natural Mam- malian Hibernation. $3.00 paper, $4.50 cloth. Peters' Check-list of Birds of the World, vols. 2-7, 9, 10, 12-15. (Price list on request. ) Turner, R. D., 1966. A Survey and Illustrated Catalogue of the Teredinidae (Mollusca: Bivalvia). $8.00 cloth. Whittington, H. B., and W. D. I. Rolfe (eds.), 1963. Phylogeny and Evolution of Crustacea. $6.75 cloth. Proceedings of the New England Zoological Club 1899-1948. ( Complete sets only. ) Publications of the Boston Society of Natural History. Authors preparing manuscripts for the Bulletin of the Museum of Comparative Zoology or Breviora should send for the current Information and Instruction Sheet, available from Mrs. Penelope Lasnik, Editor, PubUcations Office, Museum of Comparative Zoology, Harvard University, Cambridge, Massachusetts 02138, U.S.A. © The President and Fellows of Harvard College 1971. THE MILLIPED FAMILY CONOTYLIDAE IN NORTH AMERICA, WITH A DESCRIPTION OF THE NEW FAMILY ADRITYLIDAE (DIPLOPODA: CHORDEUMIDA) WILLIAM A. SHEAR ABSTRACT The milliped family Conotylidae in North America is revised; figures and descriptions of all known species are given. The new family Adritylidae is proposed for the genus Adrityla Causey. Two new genera of conotylids are described: Achemenides (type species, Conottjla pcctinata Causey) and Plumatyla (type species, Conotijla humerosa Loomis). Twelve new species of conoty- lids are described: Conotijla extonis, C. peisonata, C. elpenoi; C. smilax, C. octjpetes, C. aeto, C. vista, C. celcno, Taiyutijla napa, T. francisca, Aiistrotyla hoiealis, and A. chihuahua. Three new synonymies in the genus Conotijla are recognized. The Inology and zoogeography of the group is briefly discussed. INTRODUCTION The small niillipeds of the family Conoty- lidae are poorly represented in most col- lections, though species are fairly common in the northern parts of the United States, and at higher elevations elsewhere. That the distribution of this interesting group is a relict of a previous, colder age seems beyond question. Several species occur in caves, and many are characteristic of high altitudes, reaching near the timberline in the Rocky Mountains of Colorado and Alberta and in the White Mountains of New Hampshire, while adjacent lowlands are poor in species and in individuals. In the middle Appalachians, a j)attem of highly localized, rare, endemic species is now emerging. The importance of millipeds to zoogcographers was repeatedly empha- sized by O. F. Cook, but attempts to use them have been few, probably due to the chaotic state of diplopod taxonomy. It is hoped that studies like this one will en- courage zoogcographers and paleoecolo- gists to utilize the excellent zoogeographic information presented in many groups of millipeds. The family name Conotylidae was pro- posed by O. F. Cook in 1896 for the milli- ped genera Conotijla, Trichopetolum, Sco- feipes and Zij<;onopus, first described in detail by Cook and Collins in 1895. Since its establishment, little revisionary work has been done on the taxonomy of this family despite the large number of addi- tional names that have accumulated under it. Verhoeff (1932) realized that the Conotylidae, in the sense of previous work, was an unnatural assemblage, and removed some genera to his new family, Trichopeta- lidae; his paper was not noticed by American students. Chamberlin (1952) established the genus Taiyutyh, and several new species were added to Conotijla by Loomis ( 1939, 1943 ) and by Causey (1952). Hoffman ( 1961 ) , aware of Verhoeffs assignment of the genera Trichopetalum, Scoterpes, and Zijgonopus to the Tricho- petalidae, studied the gonopods of several species of Conotijla in detail. He reviewed previous studies of the conotylids, and Bull. Mus. Comp. Zool, 141(2) : 55-98, February, 1971 55 56 Bulletin Museum of Comparative Zoology, Vol. 141, No. 2 placed a number of the forms then in Conofijla into a new genus, Sonoratyla. Taiijutyla was also discussed, as well as Chaniberlin's enigmatic genera Zygotyh (1951), Cookella, and BoUmaneUa (1941). Hoffman made no attempt under the cir- cumstances (see section below on dubious names) to clarify the status of these latter three genera. At the same time, Causey (1961a) carried out a similar study, pub- lished earlier than Hoffman's, and proposed the genus Austrotyla for some of the spe- cies later included in Hoffman's Sonoratyla. Later in the same year, Causey (1961b) published the new genus, Adrityla, based on Conotyla deseretoe. Although further papers on conotylid genera have been an- nounced (Causey, 1961a), they have not as yet been published. T wish to thank Dr. Richard Hoffman, Radford, Virginia, for suggesting this study and providing numerous specimens and much unpublished data. His advice has been greatly appreciated. Dr. Nell Causey, Baton Rouge, Louisiana, loaned her col- lection of Conotylidae, probably the largest and most representative in existence, and made many helpful suggestions. H. F. Loomis, Miami, Florida, loaned unpub- lished drawings of new related taxa. Dr. Herbert W. Levi, Museum of Comparative Zoology, Cambridge, Massachusetts, read and edited the manuscript and loaned material from the collections under his care. Dr. Ralph Crabill, U. S. National Museum, loaned important type speci- mens. Dr. Andrew A. Weaver, Wooster, Ohio; Mr. Robin Leech, Edmonton, Al- berta, Canada; Mr. Michael Gardner, Davis, California; and, Mr. Erik Thorn of the British Columbia Proxincial Museum, Victoria, British Columbia, Canada, also loaned or donated important specimens. Mr. Stewart Peck of the Museum of Com- parative Zoology loaned specimens and unpublished data, and contributed much to my understanding of the biology and evolution of the cave forms in the Conoty- lidae. All types of new species described herein are deposited in the Museum of Comparative Zoology. TAXONOMIC POSITION OF THE FAMILY CONOTYLIDAE The affinities of the Conotylidae lie primarily with three North American fami- lies (one briefly described as new, below) and with a poorly known group of Asian and South American species. Verhoeff (1913) placed Japanosoma (Japan) in the monotypic subfamily Japanosominae, and Eudi^ona (Chile) in the subfamily Eudi- goninae of the Conotylidae. Neither of these subfamilies has been restudied since their original proposal, and their position remains in doubt. However, judging from the published data, they seem to be typical eonotylids. The Asian family Diplomarag- nidae is somewhat more distantly related to the Conotylidae (Verhoeff, 1942; Hoff- man, 1963). Recently, Buckett and Cardner (1967) have described a new monotypic family, Idagonidae, type genus Idop^ona, from caves in Idaho. They stated that the new family was related to both the Conoty- lidae and the Cleidogonidae, but as I have earlier pointed out (Shear, 1969), the Conotylidae and the Cleidogonidae are not at all closely related to each other. The single major distinction between the Conotylidae and the Idagonidae lies in the loss of the telopodite articles of the poste- rior gonopods (ninth legs) in the Idagoni- dae; except for this, Idagona might be considered as an aberrant conotylid genus. The relationship of the Conotylidae td the Trichopetalidae is somewhat prob- lematic. The two families are quite similar in genc^ral body plan, but there is a distinct morphological gap in gonopod structure. The anterior gonopods of the eonotylids arc fused into a single article; those of the trichopetalids are two-articled, with a prominent coxite. A colpocoxite is lackim: in the posterior gonopods of trichopetalids. which are two- or three-articled in all known forms; eonotylids have a large MiLLiPED Family Conotylidae • Shear 57 colpocoxite on the three-articled posterior gonopods. Milhped gonopods are developed from walking legs, and there is a well-recognized tendency for more primitive forms to have the gonopods more closely resembling legs; reduction and fusion of segments is con- sidered a specialized character. Thus, the Trichopetalidae seem a more primitive stock than the Conotylidae. Moreover, the conotylids have relatives in Asia and in South America, indicating possible disper- sal routes into North America from either of these regions (or the reverse), while the trichopetalids, like the clcidogonids, are peculiarly North American. My own feeling is that judgment should be reserved pending a study of the Trichopetalidae; it is quite likely that the resemblance be- tween them and the conotylids is due in great part to parallelism, or that tricho- petalids are nearer an ancestral stock than conotylids. The problem is aggravated by the obviously derived trichopetalid char- acters of reduction in segment number, loss of at least some ocelli, and minute size, all of which contrast with the seemingly more primitive gonopod plan. The following key will serve to separate the North American milliped families that may be confused with the Conotylidae. All of the families have in common swollen lateral segmental shoulders and greatly enlarged segmental setae; the male gono- pods superficially resemble each other.- Key to Conotylidae and Related Families IN North America la. Telopodites of ninth legs of males absent, nintli leg represented by the simple, sub- triangular colpocoxite, which curves around the larger telopodites of the anterior gono- pods; caves in Idaho Family IDAGONIDAE lb. Telopodites of ninth legs of males present but reduced, distal articles enlarged and turned dorsally 2 2a. Tenth legs of males with the coxae greatly enlarged and lobed, tenth telopodite re- duced to two or three segments; anterior gonopods a pair of cheirites formed by fusion of telopodites, sternites, and tracheal apodemes Family ADRITYLIDAE, new 2b. Tenth legs with the coxae only slightly enlarged, not lobed; anterior gonopods with telopodites free from the sternites 3 3a. Ninth male legs with a conspicuous col- pocoxite, telopodites two-articled, the distal article oval and enlarged; anterior gonopods a single article; segmental setae usually less than one half the body width Family CONOTYLIDAE 3b. Ninth male legs without colpocoxites, telop- odites usually a single article often witli conspicuous constrictions, often with a claw; telopodite of anterior gonopods two- articled; segmental setae frequently more than one half tlie Iwdy width Family TRICHOPETALIDAE A detailed study of the gonopod stmc- ture of Adrityla deseretoe (Chamberlin) has convinced me that it should not be retained in the Conotylidae, as to retain it would make the family obviously poly- phyletic. The new family Adritylidae, briefly characterized below, is probably related to the Conotylidae to about the same degree as is the Trichopetalidae. By even the most conservative criteria, the structure of the gonopods of Adrityla ex- cludes it from any known chordeumoid family, though there is a vague resemblance to Morquetia ( Opisthocheiridae; Brole- mann, 1935: 278); I have not seen speci- mens of Marquetia. Family ADRITYLIDAE new family Type genus. Adrityla Causey, 1961; type species Conotyla deseretae Chamberlin, by original designation and monotypy. Diagnosis. Large chordeumid millipeds (to 25 mm long) with 30 postcephalic segments, segmental setae and paranota prominent, head not covered by collum. Male gonopods modified from three inter- locking pairs of legs; anterior gonopods (eighth legs) consisting of pair of cheirites formed from fusion of telopodite, stemite and trachael apodeme of each side; poste- rior gonopods (ninth legs) three-seg- mented, coxae with large colpocoxite, IDartially fused to sternites and tracheal 58 BuUetin Museum of Comparative Zoology, Vol. 141, No. 2 apodemos. distal telopodite article inflated, tumt^d dorsally. Tenth legs with coxae greatly enlarged and lobed, coxal gland present, coxae fused to sternites, telop- odites reduced. Eleventh legs without coxal glands. Notes. The singk^ known species, A. deseretae, is common in canyons of the Wasatch Mts., Utah. The reader is re- ferred to the paper of Causey (1961b) for further details. Family CONOTYUDAE Cook Conotylidae Cook, 1896, Biandtia, No. 2, p. 8. Veriioeff, 1932, Zool. Jahil). Abt. Syst., Vol. 62, p. 500. Hoffman, 1961, Trans. Amer. Ent. Soc, Vol. 87, p. 263. Type iienus. Conofi/Ja Cook and Collins, 1895. Diagnosis. Small to medium-sized (9-25 mm) chordeumid millipeds with 30 post- cephalic segments. Head not covered by collum. Eyes present, in triangular patch, sometimes reduced in size and number from maximum of 22-24. Antennae with third article longest, proportions of other articles variable; not markedly clavate; long and slender. Sensory cones four or five. Mouthparts typical. Mentum not divided, mandibles witli about 12 pectinate lamellae. Post-collum segments with prominent lateral swellings bearing outer two seg- mental setae on each side. Segmental setae large, prominent, movable. Epiproct and periprocts truncate; spinnerets two. Legs long, slender, claws prominent, basal seg- ments heavily setose-pilose. Legs one and two of males reduced in size, six-segmented; legs three through seven longer than post- gonopodal legs, usually thickened and eras- sate, often with strong knobs on some of the segments. Gonopods modified from eighth and ninth legs; anterior gonopods consisting of single article, sternum variable. Posterior gonopods three-jointed, coxa free from sternum, with large colpocoxite, distal telop- odite article inflated, turned dorsad, with- out a claw. Coxae of tenth legs with large coxal glands, coxae sometimes lobed. Coxae of eleventh legs without glands, sometimes with prominent hooks. Prefemur of eleventh legs with prominent hooks. Remaining legs unmodified. Female cyphopods of two valves opening anteriorly at rest, pore of oviduct covered by single receptacle. Distri])ution. North America, Japan, Chile. BIOLOGY OF THE CONOTYLIDAE Little is known of the biology and ecology of the Conotylidae. In southeast- ern West Virginia, immature specimens were frequently taken from leaf litter and rotted ^^'ood by Tullgren funnel sampling during the summer and early fall. Ho\\'- ever, nearly 90 percent of all examined collections of mature specimens, including those from caves and high altitudes, were taken between the months of November and April. Cook and Collins (1(S95) com- mented on the surprising activity of Conotyhi during the winter, and I have seen mature specimens of Conotyla walk- ing briskly over frozen logs and ice crystals. In addition, most of the records known to me from the eastern part of the continent are either from caves, high altitudes, or cool microclimates supporting such trees as hemlock, spruce, and fir. In the Rocky Mountains of Colorado, all but a few records of AiistrotyJa coloradensis are from coniferous forest abo\e 7000 ft. (2500 m) elevation. Conotyla alhcrtana has been col- lected in Alberta in alpine meadows above 6500 ft. (2150 m), and no records below 4200 ft. are known. Aiistrotyla specus, CoJiotyla blakei, and C. hollmuni are found at low elevations in the interior of the continent, but most reliable records are from caves. Achemenides pectinatus and Plumatyla humerosa are known only from caves, and the latter is the only conotylid showing well-marked cave adaptations. Howden (1963) has pointed out that animal species populations may adjust to a warming climate by persisting in cooler microhabitats (caves, mountain-tops), or MiLLiPED Family Conotylidae • Shear 59 by adjusting their major period of activity vations on the ecology of only a single to the cooler part of the year. The evidence species, Conotyla hlakei. About 20 speci- abovc seems to indicate that, following mens of both sexes were observed and Quaternary glaciations, both adjustments collected on Mt. Equinox, Bennington Co., may have been made by various species Vermont, and Mt. Grey lock, Berkshire Co., of conotylids. Massachusetts. All specimens were taken The distributions of Conotyla hhkei from above 3000 ft. ( 1000 m ) in elevation, (Map 2) and of Austrotyla specus (Map in dense fir forests. The Mt. Equinox 3) may reflect a current period of cave populations were observed in October, invasion and northward movement follow- 1968, and May, 1969; those on Mt. Grey- ing the Wisconsin glaciation and subse- lock in May, 1969. Without exception, the quent retreat. Barr ( 1968), in an excellent animals were found under the bark of dead summary of the evolution of troglobitic standing trees or fallen logs. The bark was animals, points out that the isolation and loosened to the extent that it could easily evolution of a troglobite usually is pre- be peeled off in large slabs, and the wood ceded by a stage during which an ances- underneath, on which the animals were tral species is troglophilic. The small cave usually resting, was saturated and soft- populations are occasionally swamped water could easily be squeezed from it by genetically by invasions from the sur- the fingers. Held in the hand or exposed face population. Isolation and subsequent to sunlight, the millipeds reacted by curl- speciation occurs when the surface popu- ing the head and first few segments under lations are eliminated by some (usually the body and withdrawing the legs into a chmatic) change that does not significantly position parallel with the long axis. After affect the environmental conditions in the a few seconds, a period of rapid running caves. Because populations in caves are and searching followed. If not allowed to small, genetic drift and founder phenomena find shelter from the sun, or if not released play a great part in their evolution (Barr, from the hand, the millipeds became im- 1968: 82-84). Although present Conotyla mobile and moribund in a matter of 30 and Austrotyla cave populations seem to be seconds. Similarly, an attempt to keep C. geographically isolated from large surface hlakei in the laboratory resulted in the populations, no appreciable differentiation death within 24 hours of all individuals has taken place. Perhaps more thorough collected, even though high humidity was surface collection in the regions of cave maintained. Presumably, temperature was records will produce evidence of popu- a major factor in the laboratory and field lations that can produce the swampmg deaths. ^"^<^'*^' A single mating was obserx^ed in May, However, it must not be overlooked that 1969, on Mt. Equinox, at an elevation of it is the maxima and minima of climatic about 3800 ft.; there were numerous per- parameters, not the average, that usually sistent snowbanks in the fir thicket where restrict the range of a species. In the case the mating pair was found. The female of Achemenides pectmatus, caves may have held onto the wood of the rotted tree-trunk served as refugia from the severe peri- with the last 10 or 12 segments, while the glacial climate during the glacial maxima, male coiled about her body, using the cras- /V further difficulty is posed by the wide sate first seven pairs of legs to embrace distribution in California of Plumatyla her head and anterior segments. No thrust- humerosa, a true troglobite. More than one ing movements of any sort were obser\^ed, species may be represented in the nu- but the pair experienced difficult}' in mcrous collections of immature specimens, separating after being disturbed. I have made extensive personal obser- Schubart (1934) reported spermatophore 60 Bullet ill Mii.svuiii of Comparative Zoology, Vol. 141, No. 2 formation and transfer in some European chordeumids. In these forms, the male se- cretes sperm from the seminal pores on the eoxae of the second legs into coxal sacs on the postgonopodal legs (in the case of the conot>'lids, these sacs are foimd on the tentli ](^gs only). The sc^eretions from these glandular sacs then form the seminal fluid into a spermatophore (Fig. 34) which, during mating, is transferred to the female cyphopods (Fig. 15) at the bases of the second legs by the gonopod complex (Fig. 14) of the male. In presena^d male speci- mens of conotylids, these coxal sacs on the tenth legs are often extruded. In several cases, two large globular objects were at- tached to the extruded ends of the coxal sacs. These were easily detached, and examination under a compound microscope showed that each consisted of a cemented mass of small cells. This structure may represent a spermatophore. If this is the type of mating behavior that occurs in the family, then the distinction in gonopod function pointed out for other families by Brolemann (1935) and referred to by Hoffman ( 1961 ) may be applicable to conotylids.^ Feeding presumably takes place as in otlier chilognath millipeds, material being scraped or picked up from the substrate by a combination of movements of the gnatho- chilarium and mandibles, and ground by the dentate and pectinate lamellae of the mandibles. The finely triturated material in the foreguts of several C. hlokei was composed of wood tracheids and fungal hyphae. Molting and oviposition have not been obsei-ved, but some North American chord- eumids make silk chambers for these pur- poses. ^ Because of some uncertainty, I have not used the terms paragonopod and peltogonopod (see Hoffman, 1961) but have used the more imder- standable, tiiough less specific, terms anterior gonopods to indicate those modified from the eighth legs, and posterior gonopods to indicate those modified from the ninth legs. TAXONOMIC CHARACTERS AND GONOPOD STRUCTURE Unfortunately, as is true of many milli- ped groups, nonsexual characters are of very little use in separating species, and even genera. Causey (1961a) attempted to use the relative lengths of the antennal segments in some species of Aiistrotylo, but such meristic characters must be used with caution when few specimens are available. In species that occur both in and out of caves, ocelli number and arrangement is variable, or it is the same in related epigean species, so that this character is usually of little value. Nonetheless, since such char- acters may at some time be useful, I have described them for all the species men- tioned here. I have, however, omitted much of the usual detail from the species descriptions, particularly description of structures which are common to all mem- bers of a family or genus. Also, female specimens are difficult to assign to species, and even genus. The cyphopods, useful in other chordeumid groups, are virtually identical in different species of Conotylidae. In addition, they are generally so poorly sclerotized that they shrivel when presei"ved, further reducing their utilit\^ in taxonomy. The best taxo- nomic characters are to be found in the male gonopods. In addition, the modi- fications of some of the pregonopodal legs of the males, usually taking the form of mesal lobes on one or two segments, are quite important in separating species. Conoti/hi melinda (Figs. 1, 2) has gono- pods typical in many respects of the large group of Appalachian species which are highly endemic in their distribution. In lateral view ( Fig. 1 ) the anterior gonopods are seen to curv^e laterad of the colpocoxites (c) of the posterior gonopods (telopodite articles of posterior gonopod not shown). The apex of the anterior gonopod is di- vided into two branches, an anterior branch (ah) and a posterior branch (ph), and there is a prominent lateral shoulder (Is) on MiLLiPED Family CoNOTiLiDAE • Shear 61 ■igures 1-6. Anatomy of conotylid male gonopods. Fig. 1. Conotyla melinda, lateral view. Fig. 2. C. melinda, anterior lew. Fig. 3. C. fiicheri, anterior view. Fig. 4. C. blakei, anterior view. Fig. 5. C. smilax, anterior view. Fig. 6. C. 'trollneata, anterior view. 62 Bulletin Museum of Comparative Zoology, Vol. 141, No. 2 which, in some species, the process of the femur of the seventh leg rests (ConotyJa vista. Fig. 41). Note also the coxal sac of leg ten (cs) and the modified prefemur of leg eleven (pf 11). In an anterior semi- diagrammatic view (Fig. 2), the narrow, bandlike sternum of the anterior gonopods is visible, as well as the telopodite articles of the posterior gonopods. Hoffman (1961) showed that the narrow muscle (rcl) known as the remotor coxae longus, extending from the distal part of the tracheal apo- deme (ta) to the base of the anterior gono- pod, can be used to establish the homology between that portion and the coxa of a walking leg. Thus, the anterior gonopod is a coxotelopodite articulating directly with its sternum. In Conotijla fischeri (Fig. 3), a similar pattern is repeated, but the an- terior branch of the anterior gonopod (0^1,) is usually not developed. This branch is typically plumose when present. Conotijla hlakei (Fig. 4) is representative of a group of species with somewhat simpler gonopod construction. The anterior branch (ah) of the anterior gonopod is reduced to a poorly sclerotized plumose rod that is broken off in many specimens, and the posterior branch (ph) is a smooth continuation of the gonopod mass. There is no lateral shoulder. The colpocoxites (c) are likewise simpler in form (Fig. 17). The sterna are somewhat broader in this species than in the central Appalachian group. Conotijla similax (Fig. 5) is a slightly aberrant Ap- palachian form in which the anterior gonopods (ag) do not pass laterad of the colpocoxites (c), and bear no posterior branch. The colpocoxites (Fig. 31) are complex on the posterior surface, with a terminal process (tp), mesal tooth (mt), lateral notch (In), and pilose area (pa). Conotijla atrolineata (Fig. 6) differs from the eastern North American species of Conotijla in the simplification and re- duction of the anterior gonopods, the broadening of the sternum of the posterior gonopods (.s), and the much greater com- plexity of the colpocoxites (Figs. 48, 49, 50). This trend culminates in the related species, C. alhertana, ^\'ith an exceptionally broad posterior gonopod sternum (Fig. 7). In Taiijiitijla corvallis (Fig. 8), the an- terior gonopod sternum (s) is heavily sclerotized and not bandlike, completely encircling the simple, platelike anterior gonopods (ag), which are larger than the colpocoxites (c). In Austrotyla colorademis (Fig. 9), the sternum of the anterior gonopods (.s) may, in certain preparations, appear divided, though it is in reality a single piece en- circling the gonopods and nearly meeting posteriorly. Lateral extensions of this sternum articulate with the lateral surface of the gonopod, which is simple and leaf- like and larger than the colpocoxite. The colpocoxite (c) is usually cupped anteriad and may have two parts. The tracheal apodemes are reduced in size. Plumatijla humerosa (Fig. 10) is a unique form somewhat intermediate be- tween Taiijutijla and Austrotyla, but the lateral extensions of the solidly constructed sternum (s) are more firmly fused to the gonopod than they are in either of the preceding. The anterior gonopod (ag) is complex (Fig. 80), with two branches per- haps homologous to the two branches fovmd in many Conotijla species. The colpocoxites (c) are remotely similar to those of Austro- tyla. Achemenides pectinatus (Fig. 11) dif- fers in most respects from the other mem- bers of the family. The anterior gonopod sternum (sj is truly divided, and each sternite is doubled over longitudinally (Figs. 54, 55). The anterior gonopods (ag) are fused at the base, and are much larger than the colpocoxites (c), which are widely separated on a thin, bandlike sternum. The tracheal apodemes of the posterior gono- pods are small and bifurcate. PROBLEMATICAL NAMES IN THE CONOTYLIDAE Due to the fact that the male genitalia are the only relialjle specific characters in MiLLiPED Family Conotylidae • Shear 63 this family, names based on females and on immature specimens are extremely difficult to assign, particularly if more than one species of conotylid occurs in the region. In two cases, Conotyki wyandotte (Boll- man) and Conotyh jonesi Chamberlin, I have made some attempt to place names attached to females or lost specimens in the proper synonymy. Listed below are the problematical names, \\'ith brief discussions of their original proposals and histories. Cookella leibergi (Cook and Collins) Conotyla leibergi Cook and Collins, 1895, T- New York Acad. Sci., 9(1): 77, figs. 102-104. Cookella leibergi, Chanilserlin. 1941, Bull Univ Utah, Biol. Ser., 6(5): 13. This species was originally described from a female (holotype in U. S. National Museum, examined), which was supposed to have a small, broadly triangular pro- mentum. This mistaken original observation of Cook and Collins (1895) was used by Chamberlin (1941) as a pretext for erect- ing the completely unnecessary genus Cookella. The type locahty, [the shore of] Lake Pend d'Oreille, Idaho, could be within the range of either Conotyla atro- lincata or C. albertana, and leibergi would thus be treated either as a junior synonvm of atroUneata or the correct name for al- bertana. Loomis (personal communication, 196S) has in preparation descriptions of new taxa from the Idaho-Montana area; leibergi males may conceivably be repre- sented among them. As neither the generic or specific name can be properly dealt ^\•ith in the absence of males, I consider Cookella leibergi a nomen chibium. Trichopetalum glomeratum Harger Trichopetahim glomeratum Harger, 1872, Amer. ^J. Sci. Arts, 4: 118. '^onoUjla glomerata, Cook and Collins, 1895, ]. New York Acad. Sci., 9(1): 78. The holotNpe, from the John Day \'alley, >egon, has been lost, and as Cook and Col- ins (1895) stated, the original description is too vague to allow definite placement, though a certain similarity to Taiyntyla emerges. It is probably neither a Tricho- petahim nor a Conotyla. I consider it a nomen chibium. Bollmanello oregona Chamberlin BoUmauella oregona Chamberlin, 1941 Bull Univ Utah, Biol. Sen, 6(5): 12. The male holotype, from John Day Creek, Douglas Co., Oregon, is reputedly in the Chamberlin collection at the Uni- versity of Utah, but it could not be located by the curator there. There is nothing diagnostic about the description of the genus or of the type and only species, which even contradict one another in the matter of a lobe on the fourth article of the fifth legs. No illustrations were presented. I consider it a nomen dubium. Zygotyla phana Chamberlin Zijgotijla phana Chamberlin, 1951, Nat. Hist. Misc. No. 87: 7-8, fig. 14. The holotype, from blue River, British Columbia, Canada, is ob\iously immature, having only 28 segments and undeveloped gonopods ["The gonopods seem to have been broken off in the type (Chamberiin, 1951)."] Though the type was to have been placed in the Provincial Museum of British Columbia, they have never received it, and it could not be located in the Chamberiin collection at the University of Utah. Geo- graphically, this species could possibly be a synonym of C. atroUneata, but it may also represent an undescribed species. Nomen dubium. Conotyla jonesi Chamberlin Conotyla jonesi Chamberlin, 1951, Nat. Hist. Misc. No. 87: 7, fig. 1,3. The drawing gi\en by Chamberiin (1951) for C. jonesi is very much like Taiyntyla corvallis when properly oriented; the tvpe localities ( t>'pe of T. corvallis from Cor\allis, Oregon) are onl\- about 25 miles apart and both are in the \Mllamette River 64 BiiUetin Museum of Comparative Zoology, Vol 141, No. 2 valley. None of the eharacters given in the original deseription of C. joncsi are in the least (liagnostie; they eould apply to almost an\' eonotyloid of the region. See Hoffman (1961) for a full discussion of Chamberlin's errors of observation in the deseription of T. corvallis. The type of jonesi WAS supposedK' i-)laeed in the Provincial Museum of iiritish Columbia, but they have no ic^cord of ever having received it, and it could not be located in Chamberlin's collection at Salt Lake City. At any rate, if specimens from the type locality of jonesi prove to be identical to corvallis, the name may have to be changed to jonesi, which has a year's priority. For the present, jonesi is best regarded as a nomen duJnum. Key to Genera of Conotylidae of North America la. Anterior ^onopocl sternum divided; anterior gonopods fused to each other at base, much larger than colpocoxites of posterior gono- pods (Fig. 10); northern Illinois, nortlieast- ern Iowa, southwestern Wisconsin Achemenides lb. Anterior gonopod sternum not divided (Figs. 1-9); anterior gonopods not fused to each other, though they may be closely appressed, smaller oi larger than colpo- coxites of posterior gonopods 2 2a. Anterior gonopods flattened, platelike, often appressed in the midline (Figs. 8, 9), larger than or subequal to colpocoxites of posterior gonopods - - 3 21). Anterior gonopods never flattened, often with complex or plumose branches (Figs. 1, 32-.34), smaller than or subequal to colpocoxites of posterior gonopods 4 3a. Anterior gonopod sternum with lateral processes partially fused to lateral sides of coxotelopodites (Fig. 69); coxotelopodites complex on posterior surface (Fig. 78); colpocoxites small, cupped anteriad; Illi- nois and Missouri, and Rocky Mountains from Alberta to Chihuahua Austrotijla 3]i. Anterior gonopod sternum without such processes, lieavily scleroti/.ed, completely surrounding jjases of coxotelopodites ( Figs. 59, 61, 63); colpocoxites subequal to an- terior gonopods, fre(iuently complex and branched (Fig. 64), not cupped anteriad; Pacific Coast from central Oregon to San Francisco Bay region - Taitiiifyhi 4a. Anterior gonopod sternum with lateral process partially fused to lateral edge of gonopod; gonopod two-branched, mesal edge of major branch heavily laciniate (Figs. 80, 81); colpocoxitc two-branched; animals without pigment, ocelli irregular; caves in northern California Pltimattjla 4b. AnlcMJor gonopod sternum simple and bandlike (Figs. 1-7); gonopod with one or two small branches, usually not laciniated, colpocoxites with complex posterior surface (Fig. 52), but not two-branched; animals usually pigmented, ocelli round, black; eastern North America from Maine to North Carolina, Rocky Mts. of Alberta and British Columbia, Sierra Nevada Mts. of California Conotyla Genus Conotyla Cook and Collins, 1895 Craspedosoma, Bollman ( in part, not of Leach, 1815), 1893, U. S. Nat. Mus. Bull. No. 46, pp. 35, 183. Tiichopetalum, McNeill (in part, not of Harger, 1872), 1887, Proc. U. S. Nat. Mus., Vol. 10, p. 330; Bull. Brookville Soc, Vol. 3, p. 8. Scoterpes, Bollman (in part, not of Cope, 1872), 1893, Bull. U. S. Nat. Mus. No. 46, p. 106. Conotiila Cook and Collins, 1895, Ann. New York Acad. Sci., Vol. 9, No. 1, pp. 70-71. Hoffman, 1961, Trans. Amer. Ent. Soc, Vol. 87, p. 265. Proconotyla Verhoeff, 1932, Zool. lahrb. Abt. Syst., Vol. 62, p. 501. NEW SYNONYMY. Type species of Conotyla, Conotyla fischeri Cook and Collins, of Proconotyla, P. ])lakei Verhoeff. The generic name is a feminine Greek noun, "a cone-shaped lump," and refers to the segmental shoul- ders. Diagnosis. Anterior gonopod sternum simple and bandlike, not di\ided. Anterior gonopods \'ariously branched or simple and acuminate, smaller than or subcciual in size to colpocoxites of posterior gonopods. An- terior gonopods usually curved posteriad of colpocoxites. Colpocoxites of posterior gonopods large, complex to simple on pos- terior surface. Sternum of posterior gono- pods bandlike or conspicuously broadened and ovate, spiracles large and prominent, tracheal apodemes fused to sterna. Progon- opodal legs of males frequently with femoral knobs, appearing on all legs in one case, and on none in another extreme. Species usually pigmented; ocelli 14-23, MiLLiPED Family Conotylidae • Shear 65 0. 5 mm 0.5 mm Figures 7-14. Anatomy of conotylid male gonopods and of Conotyla blakei. Fig. 7. Conotyla albertana gonopods, an- terior view. Fig. 8. Taiyulyla corvallis gonopods, anterior view. Fig. 9. Austrotyla coloradensis gonopods, anterior view. Fig. 10. Acfiemen/des pectinatus gonopods, anterior view. Fig. 11. Plumatyla humerosa gonopods, anterior view. Figs. 12-14. Conofy/o blakei. Fig. 12. Leg 7, posterior view. Fig. 13. Gnathochilarium, ventral view. Fig. 14. Segment 7 of male, lateral view (anterior to the left). 66 Bulletin Museum of Comparative Zoology, Vol 141, No. 2 Map 1. North America, showing distribution of conotylid genera (some records of immature specimens included); dots, Conofy/a spp.; circles, Ausfrotyla spp.; triangles, Plumatyla spp.; solid square, Achemenides pectinatus; open squares, Tai- yutyla spp. usually round, black, arranged in triangular eyepatch in four or five rows. Coxae of tenth legs with large coxal glands, not lobed; coxae of eleventh legs normal; pre- femur of eleventh legs with prominent posterior hooks. Species. Fifteen known. Distribution. See Map 1. Eastern North America from Maine through \'ermont, New Hampshire and New York to Ohio and Indiana in the west and to North Caro- lina in the east; center of diversity is ap- parently in eastern West Virginia and southwestern Virginia; Rocky Mountains of the Alberta-British Columbia border; Sierra Nevada Mts. of California. Key to Species (Males) la. Pregonopodal legs only sliglitly more eras- sate than postgonopodal legs, without knobs; Nevada Co., California extorris lb. Pregonopodal legs strongly crassate (Fig. MiLLiPED Family Conotylidae • Shear 67 12) at least leg 4 or 7 with a femoral poorly sclerotized plumose rod (Figs. 4, 2a. LegVwithafem'oralknob":: 3 ^^' ^f^' ^^^'^'1^^. ^^^''l ^"""^ ^ holhnani 2b. Leg 4 without a femoral knob, leg 7 with ^^® clearly related, but the resemblance of strong femoral knob (Fig. 12); Vermont, ^- extorris to these species is probably a New York, Pennsylvania, Maryland, and result of parallelism. West \'irginia, often in caves blakei 3a. Leg 7 without a femoral knob 4 n^^^i I Li i • i\i l rr\ 3b. Leg 7 with a strong femoral knob ..::;Z: 6 ^°"°^y/° ^'^'^f ' (Verhoeff) 4a. Legs 2 and 3 with femoral knobs; Virginia Figures 12-18 ^, " celeno FwconoUjla hlakci Verhoeff, 1932, Zool. Jahrb. 4b. Legs 2 and 3 without femoral knobs 5 Abt. Syst., Vol. 62, p. 501, figs. 33-37. Hoff- 5a. Leg 5 without a femoral knob; Indiana man and Chamberlin, 1958, U. S. Nat. Mus. _ .... hollmani Bull. 212, p. 101. Hoffman, 1961, Trans. Amer. 5b. Leg 5 with a femoral knob; Virginia ... Ent. Soc, Vol. 87, p. 271. venetia Conotyla vaga Loomis, 1939, Bull. Mus. Comp. 6a. Leg 6 without a femoral knob; British Zool., Vol. 86, pp. 182-183, fig. 10. Chamber- Columbia atiolincata lin and Hoffman, 1958, U. S. Nat. Mus. Bull. 6b. Leg 6 with a femoral knob 7 212, p. 99. NEW SYNONYMY. 7a. Leg 2 with a femoral knob 8 7b. Leg 2 without a femoral knob 9 Types. Male holotype of P. hiakei, Mt. 8a. Leg 1 with a femoral knob; West Virginia Adams, Essex Co., N. Y., wherealjouts «K I'Z'T'-^'V'f V'T" r "V;." . ^.'''^" unknown; of C. voga, South Temple bb. Leg 1 without a femoral knob; Virginia r^ n ^ r^ r. i • • i jfjelimla Cave, Berks Co., Pennsylvania; ni the 9a. Leg 3 with a strong femoral knob 10 ^^- C. Z., examined. 9l3. Leg 3 with a weak femoral knob, or lack- Diagnosis. Distinct from other species T« !!^^ ^ ^^^^ --: :---: ^ of the ])IaJr .r.r^ o uicsal traHsparcut plate; pilose area limited 1 HE Blakei Group *. i i. i r -..i ^ i i i to lateral surface, without branches; mesal This group is characterized by simple, tooth and lateral notch not present. usually acuminate anterior gonopods in Remarks. This species is widely dis- vvhich the anterior branch is reduced to a tributed (Map 2). It has been collected 68 BuUctin Museum of Comparative Zoology, Vol. 141, No. 2 Figures 15-20. Anatomy of Conofy/o blakei and gonopods of Conotyla spp. Figs. 15-18. Conofy/o blakei. Fig. 15. Head and anterior trunk segments of female, lateral view. Fig. 16. Head of female, anterior view. Fig. 17. Colpocoxite of left posterior gonopod, posterior view. Fig. 18. Left anterior gonopod of male, posterior view. Figs. 19-20. C. bollmani. Fig. 19. Left anterior gonopod of male, posterior view. Fig. 20. Left posterior gonopod, posterior view. MiLLiPED Family Conotylidae • Shear 69 Map 2. Part of northeastern United States, showing distribution of some Conofy/o spp.; dots, epigean records of C. blakei; circles, cave records of C. blakei; crosses, records of C. fischeri. Question mark indicates a dubious identification. Map 3. Part of central United States, showing distribution of Austrotyla and Achemen/des,- dots, epigean records of Austrotyla specus; circles, cave records of A. specus; crosses, records of Achemenldes pecttnatus (all cave records). Ques- tion mark indicates a dubious identification. in the Adirondack counties of New York in Canadian Zone Forests and in caves; in caves in Sullivan, Orange, and Schoharie counties, and in numerous caves in central Pennsyh'ania and Maryland, and a single case in Jefferson Co., West Virginia. I have personally collected blakei in^ a scrubby fir forest at 3800 feet on the sum- mit of Big Equinox Mountain, Bennington Co., Vermont, where it was abundant under the bark of both standing and fallen dead trees. The distribution of ]}Ia]ographicall\' closer. HoNxt^ver, the form of the gonopods and especially the anterior gonopod sternum favor the present placement. If more new species with a similar combination of characters are col- lected, a new genus may be justified. The Fischeri Group This is a compact, closely related group of three species from New York, Ohio, and Kentucky, characterized by the reduction of the anterior branch of the anterior gonopods, and by their similar colpocoxites. Conotyla fischeri Cook and Collins Figures 3, 24, 25 Conotifla fischeri Cook and Collins, 1895, Ann. New York Acad. Sci., Vol. 9, pp. 71-74, figs. 55-78. Loomis, 1943, Bull. Mus. Comp. Zool., Vol. 92, p. .382 (key). Hoffman, 1961, Trans. Amer. Ent. Soc, Vol. 82, pp. 265-266, fig. 7. Chamberlin and Hoffman, 1958, U. S. Nat. Mns. Bull. 212, pp. 98-99 (list). Types. None designated by original authors; type locality restricted by Chamberlin and Hoffman ( 1958 ) to Syracuse, Onondaga Co., New York. Diagnosis. Distinct in the form of the gonopods from all other species except the closely related personata; differing from personota in the gonopods as described under that species, and also in having a much more pronounced apo^jhysis on the femur of leg 4. Description of mole from Skaneateles, Onondaga Co., New York. Length, 19.4 inm. Eyepatches triangular, 20 ocelli in four ro\x's on left side, 21 ocelli in four rows on right side. Antennal segments in order of length: 3, 4 = 5, 2, 6, 7, 1. Pre- gonopodal legs somewhat less enlarged than usual, legs 4-7 usually with apophyses on femur; distal on leg 4, mesal on legs 5 and 6; large, proximal and toothed, if pres- ent, on leg 7, but not resting on anterior gonopod in situ. Anterior gonopods (Fig. 24) short, subfiuadrate, posterior branch absent, anterior branch short, not curving around colpocoxites of posterior gonopods; lateral shoulder poorly developed. Colpo- coxites of posterior gonopods (Fig. 25) without a terminal process, but with a prominent ridge on the posterior surface, bearing near its proximal end a long, rather flattened rod; lateral notch absent; pilose area with one or two pilose branches. Notes. Cook and Collins (1895) had over 100 specimens from Onondaga, On- tario, and Wayne counties, New York (Map 2). They were collected primarily from rotting littcn- in woods, and under the moist, rotting bark of fallen trees. Hoffman ( 1961 ) reported the species from Hamilton Co., 2 miles west of Morehouseville, but gave no additional ecological data. His material, which I re-examined for this study, differs from the Cook and Collins series and from other Onondaga Co. speci- mens in having the apophysis of the seventh legs very small, or absent; the other apophyses are normally developed. The species is evidently limited to the rolling hills south and east of Lake Ontario (Map 2), but extension of its range into areas bordering on Lake Erie would be most interesting, as it might indicate the exact nature of this species' relationship to per- sonata, which occurs in similar terrain just south of Lake Erie in Ohio. Several in- triguing immature specimens from Ontario, ' northern New York, and from Potter Co., Pennsylvania, this latter locality near the southern limit of Pleistocene glaciation, may belong to this species. MiLLiPED Family Conotylidae • Shear 73 Conotyla personata n. sp. Figures 26, 27 Type. Holotype male collected by A. A. Weaver, May 5, 1960, Funk's Hollow, Wooster, Wayne Co., Ohio. The spe- cific epithet is a Latin adjective, meaning "masked, hidden," and refers to my original confusion between this species and fisclieri. Diagnosis. Very similar to fisclieri, but distinct in details of the gonopods. The lateral shoulder of the anterior gonopod is much higher (Fig. 26) than in fisclieri, and the posterior branch is longer and more pronounced; the lateral notch of the colpocoxite is almost obsolete in fisclieri; in personata it is the deepest and broadest of any species in the genus. Description of male holotype. Length, 14 mm. Ocelli in four irregular rows on each side, 21 on right side and 23 on left side. Antennal segments in order of length: 3, 4, 5, 6 = 2. 7, 1. Fourth joints of legs 4, 5, and 6 with small, distal knob; fourth joint of leg 7 \\'ith long, apically toothed proximal process that in situ rests on the lateral shoulder of anterior gonopod. Ster- num of anterior gonopods as described for genus, but lateral edge not visible in poste- rior view. Anterior gonopods (Fig. 26) short, subrectangular in posterior view; lateral shoulder prominent; anterior branch absent, posterior branch curved dorsad slightly. Colpocoxite of posterior gonopod (Fig. 27) bluntly subtriangular, all proc- esses and teeth undeveloped; pilose area extended into at least one pilose branch; lateral notch deep, extending nearly half- way up colpocoxite. Notes. The male holotype was taken in a deep, heavily \^'ooded, mesic ravine sur- rounded by rolling cultivated land. The north-facing slope of the ravine supports a heavy mixed forest with much hemlock; the south-facing slope is forested with oak and pine. A study of the humus fauna of these slopes produced several immature and female Conotyla in early fall. This species is possibly widespread in northern and central Ohio, l^ut the only males are from the type locality, and nearby. Conotyla elpenor n. sp. Figures 28, 29 Type. Male holotype collected with an- other male and two females on 26 February 1966 by Branley A. Branson, nine miles northwest of Pine Ridge, ekn'ation 800 ft. (250 m), Wolfe County, Kentucky. The specific epithet refers to one of the companions of Odysseus, changed into a swine by Circe's witchcraft. Diapjnosis. With the characters of the fisclieri group; distinct in the form of the anterior gonopods, with two well-defined branches ( Fig. 28 ) , and in the colpocoxites of the posterior gonopods having an ex- ceptionally deep lateral notch and coxal pocket. Description of holotype male. Length, 18 mm. Eyepatches subtriangular, 20 ocelli in four rows on each side. Antennal articles in order of length: 3, 4 = 5, 6, 2, 7, 1. Pre- gonopodal legs 3 through 7 with apophyses on femur; apophysis of leg 3 small, distal; of legs 4 and 5 mesodistal, larger; of leg 6, proximal and blunt, of leg 7 proximal and greatly elongated, coarsely toothed distally, resting in situ posterior of lateral shoulder of anterior gonopod. Anterior gonopod (Fig. 28) with lateral shoulder, large, lamellate; two branches of gonopod may represent a bifurcation of the posterior branch, in situ they extend laterally around colpocoxites of posterior gonopods. Colpocoxites of posterior gonopods (Fig. 29) closely resembling those of personata; but somewhat longer, bent sharply pos- teriad, blunt; lateral notch deep, pilose area with two indistinct branches. Notes. The holotype and the other known specimens were taken from beneath a rotting log in mixed hardwood forest. Though the specimens were labelled as being from Wolfe Co., the distance north- 74 Bulletin Museum of Comparative Zoology, Vol. 141, No. 2 MiLLiPED Family Conotylidae • Shear 75 west of Pine Ridge indicated on the label would place the locality in Powell Co. This species is known only from the type lo- cality. A female from Carter Co., Kentucky (MCZ), may belong to elpenor (Map 5). The Smilax Group The single species of this group, C. smilax, is unique in several ways, but may be related to the preceding group. Conotyla smilax n. sp. Figures 30, 31 Ttfpe. Holotvpe male collected ca. elev. '3000 ft. (ioOO m). Kate's Mt., above White Sulphur Springs, Greenbriar Co., West Virginia, April 1, 1967, by W. A. Shear. The specific epithet is a Latin noun in apposition and refers to Smilax, the greenbriar, abundant at the type locality. Diap,nosis. Distinct from all other related species of Conotyla in lacking the posterior branch of the anterior gonopods ( Fig. 30 ) . The anterior branch is well developed and had a deeply laciniated membrane running the length of its posterolateral edge. Description of male holotype. Length, 13.3 mm. Ocelli in four ro\\'s on each side, 20 ocelli in each eyepatch. Legs 3 through 7 enlarged and crassate, knob present on femur of legs 3 through 7. Antennal articles in order of length: 3, 4 = 5, 6 = 2, 7, 1. Sternum of anterior gonopods as described for genus. Anterior gonopods (Fig. 30) with conspicuous lateral shoulder, but knob of fourth segment of leg 7 does not rest on shoulder when gonopod is in situ; poste- rior branch absent, anterior branch large, thin, nearly half as long as remainder of coxotelopodite, extending straight ventrad, with a deeply laciniated membrane on posterolateral surface. Colpocoxites of posterior gonopods with terminal process long, aciculate; mesal tooth inconspicuous, pilose area with definite branches, ex- tensive (Fig. 31). Notes. The male holotype, the only known specimen, was collected after a light rain, crawling about on dead twigs under a Rhododendron. The type locality (Map 5) is an unusual area, dissected into adjacent areas of very mesic forest of mixed hardwoods and hemlock and "shale bar- rens," with a hot, diy microclimate and sparse vegetation of scrub oak and Vir- ginia pine. Apheloria trimaculata and Cleidogonu major were taken in abundance. The Venetia Group These species have similar anterior gono- pods in which the anterior branch is absent and the posterior branch long, giving a right-angled appearance to the gonopod {kg. 33). Conotyla venetia Hoffman Figures 32-34 Conottila veitetia Hoffman, 1961, Trans. Amer. Ent. Soc, Vol. 87, p. 267, Pi. IX, figs. 1-3. Type. Male holotype from three miles north of Clifton Forge, Alleghany Co., Virginia; type in U. S. National Museum, examined. Diagnosis. Distinct from all other species of Conotyla in bearing apophyses on the fourth article of legs 4 and 5 only; distinct in details of the gonopods from others of the venetia group. Description of male from Warm Spring Mt., Alleghany Co., Virginia. Length, 14.5 mm. Ocelli 20 on each side, in four rows. Antennal segments in order of length: 3, 4, 5, 2, 6, 7, 1. Legs 4 and 5 with small <- Figures 30-37. Male gonopods of Conotyla spp. Figs. 30-31. Conotyla smi/ox. Fig. 30. Left anterior gonopod, posterior view. Fig. 31. Left posterior gonopod, posterior view. Figs. 32-34. C. venetia. Fig. 32. Left anterior gonopod, lateral view. Fig. 33. Colpocoxite of left posterior gonopod, posterior view. Fig. 34. Coxae of 10th legs of mole, ventral view. Figs. 35-36. C. ocypefes. Fig. 35. Left anterior gonopod, posterior view. Fig. 36. Left posterior gonopod, posterior view. Fig. 37. C. oeto, left anterior gonopod, posterior view. 76 Bullet in Museum uf Comparative Zoology, Vol. 141, No. 2 apophxses on femora, distal in position. Anterior gonopods (Figs. 32, 33) with posterior branch extending at right angle posteriad from t(>lopodite, anterior branch absent; lateral shoulder well dexeloped, heavily rugose on latc^ral surfaces posterior branch extending in situ well around lateral surface of colpocoxite of posterior gonopod. Colpocoxite of posterior gonopod (Fig. 34) subrectangular in outline; terminal process definitely set off by promintnit shoulders at its base; mesal tooth reduced to small knob; pilose area lateral; lateral notch present, but small. Notes. This species is known only from two localities in Alleghany Co., Virginia (Map 5), separated bv nearlv 3000 ft. (1000 m) of altitude ( 400-3400' ft. ) . Hoff- man (1961) implies that the species may be widespread in central western Virginia. Conofyla ocypefes n. sp. Figures 35, 36 Type. Male holotype from Sugar Grove, Fairfield Co., Ohio, collected by J. H. Emerton, December 26, 1915. The spe- cific epithet is a noun in apposition, the Greek name of one of the three Harpies of mythology. Dia on the fourth joint, not two lobes on the third; leg 7 (Fig. 58) has apophyses on both the third and fourth joints, not just the third; there is no evi- dence of a "dorsal" (anterior) branch of the anterior gonopods; the deeply laciniate lamellae of the posterior gonopods are coxal, not sternal. The reduced ocelli and fairly large size of this species indicate that it is somewhat adapted for a subterranean existence. Anstrotijla specus, on the other hand, is troglophilic and shows no strong adapta- tions for cave life. Thus both could pos- sibly occur in Smith Park Cave without competing directly. The locality needs to be recollected to confirm this, and to pro- vide more specimens. The presence of this species at the edge of the classical Driftless Area, and in caves, is of great potential significance, especially since it shows a combination of conotyloid-austrotyloid fea- tures. But the problems associated with defining this area (Frye, 1965) limit specu- lation. Gushing (1965) found evidence of many disjunct arctic-alpine plants in the Driftless Area, and other plants that occur only south of the Wisconsin maximum farther east. It is not unlikely, therefore, that the region escaped glaciation, while the extreme climate drove A. pectinotiis or its ancestor to seek refuge in the moder- ated climate of caves. In view of this un- usual interest, I present a detailed list of new records below (see also Map 3). Records. ILLINOIS: Jo Daviess Co., mines in North California Diggings, 7 mi. NW of Hanover, 31 October 1965, S. Peck, $9 9; South Nicholsen Mine, 31 October 1965, S. Peck, S $ 9; Hutchings Mine, 5 mi. E of Galena, 30 October 1965, S. Peck, $ 9 . IOWA: Jackson Co., Hunter's Cave, 5 mi. N of Andrew, January-February 1966, S. Peck, (5 5 5 5. WISCONSIN: Richland Co., John Gray Cave, 5.5 mi. NNE of Richland Center, no date, C. Krekc^ler, $ . Genus Taiyufyla Chamberlin, 1952 Taiytityla Chamlieilin, 1952, Nat. Hist. Misc., \ Chicago Acad. Sci. No. 113, p. 1. Chamberlin and Hoffman, 1958, U. S. Nat. Mus. Bull. 212, p. 102. Hoffman, 1961, Trans. Amer. Ent. Soc, Vol. 87, p. 270. Type .species. Taiyutyla corvallis Cham- berlin, by original designation. The origin of the generic name, a neologism, is ob- scure. The gender is believed to be femi- nine. D/f/gno,s/.s. Small (S-11 mm in length) conotylids with the fifth antennal segment longer than the fourth, 20 ocelli or less in each eyepatch. Pregonopodal legs with apophyses on third or fourth article. Eleventh legs with prefemoral hooks. An- terior gonopods platelike, sometimes with terminal and subterminal processes or lamellae; stenium of anterior gonopods completely surrounding coxotelopodites, which articulate primarily on its posterior surface; spiracle prominent, in lateral de- pression. Posterior gonopods large, but slightly smaller to distinctly smaller than anterior gonopods, platelike or with two major branches, subterminal branch usually flattened, but sometimes spikelike, subtend- ing or bearing a plumose-hirsute area; terminal branch or process short, or long and spirally curved, often bearing accessory teeth; no coxal depression, coxal bases swollen posteriorly. Sternum of posterior gonopods broad and deep anteriorly, thin and ribbonlike posteriorly; spiracles at proximolateral margins of broad, semi- circular depressions. Telopodite articles of posterior gonopods \'ariable, in some cases subequal and with second showing signs of segmentation, or with the first twice as long as the second, which is reduced to a small faiob. Species and distrihution (Map 1). Three species in southern On^gon and northeni California. MiLLiPED Family Conot\'lidae • Shear 87 Key to Species la. Second telopodite joint of male posterior gonopods only half the length of the first (Fig. 64); leg 3 with an apophysis on the fourth joint, other pregonopodal legs un- modified napa lb. Second telopodite joint of male posterior gonopods (Figs. 60, 62) at least as long as the first; legs 4 through 7 and sometimes leg 3 with femoral apophyses 2 2a. Third leg of male with an apophysis on the femur; gonopods as in Figs. 61, 62; Marin Co., California francisca 2b. Third leg of male unmodified; southern Oregon corvallis Taiyutyla corvallis Chamberlin Figures 8, 59, 60 TaiyiityJa corvallis Chamberlin, 1952, Nat. Hist. Misc., Chicago Acad. Sci. No. 113, pp. 1-2, figs. 1, 2. Chamberlin and Hoffman, 1958, U. S. Nat. Mus. Bull. 212, p. 102 (list). Hoffman, 1961, Trans. Amer. Ent. Soc, Vol. 87, p. 270, pi. 10, figs. 8, 9. Type. Holotype from Corvallis, Linn Co., Oregon; in Chamberlin collection, Salt Lake City, Utah, could not be located. Diagnosis. Unique in lacking an apoph- ysis on the fourth joint of leg 3. Description of mole from Corvallis, Ore- gon. Length, 10.5 mm. Eyepatches tri- angular, three rows plus single ocellus, totaling 17, on right side; 20 ocelli in four rows on left side. Antennal articles in order of length: .3, 5, 4, 2, 6 = 7, 1. Legs 4 through 7 with apophyses on femora; apophyses on legs 4 and 5 veiy small, distal; apophysis of leg 6 strong, capitate, mesal; apophysis of leg 7 strong, capitate and toothed, promixal. Anterior gonopods ( Fig. 59 ) larger than colpocoxites of poste- rior gonopods, subrectangular, flattened antero-posteriorly, slightly bent mesad, with aciculate subterminal process arising From a longitudinal ridge; lateral edges raguely serrate. Sternum of anterior gono- pods completely encircling coxotelopodites, produced between them as bilobed or trilobed condyle. Colpocoxites of posterior gonopods (Fig. 60) large and bulbous at base, ending in pointed terminal process, below which is a pilose area and a pointed subterminal process. Second joint of pos- terior gonopod with a distinct basal shoulder; third joint nearly twice as long as second. Notes. Nothing is known of the ecology and biology of this species, which is known only from the type locality (Map 6). Taiyutyla napa n. sp. Figures 63, 64 Type. Holotype male collected with two females and an immature specimen by Vincent Roth, 31 December 1953, Mt. St. Helena, Napa and Sonoma cos., California. The specific epithet refers to the type locality. Diagnosis. In both corvallis and fran- cisca, the second telopodite article of the posterior gonopods is longer than or as long as the first; in napa the second joint is less than half as long as the first. Only leg 3 has an apophysis on the fourth joint in napa; corvallis has leg 3 unmodified and francisca has apophyses on legs 4 through 7, as well as 3. The low swelling on the third joint of leg 2 present in napa is absent in the other two species. Description of liolotype male. Length, 8 mm. Eyepatches irregularly triangular, 3 irregular rows with a single ocelkis totalling 18 on right side; IS ocelli in four rows on left side. Antennal joints in order of length: 3, 5, 4, 2 = 6, 7, 1; joint 5 about twice as long as joint 4. Leg 3 with large distal apophysis on femur; leg 2 with a low, longitudinal swelling on the mesal side of third joint. Anterior gonopods (Fig. 63) curv^ed posteriad distally, with prominent mesal shoulder about midway in their length, and with a thin, membranous sub- terminal lamina extending mesally. Colpo- coxites of posterior gonopods (Fig. 64) intermediate between corvallis and fran- cisca, subterminal process short, pointed; terminal process long, curved posteriad, with single accessory tooth. First telop- 88 BuUetin Museum of Comparative Zoology, Vol. 141, No. 2 Map 6. Northern California and southern Oregon, showing distribution of various conotylids; dots, Plumatyla humerosa. circle, Conotylo extorns; solid squares, Ta/yutylo francsco; open squares, 7. napa,- triangles, T. corvol/is, including 7. jonesi. Map 7. Southern Indiana, showing distribution of Conofylo bo/lmoni; dots, epigean records; circles, cave records; dottec line, limit of lllinoisan glacial drift; broken line with dots, limit of Wisconsin glacial drift. odite joint more than twice as long as second. Notes. Nothing is known of the biology of this species, which is known only from the type locality (Map 6). Mt. St. Helena has a maximum elevation of 4344 feet. Taiyufyla francisca n. sp. Figures 61, 62 Type. Holotype male collected by C. W. O'Brien, 7 Januaiy 1962, one mile SE of Inverness, Marin Co., California. The specific epithet, a noun in appo- sition, refers to the proximity of the type locality to San Francisco Bay. Diaii,nosis. Distinct from corvallis pri- marily in the form of the colpoeoxites o\ the posterior gonopods, which bear spirally curved apical processes with several acces- sory teeth, also in the stronger modifi- cations of legs 4 and 5 of the males. Description of liolotype mole. Length 9.5 mm. Eyepatches triangular, 20 ocell: in four ro\\'s on left side, 20 ocelli in foui rows and single ocellus on right side. An- tennal segments in order of length: 3, 5, 4. 2, 6, 7, 1. Apophyses of legs 3 through "i approximately same size, distal on leg 3 slowly becoming mesal on succeeding leg? and proximal on leg 7. Anterior gonopodf MiLLiPED Family Conotylidae • Shear 89 (Fig. 61) large, flattened, terminal process blunt, slightly curved; large, thin mesal flange present. Colpocoxites of posterior gonopods (Fig. 62) smaller than anterior gonopods, complex; terminal process curved in a spiral of two turns, with four accessory processes as shown (Fig. 62); subterminal process a twisted, horizontal lamella finely laciniate on the posterior side. Second telopodite joint of posterior gonopod slightly longer than first, vaguely annulated. Notes. Little is known of the biology of this species; the holotype was taken in a Berlese sample of Piniis mtiricata duff. See Map 6. Genus Austrotylo Causey, 1961 Austwtyla Causey, 1961, Proc. Biol. Soc. Washing- ton, Vol. 74, p. 260 (in part). Soiwiati/la Hoffman, 1961, Trans. Ent. Soc. Amer., Vol. 87, p. 269. Type species. Of Austrotyhu ConotyJa specus Loomis, by original designation; of Sonoratyhi, ConotyJa montivaga Loomis. The generic name is a feminine Latin- Greek neologism referring to the southerly distribution of the genus with respect to Conotyla. Diapiosis. With the characters of the family. Anterior gonopod sternum appear- ing divided in some cases, but usually contiguous in the anterior and posterior midlines, or joined by sclerotic membrane. Anterior gonopods flattened, platelike, but with complex posterior surfaces; anterior gonopods covering colpocoxites of posterior gonopods in situ. Colpocoxites of posterior gonopods much smaller than anterior gono- pods, usually with a single cupped lamella and a rodlike or platelike mesal branch, sometimes plumose. Pregonopodal legs with femoral lobes on legs 3 and 4 in all known species. Capitate lobes present on coxae of legs 10 and 11 in some species. Species usually pigmented, 20-24 ocelli in triangular j)atch. Species. Five; distributed (Map 1) through Missouri and Illinois, Rocky Moun- tains from Alberta to Chihuahua (and possibly Queretaro). Separation of specus or its ancestors from the Rocky Mountain species may have occurred as early as the Pliocene, at the time of the formation of the Great Plains. Key to Species la. Coxae 10 and 11 of males without lobes .— horealis lb. Coxae of legs 10 and 11 with lobes, or either coxa 10 or 11 lobed 2 2a. Lobe on coxa 11 only ._ coloradensis 2b. Lobe on coxa 10 3 3a. Lobe on coxa 10 only 4 3b. Lobes on both coxae 10 and 11 specus 4a. Gonopods as in Figs. 68-70 rnontivaga 4b. Gonopods as in Figs. 77-79 chihuahua Austrotyla specus (Loomis) Figures 65-67 Conotyla specus Loomis, 1939, Bull. Mus. Comp. Zool., Vol. 86, p. 184, figs, lla-c. Chambedin and Hoffman, 1958, U. S. Nat. Mus. Bull. 212, p. 99 (list). Ausfrotyhis specus specus, Causey, 1961, Proc. Biol. Soc. Wash., Vol. 74, pp. 260-264, figs. 5-10. Austrotyla specus rnontivaga, (in part) Causey, 196.1, Proc. Biol. Soc. Wash., Vol. 74, pp. 264- 265. Sonoratyla specus, Hoffman, 1961, Trans. Ent. Soc. Amer., Vol. 87, p. 269. Type. Male holotype from Rice's Cave, 3 miles northeast of Goldman, Jeff- erson Co., Missouri, in Museum of Comparative Zoology, examined. The specific epithet is a noun in apposition (Latin: "cave") referring to the habi- tat of the type series. Diagnosis. This species has much less complex anterior gonopods than colora- densis or chihuahua, and those of horealis are much simpler. In addition, the coxal processes of legs 10 and 11 are unique. See under rnontivaga for a discussion of differences between that species and specus. Description of holotype male. Length, 11.5 mm. Eyepatches quadrangular, ocelli of both sides in three rows, 21 on the left side, 23 on the right, ocelli fully pigmented, not irregular in shape. Antennal articles in 90 Bulletin Museum of Contparativc Zoology, Vol. 141, No. 2 0. 3 mm Figures 65-73. Gonopods of Austrotyla. Figs. 65-67. Austrotyla specus. Fig. 65. Right posterior gonopod, anterior view. Fig. 66. Anterior gonopods, posterior view. Fig. 67. Left anterior gonopod, lateral view. Figs. 68-70. A. montivaga. Fig. 68. Right posterior gonopod, anterior view. Fig. 69. Anterior gonopods, posterior view. Fig. 70. Left anterior gonopod, lateral view. Figs. 71-73. A. borealis. Fig. 71. Right posterior gonopod, anterior view. Fig. 72. Anterior gonopods, posterior view. Fig. 73. Left anterior gonopod, lateral view. MiLLiPED Family Conotylidae • Sheai 91 order of length: 3, 5, 4, 6, 2, 7, 1. Pre- gonopodal legs modified as described for the genus, legs 10 and 11 with anteriorly directed capitate lobes on the coxae. An- terior gonopods (Figs. 66, 67) with the sternal lobes evenly and heavily sclerotized, coxotelopodites with a lateral lobe above the sternal lobes; proximal laciniated branches large, prominent. Colpocoxites ( Fig. 65 ) with the posterior lamella cupped and rounded distally, anterior branch sig- moid, laciniated distally; third telopodite joint twice the length of the second. Notes. Both Causey and Loomis over- looked the lobes on the coxae of the Uth legs. Causey noted the variation in color- ation and ocelli of this species, and a study of other populations indicates that both pigmented and unpigmented individuals occur in the same caves. Epigean speci- mens are rare; Causey reported them as A. s. montivaga from northern Illinois and Wisconsin. It seems likely that the same situation holds in this species as in Conotyla blakei, with troglophilic populations in the southern, lowland part of the range and epigean populations in the northern part. The absence of epigean records of either species in the regions of the cave popu- lations may be due to a lack of collecting at the proper time, late fall, winter, and early spring, when these animals mature and are most active. Unpublished records kindly given to me by Stewart Peck include :-aves in Jackson Co., Iowa, and in the Following Illinois counties: Jackson, Jo Daviess, Monroe, Saline, Henderson, and Union. Causey (1961) reported it from Franklin, Jefferson, and St. Clair cos., Missouri, and Sauk Co., Wisconsin (Map 5). An immature specimen in the Museum jf Comparative Zoology from Blue Earth oo., Minnesota, strongly resembles A. >pecus. ^usfrotyla montivaga (Loomis) Figures 68-70 'Conotyla montivaga Loomis, 1943, Bull. Mus. Comp. Zool., Vol. 92, pp. 383-384, figs. 4a-d. Chambeilin and Hoffman, 1958, U. S. Nat. Mus. Bull. 212, p. 98 (list). Sonoratyla montwap,a, Hoffman, 1961, Trans. Ent. Soc. Amer., Vol. 87, p. 268, pi. 10, figs. 10-11. Austrotijla specus montivaga, (in part) Causey, 1961, Proc. Biol. Soc. Washington, Vol. 74, p. 264. Type. Male holotype from Santa Rita Mts., elevation 7500 ft., Pima Co., Arizona, in Museum of Comparative Zoology, examined. Diagnosis. This species is much smaller than coloradensis, and the posterior surface of the anterior gonopods is much less com- plex; there is no clear area on the lateral sternal lobes as there is in coloradensis and chihuahua. Distinct from specus, which it closely resembles, in the details of the gono- pods; in specimens I have examined, the third telopodite article is nearly three times as long as the second in montivaga, while in specus it is only slightly more than twice as long; there is no coxal lobe on leg 11 in montivaga, and one is present in specus. Description of liolotype male. Length, 9.0 mm. Causey ( 1961 ) described in detail the nonsexual characters of this species, but stated that the gonopods (Figs. 68, 69) were identical to those of specus. This is not the case. In posterior view (Fig. 69), the proximal laciniated lobes of montivaga are smaller, the ridges and lobes are less developed, and there is no lateral extension on the coxotelopodite above the insertion of the sternal lobes. The colpocoxites ( Fig. 68) are less developed in montivaga, and the anterior branch is thicker and blunter. Otherwise, the description given by Causey (1961) is accurate. Notes. Known from numerous specimens collected in the Santa Rita and Santa Cata- lina Mts., north and south of Tucson, Pima Co., Arizona. The type is from an elevation of 7500 ft., probably in or just below the Pinus ponderosa zone at the latitude of Tucson. The identity of specimens from Mescalero, New Mexico, could not be checked, as they were not available for study, but Chamberlin reported colora- 92 BuUctin Museum of Comparative Zoo/ogry, Vol 141, No. 2 densis from Riiidosa, in the same mountains and onl\' about 20 mil(\s to the northeast. ^^"hie]l spi^eies (or an undescribed one) aetually oecurs in soutliem New Mexico awaits clarification. Both Loomis and Chamberhn mi<2;ht have assumed tliat their species was the only one in the Rocky Mountains; coJorodensis males were not described until 1961. The records reported by Causey ( 1961 ) from Illinois and Wis- consin refer to epigean populations of specus; she stated that the gonopods were identical with speciis, and placed monti- V(ia(l 2-4 October, 1967, Jasper National Park, Alberta, "Sta. 5," 5300 ft. The specific epithet is an adjective indicat- ing that it is the extreme northerly repre- sentative of its genus. Diagnosis. This species is smaller than the other representatives of the genus, being about 9 mm long in mature males. The stenia of the anterior gonopods show very distinct angular shoulders while still attached to the telopodites, as in specus. The posterior surface of the coxotelopodites is simpler than in any other species, and there are no lobes on the coxae of legs 10 and 11. Description of holotijpe nude. Length, 9.S mm. Eyepatches triangular; 24 ocelli on left side, in four rows plus single ocellus; four ro\\s plus single ocellus totalling 22 ocelli on right side. Antennal segments in order of length: 3, 5, 4, 6, 2, 7, 1. Legs 3 and 4 modified as described for genus. Lateral lobes of the sterna of the anterior gonopods forming a distinct angular shoulder with the telopodites (Figs. 72, 73), with a distal area of very thin cuticle; spiracle on anterior surface, easily seen. Coxotelopodites of anterior gonopods simple, with single thickened ridge on posterior surface, ending in somewhat swollen knob. Colpocoxites of posterior gonopods (Fig. 71) with anterior branch long, filiform, branched. First tc^lopodite joint is less than one third length of second telopodite joint. Coxae 10 and 11 without lobes. Notes. Known only from the type lo- cality. See notes on ConotyJa olhertana for further details on type locality. Ausfrofyla coloradensis (Chamberlin) Figures 74—76 Conott/la coloradensis Chamberlin, 1910, Ann. Ent. Soc." Amer., Vol. 3, p. 237, pi. 32, figs. 7-9, pi. 33, figs. 1-3. Chamberlin and Hoffman, 1958, U. S. Nat. Mus. Bull. 212, p. 98 (list). Austrotijla coloradensis. Causey, 1961, Proc. Biol. Soc. Washington, Vol. 74, pp. 254-260, figs. 2-4. Type locolity. Colorado. Causey (1961) designated a male neotype from Allen's Park, Boulder Co., Colorado, which is deposited in the Museum of Compar- ative Zoology, examined. Diagnosis. Distinct in size and com- plexity of gonopods from all except duJiita- hua, but coloradensis has no anterior lobe on the coxa of leg 10. Distinct from monti- vaga by the larger size and much more prominent lateral shoulders. Description of neotype male. Length, 15.7 mm. Eyepatches triangular, 23 ocelli on each side in four rows plus single ocellus. Antennal articles in order of length: 3, 5, 4, 6, 2, 7, 1. Pregonopodal legs modi-' Figures 74-81. Gonopods of Austrotyla and Plumatyla. Figs. 74-76. Austrotyla coloradensis. Fig. 74. Right posterior gono- pod, anterior view. Fig. 75. Anterior gonopods, posterior view. Fig. 76. Left anterior gonopod, lateral view. Figs. 77-79. MiLLiPED Family Conotylidae • Shear 93 0.3 mm ^- chihuahua. Fig. 77. Colpocoxite of righf posterior gonopod, anterior view. Fig. 78. Anterior gonopods, posterior view, ■ig- 79. Left anterior gonopod, lateral view. Figs. 80-81. Plumatyla humerosa. Fig. 80. Right anterior gonopod, anterior lew. Fig. 81. Left posterior gonopod, posterior view. 94 Bulletin Museum of Comparative Zoology, Vol. 141, No. 2 fied as described for genus, apophysis of leg 3 sonie\\'hat larger than that of leg 4. Anterior gonopods (Figs. 75, 76) thin, lamellate, bearing posteriorly a prominent knob near the midline of each telopodite and a mesal laciniated branch; sternal lobes with a semicircular area of very thin cuticle that appears to be a hole at low magnification. Colpocoxites of the posterior gonopods (Fig. 74) relatively large, deeply cupped, posterior lamella deeply notched laterally, anterior branch small, lightly laci- niated; second telopodite joint twice the length of first. Notes. Causey's (1961a) drawings of the gonopods, made from cleared, slide- mounted material, leave something to be desired. The thin area on the sternal lobes is represented as an open space, and the tracheal apodeme is shown as being coalesced with the sternal lobe high up on the coxotelopodite. In actuality, the tracheal spiracle, not seen by Causey, is easily visible on uncleared preparations and is at the lateroanterior comer of the stemiun, whence it leads normally into the tracheal apodeme. The horizontal portions of the sternites meet in the anterior midline as well as the posterior, rather than being represented entirely by a membranous area. The colpocoxites of the posterior gonopods in Causey's illustration show neither the deep lateral cleft in the posterior lamella, nor the short, mesal plumose branch. See Causey (1961a) for emendations in Cham- berlin's original description, based on a female. There is some variability in size in this species, but it could not be connected with any geographical trend. The smallest speci- mens were about 16 mm long and the longest were close to 23 mm long. KnowTi from numerous specimens from the following Colorado counties (Map 4): Larimer, Jackson, Eagle, Pitkin, Chaffee, Gunnison, Hinsdale, Mineral, and Conejos. Probably also occurs in southern Wyoming and northern New Mexico. The majority of records are from coniferous forests above 7000 ft. elevation. Austrotyla chihuahua n. sp. Figures 77-79 Type. Male holotype from 100 m above Rio Urique, 84 km south of Creel, Chihuahua, Mexico, collected Febru- ary 28, 1966, by J. Reddell and W. Bell. The specific epithet is a noun in apposition, referring to the type lo- cality. Diaf^nosis. The posterior surface of the anterior gonopods is complex, as in colora- dc'nsis\ but the colpocoxites of the posterior gonopods of chihuahua have the posterior lamella angular and the anterior branch large, flattened and reflexed; chihuahua has a coxal lobe on leg 10, while colora- densi.s has none. Description of holotype male. Length, 14.0 mm. Eyepatches tiuncate-triangular, 23 ocelli in four rows on left side, 21 ocelli in four rows on right side. Antennal articles in order of length: 3, 5, 4, 2 = 6, 7, 1. Legs 3 and 4 modified as described for genus, apophyses slightly larger than in other species. Sterna of anterior gonopods as in coloradensis; gonopods (Figs. 78. 79) almost contiguous in midline, distinctly depressed mesally on anterior surface; posterior surface with complex knobs and plumose branches. Colpocoxites of poste- rior gonopods (Fig. 77) with posterior lamella triangular; anterior branch flat- tened, reflexed. Coxa of leg 10 \\'ith an anterior lobe. Notes. Nothing is known of the biology of this species, which is known only from the type locality. Genus Plumafyla, new genus Austrotijla Causey, (in part) 1961, Pioc. Biol. Soc, Washington, Vol. 74, p. 260. Sonorattjla Hoffman (in part) 1961, Trans. Amer Ent. Soc, Vol. 87, p. 269. Type species. Conotyla humerosc Loomis; the generic name is a Latinized MiLLiPED Family Conotylidae • Shear 95 Spanish-Greek neologism derived from the related genus Conotyla and Plumas Co., California, type localit)- of the type species, rhe gender is feminine. Diagnosis. With the characters of the Family. Anterior gonopod sternum inter- mediate between Toiyutyla and Austrotyla, leavily sclerotic throughout, with lateral obes extending laterad to coxotclopodites, out incomplete posteriorly, as in Cono- "ylo. Anterior gonopods with two major tranches, the anterior largest and set nesally with small laciniate processes; posterior branch a simple hirsute rod. Posterior gonopod sternum broadened and lepressed laterally. Colpocoxites of poste- •ior gonopods with a large mesal branch )earing laciniations as in anterior gonopod; ateroposterior lamella heavily sclerotized. ^emoral lobes on some pregonopodal legs. Jpecies troglobitic, without pigment, ocelli ibout 10, in two rows. Species. One, found in mines and caves n northern California. '/umofy/a humerosa (Loomis) Figures 80, 81 lonotijla humerosa Loomis, 1943, Bull. Mus. Comp. Zool., Vol. 92, pp. ,384-385, figs. 5a-cl. Type. iMale holot\'pe and other speci- mens collected Sunnyside Mine, 3 mi. SW of Seneca. Plumas Co., California, January 22, 1923, by H. S. Barber; deposited in U. S. National Museum, examined; immature male paratype in Museum of Comparative Zoology. Diagnosis. See generic diagnosis. Description of topotype male. Length, 6.0 mm. Ocelli in two rows, 9 ocelli on ?ft side, 8 ocelli on right side. Antennal 3gments in order of length: 3, 5, 4, 2, 6, , 1. Legs 5, 6, and 7 with prominent smoral lobes. Anterior gonopods (Fig. 0) larger than colpocoxites; posterior ranch bears a subterminal lateral hook nd is distally bifid; posterior branch rod- ke, densely pilose. Posterior basal knob ts into sternal cavity of posterior gono- pods. Colpocoxites of posterior gonopods (Fig. 81) with two branches, anterior mesal branch hooked posteriad, small laciniated branches on lateral surface; poste- rior lateral branch a subtriangular, curved, well-sclerotized lamella. Tracheal apodemes of both gonopods reduced in size. Notes. The only other mature specimens known to me, excepting the specimens from the type locality kindly lent to me by N. B. Causey, are from Indian Wells Ice Cave, Lava Beds National Monument, Siskiyou Co., California. They differ from the holotype and paratype by being smaller (13-14 mm), lacking a lobe on the femur of leg 5, and ha\dng the anterior branch of the anterior gonopod slightly more attenu- ate. Otherwise, the gonopod structure is identical to the Sunnyside Mine specimens. For this reason, I hesitate to describe it as a distinct species. Immature representatives of Plumatyla are known from a number of mines, limestone caves, and lava tubes in northern California and adjacent Oregon. Only the collection of mature specimens can iudicate the range of variation and the number of species in this genus. The lava tubes in Lava Beds National Monument may be as old as 60,000 years (Gale, 1959). The local glaciation of the northern Cali- fornia area during the late Pleistocene (Detling, 1968) may have been responsible for the cave habitat of these animals. LITERATURE CITED Barr, T. C. 1968. Cave ecology and the evo- lution of troglohites. Evol. Biol., 2: 35-102. Broleaiann, H. W. 1935. Myriapodes Diplopodes ( Chilognatlies I) Faiiue dc France, 29, Paris. 368 pp. Buckett, J. S., AND M. R. Gardner. 1967. A new family of cavemicolous millipedes with tlie description of a new genus and species from Idaho (Diplopoda: Chordeumida: Chor- deumidea). Michigan Ent., 1(4): 117-126. Causey, N. G. 1952. Four new chordeumid millipeds from the United States (Nema- tophora, Chordeumidea). Proc. Biol. Soc. Washington, 65: 111-118. . 1961a. Austrotyla, a new milliped genus ( Chordeumidea : Conot>'lidae : Conotylinae ) . Proc. Biol. Soc. Washington, 74: 251-266. 96 Bulletin Museum of Comparative Zoology, Vol. 141, No. 2 . 19611). Adrityh, a new millipecl genus (Chordeuinidea: Conotylidae). Psyche, 68: 131-136. Chamberlin, R. V. 1941. New western millipeds. Bull. Univ. Utah, Biol. Sen, 6(5): 3-23. . 1951. Eleven new western millipeds. Natur. Hist. Misc. No. 87: 1-12. 1952. Two Oregon millipeds of the Order Chordeumida. Nat. Hist. Misc., No. 113: 1-3. Chamberlin-, R. V., Axn R. L. Hoffman. 1958. Checklist of the millipeds of North America. U. S. Nat. Mus. BuU., 212: 1-236. Cook, O. F. 1896. On recent diplopod names. Brandtia, 2: 8. Cook, O. F., and G. N. Collins. 1895. The Craspedosomatidae of North America. Ann. New York Acad. Sci., 9(1): 1-100. CusHiNG, E. J. 1965. Problems in the Quaternary phvtogeography of the Great Lakes region. /h'h. E. Wright, Jr. and D. G. Frey [Ed.] The Quaternary of tlie United States. Prince- ton, N. T-: Princeton University Press, pp. 403-417. " Detling, L. E. 1968. Historical background of the flora of the Pacific Northwest. Bull. Mus. Nat. Hist. Univ. Oregon, 13: 1-57. FrYE, J. C, H. B. WiLLMAN, AND R. F. BlACK. 1965. Outline of the glacial geology of Illi- nois and Wisconsin. In H. E. Wright, Jr. and D. G. Frey [Ed.] The Quaternary of the United States. Princeton, N. J.: Princeton University Press, pp. 43-61. Gale, R. R. 1959. Geology of Lava Beds National Monument. Bull. Nat. Speleol. Soc, 21(2): 62. Hoffman, R. L. 1961. Systematic and morpho- logical notes on North American conotyloid Diplopoda. Trans. Amer. Ent. Soc., 87: 259- 272. Hoffman, R. L. 1963. Notes on the structure and classification of the diplopod family , Heterochordeumatidae. Ann. Mag. Nat. Hist., ' 13(6): 129-135. Howden, H. F. 1963. Speculations on some beetles, barriers, and chmates during tlie Pleistocene and Pre-Pleistocene periods in some non-glaciated portions of North America. Syst. Zool., 12(4): 178-201. LooMis, H. F. 1939. The millipeds collected in Appalachian caves by Mr. Kenneth Dearolf. Bull. Mus. Comp. Zool., 86(4): 165-193. . 1943. New cave and epigean millipeds. Bull. Mus. Comp. Zool., 92(7): 373-410. Moss, E. H. 1955. Vegetation of Alberta. Bot. Rev., 21: 493-567. ScHUBART, O. 1934. Tausendfiissler oder Myriap- oda. 1: Diplopoda. In Die Tierwelt Deutsch- lands, 28 Teil. Jena: Gustav Fischer, 318 pp Shear, W. A. 1969. Cave miUiped genera of tht United States. Psyche, 76(2): 126-143. Verhoeff, K. 1913. Ascospermophoren aus Japan Zool. Anz., 43: 342-370. . 1932. Diplopoden-Beitrage (124. Diplo poden-aufsatz ) . Zool. Jahrb., Syst. Arb., 62: 469-524. 1942. Ascospermophoren aus Japan unc iiber neue japanische Diplopoden. Zool. Anz. 137: 201-217. Wayne, W. J., and J. H. Zumherge. 1965. Pleisto cene geology of Indiana and Michigan, h H. E. Wright, Jr. and D. G. Frey [Ed.] Th. Quaternary of the United States. Princeton- N. J. Princeton Univ. Press, pp. 63-83. [Received 10 October 1969.) MiLLiPED Family Conotylidae • Shear 97 Valid generic and species names in italics; Achemenides, 84 Adritylidae, 57-58 aeto, Conotyhi, 76 albeitana, Conotyla, 82 utrolineata, Conotyla, 80 \twliucata group of Conotyla, 80 Atistrotyla, 89 hlakei, Conotyla, 67 Blakei group of Conotyla, 67 blakei, Proconotyla, 67 boUmani, Conotyla, 69 oollmani, Trichopetalum, 69 horealis, Atistrotyla, 92 zeleno, Conotyla, 80 :hihiiahua, Austrotyla, 94 Uleidogonidae, 56 :oloraclen.sis, Austrotyla, 92 :oloradensis, Conotyla, 92 'Jonotyla, 64-67 Donotylidae description, 58 biology, 58-60 gonopod structure, 61-62 key to genera, 64 in key to related families, 57 problematical names, 62-64 taxonomic position, 56 ^orvallis, Taiytttyla, 87 Diplomaragnidae, 56 dpcnor, Conotyla, 73 Zudifiona, 56 uuligoninae, 56 'xtorris, Conotyla, 70 ischeri, Conotyla, 72 uscheri group of Conotyla, 72 rancisca, Taiyutyla, 88 ^lomeratum, Trichopetalum, 63 INDEX only major discussions listed for species. humerosa, Phimatyla, 95 humerosa, Conotyla, 95 Idagonidae, 56, 57 Japanosoma, 56 Japanosomatinae, 57 jonesi, Conotyla, 63 leibergi, Cookella, 63 Marquetia ( Opisthocheiridae ) , 57 mclinda, Conotyla, 77 Melinda group of Conotyla, 77 montivaga, Austrotyla, 91 montivaga, Conotyla, 91 montivaga, Sonoratyla, 91 napa, Taiyutyla, 87 ocypetcs, Conotyla, 76 oregona, Bollinanclla, 63 pectinatus, Achemenides, 84 pectinata, Conotyla, 84 pectinata, Sonoratyla, 84 personata, Conotyla, 73 pliana, Zygotyla, 63 Plumatyla, 95 Proconotyla, 64 smilax, Conotyla, 75 Smilax group of Conotyla, 75 Sonoratyla, 89, 94 specus, Austrotyla, 89 specus, Conotyla, 89 specus, Sonoratyla, 89 Taiyutyla, 86 Trichopetalidae, 55, 56, 57 vaga, Conotyla, 67 venetia, Conotyla, 75 Venetia group of Conotyla, 75 visia, Conotyla, 79 Wyandotte, Conotyla, 69 Wyandotte, Scotherpes, 69 mwmM^^Mm^Wh^^:^v'/> : \ • OF THE Museum of Comparative Zoology Monograph of the Cuban Genera Emoda and Glyptemoda (Mollusca: Archaeogastropoda: Helicinidae] WILLIAM J. CLENCH AND MORRIS K. JACOBSON HARVARD UNIVERSITY VOLUME 141, NUMBER 3 CAMBRIDGE, MASSACHUSETTS, U.S.A. 4 FEBRUARY 1971 PUBLICATIONS ISSUED OR DISTRIBUTED BY THE MUSEUM OF COMPARATIVE ZOOLOGY HARVARD UNIVERSITY Bulletin 1863- Breviora 1952- Memoirs 1864-1938 JoHNSONiA, Department of Mollusks, 1941- OccAsiONAL Papers on Mollusks, 1945- Other Publications. Bigelow, H. B., and W. C. Schroeder, 1953. Fishes of the Gulf of Maine. Reprint, $6.50 cloth. Brues, C. T., A. L. Melander, and F. M. Carpenter, 1954. Classification of Insects. $9.00 cloth. Creighton, W. S., 1950. The Ants of North America. Reprint, $10.00 cloth. Lyman, C. P., and A. R. Dawe (eds.), 1960. Symposium on Natural Mam- malian Hibernation. $3.00 paper, $4.50 cloth. Peters' Check-list of Birds of the World, vols. 2-7, 9, 10, 12-15. (Price list on request. ) Turner, R. D., 1966. A Survey and Illustrated Catalogue of the Teredinidae ( MoUusca: Bivalvia ) . $8.00 cloth. Whittington, H. B., and W. D. I. Rolfe (eds.), 1963. Phylogeny and Evolution of Crustacea. $6.75 cloth. Proceedings of the New England Zoological Club 1899-1948. ( Complete sets only. ) Publications of the Boston Society of Natural History. Authors preparing manuscripts for the Bulletin of the Museum of Comparative Zoology or Breviora should send for the current Information and Instruction Sheet, available from Mrs. Penelope Lasnik, Editor, Publications Office, Museum of Comparative Zoology, Harvard University, Cambridge, Massachusetts 02138, U.S.A. © The President and Fellows of Harvard College 1971. MONOGRAPH OF THE CUBAN GENERA EMODA AND GLYPTEMODA (MOLLUSCA: ARCHAEOGASTROPODA: HELICINIDAE) WILLIAM J. CLENCH AND MORRIS K. JACOBSON TABLE OF CONTENTS Introduction 99 Acknowledgments 101 Genus Einoda H. and A. Adams 101 Key to Species 103 Emoda siJacea (Morelet) 104 £. sa^iaiana (Orbigny) 105 E. pulclicrrima pidcherrima (Lea) 107 E. ))ulchcrrima titanica (Poey) 108 E. suhmarginata (Gray) 109 E. ciliata (Poey) 110 E. cilUita f^uisana (A. J- Wagner) 111 E. emoda (Pfeifler) 112 E. cah'doniensis, new species 112 E. huijcimensis (Poey) 113 E. hcrmudczi Aguayo & Jaume _„ 114 E. najazaensis Aguayo & Jaume 115 E. hriarca (Poey) 116 E. mdijarina maijarina (Poey) 117 E. inaijdrina guticrrezi Aguayo & Jaume ... 117 £. maijarina mirandensis Aguayo & Jaume . 118 £. hlanesi Clench & Aguayo 118 Genus Glt/ptcmoda Clench & Aguayo "118 Ghjptemoda torrei torrei (Henderson) 119 G. torrei freirei Clench & Aguayo 120 Emoda (?) dementis Clench & Aguayo .. 120 References Cited 121 Index 130 ABSTRACT The present paper is the third in a series deal- ing with the Cuban representatives of the land prosobranch family Helicinidae. The studies on the genus Viana (1968) and Priotrochatella (1970) have already appeared. The genus Emoda is assumed to have originated in Oriente Province, probably from some form of tlie West Indian genus Alcadia. From there it spread westward across the island into Pinar del Rio Province. It Bull. Mus. Comp occurs in each of the six provinces, but is absent from the Isle of Pines. The specific and sub- specific taxa assigned to Emoda were investigated. Of the 27 taxa proposed, it is concluded that 16 are valid. In addition, one new species, Emoda caledoniensis, is described and the subgenus Ghjptemoda Clench & Aguayo 1950 is raised to generic rank. INTRODUCTION The family Helicinidae appears in two widely disjunct areas of distribution cen- tered in the Caribbean in the western hemi- sphere and in Southeast Asia and the Philippine Islands in the eastern hemi- sphere. The island of Cuba has an espe- cially rich helicinid fauna, possessing no fewer than four endemic genera with numerous species. In a previous study (1968), the present authors have mono- graphed the endemic genus Viana, and in 1970, their monograph on PriotwchateJki was published. In this work we take up two more such genera, Emoda and Ghjpte- moda. In this series of studies, we also plan to complete the examination of all Cuban helicinids and to investigate the Cuban spe- cies of the genera HeJicina sensu stricto, EutrochateUa, Alcadia, Lucidella, Cerato- discus, and Froserpina. Lea (1834b: 161) described the first Emoda as Ilelicimi pidcherrima and until the appearance of Wagner's work (1907- 1908), most other new species were in- cluded in the genus Ilelicimi. The name . ZooL, 141(3): 99-130, February, 1971 99 100 Bnlk'tin Miisemn of Comparative Zoology, Vol. 141, No. 3 Emocla was given by H. and A. Adams in 1856 to a conglomerate of species which had few cohesive affinities. Because of this, the name was disregarded by later students of Cuban mollusks (Arango, 1878-1880; Crosse, 1888), and it first entered into general use after Wagner (1907-1908) published his extensive monograph. Al- though this author used only conchological and opercular characteristics, and in spite of serious defects in his work, he defined the taxon, which he considered a subgenus of Alcadia Gray, so successfully that all the species he considered to belong to E7noda are still considered to do so today. He confined the group to Cuba. H. B. Baker ( 1922), f)lacing Emocla as a subgenus of Scliasicheiki Shuttle\\'orth, examined the radulae of several species and chose HeUcina silacea Morelet 1849 as the type. (See below for a further discussion of the generic and subgeneric placement of Emocla.) Aguayo & Jaume (1954) in- troduced several new taxa and briefly reviewed some of the older ones. Twenty- seven taxa have been proposed for the members of this genus; of these we recognize 16 and add one new taxon. Of the 17, 11 occur only in Oriente Province, 3 or possibly 4 in Las Villas, 3 in Cama- giiey, and 2 in Pinar del Rio. Havana and Matanzas have only the widely ranging E. .submarginata, which is also the only species occurring in more than one province. We have records of its occur- rence in each of the six provinces of Cuba. All the other forms are found in only a single province and most of them in nar- rowly limited localities ( Pis. 4, 5; Table 1 ) . Emocla is confined to the mainland of Cuba and to at least one of its satellite cayos or keys (Pi. 4). No species is known to occur on the Isle of Pines or elsewhere in the West Indies. Emocla probably had its origin in the eastern end of Cuba and especially in Oriente Province, since many more species exist in that area than else- where on the island. The most widely spread species in Oriente is £. pulcherrima and, with no opposing fossil records, it might be viewed as an early, if not the earliest, form of the genus. E. su])marg,in- ata, which, \\'ith little difficulty, can be derived from pulcherrima, was in all likeli- hood the foriu ^\'hich, because of its ability to populate lowland as well as hilly areas, spread the genus to other parts of the island. Emocla hermudezi and E. najazaen- sis, two forms found in Camagijey, the province immediately adjacent to Oriente, can be easily derived from suhmarginata and probably resulted from isolation in the mountain ranges of Cubitas and Najaza respectively. The species foimd in Oriente occupy calcareous areas isolated by intei^vening volcanic rock. The isolation in the more western parts of the island results from mountain areas separated by noncalcareous lowlands. This isolation must have oc- curred during the late Tertiary, probably dating from the early Pliocene when Cuba more or less attained its present form. Only in this way can we account for such sharp differentiation as that shown, for example, between the two neighboring species ciliata and hriarea in southern Las Villas Pro\'ince. The only area of clear overlap appears to be in tlie eastern part of Pinar del Rio Province at the Sierra del Rosaria area of the Sierra de los Organos. Here E. sub- marginata and sagraiana both occur. The small form of typical sagraiana (see below, p. 105) is more easily derived from .siib- marginata than is the larger form called "percrassa." It may be assumed that the former then is ancestral to the latter. As will be shown, however, isolation of the two sagraicma forms did not take place, and both, with numerous degrees of inter- gradation, occur throughout the range. Species of Emoda, unlike Viana, are found in each of the six provinces of Cuba and occur from the eastern tip near Cabof Maisi in Oriente Province to the westem limits of the Sierra dc los Organos in Pinar del Rio. They are largely confined to tht mountain areas, except for one species Emoda and Glyptemoda in Cuba • Clencli and Jacobson 101 with the widest distribution, E. suhmaniin- ata (Gray), which also occurs in the low- lands. Though Emoda is for the most part an upland group, it is not as closely con- fined to a limestone substrate as Viana, and it is found on the ground under rotting leaves, and on branches and vines, as well as on calcareous rocks. Alcodia striatum (Lamarck) from Puerto Rico, the type-species of the subgenus Striatemoda H. B. Baker 1940, is obviously not an Emoda, despite Baker's doubts on this point (1940: 71). Among other dif- ferences such as size, operculum, and color, it also lacks the diagonal, slightly curved axial furrows on the protoconch. The fact that it has no spiral sculpture is not signifi- cant, since most species of Emoda also lack this feature. We have not been able to examine or at least confirm the location of some pertinent type material. Dance (1966) reported that the nonmarine shells of Pfeiffer, among which were many types of Cuban land shells, became part of the Dohni collection in the Stettin (Szczecin) Museum, Poland, where it was totally destroyed during World War II. A personal communication (July 1968) from the museum director. Dr. W. Filipo\\'iak, confirmed this fact. The words "type destroyed" in the text of this study reflect this situation. The types of the species described by Poey and some described by Cundlach. are presumably in the Museo Poey^ in the University of Havana. In spite of repeated requests, we have failed to obtain permis- sion from the museum authorities to ex- amine the collection, nor has it been possible for us to have the presence of this material in the collection verified. Never- theless, we suppose that the material is there and indicate this assumption by ^ A new public museum has been established recently, with exhibits housed in tlie former Capitol building in Havana. This museum is called the Museo Fehpe Poey and should not be confused with the older Museo Poey in the Uni- versity of Havana. writing "Type, probably MP" in the body of the text. The specimens examined are in the col- lection of the MCZ, unless otherwise noted. ABBREVIATIONS USED: MCZ Museum of Comparative Zool- ogy, Cambridge, Massachusetts. USNM United States National Mu- seum, Washington, D. C. MP Museo Poey, University of Ha- vana, Havana. BM(NH) British Museum (Natural His- tory), London. ACKNOWLEDGMENTS We gratefully acknowledge the generous help of Drs. Harald A. Rehder and Joseph P. E. Morrison of the United States National Museum, who made much com- parative material available to us. Drs. Ken- neth J. Boss and Ruth D. Turner, and Mr. Richard I. Johnson, all of the Museum of Comparative Zoology, read the manuscript and offered many valuable suggestions. We are also especially grateful to Dr. Turner for providing the anatomical and radular discussions for the present study. We are grateful to Mr. J. F. Peake and Mrs. Angela Cane of the British Museum (Natural History) for supplying us with the BM(NH) catalog numbers of some of Orbigny's types and the illustration of the type of Helicina crassa Orbigny. Dr. A. Riedel of Warsaw, Poland, kindly supplied the catalog numbers of two types. Dr. Hor- tensia Sarasiia of the Academia de Ciencias de la Republica de Cuba most graciously provided us ^^'ith excellently presented alcoholic material from Pinar del Rio. This study is part of the work done under NSF Grant No. GB 1004. Genus Emoda H. and A. Adams Emoda H. and A. Adams 1856, the Genera of Recent Mollusca, London, 2: 304 [as a subgenus of Helicina Lamarck 1799], (type-species, Heli- cina silacca Morelet 1849, subsequent designation H. B. Baker 1922: 56). 102 Bulletin Museum of Comparative Zoology, Vol. 141, No. 3 Description. Shell ^cncralK about 15 to at times is strongly sculptured. Alcadia is 30 mm in diameter, imperforate, depressed generally smooth, unicolored, or white with to moderately depressed, to subglobose, shades of brown or pink; these various dull or moderately glossy and generally colors are never arranged in bands. There solid. Color brown, green, yellow, or red- are also constant differences in the oper- dish puii-)k\ banded in occasional popu- cula, especially at the raised columellar lations or indi\idual specimens. Outer lip edge, which in Alcadia has the upper por- simple to considerabK- thickened, generally tion extended internally as a low, narrowly witli l)asal notch and /or protuberance, triangular protuberance, a condition not Axial sculpture variable: smooth or with found in Emoclo. Finally, Alcadia is pri- faint, irregular grow th lines or with regular, marily a Jamaican genus wdth some forms irregular, or wrinkU^d ribs. Spiral sculpture in Cuba and elsewhere, whereas Emoda generalK- wanting, or when present, con- occurs only on the main island of Cuba, sisting of rather faint incised lines or rows The Alcadia closest to Emoda in appear- of small pits. Periostracum weak to strong, ance is the species nuda Pfeiffer 1S66, and shining or lusterless, wanting in occasional its subspecies, E. n. bagaensis Aguayo 1953, species, sometimes laised in somewhat from Oriente. Here, however, an exami- hirsute spiral ridges. Operculum auricu- nation of the operculum will immediately late, concave, thickest at columellar margin, reveal the true relationship, thmning rapidly and becoming bladelike The differences between Schasicheila at outer margin. External calcareous layer and Emoda in shell and opercular structure of operculum nearly smooth, microscopi- are even more pronounced. The shells of cally pitted, white, light browTi, or reddish; Schasicheila are colorless and much more internal chitinous layer reddish or brown- fragile, and the roundly inflated base gives ish, marked by irregular, semicircular it a shape that is never seen in Emoda. growth lines. Columellar edge raised in (See below for further discussion.) It also narrow ridge, generally straight but bent differs in the nature of the peristome at its inward at top. forming a short, obli(iue insertions in the basal angle and the notch furrow intemally. Nucleus lateral and in the upper columellar angle. The wide central on colmnellar edge. parietal callus, found in all species of Remarks. Emoda has at various times Emoda, is wanting in Schasicheila. Finally, been regarded as a subgenus, either of the operculum, though superficially similar, Schasicheila Shuttleworth 1852 (Baker differs as follows: in Emoda it always 1922; Thiel{\ 1929) or of Alcadia Gray presents some color, either in the corneous 1840 (Wagner, 1907-1908; Keen, 1960). In layer alone or in both calcareous and cor- this study, we follow Baker (1926) and neons layers, wdiereas the operculum of Wenz (1938) and accord Emoda full Sc/!«.s/c/je//« is always white, with occasion- generic rank in the subfamily Helicininae. ally only a small area of light brown near Although the shells of Emoda have the columellar ridge. The columellar niar- several f(\itures in common with those of gin of the Emoda operculum, as \'iewed Alcadia. espccialK' with the subgenus Anal- from the side, is relati\'ely straight or only cadia Wagner, they are readily distinguish- very weakly sinuous and is more or less of able. In Emoda the shells are generally the same widtli throughout, whereas in larger, frequently considerably so, than the Schasicheila it is strongly bent inward at Alcadia from Cuba; the smallest Emoda the center, and is noticeably thickest at shells are about 15 mm in diameter, the midline and thinnest at both t(>rmi- whereas many species of Alcadia have nations. shells considerably smaller. Emoda gen- The radulae of E. silacea, sagraiami, and erally has a far more vivid coloration and ciliata w^ere described by H. B. Baker Emoda and Glyptemoda IX Cuba • Clench and Jacohson 103 ( 1922 ) . In this discussion, we are using Baker's concepts and terms as defined on his page 30. He found that the radula of Emoda is typically that of the subfamily Helicininae. It possesses the central tooth complex of seven plates (one rhachidian and three paired), and a strong capituli- form complex consisting of a strong, sub- rhomboid comb-lateral plate and a smaller, but heavy, accessory plate which is only weakh' articulated with the comb-lateral. Baker found some variation among the three species he studied and felt that ciliafa might belong in a special section. The paired central teeth and the comb-lateral are ^^'ell cusped, \\ith ciliata ha\dng a somewhat smaller number of cusps on the comb-lateral than the other two species. In addition, it has a few more inner bicuspid marginals than the other t\\'o: six or se\^en instead of four or five. Troschcl (1857: S2, pi. 5, fig. 12) described and figured the radula of E. suhmarginata in much more general outline. His figure shows the ac- cessory plate completely separated from the comb-lateral, whereas Baker demon- strates that the former articulates weakly ^^■ith the latter. Isenkrahe (1867) investigated the anat- omy of HeJicina titanica ( = Emoda pidcherrima titanica). Bourne (1911), al- though he did not specifically study any Emoda, was able to make certain recti- fications in Isenkrahe's work on the basis of his investigations of the closely allied genus Alcadia. Baker (1926) provided some notes on the anatomy of E. sagraiana and E. suhmarginata. Both Bourne (1911: 777) and Baker (1926: 35) commented on the general uniformity of the genitalia in the family and the uselessness of these characteristics for diagnostic purposes. For the present study. Dr. B. D. Turner of the Museum of Comparative Zoology dissected a specimen of E. sagraiana obtained from near Sumidero, Pinar del Bio, which Dr. Hortensia Sarasua of Marianao, Havana, Cuba, most graciously provided. A com- plete report of this investigation ^^'ill be published by Dr. Turner later in this series of studies. For the moment, we are pro- \iding a iew remarks and figures of the gross anatomy (PI. 6). In alcoholic specimens the color is \&.r\- able, ranging from nearly uniformly ivory in immature specimens to dark gray or black in the adult. The base of the ten- tacles and the mantle are dark gray, the color gradually growing paler on the dorsal part of the foot and becoming almost ivoiy at the foot margin and the distal portion of the tentacles. The tentacles are long and slender, somewhat wider proximally. The eye is located on a low peduncle on the outer part of the base of the tentacle. The sole of the foot is creamy white, rounded anteriorly and bluntly pointed posteriorly. The columellar muscle at the insertion edge is long, curved, cream colored, and rounded at the interior end. The odontophore is long and cur\^ed and is pro\'ided \\i\\\ a short hook at the inner termination. Other details of the anatomy can be found in Baker (1926: 48). Key to the Species of Emoda 1. Shell large, adult generally 24-32 mm in diameter 2 Shell smaller, adult generally 16-22 mm in diameter .10 2. Shell strongly sculptured 3 Shell smooth or with weak growth lines only 6 3. Sculpture regular 4 Sculpture irregular, of strong diagonal growth lines, shell yellow SILACEA 4. Sculpture of prominent spiral furrows, red- dish color _. EMODA Sculpture of axial lines only 5 5. Shell green, generally c. 24 mm in diam- eter B AYAMEXSIS Shell brownish red, generally c. 28-30 mm PULCHERRIMA TITANICA 6. Lip strongly expanded and strongly re- flected SAGRAIANA Lip weakly expanded, not reflected 7 7. Shell wine colored, with white band at periphery and subsuturally BRIAREA Shell without band - — 8 8. Shell rufous, periostracum oli\aceous CALEDONIEXSIS Shell yellowish green 9 104 Bulletin Museum of Comparative Zoology, Vol. 141, No. 3 9. Shell with spiral lines on periostracum only CILIATA CILIATA Shell with spiral lines incised on surface .— ^ CILIATA CUISANA 10. Distinct spiral as well as axial sculpture present - 11 Spiral sculpture very weak or absent 12 11. Shell with rounded carina, generally brown- ish -- SUBMARGIXATA Shell with subacute keel, generalK- faintl\- greenish NAJAZAENSIS 12. Shell with distinct notch at parietal in- sertion of upper lip BERMUDEZl Shell witliout notch 13 13. Shell shape turbinate PULCHERRIMA PULCHERRIMA Shell shape subglobose 14 14. Sculpture of diagonal, straight, regular lines BLANESI Sculpture of vermiculate wrinkles 15 15. Shell small, adult reaching 16 mm MAYARINA MAYARINA Shell larger, adult reaching 22 mm 16 16. Sculpture strong, shell reddish green MAYARINA GUTIERREZI Sculpture narrower, color reddish _____ MAYARINA MIRANDENSIS Emoda silacea (Morelet) Plate 1, figures 20, 21; Plate 5. Hdicina silacea Morelet 1849, Testacea Novis- sima Insulae Cubanae et Americae Centrahs, 1: 20 (Cuba; type-locality, here restricted. El Yunque de Baracoa, Oriente; 3 svntypes, BM(NH) 1893. 2.4.813-15). Hdicina ochracca Poey 1851, Memorias Historia Natural Isla de Cuba, 1: 112, pi. 11, figs. 1-4 (Baracoa; type, probably MP). Ilclicina silacea Morelet, Pfeiffer 1862, Novit. Conch., 2: 197, pi. 52, figs. 11-15. Alcadia (Emoda) silacea Morelet, Wagner 1908, in Martini & Chemnitz, Conch.-Cab., (2) 1: sect. 18, pt. 2, p. 100, pi. 18, figs. 10-12, 15. Description. Shell reaching 28.5 mm in diameter, depressed turbinate, rather rough and solid. Whorls 4V2, flattened, body whorl moderately inflated. Color pale greenish yellow, occasional specimens showing a faint brownish tinge, the color stronger at the interstices than on the axial cords. Some specimens with lighter pe- ripheral band bordered by subperipheral browTi band of varying width; spire and base general!}' lighter. Spire moderately raised, rounded. Aperture semilunate, outer lip unevenly curved, columella white, short, weakly convex above, concave below. Peristome strongly flaring but \\'eakly re- flected above, unevenly thickened, widest and thickest at or near the periphery, nar- row above where it merges into the parietal callus, and below where there is a small protuberance of varying shape. Parietal callus thin, white in the columellar area, transparent near outer margin. Suture weakly impressed. Penultimate and body \\horls prominently marked by more or less irregular, diagonal, somewhat wavy axial cords, separated by wider but quite shallow intervals. Sculpture weak on the early post- nuclear whorls. Protoconch IV2 whorls, lighter in color than the rest of the shell, flattened, marked with faint, curved, diag- onal, axial wrinkles. Periostracum thin and lusterless. Operculum as in genus, reddish in color, the internal corneous layer darker than the external, moderately lustrous calcareous lamina. Height Diameter mm mm 17.0 28.5 El Yunque, Baracoa, Oriente 17.0 27.5 El Yunque, Baracoa, Oriente 15.5 23.5 Mayari, Oriente 15.5 21.5 Finca "La Caridad," near Baracoa, Oriente Remarks. This species is well character- ised by its large size, prominent diagonal axial ornamentation, and striking uniformity of color, yellow predominating, with oc- casional specimens having a brownish tinge. At the type-locality, a large number of specimens have a lighter peripheral band bordered by a brown subperipheral band of varying ^^"idth. The typical forms seem to be confined to El Yunque itself, the specimens from Mayari being somewhat smaller and the ones from Finca "La Cari- dad" considerably so. This \'ariation in size in different populations is quite common in the genus, and may be associated with the size of the area inhabited and/or variations in other features of the locality, such as the availability of food, shelter, etc. E. silacea differs from pulclwrriyna tita^iica of about Emoda and Glyptemoda in Cuba • Clench and Jacobson 105 the same size, in color, in being more de- pressed, in having a thicker hp and more irregular and lower axial costae. Poey described ochracea because he failed to find the decussate base that More- let had mentioned in his description of silacea. However, as Pfeiffer pointed out (1862: 198), Morelet was in error because he did not have a large enough sample when he wrote his description and con- sidered an occasional ^^ariation to be a consistant feature. Specimens examined. ORIENTE. El Yunque de Baracoa; Mayari; Soledad, Guandoa, Baracoa; Finca "La Caridad" on road from Baracoa to Duaba. Emoda sagraiana (Orbigny) Plate 1, figures 16-19; Plate 4, figure 1. Helicina sagraiana Orbigny 1842, MoUusques, //! Sagra, Histoire Physique, Politique et Natuielle de I'lle de Cuba, 1: 240, pi. 18, figs. 12, 13 (I'inte- rieur de I'lle de Cuba; Cerro de Cuzco [not "aux environs de Trinidad"]; 4 syntypes, BM(NH) 1854.10.4.163). Helicina sagra Sowerby 1847, Thes. Conch., 1: 3, pi. 1, fig. 10, pi. 3, fig. 126 (Cuba [error for H. sagraiana Orbigny]). Trochatella (Viana) sagra "d'Orbigny" Chenu 1859, Manuel de Conchyhologie, Paris, 1: 496, fig. 3691 [error for sagraiana]. Helicina caialinensis Pfeiffer (July) 1856, Malak. Bliit. 3: 56 (prope Catalina, provincia occiden- talis insulae Cubae; type destroyed); Pfeiffer (Dec.) 1856, Novit. Conch., 1: 83, pi. 23, figs. 1-6. Helicina cataJiniana Sowerby 1866, Thes. Conch., 3: 278, pi. 1, fig. 8 [error for caialinensis]. Alcadia (Emoda) sagraiana d'Orbigny, Wagner 1908, in Martini & Chemnitz, Conch. -Cab., (2)1: sect. 18, pt. 2, p. 94, pi. 17, figs. "1-4, pi. 18, fig. 13. Emoda sagraiana percrassa Aguayo & Jaume 1954, Torreia, 21: 12, pi. 1, figs. 1-3 "(Luis Lazo, Provincia de Pinar del Rio; holotype, MP 17341). Description. Shell 17 to 29 mm in diam- eter, depressed to moderately raised, gen- erally smooth. Whorls about 5^4, veiy slightly rounded, body whorl depressed, rather narrowly rounded at the periphery. Color generally brownish, from pale yel- lowish buff to bright orange-brown; oc- casional specimens with an olivaceous tinge, others with a lighter peripheral band, spire and callus frequently reddish. Spire moderately raised, rounded, domelike, the same color as the rest of the shell, occasion- ally reddish, rarely darker, sometimes lighter. Aperture widely semilunate, some- what extended laterally and more or less regularly rounded in peripheral region, white near lip, brownish in the interior. Lip thickened, occasionally very much so, slightly flaring, rarely reflected above. Basal callus slightly raised, white in columellar region, transparent near outer margin, as extensive as, but more rounded than, the aperture. Columella short, oblique, white, concave below; basal notch usually deep, acute or obtuse; basal tooth low, rounded. Suture moderately impressed. Axial sculp- ture of very fine, irregular, diagonal growth lines. Spiral sculpture wanting. Proto- conch l^/i \\'horls, sublustrous, slightly raised, faintly marked by curved, retractive, axial wrinkles. Periostracum thin but strong, light or dark brown or olivaceous, in some specimens bro\\'n tinged with green. Oper- culum as in genus, large, strong, calcareous lamina white, or tinged with faint reddish bro\\'n, generally on one plane but curved sharply and shortly inwardly below; in- ternal corneous lamella thin, brownish, somewhat darker at the margins. La Claraboya, Luis Lazo Ensenada de los Biuros, Cabezas Caiguanabo, Consolacion del Norte Bejarano, Canalete Galalon, San Andres Catalina, San Diego de los Bairos Sierra Paso Real, Guane Cafetal "La Villa," Candelaria Galalon, San Andres Pan de Azucar, Viiiales La Muralla, Guane Remarks. A highly polymorphic species, E. .sagriana varies from a small brownish form with an acute basal notch and rela- tively thin peristome to large greenish forms with a red spire and/or callus and an exceedingly heavy lip. In between are Height Diameter mm mm 21.5 29.0 17.5 28.3 17.5 26.0 17.5 25.5 16.0 25.5 16.0 24.5 15.5 23.5 15.5 23.3 15.0 23.5 14.5 24.0 13.5 19.5 106 BiiUctin Museum of Comparative Zoology, Vol 141, No. 3 found all kinds of inter2;radcs scattered thronghout the range. The smallest forms come from around Guane near the western terminus of the Sierra de los Organos, but almost similar forms are found near Cata- lina, toward the eastern end. The hea\'y, thicklipped form, to which Aguayo & Jaume gave the subspecific name of pcrcrassa, occurs generally in the region of Luis Lazo, Sumidero, and Cabezas, in the west central portion of the Sierra de los Organos; speci- mens which cannot be separated from those found here also occur, together with the smaller, more typical specimens toward the east at San Andres, Consolacion del Norte, and San Diego de los Baiios. The popu- lations at Ensenada de los Burros in Ca- bezas and at Finca "La Giiira" near Luis Lazo can be readily divided into the small brownish form with the acute notch and thin lip, and the large, heavy lipped form with the obtuse notch and heavy peristome. The complete lack of intergrades in these localities argues strongly that here are two distinct sympatric groups that readily deserve specific distinction. These clear differences, however, fade away in the populations from the eastern end of the range around the Sierra San Andres and San Diego de los Baiios, where many arbi- trary decisions are needed to place the majority of the specimens. Aguayo & Jaume (1954: 4) well de- scribed the confusion that seems to exist: "We are confronted by the following dilemma: either we name all the \'ariants which occupy more or less restricted areas ( 'microspeeies' or incipient subspecies), thus increasing the confusion of the present nomenclature, or we put all the diverse forms under a single specific name, thereby creating the error of postulating a uni- formity that is far from the truth" (trans- lated). They go on to suggest that only the two extreme forms be given names (sograiana and percmsso) and that the intermediar\^ colonies be designated by the "conventional" formula: Emoda sagraiana Orbigny tr. percrassa. This is obviously no solution, since the determination of just where extreme forms end and the inter- mediaiy forms begin will have to be largely subjective. Moreover, we agree with Mayr (1969: 46) who wrote, "Now that it is being realized that every local population is different from eveiy other one, even if they li\'e only a few miles apart or less .... there is no longer any excuse for a formal recognition of innumerable local subdivi- sions of subspecies." On the basis of the data and the large amount of material available to us at present, we can only conclude that we are dealing with a highly complex and variable species whose com- ponent parts cannot be completely under- stood without intensi\e biometric and ecological study. For the present, we feel that least confusion will result if all the Emoda-likc populations in the Sierra de los Organos mountain complex are considered sagraiana. This interpretation should prove as utilitarian as calling all the polymorphic populations of the littoral marine species Thais lapillus in the Western Atlantic or T. JamcUosiis in the Eastern Pacific. Specirnens examined. Pinar del Rio. Guane: Sierra de Guane; "Sierra Mendoza"; Sierra Paso Real; La Muralla; Punta de la Sierra. Luis Lazo: Sierra San Carlos; La Caoba; Sierra Los Acostas; La Cetricina; La Claraboya; La Giiira; El Potrerito; La Estrechura; El Junco; Ensenada del Bar- rio, Pica Rica, and Mogote La Giiira, Sumidero. Cabezas: Valle Isabel Maria; Ensenada de los Burros; Sierra La Maje- quilla, Petiablanca; Ensenada Chica; Valle de Cabezas; Mogote Cementerio; Sierra del Quemado. Vinales: Potrero de Manuel Sanchez near Capon; Arroyo Melindre; Kilometer 14; Mogote de la Vega No. 3; Mogote Capon; Sierra Penitencia; Sitio del Infienio; Sierra Derrumbada; Sierra Cela- das; Sierra Celadas on the road to La Mina Matahambre; Mina Matahambre; Hoyo de los Cimarrones, Sierra del Infienio; El Cuajani; Los Peladores at summit of Sierra del Infierno; Sierra del Infierno; Sierra el Martillo; El Queque; Ensenada del Valle, Emoda and Glyptemoda in Cuba • Clench and Jacobson 107 El Queqiie; Sierra de Galeras; Mogote Pequeilo; Mogote Rojas; Mogote Dos Hermanos; Puerta del Ancon; Lorenzo Lopez, 1 km N of Vinales; Hoyo Jutia; El Cejanal; Hoyo de Fania, Palmarito; Ense- nada de Martin Miranda, Palmarito; Hovo de Lorenzo Martinez, Palmarito; Las Deli- cias; Cueva de los Santos, Palmarito; Sierra de Palmarito; Mogote Quillo; El Punto, Pan de Azucar; Pan de Azucar; Hoyo Largo, San Antonio; Hoyo de los Muertos; Mogote Jose Maria Garcia. Smi Vicente: La Chorrera; Mogote la Mina Constancia; Mogote Jutia, 1 km E of Baiios de San Vicente; Bafios de San Vicente; Ensenada de los Baiios; Mogote Pequeiio, Costanera de San Vicente; Hoyo del Ruisenor, Ancon; Cueva del Rio, Baiios de San Vicente; Costanera de San Vicente; Mogote de la Resbalosa; Mogote del Marmol and Laguna de Piedras; Sierra del Abra; Punta de la Costanera de San Vicente; Ensenada de San Vicente; Las Cuevitas, Ancon; Puerta del Ancon. Consolacion del Norte: Monte Largo, Pico Chico; Mogote Cascajal near Pico Chico; Mogote Grande SW of Pico Chico; Puertecitos de Caiguanabo; Los Portales, Caiguanabo; Farallon de las Avis- pas; Pico Grande; Bejarano, Canalete; Abra de Bejarano; "Sierra Canalete"; Sierra San Andres; La Sierra; Sitio de la Sierra; Mogote de la Cidra; Mogote de la Palma; Mogote el Indio; Ensenada de la Ayua; Pasada de la Ayua; La Jagua; La Jagua de Azquiz; Sierra Gloria, N of Colniillo de la Vieja; Colniillo de la Vieja; Los Caracoles; Abra de la Colmena; Colmena de Piedra, Galalon; Sierra Grande, Gua- camaya. San Diego de los Bafios: Mogote de la Finca; Mogote el Bosque; La Catalina, N of San Diego; Cayito, 3 mi. from La Catalina; Hoyada de la Catalina; Mogote Colorado; Pinalito; Mogote de los Indios; Los Cayitos, Catalina, N of San Diego de los Baiios; Cueva del Indio. Guajaihon: San Jose de Sagua; SW of Pan de Guajaiboii; San Juan de Sagua; Sierra Chica; Ranclio Lucas, W of Guajaiboii; Sierra de Juan de Guacamaya, Baliia Honda; Hato Sagua; Caimito; La Funiia; Sierra la Giiira; El Toro, Sierra Limones. Cabanas: El Guabinacho; 2 mi. N of Santa Cruz; Las Animas, Rangel; El Retiro, Rangel; El Maiiiey, Callajabos (Caya- jabos); Loma de Gloria; Rancho Mundito, Rangel; Subida al Rangel; El Taco. Candelaria: Escuela de Frias; Cafetal "La Villa." Emoda pulcherrima pulcherrima (Lea) Plate 1, figures 14, 15; Plate 2, figures 13, 14; Plate 3, figures 6, 7; Plate 5. Helicina pulcherrima Lea 1834, Trans. Ameri- can Phil. Soc. (NS), 5: 49, pi. 19, fig. 57 (Java?); 1834, Observations on the Genus Unio, Phila- delphia, 1: 161, pi. 19, fig. 57 ("Java?"; lecto- type, probably figured specimen, here selected, USNM 104613; paralectotypes, USNM 104613a, 104609, 104595, all from Lea collection; type- locahty, here selected, Guantanamo). Helicina crassa Orbigny 1842, Molliisques, in Sagra, Histoire Physique, Politique et Naturelle de rile de Cuba, 1:'243, pi. 19, fig. 5 [not fig. 6] (interieur de I'lle de Cuba; type, BM(NH) 1854.10.4.167). Helicina pulcherrima Lea, Pfeiffer 1850, in Martini & Chemnitz, Conch. -Cab., (2) 1: sect. 18,. pt. 1, p. 17, pi. 1, figs. 5-7, pi. 6, figs. 5, 6 (die Insel Cuba). Helicina ruhrocincta Poey 1854, Memorias His- toria Natural Isla de Cuba, 1: 417, pi. 33, figs. 16-19. ( "Probablemente en el departmento ori- ental"; type, probably MP.) Description. Shell about 18 mm in diam- eter, depressed globose, strong. Whorls 4y2, flat, body whorl descending shortly near the aperture. Color varied, light green with darker green, irregular, faintly sigmoid, axial bands, or light to dark reddish brown; spire usually lighter. Occasional specimens have a narrow, light-colored peripheral band. Spire moderately raised, rounded. Aperture widely semilunate, white inside; peristome thickened, moderately reflected, widest at center, gradually narrowing at extremities, with a small, elongate, rounded, toothlike protuberance near the basal notch. Columella short, white, barely con- vex, more strongly so below the mid-point. Parietal callus strong, as wide as the aper- ture, white in the columellar region, darker 108 Bulletin Museum of Comparative Zoology, Vol. 141, No. 3 Height Diameter mill mm 15.1 20.6 15.5 21.5 13.5 19 13.0 19 11 16.5 near thc> outer, rounded edge; in oecasional speeimens the entire callus is white, a pos- sible gerontic sign. Body whorl sculptured with strong, regular, rounded axial cords, wider than the inter\'als, weaker and less n^gular on the base and the spire. Proto- conch IV2 whorls, moderately raised, with irregular, curved axial wrinkles. Periostra- cum strong, shining, usually wanting on earh' whorls. Operculum with white or light reddisli calcareous lamina, and brown or rcxldish corneous one. Lectotype ( USNM ) Mayari, Oriente Monte Turquino, Oriente Los Arroyos, Miranda, Oriente Loma del Gato, Guantanamo, Oriente Remarks. This subspecies is confined to the entire southern portion of Oriente Province. It is characterized by its gener- ally small size, green or brown color, and especially by the strong, regular axial sculpture on the body whorl. Even brown specimens, which superficially resemble suhmar^pe-locality ) ; Sitio Pcrdido; Loma Camoa; Mantilla; ^Forcart (1950: 71) identified E. suhmarginata from Pinar del Rio between the city of Pinar del Rio and Viiiales. We have not seen the specimens lint suspect that the identification is in error. Madruga; El Ingles; La Canabrava; Finca "La Belianza," Cotorro; Pefia Blanca, Sierra Anafe; Jaruco. Matanzas. Elena, El Palenque; Cotorra; Dos Cecilias, Coliseo; Jaguey Grande; Abra de Figueroa, Valle de Yumuri; Cueva el Negro, San Miguel; Canasi Rd., near El Pacito; Loma del Fine, Cardenas; Bridge over R. Canimar, Matan- zas-Limonar Rd.; 20 mi. SE of Matanzas; Ceiba Mocha; Vista Alegra (USNM). Las Villas. Soledad: Mogote el Palmar; Bo- tanical Garden; Guabairo; Loma de la Jutia, Vilche's Potrero; Sagua la Grande; Mina Carlota, about 12 mi. E of Soledad; Laza del Medio; El Mamey; San Lorenzo; Jumagua; Mogote W of La Jumagua. Trinidad: Northwest of La Vigia; Sierra de Santa Rosa. Yaguajaij: Yaguajay-Remedios Rd.; Cerros de Yaguajay; Guainabo; El Tigre; Punta Judas; Sierra Judas de la Cumagua. Remedios: El Mamey; Los Hatil- los; La Culebra; Loma Caraballo; Dos Sierras; Buenavista; La Puntilla (USNM); Cerro Guajabana. Caiharien: Finca "Cueta"; Loma de Petrera; Punta Caguanes, Buena- vista Bay; Santa Rosa; Cayo Conuco; Cayo Lucas; San Joaquin, Esperanza; El Boque- ron, Jatibonico; San Agustin, Zulueta; Piedras de Amolar, Escanibray; Villa Ciba, Carretera de Camajuani; Chicharron, Sierra Vega Alta; La Vigia, Mayajigua; Sierra de Canoa, Los Llanados; Monte Cagiieiras, Sancti Spiritus; Sierra 3 km S of Dolores. Camaguey. Punta Alegre. Oriente. Camino de la Loma de la Bandera, Pinar de Mavari; Mayari; between Sagua de Tafiamo and Cananova. Emoda ciliato (Poey) Plate 1, figures 7-9; Plate 4; figure L Hcliciua ciliaia Poey 1851, Memorias Historia Natural Isla de Cuba, 1: 109, pi. 11, figs. 5-8 (Trinidad [Las Villas]; type, probalily MP). Helicina fossulata Poey 1857, Memorias Historia Natural Isla de Cuba, 2: 25 (Trinidad, [Las Villas]; syntypes, probably MP; paratype, MCZ 263901 from the Andiony Collection collected by Dr. Gutierrez, San Juan de Letran, Santa Clara [ = Las Villas]). Alcadia {Emoda) ciliata Poey, Wagner, 1908, in Emoda and Glyptemoda in Cuba • Clench and Jocobson 111 Martini & Chemnitz, Conch.-Cab., (2) 1: sect. 18, pt. 2, p. 102, pi. 8, figs. 5-6; pi. 19, figs. 14-17 (Trinidad und Banao auf Cuba). Description. Shell reaching 27 mm in diameter, strong, widely turbinate, smooth but lusterless, rounded at periphery. Whorls 5V2, moderately convex, body whorl rounded at base, barely descending near aperture. Color predominantly light yel- low, occasional specimens faintly tinged with brown; white when decorticated, with an indistinct, whitish subsutural band con- tinued subperipherally on the body whorl. Spire depressed, dome shaped. Aperture widely semilunate, yellowish within. Palatal lip white, strongly thickened, moderately flaring. Basal denticle small and obtuse. Parietal callus smooth, thick, white in the columellar region, transparent near the rounded outer margin, more rounded than the aperture. Columella short, oblique, very weakly sigmoid, angled above, slightly rounded at insertion with basal lip. Su- ture moderately impressed. Shell smooth, marked by very fine, curved, diagonal growth lines. Protoconch VA whorls, barely raised, moderately lustrous, approximately same color as rest of shell, marked by fine, curved, axial wrinkles. Periostracum strong, closely marked by irregular spiral cords of raised periostracal material, not reflected in the shell surface beneath. Operculum as in genus, light to dark reddish brown near the outer margin. Height Diameter mm mm 19.0 27.0 Banao, Las Villas 19.0 25.5 Ciegos de Ponciano, Las Villas 15.5 19.5 San Juan de Letnin, Las Villas; paraty^e of jossulata. Remarks. This relatively large-shelled species is readily distinguishable from the others of more or less equal size by its uniformly yellow color and depressed shape. It differs from silacea, the other spe- cies in which yellow predominates, by its smooth, almost unsculptured surface. It apparently has a limited range in the moun- tains between Trinidad and Sancti Spiritus in the southeastern part of Las Villas Province. Helicina fossidata Poey is only a color form with a peripheral band. Poey wrote of ciliota, "Color es ya de un amarillo bajo, ya de un moreno rojizo," and of jossulata, " . . . . sordide rubra . . . cingulis duobus flavidis ornata, quorum alter ad suturam, alter ad peripheriam." The type locality of both forms is "Trinidad" and the only difference seems to be the presence of a spiral peripheral band in jossulata. Else- where we show that banded and unhanded forms frequently occur in single popu- lations. Specimens examined. Las Villas. Banao, Sancti Spiritus; Ciegos de Ponciano; Camino de Banao, Sancti Spiritus; San Juan de Letran. Emoda ciliata guisana (A. J. Wagner) Plate 1, figures 10, 11. Alcadia (Emoda) ciliata p.uisana A. J. Wagner 1908, in Martini & Chemnitz, Conch.-Cab., (2) 1: sect. 18, pt. 2, p. 102, pi. 18, figs. 7-9 (Guisa auf Cuba; types in Warsaw Zoological Museum 8353). Description. Shell like that of ciliata, differing from the nominate form by the presence of shallowly incised, parallel, spiral lines on the shell, as in suhmarttlcMnent — in Las Villas, or Wagner's locality is in error. The mili- tar\- map of Cuba cites no Guisa from Las Villas. We are keeping this name pro\'ision- ally since the material at hand is not suffi- cient to indicate whether the incised spiral lines are characteristic of individuals or of an entire colony. Specimens examined. "Cuba"; Guisa. Emoda emoda (Pfeiffer) Plate 2, figures 6, 7; Plate 5. Hclicina emoda Pfeiffer 1865, Novit. Conch., 2: 253, pi. 64, fijis. 6-8 (parte oriental! ins. Ciibae; [Monte Toro, based upon Pfeiffer 1860: 77]; type destroyed ) . AJcadia (Emoda) sagraiana emoda Pfeiffer, Wagner 1908, in Martini & Chemnitz, Conch. - Cab., (2) 1: sect. 18, pt. 2, p. 96, pi. 17, figs. 7-9. Description. Shell reaching 31 mm in diameter, moderately strong and relatively smooth. Whorls 5V2, moderately rounded, body xv'horl depressed, descending shaiply at aperture. Color of body whorl bright green, reddish brown when decorticated; spire reddish, peripheiy banded by a \\'hitish line margined above and below by dull, narrow, reddish bands. Spire rela- tively high and dome-shaped. Aperture wideK' semilunate, white inside; peristome thickened, moderately flaring, not reflected, basal tooth generally weak. Parietal callus smooth, rounded, larger than the aperture, white near the columella and transparent near the outer margin. Columella short, white, slightly sigmoid, the wider cun'e below, and briefly rounded near the basal tooth. Suture moderately impressed, rough- ened by the terminations of the growth lines. Sculpture of low, irregular, mod- erately strong growth lines crossed by im- pressed, closely spaced spiral furrows, most closely approximated at the base and weaker on upper whorls. Protoconch VA whorls, reddish, smooth, faintly marked by irregular, curved, diagonal axial riblets. Periostracum strong, green on body vvhorl, weaker and reddish on spire. Operculum as in the genus. Height Diameter mm mm 23.5 31.0 Manatial, Santa Fe, Oriente 22.0 30.5 Cuaso, Guantanamo, Oriente 18.0 28.0 El Codrito, Monte Libano, Oriente Remarks. This is a well-marked species from near Guantanamo in Oriente Province. Pfeiffer thought it might be related to hriarea from Las Villas Province because of its large size, but emoda has a more elevated shape, the shell is not so solid, the color differs markedly and the spiral sculp- ture is completely wanting in hriarea. Wagner (1907: 96) confused it with the larger sap'aiana from Luis Lazo in Pinar del Rio Province, but that shell is some- what smaller, more depressed, heavier, has a far thicker, more reflected lip, and lacks the spiral sculpture. Pfeiffer's figures are of decorticated, weathered specimens but they show (especially his fig. S) the spiral sculpture quite clearly. This species is remarkably constant in its color but, like many Emoda, it varies in size. The popu- lations from Monte Libano consist of smaller shells which, however, clearly be- long to emoda because of the elevated shape, the green and reddish color, and the presence of spiral lines. Although the type- specimens have been destroyed, there is no doubt regarding the identity of this species. Pfeiffer's figures are unusually good and we have examined several lots from near the type-locality. Specimens examined. Oriente. La Subida de la Hembrita; Guantanamo; El Codrito; Monte Libano; Montaiia de Guantanamo; Manatial de Santa Fe; 1 mi. NW of Guaso; N of Guantanamo City. Emoda caledoniensis, new species Plate 2, figures 11, 12; Plate 5. Ilolotype, MCZ 261352, from Mina Cale- donia, Mayari, Oriente. L. R. Rivas, col- lector. Emoda and Glyptemoda in Cuba • Clench and Jacobson 113 Height Diameter mm mm 21.5 30.5 21.0 28.0 20.0 29.0 19.5 2S.0 17.0 24.0 Para types, MCZ 261350; USNM 463790, both from the same locaHty as the holotype; MCZ 261348, from Brazo del Pino, Wood- fred, Mayan', Oriente; MCZ 261351 from Sierra del Cristal, Mayan', Oriente. Brazo del Pino, Paratype Brazo del Pino, Paratype Sierra del Cristal, Paratype Mina Caledonia, Holotype Mayari, 1 km from the pine forest, Paratype Description. Shell reaching 30.5 mm in diameter, inflated, smooth and moderately solid. Whorls 5V2, moderately rounded, body whorl inflated, weakly shouldered obliquely below the suture, rounded at the periphery, shortly descending near the aperture, where it blends into the basal callus. Color olivaceous, with irregular darker axial streaks, light brick red in decorticated specimens. Spire moderately raised, rounded, darker than rest of shell. Aperture semilunate, lip moderately thick- ened, flaring, barely reflected above, widest near the center, narrowing at each ex- tremity; basal tooth small, rounded, basal notch barely perceptible. Columella short, white, slightly sigmoid and rounded below. Basal callus smooth, larger than the aperture, white in columellar region, trans- parent at outer margin. Axial sculpture of very fine, diagonal growth lines, spiral sculpture wanting. Suture moderately im- pressed, somewhat roughened by the ter- minations of the growth lines. Protoconch IV2 whorls, faintly marked by cui-ved, axial wrinkles. Periostracum strong, olivaceous or green. Operculum as in genus, calcare- ous layer white, inner corneous layer reddish, darker at margins. Remarks. The four lots in MCZ upon which this description is based all come from the region around Mayari in Oriente Province. This new species belongs to the group of large, predominantly greenish or olivaceous shells with a reddish spire from the eastern part of the country. It differs from pulcherrima titanica from Baracoa by its higher spire and in the absence of the strong, regular, axial sculpture; it is gener- ally smaller and more rounded than E. emoda from Guantanamo and lacks the spiral sculpture of closely set incised lines. Emoda bayamensis (Poey) Plate 2, figures 4, 5; Plate 5. Helicina bayamensis Poey 1854, Memorias His- toria Natural Isla de Cuba, 1: 416, pi. 33, figs. 7-10 [not 8-11] (Bayamo [Oriente]; lectotype, here selected, MCZ 73855, Buena Vista, Bayamo, Oriente, Anthony Collection, from Poey; para- lectotype MCZ 256496, same locality. The shells in MCZ from which the type selections were made were identified as type material by Torre in June 1912). Helicitm hastklana Poey 1854, Memorias His- toria Natural Isla de Cuba, 1: 415, pi. 33, figs. 11, 12 (Puerto Principe [ = Camagiiey]; type, probably MP). Helicina bayamensis Poey, Pfeiffer 1862, Novit. Conch., 2: 199, pi. 53, figs. 1-5. Helicina jeannereti Pfeiffer 1862, Malak. Bliit., 9: 6 (Mayari [Oriente]; type destroyed). Helicina bayhamensis "Poey" Reeve 1873, Conch. Icon., Helicina, 19r pi. 2, fig. 11 (Cuba) [error for bayamensis]. Helicina bastidiana "Poey" Reeve 1873, Conch. Icon., Helicina, 19: pi. 8, fig. 61 (Cuba) [error for bastidana]. Description. Shell reaching 23 mm in diameter, depressed globose and solid. Whorls 4^4, weakly convex, body whorl moderately inflated, depressed but not carinate at the peripheiy. Color of body whorl olivaceous and sometimes tinged with brown, with narrow, irregular, dark green axial stripes; spire much lighter; occasional specimens with a yellowish peripheral band generally margined with red. Spire moderately raised, rounded, lighter in color than the rest of the shell. Aperture semilunate, irregularly rounded, inside margin of upper lip diagonal, almost straight. Lip white, well expanded, but not reflected, unevenly rounded, somewhat ex- tended just above the periphery, widest centrally, narrowing rapidly at the extremi- ties; basal tubercle small. Columella short, white and weakly sigmoid. Parietal callus 114 Bulletin Museum of Comparative Zoology, Vol. 141, No. 3 smooth, thin, \\'hite in the coluniellar region, transpaic>nt near outer margin. Suture moderately impressed. Sculpture of irregular, curved, axial cords, \\'eaker and less regular than in pulcherrima, weaker on the base, almost obsolete on earlier ^^'horls. Operculum and periostracum as in the genus. Height Diameter mm mm 15.0 2.3.0 BiKMia Vista, Bayamo, Oriente, Lcctotype 14.0 21.5 Paralectotype Remarks. Poey described hastidana and hayamensis at the same time, but in 1S65 (p. 144) he selected lianamensis as "el tipo comun" and wrote that hastidana "es una \^ariedad, y muy rara." It seems to be noth- ing but a color variant. This species differs from the larger Emoda by the irregular rather than rounded inner edge of the periostome. There is a distinct rounded angle above and below the periphciy; in other Emoda this entire area is evenly rounded. In fully adult shells there is also an extension of the central portion of the outer lip, so that the shell has a somewhat lengthened rather than a rounded appearance. This feature is illustrated by Poey (1854, pi. 33, figs. 11, 12) and is present in the lcctotype. Sub- mature shells generally lack this character- istic. The species differs from pulcherrima titanica in its smaller size and weaker axial costae. It differs from ciliata in its lively colors and comparatively stronger axial sculpture, and from sagraiami from the other end of the island, in its far weaker basal notch and tooth. We could discover no feature to distinguish it from jeannereti. Pfeiffer, in his discussion of the latter, noted several differences between jean- nereti and sagraiana but failed to compare his species with hayamensis. Wagner (1907: 99) cited this as a subspecies of pulclierrima, but it lacks the strong axial sculpture on the body whorl that is char- acteristic of pulcherrima and, at the present state of our knowledge, is probably best regarded as a species. We have not succeeded in clearly identi- fying jeannereti Pfeiffer. The species has never been figured satisfactorily. Reeve (1874: pi. 1, fig. 3a) had a poor figure of a shell with a bright green color which he referred doubtfully to jeannereti. This figure seems to have been copied from Sowerby (1847: pi. 3, fig. 126), where it appears as Helicina sagra [sic]. Sowerby (1866, pi. 1 [266], fig. 5) figured a typical brown sagraiana and referred jeannereti doubtfully to it. The green form (fig. 6) he now called hastidiana (sic) [ = haya- mensis]. Arango (1879: 46) copied Pfeif- fer's data but cited Wright as the collector instead of Jeanneret. Wagner ( 1907-1908 ) omitted the name from his monograph and Fulton (1915) failed to note this omission. The only specimens available for our in- spection come from Palma Soriano (Ber- mudez collection), about 60 km southwest of Mayari, Pfeiffer's type-locality of the species. We do not know who made the determination. In his discussion, as we have noted, Pfeiffer did not compare his species with hayamensis. Hence our plac- ing jeannereti in the synonymy of haya- mensis is provisional. Specimens examined. Oriente. La Can- tera de Miranda; Buena Vista, Bayamo (type-locality); Guisa, Bayamo; Miranda, Palma Soriano. Emoda bermudezi Aguayo & Jaume Plate 1, figures 5, 6; Plate 4, figure 1. Emoda hcrmudczi Agiia\o & Jiiinne 1954, Tor- reia. No. 21, p. 10, pi. l, figs! 4, 6 ("Paso de las Trincheras," Sierra de Culiitas, Provincia de Camagiiey; holotype, MP 17346). Description. Shell reaching 19.5 mm in diameter, rather solid, depressed conic, periphery moderately carinate. Whorls AVz, flat, body whorl descending rapidly near aperture, where it blends into the thick- ened, slightly raised basal callus. Color generally olivaceous at base, tinged with Emoda and Glyptemoda in Cuba • Clencli and Jacobson 115 brownish orange above, occasional speci- mens with a narrow, yellowish peripheral band. Spire low, rounded, same color as rest of shell. Aperture widely semilunate; peristome white, thickened, very weakly expanded. There is a distinct notch above near the parietal margin from which a short, raised marginal lamella extends to the basal callus. The basal callus as wide as the aperture, porcellanous, rounded, slightly raised in mature specimens. Colu- mella weakly sigmoid, the basal tooth c^uite weak. Suture very slightly impressed. Later whorls marked by irregular diagonal lines of growth; surface with concentric, spiral rows of impressed pits. Protoconch IV2 whorls, slightly raised, weakly marked by diagonal and regular axial sculpture. Periostracum thin; operculum as in genus, inner corneous layer light brownish, slightly darker at the margins. Fleiijht Diameter mm mm 14.5 19.5 San Francisco, near La Tinaja, Camagiiey 13.0 17.5 Santa Rita, Camagiiey 12.5 19.5 Paso de las Trincheras, Sierra de Cubitas, Camagiiey 12.5 16.5 El Cercado de Cubitas, Camagiiey Remarks. This species is very close in its relationships to the widely distributed suhmarginata, which it resembles in color and in the presence of concentric, though weak, spiral lines that consist of a series of pits. It is easily distinguished by the sharp, triangular notch near the posterior (upper) angle of the peristome, a characteristic which, as Aguayo & Jaume pointed out (1954: 10), is found in no other Emoda. The color as given above, as well as the description of the operculum, is from a lot of fresh shells from the Bermudez collec- tion labelled merely "Camagiiey." The other and better localized shells that were available were dead shells, the color hav- ing been somewhat faded or stained by reddish earth. In all specimens the concen- tric spiral lines were quite weak. The species apparently is limited to the Sierra de Cubitas in the northeastern part of Camagiiey Province. This species, as well as najazaensis, was misidentified as hasti- dana Poey and appeared in the Museo Poey and other museum collections under this name. Aguayo & Jaume (1954: 11) pointed out that hastidana is a mere color form of bayamensis, an observation made originally by Poey himself (1S65: 144). Specimens exam,ined. Camaguey. Las Cuevas; Santa Rita v la Entrada del Camino de La Guanaja [Santa Rita and the entrance to La Guanaja Highway]; Banao; Paso de los Burros; Paso de los Trincheras y Cueva de los Indios; San Francisco, N of La Tinaja; East side of Vereda de Burro, Finca San Clemente (USNM); Paso del Este (USNM); El Circulo (USNM); Los Corrales de Cangilones (USNM); North entrance to Paso Escalcra (USNM). All of the above localities are in the Sierra de Cubitas. Emoda najazaensis Aguayo & Jaume Plate 1, figures 3, 4; Plate 4, figure 1. Emoda najazaensis Aguayo & Jaume 19.54, Tor- reia, No. 21, p. 9, pi. l', figs. 7, 9, 10 ("El Cacaotal," Sierra de Najaza, Provincia de Cama- giiey; holotype, MP 16222: paratypes, MCZ 257558, from "Vereda del Telegrafo," Sierra del Chorillo, ex MP). Description. Shell reaching 22 mm in diameter, smooth, depressed conic, periph- eiy roundly carinate. Whorls 5V4, very weakly rounded, body \\'horl descending shaiply at the aperture, where it blends into the basal callus. Color varied, pale olivaceous, pale brown, buff, or faintly orange, a pale narrow band at the periph- ery. Spire low, mammiform and somewhat darker than rest of shell. Aperture widely semilunate, lip slightly thickened and barely reflected. Columella shallowly con- cave below, weakly convex above; basal callus weakly lustrous, white, as wide as the aperture, rounded at outer margin. Basal tooth quite small. Suture moderately im- pressed. Spiral sculpture of closely set, concentric cords, which are raised in the 116 Bulletin Muscudi of Comparative Zoology, Vol. 141, No. 3 pcriostracum, and rcflecttxl below as a line of minute impressed pits in the shell, most prominent on the body whorl, weaker at the base. Protoeoneh IV2 whorls, raised nipplelike abo\'e the sneeeeding whorls, scnlptur(>d with faint, rc\gular axial striae, almost the same eolor as the rest of the shell. Periostraeum thin. Opereulum closely resembles that of E. bernmdezi, but the in- ternal corneous layer is not darker at the margin. Heijiht Diameter mm mm 15.5 20.0 Vereda del Telegrafo, Sierra del Chorrillo, Camagiiey 14.5 20.0 El Cacaotal de Sifonte, Sierra Najaza, Camagiiey 13.0 19.0 El Palomar de San Jose, Cama- giiey Remarks. This species is similar to .stil)- iiiar<:,in(ita, especially in color and in the presence of spiral concentric incised striae, though in najazaensis these are more closely spaced. However, it has a generally larger shell, which is more depressed and more strongly keeled. It appears to be confined to the area around the Sierra Najaza (Najasa) in the southeastern part of Camagiiey Province. The specimens we were able to examine, including several lots from the type-locality, did not show the colors mentioned by Aguayo & Jaume (1954: 9). Specimens examined. Camaguey. Najaza; El Palomar de San Jose; El Cacaotal, Najaza; Sierra Guaicanamar; Sitio Afuera; Vereda del Telegrafo, Sierra del Chorrillo; Sierra del Cochimbo; El Cacaotal de Si- fonte, Sierra de Najaza. Emoda briarea (Poey) Plate 2, figures 1-3; Plate 4, figure 1. Helicina hriarca Poey 1851, Memorias Historia Natural Isla de Cuba, 1: 108, pi. 11, figs. 9-12 (San Diego de los Baiios [sic]; lectotype, here selected, MCZ 73854, from the J. G. Anthony Collection, collected by Gundlach [Trinidad Mountains, Santa Clara ( = Las Villas)]. The specimen appears originally in the MCZ collection as a syntype, identified as such by Torre in June 1912). Helicina briarea Poey, Pfeiffer 1862, Novit. Conch., 2: 195, pi. 52, figs. 1-5 (Trinidad). Helicina hriaraea "Poey," Sowerby 1866, Thes. Conch., 3: 278, pi. 1, figs. 1-3, [error for briarea]. Description. Shell very large for the family, reaching 34.5 mm in diameter, depressed turbinate, solid, roundly carinate, sharply so in juvenile specimens and rel- atively smooth. Whorls about 5, flat, body whorl moderately inflated, flattened ba- sally, descending shortly near the aperture. Color puqjlish brown on the body whorl and somewhat lighter on the spire. There is a variously wide, yellowish or whitish subsutural band that is extended as a supra- peripheral band on the last whorl, where it is separated from the light colored basal area by a moderately narrow, puiplish red band. Subsutural band wanting on early postnuclear whorls. Spire depressed, weakly rounded except for the slightly raised protoeoneh. Aperture widely semilunate, evenly rounded, white, ground color and spiral band showing through inside. Peri- stome expanded, barely flaring above, widest just above the periphery and slightly thickened. Basal denticle quite small, ob- tuse, somewhat lengthened. Parietal callus smooth, or yellowish white, more rounded than the aperture. Columella obli(iue, white, short, concave below. Suture weakly impressed. Sculpture of very fine, irregular growth lines, without spiral lines. Proto- eoneh IV4 whorls, somewhat raised, light brown, with faint, curved axial wrinkles. Periostraeum thin, yellowish, not carrying the shell color, with faint, well-spaced spiral ridges. Ojoerculum reddish, darker at margins, relatively thin. Height Diameter mm mm 22.5 34.5 Puriales, Las Villas 22.0 31.5 Puriales, Las Villas 18.0 25.5 Caracusey, Las Villas 16.0 22.0 San Bias, Las Villas Remarks. Most specimens of this species can be readily recognized by their large size. In pojiulations where the specimens are smaller (San Bias, Ciegos de Ponciano, etc.), they are determined without diffi- Emoda and Glyptemoda in Cuba • Clench and Jacobson 117 culty by their smooth surface, depressed shape, characteristic color, and by the sub- sutural band, which has the same color as the base. The species differs from the larger specimens of sag,raiana from Luis Lazo in Pinar del Rio Province in its color and in the far less strongly thickened and unreflected lip. It differs from E. emoda in its more depressed shape, stronger shell, and color. The species seems to be con- fined to the area around Trinidad in the south central part of Las Villas Province. As Pfeiffer noted (1862: 196), Poey's lo- cality was in error. We here select Trini- dad, Las Villas Province, Cuba, as the type-locality. Poey selected the trivial name referring to the mythological giant Briareus, because of the large size of the shells. Specimens examined. Las Villas. Trini- dad: Fomento; Finca Pitajones, Caracusey; Santa Rosa, Cafetal Los Puriales; Rio Ca- ballero; Portillo, Rio de Ay; Rio Cabunii; Ciegos de Ponciano; San Bias Dam above San Bias. Emoda mayarina mayarina (Poey) Plate 2, figure 17; Plate 5. Helicina mayarina Poey 1854, Memorias Historia Natural Isla de Cuba, 1: 417, pi. 34, figs. 6-8. (Mayari [Oriente]; holotype, MP 17034). Helicina mayarina Poey, Pfeiffer 1856, Malak. BUit., 3: 144; Pfeiffer 1858, Monographia Pneu- monopomorum Viventium, Suppl. 1, p. 184; Pfeiffer 1862, Novit. Conch., 2: 200, pi. 53, figs. 6, 7. Emoda mayarina (Poey), Aguayo & Jaume 1954, Torreia, No. 21, p. 7, pi. 2, figs. 1, 6. Description. Shell reaching 16 mm in diameter, subglobose, solid, moderately strong, sublustrous. Whorls 4, convex, the last somewhat descending at the aperture below the periphery. Color brownish red with several faint reddish wrinkles, base somewhat yellowish, with a darker periph- eral band margined above by a lighter band of approximately the same width. Spire moderately raised, submammiform. Aperture slightly oblique, subsemilunate, white at the lip, darker within. Lip simple and thin. Parietal callus very thin, white and glassy in center, transparent near outer margin. Protoconch with fine, cui^ved axial striae that grow stronger on the later whorls and become vermiculate on the entire body whorl, being obsolete only on the base and near the aperture, where they turn into straight axial lines. Spiral sculp- ture wanting. Operculum not seen, but presumably like others in this subgenus. Height Diameter mm mm 13.0 16.0 Ma>ari, Oriente, (USNM) Remarks. This species, according to Aguayo & Jaume (1954: 7), is exceedingly rare; the Museo Poey has only a single specimen. The shells of this species and of its two subspecies are easily identifiable by their globular shape and especially by the variously strong vermiculate or wrinkled sculpture. This latter feature alone will prevent confusion with any other species. Aguayo & Jaume believed that the nominate species came from the mountains of Nipe or Mayari rather than from the lowlands around the port of Mayari as Poey stated. This species was omitted by Wagner in his monograph (1907-1908), nor did it appear on the list that Fulton (1915) pub- lished, citing the species Wagner had overlooked. Emoda mayarina gufierrezi Aguayo & Jaume Plate 2, figures 15, 16; Plate 5. Emoda mat/arina fiutierrczi Aguayo & Jaume 1954, Torreia", No. 21, p. 7, pi. 2, figs. 3, 5 ("Finca [Lorenzo Cobos] Estalile," Mayari Arriba, Provincia de Oriente; holotype, MP 17342; para- type, MCZ 187562). Description. This subspecies resembles the nominate species in the globular shape of the shell, but differs in being larger and more strongly sculptured, by having stronger spiral ridges on the base, and in color, which is predominantly reddish with a greenish tinge. 118 Bulletin Mimitm of Comparative Zoology, Vol. 141, No. 3 Height Diameter mm mm 15.0 19.5 "Finca Estable," Mayari Arriba Remarks. As Aguayo & Jaume comment (1954: 8), the color in this form is not of specific importance. In this case it is the larger size and the more rugose, vermicu- late sculptiue which are definitive. Specimens examined. Oriente. Finca "Lorenzo Cobos Estable," Mayari [Arriba], Oriente (E. Pujals). Emoda mayarina mirandensis Aguayo & Jaume Plate 5. Emoda mayarina mirandensis Aguayo & Jaume 1954, Torreia, No. 21, p. 8, pi. 2, figs. 2, 4 ( "Loma de la Cantera" cerca del Central Miranda, Palmarito de Canto, Provincia de Oriente; holo- type, MP 12593). Description. "The present variety is dis- tinguished from the typical subspecies by its larger size, its somewhat more de- pressed form, reddish color and narrower and less irregular sculpture" (translated). IKiulit Diameter mm mm 16.0 21.0 Remarks. We have not seen specimens of this subspecies. Emoda bfanesi Clench & Aguayo Plate 1, figures 12, 13; Plate 5. Emoda blanesi Clench & Aguayo 1953, in Aguayo, Mem. Soc. Cubana Hist. Nat., 21: 301, pi. 35, figs. 1, 2 ( "Embarradas" al norte de Banes, Oriente; holotype, MP 17324, from Em- barradas north of Banes, Oriente, Cuba; paratype, MCZ 257553). Description. Shell about 18 mm in diameter, subglobose, sublustrous and mod- erately strong. Whorls 4V4, moderately rounded, body whorl subinflated and barely descending at the aperture. Color reddish, tinged with yellow on the penulti- mate and body whorls, the yellow color generally becoming stronger as it ap- proaches the aperture; peripheral band light colored with a darker reddish band bordering it below. Spire depressed, rounded. Aperture semilunate, white near lip, darker inside. Lip thickened inside, rounded, not flaring, basal tooth barely perceptible. Parietal callus very thin, white in the columellar region, transparent near outer margins. Suture moderately im- pressed. Postnuclear whorls sculptured with fine, curved axial cords that grow stronger on the successive whorls, the intervals being twice as wide as the axial cords on the body whorl. These cords are irregular and in occasional specimens run together, extending to the base without growing narrower. Spiral sculpture consisting of a subsutural depression on the last whorl, and composed of one to three small inci- sions on the wider and more depressed axial costae. Protoconch IV4 whorls, slightly raised and sculptured with fine, curved axial wrinkles. Operculum with a thin internal corneous layer, yellowish and darker at the outer margins. Height Diameter mm mm 13.8 16.8 Rio Seco, Banes, Oriente 13.8 16.2 La Campana, Rio Banes, Oriente 12.4 17.2 Rio Seco, Banes, Oriente Remarks. This species has a limited dis- tribution along the north central shore of Oriente Province between Banes and Gibara. The shell is easily recognized by its color, the simple, nonflaring lip, the subglobose shape, and the impressed sub- sutural line. It differs from pulcherrima in having stronger axial costae, a simple lip, and in the presence of the subsutural groove with the incisions on the summit of the axial cords. Specimens examined. Oriente. La Cam- pana, Rio Banes; Rio Seco, Banes; Antilla; Loma de Carbon, barrio de Rio Seco; N of Banes (both USNM). Genus Glypfemoda Clench & Aguayo Glyptcmoda Clench & Aguayo 1950, Rev. Soc. Malacol. 'Carlos de la Torre,' 7: 61 (as a sulv genus; type-species, Helicina torrci Henderson 1909, original designation). Emoda and Glyptemoda in Cuba • Clench and Jacobson 119 Description. Shell like that of Emoda sensu stricfo, but more depressed, exceed- ingly rough, dirty white in color, occasion- ally tinged with lemon yellow or purple, and sculptured with strong, thin lirae, separated by flat, much wider intervals. The operculum has the outer surface covered with small, more or less equisized, rounded, glassy granules. In this character it is between the operculum of Emoda with its unmarked surface and the strongly roughened surface of the operculum of ScJiasiclteila. Periostracum wanting. Remarks. Keen (1960: 286) synony- mized Glyptemoda with Emoda s. s. We cannot agree with this decision. Instead we feel that this taxon should be raised to full generic rank. Baker (1922) showed that the lateral tooth complex in Helicininae consists of a hammer-shaped comb-lateral structure with the basal pillar located at one end. In the subfamily Proserpininae, the pillar is cen- trally located, giving an anvil or T-shape to the structure. In Glyptemoda the basal pillar, consisting of two united portions separated by a deep vertical furrow, is almost as large as the top of the tooth itself (PI. 7, fig. 1). Apparently this type of pillar can be derixed more easily from Proserpininae than Helicininae but may be distinct enough to stand in a subfamily itself. There are other radular peculiarities in Glyptemoda. Five of the seven central teeth are oblong in shape and are weakly cusped. In this respect they differ from the lanceolate or obliquely lanceolate cor- responding teeth of the Helicininae and Proseipininae. The numerous uncinals are strongly curved and have one or two \\'eak cusps on the inner marginals to three to fi\'e rather strong cusps on the outer ones. These radular characteristics, together with the strong surface sculpture and de- pressed shape of the shell, as well as the granule-covered outer surface of the oper- culum, induce us to recognize Glyptemoda as a full genus. Glypfemodo forrei torrei (Henderson) Plate 3, figures 1, 2; Plate 5. HeJicina torrei Henderson 1909, Nautilus, 23: 50, pi. 4, figs. 1-3 (Los Negros, 25 miles [sic] southeast of Bayanio in the province of Oriente, Cuba, "in woods on low limestone lulls"; holotype, USNM 463791, paratypes, MCZ 20863). Description. Shell much depressed, reach- ing almost 30 mm in diameter, lustre- less, rough and veiy solid. Whorls 4y4, rapidly widening, almost flat, the last whorl descending shortly at the aperture. Color grayish white or cream, occasional specimens sho\\'ing yellow or reddish purple on the spire. Spire planiform or slightly raised. Aperture semilunate, evenly rounded, generally white inside, occasional specimens with a tinge of yellow deep within. Lip flaring, white, thickened x\'ithin, reflected aboxe, widest above, gradually narrowing at the base, where it terminates in a very small, rounded tubercle. Parietal callus white, lustrous, not raised at the outer edge. Columella white, short, convex above, almost straight below. Suture well impressed. Surface with 17 to 20 thin, sharply raised spiral cords, crossed by regular, rather strong, diagonal axial sculp- ture. This sculpture crenulates the upper edge of the spiral cords. Intervals much wider than the cords, crossed by the axial growth lines. Protoconch 2 large whorls, smooth but with closely spaced, cun'cd axial wrinkles, rather more closely spaced than in most species of Emoda sensu stricto. No periostracum. Operculum as in subgenus. Height Diameter mm mm 16.5 28.9 Los Negros, Jiguani, Oriente, para- t>'pe 16.0 29.0 Los Negros, Jiguani, Oriente, para- type 14.0 30.5 Los Negros, Jiguani, Oriente, para- type 13.5 26.0 Los Negros, Jiguani, Oriente, para- type Remarks. The number of spiral ridges varies from 17 to 20 in a series of paratype 120 Bulletin Museum of Comparative Zoology, Vol. 141, No. 3 specimens and from 15 to 20 in a series from the de la Torre collection merely labelled "Los Negros, Jiguani." This vari- ation in number may be due to the pres- ence of thinner and lower subsidiary ridges in the intervals in some specimens. The colors as given by Henderson do appear in isolated specimens, but most, even those taken alive, are white or cream colored under an encrusting layer of dirt. The spire, which is completely flat in some specimens, is moderately to well raised in others. This might be a secondary sexual characteristic. Henderson's location of Los Negros at 25 miles southeast of Bavamo is in error. Los Negros in Oriente is only about nine miles from Bayamo. Specimens examined. Oriente. Los Ne- gros, 15 km SE of Bayamo. Glyptemoda forrei freirei Clench & Aguayo Plate 3, figures 3, 4; Plate 5. Emoda {Glijptemoda) ionci freirei Clench & Aguayo 1950 Rev. Soc. Malacol. 'Carlos de la Torre,' 7: 61, pi. 12, figs. 1-3 (La Manuela, antes de llegar a la Furnia de la Mina Los Negros, Baire, Oriente; holotype, MP 12918; para- types, MCZ 18.5798). Emoda pujaisi .\guayo & Jaume 1954, Torreia, No. 21, p. 5, pi. 1, fig. 8 ("La Cloria," Union del Camino de Maffo a San Lorenzo y a Los Negros, Jiguani, Provineia de Oriente; holotype, MP 17386). Description. Shell as in torrei but differ- ing in its smaller diameter (5 to 8 mm less than the nominate species), its proportion- ately more elevated shape, and by the spiral ridges, which are lower and flattened at the top rather than sharp. The speci- mens from the Cueva de Pepin are tinged with yellow or brick red. Height Diameter mm mm 14.5 20.0 "La Gloria," Maffo (as E. puialsi) 14.3 20.0 "La Gloria," Maffo (as E. pujaisi) 14.0 21.5 La Manuela (type-locality) Remarks. According to Clench & Aguayo, Jose A. Freire, who collected the type specimens, reported that he found this sub- species in the lowlands, whereas torrei torrei lives in the higher elevations. We have not seen the operculum, but presum- ably it is similar to that of torrei torrei. Emoda pujaisi Aguayo & Jaume is a neighboring population that differs from the present subspecies only in degree. The two forms are found close together in the vicinity of Los Negros, and there is little doubt that they are the same. Specimens examined. Oriente. La Muela; Cueva de Pepin, Baire (USNM); "La Gloria," Maffo, Jiguani (type locality of E. pujaisi). Emoda (?) dementis Clench & Aguayo Plate 3, figure 5. Emoda dementis Clench & Aguayo 1950, Re\-. Soc. Malacol. 'Carlos de la Torre,' 7: 62, pi. 12, figs. 1, 5; (Cayo del Key, Mayari, Oriente; holo- type, MP 12923; paratype, MCZ 185799). Description. Shell imperforate, thin, fragile, and translucid. Color uniformly bluish green. Whorls 2 in the juvenile specimens, suture well impressed, body whorl with rounded keel. Spire depressed, little raised above the plane of the body whorl. Aperture semilunate; umbilical cal- lus microscopically granulose. Columella short, blending gradually into the basal lip. The spiral sculpture' consists of irregular microscopic striae grouped in unef(ual series; axial sculpture of fine irregular, diagonal, growth lines. Protoconch P/j whorls, axially striated, the striations crossed by spiral lines; with four curved, rounded furrows at the distal portion. Periostracum imperceptible; opercukun thin, green, smooth, nucleus submarginal, with a sculpture of concentric lines. Height Diameter mm mm 6.5 10 (juvenile) Remarks. This species was described on the basis of four juvenile specimens, and was doubtfully assigned to Emoda. These doubts are justified. The protoconch, as described bv the authors and as seen in the Emoda and Glyptemoda in Cuba • Clench and Jacobson 121 specimen in MCZ, lacks the regular, curved, axial striae or wrinkles that characterize the genus, and also shows faint spiral sculpture that is not found on the proto- conch of Emoda sensu stricto. The shell is very fragile, almost transparent, weaker than the customarily strong shell of Emoda, even in juvenile specimens. The true place- ment of dementis Clench & Aguayo must await the discovery of additional and more mature specimens. REFERENCES CITED Adams, H. and A. 1856. The Genera of Recent Mollusca, London, 1: 300-309, 3: pi. 87, (Helicinidae). Aguayo, C. G., and M. L. Jaume. 1954. Notas sobre el genero Emoda, Torreia, Habana, no. 21: 4-13. Arango, R. 1878-1880. Contiibucion a la Faima Malacologica Cubana, Havana, pp. 280 + 35 (inde.x and errata). Baker, H. B. 1922. Notes on the radula of the Helicinidae, Proc. Acad. Nat. Sci. Philadel- phia, 74: 29-67. . 1926. Anatomical notes on American Helicinidae, ibid., 78: 35-56. 1940. New snbgenera of Antillean Heli- cinidae. Nautilns. 54: 70-71. Bourne, G. C. 1911. Contribntions to the mor- phology of the gronp Neritacea at the aspido- branch gastropods. Proc. Zool. Soc. London, 1911: 759-809. Clench, William J., and Morris K. Jacobson. 1968. Monograph of the Cuban genus Viana (Mollusca: Archaeogastropoda: Helicinidae), Breviora, Mus. Comp. Zool., no. 298: 1-25. . 1970. The genus Priotrochatella of the Isle of Pines and Jamaica, West Indies. Occas. Pap. Moll., Mus. Comp. Zool., Cambridge, Mass., 3(39): 61-80. Dance, S. P. 1966. Shell Collecting. Berkeley and Los Angeles: University of California Press, 344 pp. Fischer, P. 1885. Manuel de Conchyliologie, Paris, pp. 794-799, figs. 554-557 (Helicini- dae). FoRCART, L. 1950. Von Dr. H. P. Schaub auf der Insel Cuba gesammelte Land- und Siisswasser- schnecken. Arch. Molluskenk. 79: 67-72. Fulton, H. C. 1915. On Dr. Anton Wagner's Monograph of Helicinidae in the Conchylien- Cabinet, 1911. Proc. Malac. Soc. London, 11: 237-241; 324-326. Isenkrahe, C. 1867. Anatomic von Helicina titanica. Arch. Naturg., 33rd year. Vol. 1, pp. 50-72. Keen, M. 1960. Helicinidae. In Treatise on In- vertebrate Paleontology, Mollusca, University of Kansas, Lawrence, Kansas, 1 : 285-288, figs. 186-187. Lea, I. 1834a. Observations on Naiades and Descriptions of New Species of that and other families. Trans. American Phil. Soc. (N. S.), 5: 23-119. . 1834b. Observations on the genus Uiiio, Philadelphia, 1- 1-232. Mayr, E. 1969. Principles of Systematic Zoology. New York: McGraw-Hill, 11 + 428 pp. Orbigny, a. d'. 1842. Mollusques. In Sagra, Histoire Physique, Politique et Naturelle de rile de Cuba, Paris, 2: 1-264. Pfeiffer, L. 1850-1853. Helicinacea. In Martini & Chemnitz, Conchylien-Cabinet, ( 2 ) 1 : sect. 13, pt. 1, 78 pp. . 1854-1860. Novitates Conchologicae, Cassel, 1: 1-138; ibid., 1860-1866, 2: 140- 303. 1860. Zur Molluskenfauna der Insel Cuba, Malak. Bliit, 6: 66-102. PoEY, F. 1851-1854. Memorias sobre la Historia Natural de la Isla de Cuba, Havana, 1 : 1-463; op. cit., 1856-1866, 2: 1-442. : — . 1865. Notas. In Rafael Arango, Catalogo de los Moluscos Terrestres y Fluviales de la Isla de Cuba, Repertorio fisico-natural de la Isla de Cuba, Havana, pp. 71-149. Reeve, L. a. 1874. Conchologia Iconica, London, Vol. 19, 34 pis. and letter press (Helicini- dae ) . SowERBY, G. B., II. 1847. Thesaurus Conchy- liorum, London, Vol. 1, monograph of the genus Helicina, pp. 1-16. . 1866. O)). fit., Vol. 3, second monograph of the genus Helicina, pp. 277-302. Thiele, J. 1929. Handbuch der systematischen Weichtierkunde, Jena, 1: 80-91, figs. 57-67, ( Helicinidae ) . Troschel, F. H. 1857. Das Gebiss der Schnecken, Berhn, 1: 75-85 (Helicinidae). Wagner, A. 1907-1908. Helicinidae. In Martini & Chemnitz, Conchylien-Cabinet, Niirnberg, (2) 1: sect. 18, pt. 2, pp. 46-138. Wenz, W. 1938. Gastropoda. In Handbuch der Paliiozoologie, Helicinidae, Berlin, 6: pt. 1, pp. 435-448. (Received 15 January 1970.) 122 Bulletin Museum of Comparative Zoology, Vol. 141, No. 3 PLATE 1. Figs. 1 and 2. Emoda submarginata (Gray), El Grille, Madruga, Havana, MCZ 60969, Topofype (1.5 X). Figs. 3 and 4. f. na;ozaens/s Aguayo & Jaume, El Cacaotal, Sierra de Najaza, CamagiJey, MCZ 257422, Paratypes (1.5 X)- Figs. 5 and 6. E. bermudezi Aguayo & Jaume, Banao, Cubitas Mountains, Camaguey, MCZ 262719 (1.5 X)- Fig. 7. £. ciliata (Poey), San Juan de Letran, Trinidad, Las Villas, Lectotype, MCZ 262717 (1.5 XI- Figs. 8 and 9. £. ciliata (Poey) same localify, MCZ 127230, Topotypes (1.5 Xl- Figs. 10 and 11. E. ciliata gu/sono Wagner, 'Cuba," MCZ 262718 (1.5 Xl- Figs. 12 and 13. E. b/ones/ Clench & Aguayo, Rio Seco, Bones, Orients, MCZ 1 35961 (1.2X1- Fig. 14. E. pu/c/ierrimo (Lea), USNM 104613, Lectotype (1.5 Xl- Fig. 15. E. pulcherrima (Lea), Trans. American Philos. Soc. (NS), pi. 19, fig. 57, photo- graphic copy. Figs. 16 and 17. E. sagraiana (Orbigny), Ensenada de los Burros, Cabezas, Pinar del Rio, MCZ 127148 (1.5 XI- Figs. 18 and 19. E. sograiono forma percrassa Aguayo & Jaume, La Gijira, Luis Lazo, Pinar del Rio, MCZ 127253 (1.5 XI- Figs. 20 and 21. E. silacea (Morelet), Yunque de Baracoa, Oriente, Topotype, MCZ 127229. Emoda and Glyptemoda in Cuba • Clench and Jacobson 123 124 Bulletin Museum of Comparative Zoology, Vol. 141, No. 3 PLATE 2. Fig. 1. Helicina briarea Poey, Trinidad Mountains, Santa Clara [Las Villas], MCZ 73854, Lectotype (1.4 X)- Figs. 2 and 3. E. briarea (Poey), "Puriales," Trinidad Mountains, Las Villas, MCZ 58560 (1.4 X)- Fig- 4. Helicina baya- memii Poey (Buena Vista, Bayamo, Oriente) MCZ 73855, Lectotype (c. 1.4 X)- Fig. 5. ditto, MCZ 256496, Paralectotype. Figs. 6 and 7. Emoda emodo Pfeiffer, Manatial de Santa Fe, Oriente, MCZ (1.2 X). Figs. 8 and 9. He//c/no titanica Poey, Barocoa (Oriente), MCZ 256495, Syntypes (1.2 X)- Fig. 10. Emodo pujalii Aguayo & Jaume (^ torre; treirei Clench & Aguayo), Loma La Gloria, Maffo, JiguanI, Oriente, MCZ 263902, Topotype (2.1 X). Fig, 11. Emoda caledonlensis n. sp.. Mine Caledonia, Mayari, Oriente, MCZ 261352, Paratype (1.2 Xl- Fig. 12. ditto, 1 km from pine forest, Mayari, MCZ 261349, Paratype (1.2 X)- Figs. 13 and 14. E. pulcherrima (Lea), Pico Turquino, bet. 2000 and 5000 ft., Oriente, MCZ 267577 (1.5 X). Figs. 15 and 16. £. mayan'na gutierrezi Aguayo & Jaume, Finca Estable, Mayari Arriba, Oriente, Paratype, MCZ 187562 (1.2 X)- Fig. 17. Helicina mayarina Poey, Mayari, Oriente, Syntype, USNM. (1.5 X)- Emoda and Glyptemoda in Cuba • Clench and Jacohson 125 PLATE 3. Figs. 1 and 2.^ Glyptemoda torrei torrei Henderson, Los Negros, Jiguani, Oriente, MCZ 20863, Paratypes (1.5 X). Figs. 3 and 4. Glyptemoda torrei freirei Clench & Aguayo, La Manuela near the mine, Boire, Oriente, MCZ 185798, Poratype (1.5 X)- Fig. 5. fmodo ? dementis Clench & Aguayo, photographic copy of original figure. Rev. Soc. Mai., 7: pi. 12, fig. 4. Figs. 6 and 7. Helicina crassa Orbigny ( = pukherrima Lea), Syntype, BM(NH) No. 1854.100.4.169 (1 X). (Photo courtesy of Angela Cane, BM(NH).) 126 Bulletin Museum of Comparative Zoology, Vol. 141, No. 3 fig.i PLATE 4. Fig. 1. Distribution of (1) Emoda sagraiana (Orbigny); (2) E. briarea (Poey); (3) E. ciliata (Poey); (4) E. bermudezi Aguayo & Joume; (5) E. najazaensis Aguayo & Jaume. Fig. 2. Distribution of Emoda submarg/nafo (Gray). Emoda and Glyptemoda in Cuba • Clench and Jacohson 127 o c °=, E o Uj ~ D _ — O dj" E j; E ^ >^ => ° ^ ^ c — oa • 2 D O J I r i O) O) - Z) c IT) «/i a X sagraiana silacea pulcherrima p. titanica suhmarginata ciliatu c. guisana enwda caledonietisis haijamensis bermtidezi Jiajazaetisis hriarea mcujarina m. gutierrezi m. mirandensis hlanesi torrei torrei torrei freirei X X X X > 1-1 X X ? >, :3 us o X X X •c o X X X X X X X X X X X X X 130 Bulletin Museum of Conipamtive Zoulugij, Vol. 141, No. 3 INDEX Alcadia, 99, 100, 102, 103 ba^i^eiisis, Alcadia nuda, 102 hastidaiui, llcUcina, 113, 111, 115 basUdknia, IlcUcina, 113 bayiunensis, Enioda, 113, 114, 115 bayhamensis, Helicina, 113 bennndezi, Emoda, 100, 110, 114 blanesi, Emoda, llcS hriaraea, Helicnia, 1 Hi briarea, Emoda, 101, 112. 116 hrkirca. Ilclicina, 116 caledonit'iisis, Emoda, 99, 112 cat (dine mis, Helicina, 105 catuUnianu, Ilclicina, 105 Ceratodiscus, 99 ciliiita, Alcadia (Emoda), 110 ciliata, Emoda, 100, 102, 103, 110, 111, 114 ciliiita, Helicina, 1 10 dementis, Emoda, 120, 121 crassa, Helicina, 101, 107, 108 Emoda. 99, 100, 101, 102, 119, 120, 121 emoda, Alcadia (Emoda) saf:iaiana, 112 emoda, Emoda, 112, 117 emoda, Emoda sagraiana, 112 emoda, Helicina, 112 Eutiochatella, 99 fossuUita, Helicina, 110, 111 freirei, Emoda (Glyptemoda) toirci, 120 Ircirei, (dypteiiioda torrei, 120 Glyptemoda, 99, 118, 119 gtii-sana, Alcadia (Emoda) ciliata, HI miisana, Emoda ciliata. 111 miticrrezi, Emoda mayariiia, 117 Helicina, 99 Hclicinidac, 99 llelicininae, 119 jeannercii, Helicina, 113, 114 lamellosus, Thais, 106 lapillus, Thais, 106 Liuidclla, 99 mayarina, Emoda mayarina, 117 ntaijarina, Helicina, 117 niirandensis, Emoda mayarina, 118 najazaensis, Emoda, 100, 110, 115, 116 nuda, Alcadia, 102 OL hracca, Helicina, 100, 104, 105 ])erc>a.ssa, Emoda sagraiana, 100, 105, 106 phnospira, Alcadia (Emoda) ptdcherrima, 109, 110 Priotiochatella, 99 Proserpina, 99 Proserpininae, 119 pulcherrima, Emoda pulchcrrima, 100, 107, 108, 109, 110, 114, 118 pulchcrrima, Helicina, 99, 107 rubra, Helicina, 109 rubrocinctu, Helicina, 107, 108 sagra, Helicina, 105, 114 sagra, Trochatclla, 105 sagraiana, Emoda, 100, 102, 103, 105, 112, 114, 117 sagraiana, Helicina, 105, 106 Schasicheila, 100, 101, 102, 119 silacea, Alcadia (Emoda), 104 silacea, Emoda, 101, 102, 104, 111 Striatemoda, 101 striatura, Alcadia, 101 submarginata, Emoda, 100, 103, 108, 109, 110, 111, 114, 115, 116 submarginata, Helicina, 109 titanica, Alcadia ( Emoda ) ptdcherrima, 108 titanica, Emoda pulchcrrima, 103, 108, 113, 114 titanica, Helicina, 103, 108 torrei, CUyptemoda torrei, 119, 120 torrei, Helicina, 119 Viana, 99, 100 zaya.si. Emoda, 109 Note: Synonyms are italicized. OF THE Museum of Comparative Zoology The Diadematus Group of the Orb- Weaver Genus Araneus North of Mexico (Araneae: Araneidae] HERBERT W. LEVI HARVARD UNIVERSITY VOLUME 141, NUMBER 4 CAMBRIDGE, MASSACHUSETTS, U.S.A. 4 FEBRUARY 1971 PUBLICATIONS ISSUED OR DISTRIBUTED BY THE MUSEUM OF COMPARATIVE ZOOLOGY HARVARD UNIVERSITY Bulletin 1863- Breviora 1952- Memoirs 1864-1938 JoHNSONiA, Department of Mollusks, 1941- OccAsioNAL Papers on Mollusks, 1945- Other Publications. Bigelow, H. B., and W. C. Schroeder, 1953. Fishes of the Gulf of Maine. Reprint, $6.50 cloth. Brues, C. T., A. L. Melander, and F. M. Carpenter, 1954. Classification of Insects. $9.00 cloth. Creighton, W. S., 1950. The Ants of North America. Reprint, $10.00 cloth. Lyman, C. P., and A. R. Dawe (eds.), 1960. Symposium on Natural Mam- malian Hibernation. $3.00 paper, $4.50 cloth. Peters' Check-list of Birds of the World, vols. 2-7, 9, 10, 12-15. (Price list on request. ) Turner, R. D., 1966. A Survey and Illustrated Catalogue of the Teredinidae (Mollusca: Bivalvia). $8.00 cloth. Whittington, H. B., and W. D. I. Rolfe (eds.), 1963. Phylogeny and Evolution of Crustacea. $6.75 cloth. Proceedings of the New England Zoological Club 1899-1948. ( Complete sets only. ) Publications of the Boston Society of Natural History. Authors preparing manuscripts for the Btilletin of the Museum of Comparative Zoology or Breviora should send for the current Information and Instruction Sheet, available from Mrs. Penelope Lasnik, Editor, Publications Office, Museum of Comparative Zoology, Harvard University, Cambridge, Massachusetts 02138, U.S.A. © The President and Fellows of Harvard College 1971. THE DIADEMATUS GROUP OF THE ORB-WEAVER GENUS ARANEUS NORTH OF MEXICO (ARANEAE: ARANEIDAE) HERBERT W. LEVI ABSTRACT Among our commonest spiders are the large Aiancus orb-weavers. Eighteen species are found north of Mexico. They are redescrilied, their diagnostic characters ilhistrated, their ranges mapped, and notes on their habits recorded. Three species turned out to be new: A. loashingtoni from northeastern U. S. and eastern Canada; A. yiikon from Yukon Territory; and A. pima from the Southwest. INTRODUCTION Among the species included in the present revision are some of the commonest spiders of North America, although several Araneus species are quite rare. Collections were examined not only to get an idea of the variation and distribution of common species, but also with the hope of finding members of the rarer species. The enormity of the collections available slowed down the work immensely. Because of the urgent need for identifi- cation and information on the distribution of the common spiders, I am presenting separately the diademotus group of the genus Aroneus, some members of which have never been illustrated before. It may take many years before I have examined types of all names of the numerous genera placed within the family and the many species described in the genus Aroneus in the Americas. At present I do not know the limits of the genus and have not de- cided whether certain species are best included in the genus Araneus. The recognition of reliable diagnostic morphological features of species was one of the critical questions to be resolved. My conclusions agree entirely with those of Grasshoff (1968), but differ from those published by Archer (1951a, b). Archer tliought that the species can be diagnosed by the shape of one single sclerite of the male palpus, the median apophysis, and the ventral view of the scape and epigynum of the female. Perhaps these structures will prove of value in differentiating genera; they certainly are of quite limited value for species diagnosis. Also, I agree with Grasshoff that leg spines have doubtful value in separating males of Araneus spe- cies; they are extremely variable within populations (L. D. Carmichael, in manu- script ) . As in my revisions of theridiid spiders, long established names of common species were kept; changing them does not make sense. ^ ^ I am following the purpose of the International Code on Zoological Nomenclature as expressed in its Preamble, although other authors have oc- casionally in my opinion intei-preted individual pro\isions out of the context to obligate tlie changing of names. For many common North American species older names are available, but these often have doubtful application as the types have been lost, and interpretation of the usually inadequate description depends on the experience of the reader. Doujjtful also are many of the names used bv Chamberlin and Ivie (1944) in Bull. Mas. Conip. Zool, 141(4): 131-179, February, 1971 131 132 Bulletin Mmeum of Comparative Zoology, Vol. 141, No. 4 In this study the work of the late Dr. H. Wiehle on European spiders has been of tremendous help, as has also the recent study on variation and moiphological cri- teria of several European Amneus species by Grasshoff (1968). I wish to thank the numerous biologists who have contributed specimens for study: Mr. J. D. Berman; Mr. D. Bixler; Dr. R. E. Crabill of the United States National Museum; Mr. Bruce Cutler; Dr. C. D. Dondale sent collections of the Canadian National Museum; Dr. M. G. Emsley of the Academy of Natural Sciences, Philadelphia; Dr. W. J. Gertsch of the American Museum of Natural History; Dr. O. Kraus of the Senckenbergische Naturforschende Gesell- schaft; Mr. Robin Leech; Mr. G. H. Locket; Dr. D. C. Lowrie; Dr. M. Moritz of the Zoologisches Museum der Humboldt Uni- versitat, Berlin; Mr. P. L Pcrsson of the Natural History Museum, Stockholm; Mr. T. R. Renault of the Canada Department of Forestry, Fredricton, New Brunswick; Miss Susan Riechert; Mr. V. D. Roth of the Southwestern Research Station; Dr. J. G. Sheals and Mr. D. Clark of the British Museum (Natural History); Dr. R. Snetsinger; Mr. W. A. Shear; Dr. E. Sutter of the Naturhistorisches Museum, Basel; Dr. S. L. Tuxen and Mr. B. Petersen of the Universitetets Zoologiska Museum, Copen- hagen; Mr. L Valovirta, University of Hel- sinki Zoological Museum; Dr. H. V. Weems, Jr. and Mr. K. J. Stone of the Florida Col- lection of Arthropods; Dr. T. Yaginuma; and Dr. G. Edmundson and the staff of the University of Utah collections. My wife has helped with writing and editing. This investigation was su]Oported in part by Public Health Service Research Grant AL01944 from the National Institutes of Allergy and Infectious Diseases. their attempt to resurrect names of Walckenaer for the manuscript ilkistrations of Abbot. Sub- jective judgment has to be used in interpreting tliese rather primitive ilkistrations of the dorsal aspect of spiders, tlie types of these names. While other biologists may feel sure of their judgment in these dou])tful cases, I am not, and after hav- ing obtained copies of Abbot's drawings I find that I disagree with many of the Chamberlin and Ivie assignments; many are obviously in error (Levi and Levi, 1961). The Statute of Limi- tations ( Art. 23b ) permits us now to assign many of these uncertain ancient names to the status of obsolete names, should they threaten familiar names that have been in common use for more than 50 years. Like any other tool used by the taxonomist, the Code has to be used with common sense. My request to the Commission to resolve by plenary power the problem of the two family names, Araneidae versus Argiopidae, has been withdrawn because only a few, often those in opposition and those with least experience in taxonomy and nomenclature, state their opinions in writing to the Secretary. However, the non- controversial request to place the name Argiope on tlie Official List of Generic Names in Zoology has not been witlidrawn. Although sent to the Secre- tary of the Commission in May 1967, it has not been printed yet. Once printed the name is protected by Art. 80 of tlie Int. Code Zool. Nomcncl. Instead of submitting the problem of the family name to the Commission, I have circulated a request for the opinion of colleagues. The poll included, of course, nontaxonomists working witli orb- weavers. A majority of correspondents favored Araneidae (Araneidae, 43; Argiopidae, 29; Argyo- pidae, 1; outside of North America, Araneidae, 28; Argiopidae, 25; Argyopidae, 1). I will con- tinue to use the name Araneidae. The problem of the spelling of spider generic names is puzzling. The Code (Art. 32) now considers incorrect the emendations of spellings by Thorell, almost universally adopted for 100 years. Changing the names now would conflict witli Art. 23b (since the incorrect emendations are junior objective synonyms. Art. 33a), and also with the purposes of the Code and would cer- tainly be wrong. Thus it seems to me that we have to follow the long accepted spellings of generic names as also generally adopted by Bonnet in the Bihliographia Arauconim, and at- tempt to bring critical cases to the attention of the Commission. It is possible tliat the pertinent paragraphs of the Code might be clarified by a future International Congress to a\oid unneces- sary name changes. Araneus diadematus group Orb-weavers • Levi 133 Araneus Clerck^ Araneus Clerck, 1757, Svenska Spindlar, p. 15. Type species designated by Simon, 1893, His- toire Naturelle des Araignees, 1: 829, A. angu- latiis Clerck. But the type is said to be A. diadematus Clerck by Petrunkevitch, 1928, Trans. Connecticut Acad. Sci., 29: 136 and Bonnet, 1955, Bibliographia Araneorum, 2: 408, altliough the type species is correctly stated to be A. angtdattis by Petrunkevitch, 1911, Bull. Amer. Mus. Natur. Hist., 29: 255 and by Bonnet, 1950, Bull. Soc. d'Hist. Natur. Toulouse, 85: 1-9. Aranea Linnaeus, 1758, Systenia Naturae, 10th ed. p. 619. The only genus of spiders; A. diadema is listed first. No valid type designation seems to have been made previously, thus I here designate A. diadema Linnaeus. The generic name Aranea has always assumed to be a synonym for Araneus Clerck. Epeira Walckenaer, 1805, Tableau des Araneides, p. 53. Type species designated by Latreille, 1810, Considerations Generales, p. 424, Aranea diadema Linnaeus; a second designation is by Thorell, 1869, On European Spiders, p. 53, Epeira diademata (Clerck). In 1928 Petrunke- vitch (Trans. Connecticut Acad. Sci., p. 136) indicated that E. cornuta (Linnaeus) [sic] was the type. Presumably he meant Araneus cor- nutus Clerck. Neopora Simon, 1864, Histoire Naturelle des Araignees, p. 261. A name for a subgenus. The type species is N. diadema ( = Araneus diade- matus Clerck) designated by Bonnet, 1958, Bibliographia Araneorum, 2: 3054. Burgessia McCook, 1894, American Spiders, 3: 182. A subgenus for the group comprising corticaria, miniata, honsallae, mayo, hispinosa, paeificae as well as forata, linteata, and juniperi. The type species is Epeira corticaria (Emer- ton) ( = Araneus corticarius) here designated. Bonnet, 1955, Bibliographia Araneonnn, 2: ^17, ^ Although the starting point of zoological nomenclature is Linnaeus' Sijstema Naturae, 10th edition, with the arbitraiy date of 1 January 1758, the work of C. Clerck, 1757, Svenska Spindlar, pul)]ished before, is an exception per- mitted by Article 26 of the old International Code on Zoological Nomenclature passed at the XIII International Congress of Zoology of 1948. After tlie XV International Congress of Zoology in London in 1958 adopted a new Code (1961), Clerck's Aranei Suecici was placed on the Official List of Accepted Works by Direction 104 of the International Commission on Zoological Nomen- clature [1959, Bull. Zool. Nomencl., 17(3-5): 89-91]. was in error when he said that McCook failed to include species in the sul:)genus. Euaranea Archer 1951, Amer. Mus. Novitates, 1487: 34. Type species for new subgenus; Aranea cavatica (Keyserling) by original desig- nation. Tlw structure of Araneus genitalia. The terms used for the sclerites of the x^alpus are those of Comstoclc (1910), which have become widely used. I used them in the revisions of Theridiidae and of Argiope ( 1968 ) and they were also used by Grass- hoff (1968). No terms are in general use to describe the female genitalia; I here follow Grass- hoff. The epigynum has a prominent scape (Fig. 1). The scape is attached to the base. Below and behind the base are a pair of basal lamellae which are large and extend on each side of the epigynum of Araneus marmoreus (Figs. 1-3). The slit- like openings are on the venter and lead into a funnel which continues posteriorly but is partly open on one side as a groove. The groove, as can be seen in cleared or macerated preparations, runs into a funnel- shaped chamber toward the middle of the base and then bends toward the outside and continues under the lateral sclerites ( crosshatched in Figs. 1-3). In A. diade- matus a median posterior sclerite ( Fig. 36 ) covers the grooves, while the funnels are more or less open posteriorly in A. ;?7<7r- moreus (Fig. 3). The palpus of Araneus has a large terminal apophysis (term, apoph., Figs. 4- 6, 8), a sclerite below the subterminal apophysis (subterm. apoph., Figs. 4-6, 8), and distal hematodocha ( dist. hemat. ) that are absent in Argiope and Gea ( Le\'i, 1968) . Species differences. Archer (1951) de- scribed and pictured only the median apophysis of the palpus (med. apoph. in Figs. 4-6, 8) and the scape of the epigy- num. He erected numerous araneid genera on the basis of the shap)e of the median apophysis. Unfortunately, the median apophysis turns out to be unreliable for separating species of Araneus and is pre- 134 Bulletin Mmeiim of Comparative Zoology, Vol. 141, No. 4 sumably a poor character for separating genera. Related species often have a similar median apophysis (see A. gemma group. Figs. 199, 211, or A. saevus and A. diadcmatus. Figs. 38, 55). There are also indix'idiial differences in this strvicture within species (e. g., A. nordmanni. Figs. 61, 63, 65, 67). This unreliability of the median apophysis as a species specific char- acter is of interest in view of its definite function in copulation: its median spine takes hold of the tip of the scape of the epigynum (van Helsdingen, 1965; Grass- hoff, 196S). The form of the scape also varies among individuals of a species. The distance between the spine and the distal spine or ridge on the median apophy- sis varies in different species with the length of the epigynal scape. The spine is close to the distal end of the median apophysis in species having a short scape (A. gemma, A. gemmoides, A. pima). The distance is large in those having a long scape (A. bicentenarius, A. andrewsi, A. diadcmatus; and A. saevus). In the related genus Eriopliora, in which the particularly long scape reaches to the spinnerets, the scape is matched by a shift in position and elongation of the median apophysis. The shape of the conductor is of diag- nostic value only in some species: it is an unusually large sclerite in A. corticarius (Fig. 120), and a very small, narrow one in A. trifolium (Fig. 179). The conductor in copulation embraces the scape of the epigynum, facing the median apophysis, forming a track on which the scape can glide in two directions. Of far greater diagnostic value in sepa- rating species arc the shapes of the embolus and terminal apophysis (term, apoph. in Figs. 4-6, 8), structures easily seen but often slighted in illustrations of araneid species. (Seen in all illustrations of the mesal view of the palpus in this paper.) Part of the embolus enters the female duct system (the left embolus, the left opening and ducts). But during copulation the terminal apophysis comes to lie below the base of the scape, against the abdomen, and supplies support. The cap that is present on the embolus of virgin males (Levi, in press) differs in shape in different species. The caps of some species (e.g., A. ilknidatiis) are very large and noticeable. The paracymbium differs greatly among species within the genera Meta and Zijgi- eUa (Gertsch, 1964), but is similar in almost all Araneus species. This only illus- trates the difficulty of making generali- zations or of attempting to use only one sclerite of the palpus, such as the median apophysis, for species diagnosis in all genera of the family. The paracymbium, during copulation, lies near the base of the median apophysis; it prevents the median apophysis from turning and by hooking into the tegulum border, acts as a stopping wedge between median apophysis and tegulum. In the female the best diagnostic char- acters are those of the epigynum. How- ever, the length and wrinkling of the prominent scape varies, perhaps due to different folding. In most species the scape is short; however, in some species it is always long (bicentenarius, andrewsi, dia- dcmatus, and saevus). The general shape varies among species. The scape may break off during mating in almost any species, and usually is torn off in A. corti- carius and A. groenlamlicohis. The general structure of the base of the epigynum is much less variable and differs more between species than among individ- uals of a species. As it is undesirable to cut off the scape to view the base, it is best examined in posterior view by pulling the epigynum slightly away from the abdomen. Unfortunately, most authors have not illus- trated this posterior view, exceptions being Wiehle (1963) and recently Gertsch (1964) in revising American species of the araneid genus Zygiclla. The basal lamellae are unfortunately also (|uite variable among different individuals of a species, as is shown by Grasshoff (1968, p. 46, fig. 41). Araneus diadematus group Orb-weavers • Levi 135 However, genitalia are not the only char- placement" is a useful term proposed by acters. Araneus corticarius has highly dis- Brown and Wilson (1956) for the phenom- tinctive abdominal shape and coloration enon that morphological characters of (Fig. 118). It cannot be confused with any related species having overlapping ranges other North American species. The cross- show greater difference in the area of shaped arrangement of markings on the overlap than in their allopatric distribution, dorsum of A. diadematus (Fig. 37) is diag- Examples have been found whenever taxo- nostic in North America, although not in nomic revisions are made and there is Europe. abundant material. In my theridiid re- The second tibia in males of many spe- visions I found character displacement be- cies has strong spines and is often bent, tween Tlieridion montanum Emerton and Grasshoff (1968, p. 24, fig. 17) illustrated T. lawrencei Gertsch and Archer (Levi, the considerable variation within the four 1957a, p. 72). Specimens of Steatoda hes- spccies studied. While the spine pattern pera Ghamberlin and Ivie and Steatoda is not a good character for separating spe- horealis (Hentz) are more distinct where cies of Araneus (L. D. Carmichael, in the borders of their ranges meet, although manuscript), the presence or absence of I failed to point this out in my revision spines and modifications of tibia 2 may be (Levi, 1957b). Argiope trifasciata (For- used to separate some species. skal) males have a noticeably smaller The male of most species has a hook on palpus in the area of overlap with the very the distal margin of the first coxa (Fig. similar A. florida Ghamberlin and Ivie, 183), which fits into a proximal depression which has a larger palpus (Levi, 1968, p. on the dorsal surface of his second femur, 335, 337). coupling the legs during mating. The coxal The possibility of drawing erroneous hooks are small and more posterior in some conclusions by ignoring geographic and species (A. gemma) and absent in others individual variation and picking up in- (A. cavaticus). A cone or spur is present dividual specimens is best showii in on the second coxa (Fig. 183) in some Araneus nordmanni. Araneus nordmanni is species. The hook on the first coxa and found in the same habitat, on forest trees, the corresponding depression are found in as the dark-colored A. saevus. Where the species of many araneid genera and do not ranges of the two species overlap, A. nord- seem to reflect close affinity. manni is rarely dark, and shows little vari- Any diagnostic feature of any animal ation in size and shape. However, outside species is subject to variability within a the range of A. saevus, in the southern part certain range. This is true also of spiders, of the range of A. nordmanni, it is often Guriously, this is denied by some spider black, it varies in size, and in no two males specialists who consider any differences in are the embolus (Figs. 69-75) and median structure (individual or geographic) to apophysis exactly the same shape. Araneus represent a "species difference." The vari- nordmanni is largest in the southwestern ability of many characters of four Araneus part of its range. By looking only at the species of Europe has been beautifully median apophysis of the palpus of a few illustrated by Grasshoff ( 1968 ) . My obser- specimens and noting differences in size vations support Grasshoff 's completely: and color. Archer described A. pseudo- while sclerites in the palpus ( Figs. 96-99, melaena from the southwest and A. darling- 103, 104) and all parts of the epigynum toni from southeast. The differences Archer show some variation among individuals observed are there, but by examining larger (Figs. 76-92), nevertheless there are gaps samples one can easily see that the differ- in the variability between species. ences are within the variation of A. nord- Character displacement. "Character dis- manni. Perhaps the differences among 136 Bulletin Museum of Comparative Zoology, Vol. 141, No. 4 populations in the southwest are empha- dii.s (Olivier) and tiny A. poUidus males sized by the isolation of the habitats in will court A. dkidematiis. In courtship the which this forest species can sur\dve. male plucks web threads in a rhythm char- Introp-ession. Of considerable interest acteristic for his species. Females do not is the Araneus <^emma group of species, respond to the plucking of a male of the including fi\'e species north of Mexico. The wrong species. One male of A. diadcmatus eastern A. cavaticus is very distinct, but is gave up only after five hours of courting closest to A. gemma of the Pacific coast, a female of A. paUidus. There is the widespread A. gemmoides Habits. All North American Araneus of from the central states and provinces to the diadematus group have similar life the Pacific coast. All these species build histories. There is one generation, which large webs on bams, houses, and f)orches matures in summer, mates in summer, and although their original habitat may have dies in fall after making a loose fluffy egg been chffs and entrances to caves. In sac (Plate 1). Spiderlings leave the egg addition there are A. pima in Utah and sac in spring (A. diadematus). Arizona, and A. illaudatus from Arizona to The virgin male has the embolus capped Texas. The habitats of the last two are (Levi, in press). The function of the cap uncertain, although they probably are also is not blown, but it is not believed to found on buildings. transmit sperm although a duct is at times Araneus gemma McCook was split by visible. Males are known to mate several Chamberlin and Ivie into three species: times and to survive mating. It has been A. gemma, A. gemmoides, and A. pirus. suggested that mating before sperm in- With few specimens on hand the judgment duction, as observed in some spiders, may was sound. However, any series shows that remove the cap and permit the embolus to A. gemma is quite variable, as is A. pirus; function (Kullmann, van Helsdingen, per- furthermore, at times A. gemma is collected sonal communication ) . The cap is diag- with A. gemmoides. Some females are nostically different in each species. About intermediate and cannot be assigned to half the males in collections are virgin and either species (Figs. 215-217). Populations still have the cap, half have mated and of A. gemmoides are relatively uniform ha\'e lost it. Virgin males wandering in except for coloration in the area of overlap search of a mate may be more likely to be with A. gemma. Apparently the two spe- collected. cies hybridize and introgression is taking In all collections examined, only one place with gene flow into the A. gemma male was marked as having been killed and population. It is known that males may tiy eaten by a female A. diadematus. The male to mate with females of the wrong species, when examined turned out to be an A. Broken-off embolus caps are at times found cavaticus, not the partner of the female, on the epigynum of a different species. It However, Grasshoff (1964) reports that is especially common to find large A. illau- the female of Araneus paUidus has to bite dafus tips in the epigynum of A. pima. into the male's al^domen to permit holding Apparently the cross of A. gemmoides X on during copulation. In one case the A. gemma is fertile. This would produce a female, feeding on an insect, was prevented potentially interesting study. It is not fol- from biting; the male could not hold on lowed up here. to the female, slid off, and was wrapped It is not uncommon to collect females in silk like prey, with the wrong males. This happened to All species make an almost vertical orb as careful a naturalist as Emerton. Grass- web \\ith about LS-30 radii, with a retreat hoff (1964) reports that males of Araneus above to the side of the web in leaves, diadeinatus will court females of A. palli- bark, or lichens and connected with the Araneus diadematus group Orb-\veavers • Levi 137 Plate 1. Egg sac of captive Araneus pimo sp. n. from Arizona hub by a signal line (Plate 2). During the day the spider rests in the retreat, at night usually in the center. The center has an irregular mesh. Araneus diadematus is more likely than other species to be found in the center of the web at daytime. Al- though all species mature at the same time, their variation in size is reflected in the diameter of the orb and the size of the prey handled. Also each species seems to have its own habitat, some preferring forests (A. saeviis, A. nordmanni, and the small A. corticarius), some meadows (A. trifolium, A. marmoreus), city gardens (A. diadematus in North America), build- ings or cliffs ( A. cavaticus group ) . Araneus hicentenarius, often brightly colored, makes its retreat among lichens (Plate 3), the color of which it matches. Species living in the open in herbaceous vegetation tend to have an oval abdomen (A. marmoreus, A. trijolium), while those living on trees and cliffs have humps (A. nordmanni, A. saevus, A. cavaticus). The adaptation of this curious correlation is not known, but is also believed true of 138 Bulletin Museum of Comparative Zoology, Vol. 141, No. 4 im^^n Plate 2. (Top) Web of Aroneus covaf/cus (Keyserling) powdered with corn starch, West Virginia. (Photo by W. A. Shear. (Bottom) Web of A. gemmoides Chamberlin and Ivie with adult female, Wisconsin. Araneus diadematus group Orb-weavers • Levi 139 Plate 3. Araneus bicentenarius (McCook) in retreat among lichens. West Virginia. Both spider and the lichens are the same shade of green. (From Ektachrome photo by W. A. Shear.) European species. It has been suggested (W. S. Shear, personal communication) that only those that have a retreat in curled up leaves lack humps. Western A. gemmoides and some A. gemma have black bars on the venter, which are distinct "eyes" in A. pima (Plate 4). I assume that when disturbed the spider zooms down its signal thread to appear at the center of the orb web. In South America I watched an araneid make maximum use of its ventral spots, which in the particular individual observed were in the shape of a face. The habitat observations reported here are gleaned from the labels on vials. Only mature individuals were used. Most check- lists and literature on habits are unreliable because the species determinations are in doubt and often wrong. The specimens usually have not been turned over to a museum for safe keeping as voucher specimens. In mapping distributions, only specimens examined have been used. Key to Females la. Abdomen with humps or angular anteriorly (Figs. 21, 37, 118) 6 lb. Abdomen oval to spherical, without humps or anterior angles (Figs. 144, 177) 2 2a. Posterior lamellae of epigynum showing on each side in ventral view as large curved folds; scape overhanging a depression bordered on each side (Figs. 1, 107) marmoreus 2b. Posterior lamellae not visible in ventral view or, if visible, scape not overhanging a depression bordered on each side 3 3a. Scape of epigynum with more or less parallel sides, very rarely broken off (Figs. 138, 174) 4 3b. Scape of epigynum tapered toward tip, or often broken off (Figs. 159, 167) 5 4a. Legs banded; a bordered depression on each side of epigynal scape in ventral view (Fig. 174); no median keel in pos- terior view (Fig. 176) trifolium 4b. Legs not banded; base of epigynum a domed area truncate on the posterior side; openings posterior and separated by a keeled septum (Figs. 140, 142) iviei 5a. Scape of epigynum widest at base; on each side of scape a shallow depression with a narrow rim (Fig. 167) ijukon 5b. Scape, if present, widest close to its mid- dle; scape hiding ventral depression (Fig. 159) and if scape is broken, rims wide ( Fig. 162 ) groenlandicohis 6a. Abdominal humps projecting toward sides; usually a transverse line between humps anterior of which abdomen is usually dark, posterior hght (Fig. 118) corticarius 6b. Humps dorsal, transverse markings not as contrasting 7 7a. Scape long, at least twice as long as width of base — 8 7b. Scape short 12 8a. Scape coiled (Fig. 52); southwest santarita 8b. Most of scape straight (Figs. 45, 48) 9 9a. In posterior view of epigynum there is a median hght groove (Figs. 17, 20, 29) .__.10 9b. No such groove in posterior view — -11 10a. Venter of groove (toward scape) dark and swollen, epigynal lamellae on each side large (Fig. 29); Pacific coast andrewsi 10b. Venter of groo\e not swollen toward scape, lamellae small (Figs. 17, 20); 140 Bulletin Museum of Comparative Zoology, Vol. 141, No. 4 Plate 4. Araneus pima sp. n. showing eye spots on venter of abdomen. Canada, eastern United States to South- west bicentcnarius 11a. Spider dark, abdomen very dark, black, usually with a white lanceolate cardiac mark on abdomen (Fig. 51); forests; epigynum as in Figs. 42, 44, 45, 47, 48, 50 saeviis lib. Spider light, abdomen almost always with light dorsal marks in shape of a cross (Fig. 37); city gardens; epigynum as in Figs. 34, 36 diadcmatus 12a. Epigynal scape with transverse annuli separated by grooves ( Figs. 123, 124, 187, 188) _. ,13 12b. Epigynal scape lacking transverse annuli (Figs. 207, 224, 233) 14 13a. Scape twisted (Figs. 123, 124); base on each side of scape with a diagonal rim enclosing the median depression (Fig. 123); Newfoundland to New Hampshire — - — — loashingtoni 13b. Scape straight (Figs. 187, 188); rims of depression hidden by scape (Fig. 187) ____15 14a. Width of scape more than half width of epigynum (Fig. 187); epigynum in pos- terior view with median triangular sclerite (Fig. 189); Eastern United States and Canada, on overhanging cliffs, common on buildings cavaticus 14b. Width of scape less than half base of epigynum (Figs. 87, 90); epigynum in posterior view with heart-shaped depres- sion (Figs. 89, 92); widespread in forests nordmanni 15a. Anterior end of epigynal scape constricted (Fig. 233); Texas and southwest . illaudatus 15b. Anterior end of scape is wide, scape tap- ered toward tip 16 16a. Epigynum and scape very small, scape triangular (Fig. 195); widespread in cen- tral Canada and United States to Pacific coast gemmoides 16b. Epigynum and scape otherwise (Figs. 203, 205, 218); Pacific coast and Southwest , ..17 17a. Scape without lateral rims (Figs. 218, 221); a deep transverse groove in posterior view (Figs. 220, 223); California to Texas pima 17b. Scape witli lateral rims, very variable (Figs. 203, 205); transverse groove very narrow if present (Figs. 204, 206); Pacific coast gemma Key to Males la. First coxa with a hook on distal margin facing second coxa (Fig. 183) 2 lb. First coxa without hook on distal margin ._ 9 2a. Conductor of palpus with teeth around margin (Figs. 23, 24, 31); median Aranevs diadematvs group Orb-weavers • Levi 141 o "D Q- "G If n c o Q- ^2 X o LU **- 3 !=! F ■D O c (1) > O — ' E ^ D D o CN O -o o ^ Q- c LO c (U (U > "5 o ' to dl c o E s o ■*- D "D CN D C t/T o c E 3 ^ D 3 OJ Q. C D X >. o UJ D) E W) 'q. o 1^ LU E o ^ _■ *yi E 3 . — 3 *- > 1 a> 0 a> o- c D E O Co < E 'o 'o "o £ a) D 3 D (^ lyi C^ -=t 0 o u CO 1 1 CO "^ rJ, °^ > \„ O) O) O) 142 Bulletin Museum of Comparative Zoology, Vol. 141, No. 4 apophysis with at least 2 distal spines (Figs. 23, 32) 3 2h. Conductor of palpus smooth around mar- gin; median apophysis with a keel on distal edge (Figs. 4, 110) 4 3a. Terminal apophysis with drawn-out tip (Figs. 31, 32); teeth around edge of conductor suhequal (Fig. 31); Pacific coast - andrewsi 31). Tip of terminal apophysis an acute tri- angle (Figs. 23, 24); first tooth on em- bolus side of conductor noticeably larger than others (Figs. 23, 24); Canada, eastern U. S. to Southwest hicentenurius 4a. Second co.xa with spur (Fig. 183) 6 4b. Second coxa without spur 5 5a. Palpal conductor unusually large (Fig. 120); in mesal view median apophysis about 3 times as long as wide (Fig. 119); Pennsylvania to Alaska corticarius 5b. Palpal conductor of normal size (Fig. 128); median apophysis about 2 times as long as wide (Fig. 127); Newfoundland to New Hampshire - washingtoni 6a. Terminal apophysis paralleled b\' a long subterminal apophysis (Figs. 4, 96-99, 103, 104, 110) 7 6b. Subterminal apophysis a short stub 8 7a. Embolus sclerotized, subcircular at end (Figs. 4, 103, 104, 110); holarctic _ marmoreus 7b. Embolus longer than wide as in Figs. 69- 75, 96; holarctic nordmanni 8a. Embolus a curved hook (Figs. 38, 40, 41); apical apophysis short, curved, sickle- shaped (Figs. 38, 39) diadematm 8b. Embolus otherwise (Figs. 55, 57, 58); apical apophysis a long, strongly curved prong (Figs. 55, 56) sacvtis 9a. Inner lamella of embolus a wide projecting plate of larger visible area than narrow conductor (Fig. 179); embolus with distal notch (Figs. 178, 180, 181) trifolium 91). Inner lamella of embolus, if projecting, with visible part always much smaller in area than conductor (Fig. 172); embolus otherwise 10 10a. Second tibiae modified by being swollen or curved (Fig. 184) 11 10b. Second tibiae straight, not swollen 12 11a. Median apophysis with slender, graceful proximal spine and distal projections with strong, straight spines, the lower one with a dentate edge (Figs. 171, 172) yukon lib. Median apophysis witli proximal spine stout and distal projections both bent out at right angles to main axis of median apophysis (Figs. 164, 165) .. groenlandicoliis 12a. Terminal apophysis a sclerotized prong, widened just before tip (Figs. 145-148); conductor with a pocket on lateral side (Figs. 146, 148); embolus as in Figs. 145, 149-151 iviei 12b. Terminal apophysis lightly sclerotized, fleshy, never widened near tip; conductor and embolus of different shape 13 13a. Males more than 10 mm total length; proximal spine of median apophysis placed in proximal half of sclerite; distal edge a keel (Figs. 191, 192); eastern United States, Canada cavatictis 13b. Males less than 9 mm in total length; proximal spine of median apophysis placed close to distal spine (Figs. 211, 212, 238, 239); central, western and southwestern region ____ 1 4 14a. Palpal tibia almost equal in size to bulb (Figs. 199, 200); spines of median apophysis slender; terminal apophysis truncate ( Figs. 199, 200 ) ; widespread in central and western North America — fienimoides 14b. Palpal tibia less than half area of bulb; spines of median apophysis stout; terminal apophysis more pointed; Pacific states, southwest to Texas 15 15a. Embolus pointed (Figs. 211, 214); em- bolus cap short (Fig. 213) gciutna 15b. Embolus truncate ( Fig. 232 ) ; embolus cap elongate, pointed (Figs. 231, 240); Texas to California - 16 16a. Total length 5-11 mm; spines of median apophysis stout (Figs. 228, 229) pima 16b. Total length less than 4 mm; spines of median apophysis slender, recurved ( Figs. 238, 239 ) illamlaius Araneus angulatus Clerck Figures 9—14 Araneus angulatus Clerck, 1757, Svenska Spindlar, p. 22, pi. 1, figs. 1-3, 9. Type specimens in the Natural History Museum, Stockliolm, lost. Locket and Millidge, 1953, British Spiders, 2: 127, figs. 80, 83, $, $. Bonnet, 1955, Biblio- graphia Araneorum, 2: 433 (in part). Aranea angulata, — Wiehle, 1931, in Dahl, Die Tierwelt Deutschlands, 23: 52, figs. 4, 5, $, $. Roewer, 1942, Katalog dcr Araneae, 1: 795 (in part ) . Note. Although there are many Hterature records of this species m North America, all refer to large specimens of various other species (A. anclrew.si; A. hicentenarius; A. saemis; A. marmoreus; A. nordmanni) that had been misidentified. No specimens of Araneus angulatus coming from North America have been found in collections. Araneus diadematus group Orb-weavers • Levi 143 • H O Araneus andrewsi • Araneus bicentenorius \, Map 1. Distribution of Araneus bicenfenorius (McCook) and A. andrewsi (Archer) The differences between A. angulatus and Noith American species are discussed un- der A. bicentenarius. The web, described by Wiehle (1931) has bridge threads up to 5 m long at 3 ni height. The number of spokes averages 21 and the snare region is 40-63 cm across. Araneus bicentenarius (McCook) Plate 3; Figures 15-26; Map 1 Epeira gigas Leach, 1815, Zoological Miscellany, 2: 132, pi. 109. Female type probably from America in tlie British Museum, Natural History, examined. ( Not A. gigas, — Comstock, 1912, 1940, The Spider Book and some other authors^. ) Epeira bicentennaria McCook, 1888, Proc. Acad. Natur. Sci., Philadelphia, p. 195, figs. 3, 5, $. Syntypes from "northwestern Ohio and Al- legheny mountains" lost. Epeira angulata var. bicentenaria, — McCook, 1893, American Spiders, 3: 186, pi. 10, figs. 3-5, pi. 11, figs. 2-4, 5, S. Aranea bicentenaria, — Archer, 1951, Amer. Mus. Novitates, 1487: 31, figs. 68, 78, ?, $. Aranea kisatchia Archer, 1951, Amer. Mus. ^ The International Commission on Zoological Nomenclature will be asked to place the name Epeira gigas Leach on the Official List of Re- jected Names in Zoology and the name Epeira bicentenaria McCook on the Official List of Spe- cific Names in Zoology. Novitates, 1487: 27, fig. 69, 9. Female holo- type from Grant Parish, Louisiana, in the American Museum of Natural History, ex- amined. NEW SYNONYMY. 'Note. Leach's description suggested that his Epeira gigas is A. hicentenarius. The type of E. gigas was found pinned and stuffed with cotton in the British Museum by Mr. D. J. Clark. After placing it in alcohol Mr. D. J. Clark examined the speci- men and reported it to be an unusually large specimen of A. angulatus; upon com- paring it to my drawings he found it to match Figures 18, 19 of a southern A. bicentenarius. I have since examined it myself. The name A. gigas has been used by some authors for A. marmoreiis, never before for A. Iricentenarius. A junior homonym, Epeira gigas C. L. Koch, 1830, has been renamed A. grossus (C. L. Koch) and belongs to a species also closely related to A. angulatus. The spider was collected by McCook in 1882 at the bicentennial of the city of Philadelphia. By misprint the name was spelled with two n's in the first publication. Description. Female from New Jersey. Carapace dark brown. Sternum dark brown with a lighter, branched mid-longitudinal band. Legs mottled brown with darker 144 Bulletin Museum of Comparative Zoology, Vol. 141, No. 4 bands. Abdomen very dark with folimn. Venter dark brown to black, sometimes with a median hght area. Abdomen with large hnmps. The dorsum has small sclerotized spots, the bases of setae, but all setae are short. Total length, 15 mm. Carapace, 5.9 mm long, 5.0 mm wide. First femur, 6.S mm; patella and tibia, 9.2 mm; metatarsus, 5.4 mm; tarsus, 1.8 mm. Second patella and tibia, 8.4 mm; third, 5.0 mm; fourth, 7.6 mm. Male from New Hampshire. Coloration as in female except for an anterior, dia- mond-shaped, longitudinal white mark on dorsum of abdomen. The first coxa has a hook on the distal margin, the second a spur. The second tibia is very strong with strong macrosetae. Total length, 7 mm. Carapace, 6.5 mm long, 4.6 mm wide. First femur, 6.9 mm; patella and tibia, 8.4 mm; metatarsus, 4.9 mm; tarsus, 1.7 mm. Second patella and tibia, 8.4 mm; third, 5.0 mm; fourth, 6.7 mm. Variation. Individuals of this rare species differ in size and coloration. The largest specimens come from the southern United States. Females are up to 28 mm in total length; one measuring 21 mm in total length had the carapace 10.0 mm long, 8.6 mm wide. The smallest specimen comes from Canada; the smallest female measured 13 mm in total length, carapace 6.1 mm long, 5.6 mm wide. Many northern and southern specimens are strikingly colored on the abdomen (Plate 3, Fig. 22), others are just shades of brown in alcohol (Fig. 21 ) . The scape of the epigynum is variable in length, and the median depressed area in posterior view is of variable width, narrowest in some southern specimens (Fig. 20) called A. kisatchiiis by Archer. However, intermediate epigyna are com- mon. Diagnosis. Females of A. bicentenarius and A. andreicsi differ from A. diadematus and A. saevus, both of which also have a long epigynal scape, by having the median area of the epigynum in posterior view depressed and hght (Figs. 17, 20, 29). Males of A. bicentenarius and A. andrewsi differ from males of other groups by hav- ing the margin of the conductor toothed (Figs. 23, 24, 31, and easily seen in apical view) and by the shape of the embolus, the tip of which can only be seen in ventral view (Figs. 26, 33) after removal of the conductor. The related Eurasian A. angulatus has the median, posterior area of the epigynum swollen and sclerotized, rather than de- pressed (Fig. 11), and the embolus and terminal apophysis of the palpus (Figs. 12-14) differ in shape from those of the two North American species. In females of the related Pacific coast A. andrewsi, the median posterior de- pressed area of the epigynum is distally (ventrally) black and swollen (Fig. 29); that of A. bicentenarius is not swollen and is light in color (Figs. 17, 20). Males of A. andrewsi have the tip of the terminal apophysis sclerotized and pointed (Figs. 31, 32), while that of A. bicentenarius is an acute fleshy lobe (Figs. 23, 24). The embolus can be seen under the terminal apophysis inside the palpus. The embolus differs slightly in shape in different speci- mens; its opening is ventral and is partly hidden b)' the conductor. Note. In collections and literature A. bicentenarius has often been called A. angulatus, as have large specimens of var- ious North American Araneus species. Habits. Araneus bicentenarius is found on trees in woods. One brightly marked specimen was collected by a botanist among lichens on jack pine {Pinus banksi- ana). The dorsal pattern makes the spider disappear among Evernia mesomorpha, Parmelia aundenta, and P. caperata. (L. L. Darrow, personal communication.) An- other was green when collected, and "blended perfectly with lichens, etc." on a maple tree in West Virginia. (Plate 3) (W. A. Shear, personal communication). This last specimen was brown in alcohol and did not have contrasting markings. Most collections of this rare species consist Araneus diadematus group Orb-weavers • Levi 145 o O > ^ (1) ^ > oo O — J2 . E U-) LU NO o E ^• O o ^? < "2 Si 4) CN lA « E CN E o D o E o a- en o > n (U LU :^ CN o CN 1 CO CN E 3 U"> c >^ Q. D5 Q. LU o O Q. E F o u 0) CN CN to —1 1 > CN C (U tn CN F 3 -D E F O c: ir o n D E m J. E ri > r— CN ^ *A vt D 0) Q V o 3 3 .n OO CN cn E U- CN u_ UJ 146 Bulletin Museum of Comparative Zoology, Vol. 141, No. 4 • Araneus saevus O Araneus santarita Ma p. 2. Distribution of Araneus diadematus Clerck, Araneus saevus (L. Koch) and Araneus santarita (Archer) of single specimens, very few of which are males. Distribution. Eastern North America from Nova Scotia, Minnesota to northern Florida and Texas (Map 1); one male came from Coyote, New Mexico. Araneus andrewsi (Archer) Figures 27-33; Map 1 Aranea andrewsi Archer, 1951, Amer. Mus. Novi- tates, 1487: 31, figs. 63, 64, 82, $, $. Male holotype from Claremont, California, in the American Museum of Natural History. Description. Female. Carapace brown, darker on sides. Stcrnnm mottled brown. Legs mottled dark brown, indistinctly banded. Dorsum of abdomen brownish black, mottled, with a folium. Ventrally, area between epigynum and spinnerets dark gray. Carapace hairy. Abdomen with two large humps, very hairy, and with many very small sclcrotized plates, the bases of setae. Total length, 14 mm. Carapace, 6.7 mm long, 5.9 mm wide. First femur, 6.3 mm; patella and tibia, 9.2 Araneus diadematus group Orb-weavers • Levi 147 mm; metatarsus, 5.1 mm; tarsus, 1.8 mm. Second patella and tibia, 8.4 mm; third, 5.1 mm; fourth, 7.6 mm. Male. Coloration like that of female; very dark and hairy. Abdomen with dis- tinct humps. First coxa with a hook on distal margin, second with a large cone. Second tibia strong and bent, armed with macrosetae. Total length, 11 mm. Cara- pace, 5.2 mm long, 4.4 mm wide. First femur, 5.9 mm; patella and tibia, 8.4 mm; metatarsus, 4.6 mm; tarsus, 1.7 mm. Second patella and tibia, 7.6 mm; third, 4.2 mm; fourth, 5.9 mm. Variation. Many individuals are almost black with the folium outline barely visible. Females measured 11-22 mm in total length, with carapace 5.0-7.8 mm long, 4.8-6.1 mm wide. Males measured 8-11 mm in total length; the smallest had the carapace 4.7 mm long, 3.6 mm wide. Diafi,nosis. This Pacific coast species cannot be confused with any others in California. It differs from the related eastern North American A. hicentenarius in details of the genitalia (see under A. hicentenarius) . Hahits. Specimens have been collected in a house basement. Another was in curled-up bark of a eucalyptus tree, with the web at chest height, strung toward a fence below the spider. Other specimens were collected on tree bark, and on the trunk of a walnut tree {Jiiglans calif ornica) . Distribution. Araneus andrewsi is found from Oregon to southern California along the coast (Map 1). Araneus diadematus Clerck, Cross Spider' Figures 34-41, 95, 184^186; Map 2 Araneus diademaius Clerck, 1757, Svenska Spind- lar, p. 25, pi. 1, fig. 4. The type specimens in the Natural History Museum, Stockhohn, lost. Locket and Millidge, 1953, British Spiders, 2: 127, figs. 84a, 85a, 86a, 87a, 9, i. Bonnet, 1955, Bibliographia Araneorum, 2: 486. Grass- hoff, 1968, Abhandl. Senckenbergischen Natur- forsch. Ges., 516: 1-100. Epeira diademata, — Wiehle, 1927, Z. Morphol. ^ Garden Spider in Great Britain. Okol. Tiere, 8: 492; Nielsen, 1932, Biology of Spiders, Copenhagen, Vol. 2: 421. Aranea diadema, — Wiehle, 1931, in Dahl, Die Tierwelt Deutschlands, 23: 70, figs. 103-108, 2 , $ . Roewer, 1942, Katalog der Araneae, 1 : 797. Kaston, 1948, Bull. Connecticut Geol. Natur. Hist. Surv., 70: 249, figs. 779-782, 9, $. Description. Female. Carapace yellow- brown, sternum dark brown, coxae light brown. Legs yellow-brown banded with darker brown. Dorsum of abdomen with white marks anteriorly in form of a cross (Fig. 37), a folium posteriorly. Venter with a median black band and a pair of white spots closer to spinnerets than to epigastric groove. Dorsum of abdomen often with two humps. Female from Mas- sachusetts measured total length, 13 mm. Carapace, 4.6 mm long, 3.9 mm wide. First femur, 5.0 mm; patella and tibia, 6.7 mm; metatarsus, 4.5 mm; tarsus, 1.7 mm. Second patella and tibia, 5.9 mm; third, 3.5 mm; fourth, 5.5 mm. Male. Banding of legs less distinct than in female and color generally darker. Abdomen with a distinct white cross, and folium more distinct than in female. The abdomen lacks humps. Second tibia modi- fied by being stronger and having veiy strong short spines. A specimen from Massachusetts measured 8 mm in total length. Carapace, 3.8 mm long, 3.4 mm wide. First femur, 5.2 mm; patella and tibia, 7.1 mm; metatarsus, 4.7 mm; tarsus, 1.5 mm. Second patella and tibia, 5.7 mm; third, 3.4 mm; fourth, 5.2 mm. Variation. Females vary 6.5-20 mm in total length, carapace 3.5-7.2 mm long, 3.2-6.1 mm wide. Total length of males, 5.7-13 mm. The shape of the abdomen and size of humps varies. The epigynal scape may vary slightly in length and the posterior triangular sclerite (Fig. 36) varies in shape and is often much less pointed behind. There are also small differences in the embolus cun^ature. This species and its variation were recently studied by Grass- hoff (1968). 148 Bulletin Museum of Comparative Zoology, Vol. 141, No. 4 Diagnosis. Almost all females and many males can readily be recognized by the white marks arranged in the form of a cross on the dorsum of the abdomen (Fig. 37). No other species in North America has the cross mark. In addition, females are distinguished by the posterior face of the epigynum, which has a median tri- angular sclerite with the straight edge out, pointed end in (dorsally) (Fig. 36). The homologous sclerite in A. saeviis has the outer flat edge with a median notch (Figs. 44, 47, 50) and the sclerite heart- shaped; the notch leads to a depression under the scape (Fig. 45) not present in A. diadematiis (Fig. 34). Araneus diade- matus has the openings usually distinctly visible in ventral view anterior of the tri- angular plate (Fig. 34). Males are separated from all other Araneus species by the recurved embolus, triangular and pointed at its tip (Figs. 38, 40, 41), from A. saevus by the short, slightly curved terminal apophysis, and by the shape of the conductor (Figs. 38, 39). Habits. In Europe the species is wide- spread and common in woods, gardens, and fields. In eastern North America, A. diadematus seems to do well only in cities in shrubs between houses, a habitat not to the liking of any native species of Araneus. In the west records are "from floor of food store," "web in woodpile," and many in cities. It may take two years for the animal to mature in Europe (Locket and Millidge, 1953). The female is more likely to remain in the center of the web at daytime than other American species. The species has been used for experimental research. The web has about 30 spokes and is made at up to 1.5 m height. The web has been illustrated bv Wiehle, 1927, and Nielsen, 1932; the egg sac by Nielsen. Distribution. This Eurasian species ap- pears introduced, judging by its limited distribution in North America and its pre- ferred habitat of shrubs surrounding city houses. It survives only in a narrow belt from Newfoundland south to Rhode Island, west to Vancouver and Oregon, more abundant on the milder coasts than in the Great Plains and mountain states (Map 2). Samples of records are St. Johns, New- foundland; Quebec; Toronto; Vancouver; Victoria; Boston; Newport, Rhode Island; Syracuse; Ithaca; Rochester; Detroit; East Lansing; Seattle; Portland, and some smaller towns. The oldest American col- lections are about 100 years old and come from the northeast. The species does occur in Siberia and Japan. Specimens from the USSR, 64 km SSW of Irkutsk on Lake Baikal, were examined; they differed more from European ones than do North Ameri- can specimens, presumably due to geo- graphic variation. Araneus saevus (L. Koch) Figures 7-8, 42-51, 55-60; Map 2 Epeira saeva L. Koch, 1872, Z. Ferdinandeum Tirol Vorarlberg, (3) 17: 323. Male holotype specimens from Bad Ratzes, Austria [aliove Siiisi, Trentino Alto Adige, Italy], in the British Museum, Natiual History, examined. Epeira solitaria Emerton, 1884, Trans. Connecti- cut Acad. Sci., 9(6): 299, pi. 33, fig. 11, pi. 35, fig. 3, $ . Male holotype from Peabody, Mas- sachusetts, in the Museum of Comparative Zo- ology, examined. Epeira silvatica, — Emerton, 1884, Trans. Con- necticut Acad. Sci., 9(6): 300 (in part), pi. 35, figs. 1-6, ? . Female paralectotype here desig- nated, not lectotype. Epeira nigra Emerton, 1894, Trans. Connecticut Acad. Sci., 14(3): 402, pi. 1, fig. 1, 2, $■ Male and female syntypes from Laggan, Can- ada ["5000-8500' in Rocky Mountains near Canadian Pacific Railway" a station near present-day Lake Louise], in the Museum of Comparative Zoology, examined. Aranea saeva, — Roewer, 1942, Katalog der Araneae, 1: 791. Aranea solitaria, — Roewer, 1942, Katalog der Araneae, 1: 863. Kaston, 1948, Bull. Connecti- cut Geol. Natur. Hist. Surv., 70: 250, figs. 785-786, 796-797, 9, $. Araneus saevus, — Tullgien, 1952, Entomol. Tidskr., 73: 164, figs. 10, 12, $, i. Bonnet, 1955, Bibliographia Araneorum, 2: 588. Wiehle, 1963, Zool. Jahrb. Abt. System., 90: 276, figs. 84-92, 2, i. Description. Female from near Lake Louise, Alberta. Carapace, sternum and ArANEUS DIADEMATUS GROUP OrB-%VEA\'ERS • LcVt 149 o E o O > I U P ^ . g-E o O ^ ^ CN o >- o z o -a o o u I o c >^ en Q. o u o 3 o .— u^ l-v O NO S Q. D "a uS u c ^N. j: o _N ^ < o c o NO NO c o S un 3 NO "5 XI N^" • I\ E UJ OJ o 3 vO -C 1 o s s c N» lU > NO CO Z NO NO CN Ln "o^ rNv "5 O o c o N O • — U < CN NO R 3 ^ D. , D NO "o" Q. -D — 3 O E D c c 0 3 o -Q E E JD 3 c ■o o c E UJ e" E D < <. 1 On d NO t/i o uS o NO t\ 1 1 n D LJ_ -^ . — . 152 Bulletin Museum of Comparative Zoology, Vol. 141, No. 4 Araneus nordmanni Map 3. Distribution of Araneus nordmanni (Thorel Aranea nordmanni, — Wlehle, 1931, in Dahl, Tier- welt Deutschlands, 23: 58, figs. 84, 85, $. Roewcr, 1942, Katalog cler Araneae, 1: 801. Kaston, 1948, Bull. Connecticut Geol. Natur. Hist. Surv., 70: 250, figs. 783-784, 793-795, 9, 6. Aranea darlingtoni Archer, 1951, Amer. Mus. Novitates, 1487: 25, figs. 71, 75, 5, $. Female holotype from Durbin, West Virginia, in the American Museum of Natural History; para- types examined. NEW SYNONYMY. Aranea pseudomelaena Archer, 1951, Amer. Mus. Novitates, 1487: 26, figs. 70, 79, 5, $. Male holotype from Long's Peak, Colorado, in the American Museum of Natural History, ex- amined. NEW SYNONYMY. Araneus nordmanni, — Bonnet, 1955, Bibliographia Araneorum, 2: 553. Wiehle, 1963, Zoo). Jalirl). Abt. System., 90: 271, figs. 74-83, 9,6- Note. Archer named American speci- mens A. darlinii,toni because he thought that European ones lack the basal spur on coxa II. Wiehle (1963) illustrated the spur, and all European males examined had it. Also Archer indicates that "on the endal side [of the median apophysis] below the spur is a rounded angle, a feature en- tirely missing in A. norchiumni." As dis- cussed in the introduction, the median apophysis by itself is not a good character to separate species. Araneus pseudomelaena Archer was thought to be a valid entity, but extremely variable. As more collections were ex- amined it was found that those specimens Araneus diadematus group Orb-weavers • Levi 153 ci^ fV o oo c 1 o oo 5 • ro D Os SI F c zt (U o E u 0 , n .■^ o m (U o ^ i- a> (i> 1 Ll_ t/> oo -^ Ov c CO O cs TO Q) n o 0> ^ 0 cn 0" 0 'q. CO -C LU Q. hC 4) C-i t\ u CN 1 "0 VO "5 0 c c ^ 0) OJ — > > o rj vt t--. 3 0 E "5" E 0 E 0 CN u '^ 0 ^ -c 0 t-v 0 CN c (U z Araneus marmoreus ■> Map 4. Distribution of Araneus mormoreus Clerck. patella and tibia, 6.4 mm; third, 4.0 mm; fourth, 6.0 mm. Male from Michigan. Coloration like that of female with legs more distinctly banded. The first coxa has a hook on the distal ventral rim, the second coxa a spur. The second tibia is swollen and bears macrosetae. Total length of a specimen from Michigan, 8.4 mm. Carapace, 5.0 mm long, 3.7 mm wide. First femur, 5.0 mm; patella and tibia, 6.7 mm; metatarsus, 3.6 mm; tarsus, 1.2 mm. Second patella and tibia, 5.9 mm; third, 3.5 mm; fourth, 5.0 mm. Variation. Total length of females, 9-18 mm; carapace 2.7-5.2 mm long, 2.3-4.5 mm wide. Total length of males, 5.9 mm; Carapace 2.9-4.3 mm long, 2.3-3.6 mm wide. There is considerable variation in color and pattern. One female from Texas had black bands on its legs. The abdomen has a black folium in many females collected in Alberta, Montana, Wyoming, Europe, and isolated other places (Figs. 101, 102); in females from other areas the folium en- closes symmetrical light patches (Fig. 100), or sometimes it has a reticulated pattern. The shape of the terminal apophysis and the length of the subtcrminal apophysis vary from specimen to specimen ( Figs. 103, 104). Although the median apophysis of A. marmoreus is just as variable as that of other Araneus species (Grasshoff, 1968, 158 Bulletin Museum of Comparotive Zoology, Vol. 141, No. 4 figs. 36e-h), A. marmoreus has escaped being split into several species. Diagnosis. The common A. marmoreus differs from related A. corticarius and most other North American Araneus species by having an oval, subspherical abdomen. The female can be recognized by the large basal lamellae of the epigynum which almost always extend on each side beyond the base and are visible in ventral view (Figs. 1-3, 107-109). They also extend at times in A. iviei. Males, despite the distinct palpus, are often misidentified. The terminal apophysis is sclerotized and along its proximal side is a translucent edge. The terminal apophysis is paralleled by a subterminal apophysis (Figs. 4, 6, 103, 104, 110). In A. nordmanni and A. saevus the subterminal apophysis is shorter. The embolus resem- bles that of A. corticarius, being semi- circular below the tip. The embolus lamella, if it shows at all, is in the shape of a flat hook (Figs. 4-6, 111), not a piece with parallel sides as in A. trifolium (Fig. 179). The conductor (Figs. 5, 111) is much wider than that of A. trifolium (Fig. 179) and has a clearly set off, narrower tongue facing the median apophysis. Habits. Araneus marmoreus prefers tall meadows, and places its webs in grasses or low shi"ubs, sometimes on trees. In the West it has been collected in light lodge- pole pine ( Pinus contorta ) forest, its retreat under bark, in a relatively humid place. Kaston (1948) reports the egg sac to be "13 mm in diameter, a flattened sphere of loose white silk. There were 653 orange, agglutinated eggs, each about 1.15 mm in diameter." The web has been illustrated by Comstock (1912, 1940) and Kaston ( 1948 ) , the egg sac by Comstock. The web has 24^29 spokes and is made 50-90 cm high in grass. Distribution. Araneus marmoreus is hol- arctic in distribution. The borders of its Eurasian distribution have not been criti- cally mapped. The many literature citations of misidentified specimens will give a mistaken idea of the distribution. In North America A. marmoreus is found from Alaska to the southern United States ( Map 4). Araneus corticarius (Emerton) Figures 1 14-122; Map 5 Epeiia coiiicaiia Emerton, 1SS4, Trans. Con- necticut Acad. Sci., 6: 300, pi. 33, fig. 14, pi. 35, fig. 9, 9 . Two female syntypes from Beverly, Mass., in the Museum of Comparative Zoology, examined. McCook, 1893, American Spiders, p. 176, pi. 8, fig. 7, $ . Emerton, 1909, Trans. Connecticut Acad. Sci., 14: 199, pi. 5, fig. 3, S. Epeira incestifica Keyserling, 1892, Die Spinnen Amerikas, 4: 132, pi. 7, fig. 98. Female holo- type from Sitka, Alaska [?] (G. Marx), in the U. S. National Museum, examined. Aranea corticaria, — Roewer, 1942, Katalog der Araneae, 1: 860. Kaston, 1948, Bull. Connecticut State Geol. Natur. Hist. Surv., 70: 252, figs. 800-802, 9, ci. Aranea denninp.i Archer, 1951, Amer. Mus. Novi- tates, 1487: 30, fig. 81, $. Male holotype from The Pas, Manitoba, in tlie American Museum of Natural History, examined. NEW SYNONYMY. Araneus corticarius, — Bonnet, 1955, Bibliographia Araneorum, 2: 470. Description. Female from Maine. Cara- pace brown with some hairs and some irregular marks. Sternum dark brown. Coxae light. Legs light, contrastingly banded. Anterior of dorsvmi of abdomen black with a white cross (Fig. 118). Venter black with a white bracket on each side. Abdomen with large humps. Eyes large and close together. Epigynum with a folded scape (Figs. 114, 116). Total length, 6 mm. Carapace, 2.1 mm long, 1.9 mm wide. First femur, 2.5 mm; patella and tibia, 3.0 mm; metatarsus, 1.6 mm; tarsus, 0.8 mm. Second patella and tibia, 2.7 mm; third, 1.6 mm; fourth, 2.4 mm. Male from Maine. Less contrastingly colored than female. Abdomen marked like that of female but with less contrast. Abdomen has humps laterally. First coxa with hook, second without spur. The sec- ond tibia is swollen. Total length, 4.7 mm. Carapace 2.5 mm long, 1.8 mm wide. First femur, 3.0 mm; patella and tibia, 3.5 mm; Araneus diadematus group Orb-weavers • Levi 159 o^ "t; v> 3 Q. CM , □ I — ^ Q. (U 3 "5 D Q. Z s D 2 Q. o (U D - T a- 5 t/i • — O 3 n Q. c 1 O E ■- E S C -I ^ CN o 3 ■^ o 1 F K UJ ■^ o O CN n o CN ^ ■ — c E E o 5 o LL. ,— U_ ^ 11. .— _ 160 Bulletin Museum of Comparative Zoology, Vol. 141, No. 4 Araneus corticari us Araneus washing toni Map 5. Distribution of Aroneus corticari'us (Emerton) and Aroneus wos/iingfon/ sp. n. metatarsus, 1.8 mm; tarsus, 0.(S mm. Second patella and tibia, 3.0 mm; third, 1.8 mm; fourth, 2.5 mm. Variation. Total length of females varied between 5.2-8.0 mm, the carapace 2.1-2.9 mm long and 1.9-2.3 mm wide. Total length of males, 4.2-5.2 mm. The shape of the abdomen is variable. Diagnosis. The shape of the abdomen, with the humps projecting toward the sides, is not seen in any other large North American Araneus including A. washing- toni. The epigynum of the female (Figs. 114-117) resembles that of A. marmoreus, but the epigynal lamellae are smaller, only rarely visible in ventral view, and the median field behind the scape between the rims is wider than in A. marmoreus. The scape usually breaks off during mating. The embolus shape (Figs. 119, 121, 122) resembles that of A. marmoreus, but the enormous conductor, its length half that of the bulb in ventral view (Fig. 120), and the longer median apophysis (Figs. 119, 120) separate the species from A. marmoreus and A. ivashingtoni. Halrits. Kaston (1948) reports having collected A. corticarius in moist meadows and woods. It has been collected from a cranberry bog in Massachusetts, from a swamp in Wisconsin, swept from a hem- lock seedling ( Tsuga canadensis ) in a swamp in Pennsylvania. Distribution. Araneus corticarius is found from Alaska to New England. Its known range matches the range of black spiiice (Picea mariana) and tamarack {Larix lari- cina), both bog inhabitants. The southern- most records are Ice Mine, Potter Co., Pennsylvania, and Beverly Shores, Porter Co., Indiana; the northernmost, Rampart House, Yukon Terr. (Map 5). Araneus Washington} sp. n. Figures 123-130; Map 5 Holotype. Male from 3 miles up road, Mt. Washington [no date] (J. H. Emerton), in the Museum of Comparative Zoology. The species is named after George Wash- ington, as is the mountain. Description. Female. Carapace brown, sternum dark brown. Coxae light brown, Araneus diadematus group Orb-weavers • Levi 161 I CO CO CO o - _ o E O r- ID C E 2 o -a . D Ic E .y 0) 5 5 ° <» r— < S E — ^ 3 D) -7; ^ O != •i-§ ci ■? > 7 CO t .1^ ^ •— ^ 00 i/> CO 5 ^ - c: Q. tv ? > U CO lo CO o . Q. CO (D CO E o O e" E 01 r— a. ^ o o ■B i o Lo in ,- o ^ > 162 Bulletin Museum of Comparative Zoology, Vol. 141, No. 4 legs brown, indistinctly banded. Dorsnin of abdomen with a folium and a black line between humps, darker anterior of line with a white cardiac mark and a trans- verse white line (Fig. 126). Venter of abdomen black with a white bracket on each side. Each side of abdomen has a black patch which fades toward the ven- ter. The shape of the abdomen resembles that of most other species having humps; the humps are on the dorsum rather than the sides. Total length, 6.5 mm. Carapace, 3.5 mm long, 2.5 mm wide. First femur, 3.0 mm; patella and tibia, 4.0 mm; meta- tarsus, 2.3 mm; tarsus, 1.0 mm. Second patella and tibia, 3.5 mm; third, 2.2 mm; fourth, 1.0 mm. Male. The coloration is like that of the female except that the legs are banded. The first coxa has a hook, the second lacks a spur. The second tibia is swollen and has strong setae. Total length, 5.0 mm. Carapace, 2.5 mm long, 2.0 mm wide. First femur, 2.9 mm; patella and tibia, 3.5 mm; metatarsus, 2.0 mm; tarsus, 0.7 mm. Second patella and tibia, 3.0 mm; third, 1.5 mm; fourth, 2.2 mm. Dkignosis. The female differs from A. corticarius in having the humps of the abdomen dorsal rather than lateral (Fig. 126). The anterior section of the scape of the epigynum is posteriorly directed in A. icashingtoni (Figs. 123, 124); in A. corti- carixis it is directed anteriorly and then folds back (Figs. 114-116). The male dif- fers from A. corticarius by the dorsal position of the abdominal humps, by the much smaller conductor (Fig. 128), and the shorter median apophysis (Figs. 127, 128). Hahits. Emerton collected the speci- mens on low bushes. The label was written in India ink and at some later time Emer- ton pencilled "corticaria" behind the inked Epeira. The female from New Brunswick was collected on balsam fir {Abies hal- samea ) . Distribution. Newfoundland to New Hampshire (Map 5). Records. Newfoundland: Gambo, 25. IV. 1949, 9 (E. Palmen); Eddies Cove, west St. John Bay, 30. VII. 1949, 9 (E. Palmen). Neiv Brunswick: Green River Forestry Sta- tion, 22 July 1965, 9; 18 July 1967, 2^ (T. R. Renault). Neto Hampshire: 29, 2<5 paratypes collected \\\\\\ the holotype. Aroneus alsine (Walckenaer) Figures 131-137 Aranca alsine Walckenaer, 1802, Faune Paiisienne, 2: 193. Type for the Paris, France, vicinity, lost. Wiehle, 1931, in Dahl, Die Tierwelt Deiitschlands, 23: 83, figs. 119-123, 9, $. Roewer, 1942, Katalog der Araneae, 1: 781. Epeira alsine, — Wiehle, 1927, Z. Morphol. Okol. Tiere, 8: 493. Araneus alsine, — Locket and Millidge, 1953, Brit- ish Spiders, 2: 133, figs. 85d, 86d, 9, <5 . Bonnet, 1955, Bibliographia Araneonim, 2: 429. This Eurasian species, distributed from Europe to Kamchatka, has not been found in North America. It is very similar to A. iviei of North America. Araneus alsine is found on grasses of moist woods or clear- ings, and the hub of the web is about 15-20 cm above the ground. The rolled-up leaf or leaves which are used as a retreat are pulled up from the ground (Wiehle, 1927). A web with 20 spokes was illustrated by Wiehle (1931). Araneus iviei (Archer) Figures 138-151; Map 6 Aranea iviei Archer, 1951, Amer. Mus. Novitates, 1487: 33, fig. 53, 9. Female holotype from Keene Valley, Essex Co., New York, in the American Museum of Natural History, ex- amined. Aranea sachimau Archer, 1951, Amer. Mus. Novi- tates, 1487: 33, fig. 55, 9. Female holotype from Norwell, Plymouth Co., Massachusetts, in the American Museum of Natural History, examined. NEW SYNONYMY. Description. Female paratype of A. iviei from New Jersey. Carapace, sternum, legs orange-brown without any marking. Abdo- men whitish with an orange cast, and with a few whitish spots framed by darker orange. Venter between epigynum and Araneus diadematus group Orb-weavers • Levi 163 O Araneus yukon • Araneus groen landicola Mao 6. Distribution of Araneus iviei (Archer), Araneus yukon sp. n., and Araneus groen/andico/us (Strand) spinnerets has a white square with orange center. Abdomen lacks humps. Total length, 12 mm. Carapace, 5.0 mm long, 4.2 mm wide. First femur, 4.4 mm; patella and tibia, 5.4 mm; metatarsus, 3.2 mm; tarsus, 1.5 mm. Second patella and tibia, 4.9 mm; third, 2.9 mm; fourth, 4.6 mm. Male from Michigan. Carapace and sternum brown, darker on each side. Legs brown, very indistinctly marked. Dorsum of abdomen whitish with outline of a folium. Brownish venter has a light mark followed by a dark mark between epi- gastric furrow and spinnerets. Coxa with- out hooks or spurs. Second leg not modified. Total length, 6.7 mm. Carapace, 4.2 mm long, 3.2 mm wide. First femur, 4.9 mm; patella and tibia, 5.2 mm; meta- 164 Bulletin Museum of Comparative Zoology, Vol. 141, No. 4 tarsus, 3.6 mm; tarsus, 1.4 mm. Second patella and tibia, 4.2 mm; third, 2.9 mm; fourth. 4.1 mm. Variation. Females vary from 8.5-12 mm in total length, carapace 4.1-5.0 mm long, 3.2-4.2 mm wide. Males varied from .5-7 mm in total length, carapace 3.6-4.2 mm long, 2.9-3.2 mm wide. In posterior view no two epigyna are exactly alike (Figs. 140, 142). Figure 142 illustrates a specimen designated as paratype of A. .sachimau by Archer. Diaii,nosis. Araneus iviei superficially resembles A. trifoUuni; the genitalia are unlike those of any other North American species but very similar to those of the Eurasian A. alsine. The epigynum, unlike that of any other North American species, has a strongly domed base anteriorly and is truncate posteriorly (Figs. 138, 141). In A. iviei the shape of the median septum in posterior view narrows dorsally toward the abdomen (Figs. 140, 142), while in A. alsine it \\'idens into a flat plate more or less truncate toward the abdomen (Figs. 133, 134). The male's palpal conductor has a distal "pocket" (Figs. 146, 148) unlike other Northern American Araneus species. It differs from A. alsine in the shape of the conductor, and in the shape of the embolus (Figs. 145-150). Habits. The species has been collected from a cedar (Thuja occidentalis) swamp in Michigan, sweeping old fields in open forest vegetation on Isle Royale, on base of lodgepole pine (Pimis contorta) and in lodgepole woods in Alberta, and in conifer- aspen (Populus tremuloide.s) with pockets of black spruce (Picea mariana) with dense undercover of rosebushes, grasses, and sedges in the damp area, at Lake George, Alberta (R. E. Leech, personal correspondence ) . Distribution. From Alberta southeast to Pennsyhania. The northernmost record is Riverton. Manitoba, the southernmost Len- hartsville, Berks Co., Pennsylvania (Map 6). Araneus quadratus Clerck Figures 152-158 Araneus quadratus Clerck, 1757, Svenska Spind- lar, p. 27, pi. 1, fig. 3, $. Female holotype from Sweden in the Natural History Museum, Stockholm, lost. Locket and Millidge, 1953, British Spiders, 2: 130, figs. 84b, 85b, 86b, ?, S . Bonnet, 1955, Bibliographia Araneorum, 2: 575. Aranea rcaumurii Scopoli, 1763, Entomologia Car- niolica, p. .393. Types from Austria. Scopoli's collection was destroyed about 1776. ( Horn and Kahle, 1936, Entomol. Beihefte, 3: 252.) Wiehle, 1931, in Dahl, Tierwelt Deutschlands, 23: 79, figs. 115-118, 5, c^ . Roewer, 1942, Katalog der Araneae, 1: 804. Epeira quadrata, — Wiehle, 1927, Z. Morphol. Okol. Tiere, 8: 496. Nielsen, 1932, Biology of Spiders, 2: 292. Note. This European species closely related to A. yukon has not been found in America. Japanese specimens alleged to be A. (juadratus were examined, but large differences in the shape of the conductor and median apophysis make me think neither specimen is correctly identified. Araneus quadratus reported from Green- land are A. groenlandicolus (Strand). The web is made in high grasses of moist areas, the center about 50 cm above the ground, and has about 20 spokes (Wiehle, 1931; Nielsen, 1932). Araneus groenlandicolus (Strand) Figures 159-166; Map 6 Epeira (juadrata, — S0rensen, 1898, Vidensk. Med- del. Naturhist. Foren. Kobenhavn, 1898. Not A. quadratus Clerck. Aranea reauinuri var. groenlandicola Strand, 1906, Fauna Arctica, 4: 458. New name for speci- mens of Epeira cptadrata described by S0rensen from Greenland. Female lectotype here desig- nated from Ivigtut, Greenland, 15. VHl. 1889 (Lundbeck) in the Universitetets Zoologiske Museum, Copenhagen, examined. Aranea manitohae Archer, 1951, Amer. Mus. Novitates, 1487: 37, figs. 51, 59, 62, $. Male holotype from The Pas, Manitoba, in the American Museum of Natural History, ex- amined. NEW SYNONYMY. Description. Female. Carapace light brown with median longitudinal darker band and a dark band near each margin. Distal tips of chelicerae brown. Labial Araneus diadematus group Orb-weavers • Levi 165 U- r^ n -o •o ■n ■— o ^ (U __■ o ^ o Q. O. D o (U c IT) CJ UO OJ E ' •o o ^ XI o CN o Ut (U .' o Q. ^o "5 i o CN (/i — 1 LT) CN 3 "5 XI E od E LD LU 3 "5 O) E c E Q. 3 OJ E LU C • > 'a- CD o rv! IT) LO 'd. c (/> 1 — d LU ^ (1> LO 3 CN N3 ui ^o E o o 1 D XI ,_^ j: O- C D ^ Q. O ^ "o 'q. O^ 3 0) LU LU 00 ^ s E c CS E -i; o ■^' ^O u Q. u^ vO i^ in U o ' ^0 o u vt _2 3 c o 3 o o E c z> D) N3 S3 3 i S a 0 T 3 E < 6 UJ < < o E LU CO od ^C) C CO irt to 1 E T CO o ■ — O^ t-L :\ o iC LT) o -a LO 166 Bulletin Museum of Comparative Zoology, Vol. 141, No. 4 endites almost black except anterior border. Sterauni black with a small, anterior median light mark. Legs very strongly banded. Abdomen white or red with white dorsal spots (Fig. 163). Venter is dark in front of spinnerets, light behind epigynum, and has two longitndinal bands. Abdomen without humps. Total length, 9 mm. Cara- pace, 4.0 mm long, 3.3 mm wide. First femur, 3.5 mm; patella and tibia, 4.3 mm; metatarsus, 2.7 mm; tarsus, 1.1 mm. Second patella and tibia, 4.5 mm; third, 2.3 mm; fourth, 3.5 mm. Male holotype. Carapace, legs light brown. Sternum light brown with median longitudinal white line. Abdomen all white with white pigment spots around spin- nerets. Coxae without hooks or spurs. Tibia of second leg swollen and with strong spines. Total length, 6.1 mm. Carapace, 3.2 mm long, 2.6 mm wide. First femur, 3.2 mm; patella and tibia, 4.0 mm; meta- tarsus, 2.7 mm. Second patella and tibia, 3.0 mm; third, 1.9 mm; fourth, 2.9 mm. Variation. The scape of the epigynum of most females is broken off ( Fig. 162 ) , apparently it tears when mating. Females varied from 9-12 mm in total length, males from 5.5-7.5 mm. Diagnosis. Araneus groenhndicolus lacks humps on the abdomen. Females have been confused with A. trijolium because of the reddish coloration of the abdomen and the median dark carapace stripe; however, the epigynum of A. groenlandicolus is very different (Fig. 159-162). The epigynum has wider rims than that of A. quadratus; the inside edges of the rims are covered by the scape (Fig. 159) (often broken off). The male of A. groenJandicolus has strongly bent distal hooks on its median apophysis (Figs. 164, 165) differing from those of A. quadratus and A. yidwn and other Araneus species. Habits. Almost nothing is known about the habits of A. groenlandicolus, but it is assumed to prefer open ground to forest as do other species of Araneus with a round abdomen. In Alberta it has been found in sedge. Distribution. Alberta to Greenland and south to Minnesota and Maine (Map 6). Specimens examined from Greenland came from Ivigtut and Godthaabsfjord. Araneus yukon sp. n. Figures 167-173; Map 6 Holotype. Male from Snag, lat 62° 24', long 140° 22', Yukon Territory, 24 July 1948, in the American Museum of Natural History. The specific name is a noun in apposition. Description. Female para type. Cara- pace brown with darker median and lateral bands. Sternum very dark brownish black with median longitudinal line of white pigment. Legs brown with very distinct dark banding. Abdomen with a folium posteriorly, a median light longitudinal mark anteriorly. Venter light with two dark spots side by side and a dark ring around spinnerets. Spinnerets blackish brown. Abdomen oval to subspherical without humps. Total length, 10 mm. Carapace, 4.2 mm long, 3.5 mm wide. First femur, 4.0 mm; patella and tibia, 4.9 mm; metatarsus, 3.0 mm; tarsus, 1.4 mm. Second patella and tibia, 4.4 mm; third, 2.8 inm; fourth, 4.0 mm. Male holotype. Carapace with indistinct markings. Sternum dark with a median white line. Leg banding indistinct. Dor- sum of abdomen has two pairs of white spots anterior to a dark folium and an anterior longitudinal white mark (Fig. 170). Ventral dark spots are fused to each other and the posterior one to the dark ring around the spinnerets. There are no hooks or spurs on coxa. The second tibia is slightly swollen. Total length, 6.5 mm. Carapace, 3.4 mm long, 2.8 mm wide. First femur, 3.5 mm; patella and tibia, 4.4 mm; metatarsus, 3.0 mm; tarsus, 1.3 mm. Second patella and tibia, 3.4 mm; third, 2.0 mm; fourth, 3.2 mm. Diagnosis. Araneus yukon palpus differs Aranevs diadematus group Orb-weavers • Levi 167 from that of related A. quodrotus in having teeth on the lower distal lobe of the median apophysis (Fig. 172) instead of a straight carina; the conductor of A. yukon is wider and the "upper" lobe of the embolus (left in Figs. 171, 173) is wider than in A. quadratus (Fig. 158). European specimens showed little variation in these characters. Habits. Firth River, a locality in which the species is found, is on the north slope of the British Mountains, 25 miles from the Arctic Ocean. Although north of the tree line, pockets of spruce ( Fica sp. ) to 3-4 m high occur in sheltered spots along the creeks; probably the specimen was found in these (R. E. Leech, in letter). Records. Yukon Territory. Female para- type collected with holotype; Firth River, British Mountains, 9 paratype, 24. VII. 1956 (R. E. Leech) in the Canadian National Museum. Araneus trifolium (Hentz), Shamrock Spider Figures 174-182; Map 7 ?Epeira viilpecula Walckenaer, 1841, Histoire Naturelle des Insects Apteres, 2: 69. Syntypes are Abbot, Georgian Spider Illustrations, figs. 131, 356 from Georgia in the British Museum of Natural History, examined. A nomen dubituu and a nomen ohlitum. ?Epeiia appioximata Blackwall, 1846, Ann. Mag. Natur. Hist., 17(1): 80. Specimen from vicinity of Toronto, lost. A nomen dtthium and nomen oblittmi. Epeira trifolium Hentz, 1847, ]. Boston Soc. Natur. Hist., 5: 471, pi. 31, fig. 1, 2. Types from Maine, destroyed. Emerton, 1884, Trans. Connecticut Acad. Sci., 6: 306, pi. 33, fig. 8, pi. 35, figs. 13, 14, 21, 22, $, $. McCook, 1893, American Spiders, 3: 145, pi. 1, figs. 3-6, pi. 2, fig. 3, 9, S. Kaston, 1948, Bull. Con- necticut Geol. Natur. Hist. Surv., 70: 258, figs. 823-825, 2047. Epeira aureola Hentz, 1847, J. Boston Soc. Natur. Hist., 5: 471, pi. 31, fig. 2, 9. Type from Maine, destroyed. Epeira trifolium var. candidans McCook, 1893, American Spiders, 3: 146, pi. 1, fig. 4, ?. Fe- male holotype from California, lost, the color of the abdomen was yellowish-white. Aranea trifolium, — Comstock, 1912, The Spider Book, p. 479, figs. 501-508, $ ; 1940, rev. ed.. The Spider Book, p. 493, figs. 501-508, $. Roewer, 1942, Katalog der Araneae, 1: 863. Aranea gosogana Chamberlin, 1920, ]. Entomol. Zool., 12: 8, pi. 4, fig. 6, $. Female holotype from desert region in California, in the Museum of Comparative Zoology, examined. Araneus trifolium, — Bonnet, 1955, Bibliographia Araneorum, 2: 614. Note. Epeira jospidata Walckenaer, 1837, has been placed as a synonym, first by McCook: the type. Abbot, fig. Ill, how- ever, seems to have humps. Abbot's figure is either A. nordmanni or A. bicentenarius. However, Walckenaer indicated in the de- scription of the figure that the abdomen was round. Description. Female. Carapace brown with a lighter band on each side of a dark median longitudinal band; sides dark with a lighter brown border. Sternum dark brown. Legs brown with strongly marked darker bands. Dorsum of abdomen reddish with white spots ( Fig. 177 ) , venter usually reddish brown without markings in adult. A specimen from Ithaca, New York, mea- sured 15 mm in total length. Carapace, 6.5 mm long, 5.5 mm wide. First femur, 6.3 mm; patella and tibia, 8.0 mm; metatarsus, 5.5 mm; tarsus, 1.7 mm. Second patella and tibia, 6.9 mm; third, 4.2 mm; fourth, 6.7 mm. Male. Carapace, sternum and legs brown, abdomen whitish, sometimes all white, sometimes with ventral marks. None of the coxae have hooks or spurs. The tibia of the second leg is only slightly thicker than that of other legs. A specimen from Mon- tana measured 7.5 mm in total length. Carapace, 4.2 mm long, 3.4 mm wide. First femur, 4.6 mm; patella and tibia, 5.1 mm; metatarsus, 3.5 mm; tarsus, 1.5 mm. Second patella and tibia, 4.5 mm; third, 2.7 mm; fourth, 3.9 mm. Variation. Many specimens have the dorsum of the abdomen white. Total length of females varied between 9-20 mm, with carapace 4.0-6.8 mm long, 3.6-5.4 mm wide. Total length of males, 5-8 mm, with carapace 3.0-3.6 mm long, 2.5-3.0 mm wide. 168 Bulletin Museum of Comparative Zoology, Vol. 141, No. 4 Araneus trifolium Map 7. Distribution of /Araneus trilolium (Hentz). Diagnosis. Araneus trifolium is not close to any other species, although females have been confused with A. (iroenlandicolus and A. iviei on account of the spherical abdo- men, red coloration, and dark leg bands. But the epigynum of A. trifolium is very dis- tinct: a scape (only rarely broken off) flanked on each side by a depression in the base; the margin of the depression and the depression itself are sclerotized and black posteriorly (Fig. 174). The male palpus of A. trifolium has the embolus of an unusual shape (Figs. 178, 180-182); the embolus lamella is a huge flat scale with parallel sides of about the same visible area or larger than the small, narrow conductor ( Fig. 179 ) . The lamella, always easily seen, readily separates males from all other North American species. As no illustrations of the male palpus have been readily available, many males of this common species are misidentified in col- lections. Habits. Araneus trifolium is found in meadows and edges of fields making its web between herbaceous plants, often goldenrod {Soliclago sp. ), or shrubs, with the hub about 0.5-2 m off the ground. The web has about 20 spokes. The web and egg sac have been illustrated by Comstock (1912, 1940). During the recent years of drought the spiders disappeared completely from the fields around our house in Mas- sachusetts but reappeared with the first wet season in summer of 1967, to reach unusual abundance in 1968 and 1969. Kaston (1947) describes the egg sacs after Scheffer as "white, about an inch in diam- eter and delicate enough for the mass of several hundred yellow eggs to show Araneus diadematus group Orb-weavers • Levi 169 oo u o 3 o J3 ' ■ 1 CO 0) (- 4^ o D T) n 3 CO <1) n (1) c- •~ o ^ D -v on ~o -rr > ^ oo o^ ■ , *-- ^ CO 00 oo oo o c (1) i/i : 0> c> (U > W1 -> 3 3 CN Ov D) :5 O) O) D) u_ u. u. U- 1 — o Q. >s O JH o > , -C 13 r-s! id O o CN ^ ^ U-1 o Nit E CN CN D Cvi ^ 6 O CN o d E CD 'd. m Q. D CM o E dj CD E E rv! u c ^ D "o E D »/i ^ J3 E -o D cn E 01 ^ c o o a LU CD _□ m c E -^' ^ E ^ LU On' O CN E Q. Q- 0 O > 3 o "5 u 3 O 5 E (J Cvi "o E > 0 CN E D C >N ^ 3 D d) 3 dJ N O) O a. D) O CD u_ U- 2 Ll_ kl 174 Bulletin Museum of Comparative Zoology, Vol. 141, No. 4 same as what is called A. pima here. Other specimens in the McCook collection were A. gemmoides. The specimen which may be A. pima was not chosen lectotype be- cause it either has the wrong collecting locality or comes from the very border of its range. Also the specimen appeared to be an atypical A. pima. To favor stability of names, the other specimen was chosen lectotype even though as a result A. pirus will have to be synony- mized. The range of the species whose name is now A. gemma is essentially that given by McCook and also McCook's de- scription matches this species. Araneus gemma of Chamberlin and Ivie is certainly this species although the specimens could not be located. The name A. gemma has at times been used for A. pivm on collect- ing vials only. The name gemma is a noun in apposition; its ending does not change with the gender of the genus (Bonnet, 1955). Description. Female from California. Carapace maculated brown with some gray pattern. Legs yellowish, indistinctly banded brown. Abdomen brownish with a median longitudinal white line, or line absent. Venter with a black band enclosed by white brackets. Abdomen with large humps (Fig. 210). Total length, 12 mm. Carapace, 5.3 mm long, 4.6 mm wide. First femur, 5.9 mm; patella and tibia, 8.2 mm; metatarsus, 5.0 mm; tarsus, 1.9 mm. Second patella and tibia, 8.0 mm; third, 4.6 mm; fourth, 6.8 mm. Male. Carapace, sternum, legs yellow- brown. Abdomen with a folium and an- terior median white mark. Venter gray with a white bracket on each side. First coxa with a small tubercle on distal margin. Second tibia not modified. Total length, 8 mm. Carapace, 4.2 mm long, 3.3 mm wide. First femur, 5.9 mm; patella and tibia, 7.0 mm; metatarsus, 4.6 mm; tarsus, 1.7 mm. Second patella and tibia, 6.0 mm; third, 3.3 mm; fourth. 4.9 mm. Variation. Females vary 9-19 mm in total length; carapace 4.4-6.4 mm long. 3.8-6.1 mm wide. Males vary 5.8-8.5 mm in total length; carapace 3.1-4.5 mm long, 2.5-3.6 mm wide. The color varies in alcoholic specimens from light gray to almost black, some have two bars on the venter. Hardly two speci- mens have similar epigyna. It is believed that the species hybridizes with Araneus gemmoides and introgression is taking place. The male palpi, similar to those of A. gemmoides except for proportions, were not carefully studied for variation. (See introduction under species problems.) Diagnosis. The short scape of the epigy- num differs from that of related species in having a median ridge (Figs. 203, 205, 207, 208). But some females cannot be separated from A. gemmoides. (See Vari- ation above and introduction under species problems.) The males are much smaller in size than those of the eastern A. cavati- cus. The palpus has a larger bulb and smaller tibia (Figs. 211, 212) than the palpus of A. gemmoides. The upper sur- face of the embolus differs (Figs. 211, 213, 214) from that of A. gemmoides and the terminal apophysis is more pointed. Note. McCook when describing A. gemma must have noted the problems of variation and applied one name to A. gemmoides and A. gemma. Presumably he could not readily separate the specimens which he had from California. Chamberlin and Ivie having smaller collections at hand used gemma for some specimens (Cham- berlin and Ivie, 1935, fig. 79) and called others A. pirus (Chamberlin and Ivie, 1935, fig. 81), a reasonable judgment if only very few specimens are examined. Habits. Little is known of the habits of A. gemma, although they are probably the same as those of A. cavaticus and A. gern- moides. One specimen was collected on a redwood (Sequoia sempervirens) trunk, San Jose, California. One from a porch at Walla Walla, Washington, others from Moscow, Idaho, came in the same col- lection with A. gemmoides. One A. gemma was collected from a hole in an olive tree Araneus diadematus group Orb-weavers • Levi 175 176 Bulletin Museum of Comparative Zoology, Vol. 141, No. 4 (Olea europea) feeding on a ruby-crowned kinglet (Regulus calendula) which was caught in its web (Pasadena, Cahfoniia, 5 November, 1942, Mrs. H. Michener). Distribution. Araneus gemma is found from southern Alaska to southern Cali- fornia; the northernmost collections are from Ketchikan, Alaska, the easternmost from Bigfork, Montana (Map 8). Araneus pima sp. n, Plate 1, 4; Figures 218-232; Map 8 Holotype. Female holotype from Madera Canyon, Santa Rita Mountains, Pima Co., Arizona, 29 July 1958 (A. Ross) in the American Museum of Natural History. The specific name is a noun in apposition after the type locality. Description. Female from Arizona. Cara- pace brown. Sternum brown with median area lighter. Legs brown. Abdomen gray- brown, without folium, the venter with a pair of parallel longitudinal black bars surrounded by white (Plate 4, Fig. 227). Humps very large. Abdomen covered by sparse long setae. Total length, 20 mm. Carapace, 8.0 mm long, 6.4 mm wide. First femur, 7.8 mm; patella and tibia, 9.8 mm; metatarsus, 6.5 mm; tarsus, 2.2 mm. Second patella and tibia, 9.6 mm; third, 5.5 mm; fourth, 8.9 mm. Male from Arizona. Coloration like that of female except that legs show some band- ing and abdomen has a folium and an- terior median longitudinal white line. Coxae and second tibia arc not modified. Total length, 9 mm. Carapace, 5.5 mm long, 4.3 mm wide. First femur, 7.0 mm; patella and tibia, 10.0 mm; metatarsus, 5.9 mm; tarsus, 1.9 mm. Second patella and tibia, 7.6 mm; third, 4.2 mm; fourth, 6.4 mm. Variation. The largest female measured 27 mm in total length; the carapace was not measured. Another male measured 5.2 mm in total length, carapace, 2.5 mm long, 2.0 mm wide. Although what are believed to be tips of A. illaudatus emboli are found at times in the groove of the epigynum, the species do not seem to hybridize. There is not much variation except in length of scape. Diagnosis. The flat rounded scape of the epigynum (Figs. 218, 221, 224) with a deep wide groove under it (Figs. 220, 223) distinguishes females of A. pima from A. gemmoides and A. gemma. The males are distinguished from A. gemma by the shape of the median apophysis, by having the median spine joined with the distal one, and by the truncate embolus (Figs. 228, 232). The more blunt terminal apophysis (Figs. 228, 229) and the shape of the median apophysis with the spines joined distinguishes A. pima from A. illaudatus. Habits. Araneus pima has been collected under eaves of buildings of the South- western Research Station, near Portal, Ari- zona, under a bridge in New Mexico, along a trail in the Grand Canyon, and in cave enti-ance of O.T.L. cave north of Alpine, Texas. The eggs are in an oval mass 2 cm long, 1.8 cm wide, 1 cm thick. About 700 eggs were estimated on the surface of the clump of eggs, which must have contained more than a thousand. A loose pink woolly web 4-5 cm- surrounded the eggs, quite similar to the egg sac of A. gemmoides (Plate 1). Note. In some collections specimens of this species had been incorrectly deter- mined as A. gemma. It is not A. gemma of McCook nor of Chamberlin and Ivie ( 1935, fig. 79). Distribution. Araneus pima is found from eastern California to Utah and Arizona. Localities at the border of the range are: Gateway, Oregon, Kerrville, Texas (Map 8). Female and male para types from the type locality have been collected. Araneus illaudatus (Gertsch and Mulaik) Figures 233-240; Map 8 Armiea illaiidata Gertsch and Mulaik, 1936, Amer. Mus. Novitates, 863: 19, figs. 36, 37, $. Male holotype from Edinburg, Texas, in the American Museum of Natural History, examined. Archer, Araneus diadematus group Orb-weavers • Levi 177 1951, Amer. Mus. Novitates, 1487: 36, fig. 74, S (not fig. 67 2 ). Note. Archer ( 1951 ) matched the wrong female to the male described previously as A. illaudatus. Archer's female is an A. nordmanni and as Archer correctly points out is quite different from females of A. cavaticus group. Description. Female from Arizona. Cara- pace yellowish with dark brown mottling. Sternum black, coxae yellowish, legs banded yellowish and dark brown. Dorsum of abdomen with two rows of transverse black marks pointing posteriorly towards the middle on a mottled background (Fig. 236). A median dorsal longitudinal line of white spots not always present. Venter with a black band between pedicel and spinnerets containing a pair of prominent white spots side by side and sometimes a smaller pair of white spots. Abdomen with two humps. Total length, 11 mm. Cara- pace, 3.8 mm long, 3.4 mm wide. First femur, 4.2 mm; patella and tibia, 5.5 mm; metatarsus, 3.7 mm; tarsus, 1.4 mm. Second patella and tibia, 5.0 mm; third, 3.0 mm; fourth, 4.9 mm. Male holotype. Carapace brown. Ster- num brown. Legs yellow-brown. Abdo- men yellow-brown. Dorsum with distinct folium, anterior border has a white cross. Venter with a pair of white spots side by side. Coxae and second tibia not modi- fied. The abdomen has two distinct humps. Total length, 3.6 mm. Carapace, 2.0 mm long, 1.4 mm wide. First femur, 2.5 mm; patella and tibia, 2.8 mm; metatarsus, 1.8 mm; tarsus, 0.9 mm. Second patella and tibia, 2.3 mm; third, 1.3 mm; fourth, 2.0 mm. Diagnosis. The female scape is con- stricted at its base and almost diamond shaped (Fig. 233), quite different from that of the related A. pima (Figs. 218, 221, 224). The male palpus has the spines of the median apophysis farther apart and the terminal apophysis more pointed and twisted (Figs. 238, 239) than that of Araneus pima. The cap of the embolus is the longest of any Araneus species, longer than the embolus (Fig. 240). Habits. One female collected in the Chiricahua Mts., Arizona, at 7500-9000 feet [2300-2900 m] had its web on brush under pines in August. Distribution. Araneus illaudatus is found from western Texas to Arizona (Map 8). REFERENCES Archer, A. F. 1951a. Studies in the orbweaving spiders (Argiopidae) 1. Amer. Mus. Novi- tates, 1487: 1-52. . 1951b. Studies in the orbweaving spiders (Argiopidae) 2. Amer. Mus. Novitates, 1502: 1-34. Bonnet, P. 1955-1961. Bibliographia Araneorum. Toulouse, Vols. 2, 3. Brown, W. L., and E. O. Wilson. 1956. Char- acter displacement. System. ZooL, 5: 49-64. Chamberlin, R. v., and W. Ivie. 1935. Miscel- laneous new American Spiders. Bull. Univ. Utah, Biol. Ser., 2(4): 1-79. . 1944. Spiders of the Georgia region of North America. Bull. Univ. Utah, Biol. Ser., 8(5): 1-267. CoMSTOCK, J. H. 1910. The palpi of male spiders. Ann. EntonioL Snc. Amer., 3: 161-185. Gertsch, W. J. 1964. The spider genus Zygiella in North America (Araneae, Argiopidae). Amer. Mus. Novitates, 2188: 1-21. Grasshoff, M. 1964. Die Kreuzspinne Araneus pallidas — ihr Netzbau und ihre Paanmgs- biologie. Natur. Mus., 94: 305-314. . 1968. Morphologische Kriterien als Ausdruck von Artgrenzen bei Radnetzspinnen der Subfamilie Araneinae (Arachnida: Ara- neae: Araneidae). Abhandl. Senckenbergi- schen Naturforsch. Ges., 516: 1-100. Helsdingen, p. J. VAN. 1965. Sexual behavior of Lepthyphantes lepiosus witli notes on the function of the genital organs. Zool. Med., 41: 15-42. International Code of Zoological Nomen- clature. 1964. Int. Trust Zool. Nomencl., London. Levi, H. W. 1957a. The spider genera Enoplo- gnatha, Theridion and Paidisca in America north of Mexico (Araneae, Theridiidae ) . Bulb Amer. Mus. Nat. Hist., 112: 1-123. . 1957b. The spider genera Cmstulina and Steatoda in North America, Central America and the West Indies (Araneae, Theridiidae). Bull. Mus. Comp. Zool., 117: 367-424. . 1968. The spider genera Gea and Argiope 178 Bulletin Museum of Comparative Zoology, Vol. 141, No. 4 in America (Araneae: Araneidae). Bull. Mus. Comp. Zool., 136: 319-352. .(in press). Problems in the reprodnctive physiology of the spider palpus. Bull. Mus. Natl. Hist. Natur. Paris. — , AND L. R. Levi. 1961. Some comments on Walckenaer's Names of American Spiders, based on Abbot's drawings. Psyche, 68: 53-57. WiEHLE, H. 1931. Araneidae. In F. Dahl, Die Tierwelt Deutschlands, 23(6): 1-136. . 1963. Beitriige zur Kenntnis der deut- schen Spinnenfauna III. Zool. Jahrb. Abt. System., 90: 227-298. Araneus diadematus group Orb-weavers • Levi 179 INDEX Valid names are printed in italics. Page numbers refer to main references, starred page numbers to illustrations. ahine, Araneus, 161*, 162 andrewsi, Aranea, 146 andrewsi, Araneus, 145*, 146 angulata, Aranea, 142 angulatus, Araneus, 141*, 142 approximata, Epeira, 167 Aranea, 133 Araneus, 133 aureola, Epeira, 167 bicentenaria, Aranea, 143 bicentenaria, Epeira, 143 hicentenarius, Araneus, 139*, Burgessia, 133 143, 145^ candidans, Epeira, 167 canmorus, Araneus, 171 cavatica, Aranea, 170 cavatica, Epeira, 170 cavaticus, Araneus, 169*, 170 cinerea, Epeira, 170 corticaria, Aranea, 158 corticaria, Epeira, 158 corticarius, Araneus, 158, 159* darlingtoni, Aranea, 152 denningi, Aranea, 158 diadema, Epeira, 147 diademata, Aranea, 147 diadematus, Araneus, 147, 149*, 15 Epeira, 133 Euaranea, 133 gemma, Araneus, 172, 173* gemma, Araneus, 176 gemma, Epeira, 171, 172 gemmoides, Araneus, 171, 173* gemmus, Araneus, 172 gigas, Aranea, 156 gigas, Araneus, 143 gigas, Epeira, 143 gosogana, Aranea, 167 groenlandicolus, Araneus, 164, 165"* illaudata, Aranea, 176 illattdatus, Araneus, 175*, 176 169" incestifica, Epeira, 158 insularis, Epeira, 156 iviei, Aranea, 162 iviei, Araneus, 161*, 162 jaspidata, Epeira, 167 kisatchia, Aranea, 143 manitobae, Aranea, 164 mannorea, Epeira, 156 marmoreus, Araneus, 141*, 155*, 156, 159*, 169* Neopora, 133 nigra, Epeira, 148 nordmanni, Aranea, 150 nordmanni, Araneus, 150, 151*, 153*, 155* nordmanni, Epeira, 150 obesa, Epeira, 156 pima, Araneus, 140*, 175*, 176 pirus, Araneus, 172 pseudomelaena, Aranea, 152 quadratus, Araneus, 164, 165* raji, Aranea, 156 raji, Epeira, 156 reaumurii, Aranea, 164 sachimau, Aranea, 162 saeva, Aranea, 148 saeva, Epeira, 148 saevus, Araneus, 141*, 148, 149*, 151* santarita, Aranea, 150 santarita, Araneus, 149*, 150 silvatica, Epeira, 148, 150 sinistra, Aranea, 170 sinistrella, Aranea, 170 solitaria, Aranea, 148 solitaria, Epeira, 148 trifoliimi, Aranea, 167 trifolium, Araneus, 167, 169* trifolium, Epeira, 167 tusigia, Aranea, 156 vulpecula, Epeira, 167 loashingtoni, Araneus, 155*, 159*, 160 yukon, Araneus, 165*, 166 in OF THE SuLU Museum of Comparative Zoology Evolutionary Relationships of Some South American Ground Tyrants W. JOHN SMITH AND FRANCOIS VUILLEUMIER HARVARD UNIVERSITY VOLUME 141, NUMBER 5 CAMBRIDGE, MASSACHUSETTS, U.S.A. 4 MARCH 1971 PUBLICATIONS ISSUED OR DISTRIBUTED BY THE MUSEUM OF COMPARATIVE ZOOLOGY HARVARD UNIVERSITY Bulletin 1863- Breviora 1952- Memoirs 1864-1938 JoHNSONiA, Department of Mollusks, 1941- OccAsiONAL Papers on Mollusks, 1945- Other Publications. Bigelow, H. B., and W. C. Schroeder, 1953. Fishes of the Gulf of Maine. Reprint, $6.50 cloth. Brues, C. T., A. L. Melander, and F. M. Carpenter, 1954. Classification of Insects. $9.00 cloth. Creighton, W. S., 1950. The Ants of North America. Reprint, $10.00 cloth. Lyman, C. P., and A. R. Dawe (eds.), 1960. Symposimn on Natural Mam- malian Hibernation. $3.00 paper, $4.50 cloth. Peters' Check-list of Birds of the World, vols. 2-7, 9, 10, 12-15. (Price list on request. ) Turner, R. D., 1966. A Survey and Illustrated Catalogue of the Teredinidae (Mollusca: Bivalvia). $8.00 cloth. Whittington, H. B., and W. D. I. Rolfe (eds.), 1963. Phylogeny and Evolution of Crustacea. $6.75 clotli. Proceedings of the New England Zoological Club 1899-1948. ( Complete sets only. ) Publications of the Boston Society of Natural History. Authors preparing manuscripts for the Bulletin of the Museum of Comparative Zoology or Breviora should send for the current Information and Instruction Sheet, available from Mrs. Penelope Lasnik, Editor, Publications Office, Museum of Comparative Zoology, Harvard University, Cambridge, Massachusetts 02138, U.S.A. © The Presidenf and Fellows of Horvord College 1971. EVOLUTIONARY RELATIONSHIPS OF SOME SOUTH AMERICAN GROUND TYRANTS^ W. JOHN SM1TH2 AND FRANCOIS VUILLEUMIER^ PREFACE The Tyrannidae comprise about one out of ten of the approximately 2900 species of birds hsted from the South American continent by de Schauensee (1966). Parallehng this large representa- tion is ecological versatility; tyrant fly- catchers occur in every major avian habitat from lowland rain forest to high Andean paramos. As a corollary to this radiation, many Tyrannidae have ecological char- acteristics which are typical of totally un- related Old World avian groups. There are, among others, thrushlike tyrants, chat- like, shrikelike, warblerlike, and titmouse- like tyrants, in addition to flycatchers. The existence of the tyrannid radiation has been known to ornithologists for a long time. Yet a detailed evaluation of even a part of it and of the evolutionary processes involved has not been attempted. Failure to do so is due in a large measure to the provincialism (taxonomic as well as geographic) of many published accounts dealing with Tyrannidae. It is now evident that significant progress in our understand- ing of the radiation within the Tyrannidae depends on broad comparative surveys of geographic patterns and nonmoi-phological characters, especially ethological ones. Until recently, few detailed studies of behavior in tyrant flycatchers have been 1 Published by a grant from the Wetinore Colles Fund. 2 Department of Biology, University of Pennsyl- vania, Philadelphia, Pa. 19104. ^ Biology Department, University of Massachu- setts, 100 Arlington Street, Boston, Mass. 02116. made. Among the significant exceptions, one might mention analvses of genera such as Mijiarchus (Lanyon,^ 1960, 1961, 1963, 1965, 1967), Empidonax (Stein, 1958; John- son, 1963; Mumford, 1964), Tijranmis (Smith, 1966), and Smjomis (Smith, 1969 and in press, 1 and 2). In spite of these analyses, we still know virtually nothing about a large proportion of the genera and species. Museum skins and occasional field observations by collectors are literally all the information available today on whole genera or groups of genera of Tyrannidae. During the past several years both of us have studied various tyrannids. We have approached our subject independently and with different outlooks. Smith has long been interested in the Tyrannidae from the viewpoint of behavior and evolutionary studies. Vuilleumier, on the other hand, has studied only some tyrannid genera in detail as part of an analysis of speciation phenom- ena in Andean birds. Despite our different approaches, we have found that we agree on all the more important points of re- lationships and evolution within a group of genera, most of which are called ground tyrants and live on or near the ground in open habitats of South America, especially in the Andes and Patagonia. These include Muscisaxicola, Agriornis, and Xolmis and appear closely related to Ochthoeca, Saij- ornis, and other genera which are, to vary- ing degrees, less terrestrial. bur results suggest a certain number of changes in the currently accepted taxo- nomic arrangement of these birds. More significantly, the results also suggest some ways in which flycatchers have radiated into some of the more open t\^pes of habi- tats of South America. Because we employed different methods, we shall present our information separately. Nevertheless, the results converge, and the two papers constitute what we consider to be a unit: together they represent an attempt toward a better understanding of the relationships and evolution of several genera of ground tyrants and allied genera. We do not pretend to have reached the final word about the systematics of these genera, and are aware of the gaps in our knowledge. Some of our conclusions are provisional. We hope, however, that the uncertainties will be an incentive for fur- ther research by other workers. W. John Smith Frangois Vuilleumier Chapter I. Generic Relationships and Speciation Patterns in Ochthoeca, Myiotheretes, Xolmis, Neoxolmis, Agriornis, and Alusc/sax/co/a FRANCOIS VUILLEUMIER Abstract. The open habitats of the high Andes, Patagonia, south-central Brazil and adjacent coun- tries are inhabited by several more or less closely interrelated genera of Tyrannidae. The adaptive radia- tion of these flycatchers into nonforest vegetation resulted in a variety of forms that are the ecological counterparts of species from other families (Turdidae, Laniidae) elsewhere. At one extreme of this radia- tion are birds of a rather generalized "flycatcher type" (e.g., Ochthoeca), while at the other are birds that have diverged considerably, and are now entirely terrestrial in habits and wheatearhke in morphology (e.g., Muscisaxicola) . In Part I, morphological and ecological attributes of the genera Agriornis, Neoxolmis, Xolmis, Pijrope, Muscisaxicola, Mtiscigralla, Myiotheretes, and Ochthoeca (as listed by de Schauensee, 1966) are reviewed. As a result of this analy.sis, it is suggested that (a) the genus Xolmis be enlarged to include also species previously placed in the genera Myiotheretes and Pyrope, (b) Neoxolmis be retained as a monotypic genus, (c) Xolmis murina be transferred to the genus Agriornis, and (d) tlie genus Muscigralla be made a subgenus of Muscisaxicola. Diagnoses of the genera Ochthoeca, Xolmis, Neoxolmis, Agriornis, and Muscisaxicola are given, together with descriptions of the recognized species-groups. In Part II, geographical variation, species limits, and interspecific relationships are analyzed in Xolmis, Neoxolmis, Agriornis, and Muscisaxicola. Active speciation can be detected in the genera Xolmis, Agri- ornis, and Muscisaxicola. One can find: (a) species with discontinuous geographical variation across ecological barriers (e.g., Xolmis fumigata, Muscisaxicola maculirostris), (b) geographically isolated taxa at the borderline between species and subspecies ( superspecies ) (e.g., Xolmis pernix, X. fumigata, and X. fuscorufa; Muscisaxicola juninensis and M. albilora), (c) instances of secondary contact involving mar- ginal overlap (in the Muscisaxicola alpina superspecies), or possible hybridization (in Xolmis striaticollis and Agriornis montana ) . The available data suggest tliat in this group of genera, speciation is now taking place chiefly in the high Andes, where geographical isolation occurs across several well-marked ecological barriers. Zones of secondary contact in other regions, now devoid of such barriers ( in northern Bolivia ) , indicate that geo- graphical isolation was once possible there. " Nevertheless, there is so much extensive sympatry between congeneric species, in the Andes ( e.g., in Muscisaxicola) as well as in the lowlands (e.g., in Xolmis) that the early stages of their radiation cannot now be detected. TABLE OF CONTENTS Part II: Variation and Speciation in the Genera Xolmis, Neoxolmis, Agriornis, and Introduction 182 Muscisaxicola 203 Parti: The Genera and their Relationships —. 183. The Genus Xolmis 204 Classification 185 The Genus Neoxolmis __ 208 Analysis of Some Characters 188 The Genus Agriornis 208 The Genus Ochthoeca 192 The Genus Muscisaxicola 215 The Genus Xolmis 193 Discussion 224 The Genus Neoxolmis 197 Presently Actixe Speciation 226 The Genus Agriornis 198 Older Speciation Patterns 229 The Genus Muscisaxicola 200 Literature Cited 230 Bull. Mus. Comp. ZooL, 141 (5) : 181-232, March, 1971 181 182 Bulletin Mu.seinii of Comparative Zoology, Vol. 141, No. 5 INTRODUCTION As was explained in the preface, this paper, toG;ether with the one by Smith, must be considered as an attempt to de- scribe the adaptive radiation of some groups of tyrant flycatchers into open, non- forested habitats of South America, espe- cially the grasslands and scrub of the high Andes. My study of several genera of Tyranni- dae was part of an analysis of speciation phenomena among Andean and Patagonian birds. Geographical variation and inter- species relationships were investigated in each genus selected for this work. I ana- lyzed especially, first, patterns of range discontinuities within the distribution of species, and second, areas of secondary contact between taxa at, or slightly below, the le\'el of species. Knowledge of these phenomena gives us insight about the course of speciation in these birds ( Vuilleu- mier, 1969b). This original scope has been enlarged here to include a discussion of the probable relationships of these birds at the genus level. This paper, then, is divided into two parts. In the first, the taxonomie and evolutionaiy relationships of several genera of bush and ground tyrants are examined, and a classification of these birds slightly at variance with that in current use (e.g., de Schauensee, 1966) is suggested. In the second part, evolutionaiy phenomena at the species level are discussed in the genera Xohms (which here has been enlarged to include MyiotJwrete.s and Pyrope), Neoxol- mis, A'yrionii'i (includes murina, a species formerly put in Xohnis), and Muscisaxicola ( includes Muscigralla ) . This paper is the third in a series of speciation studies on Andean birds (see Vuilleumier, 1968; 1969a). ACKNOWLEDGMENTS I am deeply indebted to Ernst Mayr for his advice and stimulation during my studies of speciation, and for having read and criticized an earlier version of this paper. The manuscript benefited greatly from the numerous suggestions of Beryl S. Vuilleumier. I wish to thank sincerely Roy F. Steinbaeh (Cochabamba), Germain Broquet (Quito), ITannes Vogt (La Paz). Pablo Legname (Tucuman), Maria and Hans-Wilhelm Koepcke (Lima), and Mario Ricardi ( Concepcion ) for their help during my field work in South America. Jan Pinowski (Warsaw), Alexander Wetmore, and Richard L. Zusi (Washington) kindly provided me with data on specimens, and O. S. Pettingill, Jr., generously lent me his field data on some Falkland Islands birds. I am grateful to the following persons who have helped me during my exami- nation of specimens in the collections under their care, or who have sent me specimens on loan: Villy Aellen (Museum d'Histoire Naturelle, Geneva), Dean Amadou and Charles O'Brien (American Museum of Natural History), James Bond (Academy of Natural Sciences of Philadelphia), Jean Dorst (Museum National d'Historie Natur- elle), Ian C. J. Galbraith (British Museum, Natural History), Antonio Olivares (Insti- tuto de Ciencias Naturales, Bogota), Claes C. Olrog (Instituto Miguel Lillo, Tucu- man), Raymond A. Paynter, Jr. (Museum of Comparative Zoology), W. H. Phelps, Jr., and the late W. H. Phelps (Caracas), the late R. A. Philippi (Museo Nacional, Santiago), Robert W. Storer (Museum of Zoology, University of Michigan), and Enist Sutter ( Naturhistorisches Museum, Basel). I acknowledge the generous financial assistance of the National Science Foun- dation (Grants G-19729 and GB-3167 to the Committee on Evolutionary Biology of Harvard University), the Bourse federale de voyages de la Societe helvetique des sciences naturelles (Switzerland), the So- ciety of the Sigma Xi, and the Frank M. Chapman Memorial Fund of the American Museum of Natural Histoiy. These grants made possible the study of bush and Evolution of Ground Tyrants • Smith and Viiilleumicr 1S3 ground tyrants during trips to South America in 1964, 1965, and 1967-1968. The illustrations of this paper were pre- pared by Margaret Estey, to whom I ex- press my thanks. Finally, I wish to thank Penelope Lasnik for her editorial assistance. PART I: THE GENERA AND THEIR RELATIONSHIPS The taxa analyzed below include eight genera recognized by de Schauensee (1966), who followed the unpublished work of Zimmer. They are Agriornis (4 species), Neoxolmis (monotypic), Xohnis ( 8 ) , Pyrope ( 1 ) , Muscisaxicola (11), Muscigralla (1), Myiotlieretes (6), and Ochthoeca (9). These eight genera are all more or less closely related to one another and probably represent a group of Tyrannidae that could conceivably be given suprageneric rank (e.g., tribe). I refrain from suggesting such a taxonomic procedure at the present time, however, because I have not made a sufficiently detailed analysis of other genera ( Tum- hezia, CoJoramphus, Ochthornis, Sayoniis, Pyrocephahts) that I believe to be more closely related to the above eight genera than their location in check-lists (Hell- mayr, 1927; de Schauensee, 1966) would imply. In this part, I will discuss the re- lationships of groups of species currently classified in the above eight genera, and will show that an alternative classifiqation to the one advocated by de Schauensee (1966) might better express the phenetic groupings of these species. The birds belonging to the above eight genera are distributed in the high Andes, from the timber line up to the upper limits of vegetation near the snow line, and from Colombia and Venezuela in the north to Patagonia in the south. In the lowlands and uplands of the continent, they occur in nonforest vegetation along the Pacific coastal plain and especially east of the Andes in Bolivia, Brazil, Paraguay, Argen- tina, and Uruguay. They are even found on the Falkland Islands off southern Pata- gonia. In view of this broad distribution in South America, it is interesting to point out that this group of genera does not occur in two other regions where the vegetation is predominantly nonforest, and where they might be expected to be found. One is the llanos of Venezuela and Colombia, and the other the high altitude grasslands of the mountains of southeastern Brazil. A first group of these genera {Xolmis, Pyrope, Myiotheretes, and Ochthoeca) can be called bush tyrants. A second group (Agriornis, Neoxolmis, Muscisaxicola, and Muscigralla) are ground tyrants: they usually live away from trees, in very open habitats such as scrub and grass asso- ciations. Ecological preferences vaiy at two levels in these eight genera. (A) Interspecific differences can clearly be seen in Xolmis, where X. ruhetra is largely terrestrial, in contrast to the other species, which are more arboreal, and in Ochthoeca, where species such as O. ciniiamomeiventris in- habit forested slopes, whereas the others live in open scrub (e.g., O. leucophrys). (B) Intraspecific variability is evident in Myiotheretes striaticoUis, which lives in wet woodlands in some parts of its range (e.g., in Ecuador), but in diy, xeric que- bradas in others (e.g., in Argentina). This ecological variability is, I believe, of evolutionaiy, and hence taxonomic, sig- nificance, and is inteqoreted here in terms of adaptive trends. In other words, these birds may have become adapted to in- creasingly open habitats during their evo- lutionary history. Such trends may have culminated in the adaptive radiation of Muscisaxicola, which has about ten species that live, often sympatrically, along the Andes in the paramo and puna grassland and scrub. Geologically speaking, these habitats are probably the most recent of any South American type of vegetation. The species of Muscisaxicola differ from most arboreal flycatchers in a series of characters that are probably all part of the 184 Bulletin Museum of Comparative Zoology, Vol. 141, No. 5 adaptations (and concomitant evolutionary changes) to such extreme habitats: long legs, inconspicuous colors and patterns, and simplification of vocal manifestations. Unfortunately, ornithologists have so far done relatively little work that might en- lighten us about the modalities of this adaptive radiation. Part of the reason for this failure lies in the taxonomic treatment of these birds. It has been customary to keep the generic limits of taxa such as Agriornis, Xolmis, Muscisoxicola, and Och- thoeca relatively constant, because these genera contain species that resemble each other closely. But the more distinct or more intermediate taxa (usually single species) have either been placed in a monotypic genus or shifted back and forth from one genus to another. Such shifts can be made one way or the other, depending on which character a given author feels has more taxonomic weight as a "generic" character. Thus the species rufiventris is in some characters intermediate between Xolmis and Myiotheretes, but is distinct from either in other traits, and so was given generic rank (Neoxolmis) by Hellmayr (1927). The species pyrope differs in several ways from species ordinarily placed in Xolmis, and was consequently isolated in a monotypic genus Pyrope by Zimmer (MS; see de Schauensee, 1966). The spe- cies nifipennis was removed from Cnemor- chus and placed in Xolmis by Zimmer (MS). Finally, the species brevicouda has been isolated in a monotypic genus for so long that no one has asked whether this taxon was possibly related to others, and which ones they might have been. Part of the taxonomic problem just out- lined can be resolved by taking into con- sideration ecological and behavioral fac- tors, together with the more traditional ones of moipliology. The classification pre- sented in this paper is one attempt to do so. I suggest that the major ecological shift that has occurred in the evolutionary history of the bush and ground tyrants has been a change both from arboreal to terrestrial habits and from a relatively closed habitat (woodland) to a much more open one (steppe). The present geographical range of bush flycatchers in the nonforested Brazilian uplands, and various other nonforest vegetation for- mations of central and southern South America coincides in part with regions of great geological age that have not under- gone major changes in their position or structure. Indirect evidence (e.g., from fossil mammals) suggests that the vege- tation of such regions has been nonforest for a long time (i.e., since at least the mid- Tertiary). One may therefore suppose that the ecological shift has been chiefly in the direction of woodland to steppe, because the South American steppes (high Andean and Patagonian) appear to be much more recent than the central South American woodland formations. Whatever the actual direction of the shift, however, I have placed taxonomic weight on characters that appear to be correlated with it. Since a classification that includes only the categories genus and species is, in my opinion, not able to express satisfactorily the multiple hierarchical relationships of taxa, I have, as in previous systematic papers ( Vuilleumier, 1965, 1968,^ 1969a). made full use of the species-group and superspecies categories, which do not burden the formal nomenclature by ad- ditional names. I have also used subgenera in one case. The classification of the bush and ground tyrants constructed on the basis of an analysis of xariation in characters is out- lined below prior to the character analysis, in order to maintain throughout the paper as much uniformity as possible in the nomenclature of these birds. I am fully aware that a linear sequence of taxa cannot express all the possible re- lationships of these taxa. The sequence adopted here is therefore a compromise. The following conventions have been adopted. Genera having entirely or mostly arboreal and woodland-inhabiting species Evolution of Ground Tyrants • Smith and Vuilleumier 185 are placed before genera having terres- trial and steppe-inhabiting species. Thus Ochthoeca and Xohnis are placed before NeoxoJmis, Agriornis, and Muscisaxicola. Ochtlweca precedes Xohnis because it may be more closely related to the two Andean species-groups of Xohnis (numbers 1 and 2 in the list below) than to other genera. Agriornis is listed between Xohnis and Neoxohnis on the one hand, and Muscisaxi- cola on the other, because in some ways, it is intermediate between them. Within genera, the species-groups are listed with the Andean group ( s ) first, and the lowland one(s) next, except in Muscisaxicola, where this procedure would have relegated the maculirosfris species-group toward the end of the list, far away from the subgenus Muscigralla, with which it is more related than with other species-groups. Within species-groups, the species are listed either randomly if sympatric, or geographically from north to south if allopatric. Species-groups and superspecies have been named after the oldest name of any taxon within each. Species or semispecies considered to be component members of a superspecies are included in braces. Classification Genus Ochthoeca Cabanis, 1847 1. cinnamomeiventris species-group O. cinnamomeiventris (Lafresnaye, 1843) 2. diadema species-group O. diadema (Hartlaub, 1843) frontalis superspecies \ O. frontalis ( Lafresnaye, 1847 ) [ O. pulchella Sclater and Salvin, 1876 3. rufipectoralis species-group O. rufipectoralis ( Lafresnaye and d'Or- bigny, 1837) 4. oenuntJioides species-group O. fumicolor Sclater, 1856 O. oenanthoides ( Lafresnaye and d'Or- bigny, 1837) leucophrys superspecies O. leucopJirt/s (Lafresnaye and d'Orbigny, 1837) O. piiirae Chapman, 1924 Genus Xolmis Boie, 1826 1. fumigata species-group X. striaticollis (Sclater, 1853) fumigata superspecies X. pernix (Bangs, 1899) X. fumigata ( Boissonneau, 1840) X. fuscorufa (Slater and Salvin, 1876) (X. signata ( Taczanowski, 1874) is main- tained in this genus and species-group provisionally, pending critical examination of the one or two extant specimens. ) 2. erijthropijgia species-group X. enjthropygia ( Sclater, 1853 ) X. rufipennis (Taczanowski, 1874) 3. pijrope species-group X. ptjrope (Kitthtz, 1830) 4. cinerea species-group X. cinerea (Vieillot, 1816) X. velata ( Lichtenstein, 1823) X. coronata (Vieillot, 1823) X. dominicana (Vieillot, 1823) X. irtipero (Vieillot, 1823) 5. ruhetra species-group X. ruhetra (Bumieister, 1860) Genus Neoxohnis Hellmayr, 1927 N. rufiventris (Vieillot, 1823) Genus Agriornis Gould, 1839 1. montana species-group A. montarm (Lafresnaye and d'Orbigny, 1837) A. alhicauda (Philippi and Landbeck, 1863) 2. hvida species-group A. Uvida (Kittlitz, 1835) A. microptera Gould, 1839 3. murina species-group A. murina (Lafresnaye and d'Orbigny, 1837) Genus Muscisaxicola Lafresnaye and d'Orbigny, 1837 subgenus Muscigralla Lafresnaye and d'Orbigny, 1837 M. hrevicauda Lafresnaye and d'Orbigny, 1837 subgenus Muscisaxicola Lafresnaye and d'Or- bigny, 1837 1. maculirostris species-group M. maculirostris Lafresnaye and d'Orbigny, 1837 M. fluviatilis Sclater and Salvin, 1866 2. macloviana species-group M. macloviana (Garnot, 1829) M. capLstrata (Burmeister, 1860) 3. rufivertex species-group M. rufivertex Lafresnaye and d'Orbigny, 1837 alhilora superspecies M. juninensis Taczanowski, 1884 I M. alhilora Lafresnaye, 1855 I 186 Bulletin Museum of Comparative Zoology, Vol. 141, No. 5 Muscisaxicola Ochthoeca Fig. 1. Three of the four groups based on differences in color pattern. Top: Muscisox/co/o ruHvertex, an example of group 1 (unpotterned species with a contrasting crown patch). Center: Agriornis montana, an example of group 2 (species with streaked throat and patterned tail). Bottom: Ochthoeca rufipectoralis, an example of group 3 (species wih superciliary stripe and wing bars). Evolution of Ground Tyrants • Smith and Vuilleumier 187 _j_. -^ O C ■"■ -o ?-5 ^ C c «n .(_ o ^ ^ -^ 2 ■= -Q 1 2 " 1 1 D D u, Q ■•" u X u >s >^ o o O t^ D O 2 1- M- o o _Q D) ^ CD D) ^ .. a c ■ X I/) "o >< _Q "^^^^^^^^1^ o c O .C -" i (1) T3 C D O c D D U> C c *- (U O O r -D -a Q. =) -^ O - ^ cn D- ^=^ u> o E ^ J2 j: -□ * 0) 5 E > o a «- D. O 188 Bulletin Museum of Comparative Zoology, Vol. 141, No. 5 4. alpina species-group alpina superspecies {M. alpina (Jardine, 1849) M. cinerea Philippi and Landbeck, 1864 5. alhifrons species-group alhifrous superspecies f M . alhifrons (Tschudi, 1844) I M. flavinncha Lafresnaye, 1855 M. frontalis (Burmeister, 1860) Analysis of Some Characters Color and Paffern The color of the bush and ground tyrants varies from black and white {Xolmis irupero) to almost entirely dark smoky brown {Xolmis fumigota) or black (Och- tlweca cinnamomewentris) and to different shades of gray and brown, and, in some species, even greenish and yellowish. Some species are unpattemed, while others have striking patterns. This variability permits one to make four groups. (1) Largely unpattemed spe- cies having uniformly colored body, wing, and tail plumage. The outer rectrices usually have a pale margin. Some species have a conspicuous, others, a faint, crown patch of a yellowish, brownish, or reddish color; some species lack a crown patch altogether. This group includes all the species of Mtiscisaxicola plus Xolmis pyrope (Fig. 1). A partial exception is Muscisaxicola hrevicauda, which has a somewhat patterned rump and tail, al- though otherwise it has dull, unpattemed body plumage and a (concealed) crown patch. (2) Species having essentially unpat- temed body and wings ( save for pale outer edges of secondaries and coverts), and a white throat with dark streaks. The tail is patterned with white in some species, but is essentially unpattemed in others. No species has a cro\vn spot. This group comprises the five species here included in Agriornis (Fig. 1). (3) Species having either pattemed or unifomily colored body plumage, and a conspicuous white, buffy, or yellow super- ciliary stripe. Many species also have con- spicuous wing bars. The tail is essentially unpattemed, although a few species have a whitish or pale buffy edge to the outer rectrices. This group includes the species of Ochthoeco (Fig. 1). (4) Species having pattemed wings (often the pattern is concealed), and either pattemed or unpattemed body plumage and tail. The underwing pattern is veiy uniform. The distal one-third to one-half of the primaries is black or dark brown, the basal part rufous, white, or buff (Fig. 2). The exceptions are a white wing tip and black basal part {Xolmis clominicana) , and a pale rufous distal part of the primaries {Xolmis nihetra) (Fig. 2). All other spe- cies of Xolmis (but not X. pyrope, men- tioned above under 1) and Neoxolmis can be included in the patterned group. Attenuation of the Outer Primaries The two outer wing feathers vary from unmodified to highly attenuated at the tip. In most, and probably all, of the species having attenuated outer remiges, this at- tribute appears restricted to adult males. The possible functions of this modification during displays, for example, is still largely unknown. Wetmore (1926a: 302) indi- cated that Xolmis ruhetra made some sort of rattlelike noise in flight, which is most likely related to the attenuation of its outer primaries. Smith describes wing displays in several bush and ground tyrants. Two groups of birds can be made on the basis of variation in the degree of emargi- nation of the remiges. In the first (no attenuation) are all species of Ochthoeca, all species of Muscisaxicola, and several species of Xolmis (Fig. 3). In the second group are species having moderate to marked attenuation. They include all spe- cies of Agriornis { adult males only ) , several species of Xolmis, and Neoxolmis. Within Xolmis, variation in emargination appears to be species-specific as depicted in Figure 3. Evolution of Ground Tyrants • Smith and Vuilleumier 189 (1> a. o c E D O T3 o c 0 E o O c o c o E 0) O >^ Q. c > c _o O) < . (1) § b! ^- .0 o ^ X D 2 Q 2 ^ o «/> o/mis pecifi — — — ^^^' 0 Q j^^ ^ X <" .y ^ a> C (1) ^— ^ M 1***"^ _ , - o U Q. O -- — . — —---„>"- 3 i: ^-""^ 1- M */) . X Q. <= X Fig. 3. Attenuation of differences in attenuatio 190 Bulletin Museum of Comparative Zoology, Vol. 141, No. 5 SIDE VIEW TOP VIEW Ochthoeca cinnamomeiventris Muscisaxicola alpina superspecies Xolmis cinerea rnis livida Fig. 4. Bill structure of Ochthoeca, Muscisaxicola, Xolmis, and Agriornis. O. cinnamomeiventris: short, brood bill, with long rictal bristles, correlated with insectivorous diet and wet, forested habitat. M. alpina-. thin, narrow bill, with short rictal bristles, correlated with insectivorous diet and dry, nonforested habitat. X. cinerea-. relatively long, broad bill, with long rictal bristles, correlated with partially insectivorous diet and moderately dense habitat. A. livida-. long, narrow, and sharply- hooked bill, with short rictal bristles, correlated with varied diet and nonforest habitat. Drawn to scale. 6/// Structure The bill of the bush and ground tyrants is a very variable organ. Although the precise diets and feeding methods of most species are still very poorly known, it is possible to correlate positively certain characters of bill size and shape with food. Thus, small and insectivorous species {Ochthoeca and Muscisaxicola species) have relatively small bills, whereas larger species, of the genera Xolmis and Agriornis, having more diversified diets and occasion- ally or regularly capturing small verte- brates, have relatively powerful bills, often with a sharp hook at the tip (Fig. 4). Bill width varies considerably in the bush and ground tyrants. Species living in wetter and structurally denser habitats seem to have proportionately broader bills than do species living in dry and open habitats (Fig. 4). This correlation is more or less independent of size. Among open- Evolution of Ground Tyrants • Smith and Vuilleumier 191 coimtiy species, small-sized Muscisaxicola and large-sized Agriornis all have relatively narrow bills. Shape of Tail Most species have a square or slightly forked tail. In some species, however, the tail is markedly forked (Neoxohnis riifi- ventris) or markedly rounded (Xolmis coronato). These differences appear to be only species-specific, and of little use for supraspecific groupings. Proporn'ons There is a tremendous amount of vari- ation in the proportions of wing, tail, tarsus, and culmen lengths within the bush and ground tyrants. In most instances, it is dif- ficult to establish correlations between proportions and other characters, or be- tween proportions and habits. The most terrestrial species, such as all species of Muscisaxicola, have proportionately long, thin legs. Arboreal species usually have proportionately shorter legs than do ter- restrial ones. More detailed discussions of the possible taxonomic significance of variation in proportions will be found in the generic accounts below. Size The flycatchers under consideration vaiy from very small ( Muscisaxicola brevi- cauda) to very large {Agriornis livida). Most of this variation appears to be inter- specific. In Muscisaxicola, some of the interspecific variability in size occurs be- tween pairs or among triplets of sympatric species living either in the same or in adjacent biotopes. These differences might represent character divergence permitting resource partitioning, whereas other dif- ferences, such as species-specific head or crown patterns, may permit recognition of conspecific individuals. In Agriornis, the two high Andean spe- cies {montana and alhicauda) are sym- patric over much of their respective ranges. They are very similar in pattern but differ in absolute size. The lowland species limda and microptera, although entirely allo- patric, differ conspicuously in size rather than in pattern. These two examples, in Muscisaxicola and Agriornis, show that ecological cor- relates of size differences may or may not exist. It seems difficult to assign any taxo- nomic significance to size variation among the bush and ground tyrants. Eye Color The iris of most species of bush and ground tyrants is dark brown or blackish, but that of Xohuis pyrope and of Xolmis cinerea is red. This color is conspicuous in the field. Such variation docs not seem to be of great taxonomic significance, except to suggest that X. ptjrope is relatively closely related to birds of the X. cinerea species-group. Nest Site When considered together with other characters, such as color pattern, pro- portions, habitat, and perhaps behavioral traits, nest site may be of help in assessing the interrelationships of some taxa. I have made use of such correlations especially in the case of Neoxolniis rufiventris and Muscisaxicola brevicauda (see below). In the genus Agriornis, however, the variation in nest site may be of less taxonomic significance. One of the three species living in high Andean steppes {montana) builds its nest on the ground in protected sites, whereas another {microptera) does not. Finally, A. livida, found in open woodlands and scrubby vegetation in lower zones along the Andes, builds its nest in trees. Data about nests and nest sites arc given by Smith. t-labifat Preferences and Distribution As I pointed out earlier, there is both intra- and interspecific variation in eco- logical preferences among the bush and 192 Bulletin Museum of Comparative Zoology, Vol. 141, No. 5 ground tyrants. Further information is given by Smith. Three species occur in rather speciahzed habitats, and the pos- sible taxonomic significance of this fact needs to be examined. Xolmis pijrope ap- pears restricted to open woodlands and the edges of Patagonian forests dominated by trees of the genus Nothofagus. In other characters this species resembles Musci- so.xicola (absence of patterning) or XoJmis (proportions, eye color, arboreal habits, behavior). Since I have placed some weight on the transition from arboreal to terres- trial adaptations, I have maintained this species in Xolmis (in a distinct species- group) because it exhibits few of the more typical adaptations to terrestrial life found in other ground tyrants. Another ecological specialist is Neoxohnis rufiventris. This species occurs at the southern tip of South America, in open boglike grasslands or steppes with low Berheris shrubs. It apparently shares this habitat with no other species of the groups discussed in this paper. In its nest site, proportions, and general behavior, this species can only be considered a true ground tyrant. But in color pattern it resembles some of the arboreal species of Xolmis, especially of the Andean species- groups. Assuming that N. rufiventris has fully evolved to the terrestrial adaptive zone, I have maintained it in the monotypic genus erected by Hellmayr ( 1927 ) . The third species is Muscisaxicola hrevi- cauda. It lives in the lowlands of south- western Ecuador and northwestern Peru, in open habitats ranging from steppelike man-made biotopes (cotton plantations) to natural savannas and open woodlands. In habits it is partly terrestrial, like other Muscisaxicola, and partly arboreal, like other birds such as Xolmis. In pattern and behavior it is closest to Muscisaxicola, but its habitat is clearly denser. I have con- sidered it here the most aberrant member of the genus Muscisaxicola, and have placed it in a subgenus of its own. The Genus Ochthoeca Diagnosis A genus of about seven to nine species of small to medium-sized flycatchers living along the Andes in high altitude steppes, open woodlands, and, occasionally, wet montane forests. All forms of this genus have a relatively short but rather broad bill, and long and conspicuous rictal bristles (see Fig. 4). The outer primaries are not emarginated. Colors are usually browns and grays, but one species is largely greenish and another black and chestnut. All species have a head pattern with the loral spot and a superciliaiy stripe of a different color than the crown and cheeks (see Fig. 1). Most species have one or two wing bars (see Fig. 1), but this char- acter is geographically variable in at least one species (O. riifipectoralis). The under- wing color patterns, so striking in the genus Xolmis, are absent in Oclithoeca. Species of Ochthoeca have a characteristic field silhouette, with large head and convex forehead (see Fig. 1). Species Included I accept here the treatment of de Schauensee (1966: 338-340), who recog- nized nine species. These species can be arranged into four species-groups, de- scribed below. ( I believe that the following taxa, now placed rather far from Ochthoeca in the lists of Hellmayr [1927] and de Schauensee [1966], are closely related to, or even possibly congeneric with, Och- thoeca: Tu77ibezia salvini [Taczanowski, 1877], CyOloramphus parvirostris [Danvin, 1839], and Ochthornis littoralis [Pelzeln, 1868].) 1. The cinnamomeiventris species-group. The only species included, Ochthoeca cin- namomeiventris, is black with a white loral spot and supercilium. The underparts vary geographically from black to black and chestnut. Some populations of this species have conspicuously white axillary feather tufts. O. cinnamomeiventris has the broad- Evolution of Ground Tyrants • Smith and V uilleumier 193 LENGTH ofCULMEN from SKULL mm 29n o Q> 28- 27- o 99 <5) q9 o 1 26- 25- 24- DDD • • • A O o o o / 23- A 1 «• 22- 21- A ♦ 1 6 1 7 1 8 1 9 1 10 1 V o X .striaticollis ^ X.pernix • X.fumigcta ♦ X.fuscorufa ° X.rufipennis ^ X.erythropygio -, WIDTH ofCULMEN 11mm at NOSTRILS Fig. 5. Length of culmen plotted against width of culmen in the Xolm/s lumigata species-group (1) and the X. erythropygio species-group (2). Note the absence of overlap in width of culmen between the two groups. est and flattest bill of any species of Ochthoeco (see Fig. 4). This species may be the most strictly forest-inhabiting species of Ochthoeca or, indeed, of the entire group of genera considered in this paper. 2. The diadema species-group. This group contains Ochthoeca diademo, and a superspecies that includes O. frontalis and O. puIcheUa. All three species have a yellow loral spot, and a yellow or white supcrcilium. The upper parts vaiy from dark brown (in pulcliella) to greenisli (in diadema). The species of this group are forest birds, but they occupy relatively open biotopes, such as clearings, rather than the forest interior. 3. Tlie rufipectoralis species- group. Och- thoeca rufipectoralis, the only species of this group, has a buffy supcrcilium and patterned underparts, with contrasting rufous chest and pale gray lower abdomen. A similar pattern is not found in any other species of the genus. O. rufipectoralis lives in more open habitats than birds of the first two species-groups. 4. The oenanthoides species-group. The four species I include in this group (O. fumicolor, O. oenanthoides, and the two species of the leucophrys superspecies: O. leucophnjs and O. piurae) are character- ized by their being somewhat larger sized than the species of the other groups, by their relatively long and thin bill, and their brown or gray coloration. O. fumicolor is almost entirely dark rufous, O. oenantlioi- des is gray above and pale rufous below, whereas birds of the leucophrys super- species are gray both above and below. The birds of this species-group live higher, as a group, than any other in the genus. They are found in wet paramos (O. fumi- color) or in dry scrub associations of the puna (O. oenanthoides and O. leucophrys) . The Genus Xo/m/'s Diagnosis A genus of 13 to 14 species of medium to large sized flycatchers distributed along the high Andes, in the uplands and low- lands of central South America, and in 194 Bulletin Museum of Comparative Zoology, Vol. 141, No. 5 \ \ LENGTH of TARSUS mm 33- 32- 31- \^ 30- ^\ 29- 28- 27- o 26- o ^ 25- o qip 24- •^ 23- \ \ Q A A '\ D D \ 0 \ o \^ ^ \ CD \ O N s \ \ \ \ o X.striaticollis ^ X.pernix • X.fumigata ♦ X.fuscorufa ° X.rufipennis ^ X.erythropygia 8 -r 9 10 -T" n LENGTH of CLAW of HALLUX 12mm Fig. 6. Length of tarsus plotted against length of claw of hallux in the Xolwis lumigata species-group (1) and the X. erythropyg/o species-group (2). Note that tarsus and claw of hallux are proportionately longer in the two species of the erytfiropyg/a group, and are separated from the measurements of the fumigafo group. Patagonia, where they inhabit forest edges, woodlands, dense scrub, and, more rarely, open steppes. All species of Xolmis, except X. pyrope, are patterned, but the details of color and color combinations vary widely from spe- cies to species. The colors making up the patterns are white, gray, and black in some species (in the c'merea species-group), gray and brown in others (the fumigata, erijthropygio, and ruljetro species-groups). The underwing patterns are relatively uni- form throughout the genus (Fig. 2), al- though the colors can be either rufous and dark brown, or black and white, as was mentioned earlier. Species of the genus Xolmis have a long and relatively narrow bill with a terminal hook (see Fig. 4). There is interspecific variation in bill width within the fumigata species-group. All species of Xolmis, except the rather terrestrial species, X. rubetra. have moderately long and conspicuous rictal bristles. The following species have emarginated outer wing feathers: X. striati- collis, X. pyrope, X. dominicana, X. irupero, and X. rubetra (see Fig. 3). Only in the latter species is there any evidence of a sound being produced in flight, which may be attributed to this emargination (Wet- more, 1926a; see Smith). Species Included The genus Xolmis as understood here includes species formerly placed in the genera Myiotheretes Reichenbach, 1850; Cnemarchus Ridgway, 1905; Ochthocliaeta Cabanis and Heine, 1859; and Pyrope Cabanis and Heine, 1859. (One species, Xolmis mtirina, has been removed from Xolmis and transferred to Agriornis, and will be discussed below under that genus. ) The five above genera have been merged into one because neither morphological nor Evolution of Ground Tyrants • Smith and Vuilleumier 195 ecological characters permitted unequiv- ocal delimitation of more than one genus. Thus, X. pyrope cannot satisfactorily be separated from Xolmis sensu Hellmayr (1927), and X. rufipennis can just as easily be included in de Schauensee's (1966) enlarged genus Mijiothereies as in Xolmis. It is as arbitrary to shift either or both of these two species back and forth from one genus to another as it is to isolate them in monotypic genera, since neither species really departs from the species composing the genera Myiotheretes and Xolmis in de Schauensee's list. Another solution (merg- ing) is preferred here, but only if sub- divisions within the enlarged Xolmis can be expressed. This is done here by means of species-groups. 1. The fumigata species- group. This species-group includes Xolmis striaticollis, X. pernix, X. fumigato, X. fusconifa, and perhaps also X. signata.^ These species have relatively broader bills than other species of Xolmis, a whitish or grayish-colored throat with dusky streaks, unpattemed smoke-brown or rusty breast and belly, conspicuously patterned wings and tail (a given species can have either the tail or the wings patterned, or both), and relatively short tarsi and claws of hallux. The species of the fumigata species- group resemble most those of the eryth- ropygia species-group in color and,^ pat- tern, but they differ from them in having proportionately broader bills (Fig. 5), shorter tarsi, and shorter claws of hallux (Fig. 6). These and other differences in proportions are illustrated in Figure 7 by means of a ratio-diagram. ^ X. signata ( Taczanowski, 1874) is known from two specimens (fide Hellmayr, 1927). One of these is now deposited at tlie Polish Academy of Sciences in Warsaw ( Pinowski, in lift. ) . I have not been able to examine this specimen and can- not, tlierefore, make any decision al:)Out its taxo- nomic allocation. I retain X. signata in the fumi- gata species-group and, indeed, in Xolmis, imtil the specimen(s) can be critically studied. The habitats favored by the species of the fumigata species-group include cool, moist, forested or wooded slopes of the tropical Andes in the ecotone between the montane forests (cloud forests) and the grasslands above. They do not occur in high altitude treeless steppes, where Agri- ornis and Muscisaxicola live. They also seem to avoid the dark montane forest interior, where Ochthoeca cinnamomeiven- tris is found. The birds of the fumigata species-group converge both moq^hologically and in some of their behavioral traits toward thrushes of the genus Turdus. Note in particular the resemblance between X. striaticollis and Turdus spp. in color pattern and in stance when on the ground. 2. The erythropygia species- group. This species-group comprises two sympatric, probably not very closely related species, Xolmis erythropygia and X. rufipennis. They differ, as a group, from the species of the fumigata species-group, which also have streaks on the throat, by the combi- nation of: relatively narrower bill, re- latively longer tarsus, and longer claw of hallux (see Figs. 5-7). They differ from the other species-groups of Xolmis in hav- ing conspicuously patterned underwings with rufous instead of white (see Fig. 2). X. erythropygia occurs in wet habitats at the upper altitudinal limit of montane Andean forests. X. rufipennis occurs in slightly drier and more open habitats, often within the puna grassland and scrub in Polylepis woodlands (Vuilleumier, 1969c). Both species, but X. erythropygia espe- cially, have some superficial resemblance to chats, rather than to thrushes. 3. The pyrope species-group. The only species included, Xolmis pyrope, is gray above, with a whitish throat and belly and pale gray breast, without wing or tail markings, thus contrasting, in this lack of pattern, with all other species of the genus. Its red iris makes it resemble X. cinerea in this character, whereas its juvenal plumage resembles that of Muscisaxicola species in 196 Bulletin Museum of Comparative Zoology, Vol. 141, No. 5 -15 H -.10 ^-h- -.05 0 .05 .10 WING <^.^ \ ^ TAIL '^''? W TARSUS ^ CULMEN <^re not deemed taxonomically distinguish- able by Hellmayr ( 1927) and Pinto ( 1944). There does not appear to exist any potential for further speciation in the cinerea species-group. On the one hand, the five species of this group do not seem to have isolated and morphologically dif- ferentiated populations that could be con- sidered as incipient species. On the other hand, there is extensive sympatiy of several species over a large part of central South America (Brazil south of the Amazon, eastern Bolivia, Paraguay, Uruguay, north- ern and central Argentina). Sympatry, combined with the morphological distinct- ness of the five species, makes it im- possible to even attempt to reconstruct the past history of speciation in this species- group. Of the five species, X. velata, X. cinerea, and X. irtipero appear to be com- mon or relatively common, but X. do- minicana and possibly X. coronata may be scarce or even rare. Hudson (1920: 142) said of X. coronata: "Old gauchos have told me that fifty years ago they were abundant all over the pampas, but have disappeared wherever the giant grasses have been eaten down and have given place to a different vegetation." Detailed study of the ecological preferences of the species of this group should be undertaken to determine to what extent the moqiho- logical differences they show can be cor- related with habitat selection, predators, feeding habits, and nest sites. 5. The rubetra Species-Group The only species of this group, Xolmis ruhetra, does not exhibit any geographical variation. Its range includes the lowlands east of the Andes in northern Patagonia (during the breeding season) and the Argentine chaco and western pampas (in the southern hemisphere winter) (Olrog, 1963; Hudson, 1920: 148. however, thought ruhetra not to be migratory). The Genus Neoxo/m/s Neoxolmis rufiventris, the only species of this genus, breeds in southern Patagonia, where its nest has recently been found by Maclean (1969); it winters in the Argen- tine pampas ( Olrog, 1963 ) , as far north as Uruguay (Cuello and Gerzenstein, 1962: 126). There is apparently no geographical variation in this species. The Genus Agriornis 1. The monfana Species-Group Agriornis montana This wide-ranging species occurs in the paramos of Colombia and Ecuador, in the Evolution of Ground Tyrants • Smith and Vuilleuinier 209 Fig. 13. Geographical distribution of Agriornis montana. Colombian and Ecuadorean populations [solitaria] are isolated from Peruvian ones (inso/ens) by the Northern Peruvian Low and Upper Maranon barriers (see text). The populations of southern Peru, northwestern Bolivia and extreme northern Chile (b) are intermediate in tail pattern between those to the north (a) and south (c). See text for further details. 210 Bulletin Museum of Comparative Zoology, Vol. 141, No. 5 puna of the high Andes of Peru, Bohvia, and Argentina, and in open habitats along the Andes (in some locahties at relatively low elevations) all the way to southern Patagonia, exeept Tierra del Fuego (Fig. 13). A. montana shows some geographical variation in size, but an assessment of this variation was not possible beeause many specimens in my samples had molting or heavih' worn tails or wings. There is geo- graphical variation in body color, tail pat- tern, and perhaps in vocalizations. Variation in body color appears to be extensive and gradual, but I cannot say \\'hether it is truly clinal or not. The samples at my disposal consisted of speci- mens that had been collected in widely different years, and I found that individuals taken prior to about 1940 were generally darker than those taken since that time. Mor(^ uniform samples would be necessary for meaningful intersample comparisons. Geographical variation in tail pattern can be summarized as follows. Samples from southern Colombia to south-central Peru (Cuzco area) (labelled "a" on Fig. 13) have the four outermost tail feathers white, each rectrix having a white shaft. The fifth outermost feather has some gray but is still extensively white. In samples from southern Peru (Puno) the extent of white on outer tail feathers is diminished, so that the third outermost rectrix has gray on the inner web. In the La Paz Depart- ment of Bolivia the samples examined show much variation, and specimens collected at localities only a few kilometers apart may differ in their tail pattern. This situation was noticed by Hellmayr (1927: 5-6), who described the variable populations as a new subspecies, interjnedio, labelled "b" on Figure 13 (see also the comments of Niethammer, 1956: 103). In 1967-1968 I attempted to collect large series of these birds in northwestern Bolivia, but the low population density of A. montana made it impossible. Neverthe- less, among the birds I collected, two taken within 4 kilometers of each other in the Ballivian area of the altiplano southwest of La Paz city exemplify this individual vari- ation very clearly. One specimen (AMNH 793198, collected 21 October 1967) has the two outermost rectrices wholly white, the third has some grav at the base of the outer \'ane and along the edge of the inner vane, and the fourth has white only in the distal half. The second bird (AMNH 793199, collected 22 October 1967) has some gray at the base of the inner vane of the outer- most rectrix, whereas the second and third tail feathers from the outermost have much gray at the base. In both specimens, the shaft of the outermost rectrix has at least some gray: at the base in AMNH 793198, and the entire proximal half in AMNH 793199. However, other specimens I collected in the La Paz Department (such as AMNH 793202, taken along the eastern front of the Andes between Viloco and Caxata) and in the Potosi and Chuquisaca Depart- ments have the same tail pattern as birds from farther south along the Andes in northwestern Argentina and in the Andes from northern Chile to Patagonia (labelled "c" on Fig. 13). In all specimens I have examined from tliese regions, the three outermost rectrices have a white outer vane and a white tip. The fourth and fifth tail feathers are only white-tipped. In short, geographical variation in tail pattern does not seem pronounced either north of central Peru or south of north- central Bolivia. The populations living in a relatively narrow zone in southern Peru and northwestern Bolivia, around the basin of Lake Titicaca (Fig. 13), are inter- mediate between two types of tail pattern. Several genera other than Aii,riornis have species that exhibit geographical xariation in the same area, or else have allopatric taxa at or immediately below the level of species coming in contact in this zone (see Vuilleumier, 1969b). Detailed analysis of larger samples than I have seen will be necessaiy before it is possible to tnteipret the significance Evolution of Ground Tyrants • Smith and Vuilleumier 211 Table 2. Geographical variation in wing-, tail-, tarsus-, and culmen-length in AgRIORNIS ALBICAUDA Populations ( Males only ) Wing-length (mm) Range (mean) (N) Tail-length ( mm ) Range (mean) (N) Tarsus-length ( mm ) Range (mean) (N) Culmen-length ( mm ) Range ( mean ) (N) Ecuador 137.0-156.0 (143.2) (5) 115.0-132.0 (120.0) (5) 41.0-44.0 (42.8) (5) 31.0-33.5 (32.2) (5) Peru 136.0-145.0 (140.3) (7) 116.0-124.0(119.2) (7) 39.0^2.0 ( 40.9 ) (7) 31.5-33.0 (32.2) (7) Bolivia 133.0-138.0(135.5) (2) 111.0 (1) 39.5-40.5 (40.0) (2) .33.0-33.5 (33.2) (2) NW Argentina 130.0* (1) 101.0* (1) 40.0 (1) 36.0 (1) Note: The single these two ] specimen from NW Argentina has worn tips to both tail and wing measurements ( * in the table ) are probably lower than those of birds : feathers, so that in fresh plumage. of the geographical variation in A. montano. Provisionally, I would consider this variation to be a case of secondaiy hybridization taking place before repro- ductive isolation was achieved by the ex-isolates. That hybridization, rather than primary intergradation, is taking place is suggested by the individual variation ob- served in northwestern Bolivian popu- lations, which thus resemble hybrid swarms in other species. Geographical variation in voice may take place in A. montana. Birds from Ecuador (see Smith) and from Bolivia (Department Potosi, personal obsei^vation ) appear to have very similar vocalizations. On the other hand, birds from northern Argentina (Tucuman, see Smith) have slightly lower pitched and shorter calls. It would be most interesting to check, by means of additional recordings, whether such vari- ation is indeed geographical, and, if so, whether it parallels the variation in tail pattern just described, or, on the contrary, is independent of it. Agriornis albicauda This species is similar to montana in both color and pattern, including the tail pat- tern. A. albicauda differs from A. montana in having broader streaks on the throat, in being much heavier (the average of three albicauda being 82.3 grams, and of six montana being 60.7 grams), and in having a thicker, more powerfully hooked bill with pale honi-colored lower mandible (the bill of montana is entirely black). A. albicauda is the only species of the genus Agriornis living strictly in the high Andes, where it occurs from Ecuador in the north to north- ern Chile and northwestern Argentina in the south. There is some geographical variation, as pointed out by Zimmer (1937: 2-3). He stated that Ecuadorean birds are larger and darker than Peruvian ones. In order to check whether the variation in size might be discontinuous, I have measured speci- mens from Ecuador, Peiii, and Bolivia. As can be seen from Table 2, there is very little variation in culmen-length, and only a slight cline of decreasing size from north to south in wing-, tail-, and tarsus-length. Agriornis alJ)icauda and A. montana are too extensively sympatric for any inference to be made about their past history, assum- ing, as I do here, that they originated from the same ancestral stock. The Andes could have been invaded at two different times by an early stock, or else speciation could have taken place in situ in the Andes. The two species share the same habitats: open 212 Bulletin Museum of Comparative Zoology, Vol. 141, No. 5 p ^ A. microptera andecola "^ A. microptera microptera (winter range) I A. microptera microptera (breeding range) ♦ A.livida Fig. 14. Geographical distribution of Agriornis livida and A. microptera. A. livida and A. microptera are fully al- lopatric. Wintering populations of A. microptera microptera overlap slightly with resident ones of A. microptera andecola in northwestern Argentina. Evolution of Ground Tyrants • Smith and Vuilleuinier 213 slopes and valley floors of high Andean Agriornis microptera valleys, and areas having sparse and xeric Agriornis microptera has two series of vegetation of low shrubs, with scattered populations, in Patagonia and in the high rocks and boulders used as observation Andes respectively, separated from one posts. In Bolivia, I found that montona another by a hiatus of over 1000 kilometers outnumbered albicaiida about ten to one (Fig. 14). The Patagonian birds (sub- in such habitats. In northern Chile, how- species microptera) breed in open brushy ever, A. albicaiida was "more plentiful" steppes from Neuquen and Rio Negro in than A. montana (Johnson, 1967: 233). the north to Santa Cruz in the south; they do not occur on Chilean territory or on the 2. The livida Species-Group island of Tierra del Fuego. This fact is Agriornis livida interesting, because it seems to suggest that ... ..11 A. livida (which does occur on Tierra del The distribution of this species mcludes p^^ ^^ ^^^ ^ microptera are replacing the dry lowlands west of the Andes in ^^^^ ^^^^ ecologically. This supposition ChUe from Atacama (up to about 2000 ^^^^^^^ confinned when the distribution of meters altitude, Goodall et al., 1957: 141) ^^^ ^^ ^p^^,.^^ .^ examined in Patagonia southward to about 40-41 S. In this ^^^^^^^ ^^^^j^. ^ ^.^.^^^ .^ restricted to region of south-central Chile, the Andes ^^^ ^^^^^^ ^^^^^^ foothills, whereas are much lower and are less of a barrier ^^^^^^^^^^^ i, fo^^d in the lowland steppes than they are farther north, and con- ^^^^^^^ ^^^^ ^^^^^^^ ^^^ Atlantic Ocean, sequently, A. livida is found m the moun- ^j^^ Patagonian populations of A. mi- tains, especially in dry valleys lying m ram ^,^p^g,„ lea^p theij- breeding grounds shadows (the Lonquimay Valley, tor ^^^^^^ February and migrate northward example). South of about 41 S, A. livida ^^ ^^.^ nonbreeding quarters in northern does not occur west of the Andes because Argentina and southern Bolivia (see Fig. this zone is clothed by Nothofagus forests, ^^^ Nonbreeding Patagonian A. microp- which constitute unsuitable habitats; it oc- ^^^^ ^^^^ ^^^^^ collected in the following curs only along the dry eastern foothills of Argentine provinces : Mendoza (March), the Andes, mostly on Argentine territory g^|^^ (March, April, May), Tucuman (see Fig. 14). A. livida thus appears to ^^^^^ j^^^ j^j^^^ j^^^^ (^^y) Santiago avoid both very xeric and very wet habi- ^^j g^^^^^ ^j^jy^ ^^^ ^^.^^ (August) tats: its center of abundance, accordmg to (specimens examined personally), Cor- Hellmayr (1932) and Goodall et al. (1957), ^i^^^^ Buenos Aires, and Santa Fe (accord- is in central Chile, from Aconcagua to Bio- ^^^ ^^ Hellmayr 1927). Bio, and again in south-central Patagonia '^^le high ' Andean populations of A. in Chubut and Aysen. microptera (subspecies andecola) are sup- Geographical variation in A. livida is p^^^^i ^^ ^^ resident. Their geographical minor, and apparently involves only wing- ^.^^g^ includes northern Chile ^(Goodall length and tail-length (as pointed out by ^^ ^j 3^957. ;^43) northwestern Argentina, Hellmayr, 1932, and as confirmed by my BoIj^I^ a„^| probably southern Peru. In examination of series of specimens). The northern Chile, Johnson (1967: 232) found samples seen, however, are not sufficiently a. microptera in tola (Baccharis) steppes, large for one to conclude whether this where it was "fairly plentiful." On the variation is clinal or not. The populations Bolivian altiplano in 1967-1968, I saw living on Tierra del Fuego, although iso- microptera in similar habitats, but only lated geographically from other Patagonian twice during several months of field work, birds, seem not to be moq^hologically and can only conclude that the population differentiated. density of the species is low. 214 Bulletin Museum of Comparative Zoology, Vol 141, No. 5 m. maculirostns Fig. 15. Geographical distribution of Muscisaxicola maculirostris. The two peripheral isolates (niceforoi in Colombia and rufescens in Ecuador) are separated from each other by an ill-defined barrier (marked ?), whereas Ecuadorean and Peruvian populations ore isolated by the Northern Peruvian Low. Evolution of Ground Tyrants • Smith and Vuilleiimier 215 The populations of A. microptera found in northwestern Argentina seem to be com- posed of migratory birds from Patagonia during part of the year, and of resident ones during the breeding season. The mixture of populations between March and July appears confusing. The only definitely breeding specimen I have seen so far is MCZ 99388, an adult male in worn plumage collected 12 December 1918 at Penon, 4000 meters, in Catamarca. Other specimens collected at lower elevations seem to be migrants: for example, those collected bv Steinbach at Cachi, 2500 meters (Salta), in March and April 1905. Olrog (1949b: 153), however, collected a specimen in the Aconquija mountains (Tucuman) at an altitude of 2900 meters in May, and allocated this bird to the local, breeding subspecies andecola. This identi- fication suggests that the Andean popu- lations breed at very high altitudes during the rainy season between November and March, then move to lower elevations, where they encounter nonbreeding indi- viduals that have migrated from Patagonia. More data are needed to fully elucidate the breeding and nonbreeding ranges of Andean "residents" in northern Argentina. In spite of the wide geographical gap between breeding Patagonian and Andean populations of A. microptera, morpho- logical differentiation between the two is minor, and involves the color of streaks on the throat and thickness of the bill (Hell- mayr, 1927: 4). 3. The murina Species-Group Agriornis murina, the only species of this group, breeds in xeric scrub in Argentine Patagonia (Rio Negro, Neuquen, and Chu- but). It migrates northward to the low- lands of northern Argentina, southern Bolivia, and western Paraguay during the nonbreeding season. There does not seem to be any geographical variation in A. murina. The Genus Muscisaxicola The Subgenus Muscigralla The similarities and differences between Muscisaxicola Inevicauda and the species of the subgenus Muscisaxicola and my reasons for considering them all congeneric were given in detail in Part 1. M. brevicauda occurs from southwestern Ecuador along the foothills of the Andes and the Pacific coastal plain of Peru to extreme southern Peru. The highest alti- tude at which the species has been col- lected is 1100 meters at Yamana (Loja), Ecuador (R. A. Paynter, Jr.). Marchant (1960: 372-373) found it "commonest in open country with or without scattered bushes," and often in "surprisingly thickly wooded places." My own observations in southern Ecuador and northern Peru agree with Marchant's. M. brevicauda shows no detectable geographic variation in color, but birds from southern and central Peru (Lima, Pisco, Nazca) are a little larger in wing- length and tarsus-length than those from the Guayas Province, southwestern Ecua- dor, and Piura, northern Peru. The Subgenus Muscisaxicola 7. The maculirostris Species-Group Muscisaxicola maculirostris This species is the most widespread of the genus, ranging from the Bogota region of the Eastern Andes of Colombia south- ward along the Andes all the way to Tierra del Fuego (Fig. 15). It is found in a variety of habitats, which all have one thing in common: they are quite xeric, or even semi-desertic. Altitudinally maculi- rostris has a wide range, from about 1000 to about 4000 meters (and occasionally above ) . As might be expected from this geo- graphical range, there is geographical variation in this species. Color xariation is not clinal, but sharply discontinuous. Ecuadorean birds (subspecies rufescens) 216 Bulletin Museum of Comparative Zoologij, Vol. 141, No. 5 altitude m 4000- 2500- 1000- o o o o o o o o P o o o o •Ecuador and N. Peru oS. Peru and NW. Bolivia ^N. Patagonia A A ^ i A 78 — I — I — r 85 — 1 wing length 91 mm Fig. 16. Altitudinal variation in wing-length in Muscisaxicola maculirostris. Andean populations from Ecuador southward to Bolivia show an increase positively correlated with altitude, but Potagonian birds do not. are very buffy, even pale rufous, on their underparts. In contrast, birds from north- cm Peru to Argentina and Chile are all very uniform (maculirostris), and of a much paler and more isabelline tone than Ecuadorean birds. This variation is shown in Figure 15. Interestingly, and perhaps unexpectedly, the birds from Colombia (niceforoi) do not resemble the geographi- cally close Ecuadorean birds, but are more like those from Peru to Argentina and Chile, much farther south. Geographical variation in size as mea- sured by wing-length is somewhat ob- scured by altitudinal variation (see Fig. 16). Birds from southern Peru and north- western Bolivia live very high (3000 to 4000 meters) and have long wings. Birds from northern Peru and Ecuador live at lower elevations ( 2000 to 3500 meters ) and have shorter wings. There is, however, no correlation between altitude and wing- length in birds from northern Patagonia. They live at low elevations (below 1500 meters) but have long wings. Geographical variation in wing-length does not seem to parallel geographical variation in color: there is apparently no more abrupt change in wing-length be- tween southern Ecuador and northern Peru (see Fig. 16) than between other ap- parently not disjunct populations (e.g., Bolivia and northwestern Argentina). Muscisaxicola fluviatilis This is the only species of the genus to live in wet tropical lowlands and Andean foothills east of the Andes in Peru and Bolivia, and along streams and rivers in extreme western Brazil. However, there are in the literature a few reports of M. fhwiatilis at high altitudes in the Andes, which are worth examining in detail. Bond (1947: 130) synonymized M. titi- cacae (Carriker, 1932: 459) with M. fluvi- atilis. The type and only specimen of titicacae is an adult from Desaguadero along the shore of Lake Titicaca, at about 3900 meters. I have seen this specimen and can confirm its identification as fluviatilis. WING LENGTH <^ 90 Mainland — Breeding RIO NEGRO(2) SOUTHERN CHILE (4) Mainland — Migrants CENTRAL CHILE(7) argentina(i) Falkland Islands- Breeding (3) Evolution of Ground Tyrants • Smith and Vuilleumier 217 100 110 115 -f WING LENGTH ? 90 Mainland — Breeding RIO NEGRO(2) SOUTHERN CHILE(3) Mainland —Migrants CENTRAL CHILE a.lz) b.(6) argentina(i) t Falkland Islands-- Breeding(3) 100 110 115 mm + Fig. 17. Geographical variation in wing-length in mainland and Falkland Islands populations of Musc/saxico/o macloviana. In both moles and females, Falkland Islands birds have longer wings than do mainland birds, whether the latter are breed- ing or migrants. Number of specimens in parentheses. Females from Central Chile: a, with MCZ 94525, labeled " 9 ," but probably wrongly sexed; b, without this specimen. Olrog (1963) mentioned that fhwiatilis occurs in the highlands of northwestern Argentina. Through Olrog's courtesy, I examined four of the five specimens upon which this statement was based. Two of these five birds had been identified as M. fhwiatilis by Hellmayr (Olrog, in lift.). The four specimens I studied are un- doubtedly M. maculirostris, and I presume that the fifth belongs to this species also, since it is identical to them ( Olrog, in lift. ) . These specimens show not only that great caution must be exercised when identifying fluviatilis, but also that the species does occasionally occur high up in the Andes. There is apparently no geographical variation in M. fluviatilis. 2. The macloviana Species-Group Muscisaxicola macloviana Muscisaxicola macloviana breeds in con- tinental southern South America, from about 41 °S, along the Andes southward to Tierra del Fuego, and in the Falkland Islands. Olrog (1948) noticed this species in flat country, along forests, and on sea- shores, as well as above the timber line in Tierra del Fuego. In the Falklands, Pet- tingill (personal communication) remarked that it was a "common resident in upland areas, usually in valleys near cliffs or rocky outcrops." He noticed further that "its particular habitat seemed to be governed by the availability of crevices in loose rock or ledges for nesting and grassy places, sheltered from the wind, for feeding and resting." The mainland and Tierra del Fuego birds (subspecies mentalis) are very uni- form in color and size, but have noticeably shorter wings than birds from the Falkland Islands {macloviana) (Fig. 17). I did not detect any parallel color variation. M. macloviana is migratory. The main- land populations move northward along the Andes in both foothills and uplands, and spend the southern hemisphere winter in Peru, where, according to Koepcke (1964), the species is regularly found in loose flocks in the coastal lomas. The Falk- land Islands populations, in contrast to the mainland ones, seem to be resident, al- though neither Bennett ( 1926) nor Cawkell and Hamilton (1961) say so explicitly. The resident status of Falkland Islands birds seems confirmed by an examination of wing-lengths among continental migrants, none of which were large enough to belong 218 Bulletin Museum of Comparative Zoology, Vol. 141, No. 5 to the Falkland Islands populations (Fis;. 17). Muscisaxicolo capistrata The breeding range of this species in- cludes extreme southern Chile and Tierra del Fuego, but very little is known about its breeding habits and ecological preferences. Neither Goodall et al. (1957) nor Olrog (194(S) give any data about its habitat. Muscisaxicoh capistrata migrates north- \\'ard along the Andes at high elevation, and spends the southern hemisphere winter in northern Chile, Bolivia, and southern Peru. There is no geographic variation in this species. 3. The rufivertex Species-Group Muscisaxicolo rufivertex Muscisaxicola rufivertex breeds from northern Peru to central Chile and adjacent Argentina. In Chile (Goodall et al, 1957), this species vmdertakes altitudinal mi- grations and spends the nonbreeding season in the coastal regions of Coquimbo, Ata- cama, and Antofagasta. In Argentina, Boli\'ia, and Peru, I found M. rufivertex to inhabit xcric slopes with a rocky or gravelly soil and sparse shrubby vegetation, and rocky quebradas or small valleys with cliffs and rocky ledges. Dorst ( 1962 ) found rufivertex where small valleys open up into pampas, so that the ecological char- acteristics of grassy plains and rocky slopes of small valleys are combined. M. rufi- vertex has a rather extensive altitudinal distribution from about 3000 meters (Koepcke, 1964) up to about 4500 meters ( personal observation ) . Muscisaxicola rufivertex has some geo- graphical variation in color and size. The detection of geographic variation is prob- ably complicated by migratory movements of several populations. Zimmer (1937: 3) stated that "one specimen from Cucha- cancha, Bolivia, belongs to palUdiceps although eleven other specimens from the same locality are occipitalis. The speci- mens were all taken in June, and there is little doubt that the individual belonging to pallid iceps was a migrant from the south." Variation seems to be as follows: north- ern Peruvian birds are the largest, whereas birds from Chile (Santiago region) are smaller. The birds from northern Chile, northern Argentina, and southern Bolivia are distinctly paler than those from central Chile and adjacent Argentina, or than birds from Peru and northern Bolivia. The palest birds live in northern Chile and northern Argentina. The albilora Superspecies Muscisaxicola juninensis. This species is found apparently only at very high alti- tudes from central Peru to Boli\ ia, northern Chile, and northwestern Argentina, where it seems to be resident. It inhabits grassy steppes interrupted by rocky outcrops and small cliffs, above 4000 meters, and usually between 4200 and 4S00 meters (personal observation ) . Muscisaxicola albilora. Muscisaxicola al- bilora breeds in the Andes of Chile and Argentina from about 30 °S to southern Patagonia, but not in Tierra del Fuego. It occurs at altitudes from about 1500 to 2500 meters (Goodall et al., 1957). It seems to prefer rocky, barren slopes with almost no vegetation, at least in Rio Negro, Argentina ( personal obsen^ation ) . There is no detectable geographical varia- tion in this species. Migrations are ex- tensive, since M. albilora vacates its breeding grounds (Johnson, 1967) and spends the southern hemisphere winter in the high tropical Andes, as far north as Ecuador. 4. The alpina Species-Group The alpina Superspecies Hellmayr (1927) separated the various populations of Muscisaxicola having gray plumage and lacking a colored crown-patch into two species: Muscisaxicola alpina, Evolution of Ground Tyrants • Smith and Vuilleumier 219 Muscisaxicola alpina superspecies back dark gray-brown back gray bock pale gray grisea (alpina semispecies) • ^ cinerea(cinerea semispecies) Fig. 18. Geographical distribution of the Muscisaxicola alpina superspecies. Dark-backed populations [alpina semispecies) have isolates in Colombia and Ecuador. In western Bolivia (and perhaps also in Peru: see text) the alpina semispecies comes in contact with pole-backed populations [cinerea semispecies). Insert: collecting localities, largely from personal field work, showing the contact area. Birds of the grisea phenotype [alpina superspecies) occur along the eastern front of the Andes. The cinerea phenotype [cinerea superspecies) occurs along the oltiplano as well as along the eastern Andean front. Localities marked with an asterisk (*) represent possible overlap. 220 Bulletin Museum of Comparative Zoology, Vol. 141, No. 5 ranging from Colombia southward to Ecua- than the Andes of Peru-BoHvia, and espe- dor, Peru, and Bohvia in the department of eially those of northern Argentina and cen- La Paz, and M. cinerea, ranging from tral Chile (Lauer, 1952; Franze, 1927; northwestern Argentina to central Chile. Knoch, 1930). One might therefore con- In a footnote (1927: 28) he indicated that elude that geographical variation in color M. alpina is larger and darker, has whiter follows Gloger's rule. edges to secondaries and wing coverts, and Geographical variation in size, especially has a broader supraloral streak than M. as measured by wing-length, is extensive, cinerea. In a later publication, Hellmayr but does not follow exactly the same trend (1932: 125-126) merged the two species as variation in color. There is an increase into one, M. alpina. His main justification in size between Colombian and Ecua- for this taxonomic decision was that the dorean birds, the latter being the largest populations from northwestern Argentina of any populations of this group. From (Tucuman) were intermediate between Ecuador southward to Peru and Bolivia, M. alpina and M. cinerea: in size, these there is a decrease in mean wing-length Argentine birds were similar to M. cinerea, among the samples I have examined. This but in color they resembled M. alpina. decrease continues into northern Argentina Subsequent authors have followed Hell- and central Chile, so that Chilean birds mayr's 1932 treatment of alpina and cinerea are the smallest of any in the group. On as conspecific (see Olrog, 1963: 246; Phi- the whole, however, the north to south lippi, 1964: 141; Goodall et al., 1957: 164; decrease in wing-length cannot be said to Johnson, 1967: 244-245; de Schauensee, be clinal, because of some irregularities, 1966: 336). which are worthy of careful examination. Geographical variation in the alpina- Indeed, within samples from central and cinerea complex is detectable in both color south-central Peru and Bolivia, variation and size, and can be described as follows, in wing-length is sharply bimodal, with There is a decrease in color saturation (of part of the birds in a sample having rela- both upper and underparts) from Colom- tively small wings and part having larger bia southward to Argentina and Chile, wings. Furthermore, the smaller birds also Samples from Colombia and Ecuador are appear to be paler than the larger ones, uniformly dark, especially on the upper- thus suggesting sympatry of differentiated parts, whereas samples from Peru and populations. Bolivia are distinctly paler, and finally, The published evidence bearing on this samples from northwestern Argentina and problem is as follows. In Huancavelica, central Chile are palest. Although this south-central Peru, Morrison ( 1939 ) col- variation is gradual, there is a clear-cut lected both large-dark and small-pale birds, break between Ecuadorean and Peruvian His six large-dark specimens (four males samples. This break corresponds to an area and two females) were collected between between the two countries where the Andes 12 and 25 October. As I judge from the are too low for birds of this group to occur, data on labels, the six males have relatively and where the dry to arid valleys, includ- large testes, but the females have small ing the Marafion valley, cut across the ovaries. All six specimens have slightly Andean ranges and are a barrier to gene \\'orn plumage. Taxonomically, the six exchange between birds living on either large-dark birds were assigned to the sub- side (Fig. 18). The geographical variation species grisea (locally breeding), and the just described appears to be positively cor- two small-pale birds to the subspecies related with rainfall. As a whole, the cinerea (migrants from Chile). It is note- Colombian and Ecuadorean Andes receive worthy that both taxa were collected in more rain, and have longer rainy seasons, October, at a time when local Andean Evolution of Ground Tyrants • Smith and Viiilleumier 221 populations normally start breeding. Thus, either the two specimens of cinerea were late in beginning their southward migration to their breeding quarters in Chile, or else they were actually breeding in Peru. Bond (1947: 130), reporting on birds collected in Peru by Carriker, noted that four small-pale specimens, which he as- signed to the subspecies cinerea, "are adult, and were taken from April 10 to May 16." Since these dates correspond to the non- breeding season of many high Andean birds (Dorst, 1955, 1962; personal observ^ation ) , Bond concluded, I think correctly, that these individuals are "winter residents." These remarks, however, would not apply to the birds collected by Morrison (1939), cited previously. In Bolivia, Bond and de Schauensee (1942: 340-341) reported that griseo (large-dark) and cinerea (small-pale) had been taken together at three different localities: Viloca (La Paz), Cerro del Juno (Cochabamba), and Padilla (Chuciuisaca). A specimen of cinerea "was taken on the Cerro del Juno, October 2, where grisea was nesting at this time." This record is similar to the one published by Morrison (1939). In Bolivia, as in Huancavelica, it seems surprising that birds should remain as late as October in their winter range. Some individuals may stay longer than most wintering birds, as is well known from migrants in the northern hemisphere. Similarly, in the postbreeding season, some migrants are earlier arrivals than the bulk of the populations. Thus, the "three full- grown but immature specimens [of cinerea] . . . taken at Viloca [Department of La Paz], March 25 and 26" (Bond and de Schauensee, 1942: 341) may have been among the first wave of migrants. I believe that immature individuals of the genus Miiscisaxicola may be migrating earlier than adults. Thus, I have observed im- mature M. capistrata, a species breeding in southern Patagonia, in central Chile on 19 March and in northwestern Argentina on 25 April. By analogy, the arrival of im- mature birds from Chilean populations of cinerea in Bolivia in March would there- fore not be surprising. Two alternative explanations are possible for the above data. The first is that the small-pale populations ( taxonomically as- signed to Chilean cinerea) found in March, April, May, and October in Peru and Bolivia are migrants and do not breed in these last two countries. The second is that, even though the Chilean populations (cinerea) may be migratory, there are two moq^hologically differentiated populations of the alpina species-group living sym- patrically in parts of the Peruvian and Bolivian Andes. The second alternative implies that M. alpina (as understood, for example, by de Schauensee, 1966) consists, in fact, of two largely allopatric species having a narrow zone of overlap in the central Andes. Between October 1967 and January 1968, I investigated the relationships of the populations of M. alpina in the Bolivian Andes. I traveled on the altiplano south- west of the Cordillera Real, and made several trips along the Cordillera Real, the Cordillera de Cochabamba, and, farther south still, in the Cordillera de Mochara. I visited these areas because they were located immediately around, as well as in, the region of possible sympatry of two differentiated populations of M. alpina. I wanted to determine in what habitats these birds occur, to collect samples of specimens in breeding condition, and to delimit the ranges of the two populations. Unfortunately, much more time was spent in just traveling large distances on ex- tremely difficult roads looking for these birds in habitats which appeared suitable, than in actually collecting and studying these birds. The extreme shyness of Miisci- saxicola, combined with their low popu- lation density even in the most suitable localities, was a powerful obstacle to col- lecting large comparati\'e series, which I had hoped to assemble. Furthermore, great physical difficulties were encountered in 222 Bulletin Museum of Comparative Zoologij, Vol 141, No. 5 Table 3. A comparison of morphological characters in populations of the MUSCISAXICOLA ALPINA SUPERSPECIES FROM THE BOLIVIAN AnDES M. alpiua semispecies ^f. cincrea semispecies Character Range (mean) (samijle size) Range ( mean ) ( sample size ) Mensural characters Weight in giams (males and females) 24.9-27.9 (26.61) (7) Weight in grams ( males only ) Wing-length ( mm ) ( males only ) Tail-length (mm) (males only) Culmen-length ( mm ) ( males only ) Tarsus-length (mm) (males only) 25.8-27.9 (26.92) (5) 115.5-122.0(119.50) (5) 78.5-90.5 (82.80) (5) 17.5-19.0 (18.30) (5) 32.0-33.5 (32.80) (5) Nonmensural characters Superciliaiy stripe Underparts Upperparts Wing coverts Whitish supercilium extends slightly beyond the eye Alidomen white, tinged with pale buffy Breast gray, darker than the whitish throat Uniform dark grayish l^rown ( mouse-gray or mouse-l^rown ) Brownish gray, without buffy wash 17.5-23.3 (20.55) (14) 18.9-22.7 (21.06) (8) 108.0-114.0 (111.87) (8) 72.0-78.0 (75.18) (8) 17.0-18.5 (17.93) (8) 28.0-30.0 (29.37) (8) Whitish supercilium does not ex- tend lieyond the eye, at least in most specimens Alidomen whitish, tinged with pale gray or grayish brown Breast and throat pale gray, almost concolor Uniform grayish with distinct pale brownish tinge in most specimens Brownish with pale buffy wash the pursuit of these ekisivc birds at alti- tudes usually over 4000 meters. In spite of these difficulties, however, my efforts proved successful. I collected a total of 21 specimens of the alpina group, as well as other species of Muscisaxicola. This number is small, when compared to the yield of similar field endeavors in localities which are easier of access. Yet, the results of this exploratory collection permit the following unequivocal con- clusions. (A) Two populations of the olpina spe- cies-group occur in the high Andes of Bolivia (Departments La Paz and Cocha- bamba). One of them comprises large and relatively dark birds, hereafter referred to as "gr/.sert phenotype." The second con- tains small pale birds, referred to below as "cinerea phenotype." The birds of the grisea and cinerea phenotypes that I collected in Bolivia can be distinguished from each other by a series of characters (see Table 3). Weight alone permits complete separation, since there is no overlap in weight ranges be- tween the grisea and the cinerea specimens. In other mensural characters, birds of the grisea phenotype are consistently larger than those of the cinerea phenotype. In Evolution of Grou>jd Tyrants • Smith and Viiilleiimier 223 color characters, birds of the griseo pheno- type appear generally darker than those of the cinereo phenotype. (B) Birds of both the grisea and the cinerea phenotypes breed between October and January. Although no nest could be located, the specimens collected include adult males with enlarged testes and fully ossified skulls for both phenotypes. The females collected did not have enlarged ovaries, however. No nestlings were col- lected, but two juvenile specimens of grisea had largely unossified skulls, either very small or unobserv^able gonads, and loose, very fresh plumage. All the other specimens, of both grisea and cinerea phenotypes, had worn plumage, especially remiges and rectrices. Food carrying and display behavior, including display flights, were seen from birds of the cinerea pheno- type in October, November, and Decem- ber. (The conditions of the gonads of the other three species collected, M. rufivertex, M. jiininensis, and M. alhifrons, were similar to those of the birds of tlie grisea and cinerea phenotypes.) (C) Birds of the cinerea phenotype oc- cur both on the altiplano and along the high mountains forming the eastern An- dean wall (Cordilleras Real and of Cocha- bamba) (see Fig. 18). Birds of the grisea phenotype, however, were encountered only in the Cordilleras Real and of Cocha- bamba (see Fig. 18). ( D ) In one of the ten localities at which I collected specimens of the grisea or cinerea phenotypes, I observed both to- gether. Along the road to Morochata, at about 4040 meters altitude in the Cordillera Tunari ( Department Cochabamba ) , I en- countered a sparse population of Musci- scixicola on gentle, grassy slopes. The birds were foraging in the densely matted, grazed grass or turf, and were very wary. Much stalking and patience permitted me to secure five specimens within a radius of about one kilometer. Two of these birds proved to be of the cinerea phenotype. They had testes 9.0 X 4.5 and 9.0 X 4.0 mm respectively, fully ossified skulls, and slightly worn plumage. The other three specimens were of the grisea phenotype. All three were males, with testes and skull conditions as follows. Testes 5.0 X 3.0 mm, skull with two tiny unossified windows; testes 6.0 X 3.0 mm, skull fully ossified; testes 4.0 X 2.5 mm, skull with a moderately large unossified occipital window. All three had slightly worn plumage, like that of the birds of the cinerea phenotype. The evidence presented above suggests that two allopatric species or semispecies come in contact along the eastern wall of the Andes of Bolivia (Department Cocha- bamba). An unequivocal decision on whether or not the taxa involved are spe- cifically distinct cannot be made from the present, still insufficient, data. More in- formation is needed on the respective distri- butional ranges of the grisea and cinerea phenotypes, on possible differences in ecological preferences, in displays or other behavior, and on the relationships in nar- row zones of overlap, as in the Cordillera of Cochabamba. Do the two taxa hybrid- ize? Do they have intra- or interspecific territories, and so forth. The taxonomic solution adopted here, pending further field work to clarify the situation, is that birds of the alpina species- group constitute a superspecies, with two component semispecies, nomenclaturally treated as species: M. alpina (including grisea), living in the wetter Andes of Co- lombia, Ecuador, Peru, and northwestern Bolivia (see Fig. 18), and M. cinerea (in- cluding argentina) living in the drier Andes of central and southern Bolivia, northwestern Argentina, and central Chile (see Fig. 18). 5. The albifrons Species-Group The albifrons Superspecies Muscisaxicola albifrons. Muscisaxicola albifrons is usually found at higher alti- tudes than are most other species of the genus, except perhaps jiininensis. In Chile, 224 Bulletin Museum of Comparative Zoology, Vol. 141, No. 5 Goodall et al. (1957) give the altitudinal range from 4000 to 5000 meters. Their photograph (page 163) shows the habitat of the species in Chile: gently undulating slopes with sparse and overgrazed scrub and bunch-grass. In Peru and Bolivia I saw the species on valley slopes where grazed grasslands were interrupted by cliffs and boulders, between 4300 and 4500 meters. M. alhifrons ranges from northern Peru to extreme northern Chile and Bolivia; it has not been reported from northwestern Argentina. As far as is known it is a resi- dent species. There is no geographic vari- ation in this species. Miiscisaxicolo flaviniicha. Muscisaxicola flavimicha is a southern South American species ranging from central Chile and adjacent Argentina to Tierra del Fuego. It occurs up to about 4000 meters, according to Goodall et al. (1957), but in the Andes south of about 40° S its altitudinal limits are lower, because the Andes themselves rarely exceed 3000 to 3500 meters. The habitat of this species consists of barren rocky slopes with small cliffs or boulders, and very little herbaceous vegetation. Goodall et al. (1957) say that fhvimicha occurs especially near streams and moun- tain lakes, but in my experience, it also breeds in regions devoid of either streams or lakes. M. flavimicha is migratory and goes as far north as northern Peru (Libertad) to spend the southern hemisphere winter. This migration seems to take place entirely at high altitudes in the Andes. There is minor geographical variation in M. flavimicha. Olrog (1949b) described the birds of Tierra del Fuego as a distinct subspecies {brevirostris) because they have smaller bills and wings, and are darker in color than mainland birds. From an evo- lutionary point of view, M. flavimicha has one weakly differentiated isolate on Tierra del Fuego. It should be pointed out that although the type locality of brevirostris is Ushuaia (Tierra Grande, Tierra del Fuego), Olrog (1949b) did not think that these birds bred on this island, but rather on the mountains of Navarino and Hoste Islands and in the Cape Horn archipelago. These populations are probably rare or un- common. It was not until 1928 that Reynolds ( 1934 ) observed and collected this species in Harberton, Beagle Canal. Later (1935) he reported that he might have seen it on Freycinet. Muscisaxicola frontalis This species breeds in Chile (Anto- fagasta) and Argentina (Mendoza) in the north and from there southward to the latitude of Colchagua and Rio Negro (Johnson, 1967). Bond and de Schauensee (1942: 340) mentioned four specimens from Viloca and La Cumbre in Bolivia, and Bond (1947: 129) cited one male collected 21 April at Pampa de los Arrieros near Arequipa in southern Peru. I agree with Bond that these birds are migrants from the south. The breeding habitat of M. frontalis seems to be stony and rocky slopes with sparse vegetation of small shrubs and bunch-grass. During the non- breeding season, the species occurs also at very high altitudes, either in bunch-grass steppes or in rocky spots near marshy areas (personal observation). M. frontalis does not show any detectable geographic vari- ation. DISCUSSION The bush and ground tyrants analyzed in Part II show clear-cut instances of various stages of the speciation process. Thus, at least the most recent events in the adaptive radiation of these birds can be traced. Yet, the earlier episodes of their evolutionary history appear shrouded in uncertainty. Indeed, these flycatchers have a rather large number of both taxonomically iso- lated species, and taxonomically closely related but extensively sympatric species. Table 4 shows this evolutionary dichot- omy. Of thirty-three species and semi- species, nine (about 27 percent) possess Evolution of Ground Tyrants • Smith and Vuilleumier 225 Table 4. Stages of the specl^tion process in the Bush and Ground Tyrants Geographical variation Species and semispecies Absent Gradual Discontinuous ( Isolates slightly- differentiated ) Member of a Super- species Second- ary contact No close relative Closest relative broadly sympatric Xolmis striaticollis — X. pernix — X. fumigata + X. juscorufa + X. signaiar + X. enjthropijgia — X. rufipeimis + X. pyrope — X. cinerea — X. velata — X. coronata — X. dominicana — X. irupero — X. rubetra + Neoxohnis rufiventris + Agriornis inoniana — A. albican da — A. livida — A. microptera — A. mtirina + Muscisaxicola brevicatida — M. macuUwstris — M. fluviatilis M. macloviana M. capistrata M. nifivertex M. junmensis M. albilora M. alpina M. cinerea M. albifrons M. fJavinucha M. frontalis Totals: 33 + + + + + + 13 + + + + +' + + ? + + + + +? + +* + + + — + +? _ + + i + i + ( + +' +' — + j + 8 ^ 9 9 4 + + 6 + + + + + + + + + + + + + + 14 1 X. striaticollis shows what appears to be a stepped cline in southern Peru and northern Bolivia. The interpretation as "secondan.- contact" is hypothetical. ^ X. signata may or may not belong in Xolmis. 2 X. pyrope shows gradual geographical variation on the continent, but has a subspecifically distinct isolate on Chiloe Island. * A. montana shows a phenomenon similar to that of X. striaticollis above. ^ The two semispecies of the M. alpina superspecies appear to overlap in a very- narrow zone of the Bolivian Andes. Semispecies belonging in the same superspecies are enclosed in braces. morphologically differentiated isolates, nine (about 27 percent) are members of super- species, and four (about 12 percent) have what appear to be secondary contacts sug- gesting former isolation. At the same time, of thirty-three species (regardless of whether they also show active speciation or not), SLX (about 18 percent) are taxo- nomicallv isolated, and fourteen (about 42 percent) are broadly sympatric with their most closely related congener(s). These two aspects of the evolutionary' history of these flycatchers are analyzed further below. 226 Bulletin Museum of Comparative Zoology, Vol. 141, No. 5 Presently Active Speciation One of the most important events of geographical speciation is the estabhshment of isolates within a species' range (see, e.g., Mayr, 1963). The corollaries of geograph- ical isolation are the cessation of gene flow between populations, and the subsequent initiation of genetic divergence in the isolates under different selection regimes. The presence of genetically distinct iso- lates is therefore an index to the speciation potential of extant species (Prakash, Le- wontin, and Hubby, 1969). In the absence of actual data on the genetics of populations of Andean birds, including the Tyrannidae of the present study, I have of necessity relied on morpho- logical differentiation to make statements about genetic differentiation. My con- clusions about speciation are therefore based on the assumption that morpho- logical differentiation between two isolated populations is positively correlated with their genetic differentiation. That this is indeed the case is suggested by the work of Hubby and Throckmorton (1968) on isozyme variability among triads of closely related species of Drosophila. Nevertheless, genetic differentiation can be extensive even in the absence of parallel morphological variation. Thus the bias of my study would be to underestimate genetic differentiation: first, by consider- ing that morphologically similar isolates are also genetically similar, and second, by failing to recognize minor, but real, geo- graphical variation. The first source of bias may prove very difficult to erase. The second, however, could be eliminated by undertaking ex- haustive studies of geographical variation of each of the species treated here, espe- cially, perhaps, those considered to have no, or only gradual, variation (see Table 4). As every systematist knows, such studies are extremely time consuming. Since my goal was to obtain broad com- parative data on speciation from a large sample of the entire Andean avifauna (see Vuilleumier, 1969b), I have deliberately chosen breadth at the expense of depth, and can only hope that the many problems uncovered will be taken up as separate projects by others. The observed instances of presently active speciation fall in two categories: (a) differentiation centered around more or less well-marked geographical and eco- logical barriers, and (b) differentiation and secondary contacts not centered around such barriers. Geographical Barriers In two species and one superspecies of Xolmis, one species of Agriornis, and four species of Miiscisaxicoki, morphological differentiation can be correlated with the presence of barriers. In all these cases the isolates have been formally recognized taxonomically by previous ornithologists, usually as subspecies, but as species in the X. fumigata superspecies. From an evolutionary point of view, the barriers isolating these populations would therefore seem to inhibit gene flow from other populations, as evidenced by the moiphological differentiation. These bar- riers are listed below in geographical sequence from north to south (several of them have already been cited in Vuilleu- mier, 1968, 1969ai 1969b). ( 1 ) Cesar Depression. The lowlands sepa- rating the Santa Marta Mountains of north- ern Colombia from the Perija Range seem to affect Xolmis pernix and X. fnmigato, two species or semispecies of the X. fumi- gata superspecies (see Fig. 12). In this case, morphological differentiation is suf- ficient to suggest that species level has been reached by the populations on either side of the gap. In Xolmis erytluopygia, a population appears restricted to the Santa Marta Range; it has been separated taxo- nomically as a subspecies {orinoma). The species is absent from the Perija Range, and also from the remainder of the Colom- Evolution of Ground Tyrants • Smith and Vuillenmier 227 bian Andes, except in the extreme south (Nariho). Whether this patchy distri- bution means that extinction has occurred in the Andes between Nariiio and Santa Marta or whether it is an artifact reflecting insufficient collecting (as a result of the rarity of the species?) cannot be resolved at the present time. Xohnis striatic oil is has an apparently undifferentiated population in the Santa Marta Mountains. (2) Rio Torhes V alley. This arid valley isolates the Andes of eastern Tachira, Merida, and Trujillo from those of western Tachira in Venezuela. The populations of Xohnis fumigata living in the Andes to the east of this valley are clearly dis- tinguishable moi-phologically from other populations, and have been separated taxo- nomicallv as the subspecies liigubris (see Fig. 12)'. (3) Rio Magdalena Valley. The Central and Eastern Andes of Colombia are iso- lated from each other by this dry valley. Muscisaxicola alpina has an isolated popu- lation in the Central Andes ( taxonomically separated as the subspecies colomhianu) and another in the Eastern Andes (sub- species qttesadae) (see Fig. 18). The mor- phological differentiation between these populations is slight. (4) Upper Rio Magdalena Valley and Soidhwestern End of Eastern Andes. The southwestern part of the Eastern Andes appears too low for paramo vegetation to occur, so that there is a distribution gap between the Andes of Nariiio in southern Colombia and the central part of the Eastern Andes. Muscisaxicola maculirostris has a well-differentiated isolate ( subspecies niceforoi) in the Eastern Andes, isolated by a very large gap (marked by ? in Fig. 15) from the Ecuadorean population (sub- species rtifescens) . The problem of whether or not the actual barrier corresponds to a gap in the vegetation preferred by this species in Colombia remains to be estab- lished in the field, hence my ? in Figure 15. A similar gap exists in the distribution of Muscisaxicola alpina (marked with a ? in Fig. 18). (5) Upper Rio Maraiion Valley and North- ern Peruvian Loiv. In the Andes of north- em Peru, the Rio Maraiion Valley cuts deeply across the relatively low-lying mountains as its course changes from a south-north to a west-east direction. The Andes west of this bend are especially low, and are probably unsuitable for many high altitude birds, because the upper vegeta- tion zones, especially those above the timber line, are either absent or of very small area. This complex barrier separates populations of Xolmis fumigata (see Fig. 12), with slight morphological differenti- ation only. In Muscisaxicola, however, two species have well-marked populations on either side of this gap. In M. maculirostris, the subspecies rufescens occurs north of the barrier, and maculirostris sovith of it (see Fig. 15). In M. alpina, the northern subspecies is alpina and the southern grisea (see Fig. 18). Finally, Agriornis montana has weakly differentiated populations sepa- rated by this gap (see Fig. 14), taxo- nomically recognized as subspecies solitaria (north) and insolens (south). Thus the Maraiion gap has various effects on fly- catchers living in this area, but is it note- woithy that none of the populations so isolated belong to a superspecies. (6) Rio Apurimac Valley. The upper Rio Apurimac and some of its tributaries (Rio Pampas, Rio Mantaro) may constitute a barrier between Xolmis fumigata and X. fuscorufa of the X. fumigata superspecies (see Fig. 12), and between the weakly marked subspecies erythropygia and boli- viana of Xolmis erythropygia. In both cases, however, much more collecting is necessary before the precise barrier area between taxa can be ascertained. ( 7 ) Andes of Western Argentina. The huge hiatus (over 1000 km) in the distribution of Agriornis microptera (see Fig. 13) is difficult to interpret in terms of a single 228 Bulletin Museum of Comparative Zoology, Vol. 141, No. 5 Table 5. Small peripheral isolates that show differentiation in morphological characters Species or suijerspecies having isolate Xolmis fumigata Xolmis fumigata Xolmis enjthiopijgia Xolmis ptjrope Agriornis moniana Muscisaxicola maculiwstris Muscisaxicola maciilirostris Muscisaxicola macloviana Muscisaxicola alpina Muscisaxicola alpina Muscisaxicola flavinucha Geographical isolate ( taxonomic rank ) Location of isolate liigubris ( subspecies ) pernix ( species or semispecies ) oriiio))ia ( subspecies ) fortis ( subspecies ) solitaria ( subspecies ) niccforoi (subspecies) rufescens ( subspecies ) macloviana ( subspecies ) Columbiana ( subspecies ) quesadae (subspecies) brevirostris ( subspecies ) Merida, Venezuela Santa Marta, Colombia Santa Marta, Colombia Chiloe Island, Chile Ecuador and southern Colombia Eastern Andes, Colombia Ecuador Falkland Islands Central Andes, Colombia Eastern Andes, Colom1)ia Tierra del Fuego "barrier." The habitats favored by this species seem to be zones of xeric, rocky plains, which are largely missing in the Andes, and the foothills between the ranges of the subspecies ondecola (high Andes) and microptera (Patagonia). Range dis- junctions similar to the one found in A. microptera are also observed in Pterocne- mia pennata, Tinamotis, and Diiica. (8) Straits Between Chiloe Island and Mainland. The Chiloe Island population of Xolmis pyrope appears slightly differen- tiated from the mainland ones. (9) Straits of Magellan. The population of Muscisaxicola flavinucha living on Tierra del Fuego or on islands south of the Beagle Canal may be slightly differentiated morphologically. More field work is needed to check whether the variation is truly discontinuous (and if so, whether the barrier is indeed the Straits of Magellan) or merely clinal. (10) Straits Separating Falkland Islands From Mainland. The Falkland population of Muscisaxicola macloviana is slightly differentiated from the mainland ones (see Fig. 17). From an ecological standpoint, the ten gaps listed above are quite variable. Aside from barriers 8-10, which are marine, all others are terrestrial. Barriers 2, 3, 5, and 6 may be roughly equivalent ecologically. They are situated around arid valleys cut- ting across high mountain ranges. Barrier 1 is a more varied lowland gap. Finally, barriers 4 and 7 are ill-defined interrup- tions in the continuity of certain habitats. Of the moqohologically differentiated isolates separated by one or another of barriers 1-10, it is interesting to note that eleven may be considered peripheral and of small size (see Table 5). These isolates may have originated as small founder populations that colonized suitable habitats at the peripheiy of the range of the species (see Mayr, 1954). Eight of the eleven isolates occur north of barrier 5, in the Andes of Ecuador, Colombia, and Vene- zuela, where the distribution patterns of high altitude habitats, especially paramos, are quite insular (see Vuilleumier, 1970). The potential importance of peripheral isolates as incipient species has been em- phasized by Mayr (1963). Recently, Prak- ash, Lewontin, and Hubby (1969) have demonstrated that genetical differentiation is marked in the small and isolated periph- eral population of Drosophila pseudooh- scura in Colombia. Absence of Barriers I have argued that in Xolmis striaticoUis and Agriornis montana (see Fig. 14), and Evolution of Ground Tyrants • Smith and Vuilleumier 229 in the Muscisaxicola alpina superspecies (see Fig. 18), secondary contacts are re- sponsible for the observed patterns of dis- tribution and variation: hybridization or secondary intergradation, and narrow mar- ginal overlap. Although these interpreta- tions need to be substantiated by more evidence, it is possible to state here that these contacts do not seem to take place in regions that can be called barriers. All three contact zones are located in the Andes between southern Peru and north- western Bolivia. In both the Muscisaxicola alhilora and M. alhifrons superspecies, a distribution gap appears to exist between the pairs of species or semispecies in the Andes of ex- treme northern Chile and of western Bolivia. Whether these gaps are real or artificial, resulting from the lack of inten- sive collecting in these areas is, of course, not possible to ascertain at the present time. In the alhilora superspecies, the two semispecies (or species) are quite similar morphologically, and secondary contact might conceivably result in hybridization. In the alhifrons superspecies, however, the differences between the two taxa are well marked, and contact might result in sym- patry (as it apparently does in the alpina superspecies). All of the supposed instances of second- ary contacts rest, unfortunately, on too few specimens having sufficient label data and on too few field data. The best field evidence, in the M. alpina superspecies, comes from my own exploratory field work, so I am better aware than anyone else of the limitations of the speculations ad- vanced here. Older Speciation Patterns The taxa of bush and ground tyrants that do not show evidence of active speciation fall in two categories: (a) taxonomically isolated species, and (b) taxonomically related, sympatric species. In the first group belong Xolmis pyrope, X. ruhetra, Agriornis murina, Neoxolmis rufi- ventris, and Muscisaxicola hrevicauda. Some of these species (e.g., Xolmis pyrope, Neoxolmis rufiventris, and Muscisaxicola hrevicauda) occupy relatively specialized habitats, as I pointed out in Part I. Nothing can be inferred of the history of these species on present evidence. In the second category (sympatric con- geners ) belong especially the species of the Xolmis fumigata, erythropygia, and cinerea species-groups, those of the Agriornis mon- tana and livida species-groups, and the species of the subgenus Muscisaxicola. Sympatry in Xolmis If the Xolmis fumigata superspecies is considered as one geographical species, it is broadly sympatric with X. striaticollis, X. erythropygia, and X. rufipennis. These species show various degrees of habitat differentiation, so that habitat co-occu- pation may only occur between X. fumigata and X. erythropygia, whereas X. striaticollis and X. rufipennis are ecologically more distinct. The precise degree of habitat overlap between these species, however, is a problem needing future field work. In any event, this extensive sympatry makes it impossible to deduce anything about past speciation within the Andean species- groups of Xolmis. In the cinerea species-group, sympatry is again too extensive for inferences to be drawn about the history of the group in extra-Andean open habitats. Sympatry in Agriornis Two species of Agriornis (montana and alhicauda) are not only sympatric over large areas of the high Andes from Ecua- dor to northwestern Argentina, but they can also be found in similar habitats, at least in Bolivia (personal obser\^ation ) . Furthermore, A. microptera is sympatric with both montana and alhicauda in parts of the altiplano of Bolivia and northwestern Argentina. 230 Bulletin Mmcinn of Comparative Zoology, Vol. 141, No. 5 Sympatry in Muscisaxicola The only species of the subgenus Musci- saxicoJa that lives in complete allopatr\' is M. fluviatilis, a lowland species. In the uplands and high Andes, especially south of central Peru, up to four or five species of this subgenus live sympatrically. Sympatry in Mtiscisaxicola has long been known to ornithologists, yet no more than passing mention of the phenomenon can be found in the literature. As a result, the following discussion is compiled from my own field notes. Altitudinal and ecological replacement is obvious in Ecuador between M. maciiU- rostris, a species that lives in xeric steppes and brushlands in the inter- Andean region from about 2000 to 3700 meters, and M. alpina, a larger species that occurs in moister habitats such as meadows and bunch-grass from about 3500 to well above 4100 meters. In central Peru, M. maailirostris ranges from about 1500 to 4000 meters, and broadly overlaps with M. riifivertex, which occurs from about 3000 to 4200 meters. At altitudes above 4000 meters in the altiplano of Peru and Bolivia, M. alpina (or M. cinerea), M. juninensis, M. albifrons, and locally also M. maculirostris and M. riifi- vertex occur together. Of the five species and superspecies, Af. rufivertex seems to occur lower down, whereas M. alJnfrons is found only at the highest altitudes. I have seen and collected M. alpina and M. junin- ensis (or M. cinerea and M. juninensis) in the same habitats in Bolivia, and have also observed these two species with M. albi- frons. In some localities, I have seen M. cinerea and M. rufivertex together, in others, M. cinerea and M. maculirostris. Thus true habitat co-occupancy involves usually two, but sometimes three, of the five species listed above. Further study of these sympatric associ- ations should be made to determine whether these species differ in habitat requirements, foraging sites, nesting habits, or in other details that might reduce competition among them. It is noticeable that the two semispecies of the alpina superspecies (alpina and cinerea) overlap in body size and bill size with juninensis and rufivertex. Thus food size partitioning may not be an important way of effecting ecological segregation. This preliminary discussion of sympatry in Xolmis, Agriornis, and Muscisaxicola shows that this phenomenon is quite ex- tensive in these genera. It would therefore seem that they have been present in the high Andes and the extra-Andean open regions of central South America for a long time, so long, in fact, that the early stages of their adaptive radiation have been obliterated by complex patterns of second- ary sympatry. LITERATURE CITED Bennett, A. G. 1926. A list of the birds of the Falkland Islands and dependencies. Ibis, ser. 12, 2: 306-333. Berlepsch, H. von. 1907. Studien iiber Tyran- niden. Ornis, 14: 463-493. Bond, J. 1947. Notes on Peruvian Tyrannidae. Proc. Acad. Nat. Sci. Philadelphia, 99: 127-154. Bond, J., and R. M. de Schauensee. 1942. The birds of Bolivia. Part I. Proc. Acad. Nat. Sci. Philadelphia, 94: 307-391. Carriker, M. a., Jr. 1932. Descriptions of new birds from Peru and Bolivia. Proc. Acad. Nat. Sci. Philadelphia, 87: 343-359. Cawkell, E. M., and J. E. Hamilton. 1961. The birds of the Falkland Islands. Ibis, 103a: 1-27. Chapman, F. M. 1927. Descriptions of new birds from northwestern Peru and western Colombia. Amer. Mus. Novitates, No. 250: 1-7. Cott, H. B. 1947. The edibility of birds. Proc. Zool. Soc. London, 116: 371-524. Cravvshay, R. 1907. The Birds of Tierra del Fuego. London: Bernard Quaritch, 158 pp. CuELLo, J., and E. Gerzenstein. 1962. Las aves del Uruguay. Lista sistematica, distri- bucion y notas. Com. Zool. Mus. Monte- video, 6(93): 1-191. DoRST, J. 1955. Recherches ecologiques sur les oiseaux des hauts plateaux peruviens. Trav. Inst. Fr. £tud. Andines, 5: 83-140. Evolution of Ground Tyrants • Smith and Viiilleumier 231 . 1962. A propos de la nidification hypogee de quelqiies oiseaiix des hautes Andes peruviennes. Oiseau Rev. Fr. Ornithol., 32: 5-14. Franze, B. 1927. Die Niederschlagsverhaltaisse in Siidamerika. Petermanns Geog. Mitt., Gotha, Ergiinzungsheft, No. 193, 79 pp. Friedmanx, H. 1927. Notes on some Argentina birds. Bull. Mus. Comp. Zool., 68(4): 139- 236. GooDALL, J. D., A. W. Johnson, and R. A. Philippi B. 1957. Las aves de Chile, su conociniiento y sus costnmbres. I. Buenos Aires: Piatt Establecimientos Graficos S. A., 441 pp. Hellmayr, C. E. 1927. Catalogue of birds of the Americas and the adjacent islands in Field Museum of Natural History. Part V. Tyrannidae. Field Mus. Nat. Hist., Publ. 242, Zool. Sen, 13: 1-517. . 1932. The birds of Chile. Field Mus. Nat. Hist., Publ. 308, Zool. Ser., 19: 1-472. Hubby, J. L., and L. H. Throckmorton. 1968. Protein differences in Drosophila. IV. A study of sibling species. Amer. Naturalist, 102: 193-205. Hudson, W. H. 1920. Birds of La Plata. Vol. 1. New York: E. P. Dutton, 244 pp. Johnson, A. W. 1967. The Birds of Chile and Adjacent Regions of Argentina, Bolivia and Peru. Vol. II. Buenos Aires: Piatt Estableci- mientos Graficos S. A., 447 pp. Knock, K. 1930. Klimakunde von Siidamerika, Handbuch der Klimatologie, II, Teil G., Ber- lin, Gebriider Borntraeger. KoEPCKE, M. 1958. Die Vogel des Waldes von Zarate. Bonn. Zool. Beitr., 9: 130-193. . 1964. Las aves del Departamento de Lima. Lima: Morsom S. A., 128 pp. Lauer, W. 1952. Humide und aride Jahreszeiten in Siidamerika und Afrika und ihre Bezieh- ungen zu den Vegetationsgiirteln. ..Bonn. Geogr. Alihandl., 9: 15-98. Leveque, R. 1964. Notes on Ecuadorian birds. Ibis, 106: 52-62. Maclean, G. L. 1969. The nest and eggs of the Chocolate Tyrant Neoxolmis rufiventris (Vieillot). Auk, 86: 144-145. Marchant, S. 1960. The lireeding of some S. W. Ecuadorian birds. Ibis, 102: 349-382. Mayr, E. 1942. Systematics and the Origin of Species. New York: Columbia University Press, 334 pp. . 1954. Change of genetic environment and evolution. In J. Huxley, A. C. Hardy, and E. B. Ford, eds.. Evolution as a Process. London: Allen and Unwin, pp. 157-180. . 1959. Trends in a\ian systematics. Ibis, . 1963. Animal Species and Evolution. Cambridge, Massachusetts: Belknap Press, Harvard University Press, 797 pp. Morrison, A. 1939. The birds of tlie Depart- ment of Huancavelica, Peru. Ibis, ser. 14, 3: 453-486. NiETHAXCMER, G. 1956. Zur Vogelwelt Bohviens. Teil II. Passeres. Bonn. Zool. Beitr., 7: 84-150. Olrog, C. C. 1948. Observaciones sobre la avi- fauna de Tierra del Fuego y Chile. Acta Zool. Lilloana, 5: 437-531. . 1949a. La avifauna del Aconquija. Acta Zool. Lilloana, 7: 139-159. . 1949b. Notas ornitologicas sobre la coleccion del Institute Miguel Lillo. I. Acta Zool. Lilloana, 8: 209-217. . 1963. Lista v distribucion de las aves 101: 293-302. argentinas. Opera Lilloana, 9: 1-377. Peters, J. L., and J. A. Griswold. 1943. Birds of tlie Hai-vard Peru\'ian expedition. Bull. Mus. Comp. Zool., 92: 281-327. Philippi B., R. A. 1964. Catalogo de las aves chilenas con su distribucion geografica. In- vest. Zool. Chilenas, 11: 1-179. Pinto, O. M. de O. 1944. Catalogo das aves do Brasil. 2a. Parte. Departamento de Zoologia, Secretaria da Agricultura, Sao Paulo, 700 pp. Prakash, S., R. C. Lewontin, and J. L. Hubby. 1969. A molecular approach to the study of genie heterozygosity in natural populations. IV. Patterns of genie variation in central, marginal and isolated populations of Drosoph- ila pscudoohscura. Genetics, 61: 841-858. Reynolds, P. W. 1934. Apuntes sobre aves de Tierra del Fuego. Hornero, 5: 339-353. . 1935. Notes on tlie birds of Cape Horn. Ibis, ser. 13, 5: 65-101. Schauensee, R. M. de. 1966. The species of birds of South America and their distribution. Narberth, Pennsylvania: Li\ingston Publish- ing Company, 577 pp. Simpson, G. G. 1941. Large Pleistocene felines of North America. Amer. Mus. Novitates, No. 1136: 1-27. . 1953. The Major Features of Evolution. New York and London: Columbia University Press, 434 pp. Taczanowski, L. 1874. Liste des oiseaux re- cueillis par M. Constantin Jelski dans la partie centrale du Perou occidental. Proc. Zool. Soc. London, 1874: 501-565. Todd, W. E. C, and M. A. Carriker, Jr. 1922. The birds of the Santa Marta region of Colombia: a study in altitudinal distribution. Arm. Carnegie Mus., 14: 1-611. VuiLLEUMiER, F. 1965. Relationships and evo- 232 Bulletin Museum of Comparative Zoology, Vol. 141, No. 5 lution within the Cracidae. Bull. Mus. Comp. Zool., 134: 1-27. — . 1968. Population structure of the Asthenes flammulata superspecies (Aves: Furnariidae ) . Breviora, No. 297: 1-21. . 1969a. Systematics and evolution in Diglossa (Aves: Coerebidae). Amer. Mus. Novitates, No. 2381: 1-44. . 1969b. Pleistocene speciation in birds living in tlie high Andes. Nature (London), 223: 1179-1180. . 1969c. Field notes on some birds from tlie Bolivian Andes. Ibis, 111: 599-608. . 1970. Insular biogeography in continental regions. I. The northern Andes of South America. Amer. Naturalist (in press). Wetmore, a. 1926a. Observations on the birds of Argentina, Paraguay, Uruguay, and Chile. Bull. U. S. Nat. Mus., 133: 1-448. . 1926b. Report on a collection of birds made by J. R. Pemberton in Patagonia. Univ. Calif. Pulil. Zool., 24: 395-474. . 1946. New birds from Coloml^ia. Smith- son. Misc. Coll., 106: 1-14. Wilson, E. O., and W. L. Brown. 1953. The subspecies concept and its taxonomic ap- plication. Systematic Zool., 2: 97-111. ZiMMEB, J. T. 1937. Studies of Peruvian birds. No. 26. Notes on the genera Agriornis, Muscisaxicola, Myiotheretes, Ochthoeca, Co- lonia, Knipolegus, Phaeotriccus, Fluvicola, and Ramphotrigon. Amer. Mus. Novitates, No. 93(): 1-27. Chapter II. Behavior of Muscisaxicola and Related Genera W. JOHN SMITH Abstract. Terrestrial species of tyrannic! flycatchers in tlie genera Muscisaxicola, Agriornis, and Xolmis appear to be closely related to the less terrestrial species of Xolmis and to the essentially nonter- restrial genera Ochthoeca, Pywcephalus, and Sayornis on the basis of characteristics assessable in mu- seum skins. Although the ecological span of the group as a whole is great, specific habitat preferences, foraging methods, patterns of locomotion and nesting behavior all vary along continua that transcend generic limits. Further, members of all of the genera are shown to employ in similar ways displays that are alike in various aspects of form, and tliat do not appear to have converged as the result of selection pressures operating in the semi-open to open habitats. It is concluded that all of the species, both ter- restrial and nonterrestrial, belong to one phylogenetic group. TABLE OF CONTENTS Introduction 233 Field Sites 235 Acknowledgments 236 Part I: Nondisplay Behavior 236 Habitat Preferences 236 ( 1 ) Muscisaxicola 236 (2) Xohim and Neoxolmis 237 (3) Agriornis 238 (4) Ochthoeca 239 Locomotion and Foraging Behavior ^, 239 ( 1 ) Flight and Aerial Foraging 239 (2) Perching and Locomotion and For- aging on the Ground 239 Flocking 24 1 Nest Construction 242 Part II: Display Behavior 242 11a. Muscisaxicola 242 lib. Xolmis 255 lie. Agriornis 258 lid. Ochthoeca ..___ 259 Comparative Sununary and Discussion — 261 Conclusions __ 267 Literature Cited 268 Bull. Mus. Comp. INTRODUCTION The markedly terrestrial genus Musci- saxicola is one of the most extreme products of the South American radiation of the Tyrannidae. It is not the only genus com- mitted to ground-dwelhng, but its trim, longlegged, Oenonthe-like species are a close-knit group, strikingly different in aspect from better known tyrannids such as the kingbirds (Tyrannus), which perch on vegetation and forage largely by pursuit of aerial prey. A superficial examination of museum skins supports the usual ar- rangement of taxonomic lists which places Xolmis (herein including Myiotheretes) and Agriornis close to Muscisaxicola and suggests relationship to nonterrestrial genera such as Ochthoeca, Sayornis, and Pyrocephalus. The purpose of this paper is first to describe behavior — primarily foraging behavior, habitat preferences, and display (stereotyped communicative be- havior)— in all but the last two genera, which are reported upon elsewhere ( Smith, 1969, in press, 1 and 2, and 1967). Secondly, it is to show that this information Zool., 141(5): 233-268, March, 1971 233 234 Bulletin Museum of Comparotice Zoology, Vol. 141, No. 5 supports the apparent phylogenetic re- lationship of all of these genera. As subjects for a comparative study of behavior patterns, the terrestrial tyrannid genera appear promising, largely because of their open habitats. Easily seen birds are more easily studied than are birds ob- scured by dense vegetation. Thus, when I began in 1959 a series of comparative be- havioral studies of a wide variety of tyran- nids, I included a program on this group. Of the many tyrannids with which I even- tually worked, however, these proved among the most difficult for reasons which had not been anticipated. In the case of Muscisaxicola, for instance, most of my field work was in the Andes of central Chile. In early spring, several species were grouped into loose flocks at accessible altitudes, frequenting widely scattered good foraging sites where melt- water collected and where a carpet of green vegetation was present. When higher slopes became more open, they moved up in pairs and scattered. In these sparse populations there was relatively little dis- play behavior. Most work in my first spring was on steep slopes where a bird could put itself out of sight by flying 50 meters, leaving me scrambling to locate it again. For my second spring in Chile I found better sites, but had less time avail- able and lost most of that to an unseason- ably late snowstorm that closed access to higher altitudes and brought the birds back down into their loose flocks. An additional problem of work in early spring in central Andean valleys was high wind in at least the early morning hours when birds are most active. This wind made tape recording difficult, yet the re- cording was essential because accurate distinctions among the various calls in the repertoires of Muscisaxicola species cannot be made by the human ear. Similar problems were met in briefer attempts to study members of this genus in areas other than central Chile, although reasonably dense populations of Muscisaxi- cola macuUrostris were found in Ecuador, and of M. brevicauda in Peru. Within the other genera, both Xolmis pijropc and Agriornis livicla were much rarer as breed- ing birds in areas accessible from Santiago, Chile, than I had surmised from the litera- ture. Xolmis pijrope was more common on Tierra del Fuego, along with Muscisaxi- cola capistrata, although neither formed dense populations and both were in late phases of the breeding cycle when I was there. Again, Tierra del Fuego is almost continuously swept by high winds, hinder- ing tape recording. Elsewhere, in northern Argentina both Xolmis and Agriornis, and in southern Brazil Xolmis species were not found in dense populations suitable for study of their displays. It seems possible that at least Agriornis species may custo- marily have veiy thinly scattered popu- lations (.see Goodall et al., 1957). Finally, I spent little time in regions where Oclithoeca species occur. Under suitable conditions, there are ad- vantages to comparative behavioral studies of several species in one season. Yet, with relatively undemonstrative species at low population densities, this tactic yields in- complete information. The alternative of studying one or two species at a time re- quires considerable financial resources and many seasons. In the present case, where the end was taxonomic and the field work was, for the most part, expected to confirm conclusions derived from less expensive methods, large expenditure's did not appear justified. Accordingly, I de\'oted what field time I could allot to continuing a broad approach, even after initial experi- ence showed this would yield quite incom- plete results. Granted our nearly complete ignorance of these birds, even incomplete comparative information could be expected to provide useful perspective. I believe the results aid in making phylogenetic interpretations and justify the choice of tactic. It should be noted, however, that Evolution of Ground Tyrants • Smith and Vuilleumier 235 the observed sample of display behavior is too small to be of more than limited use in studies of communication per se. The behavioral information is insuffici- ent, by itself, to support the taxonomic views implied by my assignments of spe- cies to genera herein. This information is not intended to stand alone, but to be considered with Vuilleumier's work (Chap- ter 1, which includes a formal taxonomic conclusion ) . This presentation is divided into two main parts; the first comprises nondisplay behavior: habitat preferences, foraging methods, methods of locomotion, and nest structure and placement. This is primarily to facilitate generic comparisons, and much detail about species has been omitted or stated briefly in a general form. While the data at hand would permit much more complete descriptions in some cases, useful species comparisons require even more in- formation, gathered under a greater variety of circumstances. The second part deals with display be- havior, and all available data judged to be useful are presented. Because these data are limited, however, formal names have been assigned to the vocal displays only provisionally, and by comparison with the more thoroughly studied genus Sayor- nis (Smith, 1969, and in press, 1 and 2). These names are often reduced after their first appearance to initials (e.g.. Simple Vocalization becomes SV)- Illustrations of vocalizations are based on analyses (Sona- grams ) made with a Kay Electric Company 6061-A Sona-Graph. In all figures the ordinates represent frequency marked in kilocycles per second, and the abscissae duration in tenths of seconds. In order to illustrate better the detail in the relativelv brief vocalizations of the Muscisaxicola species, the temporal scales have been doubled (with corresponding halving of the frequency scales) with respect to the species of the other genera. FIELD SITES (1) Chile: Most sites were in the central Andes, in two regions: {a) Lagunillas, about 60 kilometers east of Santiago, a relatively well-\'egetated region centering around a ridge at about 2300 meters; and {h) farther east at the headwaters of the Rio Yeso, a tributaiy of the Rio Maipo near the Argentine border; a relatively arid, rocky region. The two main sites on the Rio Yeso were the headwater lake at approxi- mately 2500 meters, and about 3 kilome- ters downstream on a gravelly shelf where a small tributary stream descends from Cerro Morado to the east. In ad- dition, one visit was made to sites at a similar altitude north of Santiago, Farel- lones and La Parva. Aoriornis livido was studied near Los Molles, Aconcagua Province, along the coast nordiwest of Santiago. Another coastal site was west of Santiago at Isla Negra, Valparaiso Province. A few ob- servations were made on the outskirts of Santiago at Fundo lo Recabarren on Cerro Manquehue. Other Chilean sites \\'ere on Ticrra del Fuego at the farm school Las Mercedes and at Cameron sheep station, respectively north and south of Bahia Inutil. (2) Argentina: Limited work was done in southern Tucuman Province, near La Cocha, and in the cordillera of western Tucuman above Tafi del Valle (about 2200 to 2500 meters) and in the pass of Inficrnillo ( 3040 meters ) . A brief stop was made for Xolmis irupero on a gravelly river plain (altitude about 1400 meters) about 8 kilometers north of San Salvador de Jujuy. (3) Brazil: Observations were made in the interior plains of the state of Sao Paulo near Piras- sununga and Bebedoura, and in savannas 236 Bulletin Museum of Comparative Zoology, Vol. 141, No. 5 of the Territory of Amapa near Porto Platon. (4) Peru: My main site was in cotton fields near the coastal town of Lurin, about 35 kilome- ters south of Lima. Less time was spent in a steep-sided gully ("quebrada") of a tributary of the Rio Rimac just west of Surco (about SO kilometers east of Lima in the arid foothills at less than 2000 meters ) . (5) Ecuador: All sites were reached from Quito. The lowest were all within 5 kilometers of the town of San Antonio, an arid region on the equator about 25 kilometers north of Quito on the northern slopes of Cerro Pichincha. Another set of sites ranged from the pass between Cerro Pichincha and Cerro Ataeaso about 5 kilometers southward in the paramo zone along the east face of Ataeaso (over 3000 meters), as well as eastward, down from the pass about 4 kilometers to the upper reaches of good montane forest. Finally, two mornings were spent south of Machachi in the paramo on the western slopes of \^olcan Cotopaxi, at 3800 meters. ACKNOWLEDGMENTS I am deeply grateful to Dr. Patricio Sanchez, my host and closest colleague in Chile, for aiding enormously in all of the Chilean phases of my work and for teach- ing me much about Latin America. Similar thanks are due my hosts in other countries: Dr. and Mrs. Abraham Willink, Drs. H. W. and Maria Koepcke, Dr. Paulo Vanzolini, Father J. Sartori, Mr. and Mrs. M. Piggot. and Mr. and Mrs. J. Neusenger. In addi- tion, among many persons who were very helpful, the following deserve special thanks: Dr. Martin H. Moynihan, Dr. Ernst Mayr, Dr. J. Valencia, Dr. C. Doggen- weiler, Mr. J. D. Goodall, the late Dr. R. A. Philippi B.. Sr. Luis E. Pena, Mr. A. Hundey, Sr. T. Radonich, Dr. C. C. Olrog, Dr. A. Teran, Dr. D. Lancaster, Sr. P. Leahy, Sr. P. Autino, Dr. H. Camargo, Dr. and Mrs. A. S. Rand, and Mr. L. Freire. Finally, I should hke to thank my wife, Susan, who assisted with all aspects of the field work on mv final visit to Ecuador and with the preparation of the figures and manuscript. I am indebted to various sources for fi- nancial assistance. The work began with a brief visit to Ecuador financed in part from a grant from the Chapman Memorial Fund of the American Museum of Natural History. Subseciuently the project was supported bv National Science Foundation grants NSF-G 19261 and NSF-GB 2904. The following institutions have also pro- \'idcxl help in various ways: L^niversidad Catolica de Chile; Instituto Miguel Lillo de Tucuman, Argentina; Casa Humboldt in Lima, Peru; Secretaria da Agricultura in Sao Paulo, Brazil; Industria e Commercio de Mineros in Amapa, Brazil; Harvard University; and the University of Penn- sylvania. PART I. NONDISPLAY BEHAVIOR Habitat Preferences (1) MuscisaxicoJa. Among the most ter- restrial of all tyrannids, these birds inhabit open country in the Andean chain and southern tip of the continent (for a zoo- geographic account see Vuilleumier, Chap- ter I). Some (e.g., M. capistrafa, M. mac- loviano, and M. maculirostrls) usually fre- quent grass or other low vegetation, often among scattered bushes, while others (e.g., M. cinerea, M. fhivinucha, and M. aJhUora) breed in less well- vegetated areas, usually among rocks. Rocks or bushes in all eases offer the birds commanding perches up to a meter or more tall; sometimes riparian trees are used. The three smallest (M. maculirostris, M. hievicauda, and an eastern Andean one I have not seen, M. fhwiatilis) inhabit rela- tively low altitudes for the Andean species. Evolution of Ground Tyrants • Srnith and Vuilleumier 237 Their habitats regularly include bushes and even trees. In coastal Peru, M. Ijrevicaiida lives in cotton fields, preferring sites with poor growth and foraging from the bushes or on the ground. Most of its natural habitats in Peru are also brushy ( Koepcke, 1954, 1958), although on Ecuador's Santa Elena peninsula, Marchant (1960) found it in open countiy with or without bushes, as well as in "surprisingly thickly wooded places." M. viacidirostris inhabits the lowest alti- tudes of any Muscisaxicola species in the Chilean and Argentine Andes, along the upper edge of the zone of brushy vegeta- tion among bushes up to 2 meters tall (where these are not contiguous), and ex- tending above into rocky areas with fewer, much shorter bushes. In Tucuman Prov- ince, Argentina, the one pair found was in a grazed quebrada with few bushes and a riparian line of trees 8 to 10 meters tall; the birds perched within and on top of the trees. Ecuadorean sites varied from open, sparsely vegetated, steep, eroded slopes to brushy quebradas and fields of short com, comparable to the cotton fields of M. hrevicauda. In early spring, all Andean species seen in central Chile frequented upland mead- ows and river flats with short green vegetation and copious meltwater, or foraged in mud beside melting snow. Such sites are ephemeral, and through most of the breeding season, most species have xeric sites with dusty soil and sparse vegetation. I have seen M. cinerea, M. al- J)dora, M. macidirostris, and M. capistrata foraging on the banks and gravel bars of rivers and streams, and feeding by water may be common in other species (e.g., M. fluviatdis) . Goodall et al. (1957) remark that most Andean species occur near lakes, although they also occur elsewhere. Both Vuilleumier and I noted a tendency for dorsal coloration and/or saturation of color to match general characteristics of the surrounding habitat. Species inhabit- ing high rock deserts are grayest, those of lower, grassier areas are brown (grass is dry and brown through much of the time the birds are present). The two most southerly species and M. alpina in the wet paramo of Colombia and Ecuador are deepest and richest in color, and live in regions of relatively high rainfall. Similar color correlations are known from other tyrannids (e. g., Ttjronnus, Smith, 1966) and probably reduce the conspicuousness of birds which spend most of their time perched very much in the open. (2) Xolmis and Neoxolmis. (a) While ter- restrial by general tyrannid standards, most Xolmis species are much less so than are Muscisaxicola flycatchers. Most perch above the ground and live among bushes or trees commanding a view of grass or swamp vegetation. X. pyrope and the more northern species east of the Andes often inhabit edges of woods or such brushy regions as the Brazilian cerrados, although some, like X. velata, are found on more open savannas. Species in the Andean foothills will live in brushy stream cuts where the surrounding countryside is open. Even X. coronata, which winters on the pampas, "seldom" runs on the ground (Hudson, 1920: 142) and often perches on the tall grasses; this has been confirmed for individuals wintering near Rosario, Argentina, by Dr. Juan Delius (personal communication). Wetmore (1926) reports X. cinerea, X. coronata, X. irupero, X. pyrope, and X. ruhetra as occurring among and perching on bushes. Even the most terrestrial species, X. ruhetra and N. rufiventris, do not usually breed in fully open plains. Olrog (1959) gives the breeding habitat of the latter as shrubby open country ("campos arbusti- vos"). P. S. Humphrey (personal com- munication) found its principal breeding habitat to be "flat or gently rolling country, dotted with . . . shrubs about a foot or so" tall, and it was also common in rolling country with short grass tussocks. Maclean (1969) found the breeding habitat to be 238 Bulletin Museum of Comparative Zoology, Vol. 141, No. 5 "flat, open, wind-swept pampa dotted with occasional Berheris bushes." There is less tendency for dorsal color- ation to match the habitat in this genus than in Muscisoxicolo. X. inipero, for in- stance, is a conspicuous, exposed, white bird. Others are more cryptic. For ex- ample, X. pijrope and X. cinerea are dark- gray and live in fairly wooded sites. X. velota is paler gray, overlaps with the darker X. cinerea, and tends to be in more open habitats. The most richly colored species, N. riifiventris, is again the south- ernmost. (b) I have had very little field experi- ence with members of the fiimigata species- group of Xolmis. Pairs of X. jumigata were seen in Ecuador where plantations of pine or eucaly]^5tus bordered pasture land on steep slopes at about 3000 meters. They perched in the lower limbs of trees or on telephone lines, often about 7 meters above the ground. Koepcke (1954) reports that X. striaticoUis prefers more densely vege- tated areas (around woods and small groups of trees) than do species of Agrior- nis and Miiscisaxicola in the western Andes above Lima. Todd and Carriker (1922) in the Santa Marta region of Colombia found this species from about 2500 to 3600 meters, also along the edges of forest and in some more open sites with "stunted trees and shrubbery." Olrog (1959) says that X. striaticolUs inhabits brushy quebra- das and larger river valleys up to about 2500 meters in northwestern Argentina. In Ecuador and Argentina, Vuilleumier found the species in "open sites, but in Ecuador also on steep slopes where there were scattered bushes about 2 meters tall, and in eucalyptus plantations by pastures and streams." Todd and Carriker (1922) also found the rare X. pernix at the edge of forest, and in "shrubbery." A single Xolmis (ap- parently related to X. fusconifa and to X. pernix, see below) was seen by my wife and me on Volcan Cotopaxi in dense brushy paramo up to about 2V2 meters tall in a steep-sided quebrada which extends far up into the grassy paramo at about 4000 meters. The habitat of X. erytliroptjgia is given by de Schauensee ( 1964 ) as "scrubby open slope" in paramo. From all accounts, the habitat prefer- ences of this species-group seem to be like those of other Xolmis and the higher alti- tude species of Ochthoeca. (3) Agriornis. Agriornis livida at Los Molles was in what was perhaps a natural habitat before the introduction of domesti- cated grazing animals to Chile. This sub- arid countryside had large patches of bushes, cacti, and bromeliads. Between the patches were relatively small but usually continuous areas of sparse, short, dry grass about 4 to 7 centimeters tall. In the same region the species was also on edges of relatively open, grazed fields with smaller patches of bushes and cacti not more than 3 to 10 meters apart. Perhaps the present scarcity of A. livida in Chile is due to the opening of the environment. On the pass of Infieniillo in Tucuman Province, Argentina, the closely related A. montana was in arroyos at about 2200 to 3000 meters (I did not search higher). In these were small streams running amid grass, scattered rocks, and a few bushes. Open, grazed grasslands separated the arroyos, dividing pairs from one another. These grasslands had few rocks and thus lacked elevated perches. In Chile, Goodall et al. (1957) report this species on rela- tively barren but rocky Andean slopes, and X'uilleumier has found it mostly in rocky areas. On Mt. Pichincha in Ecuador we found A. montana in a very open habitat at the lower, shallow end of a large quebrada. There was no grass, and the one indi\'idual ran on bare, dry soil much like a Miisci- saxicola species; it also perched and peered from the tops of scattered agaves and bushes. Other Agriornis (probably also montana, see below) individuals were found higher in the same quebrada on Mt. Pichincha, and in another, higher quebrada. In both Evolution of Ground Tyrants • Smith and Vuilleumier 239 sites the slopes were well vegetated with bushes, bromeliads, agaves, and grasses, and appeared similar to the coastal A. livida habitat. Similar birds were found in even lusher habitat at about 4000 meters on the northwest slopes of Volcan Cotopaxi amid tall, dense grasses. On the whole, the habitat preferences of this genus probably fall between those of northern lowland Xohnis species and the larger Muscisaxicolo. (4) Ochfhoeco. Some dark-colored, rela- tively low altitude members of this Andean genus (e. g., O. cinnamomeiventris, Vuil- leumier's observations) live in dense forests, others (e.g., O. diadema and O. nifipectoralis) in second growth or edge habitats at intermediate elevations, and species such as O. fumicoJor primarily in the transition between brushy and grassy paramo. The relatively p)ale O. leucophrys inhabits arid, brushy, stream valleys over a wide range of altitudes. Most species apparently do not depend on open habitat and usually forage within and among trees or bushes. Locomotion and Foraging Behavior (1) Flight and Aerial Foraging. While fly- ing at least several meters, a Muscisaxicolo alternately flaps and folds the wings, using just one or very few flaps before each very brief fold. The flight is strong and rapid, and not undulating. In all but the longest flights the birds do not usually ascend far above the ground. Xohnis species have a similar flight, although they usually begin with a downward plunge from their higher perches and continue close to the ground. In both genera the end of a flight is usually a rapid glide on folded wings, the bird turning abiiiptly upward on opened wings to land. In short flights, birds of either genus may flap without intennittent folding, but still use the terminal glide. The flight of Agriornis is similar, with fewer wing foldings. Members of this genus look relatively heavy-headed and short-winged in flight, more thrushlike than the previous species. Both Ochthoeca rufipectoralis and O. fumicolor have slightly undulating flights with bursts of rapid beats and brief wing folding. In at least O. nifipectoralis a conspicuous pump of the tail came at the beginning of each burst of flapping, and occasionally a whirring sound came during a burst. In O. fumicolor whirring was much louder in intraspecific fights and in one probably interspecific situation, but was not heard under other circumstances. All of the species do some aerial fly- catching, and when perched, survey the air. In most cases the choice of foraging method appears to be based on availability of prey, but some species seem more prone to aerial flycatching than others. For the very terrestrial Muscisaxicola species most pursuit flights are steeply upward, but some extend laterally 20 to 35 meters, as in much more aerial genera such as Ttjrannus. I have seen Muscisaxicola macidirostris, M. bremcauda, and Och- thoeca fumicolor flight glean (hover in front of a twig or flower head and pluck nonflying prey). (2) Perching and Locomotion and Forag- ing on the Ground. In country without woody vegetation, the species of Muscisaxi- cola customarily perch stiffly upright on clods of earth, stones, or larger rocks. Where bushes or even trees are present, these provide additional vantage points. After scanning briefly from one point, most individuals dash to another, usually within 3 meters. These rapid dashes may employ running, hopping, or both. Running seems common only if a bird goes at least a meter, and only on quite level, unbroken sub- strate. Larger species may run much more often than the smaller species, and, also in keeping with the findings of Kunkel (1962) for other passerines, all species seem to use both modes. Even a running dash may begin and end with rapid hopping. I have confirmed some observations by tracking 240 Bulletin Museum of Comparative Zoology, Vol. 141, No. 5 the birds in licfht snow or fine earth, show- ing that even extremely rapid dashes may employ only hopping. Some species (e.g., M. rufivertex) often make a short flight instead of a dash, while others (e.g., M. jhvinucha) almost always dash. The relati\ely long tarsi of these species have been thought to permit greater speed in rimning. That running often gives way to hopping on broken substrates, however, suggests that long tarsi are probably also important in making long hops. Pettingill (personal communication to Vuilleumier) comments that long legs permit M. mac- loviono to stand high and thus see farther above its grassy habitat in the Falklands. When a Musclsaxicola species sees prey, it approaches by a quick flight or dash. Prey was often discovered at the base of plants where litter had gathered. Individu- als of at least M. flavinucha and M. front- alis sometimes proceeded from plant to plant, tossing litter aside with vigorous sideways swipes of the bill and peering at the site after each swipe. Feet and legs were never used for throwing litter. In M. olhilora and M. maculirostris, in- dixiduals have been seen to flash their wings open as they leaned forward after a dash or a flight toward prey. Sometimes this appeared to prevent them from falling forward by helping to check their speed. At other times, particularly if the bird did not lean forward, a very slight wing flick looked like an intention movement of flight (seen in M. albiloro, M. moculirostris, M. frontalis, and M. rufivertex, and by Vuil- leumier in M. macloviana). At yet other times, however, the movement in M. macu- lirostris appeared to be neither balancing nor intention. It was then ver)^ conspicu- ous, and might be repeated two or three times while standing still after a dash. In these cases it may have been stereotyped for conspicuousness, and probably func- tionally convergent with the "wing flash- ing" of other ground foraging birds like the mockingbirds (e.g., Mimus polyglottos, with conspicuous white wing markings, see Selander and Hunter, 1960; or Mimus gilvus, which lacks conspicuous wing mark- ings, see Haverschmidt, 1953). Convergent evolution of wing flashing techniques has been shown by Cade (1962) for the North- ern Shrike, Lanius excuhitor, which some- times forages on the ground, and similar techniques should perhaps be expected in other ground foraging birds — with or with- out conspicuous wing markings. Hailman (1960) has reached a similar conclusion, which he does not restrict to species that forage on the ground. Muscisaxicola species frequently use in- tention movements. Most (perhaps espe- cially M. flavinucha or M. frontalis) may "scissor" the tail open and closed on land- ing, or before and/or at the end of a dash. At least M. frontalis rarely gives a very brief, small amplitude wing flick with the tail movement. The frequency of tail-scis- soring varies greatly, even in one individual over a few minutes, being most common if the bird appears nei"vous, or if other birds are foraging close by. As it often exposes otherwise inconspicuous white outer webs of the outermost rectrices, it makes a good flock signal (see Moynihan, 1960), and may occur primarily during flocking. One M. rufivertex often flicked its half- fanned tail up-down at the end of a dash while foraging alone, usually with a very brief, small amplitude flick of its wings. Wing-flicking, not correlated with tail movements, \\'as seen in a foraging M. moculirostris. Each Xolmis species observed spent much less time on the ground than did species of Muscisaxicola. As Hudson (1920) says, these birds are "ground- gazers"; they usually perch on tall ^'antage points and peer downward, dropping to the ground if they see prey, or changing perch. Wetmore's ( 1926 ) observations of at least X. cinerea and X. irupero agree. I have seen X. irupero hover above the ground where no bushes were available for perch- ing. Most species hunt on the ground at least Evolution of Ground Tyrants • Stnith and Vuillewnier 241 occasionally, though. I have seen both X. pi/rope and X. velata do this, nearly always hopping slowly. Only one pair of X. pyiope dashed and paused in the fashion of MuscisoxicoJa. According to Hudson, wintering X. rubetra and Neoxolmis rufi- ventris run on the ground much more than the other species. J. Delius (personal com- munication) is familiar with the latter in winter and says that it does not make the repeated short dashes typical of Muscisaxi- cola species. Wetmore (1926: 302), how- ever, describes X. rubetra as running "swiftly along the ground to pause and stand with head erect." I have never seen XoJmis species take prey that was difficult to kill, but Hudson ( 1 920 ) says they typically grasp beetles or grasshoppers with their feet while strug- gling to kill them. Vuilleumier has seen X. sfriaticoJlis fly down from perches and forage in 30-centi- meters tall grass. Little is known of the habits of this or other members of the fumigata species-group, but from his ob- servations they appear similar to other species of Xohnis or to Agriornis, as might be expected from their size and structure. Foraging of Agriornis livida and A. montana is similar to that of the previous genera, and in some ways intermediate. Periods of a minute or more are spent peer- ing about from the top of a bush, brome- liad, or a rock, followed by dropping and running to another perch. Sometimes food is taken from the ground during such a dash; this has been large invertebrates in the few cases I have seen. At least A. livida sometimes gleans food from branches accessible from its perch. Goodall et al. ( 1957 ) say these species take large prey (small mammals, lizards, amphibians, and birds' eggs and nestlings), as well as in- sects. Goodall's figure of A. livida shows a bird holding a mouse under one of its large feet. Peiia ( 1961 ) recovered seeds, plants, insect remains, and a wing-bone of a small bird from stomachs of A. montana; he also observed an A. montana killing a "small mouse." Birds of this genus, even more so than those of Xohnis, quite likely grasp prey in their feet and beat it with their bills. They appear to be, as Goodall et al. claim, ecological counterparts of shrikes. The diree Ochthoeca species I have seen all often pumped the tail (up-down, as in all of the species in this paper) with a simultaneous brief, small amplitude flick of the wings. These movements came on landing or when perched, and their fre- quency increased if I made a bird nervous. O. leucophrijs and O. fumicolor both did mostly aerial foraging, and the latter several times sallied out to flight glean from the paramo grass or to take prey flying over the paramo. In 1959, while an O. fumicolor kept in the edge of a grove or worked over the paramo, an O. rufi- pectoralis worked through the same grove, hopping from twig to twig and gleaning like a Vireo. Vuilleumier has seen O. leuco- phrijs fly down to the ground for prey. Although the data are scanty, it appears that no known species of these genera is primarily an aerial flycatcher. All spend considerable time as gleaners, taking much of their food from the surface of the ground or of the vegetation. Flocking In at least central Chile, while lush early spring sites are available and snow-covered higher slopes are not, large numbers of Muscisaxicola flycatchers may forage to- gether. The species of large flocks inter- mingle, to some extent, with less abundant species clumping amid the more numerous individuals of other species. I have never seen a fully stable flock, and there is al- ways some intra- and interspecies aggres- sion. Birds of larger species often supplant individuals of smaller species; individuals of different species that are about the same size occasionally fight vigorously. As higher sites open and the flocks thin out, aggression appears to increase. I have then seen temporary territories in which one individual foraged for some minutes 242 Bulletin Museum of Comparative Zoology, Vol. 141, No. 5 and drov'e out all intruding conspecific individuals. Individuals of other species might be driven out or tolerated. These appeared to be foraging territories only, and were small — three to seven meters across in most cases — in sites that were exceptionally wet and covered with low, green vegetation. Hudson (1920) mentions single-species flocks of several wintering species of Xolmis in Argentina. Further, Neoxohuis rufiventris is said by him to associate with flocks of plovers. Whether these were really mixed flocks or aggregations result- ing from patchy distribiition of habitat is not clear. The same may be said of the interspecies flocks of Muscisaxicola. I have found no indication that members of the other genera (Agriornis and Ocli- thoeca) flock outside the breeding season. Within the breeding season all are reputed to be found only in pairs. Flocking be- havior does not seem to be characteristic of this group outside of migration, and may play an important role only in Xolmis. Nest Construction The open, cup-shaped nests of Musci- saxicola species are always placed on the ground in crevices between rocks, or even within a rabbit burrow (e.g., M. capistrata, personal observation). Only M. brevi- cauda does not always employ crevices, and where its nests are hidden by grass alone they are often partially roofed with grasses (Marchant, 1960). Most Xolmis species build open cup- shaped nests in bushes or trees, although X. iriipero hides its in holes in trees (Hud- son, 1920; Wetmore, 1926), or appropriates an enclosed nest of Fnrnarins rufus (see Hudson, 1920). X. cinerea is said to nest in holes in banks in Argentina, but to build an open nest in trees in Brazil and Uruguay (von Ihering, 1904; Hudson, 1920). X. coronata nests either in a bush or in a "large clump of grass" (Hudson, 1920), but apparently not on the ground. The only described nest of Neoxolmis rufiventris was on the ground (Maclean, 1969). The species of Agriornis also build cup- shaped nests in bushes, except for A. mon- tana, which nests on the ground amid rocks like a Muscisaxicola (Goodall et al., 1957). Nests of OcJitlioeca species and Pyrocepliahis ruljinus are open cups, usu- ally in bushes or trees. Todd and Carriker (1922) report a nest of Ochthoeca rufi- pccforalis that was placed on "a projection on the side of a huge boulder, sheltered by some stunted trees," apparently com- parable to the nests of Saijornis species. PART II. DISPLAY BEHAVIOR The following accounts present what is known about the form and employment of stereotyped signals ("displays") used by these birds in communicating. A detailed description of the display repertoire of Muscisaxicola alhilora provides a basis for comparison. The displays of two small Muscisaxicola species are also described in detail: that of M. maculirostris to show its comparable structure, even though it sounds simpler to the human ear, and that of M. hrevicaucla because (a) it is at least superficially somewhat aberrant; and {h) the species has customarily been placed in the monotypic genus Muscigralla, usually not seen to be related to Muscisaxicola (Hellmayr, 1927: 94, was not even sure it belonged in the Tyrannidae). Where possible, comparisons are made with displays of Sayornis species, since these are known in much greater detail and provide a coherent frame of reference (see Fig. 11). Occasional comparisons among species are made in this section to facilitate description, but general comparisons fol- low in the Comparative Summary. Ma. Muscisaxicola Muscisaxicola albilora Most individuals of this species were still in loose flocks, although many ap- peared paired, on 5 November 1961, at Evolution of Ground Tyrants • Smith and Vnillcumier 243 Farellones and La Parva, Chile. On 7 and 8 November those at the Rio Yeso sites all appeared paired, but were foraging com- munally at the best sites and showing only slightly more frequent territorial behavior. Several kilometers farther into the interior of the Andes along the Rio Volcan on 8 November there were flocks of over 60 individuals, plus lesser numbers of at least two other congeneric species. By 17 No- vember most snow at the upper Rio Yeso site had melted, and much more green vegetation was scattered about. Most pairs had dispersed onto the slope, and I found one nest being built among the boulders on an unstable talus slope. There was some communal foraging at the best local site, however, the edge of a large meltwater pond. In 1962 the upper Rio Yeso site was al- most entirely snow-covered on 15 October, and had no M. olhilora. On this and the subsequent day only three small groups were at the lower site; the individuals as- sociated with almost no display and only occasional supplanting attacks. A few lone individuals occasionally held temporary territories, but \\'ith little display. Vocal Displays (1) In agonistic encounters increasingly prominent aggressive tendencies are ex- pressed by the more or less graded series: "tseet" and "tseek," "tut," "tsk," "tchk," and "tk." Most are known primarily from males. Both "tseet" and "tseek" (Fig. la and b) may occur very early in agonistic en- counters, and "tut" (Fig. 2c) and "tsk" (Fig. Id) at various points, but "tchk" (a more sharply peaked variant of "tsk") and "tk" (Fig. 2d, perhaps a high, brief, ex- treme variant of "tut" with different har- monic structure, an extended peak, and prominent descending terminal arm) are almost restricted to fighting, as two birds grapple. (2) As avoidance or escape become more probable, "tseet" or "tsk" calls are replaced by more prolonged vocalizations: "teek" (Fig. Ic) and "sect" (Fig. le, and with a slight tendency to show two peaks in Fig. If), respectively. During agonistic encounters, fights are often separated by periods in which the quarreling birds stand apart and usually do Wing Raising displays while one individual utters long strings of "seet." In at least some encounters, how- ever, it was this individual which later pressed the attack again. (3) A bird flying or landing alone may utter a single "tseet." This is more likely to be a "tseek" or especially a "teek" (Fig. Ic) if the commimicator is flying toward its mate. The usual reply by the mate is a "tsk," but if both individuals are in flight toward each other there is often a rapid jumble of quite variable calls (e.g.. Fig. Ig), perhaps not all uttered by one indi- vidual. These include "tseek," "seet," "tsk," and even "tchk," plus special forms such as the third and final in the illustration, which are probably the only ones peculiar to greeting usage. Sometimes in less agonistic situations, however, mates employ a simpler stereot\'ped greeting (Fig. Ih). Bursts of a simple, chevron-shaped call, "tuh" (Fig. 2a), have been heard and re- corded only from apparently paired birds foraging together when there has been no agonistic activity for at least several minutes. Bursts of a higher form (Fig. 2b) appear intermediate to some greetings. The similar "tut" call (Fig. 2c) is sometimes uttered, singlv, from mates while foraging. (4) In one Aerial Display a loud "clee-ip" (not recorded) was heard; it was quite reminiscent of the corresponding call of M. maciilirostris, but with a more abrupt termination. Most of the known vocal displays of M. alhilora are thus unelaborate calls linked by extensive sets of intermediates. Only the most different forms, such as "tuh" and "tseet" are easilv' distinguished by the human ear, although many of the others are recognizable with practice. The more agonistic vocalizations resemble in form 244 Bulletin Museum of Comparative Zoology, Vol. 141, No. 5 ■ A . a b c d e ■ 1 f ■ A ^ ^ A '\ A ^ m A A A ^ a ' > u 'i ^i A ^ ^ ^ -1 — 1— Figure 1. Vocalizations of Muscisaxicola albilora-. la. "tseet"; lb. "tseek"; Ic. "teek"; Id. "tsk" (compare with Fig. 11a); le. "seet '; If. "seet" (note tfie slight tendency toward two peaks in this case); Ig. an example of a variable greeting; Ih. an example of a greeting in a less agonistic situation. A Figure 2. Vocalizations of Muscisoxicolo albilora-. 7a. a burst of "tuh" calls; 2b. a burst of higher "tuh" calls (note general similarity to Fig. Ig); 2c. "tut"; 2d. two examples of "tk." and usage calls of Soyornis species. For instance, "tseet" is comparable to the Initi- ally Peaked Vocalization (IPV) of all Sayornls species, "teek" resembles the Simple Vocalization (SV) of S. nigricans and "tsk" the SV of S. plioebe, and "tseek" somewhat resembles the Bipcaked Vocali- zation (BV) of S. phoehe and S. nigricans. As in at least S. phoelie, greetings between mates during or shortly after agonistic situ- ations usually involve SVs, BVs, or IPVs, and the special greeting vocalization of M. albilora resembles that of S. saya and to a lesser extent, that of S. phoehe. None of these possibly homologous calls corresponds exactly to each other in spe- cific usage. The fit is probably least good for the bursts of "tuh" calls of M. albilora, which may be only superficially similar to the Chatter Vocalization (CV) of S. saya, and the Doubled Vocalization (DV) of S. phoehe. (In comparing forms, note that the scale of Figure 11, showing sample Sayornis vocalizations, is like that used for the remaining genera of this paper, but proportionately condensed along the time axis and elongated along the frequency axis as compared with the scale used to illustrate the very brief Muscisaxicola vocalizations.) Nonvocal Displays (1) Wing Raising. In prolonged aggres- sive encounters birds within a meter of one another may raise one or both \\'ings. En- counters in which neither individual is \'ery actix'e usually lack Wing Raising. The minimal form has been seen when two birds, sometimes one member of each of two pairs, approach each other during Evolution of Ground Tyrants • Smith and Vuillemnier 245 Figure 3. Some Visible Displays of Muscisox/co/o alb/loro: 3a. minimal Wing Raising with one wing, and slight Head Bow; 3b. Wing Raising with synchronous fluttering of both wings; 3c. asynchronous Wing Raising with Crown Ruffled, and ruffling of the flanks and chest. foraging. One or both may "casually" lift one wing very slightly (Fig. 3a) without other display, and not facing its opponent. In such instances no fights result, and the birds usually go their separate ways. There may be fights, however, when two or more pairs come together after prebreed- ing flocks have scattered. Simple raising of one wing by one or more individuals comes early in such encounters; later the movement becomes jerky and abrupt, and the second wing begins to be jerked slightly up. Often the two wings are lifted to different extents and asynchronously, or both are raised at once and wagged ir- regularly in an asynchronous, "semaphor- ing" pattern emphasized by the pale color of the wing undersurfaces. Less commonly, both are lifted high over the back and fluttered in unison (Fig. 3b). Rarely, a bird in this pose flies perhaps three to five meters between two rocks, still with its wings high and fluttering. This resembles part of the Aerial Display, and it is likely that Wing Raising and that display inter- grade, although I have not seen the re- maining intermediates. Any flight during a Wing Raising encounter will likely ter- minate with the wings held high, and partial raising in what appears to be a flight intention movement is common be- tween bouts of Wing Raising. The tail is fully spread in most forms of the display, but not flicked. The flanks and chest are ruffled during at least some phases, and the Crown Ruffled accompanies all but the simplest forms (Fig. 3c). Thus Wing Raising shows several special plumage features: pale wing linings, pale bases to the primaries on the upper wing surface, and the occipital crown-patch. Opposing individuals orient variously but usually laterally; a tendency to orient side- ways to the opponent seemed strongest when both wings were held high. I could detect no obvious preference to raise, or to raise first the wing closer to or farther from the opponent. A few times, up to five individuals, ap- parently including t\\'o pairs, partook in an encounter. In the most ^'igorous periods of activity all used Wing Raising. Rapid and confused chasing and fighting developed, and long chases went beyond effective observing range. In one encounter between 246 Bulletin Museum of Comparative Zoology, Vol. 141, No. 5 two pairs tlie display of two individuals proceeded as far as two-wing semaphoring with only an occasional "tseet," and no fight or chase resulted. After about a minute of stalemate one individual per- formed an Aerial Display. Sayornis soya and probably S. nigricans have displays similar in form to Wing Raising and the forms of Aerial Display closest to it, but these appear to function in what is probably pair-bonding. As I have seen neither territorial defense in those two Sayornis species, nor pair-bond- ing in M. alhilora, the functional compari- son is incomplete. (2) Aerial Display. On 5 November 1961, an individual several times silently chased an intruder from what was probably a potential breeding territory. It then be- gan to fly, climbing slowly with wings held high over its back and used in bursts of short strokes. Its long legs dangled and the feet appeared relaxed. The dark tail was held vertically down and fully spread. The bird remained silent. It often simply hovered, sometimes climbed higher, and frequently turned to a new direction, re- maining over a limited area. Finally it descended in a long, fast swoop with its wings closed until near landing, and perched stiffly upright atop a group of rocks. At another site on 17 November 1961, a bird left a stalemated Wing Raising en- counter (q.v. ) and flew up in a Tyrannus- like twisting flight of about 20 to 25 meters, then swung around and dropped c|uickly to a foraging area. During the flight it uttered several loud, abrupt "clee-ip" calls. (3) Crown Ruffled. During Wing Rais- ing disputes all birds ruffled their crowns, either entirely or just at the rear. This made the head appear larger, and also made very conspicuous the occipital crown-patch, the rusty color of which is species-specific. Establishment of a temporary foraging territorv was watched on 16 October 1962. The owner foraged vigorously for several minutes before a conspecific individual arrived. The intruder was immediately driven to the edge of the area without dis- play, but it persisted in reintruding over the next three minutes. The territorial owner always caused it to retreat by dash- ing toward it without display. The tres- passer never faced toward the owner, and ^^'ould ruffle the back of its crown as soon as it began to be pressed and on stopping after a chase. (4) Head Bow. On 5 November 1961, two individuals were observed about 60 centimeters apart, oriented almost toward one another and bowing their heads, thus revealing the ruffled patch of color at the back of the crown. Occasionally one or both uttered a barely audible "tseek." Re- fore anything further happened one saw me and ran off, the other following several meters behind. SAuscisax'fcola frontalis In 1961 individuals of this species were observed foraging by a pond where some interacted with M. alhilora. In 1962 three were observed foraging in loose association. Actively foraging individuals uttered two different calls: "trit" and "treet" (Fig. 4a and 4b). They resemble, but at much lower frequencies, "tseet" and "sect," re- spectively, of M. alhilora. One individual, apparently nest site prospecting while foraging, uttered short bursts comprising several "pt" vocalizations at two frequen- cies (Fig. 4d); these were like the "tuh" of M. alhilora, and the usage strengthens the comparison of such calls with the CV of Sayornis. Yet another foraging individual used a "tuk" (Fig. 4c, like the "pt" but higher, briefer, and uttered singly instead of in bursts — i.e., resembling the "tut" of M. alhilora). I recorded several sharp "t" calls from an M. frontalis as, and shortly after, it fought with an M. alhilora. These resemble the "tk" used by the latter species in fights, Evolution of Ground Tyrants • Smith and Vuilleumier 247 but emphasize the initial ascending arm and ahnost omit the descending arm. All known vocalizations of M. frontalis thus correspond in form, and in general usage, to calls of M. albilora. \ Muscisaxicola flavinucha Displays were observed only on 26 Sep- tember 1962, when a heavy spring snowfall at Lagunillas, Chile, brought birds down from higher altitudes. Three M. flavinucha appeared on my study area before snow accumulation exceeded 2.5 centimeters. They associated loosely while foraging, and one made occasional silent supplanting flights against the others. Once the aggressor stopped about 60 centimeters away from another without supplanting. The approached individual stood still with its head withdrawn be- tween its shoulders, and began to call. There were some short bursts such as "tsee tee tsee tseet." Other calls were more ir- regularly spaced, but sounded similar, and were like calls recorded from M. albilora during greetings. Both ruffled their upper backs. The calling one maintained an oblique orientation toward the other for about 90 seconds, then the latter wandered off, and eventually flew away. Vuilleumier ( personal communication ) has heard vocalizations from two adults feeding large nestlings at Cerro Llaima (Cautin Province), Chile: high-pitched "seeht" and "seeseeht." A fledgling he watched being fed at Cerro Catedral (Rio Negro Province), Argentina, begged with a thin "see .... see . . . ." SAuschaxicola rufivertex On 12 October 1961, several separate individuals were seen on steep slopes near Lagunillas, Chile. In mid-aftemoon two flew from opposite sides of a rocky stream bed out over a gully, somewhat toward one another without approaching closely. Both performed Aerial Displays. Each would fly about 16 meters then rear up until hanging Figure 4. Vocalizations of Muscisaxicola frontalis-. 4a. "trit" (compare with Fig. la); 4b. "treat" (compare with Fig. If); 4c. "tuk " (compare with Fig. 2c); 46. burst of "pt" vocalizations (compare with Fig. 2a and 2b). vertically and stall with wings stretched fully above its back, at this point uttering a thin, high-pitched "twee-it." This was repeated several times, then both turned back and repeated it several more times before landing on the slopes. Neither dis- played further, and later in the season they apparently ascended to higher altitudes as the snows melted. On 16 October 1962, two apparent pairs foraged on the rocky outu'ash plain of a small stream at the lower Rio Yeso site. The one pair flew to a bush about 1 meter tall, one perching on top and aligning at right angles to the pair still on the ground. This upper one began occasionally lifting and waving one wing, the wing toward the other pair. It had its chestnut occipital patch conspicuously elevated (i.e.. Crown Ruffled). The other pair was then about 10 meters away, the nearer individual fac- ing the bush and Wing Raising vigorously, usually with both wings. Within a minute they flew toward the same bush, landing within 60 centimeters of the first pair. The first indi\'idual now aligned slightly more away, but continued to raise its wing. The "aggressor" oriented its body to within five to ten degrees of the first and raised both wings. Both had Crowns Ruffled, and I could hear "tee" vocalizations poorly over the high wind. Within a minute, the individual which had approached dropped back to a second 248 Bulletin Museum of Comparative Zoology, Vol. 141, No. 5 V ( . 'i Figure 5. Vocalizations of Muscisaxicola caphtrata-. 5a. "tueee" (compare with Fig. la); 5b. "tee" (compare with Fig. le and If); 5c. a common vocalization, intermediate be- tween the previous two; 5d. clear "seeee ' of a nestling, begging; 5e. loud and harsh call of a nestling, in the hand. bush 70 centimeters farther away, and continued visiorous Wine; Raisins;. The first then flew about 6 meters followed by its apparent mate; the other pair followed to within 2 meters, and all displayed as be- fore. The first pair then flew another 6 meters away, the displaying individual raising both wings over its back and glid- ing to its perch. This ended the encounter; within 45 minutes all four birds had left the area. All known displays of this species are comparable in form and usage to displays of M. olbilora. Note that Wing Raising \\'ith both wings characterized the more aggressive bird, and wings raised high in flight appeared in a situation similar to that in which it occurred in M. albilora, again resembling an intermediate between the Aerial Display and Wing Raising. Muscisaxicola macloviana Hudson ( 1920 ) described "low plaintive whistling notes" given by this species when wintering, perhaps what Pettingill (per- sonal communication to Vuilleumier ) called a "reedy cheep." Muscisaxicola capistrafa Observations were made from 4 to 8 December 1961, on the farm "Las Mer- cedes," Tierra del Fuego, Chile. One nest was located with three young, about a week old. Other local pairs probably had nests, and there were also apparently a few independent fledglings. Vocal Displays (1) While I examined nestlings, one adult sometimes called a "tueee" (Fig. 5a), closely similar to the "tseet" of M. albilora, but witli its two peaks respectively low and high instead of the reverse. It did not ap- proach close to me and usually called only when its mate flew to perch near it. ( 2 ) A variable "tee" ( Fig. 5b ) was used by both parents while I examined their nestlings, and the one which approached me closer uttered only this call. It is similar to the "sect" of M. albilora, the main difference corresponding to that of the previous case: the second peak is higher in M. capistrata. Intermediates between "tueee" and "tee" (e.g.. Fig. 5c) were common on each oc- casion I examined the young. During five consecutive days, however, I heard no other vocalizations from any adults. (3) Nestlings removed from the nest- burrow uttered two quite variable calls (Fig. 5d and e). One was a simple, clear "seeee" ^^ith rising inflection, and the nestlings were sometimes begging when uttering it. It resembles the IPV of Saijor- nis saya nestlings. The other was loud, had an initial ascending arm, terminal descend- ing arm, and often something resembling a chevron between, but with gaps filled with harsh noise separating all three sec- tions. Nonvocal Displays (1) Feather Bufflinfi. The parent which approached me closely uttering "tee" re- mained conspicuously sleeked. Its mate usually perched 3 to 7 meters away, called "tee" frequently, and sometimes ruffled its chest, flanks, and the sides of its head. (At least the first two areas were ruffled Evolution of Ground Tyrants • Smith and Vuilleumier 249 by some M. albilora when Wing Raising, but the present species did not Wing Raise in the several intraspecific fights observed. ) (2) Wing Wliin: Very loud Wing Whir- ring was heard in some intraspecific fights, and was likely a display. Loud Wing Whirrs were also produced by the bird hovering near me at the nest. (A Wing Whirr was also heard from one M. frontalis in a situation I could not be sure was dis- play, but possibly this sound need never be produced in any wing stroke functioning solely for flight.) Its usage corresponds to that of a similar sound of Sayornis phoehe. (3) Aerial Display. On the afternoon of 4 December a lone individual made two silent display flights within three or four minutes. The bird made little headway against a wind of about 30 kilometers per hour. It simply flew up about a meter, raised its wings high over its back, and fluttered briefly until poised in a vertical stall with wings stretched maximally up- \\'ard. Then it fluttered again, repeating the procedure two or three times per flight. This was probably the sort of intennediate between Wing Raising and full Aerial Dis- play which was missed in M. aJhilora. Muscisaxicola maculirosfris I observed this small species during several periods: shortly after its arrival on breeding grounds (13 September to 16 October 1962, Chile), during nesting (9 to 17 June 1966. Ecuador, and 28 November 1962, Tucuman, Argentina), late in the breeding season (4 to 6 August 1959, Ecua- dor ) , after the breeding season ( 13 Janu- ary 1962, Chile), and in midwinter (4 to 6 February 1962, Ecuador). Although at one Chilean site (Lagunillas) and at all Ecuadorean sites near San Antonio popu- lation densities were high, there was never much interaction or much displaying. Vocal Displays ( 1 ) Its basic call is "tek," most variants of which are indistinguishable in the field. Extensive tape recorded samples show that its variations tend to correspond to the various briefer calls of M. albilora, but are insufficient to show the limits of each. The "tek" itself probably corresponds roughly to the "tut" form of M. albilora, is often repeated at similar intervals, and has the same chevron shape but not the har- monics (Fig. 6a). Foraging individuals often repeat it for minutes at a time, but are silent for similar periods. Usually it comes at or just after the end of a dash, or while the bird is stopped and peering around. A bird which has just been fighting, or which has watched a hawk or owl pass by may utter "tek" and slight variations frequently. It was occasionally used in mobbing a small predator such as the Burrowing Owl (Speofyto cunicularia) or Sparrow Hawk (Falco sparverius), at least in Ecuador \\'here these are common and were often ignored or mobbed in a "des- ultoiy" fashion without displays. Some- times when we appeared to make an individual slightly nervous it would begin to call "tek." Occasionally, slightly higher pitched forms can be distinguished. These are most often uttered by a bird about to join an- other (or about to be joined), or one which is foraging near another and appears slightly nervous. They probably cor- respond to the "teek" of M. albilora. Other variants are lower or more prolonged, or have a more prominent tail (like a small "whee-oo" call, q.v. ). The prolonged ones may be used primarily by relatively fearful birds. No known vocalization clearly cor- responds to either the bi-peaked "tseek" or the "tseet" of M. albilora. (2) I have recorded nothing as flat as the prolonged "sect" of M. albilora, but did obtain some relatively long, variable, asym- metrical, squeak)^ calls (e.g.. Fig. 6b) from an apparently partly cowed and indecisive individual. It had begun to crouch after running toward several indixiduals that were responding to the overflight of a large hawk with Aerial Displays. The area was 250 Bulletin Museum of Comparative Zoology, Vol. 141, No. 5 the site of several intraspecific encounters and was likely a boundary region among ill-defined territories. Once a grappling fight there terminated on the ground, then the attacked bird crouched low with wings outspread and uttered a "squeal" ( probably more prolonged than in Fig. 6b) in the brief interval before it was again attacked. (3) No greetings were recorded from mates meeting in agonistic circumstances. However, two probably intergrading sorts of calls were found, each with a single peak preceded by a shaiply descending arm. The forms resemble BV-types recorded in agonistic greetings of M. alhilora. The first (Fig. 6c) is relatively prolonged, and was recorded from the individual which next uttered the call shown in Figure 6b, just before it stopped approaching the others. The second (Fig. 6d) tends to occur in agonistic encounters, including fights, and may be repeated in series before an Aerial Display (Fig. 6g shows the end of such a series). Such series, however, usually be- comes composed just of "tk." In relatively nonagonistic meetings mates seldom called, and then usually used "tek" or a closely similar SV variant. I very rarely heard calls that could have been similar to the LHV-type display of M. alhilora, and obtained no recordings. (4) The "tk" is slightly briefer and lower than "tek." It may be uttered in series before launching an Aerial Display, or in rapid series while standing and watching a neighbor's Aerial Display. Other uses cor- respond to the known uses of "tuh" series by M. alhilora. The descending arm of "tk" is character- istically more emphasized than is the ascending. From two fights, one in Ecua- dor and one in Chile, I recorded versions virtually reduced to a prominent descend- ing arm (Fig. 6e and f), and thus compa- rable to its apparent homologue in M. alhilora. The Chilean example is much lower in frequency than the Ecuadorean. Shortly after the Chilean fight (between two males establishing territories), one of them did an Aerial Display and upon returning gave a similar burst of low "tk" as he landed alone. Similar brief "tk" variants, usually at fairly low frequencies and not in series, were also used in fights, and many by an individual standing outside one fight as a spectator. He became the aggressor only when the first aggressor withdrew. (5) The Aerial Display vocalization has two parts, the second of which commonly occurs alone when the display is used after dawn. The first is a series of "tk" or some- times "t," which seems to be a very brief "tk." This accelerates toward the end, then rises in frequency either through several units ( Fig. 6g ) or in one unit ( Fig. 6h). The second part is a clear, whistled "clee-oo" or "whee-oo." In one Ecuadorean bird this had a relatively prolonged tail and harmonics (Fig. 6i). When combined, the two parts resemble the Regularly Re- pcat(xl Vocalization (RRV) of Pyrocepha- his ruhinus (Smith, 1967) and the CV plus IPV of Sayornis saya (Smith, in press, 2). (6) The sole nestling examined repeated almost continuously a faint, rapid series of variable "ti" units (Fig. 6j) as my wife held it in her hand. This is not yet known from congeneric species, but in Sayornis phoehe and S. nig,ricans the "tee" of nest- lings is very similar and may be used in bursts. Almost fully grown fledglings sometimes uttered rapid series (probably correspond- ing to the nestling call) and gaped toward an adult. When they were about to be fed the series seemed to become high pitched. If I approached such a fledgling closely it would usually begin to call "tek." Nonvocal Display (1) Aerial Display. In the most complex form the bird flies out (or outward and upward if the slope is not steep) and be- gins a flat, fluttering fhght, calling a string of very brief "t" or "tk" units, sometimes Evolution of Ground Tyrants • Smith and VuiUeumier 251 jerking slightly back with each call. As the series accelerates the frequency of the last few units ascends and the bird climbs sharply into a stall. Just before stalling, the tail flicks up, then drops straight down- ward and spreads. The bird comes to hang vertically in the air, dangling its legs and feet. Its wings stretch over its back and nearly touch together as it utters the clear "whee-oo" and then pitches forward. Re- gaining speed, it retracts its legs and flut- ters on in another performance. The flight is straight or twisted, and in the latter cases remains over a small piece of ground. After from one to about fifteen ( commonly about five) stalls, the bird flies back to the slope or partly closes its wings, slightly cocks its tail, and dives to the ground. Peters (1923) gave an accurate but much less detailed description. There are various less complex forms. Just the terminal "whee-oo" may be uttered, which is very like counterparts of the dis- play known in M. albilora and M. rufi- vertex. At least in Ecuador in 1966, this was the common form after extensive predawn usage of the full form. Flights were then usually given in response to the passage overhead of an avian predator or in response to Aerial Display by a neighbor. The latter individual might be close to a boundary and the two displaying against one another, or (more commonly) might be responding to an avian predator — the first Mnscisaxicolo to go into Aerial Dis- play when a Sparrow Hawk or Buteo passed seemed to precipitate the display from most of his neighbors. Sometimes a bird calling "tek" would launch into Aerial Display without an introductory series. More commonly, a series of "tk" calls was uttered before the bird took flight — these may have replaced the usual aerial calls. Sometimes such a "tk" series would build up in speed and then slow down, the bird lapsing into "tek" without an Aerial Display. Fully complex Aerial Displays were never preceded by a "tk" series from the ground. - \ - h \ \ \ a b c d e ,^ \^ r ^ ^ ^ ^ '^ Figure 6. Vocalizafons of Muscisaxicola maculirostrls: 6a. "tek" (compare with Fig. Ic, and first vocalization of Fig. Ig); 6b. squeaky sounding vocalization (compare with Fig. le and If); 6c. vocalization recorded from an individual running toward a fight, just before stopping and crouching; 6d. a somewhat similar form, used in fights, agonistic en- counters and sometimes in series before an Aerial Display; 6e. "tk," recorded from a fight in Ecuador (compare with Fig. 2c and 2d); 6f. "tk, ' recorded from a fight in Chile (compare with Fig. 2d); 6g. vocalizations from an Aerial Dis- play: the end of a "tk " series, with the last intervals short- ening, and the last several "tk" rising in frequency before the "whee-oo"; 6h. vocalizations from an Aerial Display: the "tk" series is not shown, but ended in a single relatively high unit, followed by the "whee-oo"; 6i. a "whee-oo" vocalization with a relatively prolonged tail; 6j. three suc- cessive "ti" vocalizations from a series uttered by a nestling. In a unique variant the terminal "whee-oo" was omitted and the aerial stuttering series left intact. The bird flew up when a Sparrow Hawk pounced well upslope from him, and did ordinary Aerial Display with five stalls and vocalizations before each. His "t" series accelerated and rose slightly in frequency just before stall- 252 Bulletin Museum of Comparative Zoology, Vol. 141, No. 5 ing, but at the stall he simply used another "t" of the original frequency. Yet another individual once gave a single performance that was silent except for a sharp "clit" at the stall; shortly later he did an Aerial Display with a "whee-oo" at each of several stalls. The display's appearance also varies. In the most interesting case, an Ecuadorean bird rose to level with the lip of a quebrada then flew at that height, repeating rapidly "t t t tu-twit" in which the terminal arm of the call seemed foreshortened (not re- corded). With each "tu-twit" it arched its back, raised its wings high (but not to the customary extreme), lowered its legs, and partly cocked and partly spread its tail. This cocking of the tail is normally brief and precedes the stall. In this case, the bird did not assume a vertical orientation of its body, but used a form much more like that of M. brevicouda. The Aerial Display may not differ ap- preciably between the Chilean and Ecua- dorean populations observed. The full form most common in Chile was also most common in the preda\\Ti twilight in Ecua- dor. The Ecuadorean vocalizations are apparently slightly lower in frequency, and the terminal "whee-oo" at least sometimes descends less abruptly than in the Chilean sample. Not all factors governing the use of this display are yet known. It is more common just before and during the nesting season than at other times. It is "infectious" in that males stimulate one another to per- form ( as in the Tumble Flight of the genus Tijnmnus, Smith, 1966, and the display flights of Pyrocephahis ruhimis, Smith, in preparation). It is used during aggressive encounters and is sometimes given by spectators. Perhaps it is used by a "frus- trated" aggressor — a bird intimidated but not cowed by his opponent(s) or by an aerial predator, or thwarted by poor visi- bility in the predawn twilight. Whatever the causation, the display is suited to func- tion as a long distance advertisement and specific identification of a territorial male. (2) Wing Rai.siug cum Buzz. A pattern which precedes or replaces Aerial Displays in some agonistic encounters is almost certainly homologous with Wing Raising of the larger congeneric species. Once a bird watched two others fight about 10 meters away for over a minute, calling "tek" frequently. It began to jump up about 30 centimeters while lifting both wings, synchronously, fully over its back into the maximal position typical of the Aerial Display (revealing a large expanse of pale plumage). An abrupt, snapping, mechanical Buzz was produced as the wings either reached their full stretch or started down. After several such displays the bird paused, began Wing Raising cum Buzzes again without jumping, then did an Aerial Display with repeated stalls. In two other cases this Wing Raising cum Buzz was used by one participant 30 or more seconds after a fight. In one case it was repeated twice by a perched in- dividual about 30 seconds before giving an Aerial Display, and in the other was given as an Aerial Dis]ilay when the bird flew up 2 meters, stalled in the usual fashion once but with the Buzz instead of calling. On yet another occasion. Wing Raising cum Buzz immediately preceded Aerial Display by an individual responding to the Aerial Display of a neighbor. In several dozen recorded Aerial Displays tcith vocal- izations, I have never detected a Buzz. Prol^ably the Buzz relates to the Wing Whirr of some larger species, but in M. mocuUrostris it has a ritualized relationship to Wing Raising. The wings are sometimes raised syn- chronously into the full upright position without a Buzz by a bird on the ground in an aggressive encounter, and this is per- haps the minimal form of the display. In the one case seen clearly, the bird gave a "tk" with each wing movement, then at- tacked its ojpponent. Evolution of Ground Tyrants • Smith and Vuillewnier 253 (3) Otlier Wing And Toil Movements. Aggressive individuals facing an opponent often droop their wing tips nearly to the ground before and after fighting. The wings, and less often the tail, may be given little flicks. One aggressor cocked its tail up while letting its wing tips droop, with- out flicking either. Birds in these postures were silent and very likely to attack, dart- ing forward with rapid wing flicks, showing the upper wing surfaces to the opponent. One was seen to wing quiver slightly with each "tek" after landing from an Aerial Display. Within a minute it flew off — apparently to an agonistic encounter. Usually no wing movements accompanied "tek" calls. A slight and very rapid flick of the tail, however, often accompanied a "tek," especially if calling was rapid. An- other tail movement, a quick fanning parti- ally open and shut again, often preceded running. (4) Crown Ruffled. If I approached a fledgling and it called, the adult female might also call "tek" and ruffle her fore- head conspicuously. This is the second phase of crown ruffling in the larger spe- cies, most of which have an occipital spot of color which M. maculirostris lacks. Musc/sax/co/a brevicauda I studied this species {Muscigralla brevi- cauda of authors) at only one site in coastal Peru, from 23 to 26 January 1962. The population was locally dense, and there was much displaying. Vocal Displays ( 1 ) A loud "tchek" ( Fig. 7a ) was often heard just as birds landed; it resembles in form and usage the Locomotoiy Hesitance Vocalization (LHV) of Sayornis plioebe (Fig. lie). The communicators were usu- ally watching as I walked through their territories, and probably most were males, since they also used the prolonged vocali- zations described below. In fact, the "tchek" appears to be incorporated into the prolonged vocalization, immediately pre- ceding the ultimate descending series. A slightly briefer fonn "tchk" (Fig, 7b), hard to distinguish by ear, occasionally intergraded with "tchek." "Tchk" was some- times weak, nearlv lacking its first element, with emphasis then falling on the third or, sometimes, the second element. It was also uttered on landing, or by perched individu- als making flight intention movements, and sometimes preceded landing or was re- peated two or three times during a very short flight. But usually it was in a short series as prologue to the prolonged vocali- zations given while perched. It may cor- respond to the Doubled Vocalization (DV) of S. plioebe. Occasionally in fights a harsh "zrrt" was used that may have been a version of "tchek"; it was not recorded. No forms similar to "tchek," "tchk," or "zrrt" are known from other Muscisaxicola species, except for the apparently rare LHVs. In usage, however, they at least partially re- place the two-peaked calls of M. albilora and other large species. The reduced visi- bility in the brushy habitat of M. brevi- cauda may require more complex calls for at least species identification. (2) There are at least three briefer vocalizations. One has an initial descending arm preceding a peak ( Fig. 7c ) that sounds like "tk" and is sometimes repeated two or three times in flight toward a perch, fol- lowed by "tchek" on landing. Occasionally, a similar series of "tk" followed by a "tchk" comes just after landing. In a short flight the series may be intermediate between "tk" and "tchk," with the third element of the "tchk" emphasized. Two other very brief calls sound like "tk" (Fig. 7d and e), but do not seem to form intermediates with it, although they do with each other. They are often inter- spersed among successive prolonged vocali- zations in flight, but do not form long series \\'ithout haxing inteipolated oc- casional "tk" calls of the first sort. The more obviously chevron-shaped element 254 Bulletin Museum of Comparatice Zoology, Vol. 141, No. 5 ill a, ,1 tl —I i_ tlnit hi HW I Figure 7. Vocalizations and Pose of Muscisaxicola brevi- CGudo: 7a. "tchek" (compare with Fig. lid); 7b. "tchk" (this occasionally intergrades Into the previous vocalization); 7c. "tk" (compare v/lth the elements of a "tchk"); 7d. "tk ' (compare with Fig. 6a]; 7e. "tk" (compare with Fig. 5a); 7f. prolonged series vocalization uttered while perched (all but the last unit of the introductory series has been omitted); 7g. prolonged series vocalization uttered during Aerial Display (all but the last unit of the introductory series has been omitted); 7h. customary pose while uttering a prolonged series vocalization from a perch. ( Fig. Yd, compare with "tek" of M. inaculi- rostrls) is .sometimes uttered irregularly by a bird which alternates foraging with perching on a bush and uses this and no other calls; it also utters this call some- times on landing from a short flight. Oc- casionally a series of this version comes just after landing, followed by one or more "tchk" calls leading to the prolonged vocalization. Series of the second version (Fig. 7e) may be the more usual among successive prolonged vocalizations of a flying bird, and often continue as a bird completes a relatively long flight. Such long terminal series are broken by an occasional "tk" (Fig. 7c). On the whole, these last two versions of "tk" appear to correspond closely to "tek" and "tk" in the repertoire of M. maculirosiris, and must be related to Simple Vocalizations. (3) Prolonged series vocalizations (Fig. 7f) are uttered either from perches atop cotton bushes or during Aerial Displays. Those in flight are usually, but not invari- ably, less complex ( Fig. 7g ) . All evidence indicates that the prolonged vocalizations correspond at least in part to what is usually called "song" in oscines, or to the Regularly Repeated Vocalizations of Soy- ornis and PyrocephaJus. They arc used in countercalling among conspicuous in- dividuals remaining within their apparent territories. Prolonged vocalizations may be stereo- typed elaborations of the less complex calls used by the same individuals during flight and landing. A "tchek" or "tchk" preceded by any of the "tk" vocalizations may cor- respond to a minimal form. The more com- plex flight form is a "tk" series plus a "tchek" with a harsh section inteipolated, and a set of five or more descending ele- ments appended (Fig. 7g). In the most complex (perched) version, the initial series is two to four "tchk" instead of "tk," plus yet another new feature, a section of ascending pitch matching the ultimate descending three elements. The whole vocalization sounds reminis- cent of a slurred version of the Aerial Dis- play call of M. mociilirostris, but the form of the prolonged vocalization is much more complex. If, however, the audible rise and fall of the second portion is equivalent to the "whee-oo" of M. macuUrostris, then the prolonged vocalization may be homol- ogous. Perhaps significantly, the "whee-oo" is at least rarely replaced by three chevron- shaped elements, and is in this form more similar. The greater complexity of the pro- Evolution of Ground Tyrants • Smith and Viiilleumier 255 longed vocalization of M. brevicouda is consistent with the greater complexity of each of its other most common calls. Nonvocal Displays ( 1 ) Wing Raising. No display seen was certainly comparable to Wing Raising of the other species. Once, however, after two individuals fought and chased, the pursuer perched and partly opened its right wing several times in rapid succession. Further, one incident before launching into Aerial Display also suggests that a display like Wing Raising exists, although I may not have seen its full form: A bird was repeating "tchek" fairly rapidly. It gradually extended its wings laterally, and without fully spreading the primary feath- ers, appeared to arch them slightly down- ward. It ruffled its back, fluffed its chest and crown, and uttered a minimal version of a prolonged vocalization (roughly: "tk-turrk"), then did a low display flight with one prolonged vocalization. (2) Aerial Display. A bird may launch into display silently or after a single pro- longed vocalization. Usually he climbs rapidly to about 6 or 7 meters above the vegetation, then flies with very full wing strokes. While flying he utters a prolonged vocalization, holding both wings stretched maximally upward during the final flourish of the call, lifting his head and body axis slightly upward but not into the vertical pose of other Muscisaxicola species. Titch- ing forward slightly, and usually veering at least slightly, he then repeats the perform- ance about three or four more times. Some- times he climbs continuously throughout, but if only one prolonged vocalization is uttered, the flight usually does not ascend. Except for the lack of full stalling (also omitted, but rarely, in the other species) and the deep wing stroke, the Aerial Dis- play is similar to that of other members of the genus, and to that of Pijwcephalus rubinus (which also does not usually stall). Other Display About 70 per cent of the prolonged vocalizations heard were from perched birds (the usual stance is shown in Fig. 7h). Their bills open for each call and very widely for the ultimate flourish as the head is thrown back, and the tail and occasionally the wings are slightly flicked. While M. macidirostris never gives its full Aerial Display call from a perch, this usage by M. Irrevicauda may correspond to series of "tk" by an M. macidirostris before (or while apparently not quite ready to perform) an Aerial Display. Wing flicks are occasionally given by disturbed birds, and are very frequent if the individual is very excited. These are simple, small amplitude, forward and up- ward rotations of the wing tips. The tail of M. brevicauda is so short that adults look as if they should be fledglings. Nonetheless, it has a msty terminal bar which probably has some signal function such as making tail flicks more conspicuous. The only other movement seen to involve the tail is landing atop a bush, when the tail is cocked as if for steering or balance. The tail is probably at least partly fanned in the Aerial Display. Some plumage display has been men- tioned aboxe under Wing Raising. lib. Xo/m/s All X. ptjrope and X. irupero were seen in or immediately before the breeding season, and X. velata, X. cinerea, and X. striaticolJis only after the breeding season. Most published accounts report the birds as nearly or wholly silent, and undemon- strati\'e. I heard and saw little display. Xolmis pyrope Xobnis pyrope was observed in Chile near Isla Negra, Cerro Manquehue, and on Tierra del Fuego. Vocal Displays (1) A simple, chevron-shaped call, "pt" (Fig. 8a), sounds very like "tek" of Mtisci- 256 Bulletin Museum of Comparative Zoology. Vol. 141, No. 5 / / / / / /\ Figure 8. Vocalizations of Xo/mi's species: 8a. "pt" of X. pyrope; 8b. prolonged vocalization "wfieet whut T-T- wheeoo" of X. pyrope (an initial "pt" omitted); 8c. "puh" of X. irupero; 8d. several coupled vocalizations recorded in a fight of two X. irupero; 8e. vocalizations by one X. irupero after chasing another; 8f. "pew" of X. irupero; 8g. a nasal vocalization of X. irupero (compare with Fig. lid). saxicohi muculiro.stris and is probably an SV". I noted uses identical to those of that species (q.v. ), except that X. pyrope was not seen to mob a predator. This "pt" may beWetmore's (1926: 301) faint 'V/c/c iic/c." (2) A very brief and relatively high pitched "tseet" was heard several times in a face-to-face aerial fight. Another time, an X. pyrope pursued a conspecific bird closely through some bushes, then immedi- ately after perching called: "tseet pwut, pwut pwut, put." These "pwut" and "put" w ere probably related to the "pt." "Tseet" may correspond to at least the "teek" and "tsk" of M. alhilora; if so, all are SV forms. (3) An encounter between an apparent pair and a third individual was seen on the boundary of a transient territory during spring migration. The third indixidual and one other came to a stand-off, both uttering the l)rief "pt" frequently and countercalling with bouts of a much more complex, pat- terned vocalization: "pt wheet whut T-T- wheeooo" (Fig. 8b, which omits the initial "pt"). The ultimate portion is closely similar to the prolonged \'ocalization of M. macuUrostris, and the whole follows the basic pattern of a series of brief calls and a flourish. (4) A weak, low, monosyllabic whistle is said to be characteristic (Goodall et al., 1957: 151). It may correspond to the "tseet" (IPV) of M. alhilora, but I have not heard it. Nonvocal Displays None was certainly identified. In the single stand-off encounter, both partici- pants sometimes raised both wings quickly and slightly above the back, probably a display similar to the Wing Raising of Muscisaxicola species. The individual which called after terminating a non- aggressive chase had a conspicuously iiif- fled throat. During that chase the white in the tail of both individuals was especially conspicuous. Other Xolmis Species and Neoxo/m;s (a) In northwestern Argentina I found contiguous territories of Xolmis irupero in dry, gravelly river beds below the foothills of Salta and Jujuy, and saw several terri- torial chases. When perched, each partic- ipant would utter a "puh" (Fig. 8c) at long intervals; this call is rather prolonged and flat, with a terminal increase in ampli- tude. The usage suggests the briefer "pt" of X. pyrope, but the length suggests the "sect" of M. alhilora. Sayornis saya uses a prolonged and rather flat IP\' in many ago- nistic situations, and it and S. nigricans use IPX's in situations in which S. phoehe usu- ally uses an SV. Immediately after chasing an intruder, the victor usually gave variable calls, often suggesting the initial portion of the pro- Evolution of Ground Tyrants • Smitli and Viiilleiimier 257 longed vocalization of X. pyrope (see Fig. 8e). One such victor landed and gave a prolonged "pweeeeet" plus three couplets of modified "puh" calls resembling the calls of an X. pyrope after a chase (see above). Some calls during fights had coupled elements (Fig. 8d) and one was a broad chevron ("pew," Fig. 8f). These all ap- pear more like SV than the simple "puh," and most differ primarily in duration and slope. Calls recorded from one fight are different, a set of eight very brief, nasal elements (Fig. 8g) quite reminiscent of an LHV. (b) Little is known about the displays of other Xolmis species. Hudson (1920: 142) remarked that X. coronoto has a "long, low, whistle." Wetmore (1926) has heard a "faint swee" from X. cinerea, as well as a 'little whistled song," used just at dawn and probably corresponding to the pro- longed vocalization of X. pyrope. Of X. riibetra, he says that males "at intervals flew up to make a metallic rattle with their wings as they turned abruptly and dropped to the ground." This is the only indication of which I am aware of some sort of flight display and Wing Whirr or Buzz in the genus. It may be similar to flight displays in Knipolegus and related genera (Smith, in preparation) of a lineage probably fairly closely related to die genera discussed herein. (c) I have seen X. fumigata only in the post-breeding season, and the birds did not display. Vuilleumier (personal communi- cation) has heard a loud "tew" from forag- ing individuals of X. strioticoUis in July, in Ecuador. In Argentina he found one individual in April 1965; it uttered a loud "they-tew-thew." Other accounts of the call of this species (e.g., Todd and Car- riker, 1922; Koepcke, 1964) resemble his. One individual of an apparently un- described population of this species-group was observed by my wife and me on 15 June 1966, at about 4000 meters on the northwest slope of \'olcan Cotopaxi in Ecuador. The bird was medium brown above, paler buffy below, pale yellowish on the throat, and had a very faint super- cilium. Extremely faint traces of broken streaking marked the lower throat and flanks. The bill was black. The plumage was ver)^ fluffed, hiding most of the wings so that it was impossible to tell if these were completely unmarked, but they could not have had prominent bars. In shape, size, and color the bird was very close to specimens I have examined of X. fuscorufa of southern Peru and Bolivia, but lacked prominent wing bars. According to Hell- mayr (1927), the local X. pernix of the Santa Marta range in Colombia is the population phylogenetically closest to X. fuscorufa, but it has much less prominent wing bars and a more reddish coloration (I have seen only the one skin of this popu- lation in the M.C.Z. ). A likely explanation of our bird is that it was a member of a deme intermediate between these two dis- junct populations, and that they are ex- treme members of a linear series of populations. (The genus has another sup- posed rare and local population, X. sig- nata, in addition to X. pernix and our bird. As it is an "island" genus, such populations are perhaps to be expected.) I shall refer to this bird as X. (fuscorufa/ pernix? ) . It perched silently atop a tall bush for several minutes, then repeated a loud "cleeoo" (Fig. 10a), building quickly to a rate of 29 per minute and maintaining this for over five minutes, like the BRV of a Sayornis species. There was slight vari- ation in at least the slope of the terminal arm of the call, which is markedly similar to the RR2 of RRV ("song") of S. nigri- cans (particularly an individual recorded in Panama, Smith, in press, 1 ) . After seven to eight minutes, calling stopped, and the bird dropped to the grass where it stood like a Turdus thiTish for some tens of seconds before dashing off. We were not able to find it again. (d) Neoxolmis rufiventris, closely re- lated to the species of Xolmis, is said to 258 Bulletin Museum of Com pa rat ice Zoology, Vol. 141, No. 5 ^ ^ If _i 1 1 i_ _l I I I L. Figure 9. Vocalizations and Pose of Agriornis species. 9a. "pyuk" of A. montana, Argentina; 9b. "pyuk" of same indi- vidual A. montana; 9c. "wheet hyou" of gray Agn'ornis, Ecuador, (also A. montana?); 9d. "t-eek" of A. Itvida; 9e. "t-eek-ek" of A. livida; 9f. pose of A. livida while uttering "t-eek" calls. have a "long, low, plaintive whistle" ( Hnd- son, 1920: 138). lie. Agriornis I have seen two of the five species, but never in dense populations. Both have a loud whistle which probably serves some "song" functions, since birds utter it from atop shrubs or other relatively tall vegeta- tion at long, irregular intervals (often about 45 to 80 seconds). Agriornis montana, recorded in Tucu- man, Argentina, has the briefest vocal- ization, a relatively low pitched "pyuk" of SV form (Fig. 9a\and b). Goodall et al. ( 1957 ) say this is used in the early morn- ing, which is when I recorded it, and con- sider the species more vocal than A. livida in Chile. Dark gray birds on the northeast slopes of Mt. Pichincha in Ecuador had the longest call (Fig. 9c), a clear "wheet hyou" with a form like a Sayornis saya RRl song unit (see below, and Fig. llh). A bird using it also foraged while under observation, and sometimes countercalled with another individual using the same vocalization. Vuilleumicr has heard a nearly identical call from A. montana in the Potosi Department of Bolivia (see his discussion in Chapter 1). The corresponding vocalization of A. livida in central Chile is somewhat like a Sayornis phoehe IPV: "t-eek" or "t-eek-ek" (Fig. 9d and e, with pose of bird between calls in Fig. 9f). Fully grown, nearly in- dependent fledglings were found at Los Molles, Chile, from 7 to 9 Januaiy 1962. They sometimes perched and called a flat- ter version lacking a terminal peak: "pwut." It sometimes \\'avered or was harsh. One perched alone once and uttered a loose series of weaker "peet" calls that became slightly harsher ("pwt") before it flew off. Later it landed near me, looked at me for Evolution of Ground Tyrants • Smith and Vuilleumier 259 20 seconds, then with a faint, harsh "pwt" began to forage. Still later, when an adnlt came the fledgling gaped toward it and repeated "peet" two to three times per second, the "peet" getting harsher until the adult flew away. The known vocalizations of adults of these Agriornis populations are not closely similar to one another. All, however, re- semble different vocalizations of Musci- saxicola and Sayornis species which are used (at least in part) similarly. The one visible display known is an Aerial Display by an A. oUricouda seen by Vuilleumier (personal communication) in La Paz De- partment, Bolivia. The bird circled silently, alternately rising to a partial stall and dropping forward on closed wings, as in an incomplete Aerial Display of a Mtisci- saxicoJa species. Wetmore (1926) heard "sharp squeaky notes" (perhaps BVs) from Agriornis murina during a pursuit. lid. Ochthoeca Ochthoeca fumicolor I observed this species in Ecuador on Cerro Atacaso in 1959 and 1966, and on Volcan Cotopaxi in 1966. The following vocalizations have been recorded. (1) A very variable, loud, clear, abrupt whistle: "kleeip," "kleeep," or "kleeeh" (Fig. 10c, d, e, and f). The variations have not been correlated with different usages, but are comparable to variations of the very similar IPV of Soyornis phoehe. The call is uttered at long intervals by some foraging birds while perched, or sometimes in flight. Sometimes two countercall with it: e.g., immediately after an agonistic encounter, both participants uttered it be- fore one flew away. An adult repeated it nearly every time it fed its fully grown fledglings, and spanned all of the recorded variations. Most calls of the fledglings (e.g.. Fig. lOg) were similar to one sort of the adult's, and much fainter. All usages strongly suggest the IPV of Sayornis saya. ( 2 ) A less loud "tee-oo" ( Fig. lOb, rarely just the descending arm is uttered) is nearly indistinguishable in sound from the one known call of Xolmis {fuscorufa/per- nix?). At least once, two individuals sepa- rated after one agonistic encounter, then countercalled with this before the next. It is very similar in sound and form to the terminal calls of Aerial Displays in Miisci- saxicola and the prolonged call of Xolmis pyrope, as well as somewhat similar to the one known call of the Agriornis species recorded in Ecuador. All apparently relate to RRV units of Sayornis. (3) Ochthoeca fumicolor also uses pat- terned series of one to three variable "kleep" vocalizations followed by about eight to ten simpler "klee" calls (Fig. lOh, probably also IPV or RRV forms). The latter are nearly invariable in form, and nearly evenly spaced about 0.4 seconds apart, with a slightly longer pause before the ultimate one. In series that had as few as two "klee" units these sounded more like "tee-oo." In 1966 the only series we recorded were during and after an ago- nistic encounter, as two participants countercalled at irregular intei'vals. They eventually shifted from using both "kleeip" and the patterned series to just "kleeip." One individual in 1959 uttered short series while foraging near a calling O. rufipecto- ralis (sec below). (4) Two other vocalizations, recorded during fights, both resemble calls used in fights by at least M. albilora. One (Fig. lOi) sounds like "pw-pwt" or "pw-pw-pwt," the other ( Fig. lOj ) like "twee tik." Both may be uttered in bursts, and at least the former is sometimes used in flight with a Wing Whirr. Both may be LHV forms, and have similarities with the LHV of S. phoehe ("twh-t," see Fig. lie). Some of O. fumicolor s nonvocal displays arc also known. During and after agonistic encounters there was much wing and tail flicking that was rapid during series calls, but there was some also with at least "tee-oo" calls. When foraging with or 260 Bulletin Museum of Comparative Zoology, Vol. 141, No. 5 J ^. u 1 ^^^ .\ Figure 10. Vocalizations of Xo/m/s (fuscorufo/pernix?) and Ochthoeco species: 10a. "cleeoo" of Xo/mis (fuscorufa/ pernix?); 10b. "cleeoo" of O. fum/'co/or; 10c. through 1 Of. variations of the abrupt whistle of O. lumicolor-. "kleeip," "kleep," "kleeeh," and "kleeeh," respectively; lOg. a similar vocalization of a fledgling O. fum/'co/or; lOh. "klee ' vocalization of O. fum/co/or; lOi. a vocalization recorded from a fight of two O. fum/co/or; lOj. another vocalization recorded from a fight of O. fum/'co/or (compare with Fig. lid); 10k. "tyeh" of O. /eucophrys (Peru); 101. "tyee" of O. leucophrys (Argentina); 10m. "tee tee ti," recorded from a fight of two O. /eucophrys (Argentina). without occasional "kleep" type calls, there were occasional tail flicks (always up/ down), especially on landing. A Wing Whirr in several flights during at least one agonistic encounter had tw^o forms: several single or paired snaps about 0.2 seconds apart, or a quick burst of about seven similar sounds in about 0.4 seconds (analyzed from recordings). The individ- ual foraging near an O. rufipectoralis gave very noticeable bursts of whirring every time it flew. Other Ochthoeca Species A single foraging O. rufipectoralis was observed for an hour on 7 August 1959, on Cerro Atacaso. It occasionally uttered a rather faint "cleeoo" (not recorded) which I could distinguish from the "tec-oo" being used by an O. fiimicolor primarily by amplitude. That these two species and the Xolmis {fuscoruf a / pernix?) , all in the same general region, should have such remark- ably similar calls is probably due to more than close phylogenetic relationship. The species probably have some organized social relationship (see Moynihan, 1968, for various possible interspecies relation- ships). Perhaps it is significant that while the only two members of Ochthoeca in the copse during that hour were of different species, both kept displaying but had no encounter. One other vocalization was heard when the O. rufipectoralis was especially active: a long, rapid, series of brief "pt" calls, terminating in two "cleeoo" calls. If this corresponds to the series call of O. fiimi- color the reversed sequence in the two species may be significant in the light of the unusual interspecies similiarity of the aforementioned vocalizations. The brief "pt" calls seem likely to be SVs, and may replace the IPV in this species — such a substitution occurs among the species of Sayornis. Finally, the O. rufipectoralis tail flicked less than its slightly larger congeneric as- sociate, and Wing Whirred only twice, faintlv. However, all its flights undulated like Wing Whirr flights, being broken into segments by recurrent pumping of the relatively long tail. I recorded O. leucophrys near Lima, Peru (September 1962), and in northwest- em Argentina ( November 1962 ) . The lone individual in Peru foraged with an oc- casional abrupt, SV-like "tyeh" (Fig. 10k) when perched or during flights of 6 to 25 meters between bushes. On calling while perched, it would give a slight wing flick Evolution of Ground Tyrants • Smith and Vuilleumier 261 and single tail flick. Once it called "tyeh tuh tuh tuh tuh" in flight as it approached a perch. At the Argentine site two pairs of O. Jeucophrys were in a narro\\' quebrada, and one individnal kept intruding into the area of the other pair. Several times one chased, fought, and expelled the intruder, and in each encounter one used brief bursts of: "tee tee ti" (Fig. 10m). These resemble, respectively, an SV, BV, and SV of Sayor- nis phoebe, and, to a lesser extent, calls recorded from flying Xolmis and Musci- saxicola. Between intrusions the defender foraged with an occasional "tyee" (Fig. 101), slightly more prolonged and more IPV-like than the "tyeh" of the Peruvian bird. Vuilleumier watched two individuals foraging near Mitotambo, Peru, on 27 May 1965, and one or both uttered "wheet" and "wheeyet," calls which may correspond to the "tyeh" (SV) and "tyeee" (IPV) I have recorded. (It is very difficult to correlate onomatopoetic descriptions of different ob- servers, since bird vocalizations do not correspond closely to the vowels and con- sonants we must use to describe them; Vuilleumier and I each heard calls in similar circumstances which fell into two categories by length and were generally similar in pitch.) Vuilleumier ( personal communication ) heard an O. cinnamomeiventris call oc- casionally while foraging below the Carpish Pass, Peru, on 28-30 May 1965: "long, high- pitched seee, very reminiscent of the call of Colorhamphus parvirostris." COMPARATIVE SUMMARY AND DISCUSSION The known behavioral similarities of the above genera are here summarized, and are compared with the behavior of the genera Sayornis (for which all references below are to Smith, 1969, and in press 1 and 2) and Fyroceplialus (for which references are to Smith, 1967, plus subsequent work in preparation). Habitat preferences vary along a con- tinuum transcending generic limits. While some Ochthoeca species inhabit dense forests, others live on the wooded or bushy fringes of the open, grassy paramo, and devote much of their foraging attention to it. In opcness and general distribution of bushes for perching, the latter habitats are comparable to those of Sayornis and Pyro- cepholus. Among Xolmis species, even members of the fiimigata species-group are not found in dense forest, and range pri- marily from edge to semi-open habitats with tall perches. Other Xolmis species carry this trend further, and at least two are as terrestrial as some Muscisaxicola species, although perhaps not identical in their modes of foraging. Extremes of open habitat preference are reached in both Agriornis and Muscisaxicola, but in all cases some feature provides elevated look- out posts. The less extreme members of even these genera have breeding habitats with abundant bushes or some trees. Attention in searching tends to be di- rected downward in some Ochthoeca spe- cies, in Sayornis saya, and at least season- ally in other Sayornis species. This feature is particularly obvious in Xolmis, most species of which perch relatively high for terrestrial foragers. The larger, ground- dwelling Muscisaxicola species direct most of their searching attention downward, al- though even they have some tendency to scan the air for flying prey. Tliroughout the group as a whole, locomotoiy patterns show obvious adaptations to the form of searching, from a tendency to fly close to the ground to the development of rapid hopping and running. All of the species do some or much foraging by gleaning, and possibly a tend- ency to glean is necessary in the e\ olution of terrestrial habits, since most prey in open countiy is likely to be on the ground or clinging to plants. Particularly where open habitats are ^^'indy, flying inverte- brates are relatively uncommon. Yet aerial flycatching persists. It may be of major 262 Bulletin Museum of Comparative Zoology, Vol. 141, No. 5 significance only in Soyornis, Pyrocephahis, and some Oclithoeca species, but all of the species under consideration do it. It is behavior that characterizes the tyrannids among the other terrestrial passerine in- sectivores \\'hich share their habitats. An- other such feature is the tendency of Muscisaxicola and some Xolmis species to alternate short dashes and/or flights with pauses to scan for prey; in central Chile and coastal Peru, Anthus species (motacil- lid), and Geositta and other terrestrial fumariids sympatric with one or more Muscisaxicola species usually walk con- tinuously while foraging. The Old World flycatching muscicapids ( genus Oenanthe ) , ^vith which Muscisaxicola species are con- vergent, use the same foraging tactics, while their sympatric passerine insectivores (at least in Iceland, where these are Anthus and MotaciUa [personal observa- tion] ), are also "walkers." It is as if a fly- catcher, of whatever phylogenetic affili- ation, in taking to terrestrial foraging, preempts this tactic. All members of the assemblage build open cup nests. They employ available opportunities for concealment, placing their nests in vegetation (grass clumps or bushes), in crevices or burrows, or on partly concealed ledges. Sayornis species use only the last, including artificial sub- stitutes like bridges and buildings, and Oclithoeca nifipectoralis has been known to use a similar site. Muscisaxicola is the only genus restricted to terrestrial nesting (if we consider ledges as being off the ground), and Agriornis montana and Neoxolmis rufiventris may be the only other species nesting on the ground (some popu- lations of Xolmis cinerea may nest in holes in banks). That closely related species should differ in nesting on the ground or in bushes suggests that evolution of ground nesting has not been difficult for open country flycatchers. The nondisplay behavior patterns do not appear to set any of the genera apart, but show a general trend from some adaptation to open environments in Sayornis, Pyroccph- aliis, some Xolmis, and some Ochthoeca species, to considerable adaptation in Muscisaxicola, other Xolmis species, and Agriornis. By "trend" I do not mean to imply that any of these genera is a part of the phylogenetic lineage of another. Ob- \'iously, all are contcmporaiy end-products of adaptation to different habitats, but they serve to show the nature of possible stages in the evolution of terrestrial forms from less terrestrial ones. Comparison of the displays is less easy than of the nondisplay behavior. It can best be done by comparison with the dis- plays of the related genus Sayornis, in which all displays of at least S. phoehe appear to be known, and the relationships among thos(> displays are relatively well understood. No attempt is made here to repeat detailed descriptions of the form and usage of Sayornis displays; they are cited below only insofar as is necessary to establish points of comparison or contrast with the other species. Oversimplification is inevitable, and most of the following general statements about form and usage of displays can be qualified by a variety of exceptions. The exceptions do not ap- pear to contradict the conclusions, but a reader wishing a more complete compari- son must consult the more detailed papers on S. phoehe and the genus Sayornis. One important problem is that the amount of use of different vocalizations almost certainly varies with the stage of the breeding cycle sampled during the field work, as has been shown for S. phoehe. Since all the specific display repertoires appear comparable in many respects, sea- sonal shifts can likely be predicted from what is known about S. phoehe. (1) Simple Vocalization (SV). In S. phoehe this sounds like "tp" (Fig. 11a). It is used in a varietv of circumstances, most of ^^'hich are agonistic or potentially agonistic: foraging while patrolling, re- sponses to predators, and agonistic or Evolution of Ground Tyrants • Smith and Vuillcumier 263 potentially agonistic interactions of mates. The SV of S. nigricans and the "peent" vocalization of Pyrocepluilus ruhinus are full chevrons, similar in most known usages. In Muscisoxicolo, at least some larger species divide the SV into a number of intergrading forms, each with its own range of usage. Thus, in M. olhilora, "teek" (Fig. Ic) is usually used in approach toward an interaction which may be agonistic, whereas "tchk" and "tk" (Fig. 2d) are used in grappling fights, and almost all forms are heard when mates meet during or immediately after an agonistic situation. M. frontalis and M. flavinucha appear to have similar SVs, M. rufivertex and M. macloviana may, while M. capistrata either lacks such calls or uses them primarily at other parts of the breeding cycle than were observed. Mtiscisaxicola maculirostris has a chev- ron-shaped homologuc ("tek," Fig. 6a) which it uses much more abundantly than the SVs of any of its larger congeners. The indistinct variations of "tek" appear to cor- respond to the different SV forms of M. aJhiJora. The yet smaller M. brevicauda has only comparable "tk" forms (particu- larly Fig. 7d), and restricts their usage (see below, under LHV). All Miiscisaxicola species employ forms closely related to SVs (and here grouped \^'ith them) in short series or bursts in circumstances suggesting Chatter Vocal- ization (CV) usage of Sayornis species. These usages are either by a bird associat- ing with its mate without attempting other activities, or by males before or in Aerial Displays. The one obser\'ation of CV-type calls by a lone M. frontalis apparently pros- pecting for a nest site is significant, since it is typical of CV usage in many better known tyrannid genera, and in the three Sayornis species and P. rid)inus. The forms resemble the chevron series C\^ of S. nigri- cans (Fig. lid), and similar variants of the CV of S. say a. Xolmis pyrope has a "pt" (e.g.. Fig. Sa) U /l/l /I l\ J I I L. C d I I I I I -J I I f g -I 1 1 I 1 I I I I I I I I /> \/v- /Vs^h - h -J I I I I L J 1 1 I Figure 11. Vocalizations of Sayornis species: 11a. Simple Vocalization (SV) of S. phoebe; lib. Initially Peaked Vocali- zation (IPV, one variant) of S. phoebe; lie. Bipeaked Vo- calization (BV) of S. phoebe; lid. A portion of a CInatter Vocalization (CV) of S. nigricans: lie. Locomotor/ Hesitance Vocalization (LHV) of S. phoebe; 1 1 f. A unit of Cfiatter Vo- calization (CV) from a series of S. phoebe; llg. An Initially Peaked Vocalization (IPV) of S. soyo; llh. An RRl of the Regularly Repeated Vocalization (RRV) of S. soyo; lli. An RRl of the RRV of S. nigricans (Arizona); llj. An RR2 of the RRV of S. nigricans (Panama) (compare with Fig. 10a). which corresponds to the SV of M. macu- lirostris, and X. irupero has similar but more prolonged calls (e.g., "pew," Fig. 8f) as well as coupled elements which are SV or CV. Some calls of other Xolmis species described in the literature appear to be SVs. In Agriornis, at least the "pyuk" of A. montana is of S\^ form and usage. Calls recorded from Ochthoeca leucophrys dur- ing fights ("tee" and variants, Fig. lOm) 264 Bulletin Museum of Comparative Zoology, Vol. 141, No. 5 and apparent patrolling ("tyck," Fig. 10k) are probahK* S\'s. (2) Initially Peaked Vocalization (IPV). In Sayomis phoehe IPV (Fig. lib) is very rare and variable. It is sometimes used by males before going to roost, after an agonistic encounter, or when thwarted in an agonistic situation. S. nif:,ricans uses its honiologue more than S. phoehe, although in similar circumstances as well as in some patrolling and with RRV, and in S. saya the IPV (Fig. llg) virtually replaces the SV. Thus, \\'ithin Sayomis one species uses primarily S\' and little IPV, another uses both, and the third primarily IPV. Similar patterns may recur in the other genera. Muscisaxicola alhilora has a slightly bi- peaked "sect" (Fig. le and f) which occurs between attacks in agonistic encounters like an IPV; its "tseet" (Fig. la) is also IPV-like. M. frontalis has two correspond- ing calls ( Fig. 4a and b ) . Both species also have SVs. M. capistrata may have only IPV forms (Fig. 5a, b, and e). A "tee" heard from M. rufivertex with Wing Rais- ing may be either IPV or SV. But if M. maculirostris has an IPV (perhaps the squeak. Fig. 6b, by an individual on ceasing to approach a fight), it is very rarely used, whereas SV is used abundantly. In XoJmis irupero, "puh" (Fig. 8c) is probably an IP\^ as is the "t-eek" ( Fig. 9d and e) of patrolling Agriornis livida. The most Sfl|/o;-?i/.s-like IPVs are found in Ochthoeca, the genus most like Sayomis in habits and plumage. The variable "kleeip" (Fig. 10c to lOf) used in patrol- ling, countercalling, and after agonistic encounters is the IPV of O. fumicolor, a species with no known SV. Some other calls in this genus may be either IPV or RRV forms (see below). (3) Calls of Nestlings and Fledglings. In Sayomis, nestlings of S. phoehe beg with an SV-like "tee," those of S. saya primarily with IPV forms, and those of S. nigricans with both, thus roughly corr(>s]:)onding to differences in the frequency of usage bv adults of SV and IPV. The begging of young Muscisaxicola maculirostris and M. capistrata suggests a similar correlation in that genus, and in Ochthoeca fumicolor begging fledglings used only IPVs — as did adults. It does appear as if either IPV or SV can serve nearly the same functions in different species, and that which is more prominent is not a generic characteristic. (4) Regularly Repeated Vocalization (RRV). Each Sayomis species has a char- acteristic "song" of two units (called RRl and RR2) repeated in patterns for up to thirty minutes at a time. In S. saya, IPV units make up the bulk of most "song" bouts. The most closely similar perform- ance heard from any of the species de- scribed here was the bout in which the single Xolmis (fuscorufa/pemix?) repeated regularly a "clccoo ' ( Fig. 10a ) with a form like the RR2 of S. nigricans (Fig. llj). Ochthoeca fumicolor has a closely similar call ("cleeoo," Fig. 10b), recorded in countercalling l^outs, and a regularly repeated "klee" (Fig. lOh) which is either an IPV variant or a second RRV unit (see Fig. Hi). O. rufipectoralis has a "cleeoo," and the "tyee" (Fig. 101) of O. leucophrys is similar to an RRV unit (Fig. Hi). It is likely that the "wheet hvou" ( Fig. 9c, com- pare with Fig. llh) of gray Agriomis in Ecuador is an RRV unit, but it was never heard in series. The calls referred to as "prolonged vocal- izations" in several species of Muscisaxi- cola and Xolmis pyrope are like the RRV of Pyrocephalus ruhinus. Although they are not regularly repeated for long, even in Aerial Displays, they are used in counter- calling. Most terminate with "cleeoo" or "clee-ip," and those recorded (Figs. 6g, 6h, 6i, and 8b) appear to be simplified \'ersions of the O. fumicolor "cleeoo" (Fig. 10b), preceded by a series of brief vocalizations which are usually SV- or CV-like. The complex prolonged vocalization of M. hrevicauda (Fig. 7f and g) appears to be a secondary specialization of the common form, conxergent in harshness with the RRl of S. phoehe. Evolution of Ground Tyrants • Smith and Vuilleumier 265 Regular repetition, particularly in pre- dawn bouts, as a characteristic RRV usage, seems lost in the terrestrial genera, and its use in an Aerial Display is correspondingly increased. Sayornis species sometimes, and P. nihinns frequently, use their RRVs in flight displays, and also have predawn calling patterns, and occasional daytime bouts. (5) Bipeaked Vocalization (BV). In Sayornis phoehc and S. nip^ricans this has the form of a rather abrupt and multi- peaked IPV (Fig. lie) and sounds like "T-keet." It is uttered by the aggressor in agonistic encounters and occurs in some Flight Displays. In Miiscisaxicola the most similar vocal- ization is also IPV-like, and forms inter- mediates with IPV. The only larger species it has been recorded from is M. albilora ("tseek," Fig. lb), used by birds flying toward their mates, and in agonistic en- counters between pairs. In M. maculirostris the closest approximation was a "tek" variant (Fig. 6c) used by a bird approach- ing a fight with considerable hesitance. Similarly, in M. brevicauda a "tk" variant (Fig. 7e) may approximate this call. In Agriornis, only the call described by Wetmore from a pursuit in A. miirina may be a BV. A "tee" (Fig. 10m) used in a territorial fight by Ochthoeca leucoplirys appears BV-like. (6) Locomotory Hesitance Vocalization (LHV). Sayornis phoehe has a call ("twh-t," Fig. lie) used primarily on land- ing when there is a conflict of tendencies to continue flying or to take a perch. In early spring it is used during patrolling, but thereafter most commonly by males approaching their usually aggressive mates, or by males being attacked by their mates. Less commonly, a bird appearing "nervous" in any potentially agonistic situation may sometimes utter it on landing from a short flight. S. saya has a similar form known only from male-female disputes. A very similar "twee-tk" ( Fig. lOj ) heard in fights of Ochthoeca fumicolor is the closest ap- parent homologue in the other genera. A nasal call from a fight between two Xolmis irupero (Fig. 8g) may be an LHV. In Muscisaxicola alhiJora, apparent LHV forms were recorded only in a few greet- ings between mates. It is not certain that M. maculirostris has a homologue, but in my brief field work with M. brevicauda, an LHV ("tchek," Fig. 7a) was heard very commonly. It was used by patrolling males, almost always on landing from a flight (as in S. phoebe). No greetings were seen. Within their respective genera, both M. brevicauda and S. phoebe are promi- nent for having harsh calls. Even the CV of S. phoebe is harsh, and intergrades with its LHV. In the other species, CV may replace the LHV in some usages, and the SV and/or IPV in others. (7) Doubled Vocalization (DV). A rare display of S. phoebe, this is usually a couplet of brief chevron-shaped calls. It may represent a partial remnant of the CVs of the other Sayornis species, and is used in some instances in which an LHV might seem appropriate. Only M. brevi- cauda appears to have a call ("tchk," Fig. 7b) which corresponds in usage, and this species is like S. phoebe in having a fre- quently used LHV. (S) Flight Display. The Flight Display of Pyrocephalus is similar both in its form and its usual vocalization to the Aerial Dis- plays of Muscisaxicola species. The Aerial Displays of M. brevicauda differ primarily in having a very full wing stroke, but De- Benedictis (discussed in Smith, 1967) has seen full wing strokes used by some in- dividuals of Pyrocephalus. The conspicuous Flight Display of S. phoebe is probably comparable, but may usually be relatively fast and erratic; few detailed obsen'ations exist. All Sayornis species have fluttering flights with CVs and RRVs and/or IPVs, which differ primarily in lacking stalls and in usually being oriented toward potential nest sites. Aerial Displays may also be found in at least Xolmis and Agriornis. (9) Wing Raising. The elaborate Wing 266 Bulletin Museum of Comparative Zoology, Vol. 141, No. 5 Raising displays known in agonistic en- counters of Muscisaxicola species inter- grade with the Aerial Display. X. pyrope may also have a Wing Raising display, but there is no evidence for homologous dis- plays in the other genera. S. phoehe ap- parently lacks such a display, although its Wing Fluttering display may assume some of the functions of Wing Raising. Males of both S. saya and S. nip-icans have been seem to \\^ing Raise briefly at the end of short fluttering display flights, in what ap- peared to be pair-bonding contexts. (10) Wing Whirr. Abrupt bursts of sound produced by the wings are known in Sayornis, Pyrocephahis, two Ochthoeca spe- cies, three Muscisaxicola species, and prob- ably Xolmis ruhetra. Many other tyrannids, and many other passerines, at least rarely produce similar sounds. Wing Whirrs tend to occur in agonistic situations, sometimes in association with flight displays (e.g., in Tyranmis) or with Wing Raising (very regularly in M. maculirostris) . ( 11) Other Wing, and Tail Displays. The \Ving Shuffle of Sayornis and Wing Flirt- ing of Pyrocephaliis appear to correspond closely. Probably Ochthoeca has similar displays, but related movements of the other genera may not be ritualized (except as a specialization for foraging in Musci- saxicola). Similarly, Tail Wagging in Say- ornis, Tail Flicking in Pyrocephalus, and a tail movement of Ochthoeca species cor- respond closely. Such a display is lacking in at least the more terrestrial genera, where it is to some degree replaced by Wing Raising. Tail movements known in the terrestrial species appear to be pri- marily balancing movements associated with locomotion. (12) Croicn Ruffled. Similar forms of Crown Ruffled are known in at least Say- ornis, Pyrocephalus, and Muscisaxicola. If there are other plumage displays, I have missed them by concentrating on sound recording and by not using binoculars suf- ficiently. In sum, this comparison shows that simi- larities among the displays of all these spe- cies are prominent, while widely different displays apparently do not occur. This is not to claim that there are no novel special- izations in the group; certainly the Head Row of Muscisaxicola is a novelty, special- ized to sliow the species-specific occipital patches, but it is not an elaborate inno- vation. Other less novel .specializations appear to be modifications of identifiable components, and do not observe generic limits. For instance, at least some species in the most open habitats appear to have lost the use of RRV in bouts while perched, and to have elaborated and stylized the Aerial Display instead. And at least the very terrestrial Muscisaxicola species ap- pear to have incorporated in the Aerial Display a Wing Raising which appears to be much rarer and less stylized in two Sayornis species. The displays, to the degree that they are now kno\vii, are less useful in delineat- ing the genera than are morphological and plumage characteristics. Some display char- acteristics may be of some help at this level, however. For instance, very brief vocalizations are typical only of Musci- saxicola (including M. hrevicauda, in \\]iich the component elements of displays are very brief); Xolmis pyrope may be the n^ost comparable species from this aspect in another genus. Further, the IPV and RRV are most similar in form and usage in the fumigata species-group of Xolmis, in Ochthoeca, and in Sayornis, which also intc^rgrade in plumage pattern and color- ation, and in habitat choice and many general behavioral features. Displays vary considerably within each genus, but mostly in ways already known in other genera. For instance, within one or more of the genera described herein, in some species, either the IPV or the SV may be prominent to the virtual exclusion of the other, or both may be common in the same species. This is also the case in the genus Sayornis, and shifts of a similar sort (usually involving the RV and CV) Evolution of Ground Tyrants • Smith and Vuilleumier 267 are known in tlie phylogenetically distant tyrannic! genus Tijrannus (Smith, 1966). The display behavior of Muscisoxicola hrevicauda is at the moment the most obvious case of intrageneric distinctness, and the species is undoubtedly not fully typical of the genus Muscisaxicola. For a bird which is aberrant in so many ways, however, it is the similarities of its displays with those of other Muscisaxicola species that are striking, not the differences. Fur- ther, the displays of M. hremcauda are more like the known displays of Musci- saxicola species than they are like the known displays of any other passerines of the Andean chain. Considered alone, these displays suggest that M. hrevicauda is at least closely related to the members of Muscisaxicola, and differs largely through its emphasis on the LHV and related dis- plays. Considered in the light of ecology and other behavior (see above), plumage, morphology, and geographic distribution (see Chapter 1 by Vuilleumier), they help suggest that the species belongs in Musci- saxicola, as its most aberrant member. Finally, the behavioral similarities among tlie birds of these genera indicate that they belong to a natural group, the limits of which have not yet been determined. Other tyrannid genera which overlap with some in their habitats (e.g., Anairetes, Elaenia, and Mecocerculus in the f)aramo of Ecuador and Peru) are quite different in many aspects of their repertoires (Smith, in preparation). The various Tyrannus species have little in common with these species in the detailed forms of their dis- plays, even though there is at least general comparability in the usages (as should be expected for species with basically similar social behavior). In addition, sympatric nontyrannid passerines are generally quite dissimilar from the genera described herein in the forms of their displays, and this is true even of the superficially similar flight displays of some species of Pliryf!,ilus and Anthus. Thus there is little reason to suspect that the many similarities among the displays of the species of Muscisaxi- cola, Agriornis, Xolmis, Ochthoeca, Pyro- cephalus, and Sayornis are wholly or largely due to convergence in the face of common environmental pressures. The similarities of display behavior, like the similarities of nondisplay behavior, suggest close relationship. CONCLUSIONS General behavior patterns of the ter- restrial tyrannids in the genera Muscisaxi- cola, Agriornis, and Xolmis are similar in many respects, and no strongly divergent behavioral trends are known. In habitat preferences, foraging methods, patterns of locomotion, selection of nest sites and in the structures of the nests they build, they appear to vary along continua that trans- cend generic limits and relate these species to the less terrestrial members of Xolmis, Ochthoeca, Pyrocephalus, and Sayornis, all of which they also resemble in structure, coloration, and plumage patterns. Detailed similarities in the form and usages of the displays of members of all of these genera do not appear to be determined primarily by evolutionary pressures related to the habitats of the birds, and indicate that they belong to one phylogenetic group, distinct from many sympatric tyrannids. Some spe- cialized differences in display forms have developed among the genera, but none constitutes a major innovation. The forms of the displays give some clues as to re- lationships within the group, indicating that the less terrestrial genera may be more closely related to each other than any is to Muscisaxicola, and that Muscisaxicola is a fairly coherent genus (with the partial exception of the aberrant M. hrevicauda) that is probably closely related to Xolmis through X. pyrope. Greater detail of inter- relationships can likely be revealed by more detailed studies of display behavior, but it is not clear that this will necessarily aid significantly in the recognition of generic limits within this group. 268 Bulletin Museum of Comparative Zoology, Vol. 141, No. 5 LITERATURE CITED Cade, T. J. 1962. Wing movements, hunting and displays of the Northern Shrike. Wilson Bull., 74: 386-408. GooDALL, J. D., A. W. Johnson, and R. A. Philippi B. 1957. Las aves de Chile, su conocimiento y sus costumbres. 1. Buenos Aires: Piatt Establecimientos Graficos S. A., 441 pp. Hailman, J. P. 1960. A field study of the Mock- ingbird's wing-flashing behavior and its as- sociation with foraging. Wilson Bull., 72: 346-357. Haverschmidt, F. 1953. Wing-flashing of the Graceful Mockingbird, Mimus gilvus. Wil- son Bull., 65: 52. Hellmayr, C. E. 1927. Catalogue of birds of the Americas and tlie adjacent islands in Field Museum of Natural History. Part V. Tyrannidae. Field Mus. Nat. Hist., Publ. 242, Zool. Ser., 13: 1-517. Hltdson, W. H. 1920. Birds of La Plata. Vol. 1. New York: E. P. Dutton, 244 pp. Ihering, H. von. 1904. The biology of the Tyrannidae with respect to tlieir systematic arrangement. Auk, 21: 313-322. Johnson, N. K. 1963. Biosystematics of sibling species of flycatchers in the Empidonax ham- mondi-oherhoheri-iorightii complex. Univ. Calif. Publ. Zool., 66: 79-238. KoEPCKE, M. 1954. Corte ecologico transversal en los Andes del Peru central con especial consideracion de las aves. Parte 1: Costa, vertientes occidentals y region altoandina. Mem. Mus. Hist. Nat. "Javier Prado," 3: 1-119. . 1958. Die Vcigel des Waldes von Zarate. Bonn. Zool. Beitr., 9: 130-193. 1964. Las aves del Departamento de Lima. Lima: Morsom S. A., 128 pp. Kunkel, p. 1962. Zur Verbreitung des Hiipfens und Lauf ens unter Sperlingsvogeln ( Passeres ) . Z. TierpsychoL, 19: 417-439. Lanyon, W. E. 1960. The Middle American populations of the Crested Flycatcher Myi- archus tyrannidus. Condor, 62: 341-350. . 1961. Specific limits and distribution of Ash-throated and Nutting Flycatchers. Con- dor, 63: 421-449. . 1963. Experiments on species discrimi- nation in Myiarchus flycatchers. Amer. Mus. Novitates, No. 2126: 1-16. . 1965. Specific limits of tlie Yucatan Fly- catcher, Myiarchus yiicatanensis. Amer. Mus. Novitates, No. 2229: 1-12. . 1967. Revision and probable evolution of the Myiarchus flycatchers of tlie West Indies. Bull. Amer. Mus. Natur. Hist., 136: 329-370. Maclean, G. L. 1969. The nest and eggs of the Chocolate Tyrant Neoxolmis rufiventris (Vieillot). Auk, 86: 144-145. Marchant, S. 1960. The breeding of some S. W. Ecuadorian birds. Ibis, 102: 349-382. Moynihan, M. 1960. Some adaptations which help to promote gregariousness. Proc. 12th Internat. Ornithol. Congr., Vol. II: 52.3-541. . 1968. Social mimicry; character con- vergence versus character displacement. Evo- lution, 22: 315-331. Mumford, R. E. 1964. The breeding biology of the Acadian Flycatcher. Univ. Mich. Mus. Zool. Misc. Publ., 125: 1-50. Olrog, C. C. 1959. Las Aves Argentinas. Tucu- man, Argentina: Instituto Miguel Lillo, 377 pp. Pena, L. E. 1961. Results of research in the Antofagasta Ranges of Chile and Bolivia. 1. Birds. Postilla, 49: 1-42. Peters, J. L. 1923. Notes on some summer birds of northern Patagonia. Bull. Mus. Comp. Zool., 65: 275-337. Schauensee, R. M. de. 1964. The Birds of Colombia and Adjacent Areas of South and Central America. Narberth, Pennsylvania: Livingston. 430 pp. Selander, R. K., and D. K. Hunter. 1960. On the function of wing-flashing in mockingbirds. Wilson Bull., 72: 340-345. Smith, W. J. 1966. Communication and relation- ships in the genus Tyrannus. Publ. Nuttall Ornitliol. Club, No. 6: 1-250. . 1967. Displays of the Vermilion Fly- catcher (Pyrocephalus rubinus). Condor, 69: 601-605. . 1969. Displays of Sayornis phoebe. (Aves: Tyrannidae). Behaviour, 33: 283-322. ■ . (in press, 1). Song-like displays in Sayornis species. Behaviour. (in press, 2). Displays and message assortment in Sayornis species. Behaviour. Stein, R. C. 1958. The behavioral, ecological and morphological characteristics of two popu- lations of tlie alder flycatcher, Empidonax traillii (Audubon). N. Y. State Mus. Sci. Serv. Bull., 371: 1-63. Todd, W. E. C, and M. A. Carriker, Jr. 1922. The birds of the Santa Marta region of Co- lombia: a study in altitudinal distribution. Ann. Carnegie Mus., 14: 1-611. Wetmore, a. 1926. Observations on the birds of Argentina, Paraguay, Uruguay, and Cliilc. Bull. U. S. Nat. Mus., 133: 1-448. msmmimmmmMmu Museum of Comparative Zoology iiiiliii Systematics and Natural History of the Mygalomorph Spider Genus Antrodiaetus and Related Genera (Araneae: Antrodiaetidae] FREDERICK A. COYLE HARVARD UNIVERSITY CAMBRIDGE, MASSACHUSETTS, U.S.A. VOLUME 141, NUMBER 6 13 JULY 1971 PUBLICATIONS ISSUED OR DISTRIBUTED BY THE MUSEUM OF COMPARATIVE ZOOLOGY HARVARD UNIVERSITY Bulletin 1863- Breviora 1952- Memoirs 1864-1938 JoHNSONiA, Department of Mollusks, 1941- OccAsiONAL Papers on Mollusks, 1945- Other Publications. Bigelow, H. B., and W. C. Schroeder, 1953. Fishes of the Gulf of Maine. Reprint, $6.50 cloth. Brues, C. T., A. L. Melander, and F. M. Carpenter, 1954. Classification of Insects. $9.00 cloth. Creighton, W. S., 1950. The Ants of North America. Reprint, $10.00 cloth. Lyman, C. P., and A. R. Dawe (eds.), 1960. Symposium on Natural Mam- malian Hibernation. $3.00 paper, $4.50 cloth. Peters' Check-list of Birds of the World, vols. 2-7, 9, 10, 12-15. (Price list on request. ) Turner, R. D., 1966. A Survey and Illustrated Catalogue of the Teredinidae ( Mollusca: Bivalvia ) . $8.00 cloth. Whittington, H. B., and W. D. I. Rolfe (eds.), 1963. Phylogeny and Evolution of Crustacea. $6.75 cloth. Proceedings of the New England Zoological Club 1899-1948. ( Complete sets only. ) Publications of the Boston Society of Natural History. Authors preparing manuscripts for the Bulletin of the Museum of Comparative Zoology or Breviora should send for the current Information and Instruction Sheet, available from Mrs. Penelope Lasnik, Editor, Publications Office, Museum of Comparative Zoology, Harvard University, Cambridge, Massachusetts 02138, U.S.A. © The President and Fellows of Harvard College 1971. SYSTEMATICS AND NATURAL HISTORY OF THE MYGALOMORPH SPIDER GENUS ANTRODIAETUS AND RELATED GENERA (ARANEAE: ANTRODIAETIDAE) FREDERICK A. COYLE* CONTENTS INTRODUCTION 270 ACKNOWLEDGMENTS 270 NATURAL HISTORY 271 Antrodiaetiis unicolor 271 Antrodiaetus rohustus 284 Antrodioeius pacificiis 285 Antrodiaetus occultiis 288 Antrodiaetus pugnax 288 Antrodiaetus montanus 289 Antrodiaetus hageni 291 Antrodiaetus cerherus 292 Antrodiaetus tjesoensis 292 Antrodiaetus lincolnianus 292 Antrodiaetus stygius 293 Antrodiaetus apachecus 294 Antrodiaetus roretzi 294 Atypoides riversi 296 Atypoides gertschi 300 Atypoides hadros _ ^— - 302 Aliatypus californicus 303 Aliatypus no. 1 304 Aliatypus no. 2 305 Aliatypus spp. 306 Summary and Discussion 306 Natiu-al History Tables 312 Natural History Figiues 316 TAXONOMIC METHODS 325 Analysis of Variation 325 Material 328 Collecting Methods —- 328 Morphological Terminology — 328 Methods of Presentation 329 * Department of Biology, Western Carolina varsity, Cullowhee, North Carolina 28723. Bull. Mus. Uni- TAXONOMY 330 Family Antrodiaetidae 330 Key to Genera of Antrodiaetidae 331 Genus Antrodiaetus 331 Key to Species of Antrodiaetus 332 The Unicolor Group 334 The Lincolnianus Group 366 The Roretzi Group 370 Genus Aliatypus 372 Taxonomic Tables -.- 373 Taxonomic Maps 380 . Taxonomic Figures 381 EVOLUTION 393 Phylogeny 393 Dispersal Ability and Barriers 396 Geographic Variation 396 Speciation: Probable Historical Events 398 Reproductive Isolating Mechanisms 400 LITERATURE CITED ABSTRACT 400 A taxonomic section includes a characterization of the family Antrodiaetidae, a key to its three genera {Antrodiaetus, Atypoides, and Aliatypus), a revision of Antrodiaetus, and a preliminary study of Aliatypus. The family, except for two Japanese species of Antrodiaetus, occurs only in North America. Of the 13 species of Antrodiaetus, four are newly described: A. occultus, A. cerherus, A. stygius, and A. apachecus. Six new synonymies are recognized. Natural history data are reported for all species of the three genera. Antrodiaetid evolution, with emphasis on geographic variation and speciation in Antrodiaetus and Atypoides, is discussed. Conip. ZooL, 141(6): 269-402, July, 1971 269 270 Bulletin Museum of Cotnparative Zoology, Vol. 141, No. 6 INTRODUCTION In general morphology and behavior AntrocUaefm (Figs. 109-112) and the closely related genns, Atypoicles (Coyle, 1968), resemble the stout-bodied burrow- ing trapdoor spiders of the distantly related mygalomorph family Ctenizidae. All spe- cies construct tubular silk-lined burrows in the ground and prey nocturnally upon invertebrates which wander within reach of the burrow entrance. This entrance is well camouflaged and is closed off by a collapsible silken collar (or a rigid turret). The long-legged adult males abandon their burrows during the mating season and wander nocturnally in search of females. Most species are found in humid well- drained soil in forest or woodland habitats. Antrodiaetu.s and Atypoides together form a distinct monophylctic unit within the atypoid mygalomorph spider family Antrodiaetidae. The 13 species of An- trodiaetus and the three of Atypoides ex- hibit in toto a markedly disjunct geographic distribution with species in Japan, western North America, and eastern North America. The center of species diversity in both genera is in far western North America. These are clearly the dominant (most abundant and widespread) mygalomoqoh spiders above 35° latitude on this conti- nent. Taxonomic revisions of Antrodiaetus are nonexistant. Many species descriptions, par- ticularly pre- 1900 descriptions, are based on immature specimens or tiny samples, lack diagnostic characters, and are poorly illustrated. Misidentifications are common in the literature. Little has been written about the natural history of antrodiaetid taxa. Such a state of affairs is largely due to the relatively simple external morphol- ogy and covert behavior of these spiders. I have based the taxonomic revision of Antrodiaetus, like that of Atypoides (Coyle, 1968), upon a thorough and objective search for diagnostically useful characters, my primary tool being an analysis of vari- ation of quantitative and qualitative mor- phological characters. Because I believe that many kinds of nonmorphological data provide useful, often vital, and interesting information about populations, taxa, and their past history, I have also made a start at a comparative behavioral, ecological, and life history study of the Antrodiaetidae. Using both morphological and nonmorpho- logical data, I have discussed the evolution of Antrodiaetus and Atypoides. Included is a preliminary consideration of the re- lationship of the only other antrodiaetid genus, AUatypus, to these two genera. It is my wish that this paper will stimulate fur- ther interest and continued research on these and related spiders. ACKNOWLEDGMENTS This work is a revised version of a doc- toral thesis completed at Harvard Univer- sity (Coyle, 1969). I am indebted to my advisor. Dr. Herbert W. Levi, for his en- couragement and generous support. Like- wise I am grateful to Dr. Willis J. Gertsch for his willingness to have me study these spiders which have deeply interested him for many years. Much of the material on which this study is based has accumulated in the American Museimi of Natural His- tory collection through his efforts. Dr. J. A. Beatty has shown uncommon interest in this study, generously supplying speci- mens and field observations. Dr. W. H. Bossert wrote the computer program for the analysis of variation. I am most deeply grateful to my wife, Judy, for moral sup- port, for assistance in field work, and for help with several tasks, especially typing. I sincerely thank the numerous other helpful people and institutions that have loaned material for study; these are named in the Methods section. C. W. Sabrosky and Drs. H. E. Evans, W. A. Shear, and A. A. Weaver have provided identifications of parasites, predators, and some prey. National Science Foundation Graduate Fellowships and a Richmond Fellowship Systematics and Biology of Antrodiaetus Coyle 271 from Harvard University have been the major sources of support for this study. Funds from NSF grant GB 3167 (Reed C. Rolhns, Harvard University, Principal In- \'estigator) helped support my field work during the smnmers of 1966 and 1967. Public Health Service Research Grant AI- 01944 to Dr. Levi has helped to defray some expenses. NATURAL HISTORY Only three species of Antrodiaetidae — Antrodiaetus unicolor, Antrodiaetus roretzi, and Atypoides riversi — have had more than a few sentences written about their ecology, life history, or behavior. In this section I shall summarize the natural history litera- ture of the family and add original data. Primarilv because of mv field studies in 1966, we now know much more about A. iinicolor than any other antrodiaetid. A comparative summary and discussion of the natural history of the family is placed at the end of this section after the separate consideration of each species. An enormous amount of information is still needed be- fore a valuable comparative biological study of the family can be completed. Hopefully this paper will trigger the search for such information. My observations were made chiefly in the field during the summers of 1966 and 1967, but I also observed Hving spiders in the laboratory. Most observations oi> Alia- typus species were contributed by Wendell Icenogle. All specimens are numbered and have been placed in the American Museum of Natural History (AMNH) and the Museum of Comparative Zoology (MCZ); the correspondingly numbered field notes have been placed in the MCZ. Data on burrow architecture refer to adult female burrows unless stated otherwise. In the text many localities are represented by italicized letters; these localities are identi- fied in the records section at the end of the appropriate taxonomic species description. Antrodiaetus unicolor (Hentz) Ecology Geographic distribution and elevation range. Centered in the central and south- em Appalachian region of the eastern U. S. with peripheral populations as far west as the Ozark region and south to near the Gulf coast ( Map 1 ) . 0-6600 ft ( 0-2000 m ) . Habitat. Optimum conditions for A. unicolor are found in hnmid, cool, densely forested habitats with sandy loam soil. Such habitats are abundant and nearly continuous in the dissected terrain of the central and southern Appalachian Moun- tains and theii" foothills and westward over the Appalachian Plateau. The densest and largest burrow aggregations are usually found on the inclined surfaces of slopes and banks along ravines, streams, roads, and trails; at the base of rock (particularly sandstone) outcrops; and in other sheltered spots. Hemlock (Tsuga) and Rhododen- dron are often good plant indicators of optimum A. unicolor habitats. Over the more peripheral portions of the range of A. unicolor, one finds only scattered pock- ets of favorable habitat separated by larger areas of more or less unfavorable (dry, poorly drained, etc.) habitats. Although sandy loam is optimum, burrows also occur in other soil types, even sometimes in clay soil, decaying logs, etc. Bunows are nor- mally uncommon in rocky, shallow soil that is too well drained. Within favorable habitats burrows tend to be aggregated where the soil is con- stantly humid (at least around the lower portion of the burrow) and protected from erosion and flooding. In dry regions bur- rows are often clmnped only near springs, along stream banks, or near seeps at the bases of rock outcrops. Burrow aggre- gations are seldom found on near-hori- zontal ground, unless such ground is well sheltered under something such as a rock outcrop, a sandstone shelter cave, or the base of a large tree. Two limiting factors that may contribute to this clumping under 272 Bulletin Museum of Comparative Zoology, Vol. 141, No. 6 shelters are the temporary localized flood- ing and the buildup of fallen leaf cover on open noninclined ground surfaces. Such flooding and (usually) leaf litter accumu- lation do not occur on the steeply inclined slopes and banks where A. unicolor bur- rows are common. On these inclined sur- faces, burrows are clearly aggregated in stable spots where erosion is minimal, e.g., moss covered areas and bare parts of the bank sheltered under the overhang of ground surface vegetation, exposed rocks, logs, large roots, and tree bases. Branson and Batch's ( 196(S ) statement that A. uni- color "lives in burrows mostly beneath fallen logs and deep litter" is erroneous. Population density. The densest and largest populations were found at C, D, I, J, K, O, P, and Q within the centi-al and southern Appalachian region. Even at such favorable localities, the i^opulation con- sisted of dense aggregations in optimum habitat spots and scattered individuals in intervening areas of less suitable habitat. One of the densest such burrow aggre- gations covered an area of about 50 m^ and contained an estimated 80 to 100 adult females and several thousand immatures. Small dense aggregations were common within the larger aggregations in the above localities. A 0.1-m- area at O contained four brooding females and 25 immatures. An 0.81-m- area at P contained three adult females and 198 immatures. A 225-cm- area at B contained two adult females and 31 immatures. In many dissected regions the populations are concentrated and es- sentially continuous throughout the inter- connecting ravines, young valleys, and coves, with the surrounding matrix of dryer ridges and upland areas very sparsely populated. In the peripheral portion of the A. tmicolor range, aggregations are (like the favorable habitat) discontinuous and generally smaller and less dense. Associated bumncinfi, myfialomorph spi- ders. Throughout most of its range, A. unicolor is clearly the most abundant mygalomorph spider. Except for areas of overlap with the northern Antrodiacttis rohustus (Map 1), only in southern and western parts of its range (in lowland liabitats and in the Ozark region) is the abundance of A. unicolor equaled and sometimes surpassed by other burrowing mygalomorph species. I have found the follo\\'ing such spiders living side by side with A. unicolor at the following localities: Afypoides hadros Coyle at F (These two species have been found together in other southern Illinois localities by other collec- tors.); MyrmekiuphiJa fluviatilis (Hentz) at H, O, R, Q, and Mammoth Cave, Ken- tucky; Cyclocosmia triincata (Hentz) at Q, R, and David Crockett State Park, Tennessee; Ummidia audouini (Lucas) at Chickasaw State Park, Tenn.; Ummidia carahivora (Atkinson) at N; and Ummidia spp. at Q and Blanchard Springs, Arkansas. At F, A. hadros was common but not as abundant as A. unicolor. At both Q and R, M. fluviatilis was common but much less abundant than A. unicolor. In one low road bank at Q, A. unicolor, M. fluviatilis, C. truncata, and Ummidia sp. lived to- gether, the latter two species being rare. At all other localities cited above, the other mygalomorph spiders were very rare. Atypus species are sympatric with A. uni- color but uncommon; I have never ob- served them at the same locality. I have not found Lycosidae burrows in A. unicolor aggregations. Life History Males. Males of A. unicolor (and all other Antrodiaetidae) closely resemble fe- males in external anatomy and behavior until the striking metamorphosis of the final male ecdysis. Only a few obvious, sexually dimorphic characters — a sclero- tized lip at the male genital opening, swollen male pedipalpal tarsi, and well- developed seminal receptacles of the fe- males— appear during the later immature instars. The penultimate male undergoes the final molt ^^ithin its burrow, and the adult remains there without feeding until Systematics and Biology of Antrodiaetus Coyle 273 certain factors trigger its abandonment of the burrow. Upon emergence, it wanders over the ground surface in search of female burrows, mates, and eventually dies. The data summarized in Figure 1 in- dicate that the period of male wandering and mating for the entire species popu- lation of A. unicolor lasts from late July through late December. Because of oc- casional long-lived males, the mating sea- son may be shorter than the period during which wandering males have been col- lected. The high frequency of wandering male records during October indicates that this may be the chief mating season of A. unicolor. However, all August records of mature males collected within their bur- rows were from cooler climate localities (C, I, and P) than the October collections and therefore possibly belong to earlier (September ?) emerging populations. Evidence indicates that within any given local population the time of male emer- gence (and probably mating) occurs more nearly synchronously and is probably trig- gered by local climatic conditions. At N, J. Beny conducted a systematic, year-long pitfall trap collection resulting in a rel- atively complete record of male emergence in that population (J. Beatty, personal communication). Traps were put out for approximately 24-hour periods iwo times each month (only once a month in the winter) in each of 11 different areas. Four- teen of the 18 adult males collected. were taken from 22 to 24 October (ten of these on 24 Oct. ) following the first heavy rain- fall after a dry summer. The other four males were collected on 1 October, 7 No- vember, and 3 January. Pitfall traps main- tained continuously from 7 October to 10 November at E by J. Nelson several years later (J. Beatty, personal communication) yielded 11 males during the initial week (7-14 Oct.), two the second week, one the tliird, one the fourth, and none during the final week (3-10 Nov.). A similar pitfall collection maintained by J. Nelson the following year at G from 10 September to 2 November yielded the first male during the week 28 September to 5 October, seven males the next week, one the following week, none the next week, and one during the final week (26 Oct.-2 Nov.). Collections during August revealed a pre-emergence synchrony of adult males at localities C, I, and P. Twenty-one of the 22 adult males collected were within their own burrows. I searched hard for wander- ing adult males during day and night but found only one. It appears that very few males had emerged and that proper con- ditions could have set off a large synchro- nous emergence. A few winter records exist for wandering males of A. unicolor — an early January record from N and a literature record from D of "males observed on [the] surface during late winter" (Branson and Batch, 1968). I have not seen the latter males (from D) so only assume that they are A. unicolor. Such records are probably either aberrant late emerging individuals or indix'iduals which have extended their above-ground survival time in particularly protective microhabitats. Females. Potfern of carlii postemhnjonic development. Enough data were obtained from stereomicroscopic examination of each preserved A. unicolor brood to establish the following pattern of early postembry- onic development. Hatching involves the simultaneous (or nearly so) shedding of the chorion and embryonic or "prelarval" (Vachon, 1958) cuticle. Tlie resulting instar, which I shall call the first instar, is the "first free postembrvonic stage" of Holm (1954) or "larva" of Vachon (1958). The second instar corresponds to the "first complete postembrvonic stage" of Holm (1954) or "first nymph" of Vachon (1958). The segmentation and appendages of the late embryo, as well as the dark egg teeth, are visible through the chorion. During hatching, first the chorion splits anteriorly and peels back over the abdomen. Usually the embr>'onic cuticle splits very soon after the chorion and also peels posteriorly, and 274 Bulletin Museum of Comparative Zoology, Vol. 141, No. 6 the chorion and cuticle are shed simul- taneously from the posterior of the abdo- men and remain attached to each other. Occasionally, possibly because the interval between the splitting of these two mem- branes increases, they are shed separately. These conclusions about the timing of hatching are based on the following ob- servations of preserved material: 1) Rel- atively few spiderlings in any hatching brood were actually in the process of peel- ing back or shedding the chorion or em- bryonic cuticle, and, in the great majority of those that were, the chorion was either peeling back or was still attached to the spiderling while the embiyonic cuticle was jDeeling back. 2) Almost all shed chorions observed contained the embryonic cuticle. The embryonic cuticle is very thin and transparent, lacks setae, has blunt non- functional chelicerae, and possesses a dark egg tooth at the base of each pedipalp. The first instar (Fig. 4) possesses several tarsal setae, lacks a rastellum or fang on the chelicera, lacks spinneret spigots, has anterior lateral spinnerets, and lacks pig- mentation, except for that \\'hich slowly de\'elops around the eyes. Tlie second instar (Fig. 5) has functional chelicerae and spinnerets and can locomote rapidly. Tlie chelicerae possess a few rastellar macrosetae and fully formed fangs, the an- terior lateral spinnerets are lost, the other two pairs of spinnerets possess spigots, setae are plentiful, and the body is pig- mented. This pattern of development is the same as that described for Ischnothcle karschi (Dipluridae) by Holm (1954) and for Atypus karschi (Atypidae) and Hepia- thela kinmrai (Liphistiidae) by Yoshikura (1955, 1958). It is considered by Holm ( 1954 ) to be the most primitive pattern in spiders. Vachon ( 1958 ) reviewed these studies and is in agreement on the simi- larity of patterns in the above three species. Timing of brood development . The nu- merous field collections of A. unicolor broods yield substantial data on the timing of brood development in nature within the species population as a whole. As shown in Figure 2, the egg stage is found from late June to mid-August, the first instar from mid-July to mid-September, and the second instar from mid-August through at least December. The relatively small number of broods in the process of hatching and molting is apparently a result both of the brevity of hatching and ecdysis in in- dividual spiderlings and of the close syn- chrony of development within each individual brood. The regular temporal pattern of the brood stages and the absence of more than a single brood in the burrow of any brooding female strongly indicate that A. unicolor produces only one brood per year in nature. As Table 1 shows, brood development is not strictly syn- chronous within a local population, popu- lations C and O showing the largest amount of variation. These data also indi- cate little or no lag in timing of brood development at warm climate localities (N, O, Q, R) behind that in cooler lo- calities. Substantial evidence indicates that the spiderlings overwinter in the second instar within the maternal burrow and disperse in the spring. As indicated in Table 1, 15 brooding females were collected between 14 September and 3 December from three localities. Each brood consisted entirely of second instar spiderlings, and none of the broods appeared to be diminished. In the December collection from Q, the spiderlings were quite sluggish as a result of the cold weather. During this collection and the September collection at F, I searched hard for bun-ows constiiicted by second instar spiderlings but the smallest burrows found contained only later instars. Two samples of second instar spider- lings from the December Q broods were kept alive at a nearly constant temperature of 12°C. One sample was checked in mid- February and the other in mid-March and both were still 100 per cent second instar broods. These were then released at room Systematics and Biology of Antrodiaetus Coylc 275 temperature into a jar of sand, and they eon- structed burrows within 24 hours. Few data exist concerning longevity and the time normally required for an in- dividual to mature. Oviposition to the beginning of the third instar may take nine or ten months. I \\'Ould estimate at least three or four years from egg to sexual maturity. I have collected two brooding females (384 at Blanchard Springs, Ark., on 11 Sept.; 378 at F on 15 Sept.) with new cuticle visible under the old cuticle, thus indicating that females continue to molt after becoming sexually mature and after producing broods. The following evidence indicates that a given female may be capable of producing more than one brood. A number of brooding female bur- rows were each located in the center of a cluster of homogeneously-sized burrows of small immature spiders. Three of these brooding females (787 at Fort Mountain, Ga., 433 at Q, and 448 at F) were completely isolated from other large burrows. Brood size and egg, size. The sizes of complete broods collected at different lo- calities are given in Table 2. Brood size statistics for the species as a whole are listed in Table 3. All populations with significant sample sizes show a wide range of variation in brood size and overlap ^^'ith one another, but the means are often quite different. Table 4 indicates that there is much variation in egg size of A. tinicolor. Egg size variation within an individual brood is quite small. Behavior Burrow structure. A. unicolor constructs a roughly tubular silk-lined burrow which is widened just below the entrance and at the bottom end (Figs. 6-9). The detailed shape of the burrow, its size, and the thick- ness of its silk lining vary greatly, depend- ing on the size, sex, and reproductive state of the spider, and especially on soil con- dition. Tables 5 and 6 show the variation in burrow length and entrance diameter in A. unicolor. The latter measurement is more closely related (directly) to body size than is burrow length, which is more influenced by soil conditions. The mean size of penultimate male burrows is mark- edly smaller than that of adult female burrows, probably because of the smaller body size of the males. Soil with many pebbles, rocks, or roots usually contains crooked and /or shortened burrows. Bur- rows in hard clayey soil are shorter than those in finn sandy soil. Soil stability (and possibly humidity) appears to have con- siderable effect on the thickness of the burrow lining. Burrows in loose soil tend to have thicker linings than those of the same population in more stable soil. Usu- ally the upper part of a burrow has the thickest silk lining, this probably being, at least in part, an adaptive response to the relative instability (or low humidity) of surface soil. Tlie rest of the burrow is sometimes only thinly lined with silk. Brooding females (even with unemerged broods) often have thicker burrow linings than adult females without broods. All brooding females at O had markedly thicker burrow lining than many brooding females at other localities. Burrow inclination is strongly correlated ^^'ith the inclination of the immediate ground surface. The burrow pitch — here defined as the angle fonned between the mean longitudinal axis of the burrow and the line originating at the entrance and perpendicular to the surrounding ground surface (i.e., the deviation of the burrow from the perpendicular) — of the great majority of A. unicolor adult female bur- rows is 0-15°. Exceptions are often the result of rock or root barriers in the soil. Only one or two burrows sloped upward from the entrance. Tlie probable adaptive significance of such a near-peipendicular burro\\' pitch is that it removes as much of the burrow as possible from the surface soil, which is structurally and climatically (humidity and temperature) more un- stable than the deeper soil layers. This 276 Bulletin Museum of Comporative Zoology, Vol. 141, No. 6 might be particularly important on the steeply inclined banks often inhabited by A. iinicolor where surface erosion is relatively great. Humidity, structural sta- bility, or temperature gradients are di- rectional cues that might be used in burrow construction. Occasionally, A. tinicolor burrows have single, short, dead-end side branches. These are either close to the surface (Fig. 10) or, rarely, at the bottom end. One Y-shaped burrow (389 at F) with two entrances was found. Only once have I found a spider in a side branch. Entrance structure. The silk lining of the burrow is extended above the soil sur- face to form a flexible collar which is col- lapsed inward to close off the open end of the burrow (Figs. 11-14, 40-43, 59). This collar is composed of excavated soil particles and materials from the immediate ground surface (such as pieces of twigs, leaves, conifer needles, moss, or just sur- face soil) held together by silk so that its outer surface is camouflaged and its inner surface silk-lined. The kind and quantity of surface material incorporated into the collar depends on its abundance on the immediate ground surface. When open the collar is more or less erect, and when closed it is roughly flush with the ground surface. The relative height of the collar varies noticeably within populations. Popu- lation O appears to have collars with the largest height/diameter ratio. The spider collapses the collar in a bilateral manner by pulling inwardly on two opposing sectors (lateral to the spider), producing a condensation of fold- ing at the t\vo opposing points where these sectors meet. When the burro\\' entrance opens onto an inclined ground surface, these two points are at the ventral and dorsal sectors of the entrance opening, because of the spider's stereotyped forag- ing posture orientation. The collar (espe- cially of the larger spiders) is frequently somewhat reduced at these two points, which, if the entrance opening happens to be slightly elliptical, also correspond with the ends of the long axis of such an ellipse. Careless obsei"vation of such a collapsible collar by Atkinson (1886) led to his erroneous description: "Each door is a surface of a half circle, is hung by a semicircular hinge, and the two meet, when closed, in a straight line over the middle of the hole. ..." Unfortimately the majority of subsequent descriptions (Simon, 1892; Comstock, 1912 and 1940; Gertsch, 1949; Kaston and Kaston, 1953; Kaston, 1964 and 1966; Pimentel, 1967) of Antrodioettis bm^row entrances are ap- parently based on Atkinson's inaccurate description. Under certain conditions, a closed collar may be held shut by a small amount of silk spun over the inside surface of the collar. Such sealing is usually performed before a spider molts. Tlie burrow en- trances of brooding females are sometimes similarly sealed. Soil plugs are apparently used much more infrequently to seal the burrow entrance. I have found these soil plugs in only five medium-sized burrows (at O and /) and one adult female burrow (1115 at P). Tlie immature spiders were molting and the adult female had an egg sac. In each of these, the upper 0.5-1.5 cm of burrow just behind the closed collar was packed tightly with soil (Fig. 8). Burrow construction hehavior. Nonnally an individual probably occupies and en- larges the same burrow during its entire life span, the burrow site initially being chosen by the second instar spiderling. However, I have collected several medium- sized immatures and one adult female A. unicolor on the ground surface away from any burrow and have occasionally observed empty burrows with open collars, no spider remains, and no penultimate male molts. Under laboratory conditions, large im- mature male and female spiders and adult females are capable of constructing bur- rows from scratch. These observations indicate that the abandonment of burrows Systematics and Biology of Antrodiaetus CoyJc Til and the founding of new burrows at differ- ent sites may not be uncommon in nature. The following account of burrow con- struction behavior is based on laboratory observations of ten large immature (male and female) and mature (female) spiders burrowing into humid packed sand; it is incomplete and subject to revision when more detailed observations are completed. The initial step is the construction of a shallow enclosure or cell. Next, sometimes after a period of inactivity, the spider ex- tends this closed cell deeper into the soil until the elongate burrow is formed. Most observations were made on the initial cell construction phase, ^^'hich lasted from 15 to 30 minutes. Four rather well-defined behavioral components form cell construction be- havior: 1) burrowing, 2) excavating, 3) silk application, and 4) raking. Burrowing consists of pushing into the sand, forcing it apart, and compacting it. An initial de- pression is made by forcing the palps and first legs into the sand surface. Very soon, however, as the size of the depression in- creases, the other appendages enter into the burrowing movement. Full-formed burrowing behavior consists of the chelic- erae pressing forcibly against the side of the bottom of the depression (or burrow) with the legs flexed over the dorsum of the cephalothorax and abdomen and press- ing in the opposite direction against the opposing side of the cavity (Fig^ 56). Excavating involves digging sand from the bottom of the depression, transporting it, and releasing it on the soil surface. The chelicerae are elevated and spread apart laterally, and the fangs are extended. These are forced ventrally into the sand, flexed, and then lifted away with the sand held in between. Tlie spider then pivots to reverse its direction and moves up to the top of the cavit\^ The palps aid in holding the sand in the chelicerae. With its front legs, pedipalps, and chelicerae it reaches out of the cell opening over the soil surface (Fig. 57). It then releases the load by opening its fangs and chelicerae, and pulls away from the load while holding the load against the soil surface with its pedipalps. Silk application includes circular and lateral movements of the abdomen and the entire body (with the spider facing down into the cell) and sweeping movements of the spinnerets (particularly the elongate posterior lateral spinnerets), during which silk is applied to the cell wall and most abundantly to the rim of the cell opening (Fig. 58). In raking, the first legs, the pedipalps, and sometimes the chelicerae and fangs reach out past the rim of the cell opening and pull surface detritus and soil back to the rim of the opening and partially over the opening. Tliis material is often compacted by being pressed against the chelicerae by the first legs and pedipalps. After the spider releases it, it is held in place by silk previously applied to the rim. Cell construction usually begins after considerable searching behavior in which wandering alternates with mild burrowing, ^^'hich is usually released by depressions in humid sand. Eventually a series of bur- ro\\'ing mo\'ements is repeated in one spot. Burrowing usually continues until a fairly deep depression is formed, and then ex- cavating begins and alternates sporadically \\'ith burrowing. The frequency of bur- rowing decreases, and soon silk application begins and, sporadically, but with increas- ing frequency, accompanies burrowing and excavating. If it follows excavating, silk is usually applied to the sector of the open- ing rim and cell wall where the soil load ^^'as dropped. When the cell becomes deep enough to contain the spider, raking begins and often follo^^'S excavating. Usually only a few instances of raking will suffice to pull the rim of the cell opening shut (or almost shut) and thus end the initial stage of burro\\' constiiiction. Burrowing, excavating, and silk appli- cation continue during the rest of burrow construction, but the manner in which the load of sand is deposited was quite differ- 278 Bulletin Museum of Comparative Zoology, Vol. 141, No. 6 eiit in the only two spiders observed after initial cell construction. Instead of being deposited on the ground surface, each load \\as pressed into the burrow wall just be- low the top of the closed cell. Apparently the pedipalps held the sand against the \\'all while the chelicerae and fangs were pulled awav, then the sand was pushed into the wall with the rastellar area of the chelicerae. This behavior eventually caused the area around the entrance to bulge outward into a mound. Such internal dep- osition of soil, if it is normal behavior, would seem to be an effective adaptation for minimizing the exposure of the spider to predation. Collar formation has not been observed. Atkinson (1886) observed burrow con- struction behavior in several spiders kept in captivitv, and although his description is often difficult to understand, his ob- servations generally agree with those above. He observed too, that after the initial cell is closed over, the excavated soil is deposited and pushed into the wall within the upper end of the burrow. "Dome" apparently refers to the closure over the cell. Atkinson claimed that the spider would apply "viscid liquid" as well as silk to the excavated soil deposited at the rim of the cell opening. It is not clear what he meant by "viscid liquid." Egg sac structure and position. The egg mass of A. unicolor takes the form of a flattened sphere or clli]isoid with a maxi- mum dimension of usually 9-14 mm and a thickness of usually 5-8 mm. Apparently tliis mass is formed on a portion of the burrow lining which has been thickened with a few additional layers of silk. Several layers or sheets of silk are then apparently applied over the egg mass and hold it firmly against the burrow wall, each sheet larger and attached to the burrow wall slightly farther from the egg mass than the previous sheet (Figs. 6-9). J. R. Tripp, who has recently observed oviposition and egg sac construction in this species, con- firms that this is the method of egg sac construction ( personal communication ) . Egg sac silk has the same texture and white color as burrow lining silk. As shown in Tables 5 and 6, the position of the egg sac in the burrow is quite vari- able for A. unicolor. However, while some population samples exhibit wide variation in egg sac position, others (Q and espe- cially O) show much less variation and have all egg sacs deep within the burrow. In the populations in which brooding fe- males were often actively feeding (O, Q, and R), egg sacs were well below the sur- face. It is possible that these tsvo phenom- ena are adaptively linked, since it is likely that egg sacs near the entrance would be more easily damaged during prey capture than egg sacs attached deep \\'ithin the burrow. Possible environmental causes should be investigated. All but one of the 101 egg sacs were placed on the ventral sector of the burrow wall. Most egg sacs produced a localized constriction of the bmrow lumen, but never so much that the spider could not maneuver past it. In at least most cases this constriction is counter- acted, because the burrow is somewhat widened in the vicinity of the egg sac. Overwintering. On 28 December bur- rows were observed at B on an overcast afternoon (noon temperature = 0°C). Five centimeters of snow lay on the ground except under the shelter of a sandstone wall where most of the burrows were lo- cated. The upper 1 cm of the sand soil was dry, loose, and unfrozen; the next 9-15 cm frozen; and below this humid, packed, but unfrozen. No burrows were seen with open collars. Close examination of about 20 small and medium-sized burrows revealed only two entrances sealed inside with silk and none plugged with soil. Four burrows were completely excavated. Two of these had one or t\vo very thin silk septa spun across the narrowed central part of the burrows. Two of the burrows did not extend below the frozen zone, the other two did. All four spiders were quite slug- Systematics and Biology of Antrodiaetus • Coyle 279 gish but became active after several seconds in my hand. On the night of 2 December and in day- light the next morning, I observed a bur- row aggregation at Q. The ground was unfrozen, but the air temperature was 3°C at 9 P.M. and -2°C at 8 a.m. At night, during a light snow and rain, I was sur- prised to find that most burrows of all sizes were open, and most of these had spiders at the entrance in normal foraging posture. These spiders were more sluggish than usual, however. In the morning ( clear weather), a few entrances were still open but no spiders were in the entrances. No sealed entrances could be found. This ability to remain active at relatively low temperatures may contribute to the re- markable success of Antrodiaetus in the temperate zone. Bella vior of brood inp, females and broods. Tlie data summarized in Table 7 show that the broods emerge from the egg sac just before, during, or just after they molt from the first to the second instar. In broods that were emerging when col- lected, only a few first instar spiderlings (when present) but most of the second instar spiderlings (when present) were found outside the egg sac. Also, in emerg- ing broods and in the few completely emerged broods still with egg sacs, most of the first instar exuviae were found in the egg sac. One exception, 477 at 7, had these exuviae scattered all over the burrow wall. These data indicate that most emergence is performed by the active second instar spiderlings. Tliere is no evidence indicat- ing how the spiderlings escape from the egg sac or how the egg sac is disposed of. In 59 of the 66 burrows with completely emerged broods, the egg sac was no longer present. Table 8 indicates that while in some populations (O, Q, R) brooding females with egg sac broods do not seal their bur- row entrances and usually continue forag- ing, in others (B, C, /) they commonly maintain sealed burrows. Such sealing of burrows is in contrast to neighboring con- temporary nonbrooding females, which, unless molting, usually kept their entrances unsealed and were actively feeding. How long and for what function such burrows are kept sealed is not known. I was unable to determine whether the sealing of bur- rows containing emerged second instar spiderlings was initially produced by the adult and thus signifies inactivity or whether it is simply the result of spider- lings trailing dragline silk over the inside of the entrance. At least a number of such brooding females had unsealed burrows and were actively feeding. First instar spiderlings, extremely slow moving, are sometimes seen moving within the egg sac. Second instar spiderlings are active — they can move quickly in response to stimuli, wander over the burrow wall, and are capable of performing many behavior patterns necessary for inde- pendent existence. Several second instar spiderlings released on humid sand in the laboratory constructed individual burrows and assumed normal foraging postures. Second instar spiderlings are capable of spinning draglines and nonnally appear to do so as they move about. When a brooding female burrow with a closed entrance is first discovered, second instar spiderlings are often found in the upper end of the burrow, but these always re- treat to the bottom of the burrow as it is excavated. This may be a photonegative response, but other behavioral controls must be operative in preventing these spiderlings from escaping at night if the burrow entrance is open and the brooding female is feeding (such as at Q in Sep- tember and December). Each of the five broods observed on 2-3 December at Q was aggregated in the bottom end of the burrow, several in dense clusters; and all spiderlings were sluggish. There are no data on whether spiderlings normally feed within the parental burrow before dispersal. An observation of a sec- ond instar spiderling which had not yet 280 Bulletin Museum of Comparative Zoology, Vol. 141, No. 6 constructed a burrow but which carried a collembola in its cheHcerae, and obser- vations of other second instar spidcrhngs capturing prey after constructing their burrows, indicate that they are at least ca- pable of feeding before dispersal. Intra- brood cannibalism is possible, but the broods observed in fall and winter did not appear to be smaller on the average than those collected earlier. Moltinfi behavior. Fragmented remains of one and occasionally two old exuviae were often found to be bound with silk and compacted into the bottom end of burrows. I collected (in July, August, and September) ten immature spiders each \\'ith an incompletely sclerotized cuticle and a freshly fragmented exuvia mashed into the bottom of its burrow. Seven of these recently molted individuals had sealed their entrances with silk, and three were discovered by scraping the soil (whether the entrances were sealed could not be detennined ) . Four other immatures (also collected during the summer) were found just after ecdysis; thev were much less sclerotized and could not move quickly. These were at the bottom end of their burrows, and the intact exuviae were attached lightly to the ventral sector of the wall of each burrow at one-half to three-fourths of the distance from the en- trance to the bottom end. One exuvia was attached with its venter against the burrow wall, and its anterior end heading up the burrow. The orientations of the other three were not recorded. The entrances of three of these burrows ( 133 at /, 951 at C, and 747 at Brasstown Bald, Ga.) were sealed with silk, the fourth (355 at /) was filled with a soil plug. Another medium- sized immature spider (7040 at Q) was excavated shortly before ecdysis, which it underwent within a glass vial kept humid by a moist piece of paper towel. Ecdysis lasted ten minutes. Within four days of ecdysis, the spider had thoroughly frag- mented the exuvia and formed it into a compact mass bound together with a small amount of silk. Twenty-seven burrows of recently molted adult males were excavated. All burrow entrances were closed, but 23 were discovered by scraping the soil surface; whether they were sealed could not be determined. Three of the other four were sealed with silk; the fourth was unsealed. In all but two of these 27 burrows the penultimate exuvia was partly broken up, bound with a little silk, and packed lightly into the bottom end of the burrow. Some- times this exuvia and the end of the bur- row were covered with several thin sheets of silk. The other two burrows each con- tained a pale unsclerotized male resting motionless at the bottom end of the bur- row. In each the freshly shed exuvia was attached — venter against the wall, anterior end heading up the burrow — to the silk lining of the ventral sector of the wall, two-thirds (409 at F) and three-fourths (450 at Blanchard Springs, Ark.) of the distance from the entrance to the bottom end. Apparently, molting behavior in penulti- mate males and other immatures is similar. In summary, ecdysis usually takes place in a sealed burrow on the ventral sector of the burrow wall in the lower half of the burrow (but above the bottom end). Shortly after ecdysis, the spider moves to the bottom end. After sclerotization reaches a certain point, the spider detaches the exuvia from the burrow wall, breaks it up, and compacts it into the bottom end of the burrow, adding silk. Adult males do not usually fragment the molt as thoroughly or pack it into the end of the burrow as compactly as do females or im- mature males. Tlie sealing of burrows before molting may help protect the molt- ing spider from predation and may main- tain a higher, more constant humidity within the burrow. Defensive ])eh(ivi()r. When spiders in nomial foraging posture are disturbed by light, substrate vibration, or prodding with Systematics and Biology of Antrodiaetus Coijle 281 a stick, they sometimes retreat down the burrow without closing the collar or may quickly close the collar, often remaining for a time just below the collar holding it closed. During the later stages of excavat- ing to expose a burrow, one finds most adult female spiders at the bottom end of the burrow in an aggressive defensive pos- ture. The spider faces up the burrow with its cephalothorax reared backwards at an angle to the abdomen. Legs IV and III and possibly II hold onto the burrow wall to anchor the spider. The other legs and pedipalps are raised and spread apart. The chelicerae are also raised and spread apart laterally with the fangs extended. If prodded, the spider strikes forcibly with a sudden downward movement of the cephalothorax and chelicerae. Such a de- fensive posture within the confines of the bottom end of the burrow limits the ap- proach of an attacker to that part of the spider protected by its chelicerae. Younger individuals and adult males tend to be less aggressive and often do not exhibit such defensive behavior while being excavated. Medium and large spiders will often ex- hibit similar defensive behavior outside their burrows if prodded. Predators and parasites. Observations show that females and immature males of A. unicolor are preyed upon and parasi- tized by centipedes, pompilid wasps, acro- cerid flies, and mites. Adult males have been collected several times frorn the ground webs of theridiid spiders. Four instances of apparent centipede prcdation have been observed. At Clifty Falls, Indiana, a 50-70 mm long scolo- pcndromorph centipede escaped from a burrow containing a recently killed spider. At L a scolopendromorph centipede, Thea- tops postica, was found in a burrow with spider remains. At both O and Blanchard Springs, Arkansas, a Theatops spinicauda centipede was found in an empty A. uni- color burrow. One instance of unsuccessful centipede prcdation was observed at C where a nearly dead Lithohitis foiiicatus centipede (lithobiomorph) was found in the burrow of a large immature A. uni- color. The centipede had probably entered the burrow through the soil; the burrow entrance was sealed shut. Twenty-one cases of prcdation by pom- pilid wasps have been observed (Table 9). No larvae could be reared to adulthood, thus no specific or generic identification could be made. More than one species is probably involved since variation in cocoon size is great. Spiders with attached eggs or early instar larvae were found in the bottom of their burrows dorsal side up and partially ]3aralyzed. Leg jerking was the chief response to disturbance. The wasp egg was always attached centrally on the abdominal dorsum. Soil plugs apparently constructed by the ovipositing female wasp sealed the spider and larva into some biu-- rows (Fig. 15). Cocoons were usually attached to the burrow wall some distance from the burrow end, where the spider re- mains were usually found. The hollow usually found in the burrow wall just above the soil plug appears to be the source of the plug. One tiny, recently hatched pompilid larva and its paralyzed food source from C were placed in a narrow glass vial and observed through cocoon fonnation. Feed- ing lasted four days. The spider remains consisted of a small packet of compacted spider cuticle. On the fifth day, the larva had moved away from the spider remains. It first spun a loose maze of silk threads between the walls of the tube and then constructed the cocoon within this maze. Cocoon construction took less than 24 hours. Population P was parasitized by the dipteran Eulonchu.s marialiciae Brimley of the Acroceridae. Observations were made on 1 and 2 August. This is presently the only host record for the genus Eulonchiis, which, because its distribution is other- wise limited to western North America ( Schlinger, 1966 ) , and, because it is similar to that of the antrodiaetids, may be an im- 282 Bulletin Museum of Comparative Zoology, Vol. 141, No. 6 portant parasite of other antrodiaetid spe- cies. Considering that collected specimens of E. morioUciae are extremely rare (C. W. Sabrosky, personal communication), the population discovered at P was surprisingly dense. One freshly emerged adult was col- lected vibrating its wings just inside an A. tinicolor burrow entrance. One pupa was collected from a burrow and kept for several hours until the adult emerged. Several other burrows were found with abandoned pupal skins and spider remains. A total of about 18 adult acrocerids were seen hovering near A. uniculor burrow aggregations. The adults appeared to fly only during the daytime. The largest aggregation of adults ( about 12 ) fomied rather quickly as I was excavating several burrows and hovered close to the ground where I was digging, as if attracted by some chemical released by my digging. Twice I observed adults hovering over and momentarily landing near (or on?) a closed burrow entrance. It is probable that, as in other acrocerids, the first instar larva actively searches out the spider, penetrates the cuticle, and develops as an endoparasite. My observations indicate that the larva, after feeding on the spider in the bottom end of the burrow, climbs up the burrow wall, attaches somewhere above the bot- tom end, and completes development in this spot. ( One pupa and three pupal skins wcr(> attached here. ) Immature parasitic mites have been found on 18 adult females from nine dif- ferent localities scattered over the species range. These were usually attached to appendages as well as to the body proper. It is likely that mite parasites on other specimens have gone unnoticed. Prey capture behavior. The normal foraging posture of A. unicolor is as follows ( Fig. 59 ) : The spider is positioned inside the burrow just below the opened collar and faces out the entrance with its pedipalps and first legs extended slightly forward and touching the inner surface of the open collar. Usually the spider's venter is against the ventral-most sector of the burrow wall. Sometimes a spider may be in this posture with the collar only partly opened. Often prey capture behavior is also released in spiders which are po- sitioned farther down the burrow or which have their collars closed. Such individuals, however, appear to respond more slowly to prey stimuli. Larger immature spiders and adult fe- males usually open their collars and assume the typical foraging posture only in very reduced daylight and in darkness. Younger spiders, particularly the youngest, fre- quently leave their collars open and often assume the foraging posture during day- light hours as well. The less light (heavily shaded habitats, overcast days, early morn- ing and late afternoon hours) and the greater humidity, the more likely a given spider will be found with an open collar or in a foraging posture. A. unicolor is thus primarily a nocturnal feeder with some tendency to feed also during daylight hours, depending upon environmental con- ditions, age, and probably nutritional con- dition. Observations indicate that the prey sensing radius of A. unicolor does not ex- ceed several centimeters. Substrate vi- bration transmitted via the burrow lining and collar is apparently the only stimulus used by the spider to detennine the pres- ence and location of prey. This conclusion is indicated by numerous observations (in nature and in the laboratory) of prey capture behavior being released simply by the observer gently vibrating the collar or adjacent ground surface with forceps or sticks, thus eliminating visual and chemical cues. Additional evidence also rules out visual signals. In the typical foraging postiu-e, the spider's field of vision is only a limited inverted cone of space projecting out from the biurow entrance; when the great majority of prey is caught, it has not entered this cone. I have fre- quently observed spiders within closed Systematics and Biology of Antrodiaetvs Coijle 283 collars sensing and capturing ground sur- face prey. Prey capture is commonly per- formed in the dark; moreover, the eyes of antrodiactids are quite small. Casual observation in nature and in the laboratory, as well as motion pictures of prey capture behavior in lab animals, in- dicate the following pattern for prey cap- ture behavior. After having determined the approximate location of the prey by tactile stimuli, the spider lunges from the burrow entrance at the prey. Legs IV and, usually, III remain within the burrow anchoring the spider to the burrow wall. The pedipalps and legs I are held out in front of the spider, and legs II also usually leave the burrow. The tarsi and penulti- mate segments of the pedipalps and legs I contact the prey and hold it against the substrate (Fig. 60). Lateral rows of strong suberect macrosetae on the tarsus and tibia of the pedipalps and on the metatarsus of legs I and II aid in holding the strug- gling prey and probably also provide im- portant sensory input. Sometimes legs II aid in holding the prey. While the prey is held against the substrate, the chelicerae spread apart laterally, the fangs are ex- tended, and the spider strikes forward and down at the prey with the chelicerae. The fangs contact the prey at two well-sepa- rated points (Fig. 61). There may be one to several more such strikes and maneuver- ing of the prey with the pedipalps and legs before the prey (usually still strug- gling) is quickly pulled down into the bur- row by the pedipalps and legs I (Fig. 62). The spider does not usually close the collar until later, presumably after the prey has been subdued. Sometimes, however, friction of the prey or spider legs against the collar as the prey is being pulled into the burrow partly or completely closes the collar. The time from the beginning of the lunge to the disappearance of the cap- tured prey into the burrow for seven filmed prey captures of seven different spiders ranged from 0.8 to 4.5 seconds, with a mean of 1.8 seconds. The time from the beginning of the initial lunge until contact was made with the prey for four of these spiders ranged from 1/32 to 1/16 of a second. Atkinson ( 1886 ) described the capture of an ant. After detecting the ant, the spider (juickly closed the collar, leaving it open only a crack. When the ant crossed this opening, the spider threw open the doors and caught the ant. Such behavior was observed several times, and Atkinson concluded that the spider did not strike until it could see the ant through the opening. That vision actually did play such an important role is doubtful. Prey composition. Records obtained from prey rejectamenta and freshly killed prey found in burrows in nature indicate that A. unicoJor is markedly unselective in its choice of prey. It appears to capture and feed upon any arthropod that it is capable of catching and killing (speed, strength, and size of the potential prey probably being important factors) that chances to enter within the prey-sensing radius of the burrow entrance. The prey records in Table 10 indicate that ants and beetles form a large part of this species' diet. This is not surprising considering the abundance of these arthropods on the ground surface. Howe\'cr, thinner-cuticled arthropods are probably more common prey than these records indicate; the rejectamenta records are strongly biased toward such thick exo- skeleton forms as ants, beetles, and milli- pedes which are not as finely fragmented by the spider. The ants consisted of both formicine and myrmicine species, \\'ith the large- bodied formicine genus Camponotus com- mon. Numerous beetle families were repre- sented. Three millipede families were found: Parajulidae (five individuals), Xystodesmidae (two individuals), and Polydesmidae (one Scytonotus gramtlatus, one Pseudopolydesmiis sermtiis, and two Psetidopolydesmiis sp. ). One of the spider prey was a conspecific male; another was a male Agelenopsis. Two phalangids be- 284 Bulletin Museum of Co7nparative Zoology, Vol. 141, No. 6 longed to Leiobunum. One wasp was a Vespula. In the laboratory, A. nnicolor readily captures and eats Porcellio isopods and Tene])rio beetle larvae. Small im- niatures take wingless DrosopJiiki. Identi- fiable prey ranged in length from about 2 mm to a 16-mm carabid beetle and a 25- mm Pseiidopohjdcsmus millipede. It is noteworthy that A. unicolor is capable of killing and digesting arthropods (ants, wasps, and millipedes) well known for their stings and/or chemical defenses. Disposal of prey remains. Exoskeletal remains of digested prey are apparently disposed of in two ways: 1) packed into the burrow wall, 2) discarded on the ground surface outside the burrow. In most burrows, prey rejectamenta is found mixed with a small amount of silk and compacted into the bottom end of the bur- row. All except the toughest exoskeletal portions (ant heads and thoraxes; beetle heads, thoraxes, and elytra) are usually fragmented. Sometimes similarily com- pacted rejectamenta is also present in parts of the burrow wall closer to the entrance. These trash packets possibly mark previous positions of the bottom end of that par- ticular burrow. Some burrows with large amounts of rejectamenta in the bottom end had collembola aggregated on and near the trash. Occasionally spiders kept in the labora- tory disposed of pieces of prey remains by placing them on the soil surface outside their burrows. At least some of these pieces were only partly digested and were decaying; this disposal behavior may be an adaptive response to such decay. Other evidence indicates that prey remains are disposed of outside the l)urrow: many bur- rows lack or have very little rejectamenta in their walls; a large amount of frag- mented insect cuticle lay on the ground surface in the midst of a dense concentra- tion of burrows at B. The periodic exca- vation to enlarge a burrow and the concomitant disposal of trash packed into the burrow end very likely explain both of these latter types of observations. In summary, the primary method of trash dis- posal could involve two steps: 1) packing the fragmented exoskeleton into the bottom end of the burrow and 2) discarding this compacted rejectamenta on the ground surface during burrow enlargement (or after a sufficient amount of trash collects in the burrow end). Such a method might mean less exposure to predation than \\'Ould the surface disposal of each prey exoskeleton separately. Mating behavior. Only a single obser- vation exists on the time of day of male wandering; I collected a wandering male (637 at F) on 3 August at 9 p.m. (after dark), 90 minutes after a heavy rain. The mostly digested remains of an adult male A. unicolor were found in the burrow of a conspecific female (543 at I). Antrodiaetus robustus (Simon) Ecology Geographic distribution and elevation range. Eastern Ohio east through Penn- sylvania, northern Virginias, and Maryland (Map 1). 0-1500 ft (0-460 m). Habitat. I have observed A. robustus in only one locality; Caledonia State Park in southern Pennsylvania. Small clusters of burrows were scattered over steep, well- shaded banks on a hillside near a stream in mixed conifer-hardwood forest. White pine and hemlock were dominant near the burrow sites. Most bmrows were under the shelter of trees, tree trunks, or exposed rocks and roots, and were as common in spots with a thick surface mat of pine needles as where surface litter was absent. The soil consisted chiefly of firmly packed light tan loam with pebbles. Life History Males. The data in Figure 1 indicate that male wandering and mating occur during September and October. Females. The second instar morphology of A. robustus is like that of A. unicolor. Systematics and Biology of Antrodiaetus • Coyle 285 Four records (Fig. 2) indicate that the timing of brood development may be similar to that of A. tinicolor. The presence of a fully sclerotized second instar brood within a parent burrow at Canton, Ohio, on 19 October weakly indicates that broods may overwinter within the parent burrow before dispersing. The sizes of the two completely collected broods are given in Table 3. Behavior Burrow structure. The five adult female burrows observed (Table 6) were similar in structure to A. tinicolor burrows. Tliey were widened just below the entrance and at the bottom end, and entirely or almost entirely lined with silk. The silk lining was thickest at the upper end. Burrow pitch ranged from 0 to 10°. All burrows sloped downward from the entrance. The burrow of one immature had a dead-end side branch near the bottom end, another im- mature burrow had a similar side branch near the surface. Entrance structure. Tlie silk lining of the burrow is extended above the soil sur- face to form a flexible collar like that of A. unicolor. This collar is well camouflaged and is opened and closed in the same man- ner as that of A. unicolor. Egg sac structure and position. Both egg sacs observed were constructed and po- sitioned like A. unicolor egg sacs. Both were in the bottom half of the burraw on the ventral sector of the burrow wall (Table 6). Behavior of brooding females. Both fe- males with egg sacs (collected near mid- night) had their collars closed and sealed lightly on the inside with silk. Molting behavior. Fragmented exuviae were sometimes found to be bound with silk and compacted with prey rejectamenta into the bottom end of burrows. Defensive behavior. All individuals in foraging posture at night retreated down the burrow when disturbed by light or vibrations. Adult females and often im- matures assume an aggressive defensive posture like that of A. unicolor. Both brooding females were particularly aggres- sive. Feeding behavior. The foraging posture of A. robustus is like that of A. unicolor. Individuals assumed this stance (with the collar open) only in twilight and at night. Fragmented prey cuticle (primarily of ants and beetles) was found packed into the bottom ends of most burrows. Antrodiaetus pacificus (Simon) Ecology Geographic distribution and elevation range. The Pacific Coast of North America from San Francisco Bay north to southern Alaska, with outlying, perhaps isolated, populations in the montane areas of west- em Idaho, northeastern Oregon, and ad- jacent Washington (Map 2). 0-7500 ft (0-2280 m). The range of A. pacificus extends much farther north than that of any other North American mygalomorph spider. Habitat. The habitat requirements of this species appear to be similar to those of A. unicolor. Optimum conditions for A. pacificus are apparently found in humid, cool, rather densely forested habitats with deep loam or sandy loam soils. Such habi- tats are particularly abundant in the moist lowlands and mountains along the Pacific Coast from northern California into Alaska (Fig. 63), but are less continuous inland, there to be found in the mountains of Idaho, eastern Washington, and eastern Oregon. A. pacificus also occurs in the volcanic "sand" soil of the Cascade Mountain Range, and even constructs burrows in de- caying logs in Douglas fir forest. Burrows are sparse in rocky areas with shallow, fast drying soils. The densest aggregations were discovered in soil that was humid and firm at or just below the surface. Within favorable habitats, burrows tend to be 286 Bulletin Museum of Comparative Zoology, Vol. 141, No. 6 clustered in sheltered spots protected from erosion and flooding. Such spots are found on the stable parts of slopes and banks along ravines, streams, trails, and narrow roads. Moss-covered bank surfaces, or portions under the shelter of root, log, rock, or tree overhangs are, as in A. unicolor, common sites of burrow aggregations. The densest and largest population was found at I where one burrow cluster con- tained four adult females, one adult male, and 36 immature spiders within 0.25 m-. The largest and densest population in the eastern part of this species' range was found at U. Life History Males. The data in Figure 1 indicate that male wandering and mating occur between early June and early November, with peak activity perhaps from late July to late September. Records from a few relatively well-collected localities hint that, as in A. tmicolor, male emergence is more nearly synchronous in any given locality and is controlled by local climate. Females. Close examination of the four A. pacificus broods indicate that the pat- tern of early postembryonic development is the same as in A. unicolor. Almost all of the shed chorions examined contained the embryonic cuticle, and very few of the spiderlings actually in the process of hatch- ing had lost the chorion but not the embryonic cuticle. Therefore hatching in- volves the splitting and peeling back of the chorion and then the embryonic cuticle in close succession, followed by the (usually) simultaneous shedding of these mem- branes. The morphology of the embryonic cuticle and the first instar is as in A. unicolor. The few data (Fig. 2) indicate that the timing of brood development of A. paci- ficus in nature may be similar to that of A. unicolor. Evidence that spiderlings may overwinter in the second instar and that adult females are long-lived was provided by an adult female (2246 at R) collected on 3 July. Although she was rearing no brood that year, her burrow was in the center of a 30 cm diameter aggregation of 36 burrows of spiderlings that matched the size and moq:)hology of second instar A. unicolor spiderlings. Since she was isolated from other adult burrows, these spiderlings almost certainly belonged to her brood of the previous year. Data on brood size and egg size are given in Tables 3 and 4. Behavior Burrow structure. A. pacificus burrows (Figs. 16 and 17) are very similar in struc- ture to A. unicolor burrows. The tubular burrow is slightly widened just below the entrance and at the bottom end and is narrowed between. The silk lining is thick- est in the upper end, is often nearly im- perceptible in the lower portion of the burrow, and tends to be thicker in less stable or dryer soil. There is considerable variation in burrow length among adult spiders (Table 6). These burrows tend to be longer in dryer soil, but many other factors also affect burrow length. Penulti- mate male burrows average smaller than those of adult females. Adult female bur- row pitch ranged from 0 to 35° \\'ith most burrows between 0° and 15°. No burrow sloped upward from the entrance. Entrance structure. This species pos- sesses at its burrow entrance a flexible silken collar identical to that of A. unicolor (Figs. 18, 19, and 44). The only burrows observed with the collar closed and sealed on the inner surface with a small amount of silk (as is fairly common in A. unicolor) were those of three penultimate males ap- parently ready to undergo the final molt. Two burrows were found with soil plugs. In one burrow (2994 at Catherine Creek State Park, Ore.), the plug was in the entrance imder the closed collar, but in the other (2970 at V), the plug was positioned one-third of the way down the burrow. Egg sac structure and placement. All Systematics and Biology of Antrodiaetus • Coyle 287 four egg sacs examined were like those of A. unicolor (Fig. 16). Of the three egg sacs which I excavated, one ( 2825 at S ) was near the top of the burrow, and the other two (at /) were in the bottom half (Table 6). All were attached to the ventral sector of the burrow wall. Molting behavior. Fragmented exuviae were often found compacted like prey exo- skeletons into the bottom end of burrows. Only two recently molted spiders were observed; two mature males each had the final exuvia packed lightly into the bottom end of its burrow. The exuvia of one male (2628 at V) was slightly broken, and the nature of the entrance closure could not be detennined. The exuvia of the other (L-1 at /) was unbroken, and the collar was closed and lightly sealed inside with silk. Defensive behavior. The behavior of foraging individuals of A. pacificus when disturbed by strong light and substrate vibrations is like that of A. unicolor. Adult females, when prodded, often assume an aggressive defensive posture like that of A. unicolor. Five large immature spiders at R were observed at the bottom end of their burrows spinning a thin septum of silk across the burrow lumen between themselves and the entrance as I excavated their burrows. When I broke the septa and prodded them, they exhibited the typical aggressive defensive posture. Predators and parasites. Three recently digested males were found hanging in the webs of theridiid spiders at G. One male from Washington was taken "from the stomach of Biifo boreiis," a western toad. Three large immature specimens ( 2249 and 2640 at R, and 2114 at M) each had large nematode worms filling their abdomens. A pompilid wasp larva was found on the abdominal dorsum of a female A. pacificus on 26 July near Friday Harbor, Washing- ton. Prey capture behavior. Adult females were observed in foraging position within their open burrow entrances only after nightfall, or, rarely, in very dim light; the burrow entrances are normally closed dur- ing the daytime. Immatures usually as- sumed foraging postures earlier in the evening. Very young spiders frequently had collars open during the daytime and were often in foraging postures then. The foraging posture of this species is similar to that of A. unicolor. Prey composition. Recently captured prey and identifiable prey rejectamenta found in the bottom of burrows include ants, beetles, spiders, hymenoptera (one vespid wasp and one Argidae), and He- miptera. Ant and beetle remains were far more abundant than those of other groups. A 15-mm long carabid beetle (Broscinae) and Canjponotus major worker ants were the largest prey found. Disposal of prey remains. Fragmented exoskeletons of digested prey were fre- quently found mixed with a small amount of silk and packed into the bottom end of burrows as in A. unicolor. Tlie thicker the cuticle, the less fragmented the remains. Live collembola were observed on such trash in a few burrows. Matinii, behavior. Males of A. pacificus apparently wander in search of mates only at night. I have observed 12 wandering males, seven at L and five at K, all after nightfall. Considerable daytime collecting at both localities failed to turn up wander- ing males. Four male specimens from W are accompanied by the label "on ground at night." Wandering males usually stop moving when light is shown on them. At L, two captured males were released near female burro\\\s, and each stopped suddenly when it came close to a burrow. In one encounter, the female closed her collar just as it was discovered by the male, which then used his pedipalps and first legs in an apparent attempt to open the collar. After about 30 seconds it wandered away. In the second encounter, a female lunged out of her entrance toward the other male, made contact with its front legs, and both remained motionless for a 288 Bulletin Museum of Comparative Zoology, Vol. 141, No. 6 few seconds. Then the female retreated quickly into her burrow and closed the collar. The male moved to the entrance, briefly handled the collar with its pedi- palps and first legs, and then wandered away. Because of the handling of the males and the dim light used for observation, both encounters were probably abnormal. At /, the discovery of the first leg of a male A. pocifictis on the soil surface at the burrow entrance of adult female 2232 sug- gested a male-female struggle. No male remains were found within this burrow. Anfrodiaefus occulfus new species Ecology Ceo^rapliic distribution, elevation raniie, and habitat. This species appears to be limited to the relatively dry and warm valleys of the Willamette and Umpqua Rivers of western Oregon (Map 2). 0-600 ft (0-180 m). Life History Males. Adult males of this species ap- parently wander and mate during late Sep- tember and October (Fig. 1). The data in Figure 3 indicate that there may be a partial temporal separation of the mating season of A. occidtus from that of sym- patric populations of A. pugnax and A. pacificiis. Anfrodiaefus pugnax (Chamberlin) Ecology Geograp]}ic distribution and elevation range. Northern Oregon, southern Wash- ington, and northwesteiTi Idaho (Map 2). 0-3500 ft (0-1 100 m). Habitat. I have observed A. pugnax at only two localities. At B on 8 July, a large population was located in a deciduous woods with a dense understory of shrubs and herbs along a small, flowing stream. Sunounding the woods was nonforested land, chiefly farmland. Although burrows were scattered throughout the woods on both horizontal and inclined ground, dense aggregations were foimd only beneath the densest vegetation. The soil was hard- packed brown loam \\\\h a high organic content, and was dry at the surface but slightly humid below 10-cm depth. The densest aggregation was found under the shelter of a large shrub on horizontal ground covered with scattered leaf and twig litter. Ten adult females, five penulti- mate males, and about 150 immatures were present in a 1-m- area. The habitat at Lewis and Clark Trail State Park, Wash- ington (Fig. 64) — a dense deciduous riparian woods located at the bend of a river and surrounded by wheat fields — was very similar to the above habitat. The understory vegetation was dense, and the burrows were found in the more sheltered spots on both horizontal and sloping ground. The soil was firmly packed light brown silty loam with a slightly greater moisture content than at B. Most of the other localitv records for A. pugnax are in dry regions at low elevations near rivers or streams and therefore pre- sumably in riparian woods similar to those described above. Tlie only wooded habi- tats in the immediate vicinity of C are such riparian woods. H has similar habitats. However, one collection of A. pugnax at E was made in Douglas fir forest, a more humid habitat. Thus it appears that A. pugnax is chiefly found in the low ele- vation riparian deciduous woodlands of the dry interior portions of Washington and Oregon, and in the low, relativeh' dry forest of the vallevs between the Coast Range mountains and the Cascade Range mountains. Life History Males. Collection data (Fig. 1) indicate that male wandering and mating occur within the period from late July to Oc- tober. Females. The limited data (Fig. 2) indicate that the timing of brood develop- Systematics and Biology of Antrodiaetus Coyle 289 ment in A. piignax may be similar to that in A. iinicolor. All individuals from which the brood size (Table 3) and egg size (Table 4) data were obtained are from B and Lewis and Clark Trail State Park, Washington. The second instar moiphol- ogy is like that of A. tinicolor. Behavior Burrow structure. (All observations on bin'row and entrance structure and egg sacs were made at B and Lewis and Clark Trail St. Pk.) The burrow structure is similar to that of A. unicolor. The burrow is roughly tubular and slightly widened just belo\\' the entrance and at the bottom end (Fig. 20). The entire burrow is silk- lined, with the thickest lining in the upper portion. The lining is thicker in burrows with egg sacs, at least in the vicinity of the sac. Burrows vary considerably in size but are usually relatively small (Table 6). The longer burrows were in softer soil. All burro\\'s but one were in horizontal ground; all these were vertical except for a slight turn at the bottom end of some. The burrow pitch of the burrow on sloping ground was 10°. The five penultimate male burrows observed averaged consider- ably smaller than the female burrows, were completely and rather heavily lined with silk, and had 0-5° burrow pitches. No bur- rows sloped upward from the entrance. Entrance structure. Small immature spi- ders have flexible collar entrances similar to those of A. unicolor. The silk lining of the burrow is extended above the soil sur- face, stands erect when open, and collapses inward to close off and camouflage the entrance. Soil particles and ground litter fragments are incorporated into this collar. The burrow entrances of older spiders were sealed and iDlugged with soil (Fig. 20), making it difficult to determine the exact nature of the unsealed entrance. The closure varied from a thin septum of soil and silk only a few millimeters thick to a hard plug up to 15 mm thick, mainly of soil. Such a sealed entrance is very diffi- cult to discover visually. The inside sur- face of the closure is covered with a lining of silk continuous with the burrow lining. Egi^ sac structure and position. The egg sac structure of this species is very similar to that of A. tinicolor (Fig. 20). The egg masses varied from 7 to 9 mm maximum diameter and 4 to 6 mm in thickness. Six of the eight egg sacs were positioned on the ventral sector of the burrow wall, the other two on lateral sectors. The egg sac silk appears to have the same texture and \\'hiteness as burrow lining silk. All egg sacs but one were positioned just above the bottom end of the burrow (Table 6). Behavior of brooding females. Like the large immature and nonbrooding adult spiders, all females with egg sacs had plugged and sealed entrances. Predators. A female specimen collected at G on 15 April is accompanied by a label saying that she was the "prey of Dinocne- mis," a pompilid wasp. Feeding behavior. Many young im- mature spiders were in foraging postures within their open burrow entrances during daylight, although more exhibited such be- havior at night. Fragmented prey exo- skeletons held together \\'ith small amounts of silk were found packed into the bottom ends of several burrows. Such rejecta- menta from the burrow of a brooding fe- male at B contained parts of t\vo beetles. Mating behavior. A label accompanying a male from B reports that the male was wandering o\er the ground surface at night. Antrodiaetus monfanus (Chamberlin) Ecology Geographic distribution and elevation range. Primarily the Great Basin region from Utah and Nevada north to Washing- ton (Map 2). 2500-7200 ft (750-2200 m). Habitat. I have observed A. montanus in two localities. Near Pocatello, Idaho, 290 Bulletin Museum of Comparative Zoology, Vol. 141, No. 6 on 26 July, a small aggregation of burrows ^^•as located in the bank of a dry ravine in juniper-sagebrnsh habitat at 4500 ft (Fig. 65). Burrows of three adult females and several immatures were foimd in an un- eroded portion of the bank imder a large juniper. The soil surface was covered with a thin layer of pieces of dead grass and juniper. The soil was extremely dry hard- packed loam down to 25 cm and dry looser gravel and loam below that. The bottom ends of the large burrows were about level with the ravine bed. At Crater Lake, Ore- gon, on 30 July, I observed a sparse popu- lation of A. montanii.s in open and dry lodgepole pine, fir, and spruce forest at 6000 ft. The two burrows observed were in rather tmsheltered spots, one on hon- zontal ground. The soil was composed of \olcanic "sand," very dry and loose at the surface, becoming firm at 5 to 10-cm depth and slightly humid below 15 cm. Scattered burrows of A. pacificiis and Atypoides gertschi were also found at this locality. The locality records in the Salt Lake City, Utah, area indicate that A. montamis is there found in dry, non- or sparsely wooded habitats along the lower portions of canyons and in the foothills of the mountains. All other locality records for this species are likewise from relatively xeric habitats similar to those described above. Life History Males. Collection data (Fig. 1) indicate that male wandering and mating occur within the period from early August through early November. Chamberlin and Ivie ( 19.35, 1945 ) observed males wander- ing "in the autumn." Females. Very limited data for A. mon- tamis (Fig. 2) indicate that the timing of brood development may be similar to that of A. unicolor. The brood size and egg diameter of the single known brood of A. ))i07itanus (2160 at Pocatello) are given in Tables 3 and 4. Behavior Burrow structure. (Based on observations of three adult female burrows from Poca- tello and one large immature female burrow from Crater Lake.) Burrow structure is ((uite similar to that of A. unicolor, except for a considerably greater average length (Fig. 21, Table 6). Two of the burrows were completely silk lined with the thickest lining in the upper part of the burrow. In the other two burrows, the lining below the upper 5 cm was barely perceptible. Bur- row pitch ranged from 5 to 25°, and all burrows sloped downward. Chamberlin and Ivie (1933), who found a female A. montanus without a distinct l:)urrow and under a cottonwood log, later ( 1935 ) er- roneously stated that females of A. mon- tanus "ordinarily do not have burrows." Their observation in a later paper (1945) was more accurate: "The female lives in burrows which extend about a foot or more straight down into the soil. The upper part of the burrow is lined with silk; the bottom is slightly enlarged, but not bent." Entrance structure. (Based on observa- tions of two adult female burrow entrances at Pocatello and one large immature female entrance at Crater Lake.) A. montanus constructs a flexible silken collar like that of A. unicolor. E^ig sac structure and position. The structure and means of attachment to the burrow wall of the single egg sac observed (2160 at Pocatello) were like those of A. unicolor egg sacs. Egg mass dimensions were 14 X 10 X 7 mm. The sac was positioned on the ventral sector of the bur- row wall deep in the burrow (Fig. 21, Table 6). Chamberlin and Ivie (1945) also found egg sacs only in the lower por- tion of the burrow. Behavior of brooding females. When collected in the evening before dark, the single brooding female (2160 at Pocatello) had a closed but unsealed collar. Parasites. The abdomen of a large im- mature A. montanus (0174) from Steens Systematics and Biology of Antrodiaetus Coylc 291 Mountain, Oregon, was filled with a nema- tode worm. Feeding behavior. The two large im- mature A. montamis females from Crater Lake were collected after dark in foraging positions just inside their open collars. Fragmented prey exoskeleton was found mixed with silk and packed into the bot- tom ends of the burrows of all thiee adult females from Pocatello. Two of these trash packets contained the remains of 12 medium to large-sized beetles. Mating behavior. Apparently A. mon- tamis males wander in search of mates at night and remain under the shelter of various objects during the day. One male from Verdi, Nevada, was collected \\'alking over the ground at night. Chamberlin and Ivie ( 1945 ) often collected adult males "under stones, sticks, [and] bunches of grass" presumably during daylight hours. A label with the holotype male notes that it was resting under a piece of sagebrush. Antrodiaetus hageni (Chamberlin) Ecology Geographic distribution and elevation range. South-central British Columbia, eastern Washington, and eastern Oregon (Map 2). 1000-4000 ft (300-1200 m). Habitat. I have observed A. hageni at only one locality — at Trail, British Co- lumbia, on 13 July on the slopes along a dry stream bed in a large nonforested ra- vine one mile from its junction with the Columbia River (Fig. 66). The vegetation consisted of scattered young poplar trees and large deciduous shrubs. Much of the soil surface was without plant or litter cover. The soil was chiefly deep sand with a few rocks and very little organic material. The upper 3-8 cm of the sand was hot, dry, and loose; below that it was humid, well packed, and much cooler. Burrows were found in both steeply inclined and hori- zontal ground, but were usually clustered in more sheltered spots where the soil sur- face was stable, such as under trees and shrubs. A. hageni was not found in the cooler, humid forest habitats near Trail. Near Baker, Oregon, where A. hageni has been collected, I unfortunately searched only in humid forest habitats and there found only A. pacificus, evidence that A. hageni is here found in the drier habitats (with sparser vegetation) at and below 4000 feet. An adult female A. hageni from near Oliver, British Columbia, was col- lected on a dry "sandy-rocky slope" covered with bunch grass and sagebrush (E. Thorn, personal communication). All other British Columbia records of A. hageni are in low river valleys where dry habitats similar to this and to that along the Columbia River at Trail are common; it is probably safe to assume that the habitat at Trail is an approximately normal A. hageni habitat. Life History Males. Adult male wandering and mat- ing apparently occur within the period from late July through October (Fig. 1). Females. Two records (Fig. 2) weakly indicate that the timing of brood develop- ment is similar to that of A. iinicolor. Behavior Burrow structure. (Based on observa- tions of two adult female and two im- mature female bm'rows.) The burrow architecture (Fig. 22, Table 6) is similar to that of A. unicolor. The roughly tubular burrow is somewhat enlarged just below the entrance and at the bottom end. All burrows were entirely silk lined, with the lining of the collar and upper part of the burrow much thicker than the rest of the lining. The two penultimate male bur- rows observed were similar to the female burrows in structure. Burrow pitch ranged from 0 to 35°, and all sloped downward. Entrance structure. (Based on observa- tions of one adult female and two large immature female burrow entrances.) A. hageni constructs a flexible silken collar very similar to that of A. unicolor (Figs. 45-46). 292 Bulletin Museum of Comparative Zoology, Vol. 141, No. 6 Molting behavior. An immature spider which molted within a glass tube in cap- tivity fragmented its shed exuvia several days after ecdysis. Predators. A large immature female of A. hagcni was collected together with a pompilid wasp, Friocnemis oregona Banks (identification, H. E. Evans), on 7 April at Baker, Oregon. Unfortunately, no other data was included with the specimens. It is probable, however, that A. hageni is a host of this wasp, since the three other host species records for this wasp are mygalo- morph spiders, one being Atypoides riversi (Wasbauer and Powell, 1962). Feeding behavior. All burrow entrances were closed during the afternoon of ob- servation. Prey remains were found packed into the bottom end of the burrow of one large immature female. The prey exoskele- tons were fragmented (the thinner the cuticle the smaller the fragments ) and held together with a small amount of silk. Identifiable prey consisted of 12 ants, two beetles, one dipteran, and two adult male spiders {Tegeneria domestica and Xysticus sp.). Antrodiaetus cerberus new species Ecology Geographic distribution, elevation, and habitat. Known only from the type locality in coniferous forest in northeastern Wash- ington. Approximately 2000-2500 ft (600- 750 m). Life History Males. All four examined males of this species were collected in pitfall traps some- time during May, 1962 (Fig. 1). Three more males were collected sometime dur- ing April and May, 1962 ( W. Ivie, personal communication). It is thus likely that A. cerberus males wander and mate only dur- ing the spring. Females. Two records (Fig. 2) weakly indicate that brood development timing may be similar to that of A. unicolor. Antrodiaetus yesoensis (Uyemura) Ecology Geographic distribution and habitat. Known only from the northern Japanese island of Hokkaido, where it is presumably found in the conifer forest of that cool temperate climate (Map 3). Life History Males. The few records (Fig. 1) indi- cate that the period of male wandering and mating occurs during the summer months. Antrodiaetus lincolnianus (Worley) Ecology Geographic distriJ)ution and elevation range. Eastern Kansas and eastern Ne- braska (Map 1). 700-1200 ft (210-360 m). Habitat. In 1928, Worley stated that the holotyjoe male and the three paratype males were collected "on clay banks near deciduous forest" at Lincoln, Nebraska. Later (Worley and Pickwell, 1931), he stated that these males were found "under logs in woods" apparently in their own burrows. H. S. Fitch (personal communi- cation) collected a wandering male near Lawrence, Kansas, about 35 m from a large patch of deciduous forest through which ran an intennittent stream with high banks. This limited evidence, in addition to the fact that no specimens of A. lincoln- ianus have been collected in localities removed from the western lobes of the eastern deciduous forest, indicates that this species is probably found in somewhat humid forest habitats. Life History Males. Male wandering and mating in A. lincolnianus apparently occur during late winter and early spring. The seven adult males indicated in Figure 1 were collected by four different collectors on seven different dates during four different years. Also, tsvo other males, which I have Systematics and Biology of Antrodiaetus • Coijlc 293 been unable to examine but which are almost certainly A. lincolniamis, were col- lected by yet a different collector, Scheffer (1906), in Manhattan, Kansas, on 5 April during yet another year. Such a scattering of individual collecting events makes it more likely that these dates indicate the actual wandering and mating period of this species. Females. The single gravid female record (Fig. 2) for A. lincolniantis hints that the timing of brood development may be similar to that of A. iinicolor. Behavior Burrow structure. Worley and Pickwell (1931) state that "this species is very similar in habits to B. pocificum Simon [=A. pacificus] of the Pacific Coast, con- structing burrows in the ground six to ten inches deep and lining them with silk." Anfrodiaefus stygius new species Ecology Geographic distribution and elevation range. Missouri and northern Arkansas (Map 1). 400-1500 ft (120-460 m). Habitat. I have observed A. stygius only at Bennett Springs State Park, Missouri, on 4 September. An aggregation of ap- proximately 100 adult and immature bur- rows in a 4-m- area was found on a 20-35° slope 3-5 m from the edge of a large spring-fed stream in the heavy shade of a mixed deciduous forest (sycamore,^ elm, and oak trees common). Tlie ground sur- face had a sparse to dense cover of leaf litter. The hard-packed dark brown loam soil was slightly humid at the surface and increasingly humid downward. A wander- ing adult male was collected near a stream near Warrensburg, Missouri, in a similar deciduous forest of oak, sycamore, and elm (Peck, 1966). Life History Males. The limited data (Fig. 1) indi- cate that adult males wander and mate during the fall. Females. All three broods observed had developed beyond the hatching stage, but every shed chorion contained the embry- onic cuticle, an indication that the pattern of early postembryonic development is similar to that in A. tinicolor. Also the morphology of the embryonic cuticle, the first instar, and the second instar is very similar to that of A. iinicolor. The limited data (Fig. 2) indicate that the timing of A. stygiiLS brood development in nature resembles that of A. iinicolor. The size of the only complete brood is given in Table 3. Behavior Burrow structure. The three brooding female burro\\'s obser\'cd were completely silk-lined and shaped much like A. iinicolor burrows: roughly tubular and slightly widened just below the entrance and at the bottom end (Fig. 23, Table 6). Three penultimate male burrows were shaped similarly but had thinner silk linings and were shorter (Table 6). All six of these burrows were slightly sinuous and nearly xertical, with bun-ow pitches of 15-30°. Entrance structure. All burrows dis- covered— even those of immatures — had their entrances sealed over and usually plugged with soil (Fig. 23), making it dif- ficult to determine the exact nature of an unsealed entrance, which appeared to con- sist of a flexible collar. Tlie upper 4-10 mm of the larger burrows was packed with soil. The bottom of this plug was lined with silk continuous \\ith the buiTO\\' lin- ing, and the upper exposed surface closely matched the surrounding soil surface in texture and color. Egg sac structure and position. The shape and structure of the two A. stygius egg sacs observed appeared similar to those of A. Iinicolor. Both were positioned on the ventral sector of the wall in the bottom half of the burrow (Table 6). The burrow diameter at the region of egg sac attachment may have been greater than 294 Bulletin Museum of Comparative Zoology, Vol. 141, No. 6 normal, but the sacs still produced a slight localized construction of the burrow lumen. The silk lining was thicker in the vicinity of the egg sac than elsewhere below the upper end of the burrow. Egg sac silk ap- pears to have the same texture and white color as burrow lining silk. Behavior of brooding, females and broods-. The burrow entrances of brooding females, like those of nearly all other in- dividuals, were sealed with soil plugs. The brood composed of first instar spiderlings was still enclosed within the egg sac, while both second instar broods were entirely outside the sac. In one of these latter two broods, the egg sac was no longer present; in the other it contained all the first instar exuviae, indicating that the spiderlings emerged from the egg sac after molting into the second instar. The second instar spiderlings were quite active. Defensive behavior. Some adult females assumed a defensive posture in the bottom of their burrows similar to that of A. unicolor. Predators. A pompilid wasp cocoon con- taining a larva was found in one medium- sized A. sttjsiiiis burrow. Small pieces of spider exoskeleton were attached to the outside of the cocoon. Disposal of prey remains. Fragments of prey cuticle were found packed into the bottom end of one adult female burrow. Antrodiaetus opachecus new species Ecology Geographic distribution and elevation range. Mountains of Arizona and New Mexico (Map 4). 6100-8500 ft (1850- 2600 m). Habitat. I have collected A. apachectis at only one locality, near La Cueva, New Mexico, at 7800 ft in a somewhat open forest of yellow pine, spruce, and fir. An aggregation of six burrows was found under the overhang of a large rock on the slope of a wide ravine. Aspen and scrub oak were common in the ravine. The ground surface under the rock lacked leaf litter, and the soil was humid dark sandy loam with some pebbles. All other locality records for this species are either from transition zone forest or low Canadian zone forest. Because it is restricted to such high elevation habitats, the species population is presently fragmented into several geo- graphically isolated populations each on a forested montane "island" surrounded by unfavorable nonforested habitats at lower elevations. Life History Males. Collection data (Fig. 1) indicate that male wandering and mating occur from midsummer to early fall. BEHA^^OR Burrow structure. The burrows of only four medium- to large-sized immature spi- ders were observed. The burrow shape appeared similar to that of A. stygius, i.e., tubular but slightly widened just below the entrance and at the bottom end. The silk lining was very thin and possibly ab- sent over most of the burrow, but slightly thickened at the upper end. The burrows were slightly sinuous and nearly vertical, with a burro^^' pitch of 10-20°. Entrance structure. The burrow entrance of only one immatiue spider was observed and appeared to consist of a short thin flexible collar. Feeding behavior. A medium-sized im- mature A. apachccus was found at the bot- tom of its burrow feeding upon a dead, partly digested, 20-mm long lepidoptera larva covered with urticating setae. Tliis was at 11:00 a.ai. and the collar was closed. Anfrodiaefus rorefzi (L. Koch) Ecology Geogra))Jiic distribution. Central portion of the Japanese island of Honshu (Map 3). Habitat. According to Yaginuma (1962), A. roretzi is usually found in humid, often heavily shaded habitats. Yamamoto (1942) Systematics and Biology of Antrodiaetus Coylc 295 found a sizeable burrow aggregation scat- tered over a very steep bank composed of reddish soil. Bamboo was growing over the upper part of the bank. Komatsu ( 1942 ) reported that this species is found in shaded humid habitats, often on inclines or banks. Uyemura ( 1936 ) found a bur- row on a steep slope in moss-covered soil at the base of a pine tree. Ohe ( 1966 ) found burro\\'s chiefly on inclines (40- 90° ) of banks along a trail. Such burrows were frecjuently in the shelter of exposed tree roots near the top edge of these banks or were under the overhang of exposed rocks. All but a few burrows were located on the upper two- thirds of such banks. He foimd burrows both in heavily shaded moss-covered portions of banks and in less well-shaded spots where the surface soil was rather diy. Burrows are also often found in caves (Yaginuma, 1962; Komatsu, 1961) near the entrances. Komatsu (1942) found an aggregation of over 100 burrows on an inclined surface in humid reddish soil in dim light just inside the entrance of one cave. Life History Males. The small amount of data in Figure 1 indicates that male wandering and mating may occur during winter and early spring. More data are needed. Females. Ohe ( 1966 ) obsei-ved spider- lings in parent burrows "from September on. Behavior Burroio structure. Yamamoto (1942) ob- served the burrows of two adult males and several adult females (Table 6). Tliese were tubular and silk lined. Burrows in soft soil free of obstacles were longer than those in harder soil with roots, and male l)urrows averaged shorter than adult fe- male burrows. The burrow pitch of the two male burrows was apparently about 0-10°. Komatsu (1942) observed several N'crtical burrows in inclined ground, but most were roughly perpendicular to the soil surface. Ohe ( 1966 ) described and illustrated the variation observed in the in- clination, shape, and length of a large number of A. roretzi burrows (probably both adult and immature). Most burrows apparently were not strongly curved and were roughly perpendicular to the plane of the adjacent ground surface. Exceptions were the result of root or rock obstacles. Most burrows were approximately 20 cm long or shorter, but a few reached 30 cm. Entrance structure. A pair of photos of an A. roretzi burrow entrance has been published twice (Komatsu, 1937 and 1961). Drawings of the entrance can be found in Komatsu (1961), Yamamoto (1942), Yagi- numa (1960), and Ohe (1966). Descrip- tions, all very brief, are found in Yamamoto (1942), Komatsu (1942), Yaginuma (1962), and Ohe (1966). All these describe the structure as a "double door entrance." After studying the fine photos and draw- ing of Komatsu, it appears to me as though the double door nature of the entrance has been overemphasized and that the entrance structure is more accurately described as a collapsible collar which is reduced at two opposing portions, leaving two lateral flap- like sectors that operate roughly as sepa- rate doors. Such a structure is quite similar to that of the collars of some adult A. unicolor burrows. Clearly, additional care- ful field observations are required. Ko- matsu (1942) and Ohe (1966) observed that, in entrances on an inclined ground surface, the collar is reduced at its ventral and dorsal sectors with both lateral flaps "hinging" on opposing dorsovcntral axes. Such an orientation is common in A. uni- color. Tlie collar in Komatsu's photos is constructed of silk and soil, well camou- flasied exteriorlv, and silk lined on the inner surface, this silk lining apparently being continuous with the burrow lining. Ohe found that collars in moss-covered soil were constructed partly of moss and were difficult to discover. Although Ohe (1966) observed a few burrow entrance openings 296 Bulletin Museum of Comparative Zoology, Vol. 141, No. 6 as large as 1.5-cni diameter, most were roughly 1 cm in diameter. Molting behavior. Yamamoto (1942) collected two adult males in their own burrows. In one, the shed cxuvia lay on the ventral sector of the burro\\' wall about one-third of the way down the burrow, and a maze of silk threads occupied much of the lower half of the burrow, with the male in the bottom end. The second male had spun a thin silk septum across the inside of the entrance, and no exuvia or silk thread mazes could be found in the bur- row. Yamamoto did not say whether either collar was closed or open. Possibly the second spider had sealed the inside of the collar shut. Clearly, additional observations are needed to clarify behavior during the final male molt of A. roretzi. Afypoides riversi O. P.-Cambridge Ecology Geographic distribution and elevation range. A coastal population in northern California west of the Central Valley as far south as the Monterey Peninsula, and a Sierran population in the Sierra Nevada Mountains and their foothills (Coylc, 196S). These populations are probably geographically isolated. 200-1800 ft (60- 550 m) (coastal population). 2000-8000 ft (600-2400 m) (Sierran population). Habitat. Coastal population: Rivers (1891; O. P.-Cambridge, 1883) found A. riversi burrows in both pine and deciduous forests, often in stream banks. Smith ( 1908 ) found burrows "abundant along shaded streams and in thickets in the foot- hills and mountains" of the Coast Range. Gertsch's (1949) habitat description was similar. I have collected coastal A. riversi in three localities, and will summarize my observations below. A. riversi is found in the foothills and mountains of the Coast Range, chiefly in shaded forest habitats. These may be de- ciduous, pine, or mixed deciduous-pine forests. Interestingly, A. riversi is uncom- mon or absent from the famous "basin forest" of predominately coast redwood and Douglas fir; I searched hard in such habitats at three localities unsuccessfully. A. riversi burrows are common in the next highest vegetation zone, the tan oak-pacific madrone forest containing only scattered coast redwood trees (Fig. 67). Burrows are found chiefly on rather steep inclines of banks along trails, roads, and streams. The surface litter in these spots ranges from absent to quite thick. The soil varies from compact sandy loam to clay loam, sometimes with a high pebble content. These soils are usually dry at the surface during the summer, but below about 10- cm depth are at least slightly humid. The densest burrow aggregations are found on stable, noneroding banks. Three adult fe- males and 29 immatures were found in a 0.30-m- area at C. Sierran population. I have observed A. riversi in three different locations above 6500 ft in the Sierra Nevada Mountains. The habitat ranged from open pine-fir forest to dense pine-fir or California red fir forest. The densest and largest popu- lation was found in the red fir forest at G. Burrows are found on both steeply sloping stream and road banks and on gently in- clined ground. The surface litter varied from absent to very thick. The soil was brown sandy loam, dry and loose at the surface but becoming humid and packed at from 3 to 15 cm deep. The lower the rock content of the soil, the greater is the burrow density. Life History Males. Coastal popidations: Male wan- dering and mating in these populations may not normally begin until September, October, or November (Fig. 1), and may coincide with the onset of the fall and winter wet season. At D and F on 14 and 15 August, I searched both day and night in areas of dense burrow concentrations for adult males but found none. Sierran popu- lations: In at least the higher populations Systematics and Biology of Antrodiaetus • Coyle 297 (6500-8000 ft), male wandering and mat- ing commence earlier and probably (be- cause of harsh winter weather) terminate earlier (Fig. 1). There is some evidence to indicate that the palpal tarsi of immature males of A. riversi may be swollen during more than just the penultimate instar and that many of the immature males with s\\'ollen palpal tarsi found in late summer or fall may not mature until one year later. The shed exuvia of an immature male collected at H on 8 August had swollen palpal tarsi like those of the newly formed instar. The exuvia of a recently molted immature male collected at the same locality and time had unswollen palpal tarsi, whereas those of the new instar were swollen. Thirteen actively foraging immature males with swollen palpal tarsi were collected at G, H, and 7 from 6 to 9 August when adult males were wandering. At D and F on 14- 15 August, 14 immature males with swollen palpal tarsi were trapped in foraging pos- tures in their burrow entrances at night, and seven were dug out of closed and usually (five) sealed burrows. All 21 were kept cool and humid in an ice chest, but only the latter seven molted to maturity. Females. Coastal populations: The pat- tern of early postembryonic development of A. riversi appears to be similar to that of Antrodiaetus unicolor. All of the many shed chorions observed in each of the six egg sacs with first instar spiderlings have the embryonic cuticle attached, indicating that both the chorion and embryonic cuticle split and peel off simultaneously or in close succession and are then shed to- gether. The morphology of the embryonic cuticle, the first instar, and of the second instar of A. riversi, is also quite similar to that of A. unicolor, with the obvious ex- ception that the anterior lateral spinnerets are not lost in the second instar of A. riversi. The data on the timing of brood development in A. riversi ( Fig. 2 ) indicate that eggs are laid in summer, that brood development reaches the second instar by mid-autumn, and that the spiderlings abandon the parental burrow before the following summer. Rivers (1891) observes without further comment that "there is a period of aestivation, but the cause is not yet investigated." Data on brood size and egg size are given in Tables 3 and 4. Sier- ran populations: Only a single Sierran brood was collected (at 7). Its size and egg size are given in Tables 3 and 4. Behavior Burrow structure. A. riversi has a roughly tubular silk-lined burrow which normally increases slightly in diameter from the top to near the bottom end where there is usually a decrease in diameter so that the last 2-4 cm are narrow (Figs. 24 and 25). In the Sierran populations, this terminal constriction was less noticeable at H and 7 but was well developed at G. The white silk lining is usually quite thick in the turret and upper part of the burrow, becoming thinner in the bottom portion, and may be nearly imperceptible for the last few centimeters. Burrows ^^'ith egg sacs usually had thicker silk linings than those without. Burrow dimensions, which show considerable variation, are given in Table 6. As might be expected, larger individuals within a population have larger burrows, and this correlation is shown in the lower mean burrow dimensions of the Sierran populations, which have, on the average, smaller mature individuals than the coastal populations. Also, penultimate male burrows average smaller than adult female burrows. Burrows ranged from vertical to near horizontal, but the burrow pitch ranged within 0-15°. Rivers (1891) also observed that the burrows "were more or less perpendicular" to the ground sur- face. The descriptions of coastal A. riversi bur- row structure by Rivers (1891), Smith (1908), and Gertsch (1949) are in agree- ment with my observations, except for the statement by Rivers that an adult female burrow "is more roomv at the base than 298 Bulletin Museum of Comparative Zoology, Vol. 141, No. 6 at the opening." He apparently overlooked the terminal constricted portion. Smith, like myself, observed that "the tunnels are commonly considerably reduced in diam- eter for the last tvvo centimeters. . . ." Entrance structure. The silk lining of the burrow is extended above the soil sur- face to form a rather rigid elongate collar or turret (Figs. 26, 47-50), hence the com- mon names "Californian turret builder" ( Rivers, 1891 ) and "the turret spider" (Gertsch, 1949). This turret is smooth white silk on its inner surface and is camouflaged and kept erect by attached soil particles and usually also organic ma- terials (from the ground surface litter) such as pieces of twigs, leaves, needles, bark, and moss. Normally the turret wall is supported at its base by the thickest accumulation of soil and other materials and becomes thinner distally. The opening at the top of the turret is roughly circular, and the rim is slightly expanded to fonn a lip. The entrance is closed off by collaps- ing or folding in only the distal end of the turret. Sometimes, as when a spider is molting, this closure is sealed on the in- side with a small amount of silk. There is considerable intrapopulation variation in turret height, as well as a marked tendency for individuals of the Sierran populations to have shorter turrets than those of the coastal populations (Table 6). A possible reason for this dif- ference (other than geographic variation in genetically determined turret building behavior) is that snow accumulation may annually destroy the Sierran tiu-rets, whereas coastal turrets are probably accu- mulations of several years' additions. A few turrets in the Sierran populations in- corporated needles, twigs, or leaves con- sistently attached by their ends and radiat- ing out roughly horizontally from the turret (Fig. 50). This radial arrangement closely resembles the "twig-lining" arrangement found in some of the Australian aganippine trapdoor spiders (Main, 1957a), but at present there is no evidence that in A. riuersi it has a similar functional signifi- cance in prey detection. The descriptions of coastal A. riversi turret structure bv O. P.-Cambridge (1883), Rivers (1891), Smith (1908), and Gertsch (1949) are in agreement with my obser- vations. Both Rivers and Smith describe considerable variation in the litter material used in turret construction. Smith includes two photographs illustrating some of this variation. Rivers oliserved turrets as tall as 3 inches (=7.6 cm). Egg sac structure and position. The egg sac structure ( Fig. 24 ) and egg mass shape of A. riversi is very similar to that of Antro- diaetus unicolor. Mean dimensions for the three egg masses are 10 X 9 X 6 mm. Egg sac silk appears to have the same texture and whiteness as burrow lining silk. In each of the nine burrows with egg sacs, the sac was positioned about half way down the burrow (Table 6). In five bur- rows the egg sac was on the ventral sector of the burrow wall, in three on the dorsal sector, and in one on a lateral sector. Some burrows may have been widened some- what in the region of egg sac placement, but each sac produced a localized con- striction of the burrow lumen. Behavior of brooding females and ]) roods. All of the first instar broods were completely enclosed within egg sacs, but the thickness of the silk of three of these egg sacs had been much reduced and the spiderlings could be seen moving around very slowly within. It would be interesting to know whether these spiderlings (or the parent) possibly possess silk-digesting en- zymes. All nine l)urrows with egg sacs had open turrets. Three of these were found at night, each with the female in foraging position in the top of the turret. It there- fore appears that brooding females remain active at least until the spiderlings emerge from the egg sac. Molting behavior. Molting behavior in A. riversi appears similar to that of A. unicolor. Four recently molted im- mature spiders were collected during the Systematics and Biology of Antrodiaetus Coyle 299 second week of August. In each case the top of the turret was closed and the edges held together by a thin layer of silk spun over the inside of the closure. The recently shed exu\'ia was attached to the \'entral sector of the burrow wall two-thirds to three-fourths of the way down the burrow, venter against the wall, anterior end head- ing up the burrow, and with legs and pedi- palps extended on the wall. Tlie unsclero- tized spiders were in the very loottom end of the burrows. Five penultimate males (from D and F) that later molted in cap- tivity and several other immature males with swollen palpal tarsi were collected at other localities in similarly closed and sealed burrows. A fully sclerotized adult male ready to emerge was collected at 7 in a sealed burrow with the last exuvia packed into the bottom end of the burrow. Defensive beJmvior. Wlien spiders in foraging positions were disturbed by light or substrate vibrations, they retreated rapidly down the burrow without closing the top of the tun'ct. During the later stages of excavating to expose a burrow, one finds the spider backed up tightly into the narrowed bottom end of the burrow in a stereotyped defensive posture. The cephalothorax is tilted backwards at ap- proximately a 45° angle with the abdomen and is therefore in contact with the sector of the burrow wall dorsal to the spider. Legs IV and III, and possibly II, hold onto the burro\\' wall to anchor the spider, and the other legs and pedipalps are spread apart laterally. The chelicerae are also elevated and spread apart laterally with fangs extended. If prodded, the spider strikes with a sudden downward movement of the cephalothorax and chelicerae. Penul- timate males also exhibit this same be- havior. Such a behavior pattern, like that of A. tinicolor, obviously confines the ap- proach of an attacker to that part of the spider protected by its chelicerae, but the narrowed burrow end of A. riversi prob- ably provides more effective protection for its abdomen than the enlarged burrow ending of A. tinicolor. Smith (1908) ob- served that the terminal constriction makes "a snug fit for the spider, and here it usu- ally snuggles down tightly when the digging [of the collector] has reached the limit necessary to secure the specimen." Predators. Wasbauer and Powell ( 1962 ) observed predation on A. riversi at Felton (Santa Cruz Co.), California, on 5 May, by the pompilid wasp Priocnemis oregona Banks. (See Antrodiaetus hageni predator records. ) Feeding behavior. The foraging posture of A. riversi appears similar to that of A. iinicolor; the spider is within and facing up the turret with the pedipalps and legs I extending fonvard, touching the silk lining on or below the turret lip. No adult spider was ever found in this foraging position until the arrival of full darkness. Unlike other species of Atypoides and Antro- diaetus, most individuals of A. riversi were found to leave their turrets open both day and night. A. riversi, like A. tinicolor, apparently detects prey primarily, if not solely, through the substrate vibrations trans- mitted via the silk lining of the tiuret and burrow. I was able to elicit prey capture responses by gently vibrating with forceps the ends of turret litter. The needles, twigs, leaves, and other somewhat linear litter that are often incorporated into the turret of A. riversi probably extend the prey-sensing radius of the spider even when, as is usually the case, this litter is not arranged radially. Ho\\'ever, it remains to be shown whether or not attaching linear litter to the turret is a result of selection for improved prey sensing. Fragmented prey cuticle is often found packed into the bottom end of the burrow and sometimes in other parts of the burrow wall as well. A cursory examination of trash indicates that ants may make up a large part of the prey of this species. Identifiable portions of several Campo- notiis workers were found. Smith (1908) briefly commented on the 300 Bulletin Museum of Comparative Zoology, Vol. 141, No. 6 feeding behavior of young immatures of A. riversi collected from coastal popu- lations and kept in the laboratory. These accepted both small ants and aphids, and always maintained a hold on the lip of the turret with the claws of legs IV during the lunge for prey. Courtship and matinii behavior. There are no published observations of courtship or mating in the family Antrodiaetidae. I was fortunate enough to witness a portion of mating behavior between a pair of A. riversi at G on 8 August 1967, at 9:45 p.m., about 45 minutes after nightfall. When first discovered, the male was positioned within and just below the rim of the fe- male's turret facing down into the burrow with the tarsi of legs IV on the turret rim (Fig. 27). The female was immediately below the male and facing up the burrow in a posture very similar to the normal de- fensive posture, except that her fangs were unextended. The male's cephalothorax was raised slightly and its cheliceral apophyses were positioned as a unit between the out- spread chelicerae of the female (Fig. 28). It appeared as though the male was apply- ing substantial pressure with his chelicerae, but the female did not struggle. I could not observe exactly how the pedipalps and legs I, II, and III of the male were iDOsitioned, but the pedipalps appeared to be extended beneath the female and legs III and prob- ably legs II were holding onto the burrow wall below legs IV. After two minutes in this position the female stiiiggled and was forced down the burrow by the male, which appeared to maintain his hold with his cheliceral apophyses. I am not certain that the pair was actually copulating when discovered. Adult males apparently wander in search of mates only at night. Even though I searched during both daylight and evening hours, the Sierran males \\'ere found wan- dering on the ground only after 9:30 p.m., at least half an hour after nightfall, and the largest numbers were found between 10:30 P.M. and midnight, when the search- ing ended. Afypoides gerfschi Coyle Ecology Geographic distribution and elevation range. Cascade Mountain Range and foot- hills from southern Oregon south and east into the northern end of the Sierra Nevada Mountains of California (Coyle, 1968). 2000-7500 ft (600-2300 m). Habitat. The densest and largest popu- lations of A. gertschi were found above 6000 ft in the open California red fir forest at N and P (Fig. 68). In this habitat, ground cover vegetation is nearly absent and the surface litter generally sparse. The soil is volcanic "sand" or sandy loam, dry and loose at the surface, but becoming humid and packed 5-15 cm below. Bur- rows were found both on level ground and inclined surfaces. At lower elevations burrows are often common in open forests with little ground vegetation or in non- forested areas of manzanita chaparral in volcanic "sand" soil, as long as the soil is humid below 5-15 cm. For instance, at O I found a dense population in manzanita chaparral only along the edge of the lake. The habitat at M is rather artificial be- cause of the great variety of imported vegetation, but it is wooded and the soil is sandy. Most of the burrows there were along the side of a stream (W. J. Gertsch, personal communication ) . Life History Males. The collecting data in Figure 1 indicate that male wandering and mating occur during the summer. Females. Only a single brood of A. gertscki (from N) has been collected (Fig. 2). The absence of broods with the many other females that I excavated is puzzling, but it is possible that brooding females of this species tend to keep their entrances closed noctumally and were missed be- cause of the difficulty of spotting closed Systematics and Biology of Antrodiaetus Coyle 301 entrances. Many females — all collected in late summer — contained small developing eggs that appeared to be timed for spring or summer oviposition. The morphology of the first instar is very similar to that of A. riversi. All of the shed chorions ex- amined contained the embryonic cuticle, an indication that hatching is also similar to that in A. riversi. The size of the known brood is shown in Table 3. Behavior Burrow structure. (See Table 6 for bur- row measurements.) The burrow is roughly tubular, with a slight expansion in diameter just below the entrance and a larger increase in diameter at the bottom end (Figs. 29-30). It is well lined with silk just below the entrance in the usually dry, loose surface soil, but the thickness of the lining decreases with depth as the soil increases in humidity and stability so that there is usually only a very thin lining over much of the rest of the burrow. The long axis of the upper portion of all burrows was nearly perpendicular to the plane of the adjacent ground surface. Almost every burrow curved to nearly horizontal at the bottom end. Entrance structure. The tubular silk lining of the burrow extends above the ground surface to fomi a rather thick but flexible collar which is collapsed inward to close off the entrance (Figs. 31-34; 51- 53). The collar is composed chiefly of soil particles and small bits of surface litter held together by silk so that a closed en- trance is remarkably well camouflaged. When open, this collar normally does not remain erect, but collapses outwardly so that it is roughly parallel to the ground surface. Usually only the proximal portion of the inner surface of the collar is a clean white extension of the burrow lining. The burrow entrance opening is often slightly elliptical. As in Antrodiaetus iinicoJor, a closed collar usually gives the appearance of a "double door" (which it is not) be- cause the spider collapses the collar in a bilateral manner by pulling inwardly on two opposing (lateral) sectors, producing a condensation of folding at the two op- posing points where these sectors meet. The collar is often somewhat reduced at these two opposing points, which cor- respond to the ends of the long axis of an elliptical entrance opening. No sealed col- lars were found. Egg sac structure and position. The only A. gertschi egg sac collected appeared to be very similar in size, shape, and con- struction to those of A. riversi (Fig. 29). It was attached to the ventral sector of the burrow wall less than one-third of the way do\\'n the burrow at a point where the bur- row appeared to have been slightly en- larged, but the sac produced a localized constriction of the burrow lumen. Molting behavior. A recently molted fe- male spider (2016 at M) was collected on 31 July. The burrow entrance was closed but unsealed, and a plug of soil had been placed one-third of the way down the bur- row. The weakly sclerotized spider was in the bottom of the burrow. The exuvia, heading up the burrow and with legs ex- tended, was lightly attached to the silk lining of the ventral sector of the burrow wall several centimeters from the bottom end. Defensive behavior. When a female in foraging position was disturbed by light or substrate vibration, she either rapidly withdrew down the burrow without closing the entrance or suddenly closed the collar with her pedipalps and anterior legs and held the collar shut without retreating. Feeding behavior. Burrow entrances were closed during daylight hours. Tlie youngest spiders opened their entrances and assmned a foraging posture earlier than older spiders, often as soon as it be- came twilight. Mature females usually opened their entrances well after nightfall. The foraging posture of A. gertschi ap- pears similar to that of A. riversi, with pedipalps and anterior legs resting on the 302 Bulletin Museum of Compaiativc Zoology, Vol. 141, No. 6 basal portion of thc> collar. If the plane of the burrow opening is not horizontal, the spider usually orients this stance with its venter against the lowest sector of the burro\\' opening, which is also usually the long end of an oval entrance opening. The exoskelctons of two freshly digested Camponotus worker ants were found in one adult female burrow, and ant pieces were observed on the ground surface next to another burrow. Only occasionally was prey cuticle found in burrows. Couriship and matinii, behavior. Males of A. iiertschi apparently wander only after dark. All wandering males (14) that I have collected were found after 10 p.m. even though collecting included both day- light hours and the period from 8 to 12 P.M. A female at O was discoxered at 11 P.M. feeding on a recently killed con- specific male several centimeters below her burrow entrance. Atypoides hadros Coyle Ecology Geographic di.str Unit ion and elevation range. Southern Illinois and eastern Mis- souri (Coyle, 1968). 400^1100 ft (120- 330 m). Habitat. I have observed A. hadros at only two localities. At Feme Clyffe State Park, Illinois, burrow aggregations were found in shallow shelter caves cut into the base of steep limestone bluffs. Mixed hard- wood forest extended along the base of these bluffs. Burrows were found only in those shelter caves kept moist by water seepage in late summer. The largest aggre- gation of burrows was found on the floor of the most humid shelter, where males, females, and immatures of both A. hadros and Antrodiaettts imicolor lived side by side. The soil was a moist to wet mixture of soft clay-loam and pebbles. Surface litter was absent. Burrows were present both at the well-lighted edges of the shel- ter caves and in the dimly lit areas. Except in the most humid shelter cave, A. unicolor uas more abundant than A. hadros. At Montauk State Park, Missouri, burrows of both A. hadros and Antrodiaetus were found side by side in moist reddish clay- loam with very little surface litter at the base of a high rock outcrop on the bank of the Current Rixer in mixed hardwood forest. The other three localities where A. hadros has been collected (Little Grand Canyon, Pine Hills, and Lusk Creek, Illi- nois) are also in forested areas where ravines, rock outcrops, and other sheltered spots are common. An adult female from Little Grand Canyon was taken from a burrow in a drier, less sheltered spot than the shelter caves of Feme Clyffe State Park (J. Beatty, personal communication). A. unicolor is also apparently more abun- dant at each of these three localities than is A. hadros. Life History Males. Collecting data ( Fig. 1 ) indicate that penultimate males molt to maturity during late summer or early fall and \\'an- der and mate during the first half of fall. Two mature males were collected on 13- 14 September in their burrows shortly after the final ecdysis. Pitfall traps planted at Pine Hills by J. Nelson in 1967 collected one wandering male during the first week the traps were set out (7-14 October), nine males during the next week, none during the next week, three during the next week (27 October-3 November), and none dur- ing the next week, the final \\'eek traps were put out. Females. The limited data (Fig. 2) hint that timing of brood development in A. hadros may be similar to that in A. uni- color. Second instar morphology is like that of A. riversi. The size of only one brood could be determined (Table 3). Behavior Btirroto structure. The burrows were roughly tubular and completely lined with Systematics and Biology of Antrodiaetus Coyle 303 silk. The lining of several adult female and both adult male burrows were espe- cially thick and white. Most burrows were widened slightly just below the entrance and at the bottom end (Fig. 35). Burrows of adult males averaged smaller than those of adult females (Table 6). Probably be- cause of the pebbly soil, many burrows had abrui^t curves, with the slope of die bottom end varying from vertical to hori- zontal. Entrance structure. The entrances of only two adult females and two immatures were observed. The tubular silk lining of the burrow was extended slightly above the ground surface to form a very short and thin collar (Figs. 36-37). Soil particles were incorporated into the collars so that the external surface had the same texture and color as the surrounding soil surface. The height of each collar ^^'as quite strongly reduced at two opposing points on its circumference. When the collar is collapsed, the folding is concentrated at these two points, and the opposing higher sectors or flaps of the collar meet and overlap in the center along the line con- necting the two reduced points. Brood beliavior. No egg sac remnants were present in the four burrows with second instar broods collected from Feme Clyffe State Park. These spiderlings re- acted quickly to stimuli and could move about rapidly. They were clustered in the bottom of two burrows, and were more scattered in the other two but retreated do\\'n to the burrow bottom as I exposed each burrow. Molting behavior. The burrows of both recently matured males were heavily lined with silk. Both spiders were sufficiently sclerotized to be active. The recently shed exuviae had been pushed into the very bot- tom ends of the burrows and covered over with a thin layer of silk. Since both bur- rows were discovered by scraping the soil surface, I could not determine whether the closed collars had been sealed with silk. FeecUns, behavior. Burrow entrances were closed during the daytime. Burrows of several immature A. haclros observed 90 minutes after nightfall had the collars open only a crack, while adjacent A. unicolor immatures had their collars wide open. Fragmented prey exoskeletons were packed into the bottom ends of a few burrows. Such trash from two immature biurows was examined and found to consist of ant and beetle parts. Aliatypus californicus (Banks) Ecology Geograplnc distribution and elevation range. West central California from the San Francisco Bav region to the Monterey Peninsula. 1000^2900 ft (300-880 m). Habitat. I have collected only three specimens, all these near Felton, California, between 1000 and 1600 ft in or near aggre- gations of Atypoidcs riversi. Actinoxia burrows were also present but as uncom- mon as those of A. californicus. All bur- rows were found on steeply inclined road banks in rather dense mixed deciduous- coniferous forest (tan oak, pine, and some coast redwoods ) . The soil varied from clay to sandy loam, was quite rocky, and was dry at the surface. The holotype specimen was collected at the side of a stream (Banks, 1896). Smith (1908) collected A. californicus in the foot- hills and mountains on both sides of the Santa Clara Valley, finding burrows most commonly along banks of streams and roads in "fine compact sandy soil." Occasion- ally bun-ows were found in "sand\' adobe" soil. These banks with burrow aggregations "usually ha\e little or no vegetation upon them other than short scattered moss." Smith found A. californicus "commonly associated with Euti/chides [= Actinoxia] and Atypoides [riversi]." Life History Males. Smith (1908) collected the only two known males of A. californicus during October (Fig. 1), but failed to indicate 304 Bulletin Museum of Comparative Zoology, Vol. 141, No. 6 \\ hcther tliey were taken from burrows or collected wandering. Females. Smith ( 1908 ) observed that most burrows (at least of those found in exposed banks) were sealed shut from June to "the first rains in December." "None were found with sealed doors from the end of December to the end of April." At least some burrows in shaded stream banks were not sealed for such a long time. Such estivating beha\'ior during the dry season may be common to other spe- cies of Aliatypus. Behavior Burrow structure. Smith (190Qi (Figs. 310-312). Seminal into 2 ian partition receptacles 334 Bulletin Museum of Comparative Zoology, Vol. 141, No. 6 closely paired, stalks long and slender, bowls weakly developed. Setae absent (or very rare) on labium except for anterior and lateral edges. Japan (island of Honshu) (rorctzi group) rorctzi lb. Bursa copulatrix without median partition. Seminal receptacles not or only weakly paired, or stalks short to moderately long, or bowls moderately to strongly developed (Figs. 270-309). (Some eastern A. paci- ficxis individuals are exceptions. ) Setae scattered over much of labium. N. Amer- ica or Japan ( island of Hokkaido ) 2 2a. From southwestern U. S. (Arizona, New Mexico) or central U. S. (Arkansas, Mis- souri, Kansas, Nebraska). If from central U. S. then AMD = 0.09-0.13 mm (Figs. 135, 136). IVCTR usually 0 to 2, cheli- ceral microteeth proportionately large (Figs. 166, 167) — _ (Uncolnianus group) 3 2b. Elsewhere or if from central U. S. (A. uni- color) then AMD = 0.11-0.22 mm (Fig. 130), IVCTR usually 3 or more, cheliceral microteeth proportionately smaller (Fig. 158) {unicolor group) 5 3a. Stalk and bowl portion of seminal recep- tacles weakly sclerotized and small, border with bull:) poorly defined (Figs. 304, 305). Abdomen rather dark yellow-gray or medium brown. CMT/IVMT greater than 2.50. Central U. S. (Nebraska, Kansas ) Uncolnianus 3b. Stalk and bowl portion well sclerotized and slightly larger, l^order with bulb well defined (Figs. 306-309). Abdomen Hghter. CMT/IVMT = 1.00-2.25. Southwestern or central U. S. 4 4a. IVML/IML = 1.36-1.38. IFL/IVFL = 1.04-1.05. Carapace narrower (CW/CL = 0.76-0.77) (Fig. 118). Central U. S. (Arkansas, Missouri) stygius 4b. IVML/IML = 1.21-1.26. IFL/IVFL = 1.09-1.13. Carapace broader ( CW/CL = 0.78-0.82). Southwestern U. S. (Arizona, New Mexico) apachectis 5a. Japan (island of Hokkaido) yesoensis 5b. North America — 6 6a. Eastern U. S. (east of Great Plains) 7 6b. Western N. America (Rocky Mtns. and westward) 8 7a. Combined presence of minute AL spin- neret vestiges (Fig. 317), rather short densely distributed dorsal abdominal background setae (Fig. 316), and thin elongate setae dorsally just posterior of pedicel (Fig. 314). Seminal receptacle stalks not expanded basally (Figs. 280- 283). SL/SW = 1.08-1.16 rohustus 7b. Lacking one or more of first 3 character states above; usually no AL spinneret vestiges, usually longer more sparsely dis- tributed dorsal abdominal background setae (Fig. 315), usually shorter thicker setae dorsally just posterior of pedicel (Fig. 313). Seminal receptacle stalks usu- ally expanded basally (Figs. 270-279). SL/SW = 1.13-1.27 unicolor 8a. Setae distributed over entire upper ectal surface of chelicera (Figs. 171, 172). Carapace pale to medium dark gray- yellow 9 8b. Small to large setaless area on upper ectal surface of chelicera (Fig. 170). Carapace usually darker 11 9a. CL over 6.2 mm, IFL/IML less than 1.69, posterior abdominal dorsimi without darkly pigmented area, IVTL/CL = 0.40 or more, and seminal receptacles unpaired (Figs. 292-302) 10 9b. CL under 6.2 mm or IFL/IML = 1.69 or more or abdominal dorsimi dark or IVTL/ CL = 0.41 or less or seminal receptacles paired (Figs. 288-291) 11 10a. Seminal receptacle stalks not expanded basally (Figs. 299-300). SL/SW=1.20- 1.37 (Fig. 141) hapeni 10b. Seminal receptacle stalks expanded basally (Figs. 295-298). SL/SW = 1.06-1.19 (Fig. 140) vumtanus 11a. IVTL/IVTarL = 1.90 or more. IVML/ IML = 1.42 or more cerhcrus lib. IVTL/IVTarL less than 1.90. IVML/ IML= 1.09-1.50 12 12a. ITarL usually more tlian 1.28 mm. IVML/ IML = 1.09-1.38 pacificm 12b. ITarL usually less than 1.28 mm. IVML/ IML = 1.31-1.50 piignax The Unicolor Group Descriptive dia'. Although Simon (1892) also indicated the similarity of Mygale gracilis to Brachybothrium spe- cies, no author has subsequently expressed this opinion. Hentz describes or illustrates sufficient diagnostic characters to show that Mygale gracilis belongs to the taxon Antrodiaetus as presently defined. In addition to an antrodiaetid body form, eye arrangement, 336 Bulletin Museum of Comparative Zoology, Vol. 141, No. 6 and pcdipalp shape, a conductor is illus- trated on the palpus, the spinnerets are long, and metatarsus I is sinuous. (Hentz presumably misinterpreted this cmvature as a "notch.") Unfortunately the identity of 'Mijiiale unicolor is less certain. The eye arrangement is strange, no thoracic groove is illustrated, and spinneret form is not mentioned. The reduced endites and general body form are antrodiaetidlike. To avoid upsetting the generic and familial nomenclature, I have decided that — as the meager evidence in Hentz's descriptions indicates — these two names are synony- mous, and as first reviser I select Mygale unicolor as the senior synonym. Atkinson (1886) did not indicate the type deposition of the two species he described. Simon (1890), who was loaned the type of Nidlvalvata marxi, reported that the Atkinson types were in the U. S. National Museum. However, a small adult female labeled "Type Brocliyhothriiim an- gustata" and a probably immature female labeled "Type Brachyhothriiim marxii' are in the Cornell University collection pres- ently housed in the AMNH. Unfortunately the USNM "types" are presently unavail- able for study. The Cornell specimens do not match the carapace dimensions re- corded by Atkinson for the specimens he described. Although miable to examine the syntypes of Nidivolvata angustatum, Simon (1890) stated that this was perhaps a synonym of Brachyhothriiim acccntuo- ttim. He examined the type of N. marxi and noted its great similarity to B. ac- centuatiim. He described the pedipalpal tarsus as being swollen (an indication that the specimen may be a penultimate male), using this as the diagnostic character of B. marxi in a key to species. Description. See Tables 12-14, which contain measurements, meristic data, and diagnostic ratios for a sample of the species. Male. Carapace: Figure 120. Setae very sparse on pars thoracica, sometimes moder- ately dense along lateral border. Chelic- erae: Figures 145, 146. Large setaless area on upper ectal surface. Pcdipalps: Figures 17:3-174, 188-194, 81. Tibia moder- ately to strongly swollen in lateral view; greatest diameter in lateral view at ap- proximately 1/3 of distance from proximal to distal end. Tip of o.c.s. of palpus usually blunt ( occasionally rather pointed ) ; closely appressed to i.c.s. I.c.s. tip well sclerotized; often slightly curved; usually tapered to relatively narrow point. Leg I: Figures 2.34-240. Group of macrosetae (33% to 100% are ensiform) centered at 1/2 to 2/3 of distance from proximal to distal end of tibia on prolateral surface. Fewer more scattered macrosetae (0% to 100% are ensiforai) along retrolateral aspect of ven- tral surface of tibia; distalmost macroseta of this group positioned at 2/3 or less (rarely at 3/4) of distance from proximal to distal end. Zero to 8 macrosetae (0% to 100%) are ensiform) scattered between these groups, mainly along prolateral aspect of ventral surface. Tibia nearly cylindrical except sometimes swollen at prolateral macrosetae group. One to 3 (rarely none) macrosetae ventrally at distal end of metatarsus; rarely a macroseta present at 2/3 of distance from proximal to distal end on prolateral surface. Metatar- sus weakly to moderately sinuous in ventral view. A])domen: 3 dorsal heavily sclero- tized usually continuous patches. Posterior patch smaller than anterior 2; occasionally reduced to bilateral pair of sclerotized spots. Coloration: Carapace pale gray- yellow to brown. Sternum pale yellow to gray-yellow; pedipalpal coxae similar to or slightly darker than sternum; labium markedly darker than sternum. Patella, tibia, and metatarsus of leg I Hght gray- orange to dark red-brown, usually darker than carapace; femur dorsally slightly lighter than more distal segments, ventrally even lighter. Pcdipalps and other legs dorsally similar to carapace, ventrally lighter. Abdomen pale yellow-gray to gray- brown or purple-gray; dorsal sclerotized Systematics and Biology of Antrodiaetus Coijlc 337 patches slightly darker except sclerite of second patch usually similar to carapace. Female. Carapace: Figure 130. Setae sparse to moderately dense on pars thorac- ica; densest along lateral border. Ster- num: Figure 138. CheJicerac: Figure 158. Microteeth positioned beside last 9 to last 3 macroteeth; frequently extend beyond last macrotooth. Microteeth usually pro- portionately small to medium size. Large setalcss area on upper ectal surface. Abdo- men: Figures 313, 315. Dorsal background setae sparsely to moderately long. Con- vergent median dorsal setae just posterior to pedicel usually short and rather stout; occasionally long and slender. Genitalia: Figures 270-279. Seminal receptacles with well-sclerotized stalk and bowl; vary from moderately long and moderatelv thick to quite stout. Stalk usually expanded (some- times slightly) at base. Bowl well de- veloped; its border with bulb sharply de- fined. Floor of bursa copulatrix weakly sclerotized. Coloration: Carapace pale yel- low to dark brown; pars cephalica usually darker than pars thoracica. Sternum simi- lar to pars thoracica; labium much darker, often similar to chclicerae; pedipalpal coxae usually slightb; lighter than labium. Chclicerae pale gray-yellow to dark red- brown, usually darker than pars cephalica. Pedipalps and legs dorsally similar to adjacent portion of carapace, ventrally lighter. Abdomen pale yellow-gray to dark brown, sometimes with faint chevron pattern dorsally behind tergite; tergite usually similar to pars thoracica. Diapwsis. Males. The shapes and macro- setae distribution patterns of the tibia and metatarsus of leg I (Figs. 234-240; Table 12) are distinctive for this species and include the best diagnostic characters. A. unicolor has a proportionately long meta- tarsus I (Fig. 235), a proportionately short pedijoalpal femur, and rc-lativelv large AME's (Fig. 120) so that ITL/IML, PFL/ CL, and AMD/CL (Table 13) are some- times useful in distinguishing this species from other unicolor group species. Fe- males. Characters useful in separating A. unicolor females from those of the sym- patric A. rohustus are listed in the A. ro- hustus diagnosis. If A. unicolor were not geographically separated from the other unicolor group species, identification of females would be difficult or impossible. Because of a proportionately short femur I and a fairly small mean CMT number, A. unicolor females differ most from A. pacificus and A. cerherus females in IFL/ CL and some other IFL ratios (Table 14) and from A. pacificus and A. montanus in CMT number (Table 14). Variation. Males. Although most char- acters studied exhibit large variation, only a small number of these vary discontinu- ously. Strong geographic variation occurs with the following ratios (listed in order of degree of discontinuous variation ) : PTT/ PTL (Fig. 81), IML/ITarL (Fig. 83), IML/CL (Fig. 82), ETL/CL (Fig. 84), PFL/CL, ALD/CL, and ITL/IML. In these, all samples (B, C, E, F, G, I, P, Q) (Fig. 85) except two (A, N) are connected to others by broadly overlapping variation ranges. Sample A, clearly the most diver- gent sample, is particularlv divergent in ratios IML/CL (Fig. 82), ETL/CL (Figs. 84, 188-194), PFL/CL, and PTT/PTL. Sample N is also divergent, especiallv in PTT/PTL (Figs. 81, 174), IML/ITarL (Fig. 83), and IFL ITarL. Sample A is much more similar to N in most characters \\'ith strong geographic variation than to other samples, whereas IV is usually only slightly more similar to A than to other samples. Individuals of both samples usually have a more swollen pedipalpal tibia and a proportionately shorter meta- tarsus I and pedipalpal femur than other males. Sample A individuals have a par- ticularly small ETL. Sample N males have a proportionately long tarsus I. ITL, IML, PTL, and PFL have the least continuous geographic variation of any measurements, but other measurements correlated \\'ith body size show similar pat- terns of weaker geographic variation in 338 BiiUetin Museum of Comparative Zoology, Vol. 141, No. 6 which the small-bodied sample A is some- what divergent and the slightly larger- bodied males from N are less divergent from the rest of the samples (Figs. 81, 82). Among meristic characters the strongest geographic variation occurs in the number of prolateral tibia I ensiform macrosetae (Figs. 87, 234-240). Sample N, with few such macrosetae, is divergent, but its vari- ation overlaps that of other samples. Sample Q and several males at E possess a distinctly higher number of ensifonn macrosetae on the prolateral aspect of the ventral surface of tibia I (Fig. 240) than do most other A. tmicolor males. Male carapace color varies in a geographic pat- tern similar to that in the females (see below), but the variation is continuous. The single male from O is particularly divergent in two characters; the pedipalpal tibia is strongly swollen like that of males at A and N (Fie. 81), and metatarsal I macrosetae are absent (Table 12). The single male from L is extremely small ( Fig. 82) but in all ratios and other nonmeasure- ment characters closely resembles most males (except those from A and N). Noteworthy continuous variation occurs in several characters. The o.c.s. tip, usually blunt and rounded (Figs. 189, 191), is rarely almost pointed (Fig. 194). Ex- tremely wide variation occurs in EGS num- ber (Fig. 86), which is correlated with body size. The relative size and shape of the anterior-dorsal cheliceral prominence varies considerabK'. Figure 146 illusti'atcs an exceptionally small prominence. Figure 145 a slightly larger than nonual promi- nance. The metatarsus I macrosetae pattern also varies widely (Figs. 234, 236; Table 12). Females. Strong geographic variation occurs in five ratio characters (listed in order of approximate degree of discon- tinuity): IFL/IVFL (Fig. 88), IVML/CL, IVTL/CL (Fig. 93), ALS/CL, and IVML/ IML (Fig. 89). Samples I and D arc each divergent from many other samples in some of these characters. No measure- ments or meristic characters exhibit strong geographic variation. Unlike the males, in which each popu- lation sample is homogeneous, some fe- male samples show discontinuous \'ariation within. One of the strongest and, to the casual observer, most obvious instances of discontinuous variation in this species is in coloration, particularly female carapace coloration (Fig. 92). Most samples are homogeneously either light or dark with little overlap bet\\'een these light and dark samples. In a few locaHties (7, K, L), how- ever, distinct light and dark individuals were found living sympatrically. Although light populations appear to be more com- mon in the south and west part of the species range, dark females have been col- lected in the western parts of Tennessee and Kentucky and in Arkansas. An extensive search was made for char- acters (other than color) which might also distinguish dark and light samples. Several ratio characters separated light and dark specimens in each of samples I, K, and L, b/Ut only SW/SL exhibits such nearly dis- continuous variation in all three samples. Figures 81-87. Geographic variation in Antrodiaetus unicolor males. (All measurements in mm. For scatter diagrams, large black dots represent sample A, circles sample N, and small black dots rest of species sample.) 81. Scatter diagram of PTT and PTL; representative specimens illustrated. (X's represent Anfroc/ioetus robustus males.) 82. Scatter diagram of IML and CL. 83. Scatter diagram of IML and ITarL. 84. Modified Dice-Leraas diagram of ETL/CL variation compared witfi that of A. robustus. (Horizontal line represents the observed range, vertical line the mean, open rectangle the standard deviation, solid black rectangle the 95 per cent confidence interval for the mean, number to right of range line the sample size, and letter in left column the sample locality.) 85. Map of population sample localities. 86. Histogram of EGS number compared with that of A. robustus. (One unit of vertical scale represents one specimen.) 87. Histogram of number of ensiform macrosetae on prolateral surface of tibia I. (One unit of vertical scale represents one tibia.) Right column contains percentage of en- siform macrosetae out of total number of macrosetae in this group. Systematics and Biology of Antrodiaetus • Cotjle 339 PTL 5.C0 4.00 ■- 3.00 -- o. • • o O 00 o o 82 4.00 5.00 6.00 CL robustus unicolor 86 10 20 30 40 50 60 EGS robustus unicolor • • • • 5.00 - - • • • « • • • • • • • • • • • • • • • • • • - • • • • • • • • • « • • • • • • • • • • o 4.00 - o L • — • o • 8 • 0 OOOq 0 o o • • 83 • • 3 0(1 - 1 • H ' — 1 '— 2.00 2.50 ITarL per cent ensilorm 61 -91 37 78 57 90 E.F.G ^ 87 61 - 100 61 -92 57- 100 1.. 44-84 87 - 100 . 1 ■ I 33-93 p Q ottiers 10 20 30 ENSIFORM MACROSETAE ON PROLATERAL SURFACE OF TIBIA I 340 Bulletin Museum of Comparative Zoology, Vol. 141, No. 6 (In all subsequent population sample com- parisons, dark individuals were excluded from sample 7 and light individuals from sample K.) Marked genetic heterogeneity \\ithin the light specimens is indicated by the divergence of sample 7. This sample is clearly distinct from all other light samples in IVML/IML (Fig. 89), markedly different in a few other ratios (IVTL/CL, Fig. 93), and more similar in most char- acters to several of the dark samples. (Male sample I, which matches the light female 7 sample in color, is not divergent in any character studied.) Light samples F, O, Q, and R are quite similar to each other in almost all characters, and as a group differ most from the dark samples in ratios IFL/IVFL (Fig. 88) and IVML/ IML (Fig. 89). However, at least one of these light samples exhibits considerable overlap with one or more dark sample in every character. Sample O is clearly the light sample most distinct from the dark samples. Sample N, which is intermediate in coloration (Fig. 92), more closely re- sembles dark samples in some characters and light samples in other characters and is not divergent in any character. As with color, two setation characters re- vealed discontinuous intrapopulation vari- ation. Three females from K differ distinctly from all other females in possess- ing extremely thick and short background setae on the posterior 1/3 of the abdomen. In addition, these three females have a distinctly lower IML/ItarL (Fig. 90), lower IVML/IVTarL, higher CMT/IVMT, longer thinner setae medially on the ab- dominal dorsum just behind the petiole, shorter denser abdominal background setae, and slightly more robust chelicerae than the other K females. Also the seminal receptacle stalk bases are not expanded (Fig. 277) as in the rest of sample K (Fig. 272). These three specimens even differ considerably from the rest of the species sample in IML/ITarL (Fig. 90), IVML/IVTarL, and the shape of the semi- nal receptacle stalks. The large variance of the whole of sample K in several char- acters (Figs. 88, 89) further illustrates its marked heterogeneity. Females with very thick short setae on their sternum are found within several population samples (M, P, L, D) and as scattered indi\iduals over a large part of the species' distribution range. Most other individuals possess only slender sternal setae; however, there are a few individuals intermediate in this character. Stout sternal setae are not found on light colored in- dividuals. Comparisons of the stout sternal setae subsample to the slender setae sub- sample within each of samples M and P, and comparisons of the entirely stout setae sample D against each dark slender setae population sample were performed to dis- cover if these sternal setae forms are distinct in other characters. Within each comparison at least a few ratios separated the two foiTns (Fig. 91), but few of these characters were common to all compari- sons. Sample D is markedly divergent in several ratio characters from many other A. tinicolor population samples (Figs. 88, 93). When all stout sternal setae individuals are compared to a large sample of dark colored slender sternal setae individuals, only IVFL/CL yields rather distinct separation (Fig. 91). Figures 88-93. Geographic variation in Antrodiaetus unicolor females. (Dark colored samples identified by black dot, ligfit samples by circle, and intermediate colored sample by half-blackened circle.) 88. Mod. Dice-Leraas diagram of IFL/IVFL variation. 89. Mod. Dice-Leraas diagram of IVML, IML variation. 90. Mod. Dice-Leraas diagram of IML'ITarL variation in sample with stout posterior abdominal background setae and samples with normal (slender) abdominal setae. 91 . Mod. Dice-Leraas diagram of IVFL/CL variation in samples with stout sternal setae and samples with slender sternal setae. 92. Color variation. Frequency of pars cephalica color values given for each sample. (Key to color grid at lower right. Full explanation in Methods section.) 93. Mod. Dice-Leraas diagram of IVTL/CL variation. Systematics and Biology of Antrodiaetus • Coijle 341 B C D F H I J K 2.00 IML/lTarL 1.20 1.30 1.40 Qg ezml/iml 2 5 1 1 2 9 1 1 1 1 4 10 2 2 1 2 8 1 1 3 5 2 3 2 1 1 2 1 3 1 1 1 3 5 2 1 2 1 2 2 2 1 1 1 1 3 29 18 L M N O 1 1 1 2 1 14 7 3 1 1 5 12 8 2 P 1 1 2 6 9 7 2 7 9 3 5 20 9 4 1 4 5 4 3 14 5 3 1 1 3 Q R -+- .350 .400 gttl/cl .450 93 yellow 1 1 A B 3 6 1 1 1 2 Basic hue change I orange Increase in amount of black in color 92 342 Bulletin Museum of Comparative Zoology, Vol. 141, No. 6 Noteworthy variation without a clear geographic pattern occurs in two other qualitative characters. While the great majority of females have short con\'erging setae medially on the abdominal dorsum just behind the pedicel (Fig. 313), a few have here long slender setae (as in A. robu.stus. Fig. 314) or setae of intermediate length. Such aberrant individuals are scat- tered in several distantly separated dark- population samples and are rare in light samples. No other characters vary con- cordantly with this one. Female genitalia are quite variable in this species, but the variation is continuous with a few excep- tions. Figures 274 and 275 illustrate the wide variation apparently common even in a genetically relatively homogeneous popu- lation. Figures 270-279 illustrate much of the total variation found, and do not neces- sarily represent the modal structure for each population sample. Divergent seminal receptacle fonn was found only in the three K specimens with stout posterior abdominal background setae and in both dark specimens at L. In these, all seminal receptacles lacked expanded stalk bases (Fig. 277) and also resembled A. ro])ustiis receptacle form in some qualitative aspects. Discus.s'ion. Variation in A. iinicolor is somewhat complex and difficult to inter- pret. In addition to strong geographic variation patterns, there are also instances of different forms occurring sympatrically. It is conceivable that further studies may reveal convincing evidence that this series of populations consists of several repro- ductively isolated gene pools. The two clearly divergent population samples (A and N) exposed in the analysis of variation of male characters are ap- parently genetically quite different from the other samples (and less different from each other) and each can be readily char- acterized by a combination of several char- acters. Both are clearly geographically peripheral populations presumably ex- changing genes only at a relatively slow rate, with the sampled populations located in the region of greatest population den- sity. A study of geographically inter- mediate populations is necessary to deter- mine to approximately what degree (if any) gene exchange is occurring. The possibility was examined that char- acter divergence might be occun'ing be- tween A. iinicolor and A. rohu.stii.s populations in the vicinity of A and that this might be partly responsible for the di\'ergence of sample A. Sample A is markedly more different from the sym- patric (Allegheny Co. and Westmoreland Co.) A. rohustus males than arc allopatric samples B and C — the pattern that could result from character divergence — in 10 ratios including PTT/PTL (Fig. 81), PFL/ CL and OQW/CL. Several of the char- acters that best separate sample A from the other A. iinicolor samples (IML/CL, ETL CL [Fig. 84], IML/ITarL) either do not show this pattern or show the re- verse pattern, and two other characters also show the reverse pattern. Thus, while some of sample A's uniqueness could be the result of character divergence, some cannot. The significance of the color variation, discontinuous in the female samples, is difficult to determine. Considerable genetic heterogeneity exists within the total light sample and the entire dark sample. Upon removing the more divergent elements, the remaining light and dark samples are each still genetically rather heterogeneous and they differ, but are still not distinct in any character except color. It appears imlikely that the color forms are cither geographic variants or ecophcnotypes, since they are found together in the same general habitat in at least three localities. It is possible that these color fonns are distinct species, but the absence of any other character which in many samples consistently matches the discontinuity in coloration argues against this. Perhaps the most feasible explanation is genetic polymor- phism; if so, there would appear to be a surprisingly large number of populations in Systematics and Biology of Antrodiaetus Coijle 343 which fixation of one moiph has occurred. Analysis of larger samples from localities where the forms are sympatric, close obscr\'ation for microhabitat differences and local spatial separation of the forms, a study of color variation of adults reared from single broods, and a compari- son of the physiological basis of coloration in the two forms are kinds of studies needed to solve this problem. The specimens from K with stout poste- rior abdominal background setae are so dis- tinct from the rest of the local sample that they may be reproductively isolated. Tlie possibility of genetic polymorphism involv- ing several characters or of ecological races cannot be ruled out presently. It is unlikely that age differences could produce dis- continuity in so many characters in this sample and not in other samples. Tlie divergence of sample D and the other specimens with very stout sternal setae from the rest of the dark specimens is as- sumed to represent simple intraspecific genetic variation because of the occurrence of a few intermediate indi\'iduals and be- cause only one other character separates most of the individuals of these two forms. These divergent setation forais require further study along the lines suggested above. Particularly helpful \\'ould be col- lections of males from K and D. Disirihiiiion. Centered in the central and southern Appalachian region of east- em United States with populations extend- ing peripherally in all directions, as far west as the Ozark region and apparently south to near the Gulf Coast ( Map 1 ) . Records. ALABAMA. De Kolb Co.: De Soto St. Pk. [Q], near Fort Payne, July- Aug. 1937, 25; Oct., S; Dec, $■ 1800- 1900 ft, 389. Madison Co.: Monte Sano St. Pk., Dec. 1940, $. Marshall Co.: Little Mtn. St. Pk. [R], 600-700 ft, 24 9 . — 0.8 mi. N of Grant, 9. ARKANSAS. Stone Co.: Blanchard Springs Rec. Area (USPS), 11 Sept. 1966, 3.00- 7 A m ^^ ' . • ^m • • • • (•/ • •• • • • 1 1 , , ,_ 3.00 4.00 5.00 IML 1.50 -- Q. l.OO Fig. 175 Fig. 176 Fig. 178 \ / X "^x" X " 5 • X. X X X • • ' lo • D • • O • O • 1 • o > • o o • • • o Fig. 177 2.50 96 3.00 PTL 3.50 C D E F G H J K L IVl R, S, T, V - w X z -^^ 95 1.05 1 I.IO IFL /lYFL 1.15 99 1.00 2.00 AMS /AMD 100 1.40 H ' 1 1-50 1.60 ifl/itl 1.70 4.00 5.00 6.00 7.00 CL Figures 94-100. Geographic variation in Anfrod/aefus pacificus. (All measuremenfs in mm.) 94-97. Males. 94. Scatter diagram of ITL and IML. (X's represent samples R-Z, circles sample M, small black dots rest of samples, and large black dots Anfrod/oefus occu/fus males.) 95. Mod. Dice-Leraas diagram of IFL/ITL variation. 96. Scatter diagram of PTT and PTL. (Circles represent A. pacificus males sympatric with A. occultus, black dots rest of A. pacificus males, and x's A. occultus males.) 97. Mod. Dice-Leraas diagram of CL variation. 98. Map of population sample localities. 99-100. Females. 99. Mod. Dice-Leraas diagram of IFL.'IVFL variation. 100. Mod. Dice-Leraas diagram of AMS/AMD variation. 352 Bulletin Museum of Comparative Zoology, Vol. 141, No. 6 Figure 147 illustrates a prominent pro- tuberance (one which might function like the apparently homologous cheliceral apophysis of Atypoides), Figure 148 illus- trates a modal A. pacificus cheliceral protuberance, and Figure 149 represents the low end of the range of variation in this character. Figures 96 and 175-177 illus- trate the continuous variation in the shape of the pedipalpal tibia. Usually there are three noncontinuous sclerotized patches on the male abdominal dorsum, with the pos- terior patch smallest but varying greatly in size. Occasionally this third patch is re- duced to a bilateral pair of sclerotized spots; on rare occasions it is absent. Oc- casionally in the western samples and usually in the Blue Mountains samples the three sclerotized patches are continuous. Distribution. The Pacific coast of North America from San Francisco Bay north to southern Alaska, with outlying, perhaps isolated, populations in the montane areas of western Idaho, northern Oregon, and adjacent Washington (Map 2). Note- worthy locality records (not shown on Map 2) of immatures that are almost certainly A. pacificus are: Marin and Sonoma Counties, California, and Terrace, British Columbia (near the southern tip of Alaska). Records. ALASKA. Ketchikan [A], 18- 28 July 1951, $ ; 23 Aug. 1952, $ ■ 27 Julv 1951, $■ summer 1937, <^ . BRITISH CO- LUMBIA. Vancouver Island [B]: Victoria, 17 Aug. 1966, $ ; 30 July 1967, i ■ 3 Sept. 1966, i- 27 Oct. 1934, $; 18 Nov. 1928, $; 29. — Sidney, 10 Sept. 1967, $. — Esquimalt, 30 Aug. 1947, 5. — Tofino, 7-15 Aug. 1953, i ■ 22 June-12 July 1950, 5; 29. — Wellington, 1.5-31 Aug. 1951, 5 Oct. 1949, 15 Aug. 1949, 35; 29. — Na- naimo, 13 Sept. 1935, 6 . — French Creek, 6 Sept. 1949, S. — Caledonia, Kyuquot Sound, 9. — Kyuquot, i ; 29 July 1931 ( 6 ); 17 Aug. 1958, S, 9; 31 Sept., S. — Kyuquot, Weave's Island, 29. — Kyuquot, Spring Island, 9 . Queen Charlotte Islands: Frederick Island, 29. — Yalcoun River, 11 Aug. 1961, S. CALIFORNIA. Humboldt Co.: Miranda, 3 June 1936, (Table 14) help distinguish A. pugnax from A. hageni and A. cerberiis. The best ratio for separating A. pugnax from A. pacificus is IVML/IML, and from A. mon- tamis is IFL/IML (Table 14). Tlie small cheliceral microteeth (Fig. 161) of A. pugnax help distinguish this species from both A. hageni and A. montamis. Variation. A number of characters ex- hibit strong geographic variation in male A. pugnax. On the other hand, compari- sons of four female samples (from Cor- vallis. Ore., and from Ft. Lewis, Walla Walla, and Lewis and Clark St. Park, Wash.) reveal no marked geographic vari- Systematics and Biology of Antrodiaetus Cotjlc 357 ation. Male quantitative characters with the strongest geographic variation are (in approximate order of the degree of dis- continuity of variation): IFL/IML, IFL/ ITL, IML/CL, IFL/ITarL, IMD/IML, TTD/ITL, metatarsus I macrosetae num- ber, tibia I macrosetae number, EGS, and ALD/CL. For the first four of these characters, population sample H is distinct from the northern and eastern samples, which have a proportionately shorter leg I tibia, metatarsus, and tarsus (Figs. 103-108). Sample G is similar in these characters to sample H or is intermediate. The geo- graphic patterns of IMD/IML (Fig. 101) and ITD/ITL variation are very similar to each other and resemble those of the above characters. However, this variation appears somewhat clinal, with sample F intermediate. The thickest tibiae and metatarsi I (also with the most prominent protuberances) are in the northeastern part of the species range and are the most slen- der in the southwestern portion ( Fig. 101 ) . All sample H specimens have on metatar- sus I a macroseta F (Fig. 252), not present on other males except the one from A. Also there is a much greater tendency for sample H individuals to have metatarsal macrosetae C and D (Fig. 252) and for macroseta B to be ensiform. The south- western samples (G, H) have a signifi- cantly higher mean number of EGS (Fig. 102) and of macrosetae on the retrolateral aspect of the ventral surface of tibia I (Fig. 106) than northern and eastern samples. ALD/CL exhibits a different pattern of geographic variation, with samples E and F having markedly smaller mean values of this character than samples B and H, but only small gaps separate the variation ranges of these small samples. Clearly population H is genetically di- vergent and could be satisfactorily dis- tinguished from the known northern and eastern samples by most of the above (male) characters. However, present evi- dence does not justify concluding that population H is reproductively isolated from the rest. Geographically intermediate and larger samples need to be studied. The inteiTnediate position of sample G and (occasionally) sample F in the variation pattern of some markedly varying char- acters is some evidence that populations H and G are exchanging genes, but per- haps at a reduced rate, with the more northerly known populations. Perhaps fa- vorable A. piignax habitat is discontinuous in the Willamette River Valley area of northwestern Oregon, thus hindering gene exchange with northern and eastern popu- lations. Another factor possibly involved is character divergence. It seems highly unlikely that character divergence between A. pugnax and A. pacificiis could cause such geographic vari- ation in A. pugnax, since A. pugnax is throughout its entire range sympatric with A. pacificus (Map 2), unless possibly an uncommon situation of strong local com- petition exists between A. pacificus and semi-isolated A. pugnax populations only in the Eugene, Oregon, area (samples G and H) where their habitats are spatially intermingled. An examination of all char- acters for variation patterns that could result from such character divergence re- veals that 1/6 do have such a variation pattern (one in which the A. pugnax popu- lation samples G and H are markedly more distinct from the Eugene [sympatric] A. pacificus sample than are the other A. pug- nax samples). But almost as many char- acters had just the opposite pattern. IFL/ ITarL, ITarL/CL, FFL/CL, and IFL/ITL show the former type of pattern most strongly (Fig. 108), whereas IML/CL, ITL/lfarL, and IFL/IML exhibit the strongest reverse pattern (Fig. 107). There is considerable evidence, however, to suggest character divergence between A. pugnax and A. occulttis. Records indi- cate that A. occulfus is found only in the Willamette River Valley region, where it is sympatric with A. pugnax (Map 2). Both species have been collected from 358 Bulletin Museum of Comparative Zoology, Vol. 141, No. 6 3.00 ITL Mocroselce on relrolalerol ospecl o( ventrol surloce ol tibia I 3.50 IML pugnax occultus pcjcilicus Eugene (ol HI WailoWoMa lot Bl Porliond Ineor fl lotQl somple 107 .700 .800 IML/CL pugnux pacilicus .900 A-F G H Eugene lot HI Walla Walla lal Bl Portland (near Fl totol sample 108 I i^ I 2.00 2.20 IFL/lTarL 2.« Figures 101-108. Geographic variation in Antrodiaetus pugnax males. (All measurements in mm. For scatter diagrams large block dots represent sample H, x's sample G, and circles samples A-F.] 101. Scatter diagram of IMD and IML. 102. Histo- gram of EGS number compared with that of Antrodiaetus occultus. (One unit of vertical scale represents one specimen.) 103. Scatter diagram of IFL and ITL. (Small black dots represent A. occultus males.) 104. Map of population sample localities. 105. Scatter diagram of IFL and IML. 106. Histogram of tibia I retrolateral-ventraj macrosetae number . (One unit of ver- tical scale represents one tibia.) 107. Mod. Dice-Leraas diagram of IML/CL variation compared with that of A. occultus and Antrodiaetus pacilicus. 108. Mod. Dice-Leraas diagram of IFL/lTarL variation compared with that of A. occultus and A. pacificus. Systematics and Biology of Antrodiaetus Coijle 359 Eugene and Corvallis, Oregon. Unfortu- nately, the habitat of A. occuJtus is un- kno\\'n. Of 43 quantitative characters examined, the sympatric population sample of A. pugrwx is markedly more distinct from A. occuJtus than are the allopatric A. puii,na.x population samples — the pattern that could be the result of character dis- placement— in 20 of these characters, whereas only five characters show the opposite pattern. IML/CL, IFL/ITL, CW/CL, and IFL/ITarL are the best examples of the former pattern (Figs. 103, 107, 108); ITL/IML yields the most strongly expressed reverse pattern. Noteworthy variation, usuallv with a geographic pattern similar to the domi- nant one shown by quantitati\'e characters, arc shown by a number of not strictly quantitatively defined characters. Moles: Three specimens from B and C have only a tiny area devoid of setae on the upper ectal surface of the chelicerae, while the rest in the species sample have larger setaless areas, the largest being from the westernmost localities. There is consider- able variation (but without a clear geo- graphic pattern) in the shape of the i.c.s. tip and in the distance that it extends beyond the o.c.s. tip (Figs. 207, 209, 211). The o.c.s. tip is also quite variable in shape and averages wider and shorter in sample H (Figs. 208, 210, 212). The basal portion of the palpus is larger relative to the length of the conductor portion in most south- western specimens (Figs. 207, 209). While specimens from the eastern part of the species range have either only two or three noncontinuous, heavily sclerotized patches on their abdominal dorsum, western specimens have three usually continuous patches. Females: Most specimens from B have the upper ectal surface of their chelicerae evenly covered with setae, the rest of the eastern specimens have here a small to medium size setaless area, and in western specimens this area is large. The seminal receptacle stalks are usually pro- portionately much longer and the stalk bases less expanded in eastern than in western specimens (Figs. 292-294). Oc- casionally a part of the anterior edge of the bursa copulatrix floor is very heavily scle- rotized (Fig. 294). Distrihution. Northern Oregon, southern Washington, into northwestern Idaho (Map 2). Records. IDAHO. Latah Co.: Troy [A], 16 Aug. 1908, S. OREGON. Benton Co.: Corvallis [G], 18 Aug. 1949, 30J 18 012 cerberus 5 5 2-5 26 4 18 yesoensis 2 4 4 6-5 85 1 3 3 73-4 900 07 1 1 06 060 0 0 42-0 465 51 0 0 27-0 300 33 0 0 16-0 165 17 lincolnianus 6 8 8 4 5 4 4 4 5 6-5 06± 0-5 6 4': 6-5 00' 7 40 0 36 4 37 3 4 92-4 263 + 86 342 0 0 0 0 0 0 82-1 905 + 79-0 886' 83-1 931' 01 065 96 058 07 086 0 0 0 0 0 0 40-0 455i 40-0 435. 42-0 4 76' 53 043 47 024 56 053 0 0 0 0 0 0 21-0 230i 23-0 259i 24-0 266i 25 013 29 019 30 021 0 0 0 0 0 0 18-0 205 + 18-0 192i 14-0 195. 22 016 stygius ^3 (^ 3 4 24-4 606t 268j 22 014 apachecus 73-4 0 54 J 67 322 26 043 roretzi 2 5 5 1-5 19 3 4 4 18-4 350 52 1 1 01-1 020 03 0 0 48-0 505 5 3 0 0 29-0 295 30 0 0 13-0 135 14 AMD IFL ITL IML ITarL PFL PTL unicolor 0 11-0 20 3 84-6 40 2 56-4 22 3 01-5 35 1 80-2 78 2 33-3 92 2 14-3 50 0 158' 017 5 097 + 466 3 4381 394 4 280' 556 2 314: 219 3 215' 384 2 8781 368 robustus 0 14-0 18 5 05-6 06 3 31-3 95 3 99-4 93 2 33-2 78 3 27-3 92 2 94-3 43 0 162- 013 5 5371 321 3 6251 221 4 5 391 312 2 547-' 159 3 665- 222 3 2161 162 pacif icus 0 11-0 18 3 84-6 25 2 48-4 22 3 16-5 80 1 69-3 05 2 78-4 52 2 41-3 92 0 136 + 024 5 3 1091 99-5 586 73 3 2 2 48 1 63-3 407 69 4 3 5401 09-4 568 22 2 1 356 + 88-2 300 41 3 2 7091 78-3 407 88 3 2 1641 45-3 333 occultus f° 13-0 Ti^N 39 0 158 + 11-0 oisj 14 4 801± 436 3 133* 278 3 2 671* 86-4 331 07 2 1 153 ■ 69-2 178 75 3 2 297' 60-3 265 77 2 2 8821 33-3 250 pugnax (3 77-5 3^^ (' 56-3 54\ 20 0 0 1201 11-0 008 16 Vi 584± 421) b 113 + _28lJ 3 4 561 ' 18-5 34 7 57 2 2 2351 56-3 247 46 3 3 0651 54-4 331 67 2 798 + 240 montanus f5 31-6 TTx h 65-4 "A f3 05-4 03\ 0 0 1331; 11-0 012 12 11 5 998* 20-5 464J 65 U 1401 315; 4 3 732 ' 92-4 396 44 2 2 912 + 37-2 228 78 3 3 9991 61-4 294 07 ll 4 521 262] 3 58-3 95 hageni 1^ 27-3 54\ 0 1191 004 5 4 484 + 82-5 179 27 3 3 779' 24-3 156 61 4 3 221t 46-3 172 92 2 1 616- 96-2 159 18 3 3 837 + 09-3 139 24 il 2 377 + 56-2 107J cerberus (° 09^ 75 0 090j 11 5 4 010 26-4 56 3 3 360 01-3 20 3 605 2 032 3 2 135 71-3 01 2 607 yesoensis (^ 16-3 2 4^ fl 92-1 96) (^ 48-2 ~56) 0 110 4 4 410 37-5 .69 3 3 105 09-4 07 b 200 .07 u 940 07 2 860 li 520 __J 3 09-4 1 80-2 lincolnianus f° 05-0 .07A (3 77-4 "^ (3 39-4 26\ 0 063± 09-0 .008j .12 4 4 948' 10-5 431 .31 3 2 522- 78-3 327 .58 3 2 4651 97-3 334 95 1 1 925 + 80-2 089 41 u 2341 165J {i 774' 123J stygius 3 31-3 95 2 86-3 31 0 102± .010 4 .6601 391 3 .1571 .246 3 5201 303 2 107< 202 3 585' 215 3 050 + 163 apachecus 0 .09-0 .11 4 .41-5 .84 3 .09-3 .92 3 16-4 .33 1 80-2 48 3 43-4 44 2 82-3 61 0 .102 + .009 4 .9151 .478 3 .3421 .285 3 561 + .446 2 0391 239 3 6801 331 3 0251 250 roretzi 0 .13-0 .14 5 .08-5 .12 3 .20-3 .27 4 22-4 .37 2 11 3 61 2 78 0 .135 5 .100 3 .235 4 .295 2 .110 3 610 2 780 Systematics and Biology of Antrodiaetus • Coyle 375 Table 13 (continued) . PTT EGS AMD/CL AMD/AMS ITL/CL IML/CL IFL/IML unicolor robustus pacif icus occultus pugnax montanus hageni cerberus yesoensis lincolnianus stygius apachecus roretzi 0.98-1.58 1.291±.112 1.25-1.52 1.388±.081 1.01-1.56 1.273±.112 12-51 25.7 f40-6i\ 152.9 j 17-56 33.8 (0.022-0.035 lo.029± .003 0.024- 0.027^ 0.018- 0.024* 0.032 .002 0.030 .004 0.67-2.42 1.251±.309 0.73-1.54 1.100±.228 0.50-2.00 1.0161.316 0.54-0.69 0.6281.028 0.57-0.65 0.601+.022 0.51-0.70 0.578±.035 0.67-0.88 0.7801.043 0.72-0.83 0.7531.035 0.73-0.91 0.8081.042 1.14-1.30 1.193+ .041 1.16-1.26 1.221+.031 fl. 03-1. 22^ l4.i281.O44j (^l.ie-l.sA 10-40 ('0.026-0.033^, f0.81-1.50A fb.57-0.62"\ f0.64-0.76\ (1.25-1.38^ V,1.3491.103j 22.5 1,0.0301 . 002 j (,1 . 0951 . I62J ip.5891.018j ^0.6911.027] Ij. . 3081 . 037] 1.03-1.32 1.1651.085 1.20-1.45 1.3081 .083 1.23-1.41 1.3261.063 17-44 31.1 31-62 45.5 29-59 42.0 0.021- 0.025J 0.020- 0.023i 0.029 .002 0.026 .002 0.019- 0.020J 0.02 .00 0.60-1. 0.8831. 0.59-1. 0.7701. 0.67-0. 0.7451. 18 156 09 120 80 042 0.57-0.68 0.6351.027 fb. 67-0. 77"^ V0^_7J^2t^019J ('0.64-0.66^ fO. 70-0. 74^ 1^0. 6421. 013J lp.718±.02lj 0.64-0.81 0.7271. 04£ 0.77-0.87 0.8141.024 1.16-1. 1.2901. 1.19-1. 1.2681. 1.26-1. 1.3001. 42 083 34 036 32 020 /'l. 01-1. 09'] 32-46 ro. 016-0. 017'\ fo.50-0.60'\ U-O'iO J 41.0 (0.017 J [o.556 J 1.12-1.20 1.160 fo. 81-1.0 A Vp.8971.038j 0.90-1.05 0.9821.058 0.94-1.07 0.9801.048 1.09 1.090 30-39 34.5 22-39 30.5 0.024- 0.026 0.026 1.00 1.000 0.61-0.69 0.638 0.62-0.66 0.640 0.67-0.71 0.6951.018 0.64-0.72 0.6801.027 0.63-0.72 0.6681.030 0.62-0.63 0.623 0.65-0.72 0.683 (t 1.34-1 1.391 1.35-1. 1.378 /'0.63-0.69'N I0.66I J Co.67-0.7lA fl. 38-1. 48'^ lp.6831.020J U.4291.036J 0.71-0.79 0.7581.028 0.66-0.80 0.7101.042 1.29-1. 1.3251. 1.27-1. 1.3851. 38 029 IFL/ITarL ITL/IML ITL/ITarL IML/ITarL PFL/CL PTT/CL PTT/PTL unicolor robustus pacif icus occultus pugnax montanus hageni cerberus yesoensis lincolnianus stygius apachecus roretzi 1.96-2.43 2. 2001 .114 2.11-2.33 2.1761 .08£ 1.84-2.46 2.1741 .119 2.07-2.42 2. 2311 .115 1.86-2.24 2.0591.116 1.98-2.20 2.0611.061 2.02-2.20 2.1001.071 0.74-0.85^ .0.8051 .027J 0.77-0.83 0.7991.019 Co765'-0.8l'A \j3.717l.04lj (^.81-0. 90"^ lj).854l .O24J 1.35-1. 1.4841 . 1.36-1. 1.4241 . 1.26-1. 1.3811 . 1.38-1. 1.4561 . 62 068 53 052 53 069 55 063 1.60-2.06 1.8451.111 0.53-0 0.586 0763\ 1 .022) 1.71-1.87 0.58-0.65 1.7821.050 0.6081.022 /TTTTmPA /'0.6I-0T7T U.9291.106J (,0.661+ .02 1.60-1.84 0.59-0.65 1.7061.059 0.6201.019 0.20-0.27 0.2371 .014 0.21-0.24 0.2301 .00£ 0.20-0.25 0.2271.014 0.38-0.54 0.452t .038 0.40-0.47 0.4321.021 0.33-0.48 0.4041 .026 ('0.23-0.27^ fC lo.2541.010j ic /O. 45-0. 49 \ i0.468l.016J 0.82-0.91 0.8751 .029 f0.83-0.9lA i0.875i.018j )^.88-0.90\ Vp.895i.013j /l.31-1.60\ /'l.48-1.74\ 0.57-0.7 U. 3971. 053) [1.5971. 064j 0.625+.0 2 036 1.-36-1. 1.4221, 1.39-1. 1.4461, 50 032 51 043 1.56-1.76 /'0.63-0.73\ 1.6251.046 ij).688i .020j 1.56-1.70 0.63-0.68 1.6161.045 0.6531.021 0.22-0, 0.2381, 0.20-0. 0.2261, 0.21-0, 0.2251. 25 008 25 013 23 007 ('2.42-2.5 2.466 2.22-2.32 2.273 ^ (^T92^^o793\ (1.65-1.66") _J lo^^932__J lu_653____j fb. 95-0. 99\ f: Lo.970 ; t; 1.76-1.80 1.773 0.58-0.60 0.595 /0.19-0.20"\ Lo.l97 ) 0.37-0.44 0.4171 .013 0.35-0.40 0.3801.014 0.37-0.41 0.3931.011 0.39-0.41 0.399 1.57-1, 1.600 fl. 64-1.65') 0. lu_649__J 0. 58-0.59 589 0.23-0 0.239 .24 '2.43-2.75^ /'l.00-1.04"\ ,2.5671.107^ lj..017i.018J 1^2. 12-2. 28\ /'0.87-O.94A (1 1,2. 2141. 061J 1^.8981.022J 1^ 0 2.29-2.54 N 2.4161.074J 2.39-2.43 2.417 0.89-0.99 0.9431.044 1.71-1.96 0.81-0.87 1.7971.094 0.8371.021 1.42-1."54A 1.60-1.75 0.71-0.85 1.5011.04 2; 1.6721.052 0.774i.041 1.64-1.86 0.66-0.82 1.7461.066 0.7371.053 (q.11-0.1H\ (^T^3^^o724~N lp.l78i.005j lj).238l.003j ('oTTiToTJrA 1,0.3221.011; 1.52-1. 1.646+ . 72 007 0.20-0. 0.2121. 0.18-0 0.1961 22 007 ,21 ,011 0.29-0.34 ,0.3251.015, ■0.75-0.76\ vg.753 _j 1.51-1. 1.533 55 e 2.00-2.07 2.036 0.68-0.71 0.695 0.20-0 0.210 .22 0.39 0.392 376 Bulletin Museum of Comparative Zoology, Vol. 141, No. 6 Table 14. Measurements, meristic characters, and diagnostic ratios for adult females of Antro- diaetus species. The most useful diagnostic characters are circled. Range, mean, and standard devia- tion given for measurement characters (in mm) and ratios. Hange, mean, and mode given for meristic characters. Number of females containing large maturing eggs or with brood is given in parentheses after N. For the last 7 species the CT and CMT statistics include counts of both right and left chelicerae . N CL CW OQW ALS ALD AMS AMD unicolor robustus pacif icus pugnax montanus hageni Cerberus yesoensis lincolnianus stygius apachecus roretzi 225(158) 5.4-10.7 7.38- .95 13(5) 24(13) 12(5) 7(2) 4(2) 1(0) 2(1) 3(3) 3(0) 2(0) 6.2-9.3 7. 26-' .99 56(24) 5.4-8.1 6.74i.59 /'4.2-5.9A [5. 02 J .5lJ /6.9-9 9.2>, 6 5l 6.5-8.3 7.541.63 6.6-7.5 6.99 5.5 6.7-6.9 6.84 5.9-6.9 6.34 5.9-6.! 6.37 6.6-6.9 6.77 3.88-8, 5.434r , 4.71-7. 5.577; . 4.07-6. 5.160i, 3.20-4. 3.848; . 5.50-7. 6.589; . 5.01-6. 5.783^ . 5.16-5, 5.472 4.41 5.35-5. 5.540 4.59-5. 4.847 4.86-5. 5.097 5.50-5. 5.650 06 758 23 810 37 464 71 413 87 703 59 58£ 84 73 20 27 80 1.08-2. 1.4 47; . 1.33-1. 1.506;. 1.06-1. 1.321; . 0.87-1. 1.057; . 1.34-1. 1 . 5 3 2 • . 1.20-1. 1.386; . 1.39-1 . 1.400 1.12 1.32-1. 1.330 1.23-1. 1.303 1.20-1. 1.287 1.29-1. 1.295 02 181 83 180 69 114 29 121 83 137 58 126 41 34 45 41 30 0.45-0. 0.666* . 0.63-0. 0.780;. 0.47-0. 0.621; . 0.38-0. 0.494;. 0.67-1. 0.827;. 0.54-0. 0.659;. 0.64-0. 0.677 0.57 0.61-0. 0.635 0.56-0. 0.630 0.45-0. 0.590 0.72-0. 0.745 93 083 95 097 77 071 61 070 02 091 73 065 71 66 68 68 77 0.27-0. 0.381< . 0.29-0. 0.374;. 0.26-0. 0.355;. 0.23-0. 0.309±. 0.33-0. 0.387;. 0.34-0, 0.396; , 0.33-0. 0.340 0.31 0 0 29-0. 310 0 0 24-0. 320 0 0 32-0. 343 0 0 31-0. 315 57 054 45 063 47 043 36 031 45 042 47 041 36 33 38 37 32 0.11-0. 0. 198 ■ , 0.18-0. 0.228- . 0.12-0. 0.184; . 0.11-0. 0.146;. 0. 19-0. 0.295-. 0.19-0. 0.214- . 0.19-0. 0.202 0.20 0.24-0. 0.245 0. 19-0. 0.220 0. 14-0. 0.210 0.21 0.210 34 036 31 035 29 034 21 029 40 055 25 026 22 25 25 25 0.11-0.22 0.159;. 020 0.14-0.20 0.160;. 020 0.11-0.18 0.131;. 018 /b.09-0.14'~^ l,0.112;.012j 0.12-0.18 0.157;. 017 0.12-0.16 0.140i.017 0.11-0.12) 0.117 0.11 .09-0.11^ .100 J 0.11-0.12 0.117 0.11-0.12 0.117 0.14-0.15 0.145 SL SW IFL ITL IML ITarL IVFL IVTL unicolor 2 4 94-6 084- 03 587 2 3 45-5 444; 08 501 3 5 84-7 238; 69 672 2 3 29-4 191' 71 414 2 3 11-5 236; 08 423 13-2 629; 26 193 3 5 61-7 085- .65 .643 2 2 10-4 .970i .37 .362 robustus 3 3 3 3 2 2 3 4 3 4 3 4 27-5 980; 01-4 903' 37-3 87 8; 84-5 727-. 92-5 479; 92-4 085 08 582 90 395 54 312 .57 470 05 .393 29 3 3 2 3 2 2 3 4 3 3 3 3 01-4 532- 60-3 165- 07-2 461; 39-5 190- 20-3 551- 24-3 425 44 481 99 298 97 261 05 487 84 215 61 4 5 4 5 2 3 5 6 4 5 5 5 29-6 065' 07-6 094 • 97-4 625; 20-7 143- 82-6 614; 12-5 367 37 664 22 469 37 406 31 636 22 447 73 2 2 2 3 1 2 3 3 2 3 3 3 52-3 969; 41-3 104- 77-2 199; 16-4 804* 97-3 439; 01-3 190 84 427 69 301 63 254 52 401 84 308 39 -1 2 2 3 1 2 3 3 3 3 2 3 45-3 934; 48-3 150: .73-2 144- 24-4 938; 05-3 461- 94-3 105 77 433 88 304 56 262 .71 443 92 290 27 28-1 512; .88 179 4 4 3 4 2 3 5 5 4 5 4 5 18-6 875. 69-5 609* 86-4 470* 12-7 93 3: 75-6 537; 86-5 077 .14 629 76 392 22 366 01 564 18 474 35 -1 A. 2 2 2 1 2 33-3 763; 07-3 612. • 62-2 030; 50 361 pacif icus ri 28-1 604- .96^ 160J 16 216 pugnax 90-1 084- 28 N 116j .48 199 54-2 7 59; 07 154 montanus (3 05-4 486; 07 \ .301) 2 3 2 3 82-3 229; 90-3 012 hageni 47-1 677; 84\ I25J 61 293 Cerberus 50-1 577 65 16 yesoensis 3 24 2 71 3 99 2 37 2 33 13 3 88 2 18 lincolnianus 3 3 84-4 955 07 3 3 20-3 370 54 4 4 82-5 935 05 2 2 86-3 975 09 2 2 78-2 840 90 31-1 330 35 4 4 63-4 650 67 2 2 63-2 670 71 stygius 3 3 39-3 563 84 3 3 09-3 213 46 4 4 07-4 307 71 2 2 45-2 597 86 2 2 22-2 370 63 09-1 153 28 3 4 88-4 117 52 2 2 18-2 307 48 apachecus 3 3 61-4 .813 .10 3 3 24-3 300 .39 4 4 22-4 683 97 2 2 60-3 873 05 2 2 45-2 687 86 28-1 343 47 3 4 88-4 193 41 2 2 26-2 360 41 roretzi 3 3 .84-4 .955 .07 3 3 46-3 .535 .61 5 5 08-5 250 .42 2 3 94-3 050 16 2 3 97-3 065 16 43-1 485 54 4 4 82-5 950 08 2 2 48-2 595 71 Systematics and Biology of Antrodiaetvs • Coyle 377 Table 14 (continued) . IVML IVTarL CT CMT IVCTR PTSP PTSR IMS IVMT unicolor robustus pacif icus pugnax montanus liageni Cerberus yesoensis lincolnianus stygius apachecus roretzi 3.00-6.33 4.385: .527 3.35-4.86 3.942i .470 3.05-4.82 3.9181.348 2.56-3.69 3.018: .290 fi.52-S.99^\ ^5.201r.456j 3.99-5.24 4.810: .376 4.26-4.71 4.510 3.16 3.84-3.92 3.880 3.05-3.65 3.263 3.09-3.46 3.327 3.80-4.03 3.915 1.28-2. 1.664-. 1.31-1, 1.505- . 1.28-1. 1.531'. 1.01-1. 1.220: . 1.65-2, 1.789: , 1.47-1, 1.710- , 1.47-1, 1.512 1.24 1.47-1, 1.485 1.20-1, 1.250 1.35-1, 1.363 1.43-1, 1.465 03 163 73 125 80 114 43 109 11 135 84 139 50 35 39 50 8-15 11.4 (11) 11-14 12.1(11) 9-14 10.9 (11) 9-14 11.3(11) 10-17 12.4(12) 11-12 12.0(12) 11-13 12.0 11 12-13 12.5 12-14 13.0 12-13 12.3 11 11.0 6-23 12.0(12) 11-18 13.9(13) 9-24 16.8(19) 8-17 12.1(12)1 G 16-29 21.9(23), 13-1 15.1(15) 12-19 10-16 13.0 .6.01 1-6 3.5(3) 2-5 3.6(3) 1-6 4.0(4) 4-6 4.2(4) 3-5 3.9(4) 2-5 4.0(4) 4-5 4.5 0-2 1.3 0-2 1.0 4 4.0 5-10 6.9(7) 4-7 5.9(6) 5-12 8.2(8) 4-8 5.9(5) 6-9 7.5(7) 5-11 7.4(6) 6-8 6.7 4-8 6.0 4-7 5.3 2-7 3.8(4) 3-4 3.5(3) 3-5 3.9(4) 2-5 3.2(3) 2-4 3.4(4) 3-7 4.3(3) 3 3.0 3-4 3.5 2-4 3.0 9-15 11.6(11) 9-14 10.6 (10) 9-15 12.3(13) 9-13 11.1(11) 12-16 12.9(13) 13-15 13.7(13) 11-13 11.7 10 13 13.0 11-13 12.0 11-12 11.3 15 15.0 9-20 14.2(15) 10-17 12.9(13) 9-21 15.1(14) 8-15 12.2(12) 14-19 >, 16.5(17)] 11-17 15.0(15) 12-18 15.0 veloj 8-12 10.0 10-11 10.3 15 15.0 CW/CL AMD/CL ALS/AID SL/SVJ IFL/CL •I ML/ :l IVMCR 0 0 0 0 0 0 0 0 0 0 0 0 016-0 022: 021-0 022- 015-0 020 + .018-0 022: 016-0 019: 017-0 019' 025 002 024 001 025 003 026 002 023 002 020 001 1.04-2 1.77 6: 1.61-2 2.124: 1.37-2 1.768: 1.26-2 1.607i 1.67-2 2.157- 55 296 75 330 65 .240 04 232 63 302 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 63-0.76 710- .023 68-0.72 698+ .014 71-0.80 756: .024 68-0.77 721+0.28 70-0.79 754- .027 73-0.77 745- .017 76-0.79 768 71 71-0.73 721 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 35-0 439- 38-0 404- 42-0 467- 39-0 426: 44-0 4 83- 43-0 4 59 43-0 444 41 41-0 415 .48 .023 .42 .012 .51 .020 .48 .021 52 022 48 018 46 42 unicolor (o 0 69-0 735: 79\ 020) 13-1 187 + .27\ 036 6-10 7.7(8) 0 0 0 0 0 0 0 0 0 0 0 0 0 73-0 767: 71-0 766: 72-0 .765: 77-0 808: 72-0 766: 77-0 782 79 79 015 79 021 80 023 85 024 79 031 80 robustus pacif icus pugnax ' 08-1 126- 13-1 2 34: 10-1 169: .ie\ 021J 35 052 24 031 6-9 7.3(7) 6-9 7.1(7) 6-9 7.5(7) montanus d 06-1 130- 21-1 260: 0 33J 37\ 056j 6-8 7.3(7) hageni fl,55-l (1.669: 1.77-2 1.997 1.83 1.85-2 2.062 1.64-2 2.048 1.22-2 1.739 90\l 130J ( 15 27 70 00 6-7 6.6(7) 15-1 193 19 15-1 175 22 20 Cerberus (0 0 016-0 017 017\ 7-9 8.0 yesoensis 020 6 lincolnianus ro 79-0 809 T2\ '0 0 0 0 0 013-0 015 017-0 018 016-0 .018 016) 020 020 8-9 8.5 stygius [I 76-0 764 '3 ri 09-1 109 14-1 155 10-1 119 3 21 13 0 0 0 0 66-0.69' 679 J (0 35-0 373 3 44 46 6-9 8.0 apachecus 0 0 .78-0 .801 82 71-0.76 735 76-0.78 775 0 0 0 0 40-0 422 45-0 453 6-9 7.7 roretzi .83-0 .834 3 '0 0 .021-0 .021 022^ f2.25-2 12.367 3 6-7 6.5 378 Bulletin Museum of Comparative Zoology, Vol. 141, No. 6 Table 14 (concluded) . IVTL/CL IVML/CL IFL/IML IFL/IVFL IVTL/IVTarL IVML/IVTarL IVML/IML unicolor 0.34-0 0.403:: 0.37-0 0.381t 46 022 39 008 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 49-0 596-. 52-0 554-- 51-0 582: 56-0 6 01- 60-0 639- 61-0 638- 62-0 645 56 56 567 68 040 57 015 65 033 65 026 65 016 65 015 67 1.48-1 1.620- 81 059 0.98-1 1.030--- 1.01-1 1.0391 11 026 05 Oil 1.58-1 1.78 3- 1.77-2 1.830- 1.60-1 1.706- 92 093 02 104 05 072 2.33-3 2.632- 2.47-2 2.613- 2.38-2 2.558- 06 142 81 107 79 119 16-1 358- 28-1 34 9- 57 063 robustus fl.65-1 U.730- 1.50-1 1.618- 78\ 034^ 69 050 41 040 pacif icus ro.35-0 ^0.388- 0.37-0 0.404- 0.40-0 0.429- 0.41-0 0.428- 4A 015j 43 017 48 021 43 088 A. 04-1 tl.l05- 036j a 09-1 247i 066j 1.00-1 1.044- 1.00-1 1.035- 10 021 07 018 pugnax /'1.60-1 74^ 037J ^1.52-1 J..66 3- 1.58-2 1.953- 1.81-2 1.889- 74\ 054) 16 156 08 104 A. 26-2 V2.474- ToX lOlJ G 31-1 413- "ioX _059) 2.40-3 2.912'- 2.59-2 2.814- 18 224 92 070 (i 24-1 324-- 31-1 391--- montanus 1.48-1 U.56 2 1.58-1 1.623- 61\ 041) 68 035 39 046 hageni A. 00-1 Ij^. 0 1 4 - 1.04-1 1.057 1.02 1.04-1 1.061 03N OllJ 07 .08 46 060 Cerberus '0.42-0 0.39 0.38-0 0.390 0.35-0 0.364 0.35-0 0.371 0.37-0 0.383 "44'\ 39 38 45 39 /1.70-1 (J..728 1.71 1.73-1 1.738 7 5' 74 A.93-2 y..992 10\ A.84-3 1^2.983 ") 44-1 453 35 35-1 367 36-1 376 TtN yesoensis lincolnianus 1.75 1.75-1 1.798 84 2.54 2.56-2 2.613 2.54-2 2.607 66 70 38 stygius (s 48-0 514 T3\ 54 .58 /'1.79-1 .1.819 1.70-1 1.743 1.71 1.713 .8 6\ 80 A. 04-1 IJ..046 0 5~] 1.81-1 1.846 1.67-1 1.731 1.73-1 1.770 8 8'\ .78 81 38 0 0 0 0 50-0 522 57-0 578 apachecus fl.09-1 J.. 11 6 .„) ^2.29-2 L2^^4£0_ -!!) 21-1 239 ") roretzi 1.05-1 1.060 07 2.66-2 2.672 68 27 277 Systematics and Biology of Antrodia ETUS Coijle 379 Table 15. Measurements (in mm) and meristic character values for holotype ANT) LECTOTYPE SPECIMENS OF AnTRODIAETUS SPECIES. Leg IV segments distal of femur missing on A. rohusius lectotype. Pedipalp segments flattened out of shape on A. hncolmanus holotype * MM is metarsus I macrosetae pattern; TP is number of macrosetae (total, ensifonii) on prolateral surface of tibia I; and TRV is number of macrosetae (total, ensiform) on retrolateral aspect of ventral surface of tibia I. robustus pacificus OCCllltUS 77W71tmiUS hageni Cerberus lincolniatius stygius apachecus CL 6.95 5.34 5.57 5.27 6.17 5.19 5.12 4.66 4 66 CW 5.48 4.33 4.56 4.48 5.16 4.22 4.22 3.54 3 73 OQW 1.44 1.03 1.12 1.03 1.08 0.93 0.90 0.91 0 86 ALS 0.75 0.51 0.51 0.57 0.57 0.47 0.44 0.45 0 43 ALD 0.,35 0.31 0.35 0.29 0.33 0.25 0.23 0.25 0.25 AMS 0.21 0.14 0.14 0.18 0.18 0.16 0.22 0.20 0 17 AMD SL 0.17 3.78 0.14 0.16 0.14 0.12 0.09 0.07 0.11 0.09 SW 3.44 IFL 4.88 4.71 4.97 5.46 5.61 5.05 4.90 4.63 4.41 ITL 2.80 3.01 3.31 3.77 3.95 3..35 3.58 3.20 3.09 IML 2.83 4.22 3.84 4.29 4.37 3.58 3.54 3.50 3 16 ITarL 1.40 1.96 2.33 2.63 2.71 2.03 1.92 2.18 1.80 IVFL 4.69 PFL 3.54 3.39 3.65 3.92 3.12 3.61 3.54 PTL 2.94 3.01 3.12 3.54 2.56 3.09 2.94 PTT 1.23 1.38 1.20 1.40 1.01 0.98 0 96 EGS CT 12 36 25 59 46 31 69 GMT 12 PTSP 6 PTSR 3 IMS 10 *MM none none A AB AB AB A A *TP 38,0 35,16 35,27 44,44 17,16 *TRV 6,0 4,2 18,13 12,12 6,6 380 Bulletin Museum of Comparative Zoology, Vol. 141, No. 6 Maps 1-4. Distribution of Antrodiaetus species. Systematics and Biology of Antrodiaetus • Coyle 381 Figures 109-119. Whole body and carapace views of Antrodiaetus. (Dorsal and lateral views.) 109-112. A. pugnax, B. 109. Male. 110. Female. 111. Mole. 112. Female. 113. A. unicolor ma\e, I. 114-116. A. roretzi. 114. Male. 115- 116. Female. 117-118. ,4. stygius paratype female. 119. A. lincolnianus female; Lincoln, Nebr. 382 Bulletin Museum of Comparative Zoology, Vol. 141, No. 6 120 121 122 m 00 125 (/ ;'0; : O: 127 >s- 0^"l23^0 128 124 '^ 0^ 00^ 129 130 134 , 0.6 mm 131 135 132 ^^36^^ 00,33-00 o 0 0 137 o Figures 120-137. Antrodiaetus eyes. (Dorsal view with lateral border of carapace horizontal.) 120-129. Males. 120. A. unicolor, P. 121-122. A. pacificus. 121. Holotype. 122. R. 123. A. occullus, paratype. 124. A. pugnax, A [Brachyboth- rium shoshoneum Chamb., holotype). 125. A. hageni; near Baker, Ore. 126. A. cerberus, holotype. 127. A. lincolnianus, holotype. 128. A. stygius, holotype. 129. A. roretzi, Soitama Pref. 130-137. Females. 130. A. unicolor, P. 131. A. pacificus, L. 132. A. pugnax, B. 133. A. hageni; Trail, B. C. 134. A. cerberus, paratype. 135. A. /inco/nionus; Lincoln, Nebr. 136. A. stygius, paratype. 137. A. roretzi. Figures 138-144. Sternum end labium of Antrodlaetus females. 138. A. unicolor. Q. 139. A. robusfus; Canton, Ohio. 140. A. mon tonus; Salt L. City, Utah. 141. A. hageni; Kelowna, B.C. 142. A. yesoensis. 143. A. styg/us, paratype. 144. A. roretzi, Tochigi Pref. Systematics and Biology of Antrodiaetus • Coijle 383 \. 1 ■ ■ , , ,' ' ' ' 'I, ■ Figures 145-157. Chelicerae of Antrodiaetus males. (Retrolateral view of left chelicera.) 145-146. A. unicolor. 145. F. 146. B. 147-149. -A. pacificus. 147. 6. 148. F. 149. D. 150. -A. montanus; Idaho Falls, Ida. 151. A. hageni; Baker, Ore. 152. A. lincolnianus, paratype. 153. A. stygius, holotype. 154-156. A. apachecus. 154-155. Near Alpine, Ariz. 156. Paratype. 157. A. roretzi, Yokohama. Figures 158-169. Chelicercl teeth of Antrodiaetus females. (Ventral view of left chelicera.) 158. A. unicolor, J. 159. A. robusfus; Hagerstown, Md. 160. A. pacilicus, L. 161. A. pugnox, 8. 162. A. monfonus; Downey, Ida. 163. A. hageni; Kelowna, B. C. 164. A. cerberus, paratype. 165. A. yesoensis. 166. A. lincoln- ianus; Lincoln, Nebr. 167. A. stygius, paratype. 168. A. opochecus; Flagstaff, Ariz. 169. A. roretzi; Tochigi Pref. Figures 170-172. Setation of upper ectal surface of Anfrod/oefus female chelicerae. (Dorsal view.) 170. A. pacificus, L. 171. A. montanus; Downey, Ida. 172. A. hageni; Trail, B. C. 384 Bulletin Museum of Comparative Zoology, Vol. 141, No. 6 Figures 173-187. Anirodiaetus male pedipalps. (Retrolateral view of left pedipalp.) 173-174. A. unkolor. 173. Q. 174. N. 175-177. A. paciUcus. 175. D. 176. B. 177. J. 178. A. occultus, paratype. 179. A. pugnax, A {Brachybothrium sho- sboneum Chamb. holotype). 180. A. monianui; Idaho Foils, Ida. 181. -4. hageni; Summerland, B. C. 182. A. cerberus, holotype. 183. A. yesoensis. 184. A. linco/n/onus; Lawrence, Kan. 185. A. stygius, holotype. 186. A. opochecus, holo- type. 187. A. roretzi, Saitama Pref. Systematics and Biology of Antrodiaetus • Cotjlc 385 o.c. s. EMBOLUS Figures 188-208. Antrodiaetus palpi. (Prolateral view of entire left palpus and view of tip after palpus rotated 90° on longi- tudinal axis of distal half of i.c.s. Occasionally prolateral view does not include bulb portion of palpus.) 188-194. A. uni- co/or. 188-189. Q. 190-191. 6. 192. N. 193-194. A. 195-196. /A. robustus; near Washington, D. C. (6roc/iybofhr/um shoemaken Petrunk., lectotype). 197-203. A. pacificus. 197-198. D. 199. Holotype. 200. Ketchikan, Alaska. 201-202. S. 203. W. 204-206. A. occu/fus. 204-205. Holotype. 206. Yoncallo, Ore. 207-208. A. pugnax, A [Brachybothnum shosho- neum Chamb. holotype). 386 Bulletin Museum of Comparative Zoology, Vol. 141, No. 6 Figures 209-233. Anfrodiaetus palpi. (Same views as previous plate. Figs. 225, 227-228, and 230-232 enlarged; match 0.2 mm scale.) 209-212. A. pugnax. 209-210. H. 211. C. 212. E. 213-217. A. montanus. 213-214. Paratype; Notus, Ida. 215-216. Fish Lake, Ore. 217. Washoe Valley, Nev. 218-219. A. hogeni, holotype. 220-221. /^. cerfaerus, holotype. 222- 223. A. yesoens/s. 224-225. A. lincolnianus, paratype. 226-228. A. stygius. 226-227. Holotype. 228. Paratype. 229-232. A. opoc/iecus. 229-230. Paratype. 231. Near La Cueva, N. Mex. 232. Near Alpine, Ariz. 233. A. roretzi, Saitama Pref. Systematics and Biology of Antrodiaetus • Coijle 387 234 / ///A^y ^':'^i.";-. .-^--C 1^^^^^^^^^^^^^^^'^?^^ ^^M^fyp^0f40.'ft^ ^^^^^M^^^^ Figures 234-247. Tibia, metatarsus, and tarsus of leg I of Antrodiaetus males. (Ventral and prolateral views of left leg.) 234-240. A. unicolor. 234-235. P. 236. F. 237. A. 238. /. 239. N. 240. O. 241-243. A. robusfus. 241-242. Near Rector, Pa. 243. Distal end of metatarsus (ventral view), near Rector. 244-247. A. paciUcus. 244-245. D. 246. T. 247. W. 388 Bulletin Museum of Comparative Zoology, Vol. 141, No. 6 Figures 248-257. Tibia, metatarsus, and tarsus of leg I of Antrodiaefus males. (Ventral and prolateral views of left leg.) 248-249. A. occu/tus, holotype. 250-252. A. pugnox. 250-251. A [Brachybolbrium shoshoneum Cfiamb. holotype). 252. H. 253-255. A. monfonus. 253-254. Holotype. 255. Paratype; Strawberry Reservoir, Ida. 256-257. A. hageni, fiolofype. Systematics and Biology of Antrodiaetus • Coijlc 389 2 mm J 264 ////l //■ ^^^^^^^^^^^^^ ^ \ \ ' 266 AjP^wT''^^ 2t: Figures 258-269. Tibia, metatarsus, and tarsus of leg I of Antrodiaetus males. (Ventral and prolateral views of left leg.) 258-259. k. Cerberus, holotype. 260-261. A. yesoens/s. 262-263. A. /mco/nionus, paratype. 264-265. A. sfygius, paratype. 266-267. A. apachecus, holotype. 268-269. A. roretzi; Yokohama. 390 Bulletin Museum of Comparative Zoology, Vol. 141, No. 6 270 271 272 273 $» BULB STALK J .■- \\ Vp i./ 283 284 ■\J ^-' U' 285 0.4 mm 286 287 288 289 iir Xfiji mic: %MP % 290 291 292 293 294 Figures 270-294. Antrodiaetus seminal receptacles. (Dorsal view. Floor of bursa copulatrix outlined by broken line.) 270- 279. A. unicolor. 270. J. 271. C. 272. K. 273. /. 274-275. Q. 276. B. 277. K. 278. F. 279. O. 280-283. A. ro- busius. 280. Baldwin, Pa. 281. Hagerstown, Md. 282. Canton, Oh. 283. Caledonia St. Pk., Pa. 284-291. A. pacificus. 284. Brookings, Ore. 285. /. 286. M. 287. Queen Charlotte Islands, B. C. 288-290. S. 291. Y (Kamiah). 292-294. A. pugnox. 292. Fort Lewis, Wash. 293. 8. 294. Lewis and Clark St. Pk., Wash. Systematics and Biology of Antrudiaetvs Coijle 391 0.4 mm 297 i:s-&v 299 300 301 " — - "''''304~~~^~~ """305"^^^"- 306 mm 313 7-7^:^ ••■/. - . ^ V- . ■*" :7 .:z^-. ''.' '■ ■•, ■ '^=::i: ^^1-^S ' .' r-;r:^i^=— - _:::_ -'/ ■ ■• V - ^^ — ' r-,-—. . - •—^ — ^" — ^^^ — C z'^- „.,. — -"^IT ;r^- - ^ JS^ :^^^^g t-* ^ •- 314 315 316 Figures 295-311. Antrodiaetus seminal receptacles. (Dorsal view. Floor of bursa copulatrix outlined by broken line.) 295- 298. A. monfonus. 295. Reno, Nev. 296. Pocotelio, Ida. 297. Salt L. City, Utah. 298. Downey, Ida. 299-300. A. hageni. 299. Kelowna, B. C. 300. Kamloops, B. C. 301-302. A. cerberus, paratypes. 303. A. yesoensis (likely immature). 304-305. A. lincolnianus; Lincoln, Nebr. 306-307. A. stygius; paratypes. 308-309. A. apochecus. 308. Flagstaff, Ariz. 309. Para- type. 310-311. A. roretzi. 311. Tochigi Pref. Figure 312. A. roretzi. Posterior ventral view of exterior surface of female genitalia. Figures 313-318. External morpfiology of female Antrodiaetus abdomen. 313-314. Setation medially and dor- sally just posterior of pedicel. 313. A. un/co/or, P. 314. A. robustus; Baldwin, Pa. 315-316. Dorsal abdominal background setation (1 mm post, of tergite). 315. A. un/co/or, P. 316. A. robustus; Baldwin, Pa. 317. A. robustus, AL spinneret vestige. 318. A. un/co/or spinnerets, O. 0.6 mm scale for 312-316. 392 Bulletin Museum of Coinparalive Zoology, Vol. 141, No. 6 0.6 mm 319 0.4 mm 329 330 ^^t^T^X^ Figures 319-330. Alialypus morphology. (Some views as for Anfrod/aefus.) 319-329. Al/ofypus no. 2; Chatsworth, Calif. 319. Male eyes. 320. Female eyes. 321. Female sternum. 322. Female cheliceral teeth. 323. Male chelicera. 324. Male pedipalp. 325. Palpus. 326-327. Distal segments of male leg I. 328. Female spinnerets. 329. Seminal receptacles. 330. Aliatypus sp.; Shaver Lake (Fresno Co.), Calif. Seminal receptacles. 319-320: 0.6 mm scale. 321: 2 mm scale. 322: 1 mm scale. 323-324, 328: 2 mm scale. 325: 0.4 mm scale. 326-327: 2 mm scale. 329-330: 0.5 mm scale. Systematics and Biology of Antrodiaetus • Coijle 393 EVOLUTION Phylogeny Atypoidea. Simon (1892) first empha- sized the close relationship of the Antro- diactidae, the Mecicobothriidae, and the Atypidae, but later (1903), impressed by the resemblance of the first two families to the ctenizids and diplurids, he separated these from the atypids and grouped them with the rest of the mygalomoq)h spiders. Smith (1908) studied Atypoides and AJi- atypus and sided \\\\h Simon's earlier opinion. However, until the 1940's, many authors accepted Simon's second classi- fication uncritically. Chamberlin and Ivie ( 1945 ) pointed out that the similarities of the antrodiaetids to the ctenizids and of the mecicobothriids to the diplurids were the result of convergence. Gertsch ( 1949, 1967) agreed, and grouped the Antro- diaetidae, Mecicobothriidae, and Atypidae together with the Liphistiidae to fonn the supcrfamily Atypoidea and to emphasize their close relationship and their distinct- ness from other mygalomorph taxa. Evidence strongly indicates that the antrodiaetids, mecicobothriids, and atypids do form a distinct and monophyletic group. The homogeneity of genital structure ^\ithin this group and the distinctness of the genitalia from those of nonatypoid taxa is perhaps the strongest indication of this monophyly; the genitalia are relatively complex structures probably not strongly affected by niche-oriented convergence. The always well-developed o.c.s. (Fig. 207) and the frequently present i.c.s. are absent from all nonatypoid palpi, and the Atypoidea possess four seminal receptacles (or a condition derived from this) rather than the usual two (or a derived con- dition) of other mygalomorphs. Other characters providing similar evidence are listed by Chamberlin and Ivie (1945). It is certain that the Antrodiaetidae are phylogenetically quite distinct from the ctenizids and that the similarity of their niches, together with a large number of functionally related (niche adapted) mor- phological and behavioral similarities, is simply the result of convergent, or at least parallel, evolution. The unique ancestral characters (ves- tiges of abdominal segmentation) shared by the antrodiaetids, mecicobothriids, and atypids indicate a long evolutionary history for this group. Each of these families is distinct and contains few species — these are truly evolutionary relicts. The family Atypidae is highly specialized moqiho- logically and behaviorally in accord with the unique tube-web niche which it has entered. The mecicobothriids apparently occupy a distinct sheet-web niche and are morphologically more similar to the antro- diaetids (not necessarily more closely re- lated) than to the atypids. Antrodiaetidae. The evidence summa- rized in Table 11 shows that Aliatypus arose from an ancestral stock distinct from that of Antrodiaetus and Atypoides, and also that Aliatypus may be as closely or more closely related to the Mecicobothrii- dae. Smith (1908) arrived at a somewhat similar conclusion. The first two char- acters (Table 11), w4iich are heavily weighted for reasons given earlier, indicate that Aliatypus is more closely related to the mecicobothriids. Characters 3 to 5 probably compose a complex of characters functionally related to mating behavior and therefore are indi\'idually less heavily weighted. These provide only weak evi- dence that Aliatypus may be more closely related to the mecicobothriids. Some, per- haps even all, of the last nine characters are functionally correlated \\'ith the niche re- quirements of these taxa and therefore may be misleading. Although the trapdoor niche of Aliatypus is much more similar to (but still different from) the niche of Antro- diaetus and Atypoides than to the nonbur- rowing sheet-\\'eb niche of the mecicoboth- riids, the genital character e\idence and the other distinct character states indicate that this is the result of either parallel or con- \ ergent evolution. A complete study of the 394 Bulletin Museum of Comparative Zoologij, Vol. 141, No. 6 Table n. Comparison of Aliatypus with related taxa. Thickness of arrows is roughly proportional to weight of character and degree of siitdlarity of character states. Hexura pice a ( cf , 9 ) and Meeahexura fulva ( 9 ) are the only mecicobothriids examined. CHARACTER 1. Palpus form 2. Seminal receptacle form 3. Male chelicera /,. Male leg I modified 5. Male pedipalp patella 6. Burrow entrance structure 7. Female rastellvim 8. Pars cephalica 9. ilioracic groove Antrodiaetus & Atvnoides distinct stalk moderately to well sclerotized, rarely sinuous, usually i^th bowl apophysis or anterior- dorsal prominence usually not elongate collar K. ^ strongly elevated ■<- longitudinal Aliatypus similar to Hexura M stalk unsclerotized, very narrow, sinuous, without bowl unmodified ■^ no elongate trapdoor yes strongly elevated pit, depression, or absent Mecicobothriidae similar to Aliatypus stalk rather weakly sclerotized, narrow, sometimes sinuous, without bowl unmodified yes not elongate no burrow (sheet web) no weakly elevated longitudinal 10. No. of macroteeth rows per chelicera 11. Metatarsus IV tricho- bothria number 12. Large median seta just ant. to AME's 13. AL spinnerets 14. Length of labium and Inclination to plane of sternum many absent or non- segmented and degenerating relatively long, weakly inclined 1 (rarely 1-h) yes 2-segmented and functional moderately short, strongly inclined many functional and 1- 2-segmented short, strongly inclined mecicobothriids and consideration of addi- tional characters are required to clarify the relationships of Aliatypus. Antrodiaetus and Atypoides together form a distinct monophyletic group. They are similar to each other in most of the Table 11 characters as well as many others, and arc distinct from other related taxa in numerous characters. They differ from one another in two characters: the presence of AL spinnerets and of the male cheliceral apophysis. The former structures are clearly degenerating in Atypoides and could possibly have been independently lost more than once. The cheliceral apophysis is not degenerating in Atypoides. However, all Antrodiaetus males possess an anterior-dorsal cheliceral prominence which, because of its identical location, is more likely homologous to the cheliceral apophysis than independently evolved. That this prominence is probably either of low functional importance or a nonfunc- tional vestige is perhaps weakly indicated by the high intrapopulation variability of its form in at least some Antrodiaetus spe- cies. I shall tentatively assume the pres- ence of a functional cheliceral apophysis in the ancestral Antrodiaetus-Atypoidcs stock. Antrodiaetus consists largely of two distinct but similar species groups each Systematics and Biology of Antrodiaetus • Coyle 395 clearly monophyletic and together form- ing a monophyletic group distinct from Atypoides. The unicolor group, consisting of nine closely related species, has evolved strongly modified male first legs. The Uncohuanm group consists of three species, two of ^^'hich (A. sfygiiis and A. apachecus) are clearly cognate species. It is probable that this group has undergone a partial reduction in male leg I mating modifi- cations and a functionally related elon- gation of the male pedipalp, a change convergent to the condition in Aliati/pus. Antrodiaetus roretzi, provisionally placed in this genus, is clearly distinct from these two species groups in several characters, the most important being the male palpus structure, female genital structure, and sternum form. It is distinct from Atypoides in the latter two characters. It also lacks a cheliceral apophysis and AL spinnerets, but these are probablv rather easily lost structures of low reliability. The o.c.s. of the important male palpus is strikingly similar to that of Atypoides. but the i.c.s. base has one arm heavily sclerotized as in the other Antrodiaetus species, not as in Atypoides. The palpus is somewhat dis- tinct in other difficult-to-describe features from all species in both genera. A. roretzi is apparently a relict combining some features of both genera; probably its an- cestors branched off from or near the Antrodiaetus-Atypoides ancestral stock. Atypoides is a small, rather hetero- geneous taxon of three morphologically and behaviorally quite distinct species, each apparently an evolutionary relict. A. gertschi is particularly divergent in some characters — seminal receptacle form, male leg I macrosetation, o.c.s. length, male pedipalp shape, and burrow entrance structure. The first three characters bear noteworthy resemblance to Aliatypus. Figure 69 summarizes what I believe to be one of the most probable interpretations of Antrodiaetus and Atypoides phylogeny, given the present evidence. It is a working ATYPOIDES ANTRODIAETUS Figure 69. Diagrammatic representation of probable phy- logeny of /Antrodiaetus and Atypoides. Some of tfie more important and likely evolutionary events are indicated. Ver- tical axis roughly indicates time. Horizontal axis roughly indicates degree of difference. Point of origin of A. roretzi ancestral stock is unclear. Some probable evolutionary events: 1. Character states of the ancestral stock: Male with cheliceral apophysis; male pedipalp not elongate; o.c.s. and i.c.s. strongly sclerotized; o.c.s. broad; male leg I unmodified or not strongly modified; seminal receptacles moderately well sclerotized; AL spinnerets unsegmented; burrow entrance a collapsible collar; bottom end of burrow enlarged. 2. Seminal receptacle sclerotiza- tion reduced; AL spinneret size reduced. 3. Collar becomes elongate and rigid; bottom end of burrow narrowed. 4. Male cheliceral apophysis lost; male leg I becomes more strongly modified; AL spinnerets lost. 5. Male cheliceral apophysis lost; o.c.s. narrows; male leg I increasingly mod- ified (?); AL spinnerets lost. 6. Male leg I becomes strongly modified. 7. AME diameter reduced; male pedipalp elon- gate; o.c.s. and i.c.s. sclerotization somewhat reduced; male leg I modifications reduced; reduction of IVCT number. 8. AME diameter further reduced; male pedipalp further elon- gated; o.c.s. and i.c.s. sclerotization further reduced; seminal receptacle sclerotization reduced. hypothesis hopefully of some predictive value. General zoogeographic considerations. The present day geographic distribution of Antrodiaetus and Atypoides (Map A) is markedly disjunct, with the greatest species 396 Bulletin Museum of Comparative Zoology, Vol. 141, No. 6 diversity in far western North America. The disjunct east Asian, western North American, and eastern North American pattern and the present intimate association of these genera with the modern forest deri\'ativcs of the Arcto-Tertiary Geoflora, indicate that Antrodiaetus and Atypoides have evolved along with this well-docu- mented (Axelrod, 1960) plant assemblage. The much greater present diversity of the Antrodiaetidae in North America indicates that this continent is probably the geo- graphic origin of the family and of its genera. This was also Pocock's ( 1903 ) interiiretation. Although the present day mecicobothriid distribution seems to sup- port this idea, the Atypidae, with both its genera occurring in the Tropics and sub- tropics of the Old World {Atijpus is also found in the Palearctic and Nearctic regions), could have originated there. This and the presence of the evolutionary relict, Antrodiaetus roretzi, in warm temperate Japan weakly hint that extinction could be camouflaging an Asian origin of the Antro- diaetus-Atijpoides ancestral stock. Dispersal Ability and Barriers Although many araneomorph spiders balloon, it is generally believed that few mygalomorph species do. I am aware of only three observations of possible pre- ballooning behavior in mygalomorph spidcrlings: Main (1957b) and Baerg (1928) on two ctenizid species, and Knock (1885) on the atypid, Ati/pus affinis. None of these observers actually saw ballooning. Antrodiaetu.^ and Atypoides spidcrlings have not been observed during dispersal away from the parental burrow, but some indirect evidence indicates that if they balloon at all, only short distances can be covered. Second instar spidcrlings (the dispersal stage) desiccate quickly when not maintained in high humidity. Antro- diaetid species have considerably smaller geographic ranges than do many araneo- morphs that are known to balloon. Several species exhibit marked geographic vari- ation within relatively very small areas. Few to many offspring burrows are com- monly clustered close to the parental bur- row, indicating that at least some spider- ling dispersal is short range. Active dispersal of older immatures and adult females is probably prompted only by burrow microhabitat deterioration and is certainly short range. Passive rafting down streams, rivers, and perhaps across large bodies of water are likely, for bur- rows commonly occur within dense root systems at the bases of trees, in partly decaying logs, and near water. The coloni- zation of the British Columbia coast and its islands by Antrodiaetus pacificus since the retreat of the Wisconsin ice might be largely the result of such rafting. The noctumally wandering adult males can move rapidly (and probably far) over the ground surface. Such wandering cannot alone found new populations, but is cer- tainly important in maintaining gene flow and probably also in fertilizing founder females in nearby localities. Low humidity habitats (with soil humid- ity the immediate limiting factor) are apparently the outstanding barriers to dis- persal and thus to gene flow in Antrodiaetus and Atypoides species. Geographic gaps and genetic variation discontinuities within a species population predominantly coin- cide with habitats drier than those oc- cupied by the species in question. None of the likely means of antrodiaetid dispersal could bridge more than a narrow dry habitat barrier. Geographic Variation All relatively well-sampled species of Antrodiaetus and Atypoides exhibit geo- graphic variation in numerous characters. Five of these species (Antrodiaetus iini- color, Antrodiaetus pacificus, Antrodiaetus pugnax, Atypoides riversi, and Atypoides p,ertschi) exhibit strong and sometimes dis- continuous geographic variation in at least several characters. The reader is referred to the Taxonomy section and to Coyle Systematics and Biology of Antrodiaetus • Coijlc 397 ?Pu7 MAP A Map A. Approximate known distribution ranges of Antrodiaetus and Atypoides species. ( 1968 ) , where a description and discussion each of these species, i.e., they vary con- of the geographic variation in each of these cordantly. Therefore divergent populations species follows each species description, are often defined by several characters. In Different characters frequently show sim- some cases — particularly in A. iinicolor, A. ilar patterns of geographic variation in ]mgnax, and A. gertschi — geographically 398 Bulletin Museum of Comparative Zoology, Vol. 141, No. 6 and genetically intermediate populations are known and more will probably be discovered. The clearest cases of incipient speciation (perhaps spcciation is com- plete?) are found in A. pacifictis and particularly A. riversi, where the geo- graphic barriers to gene flow also appear most formidable. Interestingly, the pre- dominant pattern of geographic variation in A. pugnax may be largely the result of character displacement, i.e., the selection pressure resulting from competition or mating errors with Antwdioetiis occultiis in tlie area of sympatry may have accelerated the genetic divergence of ( probably ) poorly connected A. pugnax gene pools. The date of disjunction (and thus the rate of phenotypic divergence) can be roughly estimated for only a few divergent populations. It is nearly certain that Pleis- tocene climatic fluctuations isolated the present divergent populations of A. nni- coIo)\ A. pacificus, and A. riversi. These disjunctions probably occiured as recently as the end of the last glacial maximum ( see below ) ; perhaps these divergent gene pools have been evolving as separate entities for as little as 15,000 years. In each of th(^ above five species there is greater geographic variation per unit area than in many well-studied araneo- morph species. Presumably, two important factors contributing to this situation are the somewhat narrow habitat requirements and particularly the poor, long range dis- persal ability of antrodiaetids. The former factor favors population fragmentation and the latter a slow rate of gene exchange among such fragments. There is some in- dication that burrowing mygalomorph spiders may in general exhibit much geo- graphic variation per unit area for perhaps these same reasons (see for example Main, 1957a; Loksa, 1964; Forster and Wilton, 1968). Speciation: Probable Historical Events Unicolor group of Antrocliaetus. Probably the last connection between eastern and western North American populations of this species group was severed in the late Tertiary during the formation of the broad semi-arid grassland of central North Amer- ica. (Map A should serve as a reference throughout this section.) It is likely that one species disjoined in this manner was morphologically and ecologically similar to, and a direct ancester of, both A. pacificus and A. unicolor (and perhaps A. rohustus) and that therefore relatively little pheno- typic divergence has occurred between these species over the last 10 to 25 million years. More recent east-west contact, such as during the cool, moist, glacial maxima of the Pleistocene, appears less likely. Present evidence indicates the absence of any widespread Pleistocene forest on the Great Plains, and that probably even nar- row forest corridors allowing east-west dis- persal of humid forest animals were nonexistent (Dillon, 1956; Frey, 1965; Wade, 1966). However, some controversy exists regarding this last conclusion (Blair, 1965). The last Bering land bridge favorable for Antrodiaetus dispersal existed no more recently than Miocene or perhaps Pliocene times. Pleistocene Bering bridges were unforcsted (Simpson, 1947; Flint, 1957; Pewe, Hopkins, and Giddings, 1965), but favorable Arcto-Tertiary forest extended across the bridge during its frequent and lengthy exposures from Eocene into at least Miocene times (Simpson, 1947, Axelrod, 1960). A. ycsocn.sis is probably descendant from a population that extended west from North America when the bridge was last favorable ecologically. An east to west dispersal is inferred because all other uni- color group species are presently North American. Thus the evidence for disjunc- tion time of all three geographic fragments of the unicolor group indicates that this species group had originated by early Pliocene. The presence of the evolutionary relict, A. rorctzi, in Japan indicates another connection over perhaps an earlier Terti- ary bridge. Systematics and Biology of Antrodiaetus Coijle 399 The relatively great diversity of tinicolor group species in the Pacific Northwest, the marked geographic variation in certain of them, and the strong habitat differences among some of them are perhaps largely attributable to the remarkable past physio- graphic, climatic, and thus ecological fluctuation in this region (Detling, 1968), and to the great diversity of these factors at any one time. Such fluctuation should favor speciation by causing expansions, shifts, contractions, and disjunctions of l^opulations. Habitat diversity and fluctu- ation have perhaps accelerated the genetic divergence of disjunct gene pools and favored shifts into different habitat niches. It is unclear whether most of this speci- ation occurred during the Pleistocene or late Tertiary. A possible case of incipient speciation exists in A. pocificus, which consists of a widespread coastal population and a pos- sibly disjunct interior montane population. At least the Oregon-Washington portion of this interior population has undergone much genetic divergence since the reduc- tion of gene exchange with the coastal population. The barrier to gene flow be- tween these two forest populations is the broad arm of dry Juniper-sagebrush wood- land plant formation (Detling, 1968) ex- tending northward from the Great Basin through central Oregon and Washington into southern British Columbia. Gene flow reduction and resulting genetic divergence probably began \\'ith the climatic drying and warming just after ( about 15,000 years ago) the last Wisconsin glacial advance. During this maximum, a boreal forest favorable for A. pacificus extended at least along the ice front from the Washington coast to the mountains of northern Idaho and possibly even extended across the Oregon plateau to connect directly the coastal and northeast Oregon populations (Heusser, 1960; Detling, 1968). If the interior population is now reproductively isolated, then disjunction and speciation could also have occurred earlier in the Pleistocene under similar conditions. It is noteworthy that this distribution pattern and probably the historical causes are very similar to those of the salamanders FJetho- don vanch/kei (Highton, 1962; Wade, 1966) and Taricha gramilosa (Riemer, 1958). Antrodiaetus tmicolor is a widespread species consisting of a dense, roughly con- tinuous central population and numerous geographically semi-isolated (some f)er- haps isolated) peripheral populations. Pre- sumably the eastern, southern, and western peripheral populations were continuous with the central population during the cooler, more humid Pleistocene glacial maxima when the favorable cool humid forest habitat expanded outward over the lowland areas. Some of these peripheral populations have since undergone con- siderable genetic divergence as a result of the increasing discontinuitv of favorable habitat peripherally and the resulting re- duction of gene influx from the central population. Lincolniomis group of Antrodiaetus. The morphologically very similar A. stygius and A. apachecus are almost certainly sister species. Perhaps the parent population split into eastern and western fragments during the formation of the grassland in the late Tertiary, the same event postu- lated to have last disjoined the nnicoJor group. However, the small amount of phenotypic divergence of these two species suggests a more recent disjunction during a Pleistocene interglacial. Blair ( 1958, 1965) has suggested this for two south- western montane relict species of lungless salamanders, but there is considerable evi- dence against an east-west Pleistocene forest corridor (Frey, 1965; Wade, 1966). Pollen analysis evidence for an extensive Parkland-yellow pine association extending eastward from Arizona well into Texas at the Wisconsin maximum (Martin and Mehringer, 1965) adds some support for a possible Pleistocene connection. Antrodiaetus apachecus is presently frag- 400 Bulletin Museum of Comparative Zoology, Vol. 141, No. 6 mented b) intervening arid habitats into a number of disjunct montane populations that have been sampled too poorly to indicate how much they have diverged Irom one another- genetically. These were probabK' last continuous near the end of the Wisconsin glacial maximum roughly 17,000 years ago when all known present populations were apparently connected by habitats no dryer than Ponderosa (yellow) pine forest (Martin and Mehringer, 1965). Atypoides. Disjunction of Atypoides into eastern and western North American seg- ments occurred at least as long ago as the fonnation of the late Tertiary grassland. A. riversi reveals a pattern of incipient speciation (Perhaps speciation is com- plete?) with two probably disjunct and genetically quite divergent populations on both sides of the dry Central Valley of California. These \\'ere last connected by gene flow across the valley floor (or per- haps to the north of the valley) during one of the cooler, moister Pleistocene gla- cial maxima, perhaps during the recent (Wisconsin) one. Similar histories have been postulated for cognate pojmlations of salamanders with similar distributions (Stebbins, 1949; Riemer, 1958). Rephoductiv^ Isolating Mechanisms I .should emphasize that although the study of geographic variation and the fore- going discussion may help in understand- ing geographic isolate fonnation in these spiders as well as the rate of phenotypic divergence of such isolates, nothing can be inferred about the time required for the e\()lution of reproductive isolating mecha- nisms, the crucial part of speciation. More data is needed on the relative im- portance of different reproductive isolating mechanisms in antrodiaetid species. Field observations suggest that ecological isolat- ing mechanisms are important in some sympatric unicolor group species in the Pacific Northwest. Temporal (seasonal) isolation exists between A. lincolnianus and A. sty gins and is suggested for other species. The importance of ethological isolating mechanisms may be suggested by the species specific morphology of the male first leg of Antrodiaetus species. The close similarity of the male palpus and of the female genitalia of all unicolor group species indicates that perhaps mechanical isolating mechanisms are unimportant in these species. LITERATURE CITED Atkinson, G. F. 1886. Descriptions of some new trap-door spiders; their notes [sic] and food habits. Ent. Amer., 2: 109-117, 128-137. AussERER, A 1871. Beitriige zur Kenntniss der AracIiniden-FamiHe der Territelariae Thorell. Verh. zool.-bot. Ges. Wien, 21: 117-224. AxELROD, D. I. 1960. The evolution of flowering plants. Pp. 227-305. In S. Tax (ed.) Evo- lution after Dai-win. Chicago, Univ. of Chicago Press. Baerg, W. J. 1928. Some studies of a trapdoor spider. Ent. News, 39(1): 1-4. Banks, N. 1896. New Cahfornian spiders. J. New York Ent. Soc, 4(4): 88-91. Blair, W. F. 1958. Distributional patterns of vertel)rates in the southern United States in relation to past and present environments. Pp. 433-468. 7/1 C. L. Hubbs (ed.), Zoo- geoi^raphy. American Assoc. Adv. Sci., Publ. 51. . 1965. Amphibian speciation. Pp. 543-556. In H. E. Wright and D. G. Frey (eds.), The Quaternary of the United States. Princeton, Princeton Univ. Press. Branson, B. A., and Batch, D. L. 1968. An ecological study of valley-forest spiders from northern Kentucky. Proc. Biol. Soc. Wash- ington, 81: 197-208. BucHLi, H. 1962. Note preliminaire sur 1' ac- couplement des araignees mygalomorphes Nemesia caementaria, Nemesia duhia, and Pachtjlomenis piceus (Ctenizidac). Vie et Milieu, 13(1): 167-178. Chamberlin, R. V. 1917. New spiders of the family Aviculariidae. Bull. Mus. Comp. Zool., 61(3): 26-75. . 1935. Miscellaneous new American spiders. Bull. Univ. Utah, 26(4): 1-79. . 1945. On some nearctic mygalomorph spiders. Ann. Ent. Soc. America, 38(4): 549-558. -, and Ivie W. 1933. Spiders of the Raft River Mountains of Utah. Bull. Univ. Utah, 23(4): 1-74. CoMSTOCK, J. H. 1912. The Spider Book. New York, Doubleday, Doran and Co. 721 pp. Systematics and Biology of Antrodiaetus • Cotjle 401 . 1940. op. cit. (rev. by W. J. Gertsch) Ithaca, New York, Cornell Univ. Press. 729 pp. CoYLE, F. A. 1968. The mygalomorph spider genus Atijpoides (Araneae: Antrodiaetidae). Psyche, 75: 157-194. . 1969. Systematics and biology of the mygalomorph spider genus Antrodiaetus and related genera. Thesis. Harvard University. Crome, W. 1955. Die Beziehungen zwischen dem dorsalen Zeichnungsmuster rmd der Metamerie des Spinnenalidomens. Part 2. Zool. Jahr., 83 : 541-638. Detling, L. E. 1968. Historical background of the flora of the Pacific Northwest. Univ. of Oregon Mus. Nat. Hist. Bull. No. 13: 1- 57. Dillon, L. S. 1956. Wisconsin climate and life zones in North America. Science, 123: 167-176. Knock, F. 1885. The life history of Atijpus piceiis. Trans. Ent. Soc. London, (4): 389^20. Flint, R. F. 1957. Glacial and Pleistocene Geology. New York, J. Wiley and Sons. 553 pp. Forster, R. R., and Wilton, C. L. 1968. The spiders of New Zealand. Part 2. Otago Mus. Bull. No. 2: 1-180. Frey, D. G. 1965. Other invertelirates — an essay in biogeography. Pp. 613-631. In H. E. Wright and D. G. Frey (eds.). The Quater- nary of the United States. Princeton, Prince- ton Univ. Press. Gertsch, W. J. 1949. American Spiders. Prince- ton, D. Van Nostrand Co. 285 pp. . 1967. A new liphistiid spider from China. J. New York Ent. Soc, 75(3): 114-118. Gertsch, W. J., and Jellison, W. 1939. Notes on a collection of spiders from Montana. American Mus. Novitates, No. 1032: 1-13. Gertsch, W. J., and Mulaik, S. 1940. The spiders of Texas. I. Bull. American Mus^ Nat. Hist, 77(6): 307-340. Gleason, H. a., and Cronquist, A. 1964. The Natural Geography of Plants. New York, Columbia Univ. Press. 420 pp. Heusser, C. J. 1960. Late Pleistocene environ- ments of north Pacific North America. Amer- ican Geog. Soc. Spec. Publ., 35: 1-308. HiGHTON, R. 1962. Revision of North American salamanders of the genus PJcthodon. Bull. Florida State Mus., Biol. Sci., 6(3): 235-367. Holm, A. 1954. Notes on the de\elopment of an orthognath spider, Ischnothcle karsclii Bos. and Lenz. Zoo. Bidrag. Fran. Uppsala, 30: 199-222. Kaestner, a. 1968. Invertebrate Zoology. Vol. 2. New York, J. Wiley and Sons. 472 pp. Kaston, B. J. 1948. Spiders of Connecticut. Connecticut Geol. Nat. Hist. Surv. Bull., No. 70: 1-874. . 1964. The evolution of spider webs. American Zool., 4: 191-207. . 1966. Evolution of the web. Nat. Hist., 75(4): 26-33. -, AND Kaston, E. 1953. How to Know the Spiders. Dubuque, Iowa, W. C. Brown Co. 220 pp. KiSHiDA, K. 1930. A new scheme of classification of spider families and genera. Lansania, 2 (13): 33-43. KoMATsu, T. 1937. (Photo of Acattyma roretzi burrow entrance.) Acta Arach., 2(4): pi. 7. . 1942. (Title not translated.) [in Japan- ese] Acta Arach., 7(2): 54-70. 1961. Cave spiders of Japan, their taxon- omy, chorology, and ecology. Arach. Soc. E. Asia. 91 pp. Levi, H., and Levi, L. 1968. Spiders and Their Kin. New York, Golden Press. 160 pp. LoKSA, I. 1966. Nemesia pannonica O. Herman (Araneae: Ctenizidae). Ann. Univ. Sci. Budapest, Sect. Biol., 8: 155-171. Main, B. Y. 1957a. Biology of aganippine trap- door spiders (Mygalomorphae: Ctenizidae). Austrahan J. Zool., 5(4): 402-473. . 1957b. Occurrence of the trapdoor spider Conotliele maijlaijana ( Doleschall ) in Aus- tralia. W. Australian Nat., 5(7): 209-216. Martin, P. S., and Mehringer, P. J., Jr. 1965. Pleistocene pollen analysis and biogeography of the Southwest. Pp. 433-45L In H. E. Wright and D. G. Frey (eds.). The Quater- nary of die United States. Princeton, Prince- ton Univ. Press. Marx, G. 1883. Order Araneina. Pp. 21-26. In L. O. Howard (ed.). Invertebrate Fauna of South Carolina. Charleston, S. Carolina. Mearns, E. a., and Irwin, J. 1902. Correspon- dence: a biographical and autobiographical letter. The Auk, 19(1): 116-117. Merriam, C. H. 1897. Charles E. Bendire. Science, 5: 261-262. Merrill, J. C. 1898. In memoriam: Charles E. Bendire. The Auk, 15(1): 1-6. Ohe, H. 1966. Observations of Antrodiaetus roretzi. [in Japanese] Atypus, No. 39: 29-36. Peck, W. B. 1966. The population composition of a spider community in west central Mis- souri. American Midi. Nat., 76(1): 151-168. Petrunkevitch, a. 1911a. Sense of sight, court- ship, and mating in Dugesiella hentzi (Girard), a theraphosid spider from Texas. Zool. Jahr., 31: 355-376. . 1911b. A synonymic index-catalogue of spiders of Nortli, Central, and South America 402 Bulletin Museum of Comparative Zoology, Vol. 141, No. 6 with. . . . Bull. American Mus. Nat. Hist., 29: 1-791. . 1925. Descriptions of new or in- adequately known American spiders. Ann. Ent. Soc. America, 18: 313-322. 1939. Catalogue of American spiders. Trans. Connecticut Acad. Arts Sci., 33; 133-138. Pewe, T. L., Hopkins, D. M., and GmoiNcs, J. L. 1965. The Quaternary geology and archeol- ogy of Alaska. Pp. 355-374. In H. E. Wright and D. G. Frey (eds.). The Quater- nary of tlie United States. Princeton, Prince- ton Univ. Press. Picicard-Cambridge, O. 1883. On some new genera and species of spiders. Proc. Zool. Soc. London, pp. 352-365. PiMENTEL, R. A. 1967. Invertebrate Identifi- cation Manual. New York, Reinhold Publ. Co. 151 pp. PococK, R. I. 1901. On some new trap-door spiders from China. Proc. Zool. Soc. London, pp. 207-215. . 1903. On the geographical distriliution of spiders of the order Mygalomorphae. Proc. Zool. Soc. London, pp. 340-367. RiEMER, W. J. 1958. Variation and systematic relationships within the salamander genus Taricha. Univ. of California Publ. Zool., 56; 301-390. Rivers, J. J. 1891. Description of the nest of the Californian turret Ijuilding spider, witli some references to allied species. Zoe, 2: 318-320. Scheffer, T. H. 1906. Additions to the list of Kansas Arachnida. Trans. Kansas Acad. Sci., 20: 121-130. Schlinger, E. I. 1966. An analysis of the dis- tribution of Eulonchus Gerstaeker ( Diptera : Acroceridae ) . Bull. Ent. Soc. America, 12 (2): 112-113. SrNiON, E. 1884. Note sur le groupe des Meci- cobothria. Bull. Soc. Zool. France, 9: 313- 317. . 1886. Arachnides recueillis par M. A. Pavie dans le royaume de Siam, au Gamboge et en Cochinchine. Actes Soc. Linn. Bordeaux, 40: 137-166. . 1890. Liste des especes de la famille des Aviculariides qui haliitent I'Amerique du Nord. Actes Soc. Linn. Bordeaux, 44: 307- 326. . 1892. Histoire Naturelle des Araignees. Paris, 1(1): 1-256. . 1897. Materiaux pour servir a la faime arachnologique de I'Asie meridionale. Mem. Soc. Zool. France, 10: 252-262. . 1903. Histoire Naturelle des Araignees. Paris, 2(4): 669-1080. Simpson, G. G. 1947. Holarctic mammalian faunas and continental relationships during the Cenozoic. Geol. Soc. America, Bull. 58: 613-687. Smith, C. P. 1908. A preliminary study of the Araneae Theraphosae of California. Ann. Ent. Soc. America, 1(4): 207-249. Stebbins, R. G. 1949. Speciation in salamanders of the plethodontid genus Ensatina. Univ. of Cahfornia Publ. Zool., 48: 377-526. Todd, V. 1945. Systematic and biological account of tlie New Zealand Mygalomoiphae. Trans. Roy. Soc. New Zealand, 74(4): 375-407. Uyemura, T. 1936. On a Japanese trapdoor spider, Acattijma roretzi, from Tenranzan. [in Japanese] Acta Arach., 1(4): 137-139. . 1942. Description of a new acattymid spider from Hokkaido, [in Japanese] Acta Arach., 7(1): 31-38. Vachon, M. 1958. Contribution a I'etude du developpement postembryonnaire des arai- gnees (Deuxieme note) Orthognaths. Bull. Soc. Zool. France, 83: 429-461. Wade, D. B. 1966. Comparative osteology and evolution of the lungless salamanders, family Plethodontidae. Mem. S. California Acad. Sci., 4: 1-111. Wasbauer, M. S., and Powell, J. A. 1962. Host records for some North American spider wasps, with notes on prey selection ( Hyme- noptera: Pompilidae). J. Kansas Ent. Soc, 35(4): 393-401. Whitmyre, G., and Snetsinger, R. 1967. A folding-door tarantula from Pennsylvania. Melsheimer Ent. Series., No. 1. WoRLEY, L. G. 1928. New Nebraska spiders. Ann. Ent. Soc. America, 21(4): 619-622. WoRLEY, L. G., AND PiCKWELL, G. 1931. The spiders of Nebraska. LTniv. Nebraska Studies, 27(1-4): 1-129. Yaginuaia, T. 1957. Spiders from Hokkaido and Rishiri Island, [in Japanese] Acta Arach., 14(2): 51-62. . 1960. Spiders of Japan in Colour, [in Japanese] Osaka, Japan. 186 pp. 1962. Gave spiders in Japan, [in Jap- anese] Bull. Osaka Mus. Nat. Hist., 15: 65-77. Yamamoto, G. 1942. Observations on the male of Acattt/ma roretzi. [in lapanesej Acta Arach., 7(3-4): 129-133. YosHiKURA, M. 1955. Embryological studies on the liphistiid spider, HeptatJiela kimurai. Part 2. Kumanioto J. Sci., Ser. B., 2(1): 1-86. . 1958. On the development of a purse- web spider, Atijptis karschi Donitz. Kuma- moto J. Sci., Ser. B., 3(2): 73-86. ^/. fjj .J OF THE Museum of omparative Zoology A Monograph of the Genera Calidviana, Ustronia, Jroschely'iana, and Semitrochatella (Mollusca: Archaeogastropoda: Helicinidae) in Cuba WILLIAM J. CLENCH AND MORRIS K. JACOBSON HARVARD UNIVERSITY VOLUME 141, NUMBER 7 CAMBRIDGE, MASSACHUSETTS, U.S.A. 6 AUGUST 1971 PUBLICATIONS ISSUED OR DISTRIBUTED BY THE MUSEUM OF COMPARATIVE ZOOLOGY HARVARD UNIVERSITY Bulletin 1863- Breviora 1952- Memoirs 1864-1938 JoHNSONiA, Department of Mollusks, 1941- OccAsiONAL Papers on Mollusks, 1945- Other Publications. Bigelow, H. B., and W. C. Schroeder, 1953. Fishes of the Gulf of Maine. Reprint, $6.50 cloth. Brues, C. T., A. L. Melander, and F. M. Carpenter, 1954. Classification of Insects. $9.00 cloth. Creighton, W. S., 1950. The Ants of North America. Reprint, $10.00 cloth. Lyman, C. P., and A. R. Dawe (eds.), 1960. Symposium on Natural Mam- malian Hibernation. $3.00 paper, $4.50 cloth. Peters' Check-list of Birds of the World, vols. 2-7, 9, 10, 12-15. (Price list on request. ) Turner, R. D., 1966. A Survey and Illustrated Catalogue of the Teredinidae ( Mollusca: Bivalvia ) . $8.00 cloth. Whittington, H. B., and W. D. I. Rolfe (eds.), 1963. Phylogeny and Evolution of Crustacea. $6.75 cloth. Proceedings of the New England Zoological Club 1899-1948. ( Complete sets only. ) Publications of the Boston Society of Natural History. Authors preparing manuscripts for the BtiUetin of the Museum of Comparative Zoology or Breviora should send for the current Information and Instruction Sheet, available from Mrs. Penelope Lasnik, Editor, Publications Office, Museum of Comparative Zoology, Harvard University, Cambridge, Massachusetts 02138, U.S.A. © The President and Fellows of Harvard College 1971. A MONOGRAPH OF THE GENERA CAUDVIANA, USTRONiA, TROSCHELVIANA, AND SEMITROCHATELLA (MOLLUSCA: ARCHAEOGASTROPODA: HELICINIDAE) IN CUBA WILLIAM J. CLENCH AND MORRIS K. JACOBSON CONTENTS Introduction 404 History 404 Radula 405 Distribution 406 Nomenclature 408 Enemies 409 Types 409 Abbreviations 409 Maps 409 Acknowledgments 410 Systematics 410 Key to Genera and Subgenera of Tribe Vianini "410 Genus Calidviana H. B. Baker . 410 Calidvimui littoricola (Pfeiffer) 411 Key to Species of U stroma 412 Ustronia Wagner 412 Ustronia sluanii sloanii (Orbigny) 413 U. sloanii aedilii Aguayo & A. Torre 414 U. acuDiinata actnninafa (Poey) 415 U. acuniinata columcUaris (Gundlach) ,.. 417 Key to Species of Tmschelviana S. S. 418 Genus Troschcl liana H. B. Baker 418 Subgenus TioscJiclviana S. S. 418 Troschclviana ( TmscJielviana) cnjthiaca (Sowerby) 418 T. ( Troschclviana ) chrijsochasma (Poey) 419 T. i Troschclviana) in-pe-locality, bei Baracoa an der Meereskiiste an Steinen; lectotype, here selected, MCZ 86610, Baracoa, ex. Gundlach; paralectotypes, same data, MCZ 273211). Helicina littoricola Gundlach. Pfeiffer 1865, Monographia Pneumonopomorum Viventium, Suppl. 2, p. 219; Arango 1879, Contribucion Fauna Malacologica Cubana, p. 45. Helicina littoricola Gundlach. Sowerby 1866, Thes. Conchyl., London, 3: 283, pi. 269, figs. 121-122. Eutrochatella (Eutrochatella) littoricola Pfeiffer. Wagner 1908, in Martini & Chemnitz, Conchyh- Cab., (2)1: sect. 18, pt. 2, p. 114, pi. 21, figs. 22-23. Eutrochatella (Bakerviana) littoricola maisiana Aguayo & Jaume 1957, Mem. Soc. Cubana Hist. Nat.,^ 23: 118, pi. 1, fig. 5 ( type4ocahty, Maisi, Baracoa, Provincia de Oriente, Cuba; holotype, MP 17399). Description. Shell about 4 mm in height, subglobose, solid, lusterless, base Httle rounded. Color grayish white or very pale flesh colored, apex lighter, aperture yellow. Wliorls 5:^, gently rounded, body whorl with a very obtuse keel, barely descending at aperture. Spire unevenly domelike, apex sharply raised. Aperture oblique, widely semilunate, yellow or orange within. Parie- tal wall with a strongly raised, somewhat glossy parietal callus, rounded at exterior margin, about the same extent as the aperture, joining the dorsal and basal lip margins. Palatal lip double, the inner lip being an extension of the margin of the parietal callus, raised sharply at the um- bilical region where it is reflected to form a pseudo-chink. Columella weakly sig- moid, rounded at the insertion with the basal lip. Sculpture of diagonal, irregular growth lines crossed by much stronger, even, rounded, raised spiral cords, wider than their intervals, becoming obsolete near the umbilical region. Protoconch 1^ whorls, white, microscopically pitted, sharply raised above the succeeding whorls. Periostracum wanting. Operculum calcareous, whitish, with a low, rather wide external columellar ridge; outer margins with a narrow, low ridge. height Width mm mm 4.1 4.3 Baracoa, lectotype 4.2 4.0 II paralectotype 3.5 4.0 II II 3.5 4.3 EI Baga, Baracoa 4.5 4.3 II II Remarks. This is an easily recognized fonn. The subglobose shape and the strongly sculptured surface are found in no other Cuban members of the Vianini. Nevertheless, it possesses the sharply raised apex which is characteristic of the tribe. The genus has had its greatest develop- ment in the various islands of the Bahamas, C. littoricola is closely allied to calida (Weinland, 1862) from the Crooked Island group and Creat Inagua, Bahamas, but the latter species has a much larger shell and has pronouncedly weaker spiral sculpture. The fact that in Cuba the range of littori- cola is confined to a strip on the north coast near Baracoa argues strongly for its having had its origins in the nearby Ba- hamas. In the colonies of this species, there are some variations in size and in comparative dimensions, some shells being smaller than others and greater in width than in height, others being larger and greater in height than in width. This may prove to be a secondary sexual characteristic. Aguayo & Jaume (1957: 118) gave the subspecific name maisiana to a population at Maisi with an orange instead of yellow aperture. They stated that the nominate form comes from Cayojuin (Cayo Juin) and other areas north of Baracoa, whereas maisiana comes from the region to the south. However, a series of littoricola in MCZ (127457, Bennudez collection) comes from El Baga, well to the south of Baracoa (PI. 8). Hence zoogeographical reasons 412 Bulletin Museum of Comparative Zoology, Vol. 141, No. 7 do not warrant the separation of the two forms which differ only in the color of the aperture. Moreover a large series of C. calida (Weinland) (MCZ 189022) from Great Inagua Island in the Bahamas, a possible ancestor of littoricola, shows a mixture of specimens with yellow and orange apertures. We must conclude that the color of the aperture is insufficient to establish taxonomic distinctions. Specimens examined. Oriente. Baracoa; El Baga. Key to the Species of USTRONIA 1. Shell, C.15 X 18 mm in size, rather thin shelled - 2 1. Shell, C.12 X 13 mm in size, more solid 3 2. Shell witliout columellar blotch —.a. acuminata 2. Shell with columellar blotch a. cohmellaris 3. Shell up to 12 mm in height s. sloanii 3. Shell less than 10 mm in height s. acdilii Ustronia Wagner Ustronia Wagner 1908, in Martini & Chemnitz, Conchyl.-Cab., (2)1: sect. 18, pt. 2, p. 116 (type-species, Helicina sloanii Orbigny 1842, by subsequent designation, H. B. Baker 1922, Proc. Acad. Nat. Sci. Philadelphia, 74: 62; not H. sloanei Wagner et atict.). Description. Shell moderately large, low to elevated turbinate, relatively thin, smooth, lip simple, moderately expanded, shallowly excavated in umbilical region; parietal callus narrow, often brightly colored. Axial and spiral sculpture present. Central and T-lateral radular teeth without cusps, as in Eiitrocliatelh, but with a .smaller number of unicuspid marginals. Remarks. This group has the radula of Etitrochatella (see Troschel, 1857: 79, pi. 5, fig. 6) and the smooth shell of Troschel- viana. In this way it occupies an analogous position to CaUdviana, which has a Tro- schelviana-Mkc radula and a Etitrochatella- like shell. In this genus we include the larger, rather thin-shelled species of west- ern Cuba. The radula of T. chnjsostoma Pfeiffer {=,sloanii Orbigny) as pictured by Troschel (1857: 78, pi. 5, fig. 4) shows the uncusped smooth teeth of Eutrochatella s. s. Thus, we employ Ustronia Wagner as it was limited by Baker (1922). Aguayo (1962: 9) tried to show that Baker's designation of .ecies, Heli- cina polituhi Poey 1852, [=H. pyramidalis Sowerlw 1842], original designation). Description. Shell broadly conic, moder- ately elevated, smooth, with faint growth lines and relatively weak spiral lines. Radula cusped as in Troschelviana s. s., but more weakly, the cusps of the T-lateral fewer and noticeably weaker. Remarks. This section was set up by Baker to include the smooth Etdrocha- tella-\ike species of Cuba with lower shells and with less strongly cusped radular teeth. Spiral sculpture, though still weak, is more noticeable than in Troschelviana s. s. The shells generally are somewhat larger than those of Troschelviana s. s. Troschelviana {Cubaviana) pyramidalis (Sowerby) Plate 2, figures 7, 1 1 Helicina conica Orbigny 1842, Mollusques, in Sagra, Histoire Physique, Politique, et Naturelle de rile de Cuba, Paris, 1: 249, pi. 20, figs. 7-9 (not fig. 10); [not Pfeiffer 1839] (tyi^e- locality, I'interieiu- de I'ile de Cuba; type, BM(NH) 1854.10.4.170). Helicina piiramidalis Sowerby 1842, Thes. Conchyl., 1: 9, pi. 3, fig. 104; 1866, op. cit., 3: 284, pi. 270, fig. 157. Helicina straminea Morelet 1851, Testacea Novis- sima Insulae Cubanae et Americae Centralis, 2: 18 (type-locality, montes insulae Cubanae Guajaibon dictos; 3 syntypes, BM(NH) 1893. 2.4.1615-17). Helicina politida Poey 1852, Memorias Historia Natural Isla de Cuba. 1: 113, pi. 5, figs. 4-6 (type-locality, la cordillera frente a Santa Cruz, en tierra de D. Francisco Adolfo Sauvalle; type, MP?). Helicina exactita Poey 1852, op. cit., 1: 114, pi. 5, figs. 7-9 ( type-locality, en las lomas de Cayajabos; type, MP?). Cuban Helicinids • Clench and Jacobson 425 Eiitrochatella (Ustionia) pyramidalis percarinata Wagner 1908, in Martini & Chemnitz, Conchyl- Cab. (2)1: sect. 18, pt. 2, p. 121, pi. 23, fig. 25 (type-locality. Range! auf Cuba; type, IZW 8540). Description. Shell 6 to 8 mm in height, broadly turbinate, moderately solid to rather thin, periphery sharply carinate, oecasionally extremely so, base flat to moderately inflated, shell sublustrous. Color bluish white or yellowish, occasion- ally very faintly nifous-tinged, base green- ish or greenish yellow, lustrous, frequently translucent, carina with a white band be- low, margined by a somewhat wider, glassy, olivaceous line; colored spot on columella wanting. Suture well impressed, occasionally subchanneled. Spire broadly conic, sides straight or faintly convex, apex sharpely raised, nipplelike. Aperture oblique, subrhomboid, outer angle acutely rounded, sometimes yellow internally. Pari- etal wall flat or moderately inflated, shallowly excavated in umbilical region, with a thin, gleaming parietal wash that extends shortly beyond the aperture, and with a low, curved, rounded lamella at the columellar insertion of the basal lip. Palatal lip weakly flaring, except at the extremi- ties, set off from the body whorl by a .shallow, rounded groove that iims along the base of the peritreme. Columella short, shallowly concave below, rounded convex above. Sculpture of diagonal growth lines, crossed by variously strong, regularly spaced spiral cords which are generally weaker or obsolescent on the body whorl and the base. Protoconch 1^ whorls, white, rounded, microscopically punctate, shaq^ly raised. Periostracum wanting, but oc- casional specimens are encrusted with a thin layer of inorganic matter. Operculum as in genus, strongly punctate on the sur- face, outer layer glassy white, subtrans- parent in center; concentric growth lines unevenly spaced; columellar margin weakly twisted, armed with a raised, rounded lamella with a punctate surface; inner layer thin, light brown, barely darker at margins, extending beyond the edge of the calcareous layer. eight Width mm mm 6.5 6.8 Rangel 6.0 7.1 Rancho Lucas, Guajaibon 6.2 7.3 Callajabos ( Cayajabos ) 6.8 7.2 Callajabos ( Cayajabos ) 7.6 7.6 Rangel 6.3 6.2 Santa Cruz de los Pinos, locality 6.2 6.0 El Retire, Rangel type- Remarks. The specimens we have ex- amined indicate that the range of this spe- cies reaches from the region of Cayajabos near the Sierra del Rosaria and Bahia Honda westward to the Pan de Guajaibon and Rangel. In these latter two localities the range apparently overlaps that of Troschelviana nihromarginata (PI. 6). The shells of this species are easily recognized by the variously sharp periph- eral carina, marked by a white band which below is margined by a subequal, oliva- ceous, generally subtranslucent line. The carina varies slightly in strength and in the prominence of the two color bands, but these are always prominent enough to identify the species. Some populations of nihromarginata in the vicinity of Vinales (Kilometer 14 and El Cuajani) have shells that are also carinate, but the keel is always less acute and is not marked by the two bands as in pyramidalis. In addition to the presence of the keel, the present species differs from the neigh- boring nd)romarginata in its lower and wider shape, its generally smaller size, its flatter w horls and in the consistent absence of the colored columellar spot. There is little doubt in our minds that all the names listed in the synonymy refer to the same species. When attempting to distinguish their species — in the few cases where the attempt was made — the various authors refer to such characters as shell color, the strength of the carina as well 426 Bulletin Museinn of Comparative Zoology, Vol. 141, No. 7 as size, shape, and sculpture, all of which we find to be variable in this group. Poey's exacuta is very sharply keeled, Init it is not difficult to find specimens in which the keels are weaker and clearly intergrade with the weaker keel of typical pyramidalis. Poey (1852: 114) admitted that his form "tal ve/. no es mas que una vari(xlad de la //. strominea Morelet," dis- tinguished by its smaller size (6 mm in- stead of 8) and the greenish rather than yellow color, both variable characteristics. Poey's politula has longitudinal striations confined to the upper whorls, a flattened base, a deep suture, all as in pyramidalis. It differs, according to Poey, in the swell- ing (hincha/on) of the last three whorls "which makes it appear more globulose when viewed from above" (translated). This characteristic seems to be associated with specimens which have a less sharply keeled carina and is of no diagnostic value. From the material available for this study, we are not able to judge whether the geographically separated populations are consistent in their variations or are com- posed of forms which show differing char- acteristics. Specimens' examined. Pinar del Rio. El Taco; Las Animas; El Retiro (all Rangel); Guajaibon; Rancho Lucas, Gua- jaibon; El Mamey, Sierra del Rosario, Cayajabos; Quiiiones, Bahia Honda; Santa Cniz de los Pinos; Ingenio Quinones, Sierra de Guacamaya. Troschelviana iCubaviana) rubromarginafa (Gundlach) Plate 2, figures 3, 4, 5, 6, 10 Helicina nihromarginata Gundlach in Poey, 1858, Memorias Historia Natural Isla de Cuba, 2: 15, pi. 1, figs. 17-18 (as Culindrella [sic]; type-locality, in monte Guajaibon insulae Cubae; lectotype, here selected, MCZ 90024, Guajaibon ex T. Bland Collection; paralecto- types, MCZ 273215, same.) Helicina (Helicina) nodae Arango 1862, Jour, de Conchy]., 10: 409 (not Helicina nodae Sowerby 1866; type-locality, rupes prope Guane Cubae occidentalis; lectotype, here selected, MCZ 73781, Guane, ex Arango). Helicina wrighti Pfeiffer 1863, Malak. Blat., 10: 195 (type-locality, Vignales in westlichen Theil von Cuba; lectotype, here selected, MCZ 73865, ex Wright, J. G. Anthony Collection). Helicina festa Gundlach in Sowerby 1866, Thes. Conchyl., 3: 284, pi. 270, figs. 152-153. Helicina rubromarginafa Gundlach, Pfeiffer 1858, Monographia Pneumonopomorum Vixentiinn, suppl. 1, p. 213; Arango 1879, Contribucion Fauna Malacologica Cubana, p. 56. Eutrochatella (Ustronia) straminea ruhromargin- ata (Gundlach) Poey. Wagner 1908, in Martini & Chemnitz, Conchyl.-Cab., (2) 1: sect. 18, pt. 2, p. 122, pi. 23, figs. 22-23. Eutrochatella (Ustronia) wrighti xanthacme Wag- ner 1908, ibid., p. 119, pi. 24 (not 25), figs. 10-11 ( type-locahty, Guajaibon in Cuba; type, IZW, 8.542). Eutrochatella (Troschelviana) (sect. Cubaviana) straminea rubromarginata ("Gundlach" Poey). H. B. Baker 1922, Proc. Acad. Nat. Sci. Phila- delphia, 74: 60. Description. Shell reaching about 8 to 10 mm in height, raised turbinate, rather solid, sublustrous, generally obtusely cari- nate. Color generally white, occasionally with a faint bluish, yellowish or flesh- colored tinge and with a yellow or orange- yellow irregular spot in the columellar region. Whorls about 7, gently inflated, body whorl with a rounded carina, not descending at the aperture; the three earlier whorls strongly shouldered. Suture well impressed, channeled in the earlier whorls. Spire broadly conical, sides some- what swollen, lending a bulbous appear- ance to the lower portion of the .shell. Aperture oblique, roundly triangular, the palatal angle widely rounded, with a pale yellow or yellow-orange spot in the colu- mellar region. Parietal wall moderately inflated, generally marked by the faint spiral cords as well as the axial growth lines, with a raised, lustrous, generally yellow or yellow-orange callus which is widest in the umbilical region. Palatal lip thin, entire, weakly flaring except at ex- tremities, generally white, in occasional populations colored like the callus, shortly angled at columellar insertion. Columella almost perpendicular below, slightly swol- len near base, obtusely angled at insertion Cuban Helicinids • Clench and Jacobson 427 of basal lip. Sculpture of oblique, diagonal growth lines, crossed by spiral cords which are strongest on early postnuclear whorls, becoming weaker, occasionally obsolescent on body whorl and base. Protoconch 1^ whorls, white, rounded, microscopically punctate, sharply raised. Periostracum wanting; shells occasionally encrusted with a thin, amorphous layer of inorganic mat- ter. Operculum as in genus, subopaque, outer layer lustrous, minutely punctate. Internal chitinous layer thin, pale brown, darker marginally. Parietal margin twisted, with a low, rounded lamella, highest near dorsal end. Outer margin thin, slightly twisted. Guane, lectotype of nodae La Furnia, Sierra La Giiira Sierra Galalon Mogote Capon Cayos de San Felipe Viiiales, lectotype of tvrighti Guajaibon, lectotype of rubro- marginata Guajaibon, paralectotype of rubro- marginala Height Width mm mm 8.1 7.2 7.6 6.8 8.8 8.1 9.0 7.4 10.7 9.6 10.4 9.4 8.0 7.2 6.8 6.9 Remarks. This species ranges from Men- doza in the west to San Diego de los Banos in the east, but it appears to be absent from the large intervening area about Luis Lazo, Isabel Maria, and Cabezas (Pi. 6). In this way it almost duplicates the range of Ustronia acuminata and TroschclAjiana chnjsochasma (q. v.). At the Pan de Guajaibon its range overlaps that of T, pyramidalis Sowerby. The shells can be recognized by the rather high conic shape, the generally rounded periphery, the spiral cords oc- casionally covering the entire .shell, and the colored spot in the columellar region. The shells of Ustronia acuminata (Poey) and its subspecies coIumcJIaris (Gundlach) are larger, generally less solid and less glossy, more depressed and ^^'ider in shape, and have less or no spiral sculpture. In addition most specimens of rubromarpe-]()calit>', die Insel Cuba; type, IZW 8572). Description. Shell varying in width from about 6 to about 13 mm, widely depressed conic, moderately shining or dull, thin. Cuban Helicinids • Clench and Jacohson 435 Color generally white to pale yellow, occasional specimens pale reddish, lip white. Whorls from 5^ to 7^, narrowly shelved, moderately rounded or more or less sharply carinate, the carina not reach- ing the area of the palatal lip; body whorl about equal to the spire in \\'idth, not de- scending, or else gently rising at the aperture. Suture well impressed. Spire depressed conic, apex sharply raised. Aper- ture subtriangular, outer angle widely rounded, yellow or reddish internally with a narrow white zone near the palatal lip. Parietal wall somewhat inflated, shallowly excavated in umbilical region; at the colu- mella there is a low triangular callus, with a strongly punctate surface. Palatal lip widely flaring but not reflected, occasion- ally gently undulate, widest above, nar- rower below the periphery, fluted in larger specimens, almost alate at the upper and basal insertions. Columella oblique, scarcelv concave below, moderately con- vex above, set off from the umbilical excavation and the punctate callus by a low, rounded lamella. Spiral sculpture of varying strength, generally somewhat weaker on the base. Axial sculpture of diagonal, irregular growth lines which sometimes lend a wavy appearance to the spiral lines where they cross. Protoconch 1^ whorls, rounded, closely punctate, sharply raised. Periostracum wanting. Cal- careous plate of the operculum of mature individuals thickly set with \o\\', blunt, glassy papillae, smaller and more thickly set than in hoI