nopdnnpeuryenee o— etd eee = paneer one PO ee hg Pk ome pa 4 ~ > oy = . . - vereomseny ie alten ears ; om rn hee a Sr mane Oe ee o Spas ee ant ne re ire POPs Bete treme ~~ a earn ers - nel . > _—. + we ah ea cn enema tO Outen tinea on pe Was rt eee ~— Vartes ieee eetann eons step aie ad OUIy] tYpLaapUR (eummouy) snjpfiszog jo uumjoo & Burmopoy ‘(g “Biyq) UereUI0g UvA epurbn eruixa epag pus “(Z “Biq) wuvUMEN sisuaU \ -ngeb snfna snydhissov0an ‘({ Big) YURIN) tuRUsOON kaUR]SDI 9Y)2)Y :4S210] LANA] OY JO Spalq UB-IOALIP WOUTUTOD 90141, [ 441g ‘ATX “10, HN ‘AY Slatig BULLETIN. OF THE AMERICAN MUSEUM OF NATURAL HISTORY VOLUME XLY, 1921-1922 NEW YORK PUBLISHED BY ORDER OF THE TRUSTEES 1921-1922 THE AMERICAN MUSEUM OF NATURAL HISTORY SEVENTY-SEVENTH STREET AND CENTRAL PARK WEST - New York Ciry BOARD OF TRUSTEES (As of September 30, 1922) PRESIDENT HENRY FAIRFIELD OSBORN Finest Vice-Presipent Seconp Vice-PresipEent CLEVELAND H. DODGE J. P. MORGAN TREASURER SECRETARY GEORGE F. BAKER, Jr. PERCY R. PYNE EX-OFFICIO THE MAYOR OF THE CITY OF NEW YORK THE COMPTROLLER OF THE CITY OF NEW YORK THE PRESIDENT OF THE DEPARTMENT OF PARKS ELEcTIvVe GEORGE F. BAKER ARTHUR CURTISS JAMES FREDERICK F. BREWSTER WALTER B. JAMES THOMAS DeWITT CUYLER OGDEN MILLS WALTER DOUGLAS A. PERRY OSBORN CHILDS FRICK GEORGE D. PRATT MADISON GRANT THEODORE ROOSEVELT WILLIAM AVERELL HARRIMAN LEONARD C. SANFORD ARCHER M. HUNTINGTON JOHN B. TREVOR ADRIAN ISELIN FELIX M. WARBURG ADMINISTRATIVE OFFICERS (As of September 30, 1922) Dmecror Executive SecreTary FREDERIC A. LUCAS GEORGE H, SHERWOOD Assistant TREASURER THE UNITED STATES TRUST COMPANY OF NEW YORK eee SCIENTIFIC STAFF (As of September 30, 1922) Freveric A. Lucas, Sc.D., Director Rosert C. Murpny, Assistant to the Director (in Scientific Correspondence, Exhibition, and Labeling) James L. Ciark, Assistant to the Director (in Full Charge of Preparation) DIVISION OF MINERALOGY AND GEOLOGY W. D. Marruew, F.R.S8., Curator-in-Chief GEOLOGY AND INVERTEBRATE PALZONTOLOGY Epmunp Otts Hovey, Ph.D., Curator Cuester A. Reeps, Ph.D., Associate Curator of Invertebrate Paleontology MINERALOGY Hersert P. Wurriock, C.E., Curator Georce F. Kunz, Ph.D., Research Associate, Gems VERTEBRATE PALZHONTOLOGY Henry Fatrrretp Ossorn, LL.D., D.Sc., Honorary Curator W. D. Marruew, Ph.D., Curator Water GRANGER, Associate Curator of Fossil Mammals Barnum Brown, A.B., Associate Curator of Fossil Reptiles Wim K. Grecory, Ph.D., Associate in Paleontology DIVISION OF ZOOGEOGRAPHY Frank Micaiter Cuapman, N.A.S., Curator-in-Chief LoweER INVERTEBRATES Roy W. Miner, A.B., Curator Wutarp G. Van Name, Ph.D., Assistant Curator Frank J. Myers, Research Associate, Rotifera Horace W. Stunxarp, Ph.D., Research Associate, Parasitology A. L. Treapwe ti, Ph.D., Research Associate, Annulata ENTOMOLOGY Frank E. Lutz, Ph.D., Curator A. J. Mutrcuuer, Assistant Curator in Coleoptera Frank E. Watson, B.S., Assistant in Lepidoptera Joserx Bequaert, Ph.D., Assistant in Congo Zodlogy Cuares W. Lena, B.S., Research Associate, Coleoptera Hersert F. Scuwarz, A.M., Research Associate, Hymenoptera Wituram M. Wuee er, Ph.D., Research Associate, Social Insects iv Scientific Staff v IcHTHYOLOGY _ Basurorp Dean, Ph.D., Honorary Curator Joun T. Nicnous, A.B., Associate Curator of Recent Fishes | — Associate in Ichthyology HERPETOLOGY G. K. Nose, A.M., Associate Curator (In Charge) Artuur OrTENBURGER, M.S., Assistant Curator ORNITHOLOGY FRANK M. Cuapman, Se.D., Curator DeW. Muter, Associate Curator oB 7 CusHMAN Murpny, D.Sc., Associate Curator of Marine Birds oP. Cuaprn, A.M., Assistant Curator, African Birds Griscom, M.A., Assistant Curator NATHAN Dwicurt, MD., Research Associate in North American Ornithology rs. Evste M. B. Retcuenpercer, Research Assistant : MAMMALOGY OY >. Anprews, A.M., Associate Curator of Mammals of the Eastern Hemisphere . E. Antony, A.M., Associate Curator of Mammals of the Western Hemisphere E a9,AnG, Assistant Curator, African Mammals oR Axnsy, Associate in Mammalogy CoMPARATIVE ANATOMY DIVISION OF ANTHROPOLOGY Criark WissLer, Ph.D., Curator-in-Chief ANTHROPOLOGY i Puy E. Gopparp, Ph.D., Curator in Ethnology N.C. Nevson, M.L., Associate Curator of North American Archeology _ Cuares W. Mean, Assistant Curator of Peruvian Archeology _Lours R. Suitivan, A.M., Assistant Curator, Physical Anthropology Crarence L. Hay, A.M., Research Associate in Mexican and Central American Archeology CoMPARATIVE PHysIOLOGY Ratru W. Tower, Ph.D., Curator vi Scientific Stag Companamive Anatomy Wituiam K. Grecory, Ph.D., Curator | J. Howarp McGrecor, Ph.D., Research Asuit in Human Anatomy DIVISION OF EDUCATION, BOOKS, PUBLICATION, AND Grorce H. SueRwoon, AM. comand | LIBRARY AND ‘Posucanoms | ie Raupx W. Tower, Ph.D., Curator Iba Ricuarpson Hoop, A.B., Assistant Lib Grorce H. sits A.M., Curator G. Crype Fisner, Ph.D., sa Rorsa Crospy Noste, "a - PusBLic feast” CHarLes-Epwarp Amory vere D.P.H, Mary Greic, Assistant — 4 . . . . . ’ . . . . \ . . ; . . . . . . . . . . . . . . . . ‘ . . . . . . . . . . . . . . . . . . . . . . . es me Oe = t27 Oe ee ee ae nee P nee ee ee ee a. wae "m REA Se SES Soe 6,5 8S Ae 88,8 Sine ee Be Sb le os 6.8 eee eee ew we x of te Ameria Monn Cong sion A Contribution to ool aie cepa (Plates I-XLV, 47 maps, one DATES OF PUBLICATION of separates is 300 copies, of which about 100 ure sihafled on the Art. 1, pp. 13-269, February 10, 1922. © 1 pp. I-11. 271-1139, October, 25, 1922. LIST OF PLATES I.—Alethe castanea woosnami, Neocossyphus rufus gabunensis, and Bleda eximia ugandz: three common driver-ant birds following a column of Dorylus (Anomma wilverthi. (Frontispiece.) II.—Dorylus (Anomma) wilverthi: temporary nest. II1I.—Dorylus (Anomma) wilverthi: army on the march. IV.—Dorylus (Anomma) wilverthi: dense masses of worker ants covering the larve and pup# among the leaves of pineapple and grass; part of an army with workers swarming over low vegetation. V.—Dorylus (Anomma) nigricans: army crossing a ditch and overwhelming a white rabbit. VI.—Dorylus (Anomma) nigricans: army crossing a path. Megaponera in entrance to nest. VII.—Pheidole saxicola: nests. VIII.—Myrmicaria eumenoides opaciventris: crescent-shaped craters of excavated earth at the entrance of nests. _1X.—Myrmicaria salambo attending scale insects on a tree of the genus Protea. X.—Crematogaster (Atopogyne) depressa fuscipennis: nest. X1.—Crematogaster (Atopogyne) theta: nests. XII.—Crematogaster (Alopogyne) theta: nests. XIII.—Nests of Crematogaster (Atopogyne) theta and Crematogaster (Nematocrema) stadelmanni dolichocephala. XIV.—Crematogaster (Nematocrema) stadelmanni dolichocephala: nest. XV.—Landscape near Niangara, showing hillocks of Termes natalensis. XVI.—Acanthotermes militaris and Pxdalgus termitolestes. XVII.—Macromischoides aculeatus and Tetramorium sericeiventre continentis: nests. XVIII.—Tetramorium setigerum querens: entrance to nest. XIX.—Plagiolepis (Anoplolepis) custodiens: nests on narrow beach of white sand. XX.—cophylla longinoda: nests of leaves woven together with silk. XXI.—Camponotus (Orthonotomyrmez) vividus: interior and outside views of nest. XXII.—Camponotus (Myrmosericus) rufoglaucus syphax: craters of white sand at entrances to nest. Polyrhachis (Myrma) laboriosa: nest. XXIII.—Polyrhachis (Myrma) gagates: nest. XXIV.—Orycteropus afer: freshly killed female. Manis gigantea: freshly killed female. : XXV.—Manis tetradactyla, Manis gigantea, and Bdeogale nigripes, males. XXVI.—Nest of harvesting ant (Messor species) and bushes of Acacia with galls, often inhabited by ants. XXVII.—Branches of Scaphopetalum Thonneri and Cola Laurentii, showing ant- pouches at base of leaves. XXVIII.—Tree and branches of Barteria fistulosa inhabited by Pachysima xthiops. XXIX.—Barteria fistulosa in forest path near Medje. XXX.—Viter Staudtii: transverse sections of a young stem. XXXI.— Vite littoralis and Vitex Staudtii: transverse sections of stems. XXXII.—Viter Staudtii: sections of stems. XXXIII.—Viter Staudtii: sections of stem, tissues, and “nutritive” layer. XXXIV .—Viter Staudtii: sections of stem. \ viii Plates ix XX Il 1.—Cuviera angolensis: sections of normal internodes, myrmecodo- mr eeeeetive bayer, and central cavity. 4 ies; section of stem, portion of ant pellet, portion of detritus, lec Saiiseebaes Lasaeratecibeticnis of taternsde and stean. c rand Plectrenia species sector of a transverse section of a young stem, S Lawrentit: transverse section of quadrangular, hypertrophied “ -myrmecodomatium. ‘ia fistulosa: transverse sections of normal and swollen stem. . a Dewevrei: transverse section of stout stem. Sarcocephalus species: 1 tof pellet from the pocket of Pachysima sthiops; transverse section natium of Barteria Dewevrei; and portion of a coccid (Stictococcus ‘ Sy Bots larvol pellet 6f Peckysima athiops. nents of nutritive layer of heteroplasia in larval pellet of Viticicola tess- ERRATA Page 11, line 10 from bottom, for var. santschii read var. santschiellum. “ee 11, at end of list of new forms insert Dorylophila schwabi (Efulen to Elat, Cameroon), p. 626; and Ocyplanus kohli var. niger (Mful Aja, Cameroon); p. 627. 18, first line, for mend read ment. 21, line 7 from top, for var. molesta read var. molestus. 27, twelfth name from top, for Pseudatta read Pseudoatta. 39, line 18 from bottom, for Ecitini read Ecitonini. 40, line 11 from top, for Ecitini read Ecitonini. 52, line 15 from top, for Cerapachyniz# read Cerapachyine. | 125, lines, 15, 16, and 21 from top, insert article “I” before verb “have.” 125, lines 17 and 19 from top, for Pheidologetini read Pheidologetonini. 125, line 22 from top, for Dacetini read Dacetonini. 165, line 9 from bottom, p. 167, line 11 from bottom, and p. 168, line 10 from top, for nossindambo read nosindambo. 170, line 12 from top, for Pheidologetini read Pheidologetonini. 202, line 10 from top, for okiavoénsis read okiavoénse. 226, line 8 from top, for annectans read annectens. 233, line 3 from top, strike out “‘ Myrmothriz (one species, probably introduced). c This species, C. immigrans Santschi, is now placed in the subgenus Dino- myrmex. 260, line 20 from top, insert “of” between species and Myrma. 281, line 4 from bottom, for Krober read Kréber. 316, line 6 from bottom, for extensive read extensile. : Explanation of Pl. XXII, line 2 from bottom, for emited read emitted. \ BULLETIN OF WERICAN Museum OF NaTuRAL History VoLuME XLV, 1922 59.57,96(67.5) | RiesiROOLOGT OF AFRICA! By Witu1am Morton WHEELER CONTENTS seal EO NC EMO. icv + 69.0: 9 seen Cleredendron LAnngpUB, ...502 sc ses scence ci ceehs sce Oe a a er Eoology of Vélew Stau@lld. 000.0 cccs, ccvcesecsseeee Rubiacem. «ois \s Sine Oa NS eos Sh eee na ae ais ee Uncaria Schreber........ Ecology of Uncaria africana... ..........+-se0ee Sarcecephalue Afnelius. «0.0266. <0660 + cece ss0is +0030 nee Ecology of Sarcocephalus Species... ..........00000005 Randia Houston... .... 050.020 scbaaiac sae 6 6¥peeeeeeenel Ecology of Randia Taga... 5.02.0 ..+00%3 r0enyeereeee Ecology of Randia Ecology of Randia physophylia... é Plectronia Linnaeus. .%s\00 3 ccce ste bs 2s2 634550 ee Ecology of Plectronia Laurentii. .. Ecology of Unidentified African Species of Plectronia .. 474 Cuviera de Candolile,.. 22. 5.5.55 sade paves oa ae we 477 Ecology of Cuviera in the Belgian Congo............... 488 3. Synopsis of Recorded Myrmecophytes...................00.00 494° Pteridophyta.255.65... 0's. 2 ods Foe € ap eRe We oe 497 Polypodiaoee’........ 04000009058 5 u0sh sp bial 497 Monocotyledomem®. : «0... 65 See ee Vii eek eo eaele nn 499 |g I Lr i 499 Orchidacee.. 5... os se Bp ekil es els als are ee 500 Dicotyledones... 2... Gin soneass.505 nn . & Platythyrea gracillima. Avvakubl........c5ccsccesbevccscevsccncs sso hune | Bothroponera pachyderma var. funerea. Medje..............0.0000eeecncneeee Phrynoponera gabonensis var. esta. Medje... obese cleweegels oss 60 5 ean a var. fecunda. Alene: St a a var. umbrosa, Medje... POT a helerodus. Stanleyville.... 5062207 oes 0sscces sentence 78 sd bequaeréé. NgAyu......cccuvc ss 000s ons lees cai sy sunt ae een 79 Buponera (Mesoponera) ingesta. Akenge....i/.6.080i3. os ies eke 82 subiridescens. Akonge:..:.: 0002.0. 60060. se eee 83 Pealidomyrmez obeous., Medje......000<< 5s) sir :nsinsseue oases pene Cannan 92 Leptogenys stuhlmanni subsp. camerunensis var. opalescens. Akenge......... ee ‘“ (Lobopella) ergatogyna. Medije...........-....+ceerececseueseess 195 Anochetus cstus. Aksonge... .....5. 0 soe cveseenesn ce neha da henna { 98 “ apaciventrie. Akemge.....0..50. 60000000 bV labs hve’ pee 698 Odontomachus assiniensis var. furvior. Faradje......................+++----- M01 a “ var. aterrinsus. - NiaQUQ, «0.05: 04s tneeeics eae iP hzematoda var. stanleyi. Stanleyville............................ 102 Tetraponera mocquerysi var. lepida, Faradje..............6.00 0.00 eeeeeeee es 106 Viticicola tessmanni var. castanea. Avakubi..................222ccecsceceeees AAS Pheidole batrachorum. . Akenge........6... 64s 60 bse A ee 128 “ — eaffra subsp. bayeri var. thysvillensis. seer ctis cco on CP w eee ‘ “subsp. senilifrons. Yakuluku.. inal ccuekee he *« minima subsp. malelana. Malela.. «<6 siaieieve ah 6a acne emai « myflognatha. Banans. . oi... .:00:5.5 5 hesoc ss dukes bok pss 134 « = niapuana. Niapu...:......0.5..0? Oe ee Sa or ee 2 onomven | + ++ a h a) pico, be Ia ee ++ ° + uerpeaysny eb. + ES, Spee ma — uendeg Reo ee ies ge! Py eR Sle ek 4 uvcopeurspuy ad bg ge eet 2 So et et eR Seager : yes ae & es | tne Fee See eS Me Tic Ns Rie reese pleas 8 a +l thee” + Bulletin American Museum of Natural History Table I (continued) — General Distribution of the Genera of Ants i i s Lttt. +. +. ie to i: UA +. $+. in aie aly, at sat es ion pres +. t+4+4+4+4+44+.. sn dere ies» ee eens ess +. - +. +. ++ | Australian ep [wordosj00 Ny Pl SU tte nae” Bo Ea Se Rs Se ae rca ae” Se Cae cae Bh be er cde Orjo1BaN Ec Ri BEN Gee Ce NE IEN a Sok ee Ae a eRe ee ket eI aha O1joIBET BT mM “RANR . . . . aC. ces . . . a . me . "mM . . . . . . . . . . “472 Spey easter eee eile Te ae Ya eee OX decd ne een Table I (continued) General Distribution of the Genera of Ants ‘: + oe it + "Wheeler, Ants of the Belgian Congo 26 Bulletin American Museum of Natural History Table I (continued) — General Distribution of the Genera of Ants 4 Archxomyrmex " : + Mayriella : ‘ : ; Promeranoplus ; : ; i Meranoplus + + + + Calyptomyrmex + ; + Te, Macromischa 3 ; Macromischoides + : : 3 Leptothorax - + + neh i Harpagorenus : ; : 4 Myrmozenus ; , 2 Formicorenus ‘ ; : ; Epimyrma y . a ; Symmyrmica ; ; : é Rogeria ; : ; + Lachnomyrmex * : Apsychomyrmex ; - Ss ea? Adelomyrmex ; . | + Ocymyrmex + ; ‘ Tetramyrma + ; : : Tetramorium > ob + oa Rhoptromyrmex oe : F Acidomyrmex ; + ; Strongylognathus : ; ‘ ; Xiphomyrmex +}| +1 +t + Decamorium + ; 4 : Triglyphothrix + 64th + Eutetramorium Z + ; Wasmannia > : ; ‘ Cataulacus + + + | + Zacryptocerus : ‘ : ; Cephalotes ; ‘ ; 4 Acanthognathus : : : 2 [vordor}00 Ny “+H 44+. -+4+4+4+44+4+44+ ONOAvON + onpoawapad + uurpeaysny ie 3 uendeg + oution of the Genera of Ants Table I leonilnedd) “+H +4 4+ 28 Bulletin American Museum of Natural History Table I (continued) pe General Distribution of the Genera of Ants ee : : eS » id nop PIAN Wheeler, Ants of the Belgian Congo | amwee > St Om “ : 4 g°R sist ih be 3 eg peice Ea teeiat le ab Hc Dent Sap Dont~ Qsete ie oneade PPh ee Bulletin American Museum of Natural History (Vol. XLV Table V (ocattinnied) Number of Species, Subspecies, and Varieties of Each Genus Hitherto Recorded from the Ethiopian Region and the Belgian Congo _ Ernrop1an ReGIon Beiotran Conco Species | Subspecies} Varieties | Species | Subspecies| Varieties REma: Ba: S cs ne s : S: @. ao: . * * } * . > Tee ee ee ee ' ; . . ee , ah oe ae ty ee ee Tetramorium Decamorium Xiphomyrmex Telramyrma Rhoptromyrmex Triglyphothriz *Wasmannia “ataulacus Microdaceton Strumigenys Epitritus cn mw Om Se et ee o be | \\ E wm See wae ee Saaonwise pwn wnoernen eee Sew ondarnre _ or —_ oa — o tlk ge ae ola ap. a we OD Wheeler, Ants of the Belgian Congo 33 F Table V (continued) er of Species, Subspecies, and Varieties of Each Genus Hitherto ecorded from the Ethiopian Region and the Belgian Congo Erxioptan REGION Be.ta1an Congo Species | Subspecies | Varieties | Species | Subspecies | Varieties 1 =f a a 10 2 2 8 1 2 10 3 4 1 1 sft 9 2 4 6 1 1 1 1 ‘ 26 3 8 6 1 1 20 il 16 5 2 4 13 a 1 5 — 4 ae, 1 3 a 1 2 4 2 oa ee 2 re 4 1 45 3 oF 81 76 37 33 19 1 ? 1 G3 le 37 17 22 23 9 11 920 389 540 318 122 158 ri ‘THe PECULIARITIES OF THE MaLaGAsy ANT-FAUNA ne a, Region includes Madagascar and a number of small _ neighb islands known as the Comoros, Seychelles, and Chagos. ‘ he ant-fauna of Madagascar was first studied by Forel in a splendid ume in Grandidier’s large work on the physical and political history : ’ island. More recently, the Swiss myrmecologist has contributed . on the ants of the smaller islands. Turning again to Tables I, IT, 1 III, we note the following facts. 1.—The Malagasy ant-fauna comprises 40 genera, somewhat less than ‘the number known from the Ethiopian Region, but only four of ese (Champsomyrmex, Parapheidole, Brunella, and Eutetramorium) are ndemic, or precinctive. _ 2.—Of the 40 genera, 34 are common to the Ethiopian and 32 to the eeaayes so that the affinities appear to be about equally divided ' een these two regions. 34 Bulletin American Museum of Natural History [Vol. XLV 3.—The generic affinities with the Papuan and Australian Regions are somewhat less pronounced (29 and 26 respectively), but considerably more than with the Palearctic, Nearctic, and Neotropical (20, 18, and 19 respectively). 4.—The subfamily Doryline is completely absent from the Malagasy Region. 5.—The Dolichoderinz are poorly represented by two genera. 6.—The Formicine are represented by only four genera (Acantho- lepis, Plagiolepis, Prenolepis, and Camponotus). 7.—The Malagasy possesses only two genera (A phznogaster and Vollenhovia) which are not known to occur in the Ethiopian Region. 8.—On the other hand, there are 13 genera (Discothyrea, Centromyr- mex, Ectomomyrmex, Dorylus, Anictus, Myrmicaria, Carebara, Pedalgus, Calyptomyrmex, Semonius, Pseudolasius, C2cophylla, and Polyrhachis) which occur in the Indomalayan and Ethiopian Regions but are not known to occur in the Malagasy. 9.—Three peculiar genera (Simopone, Melissotarsus, and Terataner) are known to occur only in the Ethiopian and Malagasy Regions. The following remarks on particular genera are of general interest. Of the four endemic, or precinctive, genera, Champsomyrmex, Para- pheidole, and Brunella are monotypic and Eutetramorium contains only two species. Champsomyrmex is very close to Odontomachus; Para- pheidole seems to be very close to and parasitic on Pheidole; the species of Brunella was originally described as an Aphzxnogaster, and Eutetra- morium, as the name indicates, is allied to Tetramorium. These four genera, therefore, lend nothing very striking to the complexion of the Malagasy ant-fauna. Its distinctive features are due to the peculiar development of species within certain genera which it shares with the faunas of other regions. Among the Ponerine, the very ancient and primitive genera Simo- pone with 3 and Mystrium with 5 species show a greater development than elsewhere, though the former occurs also in Africa and the latter both in Africa and the Indomalayan Region. The only known species of Euponera, sensu stricto, (E. sikore Forel) is found in Madagascar. Leptogenys is beautifully represented by no less than 16 species, includ- ing three species of an endemic subgenus, Machexrogenys. The remaining genera of the subfamily show nothing unusual. Among the Myrmicine, we find Tetraponera represented by 12 species and Aphznogaster, which does not occur in the Ethiopian Region, by 2, one of Aphzenogaster, sensu stricto, and one of the subgenus Wheeler, Ants of the Belgian Congo 35 my emneoentter has 21 species, distributed among the follow- subgenera: Crematogaster, sensu stricto, 11; Orthocrema, 1; Oxygyne, 5; sacrema, 4. on of the species of Vollenhovia (levithorax Emery) and single s ss of Triglyphothrix, T.striatidens (Emery), are really - ramp from the Indomalayan Region. Pheidole is well ited by 17 species, some of which have a peculiar habitus. The hala group comprises a number of forms as in the Ethiopian Terataner with 5, Xiphomyrmex with 13, and Cataulacus with sare unusually well represented, considering the small size of the i olichoderine, as previously stated, have a very poor repre- i tion _Tapinoma by only one species, melanocephalum (Fabricius), a mon tropicopolitan tramp, and Technomyrmex by 4 species, 3 of re confined to Madagascar, while one, albipes (Smith), is widely sd over the Indomalayan and Papuan Regions. b greatest representation of Formicine is furnished by the genera : with 11 and Camponotus with 47 species. The latter genus is kable on account of the great number of subgenera represented ponotus, sensu stricto, with 1; Myrmoturba with 4; Dinomyrmex 5; Myrmosaga with 7; Myrmosericus with 1; Mayria with 1; 'yrmot with 2; Caoboveis with 1; Mernincsdis with 6; aes »pytia with 1; Mirmorkachis with 1; Orthonotomyrmex with 15; yrmosaulu with 1: and Myrmopiromis ath! spcctie) The subgenera i iurincenga, Myrmonesites, and Myrmopytia are confined to adagascar. The single species of Brachymyrmezx, B. cordemoyi (Forel), wn to occur in the island of Réunion, has been introduced by com- ce from South America. In Table VI the Malagasy genera of ants ding to Forel,! the fauna is made up of groups of species having the »wing provenience and affinities. - A.—Imported Forms RU iiihats Serias, inapurted cm various octesioas by ships i eg een 8 _ American forms, evidently of recent importation: Brachymyrmex cordemoyi 7 NE 2 _Of recent importation from Indomalaya: Plagiolepis aieaseies (Jerdon) . . 1 Si oe es oncient Ln Strumigenys godeffroyi Mayr... ‘La faune des fourmis et ses rapports avec les faunes de I’ Afrique, de I'Inde, de e, ete.” pag cog Fo XV, pp. 1-6. 36 Bulletin American Museum of Natural History [Vol. XLV B.—Malagasy Forms 7. . 8. With Indomaleyan affimition os... ois... ccc cede cues soe et t+ eee a? ee 9. With Ethiopian affinities. . 9 10. With very distinct Australasiatic (OMoluccan and ‘Australian) affinities. “To these many might be added from the following group which, on the whole, have Moluccan and Australian affinities............... Ae coe & 11. Malagasy forms proper... Mey Total. . nity sticace #-tSaillvees 5 + stein ll iis ge .. 238 Forel summarizes his views on the Malagasy ant-fauna as follows: To sum up, the local Malagasy fauna is a fauna of extremely ancient relicts, which have been evolved in certain intercontinental groups (Camponotus, Pheidole, Crematogaster, ete.) to form a very peculiar fauna, the most ancient and primordial affinities of which connect it with the ancient fauna of the Moluceas and Northern Australia. But whereas the Indomalayan genus Polyrhachis has invaded Australasia and Australia, it no more exists in Madagascar than does the subfamily Doryline. Subsequently, invasions from East Africa and India confused the situa- tion; still analysis is possible, although it is sometimes difficult to distinguish the direction of the invasions, especially that of the Malagasy fauna into Africa and India. Finally, very recent invasions of cosmopolitan and even of American species, introduced, without doubt, by shipping, have still further complicated the situation, especially in the small Malagasy archipelagos and along the coast. Nevertheless, it is on the whole easy in these cases to detect the invasions and to avoid erroneous interpretations. The genus Brachymyrmex admits of no doubt, and an eye-witness, M. Vinson, of St. Denys,. was able to give me-exact information, through M. de Cordemoy, on the invasion of Plagiolepis longipes into Réunion some twenty-five years ago. With cosmopolitan affinities (varieties of intercontinental species)... . . re 9 = \y 37 Table VI a, a ‘Wheeler, Ants of the Belgian Congo | _Recorded from the Malagasy Region sties of Each Genus Hitherto x) rf ts tt “NANN oR 200s Ota? —e = “gif zs 3 * ‘ F : et Oe eR er ri 09) Fe ert (AN i ange te 1s A S { ° ri ee eee ae ee ae ee ee a Ste E ANTS COLLECTED BY THE AMERICAN MUSEUM CONGO . _— > EXPEDITION By Wa. M. WHEELER Doryline KER AND Souprer.—Clypeus as a rule very short and not limited by sutures. caring vertical, not covering the insertions of the antenne. Antenne in- the mouth and close to each other, often less than 12-jointed. Palpi at ed, in Leptanilla only one-jointed. Ocelli and eyes often absent (without n all African genera). Sutures of the thorax more or less vestigial; mesono- hing the epinotom on the dorsal face, without interposed metanotum. Spurs tibia pectinate or rudimentary. Postpetiole not always separated by a con- n from the third segment; however, in Eciton, Anictus, and Leptanilla, into the second joint of a two-jointed pedicel. Sting developed. =.—Permanently apterous, with the abdomen much enlarged and swollen; fferent morphologically from the worker. Clypeus as in the worker. Frontal more or less separated. Antenne 10- to 12-jointed. No ocelli; eyes not more M d than i in the worker; female blind when the worker is so. Segmentation of * more or less rudimentary; no traces of wings or a rudiment left at the aposed of one segment. Gaster long and voluminous. Seiepews and frontal carine much as in the female. Mandibles de- l, asarule large; in Leptanilla very short. Antennz 13-jointed; scape long, in | only slightly longer than the second joint. Eyes and ocelli well developed. -with normal segmentation, winged. Postpetiole and pedicel much as in the . Genitalia completely retractile (Dorylini and Ecitini) or exserted and not e (Leptan'llini); subgenital lamina split or furcate; cerci absent. Larv# more or less cylindrical, with short hairs, without hooked sete; mandibles , slender, faleate. - winged males were the first to be known and were originally lace i with the Mutillide. The workers and females were recognized as ants but at first classified in genera by themselves. Though their rela- jions were more or less suspected by Lepeletier de Saint-Fargeau, Hali- and Shuckard, the true affinities of the male and worker became oly gradually known after 1850, when Savage observed for the first time West Africa Dorylus males walking in an army of Anomma workers. The females, leading a permanently subterranean life, are still excessively ‘in collections and known only for a few species; their capture in the jaller species is rather fortuitous, whereas in such fierce army ants as 2 it is a very troublesome operation. 30 40 Bulletin American Museum of Natural History (Vol. XLV G. Arnold' gives the following general account of the habits of this subfamily : The members of this subfamily are commonly known as driver or legionary ants. The males, which are winged and provided w:th eyes, are frequently taken at lights; on the other hand, the workers are blind, with the exception of some species of Eciton, in which there is a pair of single-faceted eyes, and the females (excepting one species of Eciton) are both blind and wingless. The members of the genus Dorylus are almost entirely subterranean in their mode of life, rarely coming to the surface except in dull, cloudy weather. The species of the subgenus Anomma, which live in the more ' tropical and forested regions of Africa, and to which the term driver ants was origin- ally applied, and the Ecitini of South America, are , however, usually seen above the surface, although, should the rays of the sun prove too powerful, they will construct temporarily tunnels with particles of earth held together by their saliva. The species of #nictus are not so shy of the light and may be seen foraging about even in bright sunlight. It is probable that all, or at least the majority of the species are carnivor- ous, although D. orientalis has been shown by Green to feed also on tubers and the bark of trees. As far as known the members of this subfamily do not as a rule make permanent nests. This course is determined by their exceedingly predatory habits, which com- pel the adoption of a migratory form of life together with the formation of temporary nests in localities which are sufficiently productive of animal life to detain them for any length of time. Ranging far and wide in search of prey, which consists of any animal they are strong enough to overpower, these ants must sooner or later exhaust the areas round their nests, and are forced to remove the latter to new and more pro- ductive hunting grounds. But little is known of the habits of the Leptanillini; all species are hypogzic. Santschi found the nest of Leptanilla nana Santschi 40 cm. beneath the surface in clay soil; he caught females and workers by in- undating the soil so as to force them to come out of their burrows; work- ers have also been taken by sifting decayed leaves. The males are attracted by lights. A detailed account of the migrations and habits of some of the African species is given below (see under Dorylus bequaerti, D. opacus, D. kohli, D. nigricans, D. wilverthi, and D. fulvus). The Doryline are abundantly found in all tropical parts of the world, with the exception of the Antilles and the Malagasy Region; they are absent from the larger part of Australia. A few species reach North Africa, the coasts of Asia Minor, and the central and southern United States. 11915, Ann. South African Mus., XIV, p. 110. Wheeler, Ants of the Belgian Congo 41 Dorytus Fabricius ‘Workers small or of medium size, without eyes or ocelli, highly polymorphic, ‘ituting a series of forms which may be grouped as maxime, or soldiers, medize maxima the head is very large and usually broader in front than d, the mandibles are long and narrow, with a small number of teeth on the cider, the clypeus is very short and not marked off from the remainder of the by autres Frontal carine very short, erect, close together, not concealing Dn of the antennz. Antennz short, inserted very near the mouth, 9- to 12- d Stillen’ 40 the mpecies. Mediz smaller, with much smaller and shorter , but the latter not narrowed in front; anterior border of clypeus more or less ecting in the middle over the mouth. Antenne as in the maxima. Minima very ll, with the head narrowed anteriorly and the anterior border of the clypeus projecting i in the middle. Number of antennal joints reduced, seven being ¢ minimum. Promesonotal suture distinct in all three forms of worker; meso- ‘suture obsolete. Epinotum always unarmed. Petiole nodiform; post- : narrowed anteriorly, not or only indistinctly separated from the first gastric me Pygidium with a dorsal impression and terminating in three points. Pos- wr tibie each with a pectinated spur. Femaue very much larger than the worker, dichthadiiform, i. e. wingless, with and voluminous abdomen. The head has the occipital lobes swollen and rounded, ed by a median longitudinal furrow. Eyes and ocelli absent, as in the workers. ; as in the worker maxima, or soldier. Mandibles narrow, edentate. Antenne ated (12-jointed in the subgenus Dichthadia). Thorax segmented, but the without differentiated scutum and scutellum; alar insertions vestigial. vas apa its posterior corners prolonged as blunt points. Postpetiole shorter an the first gastric segment, but not followed by a constriction. Pygidium and gaping or separated so as to expose to view the eighth pair of abdominal , the anal segment and sting; the pygidium not impressed; the hypopygium the pygidium considerably and terminating in two lobes or appendages. _ Mate very large, with very large eyes and ocelli. Clypeus short, prolonged back- ‘ Salant Between the short, diverging frontal carine. Mandibles edentate. Antenne _ 18-jointed; scape one-third or one-fourth as long as the funiculus which is filiform. _ Legs short; femora flattened, tibie narrow. Wings with narrow, poorly defined _ pterostigma, placed near the apical third; radial cell elongate and open; one closed al Ceubiial cell, usually one recurrent nervure (two in the subgenus Rhogmus and in some _ anomalies). Petiole nodiform or saucer-shaped, its concavity turned toward the _ postpetiole, the latter not separated from the gaster by a constriction. Gaster long, ¢ylindrical or club-shaped. Pygidium rounded or split at the posterior border (Rhog- __ mus fimbriatus). Genital armature voluminous, completely retractile; annular _ lamina narrow; stipes and volsella simple; lacinia absent; subgenital plate deeply fureate. Emery, who has devoted much careful study to the Doryline, divides Dorylus into six subgenera (Dorylus, sensu stricto; Dichthadia Gerstecker; Anomma Shuckard; Typhlopone Westwood; Rhogmus _ S$huckard; Alaopone Emery) mainly on the number of antennal joints _ and structure of the pygidium in the worker, the number of antennal 42 Bulletin American Museum of Natural History [Vol. XLV joints and shape of the hypopygium in the female, and the shape of the mandibles and petiole in the male. The genus (Map 4) occurs through- out Africa, India, Indochina, the Malayan Region, and Indonesia (Borneo, Java, Sumatra, and Celebes). All but one of the subgenera and most of the species are found in Africa; in Asia there are less than half a dozen species belonging to the subgenera Dichthadia, Typhlopone, and Alaopone. In the ‘Genera Insectorum’ (Doryline, 1910, p. 7) Emery makes the following statement on the ethology of the genus Dorylus: is Y) Ss w» 45 60 i] 90 105 120 as 160 | a es el ca a 7) 45, a4 ry ae ¢ = wa oo ABS S05! A aie * s - . —8 ‘. TS SSEERAHB WY Ss L_--- 8 ° 2 \ °. x Ts. “Ss ° po on “~ 0 4 ° Pai £ ‘\ Ve a ” 1 - ~ 1s ‘ AE ae a eo PN es 7 coieieaieais Aeukesiesienens 5 trae e \ Bh ¢ u Te * ? is 0 is » 4s 60 5 90 105 120 ms 160 Map 4. Distribution of the genus Dorylus. Apart from the subgenus Anomma all the species of Dorylus lead a subterranean life and come to the surface of the soil only on exceptional occasions, as, ¢. g., during inundations or in order to accompany the males when they take flight. Their societies are very populous. The soldiers and workers make subterranean expeditions for the purpose of capturing insects and other small animals, and exploit manure piles, cadavers and probably also the nests of termites. The males come to lights at night. Search for the heavy and voluminous apterous females is beset with difficulties so that they are rare in collections. It may be noted that in all the specimens hitherto described, with the exception of the female of D. fimbriatus described by Brauns, the terminal tarsal joints are lacking. I infer that the workers tear them off during the underground forays, while they are dragging the colossal queen by all her legs through the narrow galleries. Wheeler, Ants of the Belgian Congo 43 Dorylus atratus F. Smith ae from Stanleyville (Lang and Chapin). " —— brevipennis Emery variety marshalli Emery Pa. _— from Medje (Lang and Chapin). Dorylus bequaerti Forel to this species, originally taken by Dr. Bequaert at Sankisia ‘Katanga, numerous workers from two colonies, one taken by Mr. Ra edi, the other by Dr. Bequaert at Pasaconde near Zambi leries under ground and in a fallen trunk of Hyphzne.” The larg- orke a of the former colony are only 4 mm. long and therefore some- ts aller than those seen by Forel (5 mm.) and the color is paler. th e probably not the largest workers of the colony. The t indi s taken by Dr. Bequaert are fully 5 mm. long and ei n aloe. The head is deeply and broadly excavated behind and rht, subparallel sides; the first funicular joint is distinctly longer ad, the remaining joints, except the last, broader than long, and petiole is also slightly broader than long. The whole body is evenly sharply punctate, the punctures on the gaster somewhat smaller but listinct. The large workers are rich ferruginous red, with some- t paler gaster; the smaller workers are decidedly paler, like those Mr. Lang at Banana. * Dorylus depilis (Emery) FParadje, 7; Medje, @; Stanleyville, @ (Lang and Chapin). yen specimens, all belonging to the typical form of this well-known Dorylus mestus (Emery ) iaakte from Stanleyville (Lang and Chapin). | Dorylus staudingeri Emery single male from Medje (Lang and Chapin). ‘Dorylus (Anomma) emeryi Mayr subspecies opacus Forel __ A fine series of workers of all sizes from a single colony taken at Ngayu (Lang and Chapin). “They appeared during the night, appar- ently attracted by some bones of large mammals, which they completely vered.”” The sides of the head of the largest workers are less convex an indicated by Santschi’s figure and like that which he gives of D. 44 Bulletin American Museum of Natural History [Vol. XLV emeryi, though slightly narrower and much more deeply excavated behind. The preapical tooth of the mandibles is lacking in the largest, though present in the medi and smallest workers. There are also three workers from Medje, taken from the stomach of a toad (Bufo funereus). Dorylus (Anomma) funereus Emery Medje, o; Stanleyville, @; Bolobo to Lukolela, @ (Lang and Chapin). Single specimens from each of these localities agree —_ with Emery’s description of the types from the Gold Coast. Dorylus (Anomma) kohli Wasmann Twenty workers from Akenge and Niangara, taken from the stomachs of toads (Bufo funereus) and frogs (Kassina senegalensis and Hemisus marmoratum), and a fine series of workers of all sizes from Avakubi (Lang and Chapin) with the following note: “They usually appear in great masses, coming right out of the ground, underneath a piece of meat. Even palm oil, poured on the floor, will attract them in the same way.” This observation shows that the species is hypogeic like the species of Dorylus, sensu stricto, and not epigwic like Dorylus (Anomma) nigricans and its various subspecies and varieties, and agrees with the observations of Father Kohl, quoted by Wasmann: “This species seems to be intermediate between the subterranean sensu stricto, and the driver ants. Its discoverer, Father Kohl, wheleitad it at St. Gabriel near Stanleyville on the Upper Congo, writes as follows: ‘The ants just mentioned seem always to wander about beneath the surface of the ground; at any rate, I have seen them on the surface only on three occasions and always after a rain.’”” Wasmann adds the interesting statement: ‘‘The subterranean mode of life of D. kohli may also be inferred from its guests, which are much less like those of Anomma than of Dorylus helvolus L. The development of the eyes of Pygostenus pusillus Wasm., which lives with D. kohli, is about half way between the small eyes of P. raffrayi Wasm., a guest of D. helvolus L., and the very large eyes of the Pygostenus species which live with Anomma wilwerthi Emery. Here, too, there is a hint in regard to the habits of the host.’”’ The remarkable wingless phorid Hexacantherophora cohabitans, recently described by H. Schmitz, was also found with Dorylus kohli _ by Father Kohl at St. Gabriel near Stanleyville. 11914, Zool. Jahrb. Abt. Syst., XXX VII, pp. 512-515, Pl. xxrx, fig. 1. Wheeler, Ants of the Belgian Congo : 45 7 Dorylus (Anomma) kohli variety congolensis Santschi _ Two series of workers, one taken at Leopoldville by Mr. Lang, the other at i Dr. Bequaert, evidently belong to this variety, in hich the head of workers measuring 7 mm. is as broad as long, whereas n the typical kohli it is longer than broad in individuals of the same size, ‘somewhat less pointed posterior angles. _ The Leopoldville specimens were found “under a piece of tin on ® shore of Stanley Pool,’ those from Thysville were “marching in a nean burrow in a forest gallery.” ie Dorylus (Anomma) kohli variety langi, new variety __ A series of more than a hundred workers from Malela (Lang and Chapin), taken beneath the prostrate trunk of a palm, represent a new sty near variety frenisyi Forel and variety minor Santschi. They ige in size from 3 to 8 mm. The largest are very probably the true xima workers as they lack the preapical mandibular tooth. In frenisyi » largest workers attain a length of 8.5 mm., in minor 8 mm. - The head of langi is nearly as broad as long, its sides convex and distinctly con- ‘ behind so that the occipital border, which is deeply and rather angularly isasieed, fo about three-fourths as long as the anterior. The dorsal and ventral surfaces of the head are somewhat flattened. The whole body is finely, sharply, and rather uni- rmly shagreened or minutely and densely punctate and subopaque; the mandibles yth and shining; the gaster behind its first segment feebly shining. The upper of the head, thorax, and gaster are uniformly but sparsely punctate, the punc- tures nonpiligerous for the most part. The suberect, yellow hairs are very sparse and confined to the gaster and the same is true of the dilute appressed pubescence. Legs and _Scapes with short stiff and appressed hairs, absent or very sparse on the extensor ‘surfaces of the femora and tibia. In some specimens a few very fine short hairs can » detected, under a magnification of 20 diameters, arising from the coarse punctures . the vertex or posterior corners of the head. Color rather bright reddish ferrugin- gus, with the legs paler and the mandibles and the upper surface of the head, except 2 the cheeks and occiput, dark brown or blackish. The upper surface of the thorax ee. except the posterior borders of the segments of the latter, are darker and e brownish than the pleure and venter. The petiole is scarcely longer than its ventral tooth small, compressed and directed backward. The smaller aa head cf nearty the cxase chaye an proportions on the larger but low _ deeply excised behind and more shining, as is also the body. The pubescence is also a little more abundant. The color is very similar but paler in the smallest individuals. Dorylus (Anomma) kohli variety chapini, new variety This is a very distinct form, represented by a series of two dozen - workers from Stanleyville (Lang and Chapin), without further data. measure 1.5 to 6 mm. in length. The largest specimens are prob- _ ably not the maxima forms as they have a preapical mandibular tooth. us = 46 Bulletin American Museum of Natural History [Vol. XLV The body is only slightly shining and very similar in sculpture to the preceding variety except that the punctures are coarser, sharper and piligerous. They are evenly distributed over the dorsal surface of the head and pronotum, similar but smaller and shallower on the epinotum and gaster, and very indistinct or absent on the petiole. Mandibles and legs smooth and shining. The head, pro- and mesonotum, gaster, scapes, and legs are covered with short, subappressed, yellow hairs arising from the punctures and forming a conspicuous, rather abundant, coarse pubescence. The body is brownish ferruginous, the head slightly darker, and appendages paler, the mandibles blackish. The head is scarcely longer than broad in front, the sides very feebly convex and converging to the posterior border, which is only slightly excised and about four-fifths as long as the anterior border. The petiole is as broad as long. The smaller workers closely resemble the larger, except that the head is a little longer and the color paler. Dorylus (Anomma) nigricans Illiger subspecies arcens (Westwood) Eleven maxima and media workers from Medje (Lang and Chapin), taken from the stomach of a toad (Bufo funereus), are very dark, almost black, and are evidently referable to this subspecies, though the largest specimens are only about 10.5 mm. long, whereas the largest workers, according to Emery and Santschi, measure 13 mm. The surface of the body is very shining, the head more opaque in front. Dorylus (Anomma) nigricans subspecies burmeisteri (Shuckard) Seven workers from the stomach of a toad (Bufo regularis) taken at Stanleyville; a series of workers of all sizes from Stanleyville and Lukolela to Basoko (Lang and Chapin); also workers from Katala (J. Bequaert). Dorylus nigricans is the famous driver or army ant, which has so greatly impressed all the African explorers. In my ant-book I have quoted some of the accounts of the earlier observers. To the field naturalist the various races of D. nigricans and D. wilverthi are so similar in appearance and habits that he designates them all as “driver” or “army” ants. It is not surprising therefore that Mr. Lang’s notes refer indifferently to both species. The four fine photographs (Pls. II, III, and IV) belong undoubtedly to D. wilverthi (vide infra) but the following note probably refers to both species: ‘‘Wherever they go, even though the file be very small, the army ants clear a road that can be easily seen. But when a large army is passing, they not only build a road but also bridges and frequently even fill in all the depressions between the dried grass with particles of sand or soil until a perfect road has been constructed. Across a pathway used by pedestrians, where they are often disturbed, they build walls and regular tunnels even in Wheeler, Ants of the Belgian Congo 47 hart ground. Particle by particle is carried out by the steady aug small workers and the soldiers, large and small, watch on both eran; ever ready to attack anything that may approach. ey a 2 a very peculiar attitude, with mandibles wide open and red thorax bent up and back till it forms a right angle with the ( =. . When they seize anything, the abdomen can be torn off 1out their loosening their grip. They are greatly feared by the natives n the greatest laggard moves rapidly when passing ‘the line.”’’ connection with the fact cited by the early explorers, that the are able to kill large animals when confinement prevents their , Santschi’s quotation of the following observation of Cruchet : es ig D. nigricans in Benguela is of interest: “Twice during the se of the year we have been compelled to take the cows out of the 1 and drive them elsewhere, because they bellowed so piteously. aking into the matter we found that the Anommas caused all this ce by crawling into the natural orifices of the animals, especially > a un Baad vulva. A brooding hen had her head half eaten away, but would i not abandon her eggs. On three occasions one of my comrads ha mh 48 Bulletin American Museum of Natural History [Vol. XLV Dorylus (Anomma) nigricans subspecies sjestedti Emery Three large workers from Faradje, Niangara, and Medje, taken from the stomachs of toads (Bufo regularis and B. superciliaris) and a frog (Rana occipitalis); also a large series of workers from Faradje (Lang and Chapin). This form closely resembles subspecies burmeisteri variety molestus (Gerstecker) in having the inferoposterior angles of the petiole prolonged outward as distinct tubercles, but is readily distinguished by having the head of the larger workers (7 to 12.5 mm.) opaque instead of shining and that of the smaller workers elongate. An interesting account of the habits of rubellus and sjestedti has been published by Sjéstedt.' Dorylus (Anomma) wilverthi Emery Plates II, IIT, and IV This fine species, the workers of which are easily recognized by the elongated and divergent posterior corners of the head (Fig. 1b), is repre- sented by a large series from Avabuki and a single small worker from Fig. 1. Dorylus (Anomma) wilverthi Emery. a, head of soldier; b, head of worker. Faradje; also by five workers from Medje and Akenge taken from the stomachs of toads, Bufo polycercus and B. funereus (Lang and Chapin). The temporary nest is shown in Plate II, the ants massed on the ground in Plate III. Concerning these ants Mr. Lang says: “We had considerable trouble with them, for they started a nest near our camp at the base of a coffee bush where some pineapple plants were growing. I took two photos before burning the place. One shows the masses of army ants heaped on top of the other. It was impossible to see what they — had beneath them, but after the fire, we found that they covered in- 11908, 'Akaziengallen und Ameisen auf den ostafrikanischen Steppen.’ In Sjéstedt, Exped. Kili- mandjaro, Meru, etc., II, 8, pt. 4, pp. 111-114. ial bi ——— —o 1922) Wheeler, Ants of the Belgian Congo 49 numerable eggs and larve. The other photo shows the mounds or heaps _ of earth particles carried out by the workers. They come on steadily, each one with a particle of soil in its jaws, and, as soon as they arrive at the summit of the mound, they open their mandibles and the grain of sand rollsinto place. After the fire they began to emigrate in enormous _ numbers, building their roads as they proceeded. There was one main line about an inch wide, excluding the soldiers. I followed this particular line for a distance of about 500 yards into the forest. Sometimes the ants seemed to have disappeared entirely into the ground, since they traveled in tunnels, but by searching I discovered their course some dis- tance beyond. I was unable to ascertain where the huge army deposited its eggs and larve. For three days the workers carried larve and eggs out of the old nest. The brood was carried under the body so that it could not be seen by the superficial observer.”” These observations were made at Avakubi. _ Dorylus (Typhlopone) fulvus (Westwood) subspecies badius (Gersteecker) variety obscurior Santschi Vankerckhovenville, 4,8, o&%; Faradje, a, %; Garamba, 8; -Batama, @; Stanleyville, @ (Lang and Chapin); Avakubi, 7 (Lieut. Boyton). Both the worker and male of this form have a characteristic color. Santschi described only the worker from Konakry, French Guinea. The Congo specimens measure 5 to 13 mm. and have the head, thorax, petiole, and legs rich chestnut brown, the gaster brownish yellow, the mandibles and antennz nearly black. The smallest workers are more uniformly brownish yellow. The differences in form between this and the typical dadius of South Africa are slight. Santschi describes the head, the base of the epinotum, and the petiole as broader in obscurior. In my specimens the head of the soldier (Fig. 2a) closely resembles that of the variety eurous from East Africa as figured by Emery. The males (Fig. 2b-f) taken from the same colony as the workers are also much darker than those of the subspecies badius and variety eurous or the typical fulvus from North Africa. They measure 33 to36mm., with ___ the thorax somewhat less than 6 mm. broad, and are chocolate brown, with the head blackish and the gaster a shade paler than the thorax and petiole. The wing membranes are also of alittle duller and deeper tint. The hairs and pubescence are less golden and less shining, more grayish. The male genitalia are intermediate in the structure of the stipes between those of the typical fulvus and the subspecies badius, as will be seen by 50 Bulletin American Museum of Natural Hist ry (Vol. XLV Fig.2. Dorylus (Tuphlopene) fulrus subspecies badius variety obecurior Santschi. a, soldier; b, head ay male: e-f, genitalia of male. comparing my figures with Emery’s.' The specimens from Batama and Avakubi are distinctly paler than the others in the series but ean hardly be regarded as belonging to a different variety. Concerning the Vankerckhovenville colony from which both workers and males were taken, Mr..Lang writes: “These ants were collected on the floor of an Azande hut. The workers and big males were swarming out of a hole in the ground during the night. These driver ants are not + annoying to human beings, but have subterranean habits. They never 11895, Zool. Jahrb. Abt. Syst., VIII, figs. Q and R, pp. 727, 728. ae ————— —.. or - \ oe ee ae eee Se eee - 4 Wheeler, Ants of the Belgian Congo 51 in columns on the surface like the others, but whenever a piece of bbap a jar of oil is deposited on the ground they immediately m below, without a tunnel ora gallery being visible from the = - , us (Alaopone) atriceps Shuckard Text Figure 3 e nae from Faradje and two y nley » (Lang and Chapin). Fig. 3. Dorylus (Alaopone) a atriceps Shuckard. Head of male. Dorylus (Alaopone) conradti Emery ery’s specimens attained a length of 6.5 mm. The soldier is easily zed by the coarsely punctate thorax and the very elongate head, h the closed mandibles, is nearly twice as long as broad. Cerapachyine have recently proposed to regard Forel’s tribe “Cerapachysii’’ as uting an independent subfamily, the larvze of these ants being so 1t from those of the true Ponerine and much more like the larve ba Doryline: The limits of this new subfamily agree with those of ieery ’s section Prodoryline, and Emery was probably right in contend- Bq MOnKER caste has a ponerine habitus, but is often long and slender. The postpetiole is separated from the third abdominal segment ya » well-marked constriction, and as broad as the third segment. In he Indoaustralian Eusphinctus even the gastric segments are marked off one another. A powerful sting is present. e characters of the FEMALE in the various genera are peculiarly ive In some cases (Phyracaces), this caste is winged and not unlike females of certain Ponerine; in others (Parasyscia, Eusphinctus), emale is wingless and ergatomorphic; and, in still others (Acantho- J tens Wm. M., 1920. ‘The subfamilies of Formicide, and other taxonomic notes.’ Psyche, 52 Bulletin American Museum aj Natural History [Vol. XLV stichus, Nothosphinctus), the female is so much like the corresponding caste in the Doryline that it might be regarded as a dichthadiigyne. The MALE, on the other hand, though lacking the cerci, has a decidedly ponerine habitus. The male genitalia are completely retractile; the subgenital lamina deeply and broadly furcate. The LARV# are extremely like those of the Doryline; they are elongate and almost cylindrical, uniformly covered with short hairs, and without piliferous tubercles. The mandibles are small, narrow, pointed, and rather feebly chitinized, and I have failed to find a trophorhinium, or triturating organ in the mouth. Apparently the young are fed only on soft food. Moreover, the foraging habits at least of certain Australian Cerapachyine (Phyracaces) resemble those of the Dorylinz.' Dr. W. M. Mann has recently sent me specimens of his Cerapachys majusculus from Fiji, with several worker pupe which are enclosed in well-developed, brown cocoons. The Cerapachynie seem, ae to agree with the Ponerine in this character. Cerapacnys fF. Smith Worker.—Small ants with peculiar, long, subcylindrical body; the head ex- cavated behind, with prominent, depressed posterior corners and very short clypeus, with which the closely approximated frontal carinw are fused. The latter are erect, leaving the articulations of the antennw exposed. The antennal fovea is bounded externally by a distinct carina. Mandibles with distinct, obscurely denticulate apical border. Antenne stout, 9- to 12-jointed, the scape incrassated distally, the terminal funicular joint large, swollen, oval or glandiform, at least as long as the three preceding joints together, thus forming a one-jointed club. Eyes small, sometimes wanting. Thorax with the promesonotal and mesoépinotal sutures absent or in- distinct. Petiole and postpetiole not marginate on the sides, the latter strongly con- stricted off from the gaster which is largely formed by its first segment. Fema.e scarcely larger than the worker and very similar, sometimes apterous and ergatoid. Fore wings when present with a discoidal and a single cubital cell. Mate with the clypeus and frontal caring much as in the female. Antenne filiform, 13-jointed; basal funicular joints short. Mesonotum without Mayrian furrows. Wing venation like that of the female. The genus has been divided by Emery into four subgenera, distin- guished by the number of antennal joints: Cerapachys, sensu stricto, having 12; Parasyscia, 11; Oocerwa, 10; and Syscia, 9. The distribu- tion of these subgenera is peculiar. The species of Cerapachys, sensu stricto, are known to occur only in the Ethiopian, Malagasy, Indomalayan, ~ and Papuan Regions; those of Parasyscia occur in Texas, Guatenal, ‘Wheeler, Wm. M. 1918. a Australian ants of the ponerine tribe Cerapachyini.’ Proc. American Ac. Arts Sc., LITI, p. 223 4 a: aL Wheeler, Ants of the Belgian Congo 53 ia, Ceylon, India, and Burma; those of Syscia have been recorded Singapore, New Guinea, Queensland, and Hawaii; while is known only from Ceylon. As these ants form small colonies ye a subterranean life, they are very rarely seen and this probably ants for the peculiar discontinuous distribution in the accompanying 5). It seems hardly possible that species of Cerapachys, sensu re, are entirely lacking in South America, but none has been found in of the many extensive collections that have been made on that con- nt. Practically all that is known of the habits of the genus is con- in a paper which I published many years ago on the Texan usyscia august Wheeler.! {2 4 ¢ ys | iat a mw. 5 5 Sit’ i LP bet : x Ti, a | PRUVERIO Kat AW | ot hem? |) J ae + . ct Su 4 Dy a. 4 Map 5. Discontinuous distribution of the genus Cerapachys. Ae ta Phe Cerapachys cribrinodis Emery ol ‘Two workers found in the stomach of a toad (Bufo funereus) taken ; ng and Chapin at Medje. Puyracaces Emery Closely related to Cerapachys. The worker and female have 12-jointed antenne. The terminal funicular joint, however, is not enlarged but tapers from the base to : tip and is not longer or scarcely longer than the two preceding joints together. _ The eyes of the worker are much larger than in Cerapachys and the sides of the petiole _ and often also of the postpetiole are strongly marginate. The female is winged or lerous and ergatoid; the male is known in certain Australian species. e ‘An American C. with remarks on the affinities of the h * Biol. Bull Tipp 181-191, 7 aime cas the Cerapachyinw.’ Bio 54 Bulletin American Museum oy Natural History [Vol. XLV This genus is known only from the Ethiopian, Malagasy, Indo- malayan, Papuan, and Australian Regions (Map 6) and is represented by the greatest number of species in Australia. The little thatisknown concerning the habits of the species is recorded in my paper entitled 2 ‘The Australian ants of the ponerine tribe Cerapachyini.’' The workers forage in small armies on the surface of the soil, like many Doryline, and prey on other ants or possibly on any small insects they may en- counter. 15 60 7S 90 105 120 135 160 Cc — ee i ty | 7% i Y ee PD. 1S we N b ae 7 0 —— Ptr i 0 _ = ’ Peace See i ee ah ES fee ee are eee.) Ue -- - ~ 2 1s 0 Is 3%” 45 60 5 90 105 120 135 160 Map 6. Distribution of the genus Phyracaces. Phyracaces langi, new species Worker (Fig, 4).— Length 4 to 5 mm. Head subrectangular, a little longer than broad and a little broader behind than in front, its sides feebly and evenly convex, its posterior border broadly and rather deeply concave and somewhat truncated, the occipital border sharply marginate with the margination surrounding the blunt but projecting inferoposterior corners and continued forward along each side of the gula to the insertion of the mandible. Eyes moderately large, flat, in front of the middle of the head. Mandibles with slightly concave external and very finely and evenly denticulate apical borders. Carine of cheeks very prominent, in the form of blunt, rectangular teeth. Frontal caring erect, subparallel in front, more approximated but not truncated behind. _ Antenne rather robust; scapes three-fifths as long as the head, slender at the base but rather abruptly enlarged before the middle; joints 2 to 9 of funiculus 11918, Proc. American Ac. Arts Se., LIII, pp. 215-265, 17 figs. Wheeler, Ants of the Belgian Congo 55 der than long, tenth joint larger, distinctly longer than broad, terminal joint 1g, not broader than the preceding and not longer than the two preceding .. Thorax subrectangular from above, about twice as long as broad, ie broader through the « epinotum than more anteriorly, evenly convex above, t traces of dorsal sutures, truncated and sharply marginate anteriorly and er orly. The margination separating the base and declivity of the epinotum is to form a small blunt tooth on each side. The eral borders of the dorsum are indistinctly marginate, all in the epinotal region, but the sloping epino- is sharply marginate laterally. Petiole as Betemincee rectangular, about one and two- ; broad as long, with bluntly dentate posterior 1a pin front and on the sides, with trun- , htly concave anterior, feebly convex dorsal oping posterior surface. Ventrally in front it bears e, triangular, compressed, subtranslucent tooth. ic ee peced as the petiole, as long as broad, ularly rectangular, flattened above, with only erio et marginate. First gastric segment larger than the postpetiole, of a similar shape ; than long, anteroventrally with a blunt | or tubercle. Pygidium subcircular, truncate, ely and indistinctly spinulate on the sides. Legs per slender, hind cox with a large rounded, trans- Are cae em Cee inner, side, ; en ng; mandibles coarsely and sparsely punctate. ad with tates, emooth and very shining space on ee frontal carine; remain- 4 » with coarse, elongate punctures or foveole Fig. 4. Phyracaces langi, and iorly with a few coarse ruge. Thorax above and new species. Worker. tl las rather regularly longitudinally rugose, with dications of elongate foveol# on the humeri and truncated anterior surface; epinotal vity more finely and regularly longitudinally striated. Sculpture of petiole above r to that of the thoracic dorsum but with more numerous elongate foveolz in it gal spaces; on the postpetiole the foveol# are larger and more abundant an pete ruge much less distinct; first gastric segment, pygidium and _ posterior portions of remaining segments coarsely and evenly punctate, the basal ” portior of these segments more shining and very evenly striolate. Scapes finely, lees $ more coarsely and much more sparsely punctate. Hairs grayish, bristly, suberect, moderately long, rather evenly distributed on he body, more abundant on the tip of the gaster, more appressed on the legs; tibix nd scapes with a few long, suberect hairs. Pubescence short, visible only on the netate portions of the gaster. - Black; mandibles, antenne, legs, tip of gaster and sting piceous, cox and middle 56 Bulletin American Museum oj Natural History (Vol. XLV Very similar to the worker. Pronotum coarsely foveolate; mesonotum small, flat, somewhat pointed anteriorly, with its ruge converging in front. Postpetiole distinctly broader than the petiole and a little broader than long. Wings whitish hyaline, with very pale yellow veins and large, conspicuous, dark brown pterostigma. Described from seven workers and eight females taken from a single colony at Lubila, “nesting in a mushroom-shaped termitarium against a tree in the forest” (Lang and Chapin). Of the four described Ethiopian species: of Phyracacee, langi is most closely related to P. foreli Santschi of the Gold Coast. The worker . of this species,. however, measures only 3.5 mm. and, judging from Sant- schi’s description, has a nearly straight occipital border, shorter antennal scapes, and ‘different sculpture, especially of the head, petiole, and post- petiole. His'figure of the petiole shows much longer posterior teeth than in langi.. ‘Tie specimen from Samkita, Gaboon, described by Santschi as the female of foreli measures 4 mm. and is so different from the worker in the shape of the petiole that I feel sure that it belongs to a distinct species, which may be designated as Phyracaces santschii, new species. Ponerinse Postpetiole separated from the third abdominal segment by a constriction which is more or less marked (except in the Odontomachini and in certain males of Ponerini), almost always as broad as the third segment (except in Myrmecia and a few others). Worker and FreMae with a powerful sting. As a rule there is a stridulating organ on the basal surface of the tergite following the postpetiole; it consists of very fine transversal striw of the articulating surface. Median spur of the tibie pectinate, : when present, except on the middle tibis of a few genera; lateral spur simple. Fore wing as a rule with two closed cubital cells; but there are many exceptions. The dimorphism of the worker is feebly marked (except in Megaponera fetens, where it is very pronounced) and the female as a rule is not very different from the worker; ergatoid females exist in many genera, In a few cases the MALE has no con- striction behind the postpetiole; such males can usually be recognized from male Doli- choderine by the feeble development of the mandibles. Ergatoid males are known for certain Ponerini. Larv with the mandibles powerfully developed for ant larve; the anterior portion of the body long, slender and neck-like, folded over the swollen abdominal portion; the segments are either densely hairy all over or covered with rows of peculiar tubercles beset with more or less prominent bristles; the larvae of Megaponera and Bothroponera are hairless. Nympus enclosed in a resistant cocoon, which may be opened by the adult without intervention of the worker. The West African Discothyrea oculata Emery is the oly case in which the nymphs are described as having no cocoon. In the Ponerine the larve are nearly always fed with pieces of sold food, which is almost invariably animal matter. Arnold says that Euponera sennaarensis (Mayr) is possibly an exception to the rule: - — ee ae ee ae Wheeler, Ants of the Belgian Congo 57 a ca tatckitae, but its nest very often contains consider- + accumulations of grass seeds, which may perhaps be used as food.' : eee Sern coenies can Rent bo over- stir pé safely asserted that at least 80 per cent. of their food con- ist of termites, and they thereby constitute one of the chief checks to these pests of tropics. Certain species are exceptional, such as Plectroclena mandibularis, which efly on millipedes and beetles, and Aerstg arnoldi Forel, whose food con- ities Patetigrne tarsatus, Megaponera fatens, Ewponera sennaarensis, of Leptogenys and Odontomachus hematoda. BEE forccag in fies has boon checrved fn covers! epedie o Ponerine t parts of the world. In our region this habit is displayed by Megaponera and to a slight extent by Paltothyreus tarsatus. The former marches in double ee striking disparity in size between the two forms composing the colony as a very singular appearance. Their prey consists entirely of termites, and when a hunting-ground containing these animals has been found, the columns break i pour into every hole and crack which leads to the invaded galleries. The hod then adopted is as follows: each ant brings to the surface one or more termites, i then re-enters the galleries to bring up more victims. This is continued until ant has retrieved about half a dozen termites, which, in a maimed condition, are t struggling feebly at the surface. The whole army reassembles again outside and picks up as many termites as it can conveniently carry, usually 3 or 4. SNRs bee Gia ro-formediand merch home. Less order is shown by P. tarsatus, Ihave often seen this ant carrying termites, in short single files composed of about ozen workers. (G. Arnold, op. cit., pp. 7-8). PLaTyTHYREA Roger Woan —Small or medium-sized, slender, monomorphic, opaque black ants, ‘surface and very poorly developed pilosity, with flat clypeus often wit! ; 4 posterior suture, indistinct frontal area and large, thick, expanded and widely _ separated frontal carine. Mandibles large, triangular, with edentate or finely denti- jut, funiculi without a distinct club. Eyes rather large; ocelli absent. Promesono- uti suture distinct, other thoracic sutures feeble or obsolete. Petiole massive, not safer its posterior articulation at the middle of the anterior surface of the ‘iole.. The constriction between the latter and the gaster moderately pronounced. idle and hind tibiw with two spurs; claws with a single tooth. FeMa.e winged, very similar to the worker and but little larger; eyes larger, but ae Matz more like the female than in most genera of the subfamily; clypeus more _ convex than in the worker and female; frontal carine not dilated anteriorly. Man- p oditles triangular, with sharp apical border. Antenne 13-jointed; szap2a little shorter as _—-«-*4915, Ann. South African Mus., XIV, p. 7. 58 Bulletin American Museum of Natural History {Vol. XLV than the second funicular joint. Eyes and ocelli very large. Pronotum large, not overarched by the mesonotum, the latter convex, with indistinct Mayrian furrows. Petiole much as in the worker. Pygidium rounded; cerci developed. This genus, of which more than 35 species are known, ranges over the tropics of both hemispheres (Map 7) and is represented by more species in Africa and Madagascar than in the Indoaustralian or Neotropi- cal Regions. Our American and many of the African species seem tofeed _ largely or exclusively on termites. I have found P. punctata (Smith) of the West Indies nesting in termitaria. Arnold gives some notesonthe habits of two of the African forms. Of P. lamellosa (Roger) subsp. . longinoda Forel variety rhodesiana Forel he says: Seo g ¢ T pale fo ke SE TN RaW 5 iat on ; ye At TU et eRe : CLS Lt Eva o y re b hw Bi) 4) AN “a Wt tJ Sled | K = = ' ‘6 xX Z_ Bs al ! — bd . 5 Bias % Pn : & | mY a or a . re | i oa 2 | . - - oe - a " * 7 > J a - : 2 oe Map 7. Distribution of the genus Platythyrea. The nest of this species is so distinctive that it cannot be mistaken for that of any other Ponerine. The entrance is surmounted by a dome, from 6 to 8 inches high, by about 12 inches broad at the base. The dome is built up of very even-sized small pebbles, about 5 to 8 mm. in their largest diameter. The entrance is situated in the center above, and this is generally the only entrance, very exceptionally there may be a smaller and less regular opening at the base of the mound. He gives the following account of P. arnoldi Forel: I have met with this species on only one occasion. The nest, situated on an open piece of ground, was surmounted by a mound with the entrance at the apex, as in lamellosa variety rhodesiana, but unlike that species the mound of arnoldi contains no~ large pebbles. The surface of the mound was covered with the elytra and carcasses of hundreds of beetles, mostly Tenebrionide. Workers were seen carrying live beetles to the nest; the prey being held by its mandibles in a position above and parallel to the body of the ant. Since a careful examination of the rubbish-heap of this nest ee Wheeler, Ants of the Belgian Congo 59 border. Reet Gets hae eedeety caer that H about 10 distinct teeth. Clypeus large, rather flat, more convex Platythyrea gracillima, new species. Worker. a, lateral view of body; 6, head of same from le behind, with broadly rounded, entire anterior border and distinct suture. Frontal carine very prominent, fused posteriorly. Eyes small, a front of the middle of the sides of the head. Antennal scapes long and stout, ¢ fully one-fourth their length beyond the posterior corners of the head; ue eres ee res tense av long ss beced Thorax , laterally compressed, especially in the meso- and epinotal regions; h the pronotum, which is as broad as long and as broad as the head, d in front and on the sides. Mesonotum longer than broad. Promesonotal re distinct, mesoépinotal suture obsolete. In profile the dorsal outline of the BEG cearty straight thd boriscnta; the base of the epinotum nearly twice as as the declivity. The latter is abrupt, submarginate on the sides, which are ly angulate. Petiole laterally compressed; seen from above a little more than as long as broad, with straight, parallel sides; in profile evenly rounded in 60 Bulletin American Museum of Natural History [Vol. XLV a front, straight above and very sharply and deeply concave behind, the ridge between the dorsal and posterior surface being narrow, transverse and feebly emarginate in the middle. At its posterior end the petiole is fully three-fifths as high as long. Postpetiole distinctly longer than broad, as broad as the gaster behind and not separated from it by a perceptible constriction, narrowed to the breadth of the petiole in front. First gastric segment as long as broad, the remaining segments short, telescoped into it. Legs rather long. Slightly shining; mandibles more shining, finely and densely punctate; re- mainder of body even more finely and densely punctate; with a few larger, but very shallow and indistinct, superadded punctures on the head, thorax and petiole. Hairs absent; pubescence yellowish gray, very short and fine, rather evenly distributed like dust over the whole body and the appendages, longer and more oblique on the mandibles. Black; mandibles, clypeus, frontal carinw, antenna, legs, including the coxm, posterior corners of the head, dorsal surface of pronotum, epinotum and petiole, and posterior border of postpetiole and first gastric segment, red; remaining gastric seg- ments yellow. Described from a single rather poorly preserved specimen from Avakubi (Lang and Chapin), taken from the stomach of a toad (Bufo regularis). . This species is unlike any of which I have seen specimens or descrip- tions in the shape of the head and body and especially of the petiole and gaster. In certain respects it approaches viehmeyeri Santschi of German East Africa, but is much smaller (viehmeyeri measures 13 mm.), and has densely punctate instead of striolate and sparsely punctate mandibles; the latter are denticulate; the head and antennz are longer; the thorax not submarginate on the sides; the pronotum is not longer than broad; the mesonotum is longer than broad; the color is very different; etc. PaLToTHYREUS Mayr Worxker.—Large black ants, with monomorphic workers. Clypeus in the middle with an elevated lobe, which is truncated anteriorly and projecting over the anterior clypeal border, marginate on the sides, excavated in the middle and extending back like a spearhead between the frontal carinz which are moderately dilated and sub- triangularly lobate. Mandibles elongate, triangular, their apical borders finely denticulate. Antennal funiculi slightly thickened distally. Eyes situated in front of the middle of the head. Thorax unarmed, not impressed dorsally; promesonotal suture distinct, mesoépinotal suture obsolete dorsally. Petiole surmounted by an erect scale. Constriction between the postpetiole and the gaster feeble; gaster rather long. Claws with a tooth in the middle. FeMate very similar to the worker but considerably larger and winged; thorax— depressed, pronotum broadly exposed. Mate with triangular clypeus furnished near its anterior border with a small conical tubercle; its posterior portion not prolonged backward between the antennal insertions. Antenne long, scape much shorter than the second funicular joint. ee Map 8. Distribution of Paltothyreus tarsotus (Fabricius). TT j “a <—- Wheeler, Ants of the Belgian Congo 61 — ~ ae B ie nd \ ae —_ um above largely exposed; mesonotum with traces of Mayrian furrows. Petiole by a thick node, its ventral surface convex, but not toothed. Postpetiole astrong tooth. Pygidium acutely pointed but not prolonged into a spine. om genus is monotypic, the single species P. tarsatus ranging over wi le of the Ethiopian Region (Map 8) as one of its most conspicuous F 2 - Paltothyreus tarsatus (Fabricius) a Text Figure 6 rt {this species, which has been repeatedly described by previous s, Lang and Chapin collected a number of single specimens from Dierinn localities: Yakuluku, 8; Stanleyville, 8, 7; Medje, 8; simu, 8; Leopoldville, 9; Bafwasende, #; Bafwabaka, 8; Faradje, —o 8, a. In addition to these, 135 workers and 5 deiilated females were taken om the stomachs of four species of toads (Bufo funereus, tuberosus, (j2 Bulletin American Museum oj Natural History [Vol. XLV superciliaris, and polycercus) captured by Lang and Chapin in the follow- ing localities: Niapu, 8; Niangara, 8; Ngayu, 8; Medje, 8, 9, o; Avakubi, 8; Akenge, 8, 2; Garamba, 8; Gamangui, 8; also a single worker from Faradje taken from the stomach of a frog (Rana occipitalis). It is surprising to find that this large ant is represented by a greater number of specimens than any other species in the toad stomachs ex- amined, for the insect is provided with a very formidable sting, is swal- lowed without mutilation, and can hardly be killed very quickly by the weak gastric fluids of the amphibians. Concerning the habits of P. tarsatus, Arnold writes: “This species is widely but locally dis- tributed. Generally the worker and female go about singly, but occasionally forage in short col- umns, in single file. The food is varied but consists largely of termites. The nests have several entrances, which are sometimes surrounded by large heaps of finely divided earth. The species Fig. 6. Paltothyreus tarsatus (Fabricius). a ° Head of worker. sive smell, which appears to me to resemble that of the juice in a foul tobacco pipe.”’ According to Santschi this species “répand une abominable odeur de charogne.”’ One of the worker specimens from Medje and one from Niangara had a long Cordyceps growing out of the side of the thorax. These ants were attached to sticks with their mandibles, a common condition in — ants that die from the attacks of these and other fungi. Dr. Bequaert says that ‘‘dead specimens of Paltothyreus tarsatus thus parasitized are sometimes found, fixed with the mandibles to a leaf or grass-stalk. The fungus has been referred to Cordyceps myrmecophila (Cesati), of the family Hypocreacee. Its fructification usually grows out between the coxal articulations, on a slender stalk about 2 em. long and ending in a club-shaped organ which bears the ascocarps’ (See part IV). has a most powerful and offen- — | Wheeler, Ants of the Belgian Congo 63 MEGAPONERA Mayr —Rather large black ants with distinctly dimorphic workers, the he antehne shorter and with more transverse funicular ew long, continued posteriorly to a level with the eyes and moderately ‘ober anteriorly. Cheeks carinate. Mandibles, long, deflected, tri- NS; ¥ as wi Naa TE SS S\\ SA SoS SQ SS > QA WS S (27 sd Map 9. Distribution of Megaponera fatens (Fabricius). ater ther feeble. Middle and hind tibie with two well-developed spurs, one ‘is pectinated; claws with a tooth near the base. ALE wingless and ergatomorphic, larger and somewhat more coarsely sculp- han the worker major, with much more voluminous gaster and the petiole nce and inclined forward. 64 Bulletin American Museum of Natural History [Vol. XLV Mate nearly as large as the worker major, with convex clypeus, not prolonged backward between the frontal carine. Mandibles very short, blunt and edentate. Antennal insertions farther from each other than from the sides of the head; scape longer than the second funicular joint. Eyes occupying less than half the sides of the head, their inner orbits slightly emarginate. Posterior border of head strongly marginate, somewhat colliform. Mesonotum prominent, twice as long as the prono- tum, without Mayrian furrows. Ventral lamella of petiole with an acute posteriorly directed tooth behind the middle. Pygidium not spined. Claws with three or four minute basal teeth. Wings short, with a discoidal cell, two cubital cells and a closed radial cell. This genus, like Paltothyreus, is monotypic and has much the same distribution, the single species, M. faetens (Fabricius), ranging over a large part of the Ethiopian Region (Map 9). Megaponera fotens (Iabricius) Plate VI, Figure 2 Zambi, 8; Niangara, 8, 9; Rungu, 8; Avakubi, 8; Faradje, 8; Panga to Banalia, 7; Boyulu, 8; Niapu, 8; Garamba, 8; Akenge, 8; Gamangui, 8 (Lang and Chapin); Malela, 8 (J. Bequaert). hy er. Fig. 7. Megaponera fetens (Fabricius). a, head of large worker; b, head of small worker; both Pratl. to the same scale. Seventeen of the specimens from Boyulu, Niapu, Garamba, Akenge, and Gamangui were taken from the stomachs of four species of toads (Bufo funereus, superciliaris, regularis, and polycercus) and a male from Faradje was taken from the stomach of a frog (Rana occipitalis). The smaller individuals have the vertex and pronotum very shining, the mandibles toothless, and the funicular joints of the antenng much shorter and more transverse than in the larger workers (Fig. 7a and b) and were therefore formerly regarded as a distinct species (M. crassi- cornis Gerstecker). A worker media was also described by Emery as a distinct species, M. dohrni. At one time he interpreted the smaller J : ; 5 7 A 7 : Wheeler, Ants of the Belgian Congo 65 Js as the true workers and the larger as ergatomorphic females. who grand this view improbable for the reason that the large are times as Numerous as the small individuals in the colony, has i rered the true female.’ It is of the ergatomorphic type, wingless thorax like the large worker and measures 18.5 > petiole, however, i is squamiform and not cuboidal as in the ts of this ant.2_ Wellman observed it in Benguela and informed its habit of marching in populous columns.’ In a later paper‘ blished some observations of Prell on the same ant in German . He found it running in single file on the road. Most of the db: its strong odor, which resembled that of oil of bitter almonds and ed to the alcohol of the vial in which the specimens were Similar observations were made by Bequaert in the au River, we encountered a troop of several hundred Megaponera marching in a a several abreast, apparently moving with decision to a predetermined goal. d the bank of the stream, stridulating loudly. We were unfortunately paneupied at this spot collecting a lot of large Papilio which came down to the drink, so that we did not think of following the Megaponera army. An hour ese ants returned in good order in the reverse direction, each of them carrying Missionary travels and researches in South Africa,’ 1859, pp. 576- the earliest account of the termite hunting Ponerine of Sentral poe yh “Nat rire Pp. >. 361. See also Prell, H., 1911, ‘Biologische Beo ae ees eed , XXXVI, pp. 243-253. Rev. Zool. Afr., ie p. 22. 66 Bulletin American Museum of Natural Hist Vol. XLV. a in its mandibles a whitish pellet consisting of dead termites glued together widhsptindcal Some of them carried as many as ten to twelve termite workers thus agglutinated, others only two or three soldiers; one carried a dedlated male, possibly the king of the plundered termitarium. The number of termites in a pellet varied with its size, but not an ant returned without something. While collecting a number of these Megaponera fatens with their booty we experienced the effect of their sting, which is lancinating and very painful but very transitory. In his monograph of the Formicide of South Africa (loco citato, p. 47) Arnold says: It is a common ant in Rhodesia and lives almost exclusively on termites, which are carried off by means of carefully arranged raids in which the ants march in double file. This is the species which is popularly called the “ Matabele” ant, and like its cousin Paltothyreus, it is also endowed with a very offensive odor. They stridulate very loudly when disturbed, and their sting is exceedingly painful. The entrance to the nest consists of one or more simple holes without any mounds of earth around them. In the Proceedings of the Rhodesian Scientific Association, XIII, 1914, p. 26 et seq., Arnold has recently published a fascinating descrip- tion of the extraordinary way in which the Matabele ant changes its nesting site and is followed by its numerous guests. I quote the greater part of his account, as the journal in which it appeared may not. be accessible to my readers: This is eminently a termitophagous species, and it is likely that it changes the _ site of its nest more often than is the case with the majority of our ants. When we bear in mind how continuous their assaults are on the colonies of termites, it seems very probable that the supply of the latter insects may be so diminished within.the practical range of the camp of the raiders that the latter may find it advantageous to move their quarters from time to time to new and more fruitful country. The migra- tion of this ant which I am about to describe is of particular interest, apart from the behavior of the guest insects, because it was the occasion of the discovery of the tquy queen of the species. * * * My attention was attracted to this migration by seeing a mass of these ants assembled together with their larvee and pupa, in the open. On one side, many workers were to be seen bringing along the larve in their jaws, on the other side of this mass a few workers were moving in the other direction, in a somewhat hesitating manner. Following the track backwards, I came to the site of the old nest, situated _ about 15 feet away. Returning to the camp, it was seen that some workers had wi started to pick up the larve again, and were carrying them yet further away fromthe original nest, only to be laid down again at about another 15 feet further away. Sub- sequent observations showed that the migration was carried out in three stages, three temporary camps being formed between the old and the new nests, which were about 60 feet apart. The method adopted by the insects was as follows. First of all, the eggs, larvee, pup and males were taken from the old nest and put down at the first camp, from which many workers were to be seen hurrying back to fetch away the rest of their charges. In the meantime, a few workers were to be seen pacing up and down on the other side of the camp. They did not carry any larve and it would almost . Wheeler, Ants of the Belgian Congo 67 though they had some idea of the numerical composition of the colony, and of the volume of the first camp should be, before the old nest could be considered been emptied by its inhabitants, and the proper moment to have arrived for start.to-be made. However, after about six or seven minutes, the march enced; and within a short time the second camp had been made at a distance ‘15 feet from the first. Similarly a third and last camp was formed further V hile the first camp was about to break up that I saw an insect then gel than the largest worker, and which, when captured in the third camp, my surprise, to be the queen. entrance to the old nest was a hole about 1 inch across, which ran down ally for about 5 inches and then branched off at an angle. Looking down this e various guests and parasites could be seen climbing up the walls in an almost ous stream, hastening to join their hosts in their new home. These insects da pens, two species of staphylinid beetles, a histerid beetle and an hagous beetle; there was also a spider. The Lepismas as usual were very ; of the larger staphylinid I saw only one specimen, but of the smaller sort the other beetles very many examples occurred, and during the half hour or so which I watched the procession, about two dozen specimens of the spider unted. Had it been possible to have cinematographed the scene, it would have ed us with a film of surpassing interest. Here, as in the case of Myrmicaria, mecophiles were able to follow the tracks of their hosts without any delay or r: ‘in ty. Occasionally one of the smaller staphylinids would leave the beaten for a short distance and then return to it again a little further on, but to the rity of these commensals, the odour of their hosts had laid down a path as clearly sd as a macadamized road would be to our eyes, so that with the above excep- it was rare to see any of these insects swerve from the line of march by as much as motley crew of cringers, thieves, murderers and body-snatchers did not to attract the slightest attention from their victims the ants, which were too with the work in hand to waste any time on the rabble following in their wake. . this crowd, the spiders alone were able to keep pace all the time with the ants, eerees’, the very small histerid, even at its most feverish pace, did not succeed ove more than 2 inches per minute, so that it would have arrived at the new about six hours after leaving the old. Those beetles which managed to reach the rent camps, while these were still intact, buried themselves in the heap of larvee Tims, where they remained until the gradual depletion of the mass made it t that they had not arrived at the site of the real nest and that another wearisome journey had to be made to attain their goal. i The spiders moved about in the camps in a very easy and unconcerned manner, aking no attempts to hide under the piles of cocoons. They ran over the backs of the _ ants, mingling in a friendly way with the crowd; yet even in the hurry and bustle of ; s march, it was not possible for these animals to conceal entirely their method of _ earningaliving. A worker ant, carrying a larva in its jaws, was seen just about to pass : der standing on the edge of the camp. The spider ran up to the worker, stroked with its front pair of legs for a second or two, and then plunged its fangs into the z= va. The latter was released by the ant after a little hesitation, and within five nutes had been sucked dry by the spider. We know that there are many ant mguea which live chiefly on the young of their hosts; but usually these insects 68 Bulletin American Museum of Natural History [Vol. XLV offer, on various parts of their bodies, those bribes in the shape of trichomes which — make the ants careless of, or oblivious to the true nature of their guests. On the other — hand, there are the synceketes, or indifferently tolerated guests, with which perhaps — the histerid and onthophagous beetles found on this occasion should be clas J which do not usually bear trichomes. They owe their immunity from attack on the — part of the ants, either to their insignificant size, or to their awkward shape, which — prevents the ants from seizing hold of them. But it is difficult to understand how the - spiders can live unmolested in the nests of such a powerful and vicious antas Mega- — ponera fatens and be allowed to feed on the larve, without apparently the mildest — protest. They do not possess trichomes, nor are they so constructed, by smoothness or hardness of texture, as to prevent the ants from seizing hold of them. 4 The staphylinids are probably to be placed in the category of synechthrans, or inimically persecuted intruders, which in- cludes a number of insects which skulk about ants’ nests, and get a living by rummag- ing about in the refuse heaps or kitchen- middens, or by attacking solitary workers in the lonely comers ans Se ee ment. 348 2H In conclusion, it should be pointed out that in these latitudes, migrations of antscan _ be expected to take place only after sunset, — was the case with Megaponera, since the deli- — cate larva cannot bear a lengthy exposure to qi the rays of the sun. Two of the vials of Megaponera collected by Mr. Lang contained a number of cocoons and larve in vari- ous stages, so that, on reading Arnold’s account, it seemed probable that the e brood might show adaptations to being Fig.8. Megaponerafatens (Fabricius). carried about and exposed to the sun- pp ak view ia Nie light. A study of the material shows that such adaptations can be detected. The larve (Fig. 8a and b) are grayish white, long and subcylindrical, and only slightly curved, with strongly marked segments and with smooth, re- markably tough integument, which is quite hairless in all stages, a condi- tion I have never observed in any other ant larva. The head is very large, rounded, strongly chitinized, and terminal, with long, acute, faleate; edentate mandibles, minute vestiges of antenne, and very prominent tactile sensille on the maxilla and labium. The size of the head and mandibles shows that the larve are fed on pieces of termites and not with or if earlier, only on dull and cloudy days, as — Wheeler, Ants of the Belgian Congo 69 ed liquid food, and the strong integument is evidently an m to exposure to the air and light and to the exigencies of fre- d protracted transportation in the powerful denticulate jaws of s. The nudity of the integument indicates that even the very e are carried singly and not in bunches held together by inter- rs as in most other species of ants. The cocoons are black y tough, characters which I have observed in certain ions to exposure to sunlight.'_ This interpretation is confirmed Lang, who, without knowing of my observations, informed me as s surprised to find Megaponera often exposing its dark cocoons to the sunlight. ntly, i in a letter to Prof. Poulton,? G. D. H. Carpenter records itional observations which he was able to make on M. faiens of Lake Victoria: ' a good deal of the ant Megaponera fetens here: one is always coming across ‘solemn, slowly marching, black processions—of any number from 50 to — 500 ors -‘Thave never seen them carrying any other booty but the species of termite (abounds here—the one I have alluded to before. It lives underground and no hills—coming out of little holes and running about, uncovered, in the open, bit 3 of live or dead grass which it carries down the holes. Presumably in corre- t h its open-air habits, its color is much darker than the large termite whose i to destroy on the islands, and which devoured my house. This one does not soldiers, which is perhaps why M aqapeners wages such ceaseless war against termed when it goes out in column, wanders about looking for the termite holes. itely one is found there is great excitement. The little bits of grass which € s plug the entrance are dragged out, and the ants scramble down the hole ily reappearing with termites, feebly struggling in their jaws. Sometimes its of dead grass, etc., as if they were breaking down hastily erected barricades! almost picture the termites hastily throwing up partitions of grass and earth i BoTHROPONERA Mayr . — —Small, medium-sized or large, opaque or subopaque, usually strongly 1 d black or dark brown ants. Workers monomiorphic. Head subrectangular, the eyes usually well developed, rarely vestigial, placed at or in front of the srior third of the head. Mandibles subtriangular, with coarsely dentate apical Cheeks without a carina. Clypeus with rounded, obtusely angular or feebly tat Ent. Bee. gc. America, arice, VIII. po. 335- rn exxix, 70 Bulletin American Museum of Natural History {Vol. XLV and sinuately marginate anterior border, prolonged backward as a narrow point between the frontal carinsz, which are broadly and lobularly expanded, incrassated and covering the insertions of the antennz. Frontal groove distinct. Antenna stout, 12-jointed. Thorax with distinct promesonotal suture, but with the mesoépinotal suture and that between the mesosternum and mesepisternum absent or obsolescent. Pronotum not marginate on the sides; epinotum usually unarmed. Petiole with a thick, more or less transverse node, in a few species somewhat compressed and dentate above or behind. Gaster subcylindrical, with pronounced constriction between the postpetiole and succeeding segment, the postpetiole truncated in front; sting rather short and blunt. Middle and hind tibiw each with a large pectinated and a simple lateral spur; claws simple. Fig. 9. Bothroponera sublevis Emery. Australia. Adult larva. a, ventral view; 6, lateral view; c, head, dorsal view; d, head in profile. FEMALE only slightly larger than the worker; winged; in other respects very similar to the worker; ocelli small; pronotum broad and exposed; mesonotum small, flattened, broader than long. Wings rather broad; with a discoidal cell, two cubital cells and a closed radial cell. MALE nearly the same size as the worker. Head short, rounded behind; eyes and ocelli very large; mandibles small, flat, edentate. Palpi long, the labial pair 3-jointed, the maxillary pair 5-jointed. Frontal carine short. Antenne very long, filiform, 13-jointed; the scape short, scarcely twice as long as broad; the first funi- cular joint not longer than broad, the remaining joints long and cylindrical. Pronotum transverse, truncated in front; mesonotum without Mayrian furrows; scutellum very Wheeler, Ants of the Belgian Congo 71 . Abdomen strongly constricted behind the postpetiole; pygidium terminat- directed spine. In some species the penultimate sternite of the er is notched and prolonged on each side as a prominent lobe. Genitalia retracted. layr described Bothroponera as a genus; but Emery, Forel, and i have been treating it as a subgenus of Pachycondyla. I return yr’s conception for the following reasons: First, the larve of onera (Fig. 9a-d) are quite different from those of Pachycondyla, I have shown in a former paper.! Second, Bothroponera, being a ita group may be advantageously separated as a Map 10. Distribution of the genus Bothroponera. a es ° 1S 30 45 Cy 7S EY 105 120 135 160 — - = = a Be * a ; wi af: fy » te — f » ra a - 4 ny NO J a |‘ q ai . A. bras Eee Be i — a » ~' ts ‘ *» a, “2 a iS # 4 a i br 8 Re: ‘ % ee -— “ wd a 3 See}. ie » ae al \ os 1s} ge ia N re , oe ae obi TMS SR nee ae Sex sordl f : 4 is ue : ~\ f 30 » a | + ‘“ / | an Whe, ee ee ae rns o Tec. > Fs ee Pd pe 2 Poo 8 er ° 1s » 45 Cy 15 90 105 120 BS 160 ore " net genus from the purely neotrovical Pachycondyla. Ectomomyrmex y be regarded either as a subgenus of Bothroponera or as an independ- renus. I prefer to adopt the latter course. I also separate out a small » of species of Bothroponera (gabonensis Ern. André and sveni Forel) a distinct genus Phrynoponera (vide infra). Third, there are certain suliarities in the habits of Bothroponera which indicate that the species ¥ y distinct. Like Pachycondyla, they form small colonies er stones in rather moist, clayey soil, but are more sluggish and do not ig readily when captured and instead emit from the posterior end of he body a peculiar mass of frothy substance. I have observed this in of the Australian species, and Bingham and Taylor have seen behavior i in the Indian B. rufipes (Jerdon), according to mamaght a ; SONYA indy of come ant rv. ete.’ Proc. adie: Phil. Soc., LVI, p. 299. ¢ 72 Bulletin American Museum of Natural History {Vol. XLV on.' Bingham says that this ant “blows a whitish, acrid smelling, rather gelatinous froth when seized” and according to Taylor it exudes when seized “‘a milky substance of a frothy nature which hardens on exposure to the air and resembles fine cotton; it is called ‘domona chunti’ or ‘gendu,’ the ‘domonas’ being the weaver caste in Orissa.” B. triden- tata (F. Smith) of Borneo seems to have the same habit, according to Beccari.* The genus Bothroponera is widely distributed over the Ethiopian, Indomalayan, Papuan, and Australian Regions (Map 10). Africa is very rich in species but Australia possesses almost as many. | The following table may be of some assistance in identifying the workers and females of the Ethiopian species of Bothroponera. 1. Head, thorax, petiole and postpetiole coarsely punctate, purtctate-rugulose or atrinted., <... i... avs svieo0ve ss 00 0-0:9:\8 e's oleate MeRtbniN Res nanan » 2 These regions finely and densely punctate, sometimes with superimposed, larger but shallow punctures. . 20.6... 006s ces cee ce so eee es 5 en 9. 2. Mandibles striate. .....5...55.ceb sees ee eee ees oe op ew s)ethie alana 3. Mandibles smooth, sparsely punctate...................200005- tees 7. 3. Petiolar node broadly excised posteriorly....................- cariosa Emery. Petiolar node sharply truncated posteriorly... ...................2000005- 4. 4. Length 8 mm.; testaceous yellow....................005: cribrata (Santschi). Length not leas than 9 mm.; black or brownish black..................... 5. 5. Antennal scapes reaching ‘to occipub....: .4.s, eaeepeee one VRS Antennal scapes not reaching to occiput; eyes small..................05. 6. Length 9 mm.; golden pubescence on body, especially on head, ass» sculpture less pronounced. .............. 0600000 eeeee talpa Ern. André. Length 12 to 15 mm.; golden pubescence less pronounced; sculpture coarser. pachyderma (Emery). 7. Petiolar node broadly excised posteriorly; body covered with golden pubescence. granosa eisis § Petiolar node truncated behind; body without golden pubescence.......... 8. 8. Gaster opaque, finely striated...................ececeees strigulosa Emery. Gaster more or less shining.......)............-+0+: pumicosa (Roger). 9. Eyes well developed in the workers. ............0scccccessveceuveceeuss 10. -; Byes vestigial in the workers. - ....« «.. o6«+08 secste.e 080s aes eens cen 15, 10. _ Lerigth 5.5, 20M. ,. .5:<0+53xseb 00000002 0nuhen asian eles 08 hin's bie 0 stain gies Length af least 7mm... . .... 0... eee us Wien 13. 12., Opaque; head. ovoid. ... .....005sanaecunes cae nee eae kriigert (Forel). Subopaque or shining ; head subrectangular.....-....:.. «+s sven uae 14. 11891, ‘Our Ants.’ Journ. Bombay Nat. Hist. Soc., Phd? ox * Nelie foreste di Borneo,’ , Firenze, 1902, p. 237; teste mery, 1911, ‘Genera Insectorum, Pon- ering,’ p. 75. Wheeler, Ants of the Belgian Congo 73 ning; length 12 mm.; clypeus angularly produced in middle; eyes see i a ie oe aa ik Sed Anta evese levissima (Arnold). ; length 7 to 7.5 mm. ; clypeus feebly sinuate in middle; eyes larger. ae crassa Emery. Length only 4.5 to 5.5 mm.; mandibles smooth, sparsely punctate; eyes very a aabigeenttesgernaeam oe po ee Pee grea nd under a log. When it was lifted the ants feigned death.” The “ from Bafwasende is very small. The specimens from the other 2 of the posterior border of each gastric segment. I regard Sant- "8 ‘iB. sculpturata, described from a female, as synonymous with Length more than 13 mm. Differing from the typical form in its somewhat greater size and in color, being | black, with only a slight brownish tinge to the legs. Even the frontal carine nn are’black and there is no red on the gastric segments. The erect hairs on rsal surface are also black, at least in certain lights, not fulvous as in the typical but the hairs and pubescence on the tibie and tarsi are of the latter color. The oveole on the gastric segments, especially behind the anterior portion of the first ent, seem to be shallower and both they and the spaces between them to be less ttly striated than in the typical pachyderma. A single specimen from Medje (Lang and Chapin) taken from the uch of a toad (Bufo polycercus). | Bothroponera talpa Ern. André orga 8; Niangara, 8; Avakubi, 8; Medje, 8, 9 (Lang and =). Fight specimens, all taken from the stomachs of toads (Bufo , polycercus, and superciliaris) | and agreeing well with André’s ‘ iption. . 74 Bulletin American Museum of Natural History [Vol. XLV Bothroponera soror (Emery) Akenge, 8, 2; Medje,8, 9; Ngayu, 8; Niangara, 8; Avakubi, 8; Niapu, 8; Faradje, 8 (Langand Chapin). Forty-one workers and three deiilated females. All but three of these specimens were taken from the stomachs of toads (Bufo superciliaris, polycercus, funereus, tuberosus, and regularis); one from Faradje was taken from the stomach of a frog (Rana occipitalis). Arnold records this as a rather rare species in Rhodesia. “Tt usually nests under stones, and has a very strong smell of cockroaches. The colonies do not usually comprise more than two dozen individuals.” Two of the specimens from Medje were taken by Mr. Lang while they were crawling on tree trunks and also on the tents of the expedition. He notes that, “‘when crushed, they gave off a stench reminding one of a bug.” Bothroponera soror variety ancilla (Emery) A single worker from Isangi (Lang and Chapin) differs from the typical soror in its smaller size (less than 7 mm.). It differs from Emery’s description of the variety ancilla, however, and agrees with the typical form in having a trace of the mesoépinotal suture. Bothroponera sjéstedti (Mayr) Text Figure 10 Eight workers taken by Dr. Bequaert at Malela agree very closely with Mayr’s description of the types from Cameroon except in being ‘\ Fig. 10. Bothroponecra sjéstedti (Mayr). Worker. a, lateral view of body; 6, head from above. smaller. They were nesting “‘under the fallen trunk of a palm in swampy ground.” The type specimens were found by Sjéstedt “in a rotten palm trunk,” according to Mayr. The species is peculiar in its very small size, pale coloration and in having the eyes reduced to a few ommatidia. Wheeler, Ants of the Belgian Congo 75 _ Purynoronera Wm. M. Wheeler ——————— a ‘Worxer.—Allied to Bothroponera but distinguished by the following characters: diy shorter and stouter; mandibles narrower, not triangular, their basal and external rders parallel, the apical border oblique, bluntly dentate, not forming a distinct : with the basal border. Clypeus short, elevated in the middle, with a median row and a ridge on each side, the anterior border broadly rounded and entire or bluntly bidentate, posteriorly extending back between the frontal carinz as a narrow acute point. Frontal carine expanded as lobes but the latter are not thickened as in ) mera, but depressed except at the edges which are smooth and slightly ele- va Gekscaling the insertions of the antennz as in Bothroponera. Eyes rather large id convex, broadly elliptical, placed just in front of the middle of the head. Anten- ‘nw stout, 12-jointed as in most Ponerine. Thorax with broad pronotum; promeso- ‘notal suture distinct, arcuate; mesoépinotal and mesepisternal sutures obsolete. with two stout spines. Petiole surmounted by a flattened scale which eurves back over the postpetiole and terminates in a comb consisting of five acute, ‘flattened teeth. Remainder of abdomen very short, oval, the postpetiole which forms nearly half of it, not truncated but rounded in front and not separated by a constric- on from the first gastric segment, though the stridulatory surface is well developed as 1 Bothroponera. Sting very long; longer, more slender and more acute than in the latter genus. Legs rather long and stout; middle and hind tibie each with a long __ peetinated and a simple lateral spur; claws simple. Sculpture of body coarse; . ‘PRonity short, abundant, coarse and erect. Fema.e winged, but wings unknown; in other respects very similar to the worker 3 Ged searcely larger. Ocelli small. Pronotum broad and exposed; mesonotum and a Te flat, together nearly circular, each being broader than long. Mare unknown. _ Genotype: Bothroponera gabonensis Ern. André. _____ In my opinion this is a very distinct genus and would probably long _ since have been recognized as such had it not been that only one or two 7 ‘species were known and these very imperfectly, that one species of . _Bothroponera [B. bispinosa (Smith) of India] has a spined epinotum, and that another Indian species [B. rufipes (Jerdon)} has the petiole antero- s posteriorly compressed above and the border denticulate, thus suggesting E the conditions in Phrynoponera. In reality the latter genus is distinct, i not only in the structure of the petiole but also of the mandibles, frontal caring, and postpetiole, in the absence of any constriction between the postpetiole and the gaster, and in the abbreviation of the latter. The genus seems to be confined to a narrow region in West-Central Africa Map 11). The species probably all live in the humus of the rain forest. ‘The workers and females of the forms which I have seen from the Congo may be separated with the aid of the following dichotomy. nay Clypeus with two large blunt teeth... . 2... «2.06.0 2. my aappeus without teeth... 0... 5.5... eee cece eee on ec aes 7. 7 e 3 Length 6.5 to 7.5 mm.; mandibles 4-toothed.. re 7. —- Length 9 mm.; mnéiiadlibles EG iw ae sve de’ win os - helinidie, new species. 76 Bulletin American Museum of Natural History [Vol. XLV = =— Y > > % = = = = = ‘Se - / +*%e 4 pT oe | ce hehe Le Br ~ SS Map 11. Distribution of the genus Phrynoponera. 3. Gaster shining, feebly sculptured (gabonensis)... + 54.0.6 hie ie Voor Signe nn Gaster opaque or subopaque, strongly sculptured: Ry cles seca es op Sythe nn 5. 4. Mandibles, frontal carine, antenne, and legs red. ag gabonensis (E. André). Mandibles, frontal carine, antenne and legs blackish. ... . var. esta, new variety. 5. Mandibles smooth, sparsely punctate...,.. 6.0... sce+s0+s4p 00 sae 6. Mandibles striated and sparsely punctate.......... var. striatidens (Santschi), 6. Mandibles, frontal carine, antennz and legs red... .var. fecunda, new variety. Mandibles, frontal carine, antenne and legs blackish. .var. umbrosa, new variety. 7. Small species (6 mm.); funicular joints 2 to 10 much broader than long; mandibles and appendages black.................. bequaerti, new species. Large species (8 to 9 mm.); funicular joints 2 to 5 at least as long as broad; 6 to 10 slightly broader than long; mandibles and appendages red. sveni (Forel). Phrynoponera gabonensis (Ern. André) There are specimens of five different forms of this species in the collection. To gabonensis, sensu stricto, I refer a single worker from Baf- wasende, one from Medje (from the stomach of a toad, Bufo funereus), Wheeler, Ants of the Belgian Congo 77 deiilated female from Akenge (from the stomachs oly -and funereus). In these specimens (Fig. lla-c) the scapes extend somewhat beyond the posterior border of the the first funicular joint is as long as broad and distinctly longer ed but not flattened. The lateral petiolar spines are nearly long as the three others, the median but little longer than the diate pair. The gaster is distinctly shining, and the mandibles, head, frontal carine, legs, and posterior borders of gastric seg- : a _Phrynoponera gabonensis (Ern. André). Worker. a, lateral view of body; 6, bead from d, as André states. The mandibles are shining and coarsely tate, without strie. The postpetiole has in some specimens a inet opalescent blue reflection not mentioned by André. The speci- me asure 6.5 to 7.5mm. The female has the pronotum indistinctly semicircularly, the mesonotum and scutellum longitudinally rugose. _ Phrynoponera gabonensis variety esta, new variety and Fema.e (deilated).—Differing from the preceding form in color, aandibles being nearly black, the frontal carine, antenne and legs blackish brown, -mearly as dark as the remainder of the body, the posterior borders of the ab- ginal segments slightly paler and more reddish. ‘The mandibles and gaster are ning as in the typical form. 78 Bulletin American Museum of Natural History (Vol. XLV Described from eight workers and a female from Medje (type locality), two workers from Ngayu, and a female from Gamangui (Lang and Chapin), all found in the stomachs of toads (Bufo superciliaris, polycercus, funereus, and tuberosus). Phrynoponera gabonensis variety fecunda, new variety Worker and Femate (deiilated).—Having the coloration of the typical form, i. e., with the mandibles, frontal carine, antennz, legs, and posterior borders of the abdominal segments red, but with the postpetiole and gaster opaque, densely and finely punctate, and with superadded coarser longitudinal punctures, or aciculations, having sharp anterior edges. The legs are somewhat more opaque and more coarsely coriaceous than in the typical gabonensis. The mandibles are shining and sparsely and coarsely punctate, as in the two preceding forms. Described from eleven workers and one female from Akenge (type locality), eighteen workers from Medje, two from Ngayu, and one from Avakubi (Lang and Chapin). All the specimens were found in the stomachs of toads (Bufo superciliaris, polycercus, funereus, and tuberogus). Phrynoponera gabonensis variety umbrosa, new variety Worker.—Coloration like that of the variety esta, black throughout, the post- petiole and gastric segments with narrow brown posterior border. The sculpture of the gaster is that of the variety fecunda. Two specimens from Medje (Lang and Chapin) from the stomach of a toad (Bufo polycercus). Phrynoponera gabonensis variety striatidens (Santschi) Medje, 8; Akenge, 8; Ngayu, 8 (Lang and Chapin). Four speci- mens, all from the stomachs of toads (Bufo polycercus, funereus, and tuberosus). These specimens have the coloration of the typical gabonen- sis and variety fecunda and the abdominal sculpture of the latter, but the mandibles are subopaque and finely striated, except at the base, in addition to having the coarse, sparse punctures of the other varieties. The epinotal spines seem to be a little longer and more acute than in any of these forms. Phrynoponera heterodus, new species FEema.e (deilated).— Length 9 mm. Very closely related to gabonensis but differing in its larger size and in the follow- ing particulars: the apical borders of the mandibles are 7-toothed and, in addition to the coarse punctures, are finely striated on their apical halves. The antennz are somewhat longer, the funicular joints 2 to 7 being as long as broad. The ruge on the front and vertex are distinctly coarser and more divergent, the eyes somewhat Wheeler, Ants of the Belgian Congo 79 smaller, the ior-cofners of the head more acute, the clypeus bluntly bidentate as in gal is. The sculpture of the thorax and petiole is also very similar, the postpetiole and gaster sculptured as in the variety striatidens but even more sharply, _ go that the whole surface is opaque. The epinotal spines are broad and flat as in nensis but the median petiolar tooth is nearly twice as long as the intermediate th. The pilosity is, if anything, a little more abundant than in gabonensis and its _ The color is black, with the mandibles, legs, and posterior borders of the segments dark castaneous brown. ¥ Pe le specimen from Stanleyville (Lang and Chapin), without _ further data. This form might be regarded as a large subspecies of jabonensis but its precise status can hardly be determined without worker Phrynoponera bequaerti, new species ae Text Figure 12 ___ Resembling gabonensis and heterodus but . “much smaller. Head, excluding the mandibles, fully as broad as long, the posterior border ly straight; the sides very feebly and evenly convex; the eyes large, moderately convex, with sir posterior orbits at the middle of the sides. les shaped as in gabonensis, with ob- bluntly 4-toothed apical borders. ypeus short, with broadly rounded, entire anterior border, the elevated central portion somewhat concave behind in the middle, with a _ ridge on each side. Antenne short and thick, ye Seapes scarcely extending beyond the pos- or border of the head; first funicular joint rly as long as broad, remaining joints, except last, decidedly broader than long. Thorax as vad as the head, short, shaped much as in ensis but the epinotal teeth are proportion- eet Se ea alma longer, being longer than broad at ‘heir and as long as the distance between the latter, flattened dorsoventrally, with round lobe-like tips. Petiole with longer __ spines than in gabonensis, the lateral spines being as long as the remainder of the __ segment and the median spine as long as the lateral. . __ Mandibles smooth and shining with very coarse, sparse punctures, most numer- ous near the inner border. Remainder of body subopaque except the borders of the frontal carine which are smooth and shining. Head reticulate-rugose, rather coarsely | the sides, on the front and vertex more finely, the ruge scarcely longitudinal. Thorax covered with coarse umbilicate foveole, which are largest on the mesonotum but everywhere so close together that the surface may be described as reticulate- rugose. Anterior surface of petiole with similar sculpture, but the meshes of the 80 Bulletin American Museum of Natural History [Vol. XLV reticulum elongate. Postpetiole and gaster appearing longitudinally striate owing to their having a sculpture like that of P. heterodus and several of the varieties of gabonensis. Legs and antennal scapes nearly opaque, closely coriaceous. Pilosity and pubescence much as in gabonensis and heterodus but the former more reclinate on the head, thorax, and abdomen. Black; mandibles, frontal carine, and legs dark brown. Described from a single specimen taken from the stomach of a toad (Bufo superciliaris) from Ngayu (Lang and Chapin) This is a very distinct species, easily characterized by its small size, edentate clypeus, long median petiolar spine and peculiar cephalic and thoracic seulpture. Phrynoponera sveni (Forel) Three workers from Medje (Lang and Chapin), agree perfectly with Forel’s description. They all show, however, a beautiful blue opales- cence, like that of Lobopelta iridescens, on the smooth declivity of the Fig. 13. Phrynoponera sveni (Forel). Worker. a, lateral view of body; 6, head from above; c, Petiole, dorsal view. epinotum, the sides of the petiole, and the whole surface of the post- petiole. This may have been overlooked by Forel, as the surfaces of these ants are often covered with a layer of dirt. It was only after my — specimens had been thoroughly washed in caustic potash that the blue coloration of the parts above mentioned was revealed. P. sveni is a strongly marked species, characterized by the long antennz, toothless clypeus, and slender, pointed and upwardly directed epinotal spines (Fig. 13a-c). 3) 1922] _ Wheeler, Ants of the Belgian Congo 81 _.-Euponera Forel = ee Bother re ae but smaller and much more finely sculptured. s Z a Worker mo cmorphic, with subtriangular mandibles the apical margins of which are dentate. Cheeks not carinate. Frontal carine closely approximated, expanded and lobular in front and concealing the insertions of the antenne. Eyes 4 ‘Placed near or in front of the anterior third of the head, sometimes vestigial or even absent. Clypeus rounded and obtusely pointed in front, usually carinate. Antenne ae. _ slender, 12-jointed, the scapes slightly thickened apically but notclavate. Thorax ag _ shaped somewhat as in Bothroponera but with distinct mesoépinotal suture and usually distinct mesoépinotal constriction. _Petiole surmounted by a thick transverse ale. _ Middle and hind tibie with two spurs; claws simple. a ‘Femaue winged; in some of the subgenera scarcely larger, in one (Brachyponera) | considerably larger than the worker in other respects similar. : “3 " £ ae P Teg a in’ : \ [ oe ; LS 5 Ady in ; ae 7 Pa vl Sh ae 5% arama ss pL ee Ne tb yN LN "A aN PREV ARIA : 77 f e 4-t Caf of wa Y . | 3 2 5 § , k “=== -4--[L- “ * Oe - C Distribution of the genus Euponera:(simple crossing) and of Euponera (Brachyponera) is (Mayr) (double crossing). . the various subgenera. penery has divided this genus into four subgenera: Euponera, sensu to; 3 Mesoponera; Brachyponera; and Trachymesopus. Euponera, with a single species, is confined to Madagascar; the other subgenera lave a wide distribution over the tropical and subtropical portions of : ( B hemispheres (Map 12). The species live in the ground, either in ter nests or under stones, logs, ete. Eu. (Mesoponera) castanea (Mayr) New Zealand lives, as a rule, in rotten logs and stumps. The colonies Bisctypones are rather large and populous, those of the other sub- 82 Bulletin American Museum of Natural History [Vol. XLV genera much smaller. In the subgenus Trachymesopus there is a pro- nounced tendency to hypogzic habits and also, therefore, to a degenera- tion of the eyes in the worker. Euponera (Mesoponera) ingesta, new species Text Figure 14 WorkER.— Length 5.5 to 6 mm. . Head somewhat longer than broad and about as broad in front as behind, with evenly convex sides and feebly excavated posterior border. Eyes small, flat, broadly elliptical, placed at the anterior fifth of the sides of the head. Clypeus carinate, i anterior border entire, rounded and projecting in the middle, sinuate at the sides. rs Fig. 14. Euponera (Mesoponera) ingesta, new species. Worker. a, head from above; b, thorax and abdomen in profile. Mandibles moderately long, convex, their apical borders with 8 or 9 subequal teeth. Antennal scapes extending slightly beyond the posterior border of the head; first and second funicular joints subequal, about one and one-half times as long as broad, joints 4 to 6 somewhat shorter, remaining joints, except the last, as broad as long. Pronotum as long as broad, somewhat depressed above and very bluntly submarginate on the sides. Mesonotum convex, transversely elliptical, nearly twice as broad as long, completely surrounded by a strong suture; mesoépinotal constriction distinct. Epinotum nearly as long as the pro- and mesonotum together, but somewhat lower, the base and declivity straight, subequal, forming an obtuse angle with each other, the former horizontal in profile, the latter flat; marginate on the sides. Petiolar scale in profile high and cuneate, its anterior surface feebly convex from side to side, its posterior surface flat, with a shallow longitudinal impression in the middle; the border evenly rounded, semicircular from behind, slightly narrowed ventrally. Gaster short, postpetiole sharply truncated in front, the constriction between it and the gaster feeble. Legs moderately long. Mandibles shining, finely and rather indistinctly punctate; remainder of body subopaque; clypeal carina and legs more shining; very finely and densely punctate, especially the head, scapes and thorax. Wheeler, Ants of the Belgian Congo 83 r jackin ahah actin, oh wicendiiaes: on the gaster pale iaparee, erect, slender, and rather uniformly distributed; pubescence very vish, moderately abundant, investing the whole body, including the as; legs somewhat paler; mandibles deep red, with black teeth; in ns the extensor surfaces of the tibie are yellowish. d from six specimens taken from the stomachs of toads reus and polycercus) from Akenge (type locality), one from also from a toad’s stomach (B. polycercus), a single specimen from , and another from Lubila (Lang and Chapin). _ Euponera (Mesoponera) subiridescens, new species Text Figure 15 ngth 6.5 to 7 mm. ad longer than broad, as broad in front as behind, with feebly and broadly posterior border and feebly convex sides. Eyes rather large, feebly convex, with hs poserir ott just infront of the mide ofthe se Mandibles long, narrow, with feebly concave external borders, the apical border very long, *s except at the tip where there are four small, blunt, oblique teeth. Clypeus 6 i . Euponera (Mesoponera) subiridescens, new species. Worker. a, head from above; }, ee fe, its anterior border broadly projecting, sinuate on each side of the middle and ore deeply at each mandibular insertion. Frontal carine short, their upper es rather concave. Antennz slender, the seapes extending beyond the posterior the head a distance nearly equal to twice their greatest diameter; funicular . i and 2 eubequal, almost twice as long as broad; joints 3 to 5 somewhat shorter; i (ne ortechene be spesdhandernasn ath mma Pronotum rather ‘and rounded, as long as broad; mesonotum transverse, semicircular, sur- — eagraselteanding Mesoépinotal constriction distinct. Epinotum as as the pro- and mesonotum together, the base rounded and convex, somewhat ca? fe 84 Bulletin American Museum of Natural History [Vol. XLV lower than the mesonotum, passing gradually into the somewhat longer, sloping declivity, which is flat, bluntly marginate on the sides. Petiolar scale shaped as in ingesta, but not so thick, with the anterior surface more flattened and the posterior not impressed in the middle. Gaster short and stout, convex above, the postpetiole truncated in front, the constriction between it and the succeeding segment very feeble. Legs moderately long. Shining; mandibles more so than the remainder of the body, smooth, with only a few large punctures along the apical margin. Remainder of body very finely but not deeply punctate and less densely than in ingesta. Hairs lacking, except on the mandibles, clypeus, pygidium, and hypopygium, where they are pale yellow and rather long; the pubescence, too, is yellowish and rather long and abundant on the body and appendages, longest on the gaster. Deep castaneous, almost black; the head and thorax with a more or less distinct blue iridescence as in some species of Lobopelta (iridescens, chinensis); inner borders of mandibles, the legs, antennz, and tip of gaster somewhat paler and more reddish. Described from six specimens, all from the stomachs of toads; four from Akenge (type locality) from the stomach of Bufo polycercus, one from Medje from the stomach of B. superciliaris, and one from Ngayu from the stomach of B. tuberosus (Lang and Chapin). Both this and the preceding species seem to be very distinct from any of the previously described African species of Mesoponera. Euponera (Brachyponera) sennaarensis (Mayr) Thysville, 8, o&, 2 (Lang and Bequaert); Avakubi, 8; Leopold- ville, 8, o; Faradje, 8, 7; Medje, 8; Zambi, o; Stanleyville 9, o; Niapu, 8 (Lang and Chapin). One of the specimens from Medje was taken from the stomach of a toad (Bufo funereus). This is a well-known ant which seems to be common throughout a large part of the Ethiopian Region and even ranges into Asia (Arabia). Concerning its habits Arnold writes that it is ‘the commonest ponerine ant around Bulawayo (Rhodesia). A crateriform mound of fine earth surrounds the entrance to the nest, which is as often situated in the open as it is under stones. The economic value of this little species can hardly be overestimated, since it is exceedingly plentiful and preys un- ceasingly on termites. It is, however, omnivorous, since it will eagerly. collect bread-crumbs, insects of all sorts, and seeds of grass. Heaps of the. latter are often found in the nests.” Escherich, in Abyssinia, and Be= quaert, in the Katanga, had previously noted its fondness for are grass seeds, a very unusual habit in the Ponerine. The following note by Mr. Lang accompanies the specimens from, Avakubi: “T have generally seen this ant, which the natives call ‘tussi- somee,’ singly or two or three together, running swiftly over the sandy Wheeler, Ants of the Belgian Congo 85 ground, from which they throw up tiny craters about one inch wide and two-thirds of an inch high. These consist of excavated particles of ground loosely put together. From the crater slender channels, two to three millimeters wide, run laterally or vertically into the hard soil. When a knife is stuck into the ground near the crater, one or even three ants may be seen hurrying away. I never saw any of the larve. The BAG craters are often quite numerous. Today I counted about 60 over an area of 500 square yards. The natives say that these ants bite (sting?) | id fear them even more than the ‘siafu’ (army ants), though they never occurin masses. They build their craters in cleared ground, chiefly _ after rainy nights, and are seldom seen during the day time.” These accounts indicate that the habits of sennaarensis are very similar to _ those of the Australian E. (B.) lutea, which I have studied in New South _ Wales and Queensland. The latter species, however, prefers to nest under stones and logs and is, if anything, even more abundant than its __ African cousin. < _ Euponera (Trachymesopus) darwini (Forel) variety africana Forel A single deilated female from Stanleyville (Lang and Chapin). _ This species has an extraordinary range, from Northern Australia and > Philippines through India to Nigeria. It is very probably hypogzic habit as the worker of most of the varieties, including the African, is PLecTRocTENA I. Smith _ Large or medium-sized black or castaneous ants, with shining surface, sparsely __ Worker monomorphic, with large, rectangular and rather flat head, with small, at, anteriorly situated eyes. Clypeus very short, its anterior border straight in the iddle, emarginate on each side at the mandibular insertion, apparently not extending ck between the frontal carine, the latter overhanging the clypeus and forming th the front an elevated lobe, longitudinally sulcate in the middle. Mandibles ig, linear, feebly curved, with a deep narrow furrow running nearly their full wth on the dorsal surface, their tips blunt, the inner margin armed with a strong oth at the basal third and another obtuse tooth, sometimes indistinct, between the latter and the tip. Antenne 12-jointed, the funiculi somewhat thickened towards _ their tips, the first joint shorter than the second. Thorax large and depressed, pro- sonotal suture distinct, mesoépinotal suture obsolete, epinotal declivity marginate he sides. Petiole with a laterally compressed node, with the anterior and posterior ces vertical in profile, the dorsal surface horizontal. Constriction between the x stp iole and gaster pronounced, with well-developed stridulatory surface. Gaster t, formed largely by the first segment. Median spurs of middle and hind tibie ‘and pectinated, lateral spurs lacking. 86 Bulletin American Museum of Natural History [Vol. XLV Fema e winged, apterous or ergatomorphic, larger than the worker but otherwise similar. Eyes and ocelli small. Anterior wings with a discoidal cell, two cubital cells and the radial cell closed. Mate about the size of the worker. Frontal carinw short, erect, closely approxi- mated, bringing the insertions of the antenne close together. Antennal funiculi fili- form, their first joint very short; scapes stout, shorter than the second funicular joint. Mandibles small, linear, parallel-sided, edentate, with rounded tips. Mesono- tum with distinct Mayrian furrows; scutellum longitudinally grooved in the middle. Genitalia retracted; pygidium terminating in a blunt or truncated point. Wings short. = z. | . ~ ” ‘te fo ~ aS ay 6 i | Reese : | $.\ as . / A) fb Mf i, ‘ . X pe ! a. iat e” a > ” = : Ss = >» » Map 13. Distribution of the genus Plectroctena. This singular genus is confined to the Ethiopian Region (Map 13). _ Arnold has observed the habits of the type species, P. mandibularis, in South Africa. ‘‘The entrances to the nest are generally indicated by large heaps of earth. The chambers are placed deep below the surface, seldom less than 2 feet, and the number of individuals seldom exceeds 50. It is a sluggish and timid ant, the workers foraging singly. The food includes termites, but consists chiefly of millipeds and beetles.” Another Pe ee he ee Pe Wheeler, Ants of the Belgian Congo — 87 ‘ican species described by Arnold as P. subterranea is castaneous ures only 7.5 to 10 mm., and has exceedingly small eyes. It, probability, belongs to a different genus. In the generic key it own to Myopias and is provisionally referred to that genus. rranea are known, but no winged females of mandibularis. to Arnold, this species has ergatoid females differing ‘from - chiefly in size, but the head and abdomen are proportionally longer. The longitudinal impression on the pronotum is while that of the dorsum of the epinotum is deeper and wider. ns are to be found, and usually only one.” It seems that Forel f these ergatoid females and described it as a subspecies (major) ularis. There is, however, still another type of female, at least yr, of which I describe a specimen below, with ocelli and slightly ss than the worker and with the thorax essentially like that of «d female, but. without the slightest indications of ever having 88 Bulletin American Museum of Natural History [Vol. XLV Plectroctena cristata Emery Text Figure 16 Medje, 8; Akenge, 8 (Lang and Chapin). Eight specimens, all taken from the stomachs of toads (Bufo superciliaris, polycercus, and - funereus). f Plectroctena minor Emery A single apterous female from Akenge from the stomach of a toad (Bufo polycercus); a single worker from Niapu from the stomach of a frog (Xenopus tropicalis); Stanleyville, 8, 2, @ (Lang and —— FEMALE (apterous).— Length about 12 mm. Smaller than the winged female and with slightly smaller eyes. Ocelli present. The thorax of the same shape as in the winged female but without wing insertions. The tint of the body is a little more reddish than in the winged female. Mate (hitherto undescribed) .— Head broader than long, broadly rounded behind, the eyes large, moderately convex, about half as long as the sides of the head. Mandibles very small, blunt, edentate. Clypeus rather convex, with feebly and broadly excised anterior border. Antenne long, filiform; scape about two-thirds as long as the seeond funicular joint, first funicular joint broader than long. Thorax broader through the wing insertions than the head, narrowed in front; promesonotal suture very deeply im- pressed. Mesonotum rather flat, with a median pit in front and well-developed Mayrian furrows. Scutellum convex, with a median sulcus so that it appears bi- tuberculate. Base of epinotum somewhat longer than the declivity which is concave and strongly marginate on the sides and above. Petiole narrower, higher than long, the node truncated anteriorly and posteriorly and rounded above and on the sides; its ventral tooth triangular, short and rather acute. Postpetiole broader than long, convex above and sharply constricted off from the gaster, its anterior ventral border projecting as a transverse welt. Gaster of the usual shape, pygidium bluntly pointed _at the tip. Legs moderately long and slender. Wings rather short (7.8 mm.). Shining, finely punctate; thorax more or less rugulose, the pronotum finely, the pleurez more coarsely, the scutellum and upper portion of the base of the epinotum reticulately rugose, the latter very coarsely. Upper portion of petiolar node very smooth and shining. Hairs yellowish, present only along the posterior borders of the gastric segments. Pubescence grayish, very fine, covering the gaster, head, and legs. Black; mouth, mandibles, tibial spurs, and articulations of the legs, ventral portion of petiole, posterior and especially lateral, margins of the gastric segments, red, Wings uniformly brownish, veins and pterostigma dark brown. st The series from Stanleyville consists of a single worker, three fe- males, and two males, all fromm the same colony. Another male from the same locality and with a different number is considerably larger (13 mm.) and evidently belongs to the same species but probably represents a distinct variety which cannot be named without the worker or female. 922) Wheeler, Ants of the Belgian Congo 89 ne _.- Psatrpomyrmex Ern. André SRE Tange, dark reddish Yieown ‘or black ants with the surface of the body yered with scattered umbilicate foveole, the spaces between which are in part at ; densely striolate. . ____ In the worker and remae the clypeus is short and broad, its anterior border te and entire. The frontal carine are approximated, dilated and lobular and the antennal insertions. The mandibles have a very peculiar shape, ng long, falcate, and toothless, ending in a long acute point and broadest just | their basal third where the basal and apical borders meet without forming a ee es et other ante The antennal funiculi are slightly thickened Tg = Paris * > ue f . , ly — 4 * > SS lees to a ae) eae So Bhi PoE. i, mes = “ <= ery id S)~ th tw ee, Se ae te” FF & a Te oe eB ee ce ee oF oe 7 ee ee a eee ay, a ee ! ih ay P tm & “to — a as - = -~ ae | mt els © oe mete eae © = > = = . SE SS ; aA ; Map 14. Distribution of the genus Psalidomyrmer. lly, with the second joint conspicuously longer than the first. Eyes small and te cttcay third of tie aides of the bead. Base of . tum with a narrow longitudinal sulcus in the middle. Promesonotal suture very istinct, impressed. Mesoépinotal suture indistinct or obsolete. Petiole with high ounded node, subtruncate in front and behind. Constriction between postpetiole 1 gaster pronounced and provided with well-developed stridulatory surfaces. er small. Middle and hind tibie without lateral spurs, with a large pectinated spur. The female is winged but in other respects closely resembles the “) i = 90 Bulletin American Museum of Natural History [Vol. XLV The MALE resembles the male of Plectroctena but has smaller eyes and the mesono- tum is without Mayrian furrows, the scutellum with a deep longitudinal suleus. Only four species of this interesting genus have been described. The Lang-Chapin collection contains a fifth, which is described below. They are all rare ants, inhabiting the virgin forest and apparently restricted to Western Equatorial Africa, from French Guinea to the Northeastern Congo (Map 14). The workers of four species of Psalidomyrmez' can be readily identi- fied by means of the following table. 1. Mandibles narrow, without distinct basal and apical borders, broadest near the middle, where they are scarcely more than one-eighth as broad as long; scapes not reaching to the posterior corners of the head; petiole longer than broad; dorsal surface of body smooth and shining between the foveole. Mandibles much broader, with distinct basal and apical margins meeting at a right angle, broadest at their basal third and about one-third as broad as long; scapes reaching or surpassing the posterior corners of the head; petiole broader than long; interfoveolar surface of head, thorax and pare at least, finely striate. .. 2.0... iis enc tae es Suee'a sds be enn 2. Length 9 to 10mm. Reddish castaneous; lobes of frontal carinw smooth po shining; striz on the postpetiole longitudinal. .... . foveolatus Ern. André. Length about 12mm. Brown-black or black, with brown antennz, mandibles, clypeus, and legs; stris on postpetiole arcuate................00.00005 3. 3. Head longer than broad; antennal scapes reaching beyond posterior corners of head; stris: on the head, thorax, and abdomen sharp; pronotum without a median longitudinal groove; mesoépinotal suture obsolete; petiole slightly broader. than ‘lohg.. .. ¢<.. 6.65. bss Ue 6) ee eee procerus Emery. Head as broad as long, antennal scapes shorter; stria on head and thorax less distinct, foveole smaller; pronotum with a median longitudinal groove ; ‘ mesoépinotal suture distinct; petiole broader... ..... obesus, new species. Psalidomyrmex procerus Emery Text Figure 17 Medje, 8; Akenge, 8; Niapu, 8 (Lang and Chapin). Nine speci- mens, all taken from the stomachs of toads (Bufo superciliaris, funereus, and polycercus). Psalidomyrmex reichenspergeri Santschi Text Figure 18 = A single worker from the stomach of a toad (Bufo polycercus) taken at Akenge (Lang and Chapin). This species is easily distinguished from procerus Emery and foveolatus André by its more slender form, smoother surface between the \P. longiscapus Santschi is only known in the female sex. 91 92 Bulletin American Museum of Natural History [Vol. XLV foveole, the more rectangular head, more elongated and narrower mandibles, longer funiculi, longer petiole, and more distinct mesoépinotal suture: . Psalidomyrmex obesus, new species Text Figure 19 . WoRKER.— Length nearly 12 mm. Very similar to procerus but differing in the following chaxunanal the body is distinctly more robust, the head being rectangular, and without the mandibles as broad as long, the thorax with more rounded surfaces and a swollen appearance. The mandibles are like those of procerus but slightly broader at the angle between the basal and apical borders and the tips are less curved. The antennal scapes reach the Fig. 19. Psalidomyrmex obesus, new species. Worker. a, head from above; 6b, thorax and abdomen in profile. posterior corners of the head; funicular joints 3 to 8 as long as broad, 9 and 10 slightly longer than broad. On the thorax the mesoépinotal suture is more distinct than in procerus and there is a narrow median longitudinal furrow on the posterior half of the pronotum as well as on the base of the epinotum. The petiole in profile is much shorter and higher and, seen from above, much broader in proportion to its length than in procerus, being very distinctly broader than long, flat and truncated posteriorly, — more rounded in front, with the anteroventral tooth long and rather acute. The sculpture differs from that of procerus as follows: the longitudinal ruge covering the mandibles are distinctly coarser, the surface of the head and thorax is more opaque, the foveole being somewhat smaller, shallower and less shining, though about as numerous and the striole of the interfoveolar surface less sharp. The petiole and postpetiole are smoother and more shining than the head and thorax 2] Wheeler, Ants of the Belgian Congo 93 ee the interf ure is so feeble as to appear more or less coriaceous or ceous. The first gastric segment is longitudinally, not arcuately striolate. The ra are transversely, the scapes and tibiw longitudinally striolate as in procerus. hairs somewhat more numerous on the dorsal surface of the head and m and on the antennal scapes. coal black, darker than procerus, legs, excluding the coxe, mandibles, See eae, antenna, soar Sete Sean enmeneoun ae in D r from two specimens from Medje from the stomach of a | ifo superciliaris) collected by Lang and Chapin. This form is y distinct and is, in my opinion, more than a subspecies of : 2 zy Q x lay = a i a : as 8 Ai ENT 5 S| eit on iN { A, mt SES TCT a ah SAGUSIN sd c PL | pee Bae 7 ~ "4 ay | - a sgh Y ot ba rae Map 15. Distribution of the genus Leptogenys. This genus also occurs in Georgia. black or reddish ants, of small or medium size, sometimes with bluish e WORKERS are monomorphic and vary little in size. Mandibles articulated terior corners of the head, almost or quite toothless and either long and linear and subtriangular, usually with the angle between the basal and apical in rounded or absent. Clypeus usually carinate and projecting in the middle e form of a lobe or angle. Antenne long and slender, the funiculi not enlarged or apically. Thorax usually with the mesoépinotal suture distinct. Petiole w laterally or, in a few species, anteroposteriorly compressed. Abdomen small ‘sle , the constriction between the postpetiole and oo not very pronounced. » slender, claws pectinated. Of Bulletin American Museum of Natural History [Vol. XLV The FEMALE is wingless and scarcely larger than the worker, either highly ergato- morphic, without ocelli, with the thoracic structure as in the worker but with more voluminous abdomen, or ergatogynous, as in the case of L. ergatogyna described below, with ocelli and the thorax more like that of the winged females of other geneva, but with the mesonotum and scutellum small and depressed. The MALE is somewhat smaller than the worker and in some species much paler in color and nocturnal, with very large eyes and ocelli, very long antennz, small mandibles, and pronounced Mayrian furrows on the mesonotum. The claws are pectinated as in the other phases. Emery has divided the genus into four subgenera: Leptogenys sensu stricto; Lobopelta; Odontopelta; and Machzrogenys. The species of Leptogenys, sensu stricto, are generally distributed in the tropics of both hemispheres. One Lobopelta, L. elongata (Buckley), occurs in the Gulf States from Central Texas eastward to Florida. Odontopelta is mono- typic and confined to Queensland. Of Machzxrogenys, three species are known, all from Madagascar (Map 15). Most species of Leptogenys form small colonies, each with a single female, and nest in the ground, usually under stones or logs. The workers are timid and extremely quick in their movements. Some species make organized raids on termites; others, like our North American elongata, forage singly and apparently only at night. Leptogenys stuhlmanni Mayr subspecies camerunensis (Stitz) variety opalescens, new variety Worker.—Agreeing with the variety angusticeps Forel in all respects, except that the head, thorax, petiole, and to some extent also the gaster, have a peculiar opales- cent blue reflection like that seen in L. iridescens (F. Smith) and chinensis (Mayr). Thirteen workers taken from the stomachs of toads (Bufo funereus and polycercus) from Akenge (Lang and Chapin). Forel drew his deserip- tion of angusticeps from a single specimen taken at St. Gabriel, near Stanleyville. He says nothing about the blue reflection, which is very striking, so that I am unable to refer the specimens to his variety. The habits of the typical stuhlmanni have been studied by Arnold.! He says: I have met with this species only in Natal, where it appears to feed exclusively on woodlice; the entrance to the nest can be plainly distinguished by the accumula-— tion of the remains of their prey, bleached a dead white, scattered around it. The nest is not indicated by any mound or other accumulation of earth; but in the neighbor- hood of Durban at least, it is very frequently found in, or immediately adjacent to, the nests of Myrmicaria eumenoides Gerst. I am inclined to think that this Lepto- 11915, Ann. South African Mus., XIV, p. 93. ee a a Wheeler, Ants of the Belgian Congo 95 pecies of a part of their large nest, rather than take the one for itself. It also has a very noticeable smell, resembling sly on slaters (Oniscus and Armadillidium) and that “the earth xd g the entrances to the nests is invariably white with innum »r- g limbs and segments of the crustaceans.” The use of the 2 _ Leptogenys (Lobopelta) ergatogyna, new species ‘3 Text Figure 20 — |—6C«€#B Len, th 73 mm. eac longer than broad, narrower behind than in front, with feebly convex and rge eyes, placed a little in front of the middle, and three small ocelli, the distinctly smaller than the anterior. Mandibles rather broad, their basal een averesl, not forming an angle with each other. Clypeus carinate, tetas aoe on tao tose, the third and fourth each nearly as long as the second. Thorax long and narrow, elongate elliptical, scarcely than the head through the eyes, laterally compressed; pronotum large, as broad, depressed in profile; mesonotum, tegule, paraptera, and scutellum ed as distinct but small sclerites, without traces of wings. Mesonotum scarcely han the pronotum, somewhat longer than broad, with distinct parapteral rows. Epinotum long and sloping, without base or declivity. Petiole as high as ir aed like the quadrant of a circle, its anterior surface evenly arcuate, 96 Bulletin American Museum of Natural History [Vol. XLV its posterior surface sharply and vertically truncated, its ventral surface anteriorly with a coarse tooth. Seen from above, the petiole is only one and one-fourth times as long as broad, slightly broader behind than in front, with straight, subparallel sides. Abdomen slender, like that of a normal worker, not enlarged as in the ergato- morphic females of other species. Sting long. Legs long and slender. Subopaque; mandibles somewhat more shining, finely shagreened and coarsely and sparsely punctate. Clypeus finely longitudinally rugulose; head, pronotum, mesonotum, paraptera, and scutellum densely and finely punctate; postpetiole and gaster more shining, even more finely but a little less densely punctate; pleure finely and longitudinally, epinotum transversely and somewhat more coarsely rugulose. Petiole finely and rather irregularly rugulose. Hairs and pubescence whitish, the former very sparse, erect, delicate, confined to the head, fore cox, and tip of gaster, short on the last; the pubescence rather short and abundant on the head, postpetiole, gaster, and appendages. Black; mandibles, antennz, and legs, including the cox#, dark brown; tarsi and funiculi scarcely paler. Described from a single specimen taken from the stomach of a toad (Bufo polycercus) from Medje (Lang and Chapin). This remarkable insect I regard as the normal female of a species which must be very closely related to L. havilandi Forel, known only from the worker. In all the species of Leptogenys [elongata (Buckley), diminuta (Smith), fallax (Mayr), arnoldi Forel] of which the female is known, this phase is like the worker in the structure of the thorax and in lacking ocelli, but has a more voluminous abdomen. Of the female arnoldi, Arnold says that “‘the mesonotum is also larger and longer than in the worker,” and I have found the same to be true of the Australian fallaz. It would seem, therefore, if I am correctin my interpretation of the speci- men above described, that it must be regarded as representing a stage in the degeneration of the formicid female intermediate between the com- mon winged and the extremely ergatomorphic form, the only form of fertile female that has been seen hitherto in the genus Leptogenys. AnocHetus Mayr Worker.—Small ants with monomorphic workers. Head irregularly hexagonal. Mandibles inserted close together at the middle of its anterior border, linear, flattened, © with three large terminal teeth bent inward at a right angle and with the inner border toothless or furnished with a row of minute denticles. Eyes usually well developed, rarely vestigial, in front of the middle of the sides of the head. Clypeus small, sub- triangular, anteriorly projecting over the insertions of the mandibles and extending backward as a narrow process between the short frontal carine, which are lobularly expended in front and more or less convergent posteriorly. Antennal fovez not con- fluent behind; head without an oblique welt or swelling on each side starting from the eye and bounding the antennal fovea; sides of head without a marked impression be- hind the antennal fovea. Antenne slender, 12-jointed; funiculi long, filiform, not. enlarged apically. Thorax long and narrow, with distinct promesonotal and some- P ke Wheeler, Anis of the Belgian Congo 97 . wo mesoépinotal su ;-epinotum usually dentate. Petiole with a node or and may terminate in a spine, or in two teeth or spines. . convex above, without a constriction between the postpetiole and the r g segment. Legs slender; middle and hind tibie each with a single pecti- ; claws simple. J # very similar to the worker; usually winged, but in some species wingless » FE uz with the head of the usual shape, large eyes and ocelli and very small } antenne very long; scape short, first funicular joint broader than long. . more or less pointed or bidentate. No constriction between the le and the succeeding nema. Pygidium usually not terminating in a spine. |. 4 rR ee a ns SERN} (Pret Ky Of Tee eh ae PV ul oT = a A 5 ee ig 4 ‘oan Map 16. Distribution of the genus Anochetus. e genus comprises numerous species which form small colonies in the ground under stones or in vegetable mould. Little is of their habits. They range over the tropics of both hemi- = 16), one species, A. ghilianii (Spinola), even entering Spain M . The subgenus Stenomyrmez, of which only two species a. 1, is confined to the Neotropical Region. Anochetus africanus (Mayr) A worker and a deiilated female from Medje (Lang and Chapin) it further data. 98 Bulletin American Museum of Natural History [Vol. XLV Anochetus estus, new species WorkKER.— : Length about 4.5 mm. . Closely related to A. africanus. Head, excluding the mandibles, a little longer than broad, the posterior margin deeply and arcuately excised. Eyes small, like those of africanus. Clypeus deeply emarginate in the middle, its posterior portion long and cuneate. Frontal groove distinct between the clypeus and the middle of the head. Mandibles about half as long as the head, shaped much as in africanus but with the terminal teeth shorter and the bases somewhat narrower. Antennal scapes not reaching to the posterior corners of the head; second funicular joint not longer than broad, third scarcely longer, joints 4 to 7 not twice as long as broad. In africanus all the funicular joints are much longer. Thorax shaped as in africanus, the pronotum rounded but not convex above, the mesoépinotum long, narrower and subcylindrical, with broad blunt epinotal teeth. The petiolar scale is high and com- pressed anteroposteriorly as in africanus, with feebly excised superior border, but the latter is more acute and the sides are nearly straight and subparallel (in africanus rounded). Gaster and legs of the usual type. Shining; the upper surface of the head, except the impressions, sides and posterior corners subopaque and longitudinally rugulose, the rugules being regular and spread- ing fanwise from the frontal carine. Thorax subopaque, the pronotum longitudinally and arcuately rugulose, except in front where the rugules are transverse, the meso- and epinotum transversely rugulose. The sculpture is distinctly finer than in africanus. Petiole and gaster very smooth and shining. Mandibles very indistinctly and finely punctate, smoother than in africanus. Hairs slender, yellowish, erect, sparse on the body, absent on the appendages, which are very finely pubescent. Deep castaneous brown, almost black, with the appendages, sides and posterior corners of head, mandibles, clypeus, and tip of gaster paler brown. — A single specimen from Akenge (Lang and Chapin) taken from the stomach of a toad (Bufo funereus). Anochetus opaciventris, new species W orKER.— Length 6.5 to 7 mm. Allied to africanus. Head longer than broad, deeply and arcuately excised be- hind, with small eyes as in africanus, clypeus, frontal groove, and antenn# much as in that species, the scapes extending beyond the middle of the occipital border a distance equal to the length of the first funicular joint, the funicular joints even longer and more slender than in africanus, the third fully twice as long as broad, the second some- what shorter. Mandibles fully three-fifths as long as the head, narrowed at the base,— broadened apically as in africanus and estus, with straight internal border, the apical and preiipical teeth long and slender, the subapical very short, triangular, not longer than broad, arising from the base of the apical. Thorax and petiole similar to those of africanus but the teeth in the former longer, more acute and erect, the latter narrower, with more deeply excised superior border so that it terminates on each side in a larger and sharper tooth, and with more nearly straight, subparallel sides. In profile the anterior and posterior surfaces of the petiole are distinctly convex, the ventral surface without a tooth. — Ants of the Belgian Congo 99 — ples aiden smooth; head subopaque, finely and neguiariy longitudinally , the rug irregular but with a feebly longitudinal tend on the prono- se on the mesonotum, more vermiculate on the epinotum. Petiole ny coarsely coriaceous ; _— subopaque, densely punctate, the posterior RE ra cack tee wh ae le ry Bw and hicnt bo a8 long, pubescence. k ; mandibles, clypeus, cheeks, gular surface of head, antennz, and legs, the coxze, dark brown, the middle portions of the femora darker. Posterior gastric segments golden yellow. ed from three specimens taken from the stomachs'of toads eus and polycercus) from Akenge (Lang and Chapin). Anochetus bequaerti Forel in le specimen taken from the stomach of a toad (Bufo regularis) alr ba (Lang and Chapin). ns Anochetus punctaticeps Mayr iteer n workers from Babeyru, forming eT of a colony “found 4 Opontomacuus Latreille jum-sized or large ants closely resembling Anochetus. 1¢ WORKER, however, the antennal fovex are confluent, being united by a ion of the front behind the frontal carinz, and there is a welt or swelling which al teeth of the mandibles acute, the predpical troneated or acute, according to ; the inner border of the mandibles usually minutely and serrately toothed. ul ‘palpi 4-jointed, labial palpi 3-jointed. Eyes always well developed. Petiole i by a conical node usually terminating in a spine which is inclined back- us: winged, with large eyes and ocelli, but in other respects like the worker. 1% with the head of the ordinary shape and with very large eyes and ocelli; »s very small; maxillary palpi 6-jointed. Antennw as in Anochetus. Petiole ily with a pointed or conical node, but without terminal spine. Postpetiole d from the succeeding segment by a rather pronounced constriction. Pygi- tinating in a spine. Claws simple. tomachus is a tropicopolitan genus with apparently two centers ion, one in the Neotropical, the other in the Indonesian and Pilseicies (Map 17). One species, O. hematoda, represented by ee en 4 C eg ae he 100 Bulletin American Museum of Natural History [Vol. XLV numerous subspecies and varieties, is found in all the warmer regions of the globe, even in the Southern United States, though not in the Mediter- ranean Region. The species all nest in small colonies in the ground or in rotten wood and the workers of some of the species are very aggressive and sting severely. They are able to leap backward a distance of several inches by suddenly closing their divaricated mandibles against any hard body that happens to be in the environment. The genus is poorly repre- sented in Africa. ——— = “3 h ¢ we , aia! > re ae | te : 5 : eiat on j 4 Am “PY et: ae a! cis be v a at A bi aa PS | a a it ioe Fi es Se nS : — = ade . eee _ a 4a « ya N to hy 4 t- & , —- Map 17. Distribution of the genus Odontomachus. Odontomachus assiniensis Emery Text Figure 21 Akenge, 8; Medje, 8; Ngayu, 8; Niangara, 8; Niapu, 8. Eighty- six specimens, all taken from the stomachs of four species of toads (Bufo polycercus, superciliaris, fags. and tuberosus) collected by Lang and Chapin. Stitz has described an O. intermedius which differs from the typical assiniensis only in having the striz on the pronotum of the worker more~ arcuately concentric and therefore more as in O. hematoda and not simply transverse. A study of the long series of specimens before me shows that there is great variation in the pronotal striation, many specimens agree- ing with Stitz’s description; others having the strie in an asymmetrical whorl like that exhibited by the ridges on the tips of the fingers, and in a considerable number the strie are simply transverse, as described by assiniensis variety furvior, new variety a ¢.—Length 9 to 12 mm. Differing from the typical form in its darker abdomen is black, the head and thorax deep castaneous, almost black, ¢ ses re reek tes Rend and omonestien of ite upper legs, and antennw somewhat paler and more reddish; in some _ the tibie are stil paler t (deiilated)—Length nearly 15 mm. Darker than the worker, black, he legs and antenns dark brown. Mandibles smoother and more shining than worker. The transverse ruge on the pronotum and the longitudinal ruge ‘mesonotum much coarser than in 0. hamatoda and the whole thorax more shin- ce is lacking, even on the gaster, which is very shining. Described fe nine workers from Faradje (type locality ) and numer- ers from Yakuluku, Stanleyville, Bafwasende (Lang and Chapin), hysville (J. Bequaert). In addition to these, there are thirty 102 Bulletin American Museum of Natural History [Vol. XLV workers and one female (the one described above) taken from the stomachs of toads (Bufo polycercus, funereus, and superciliaris) from Medje, Ngayu, Akenge, Boyulu, and Niangara. Odontomachus assiniensis variety aterrimus, new variety Worxer.—Length about 10 mm. Differing from the variety furvior in being entirely jet black, including the appendages. The sculpture of the head and thorax is distinctly finer than in any of the other forms of the species, so that the surface is more shining. The legs are smoother and also more shining, especially the femora, than in any of the other forms. The unsculptured surfaces of the body, viz., the antennal foves, the mesopleure, lower portion of epinotum, and the gaster have a distinct blue opalescence. The longitudinal groove on the dorsal surface of the epinotum is continuous. Described from a single specimen found in the stomach of a frog (Rana albolabris) from Niapu (Lang and Chapin). j Odontomachus hematoda (Linnzus) Stanleyville, 8; Malela, 8, 2 (Lang and J. Bequaert); Faradje, 8; Zambi, 8, 9; Avakubi, 8; Leopoldville, $; Vankerckhovenville, 8; Garamba, 8; Akenge, § (Lang and Chapin); Matadi, 8; Katala, 2 (J. Bequaert). All this material belongs to the typical tropicopolitan form, distributed apparently throughout the Ethiopian Region. The specimen from Akenge was taken from the stomach of a toad (Bufo funereus) and a specimen from Faradje was taken from the stomach of a frog (Rana occipitalis). In connection with the well-known leaping habit of this ant, Mr. Lang makes the following remark: ‘‘ This leaping may be of some practical use to the ants when scaly ant-eaters (Manis) open their nests. Those jumping out of the immediate range of its glutinous tongue would be fairly safe, since the Manis feeds only where the ants and their larve are thickest and seldom looks for single individuals.” Odontomachus hematoda variety stanleyi, new variety Worker.—Length 7 to 8 mm. Distinctly smaller than the typical hamatoda,— with a distinctly narrower head and the mandibles, antenn, thorax, legs, and gaster paler and reddish castaneous brown. In many specimens the cheeks, clypeus, anten- nal fovew, gula, and borders of the mandibles are yellowish. Petiole with longer and more uniformly slender spine. Sculpture of the head and thorax as in the typical hematoda, but with the gray pubescence on the gaster distinctly longer and more conspicuous. The sides of the head are much less smooth and shining than in the Sr subspecies insularis (Guérin), which is of the same size though paler in color. a ee! eee ’ Wheeler, Ants of the Belgian Congo 103 — i ‘Described from numerous specimens from two colonies taken at yville (Lang and Chapin). The cocoons are also distinctly paler ngation ach, however, is the J aiahaeat of the frontal area and is often sep- d from the clypeus. Antenne 12-jointed, short. Ocelli usually developed. 5 P usually long, formed by the petiole and the postpetiole. Gaster with well- q developed sting. Middle and hind tibie with pectinate median spurs. The proven- . or “gizzard”’ is much more specialized than in the Myrmicinz, being an- AB ecloped as an apple- or quince-shaped ball, covered with longitudinal and ¢ireular muscles and with four distinct, connate sepals, bluntly rounded and finely hairy at their tips, and posteriorly as a very short, tubular, constricted portion which _ projects as a button into the cavity of the ventriculus. FEMALE very similar to the worker, also with 12-jointed antennz; either winged, and wingless, or subapterous. All three forms of Seales occur together n the same nest of Viticicola. Wings with a discoidal and a closed radial cell; two sed cubital cells, rarely one (Viticicola). ; _ Mate also rather similar to the worker; the antenne 12-jointed. External ‘a genitalia well developed, exserted; cerci present. : “The adult Larv of all four genera of Pseudomyrminz are much alike. The is long, straight and cylindrical, not broader posteriorly as in nearly all other t larve. The anterior and posterior extremities are blunt and rounded and the ments are all sharply defined. The integument is uniformly thin and perfectly nt, though tough, only the mandibles, as a rule, being strongly chitinized d the ning of the buccal cavity somewhat pigmented. The prothoracic segment compara and in certain species can be drawn down over the head; —_ »- and metathoracic segments are narrowed ventrally, the head is large, some- Si Mittened, usually subrectangular, about as broad as long and embedded in the ral portions of the thoracic segments. The antennal rudiments are always net as small, rounded papilla, each bearing three sensille. The mandibles are . Sabet and bidentate, sometimes with a vestige of a third tooth, their upper sur- ( Beovered with regular rows of subimbricate papilla. The maxille are large, n and rounded, lobuliform, the labium short and broad, with the transverse, haped opening of the salivary duct in the middle. The sensory organs which in other ants have the form of papille or pegs on the maxilla and labium are in Soe SE (TTT meme oup of sensille. The anterior maxillary organ has five, the posterior two and each ul organ has five of these sensilla. The buccal cavity is broad and transverse, its lor land ventral walls = in contact and both furnished with fine, regular trans- 7 abdominal segment is transversely y elliptical, swollen, protuberant and furnished w ‘ith a food-pouch, the trophothylar, opening forward, i. e., towards the mouth-parts. The 104 Bulletin American Museum of Natural History {Vol. XLV hairs on the body of the larva are of three kinds: first, short, stiff, very acute hairs, generally and rather evenly distributed over the whole surface (microchxtx#); second, much longer, stouter, more gradually tapering, lash-like and somewhat curved hairs of unequal length, singly or in a row or loose cluster on each ventrolateral surface of each abdominal segment (acrochetx); and third, long hairs, of uniform length; only slightly tapering, with hooked tips (oncochzt#). These are normally present in trans- verse rows of four to eight on the dorsal surfaces of the three thoracic and first three to eight abdominal segments. On the more posterior segments they are often repre- sented by simple, i. e., pointed hairs.’" Nympus not enclosed in a cocoon. In 1899 Emery,” after a comparative study of the larve of several formicid genera, proposed to separate Tetraponera and Pseudomyrma from the remainder of the Myrmicine to form the new subfamily of the Pseudomyrmine. His arguments, however, based on fragmentary material, seemed not convincing at that time; long since Emery himself has reunited these genera with the Myrmicine and in this he has been followed by all other myrmecologists up to the present. Arecentstudy of numerous larve of this group, belonging to the four known genera, has convinced me that we must return to Emery’s conception of 1899. I have endeavored to show in a recent paper* that neither the larval nor the imaginal Metaponini can be regarded as at all closely related to the Pseudomyrminz; consequently that tribe should be retained among the Myrmicine. Like the Doryline and Cerapachyine, the Pseudomyrminz are typically inhabitants of the warmer parts of the world; a small number of forms enter the southernmost portions of the Nearctic and Palearctic Regions. TETRAPONERA |. Smith WorkKerR.—Small, monomorphic or very rarely (in one South African species, 7’. ambigua Emery, according to Arnold) with the head dimorphic. Body long and slender. Head subrectangular, with large or very large, moderately convex eyes, one- third to two-fifths as long as the head; ocelli vestigial, often absent. Mandibles short and stout, with distinct basal and apical border, the latter with a small number of subequal teeth. Clypeus extremely short, steep, elevated in the middle but not extending back between the frontal carinz, the anterior border emarginate, dentate or crenulate. Frontal carine small, short, closely approximated, lobular anteriorly, often slightly diverging behind. Maxillary palpi 5-jointed; labial palpi 4-jointed. Antennz short, 12-jointed, the funiculi somewhat thickened at their tips, without distinct clava. Thorax narrow, with well-developed promesonotal and mesoépinotal ‘Wheeler, W. M. and Bailey, I. W., 1920. * The feeding habits of poestomyenine and other ants.’ Trans. Amer. Phil. Soc. Philadelphia, N.8., XXII, pt. 4, pp. 235-279, Pls. 1-v. me, 21899, ‘Intorno alle larve di aleune formiche.’ Mem. Accad. Sc. Bologna, (5) VIII, pp. 3-10, 2 "Wheeler, W. M., 1919. ‘The ants of the genus Metapone Forel.’ Ann. Ent. Soc. America, XII, pp. 173-191, 7 figs. Wheeler, Ants of the Belgian Congo 105 es and a disti tal sclerite, often constricted in the mesoépinotal region. - Epinotu and rather high, always unarmed. Petiole and often also the post- ‘iole pedunculate, rather long and slender, both with low, rounded nodes, their ral portions not swollen or with stout teeth. Gaster narrow and elongate oval, 1 well-developed, exserted sting. Middle and hind tibia with pectinated median Stites similar to the worker and scarcely larger, winged; the wings short, the rior pair with a disceidal, two closed cubital cells and a rather narrow, closed radial. Matz scarcely smaller than the worker and very similar except for the wings. i shorter. Eyes and ocelli well developed, convex. Mandibles well developed, : dentate apical borders. Antenne 12-jointed, the scape but little longer than the nd 1 funicular joint, the first joint much shorter than the second, not swollen. sc 1otum depressed, not overarching the pronotum, without Mayrian furrows and 1 very feeble parapsidal furrows. There is, at least in some species, a concavity in pro- and mesosterna, extending dorsally nearly to the mesonotal scutum. Ex- 1 genitalia well developed, exserted. Cerci present. Wings as in the female. ‘Larva hypocephalic, with papillary exudatoria on the three thoracic and first on segments. Dorsal surface with long straight hairs, hooked at their tips. -Donisthorpe (1916, Ent. Record, XXVIII, pp. 242-244) has shown Sima Roger, the name used by most authors for this genus, must be ink as an isonym of Tetraponera F. Smith, contrary to Emery’s con- sntion (1915, Zool. Anzeiger, XLV, p. 265). The case seems to be very lear, as Smith founded his genus Tetraponera (1852) on two species, rata (= Eciton nigrum Jerdon) and testacea. The latter he afterwards 858 5) placed in the genus Pseudomyrma. Roger founded his genus Sima 1863 on S. compressa Roger (= Pseudomyrma? allaborans Walker). a (1900) Emery separated the genus Sima into two subgenera, Sima, su stricto, and Tetraponera, the former with, the latter without ocelli he worker and selected Eciton rufonigrum Jerdon as the type of Sima, stricto. This was an improper procedure, since the worker of ’s type species, S. allaborans has no ocelli. Examination of the males of several of the Indomalayan species of raponera shows that they all have 12-jointed antennz. This is also e of the males of Pachysima, Viticicola, and even of Pseudomyrma ind, hence, of the whole tribe Pseudomyrmini of Emery. Nevertheless, in his recent classification of the Myrmicine (1914, Rend. Accad. Sc. ologna, p. 34) he cites the males of this tribe as having 13-jointed antennz. Bingham and Arnold also give the same number for Tetra- ra, and Santschi, who was the first to describe the male of Pachysima »ps, failed to notice that it has 12-jointed antenne. The agg Tetraponera is distributed over the Ethiopian, Malagasy, , Papuan, and Australian Regions (Map 18), being best ated in the Ethiopian and Indomalayan. One species, 7’. bifoveo- 106 Bulletin Amervcan Museum of Natural History (Vol. XLV lata (Mayr), was taken by Dr. W. M. Mann as far north as Palestine. The species all nest in plant cavities (dead wood, twigs, stems of lianas, acacia spines, etc.) and are very quick in their movements. Their habits throughout are very similar to those of the allied Neotropical genus Pseudomyrma. The species of the latter, however, are much more numerous and constitute an abundant and conspicuous part of the Neo- tropical ant-fauna, whereas the species of T'etraponera are comparatively rare ants. 15 Y) 1S 30 45 6 i) 90 105 120 135 160 Wom wwe gee = ; en RO cad EASES EA a . 4 Q = oNe zn 3 4 as 7 30 tas, "g Be Ea ae oe te N “S ory = . = Is O $ ° _ o}—- > 0 iN o~= J ‘ ~ Pe ee * H 1S ' 3 -" \ “ td DAW OMA we ‘, ae Loe Se Se ? - iS 0 1S » 4s 6 15 90 105 120 135 160 Map 18. Distribution of the genera Tetraponera (crossed area) and Viticicola (known localities indicated by crosses). Tetraponera anthracina (Santschi) Stanleyville, 8 (Lang and Chapin); Lubutu, 8; Thysville, 8 (J. Bequaert). Five specimens which agree perfectly with Santschi’s description of the types from the French Congo. Kohl found this species nesting in the hollow twigs of Barteria fistulosa and Bequaert’s specimens from Thysville bear the note, ‘‘running on leaves and twigs of Barteria fistulosa whose cavities were apparently not inhabited by ants. Forest gallery in savannah. I have not seen their nest.” a Tetraponera mocquerysi (Ern. André) variety lepida, new variety Worker.—Length 6.5 to 7 mm. Differing from the typical form of the species in color, the thorax, petiole, gaster, and cox being very dark brown or black; the head, mandibles, antenne, legs, anterior and posterior ends and ventral surface of the petiole, brownish yellow. Vertex with a large, transversely elliptical black spot ee Wheeler, Ants of the Belgian Congo 107 co tage In one specimen the posterior portion of the e of head and thorax, sculpture and pilosity of the body very ch as. is vical foam. AL a ee Very similar to the worker. Posterior borders of brownish. . Wings grayish hyaline, with pale brown veins and dark de mb ibed from two workers from Faradje (type locality) and one akuluku and a single female from Garamba (Lang and Chapin). es iiiasers mocquerysi subspecies emacerata (Santschi) 8 aleyville, $; Faradje, 8 (Lang and Chapin); Lubutu, 8; nse c yon the Semliki River, 8 (J. Bequaert). In the narrow head di coloration, the workers agree with Santschi’s figure and descrip- om. The females, two in number, are deiilated and have the head nar- ww as in the workers, but with the cheeks more concave, the anterior oe x more dilated, and the posterior corners more rectangular and less c zt . They measure 7 to 7.5 mm.; the workers about 5 to 6 mm. . 22. Tetraponera ophthalmica (Emery). Worker. a, body in profile; b, head from above. Tetraponera ophthalmica (Hmery) Text Figure 22 N ine workers taken by Dr. Bequaert at Thysville agree perfectly th Emery’s description. The species is very easily recognized by its ge eyes. The specimens were found “running on limbs of Barteria ulosa, whose cavities were not inhabited by ants.” Viticicota Wm. M. Wheeler a - Worxer.—Closely related to Telraponera. Head convex, and rounded behind, th the clypeus and mandibles shaped as in some species of T'etraponera, the external border of the mandibles deeply emarginate at the base. Eyes much smaller, only abo t one-sixth as long as the sides of the head, flat; ocelli usually absent, sometimes _the anterior present. Frontal carine short, farther apart than in T'etraponera but and closer together than in Pachysima. Frontal area and frontal groove 108 Bulletin American Museum of Natural History [Vol. XLV obsolete. Antennw short, 12-jointed; the funiculi with distinct 3-jointed club, the first funicular joint very long, joints 2 to 7 very short and transverse. Both maxil- lary and labial palpi 3-jointed. Thorax much as in Tetraponera but more thickset, the pronotum convex and rounded, not marginate or submarginate on the sides, the epinotum very high and convex, hemispherical, with the epinotal gland on each side very long and narrow, extending obliquely upward and forward to the middle of the lateral surface of the segment. Petiole and postpetiole stout, without peduncles, the nodes from above not longer than broad, their ventral portions swollen, without teeth. Gaster and tibial spurs as in Tetraponera but the tarsal claws are simple, not toothed. FEMALE winged, or ergatoid and wingless, exhibiting also subapterous forms. Even the winged form is much like the worker, but has well-developed ocelli, though the eyes are small and flat. Pronotum large and well developed; mesonotum de- pressed, flat. Petiole and postpetiole even broader and stouter than in the worker; both broader than long. Ma.e.—Clypeus longer than in the worker and female; mandibles similar with dentate apical borders. Antennz short, 12-jointed, the second funicular joint much shorter than the scape, not longer than the first, which is slightly swollen. Eyesand | ocelli rather large and convex. Mesonotum flattened or depressed, without Mayrian furrows and with very indistinct parapsidal furrows, not overarching the pronotum. There is a very deep and wide excision, separating the pro- and mesosterna and ex- tending dorsally nearly to the mesonotal scutum. Petiole and postpetiole much as in the worker and female, but with their ventral portions even more swollen and convex. Genitalia extruded, less robust than those of Pachysima and Tetraponera. Wings with a discoidal cell, a rather broad, closed radial cell and only one cubital cell. Larva hypocephalic as in Pachysima and Tetraponera and like that of the latter genus in the development of the exudatoria and dorsal hairs. GENOTYPE.—Sima tessmanni Stitz.. This monotypic genus seems to me to be sufficiently distinct from Tetraponera. The single species is highly specialized in adaptation to life in the stem cavities of a peculiar liana, Vitex Staudtii (vide infra). The eyes have dwindled and the ocelli have disappeared; the venation of the wings has become more simple and there is a pronounced tendency for the production of wingless and subapterous females—a condition unknown in any species of Tetraponera. This peculiarity, the pale color, and the small eyes indicate that the ants never leave the cavities of their host plant, except when the latter is disturbed or during the marriage flight, and the very pale color of the males indicates that this flight must occur at night. The conspicuous development of the epinotum and of its glands suggests conditions like those in some species of Crematogaster of the subgenus Physocrema (inflata, difformis, vacca, stethogompha, etc.) of the Indomalayan Region, the workers of which are supposed to feed on the secretions of one another’s epinota (Bingham). As at present known, the distribution of the new genus is restricted to Spanish Guinea and the Ituri Basin of the Belgian Congo (Map 18). It probably also occurs in Cameroon. Wheeler, Ants of the Belgian Congo 109 ____Witicicola tessmanni (Stitz) Text Figures 23 and 24 Head longer than broad, a little broader behind than in front, with feebly concave ks, rounded posterior corners and nearly straight posterior border, and, on the ex, with a short longitudinal impression at one end of which the anterior ocellus ‘imes distinctly developed. Posterior ocelli absent. Eyes very small, flat, than half their distance from the mandibular insertions, placed a little in _of the middle of the head. Mandibles short, rather strongly angulate at the . externally, their apical margins oblique, with 5 or 6 denticles, those at the base n indistinct. Clypeus convex and evenly rounded in the middle, its anterior er projecting, entire, strongly emarginate on the sides. Frontal groove absent. nz short, scapes not reaching to the middle of the head, first funicular joint . longer than broad, joints 2 to 8 much broader than long, crowded together, s 9 to 11 forming a three-jointed club, the last joint being as long as both the rs, which are subequal and somewhat broader than long. Thorax narrower than head, constricted in the mesonotal region. Pronotum from above a little broader | long, evenly rounded and convex; mesonotum transversely subelliptical, feebly vex, surrounded by impressed sutures. Metanotum nearly as long as the mesono- n, concave, with uneven surface. Epinotum very convex and rounded, egg-shaped ym above, semiglobose in profile, as high as the pronotum or slightly higher, with e slit-shaped epinotal glands shining through the integument and conspicuously larg Petiole short, scarcely longer than broad, broader behind than in front, mvex and rounded above. In profile, its ventral surface is also convex and protuber- ; b, with a small, compressed, blunt, translucent tooth anteriorly. Postpetiole a » brodder than the petiole, scarcely broader than long and scarcely broader be- id than in front, convex and rounded above and below. Legs and gaster of the usual shape, the latter with well-developed sting. _ Very smooth and shining, including the mandibles; impunctate under a magnifi- tion of 20 diameters. Hairs golden yellow, erect, of uneven length, sparse, most numerous on the especially along its sides. These regions also have more numerous short s or suberect pubescence. Antennz and legs with shorter, more appressed hairs. 's and clypeus densely and conspicuously pubescent, the latter without a fringe Clear brownish yellow, with the borders of the mandibles, clypeus and frontal » brown. Femae (deilated).— Length 4.5 to 5 mm. Very similar to the worker. Thorax elongate elliptical, somewhat flattened ove. Mesonotum as long as broad; epinotum subcuboidal, with subequal base | declivity meeting at a rounded right angle in profile, rather sharply marked off = peg sutures from the more anterior portion of the thorax. Petiole and post- + from above subequal and of similar shape, broader than long. Gaster propor- mally larger than in the worker. . Siestptace, ploaity and onor asin the worker but the hairs and puboesence longer _ and more abundant. The pubescence is very conspicuous, extending back over the 110 Bulletin American Museum of Natural History (Vol. XLV, Fig. 23. Viticicola tessmanni (Stitz). a, body of worker in profile; b, head of worker from above; c, male in profile; d, body of deiilated female in profile; e, thorax and peduncle of worker from above; fand g, thorax and peduncle of two subapterous females from above; h, thorax and peduncle of deflated female from above; i, antenna of male; j, antenna of worker. sides and front of the head and especially on the pleure, epinotum, and nodes of the pedicel. As in the worker, the hairs and pubescence are longest on the sides of the gaster. FEMALE (ergatoid).— Length 3.5 to 4.5 mm. Intermediate in the structure of the thorax, head, and abdomen between the the worker and true female, possessing ocelli and with the mesonotum varying in size,- as shown in the figures (Fig. 23f-g), as the specimen approaches the worker or female type more closely. The wings are represented by minute brownish or blackish tubercles, the anterior pair with vestigial tegule at their bases. Some specimens Wheeler, Ants of the Belgian Congo 111 ) have the fore wings more developed as a pair of triangular pads with in- ¢ veins, and folded back over the anterior corners of the epinotum. ty and pubescence are also intermediate between the worker and female; the same. h 2.6 to 3 mm. ead, including the eyes, distinctly longer than broad, rounded behind and im- ed in front of the anterior ocellus. Cheeks short. Eyes and ocelli rather large, x. Mandibles small but with distinct, denticulate borders. Clypeus convex, d rior border rounded and somewhat projecting. Frontal carine very short. scapes about three times as long as broad, funicular joints all distinctly than broad, cylindrical, very gradually increasing in length to the tip. Thorax rOv pend long, flattened above, peculiarly and deeply excavated on the ventral b i the insertions of the fore coxze; mesosterna swollen. Epinotum resembling 0 the female. Petiole and postpetiole much as in the worker, but the former sub- culate, merging more gradually into the node, without a tooth on its ventral Gaster long and slender. Fore wing with a single cubital cell. ‘Color Age yellow of a distinctly lighter tint than in the worker and female. 8 grayish hyaline, with pale brown veins and pterostigma. ies of several hundred specimens taken at Medje from the hollow is of Vitex Staudtii Guerke. The relations of the ant to the plant _ Stitz described and figured only the worker of this species from speci- mens a by Tessmann in Spanish Guinea. He gives the native ngwe name as “odschigeso” and says that the insect stings more _ He also describes one of the ergatoid females but seems to egard it as an unusual worker. In my material about 4 to 5 per cent of specimens are ergatoid females, so that they must form a normal ustituent of the colony. They probably function as egg-laying individ- sand thus supplement the reproductive activities of the true females, ch, judging from my material, are much less numerous. ‘The adult specimens of V. tessmanni collected by Mr. Lang are ace ied by numerous eggs, larve, and pupe in all stages. I have vie adult larva (Fig. 24) because it is interesting in connection _ with the extraordinay larve of the two species of Pachysima described 4 below. It resembles the larva of Tetraponera natalensis figured by x | but is longer and more slender and two of the postcephalic =! —— 41899, Mem. Accad. Se. Bologna, (5) VIII, Pl. 11, fig. 7. 112 Bulletin American Museum of Natural History [Vol. XLV segments bear appendages, the significance of which is more fully explained in my remarks on Pachysima. The prothoracie segment bears a rounded appendage on each side and applied to the side of the head, which, as in the Tetra- ponera larve, isoverarched by the protuberant, cowl-like prothoracic segment. The first ab- dominal segment bears ventrally two large and very protuberant appendages which are fused with each other in the middleline. Theanterior segments of the body have on their dorsal sur- faces clusters of long hooked hairs, as in T. natalensis, and the more posterior segments have simple stiff hairs of very unequal length on their ventral surfaces. There are also numerous short, sparse hairs, scattered over the whole body. The young larve are essentially like the oldest in form and pilosity. The mandibles are well chitinized and minutely bidentate at the tip as in natalensis, and the head bears minute rudiments of antennz onits dorsal sur- face. I find also that the larve of certain East Indian Tetraponerz, e.g., T. allaborans (Walker), have a similar structure. Viticicola tessmanni variety castanea, new variety : WorkeER and Fema.e (deiilated).—In all respects Fig. 24. Viticicolatessmanni like the typical form except in the color of the body and (Stitz). Adult larva in profile. legs, which are pale chestnut brown, with the antenne paler and more yellowish. Of this variety Mr. Lang took numerous workers and females from two colonies at Avakubi. They’were nesting in the same species of liana as the typical form. Pacuysima Emery Worker.—Closely related to Tetraponera but larger and more robust, with smaller eyes but distinct ocelli and the frontal carine decidedly longer and farther apart. Maxillary palpi 5-jointed; labial palpi 4-jointed. Both the petiole and post- petiole armed beneath with stout teeth. Claws toothed as in Tetraponera. FemaLe.—Much like the worker. Wings very long, with venation like that of Tetraponera; radial cell long and narrow. —— a) a , Wheeler, Ants of the Belgian Congo 113 - female and with very similar wings. Antennz 12- es only a little longer than the second funicular joint; first funicular pollen, much shorter than the second. Mesonotum without Mayrian fur- ‘and mesosterna not separated by a deep concavity. Petiole and post- tate beneath. External genital valves large and stout, strongly geni- t inturned points. wi hooked dorsal hairs; the exudatoria on the three thoracic seg- ‘fi ‘abdominal segment in the zee stage (trophidium) long and “~ - = a, 4 oad | 4 om Rs ~ é, t Fe. = oy TS 7 ALS Fe | Ney | Cy 5 Te V4 vs c * = ‘ | NS . a . ao S xa Ba : 3 ae es . . ws nas i : | Ser | we, r- . Distribution of the genus Pachysima and ita host plant Barteria: crossed area, distribu- jops (F. Smith); crosees, known localities of P. latifrons (Emery); heavy interrupted line, ge of the genus Barteria. cos © it is genus comprises only two known species and was originally ed by Emery as asubgenus. It is confined to West Central Africa 19), its limited range being due to the fact that it lives in the hollow bot Barteria, a genus of plants confined to the area marked on the map. 114 Bulletin American Museum of Natural History (Vol. XLV Pachysima xthiops (Smith) Text Figures 25, 26, and 27 Avakubi, 8, o; Stanleyville, 9; Ambelokudi, 8; Isangi, 8; Panga, 2; Medje, 8, 9, co; Bafwabaka, 9 (Lang and Chapin). This shining, jet-black ant, the worker of which measures 9 to 10 mm., the male and female 13 to 14 mm., is the largest, most widely distributed, and therefore best known to taxonomists of all the Ethiopian species formerly included in the genus Sima. It is represented in the a Se a Fig. 25. Pachysima zthiops (F. Smith). a, head of male; b, head of female; c, body of worker in profile; ¢, head of same from above. collection by numerous adults and larve and pupe in all stages. The specimens from Medje and Ambelokudi were living in the twigs of Bar- teria fistulosa (Plates XXVIII and XXIX). ‘When disturbed the workers came out in great numbers. The natives, who call them ‘guma- guma,’ fear them on account of their sting.” Referring to specimens taken by Tessmann in Spanish Guinea and © the Cameroon, Stitz says that ‘this ant is often found on the trunks of — Epitaberna myrmecia K. Schum., the thickened twigs of which it in- habits. It is called ‘engunkun’ by the natives and its sting is greatly ‘ feared as it is supposed to cause fever.” Father Kohl (1909, Natur u. Offenbarung, LV, p. 97, ef seq.) gives’ a much more extensive account of the habits of P. zthiops and especially Wheeler, Ants of the Belgian Congo 115 hi ia fistulosa which it inhabits. According to his obser- ons in the Congo, it is restricted to this plant and an allied species, * De Wildeman and Th. Durand. It inhabits the peculiarly lateral branches and keeps large coccids in their cavities. The 8 to the cavities are not made at definite points predetermined histological structure, as in the case of the Neotropical ened with species of Azieca. After the marriage flight $ queen gnaws its way into an already hollow twig and while Sablishing her colony the orifice, as in Cecropia, closes by growth Fig. 26. Pachysima zthiops (F. Smith). First stage larva or trophidium. a, ventral; 6, lateral view. of the plant tissue, so that it has to be reopened from within by the _ Worker s of the young colony. As several queens enter different inter- 7 des of the same plant, their various colonies probably eventually te to form a single huge colony possessing all the cavities in common, er Ay ! Biss aro extremely pugnacious and always ready for a fight as they are a equipped with excellent weapons, their stings and mandibles. If a Barteria tree is _ roughly handled or even shaken, innumerable hosts of the ants rush out of all the = and woe to him who approaches them too closely! I have had many sore experiences with their pointed stings while studying or amputating the branches. 116 Bulletin American Museum of Natural History [Vol. XLV The pain spreads instantly over the whole affected limb and continues for a long time and on the following morning returns with full intensity during one’s ablutions. One day my black servant told me that it was customary in his part of the country to punish unfaithful wives by tying them to plants inhabited by the Sima. On examining the series of xthiops larve, I was struck with their extraordinary appearance. A further study of them and of the larve of the only other known species of Pachysima (P. latifrons) throws consider- able light on the raison d’étre of the peculiar ethological relations of larval ants to their nurses, as I have shown in a recent paper. Four distinct stages, probably separated by moults or ecdyses, may be recognized in the zthiops larva. The first stage larva, just after hatching, is represented in Fig. 26a-b as it appears in ventral and lateral view. The body is curved, convex dorsally and concave ventrally, and terminates behind in a cylindrical projection, with the anus shifted to the - ventral surface near its base. The creature is strongly hypocephalic like the larve of Tetraponera , Viticicola, and Pseudomyrma, i. e., with the head on the ventral side. The head is surrounded by a cluster of promi- nent, tubercle-like appendages. On the prothorax, which is large and forms a hood over the head, there are three pairs of these appendages, an anterior truncate pair, a median pointed pair and a large posterior pair, which are swollen and rounded and embrace the sides of the head. These correspond to the single prothoracic pair figured in the larva of Viti- cicola tessmanni. The mesothoracic segment has a pair of smaller appendages nearer ‘the midventral line. Between them arises a very peculiar organ, with a swollen, pear-shaped base prolonged into a slender, apparently erectile, tentacle-like process which extends up in front of the head and terminates in a small ampulla. The first abdominal segment bears a pair of large swollen appendages, which lie at the lateral bases of the mesothoracic pair and are united witha large and very prominent midventral tubercle. This tubercle and its lateral appendages are represented in the larva of V. tessmanni but the others, with the excep- tion of the third thoracic pair, are absent. Sections and stained, cleared preparations of the whole larva show that the various tubercles contain portions of the fat-body, at least in the basal portions of their cavities, and next to the hypodermis a dense, granular substance, evidently a coagulated liquid produced by the adipocytes or trophocytes. The liquid also fills the impaired tentacle, except its pear-shaped base, which con- tains fat-cells. Around the bases of the tubercles are muscles so arranged that their contraction increases the pressure of the fat and granular 11918. ‘A study of some ant larve, with a consideration of ait origin and meaning of the social habit among insects.’ Proc. Amer. Phil. Soc., LVII, pp. 293-243, 12 figs. Wheeler, Ants of the Belgian Congo 117 J on the appendages and in all probability causes the liquid to exude th the hypodermis and delicate chitinous cuticle onto the surface. The w 10le arrangement of the tubercles, in fact, constitutes a system of exudate organs or “exudatoria,” as I shall call them, adapted to produce substance that can be licked up by the ants when they are feeding and ing for the larve. In this stage the mandibles are small, soft, and un- nized, so that the ants must feed the larva by regurgitation on liquid _ The labium of the larva has a peculiar pair of swollen appendages, wn just beneath the mandibles in the figure. The body is naked, pt for a few sparse, pointed bristles on the dorsal surface and the lian pair of prothoracic appendages. As nothing like this larval stage h among ants or indeed among the Hymenoptera, I propose to call it the “trophidium.” _ The second stage larva is shown in Fig. 27a. The various exudatoria e small in proportion to the remainder of the body but are still much 118 Bulletin American Museum of Natural History [Vol. XLV like those of the trophidium. The body is more elliptical, the mandibles are more pointed and distinctly faleate but, even in this stage, they are unchitinized and therefore nonfunctional. The coarse hairs are visible on the dorsal surface but a more uniform investment of small hairs has made its appearance. They are blunt or even clavate, especially on the prothoracic segment. In this and the trophidium stage, I am unable to find any salivary glands in cleared preparations, though rudiments of these organs may, perhaps, be present. The third stage larva (Fig. 27b) is larger and very regularly elliptical. The exudatoria can all be recognized, except the impaired tentacle. It is, however, present in some of the younger individuals but in a greatly reduced and vestigial condition at the bottom of the deep depression which now forms a definite pocket just back of the mouth and under the midventral swelling of the first abdominal segment. In many larve I found in this pocket a small rounded, dark-colored pellet which puzzled me at first. In sections it was at once seen to consist of triturated and compacted bodies and parts of small insects. It is, in fact, a food-pellet placed by the worker ants in the pocket just behind the larva’s mouth and proves to be merely the pellet which is originally formed in the in- frabuccal pocket of the adult ants. In this stage, therefore, the larva is fed on solid food and the strongly chitinized, acute, and bidentate mandibles corroborate this statement. Slender salivary glands may also be detected in this stage indicating that the substance of the food-pellet is subjected to extra-intestinal digestion. The longer hairs on the dorsal integument have almost completely disappeared. The first pair of appendages on the prothorax have disappeared and the second pair is smaller or obsolescent. In the fourth or adult stage (Fig. 27c) the larva is more elongate and cylindrical and much more hypocephalic, the prothorax forming a great protuberance in front of the head. The exudatoria are still recog- nizable, with the exception of the first and second prothoracic pairs, which have disappeared completely. The labial appendages are reduced. A food-pellet was found in the postcephalic pocket in several of the larvee of this stage but is not represented in the figure. The coarse hairs have disappeared from the integument, which is now uniformly covered with very short, delicate hairs and the structure of the posterior end of the body is very different from that of the preceding stages. The conclusions which I draw from the study of these lave ea from those of P. latifrons and Pxdalgus infimus (vide infra) are that the young larve are fed by regurgitation, the older larve with pellets of Wheeler, Ants of the Belgian Congo 119 _ erushed insects, and that,-especially during their younger stages, the - larve are so assiduously fed and cared for because they furnish liquid exu- dates, small in quantity, to be sure, but of such a quality as to excite the appetite of their nurses and induce regurgitation. I believe that the z salivary glands, as soon as they develop, take on the function of supplying exudates and at the same time aid in the extra-intestinal digestion of the food placed in the postcephalic pocket. That the salivary glands may be important as exudate organs throughout life is indicated by certain genera of Myrmicinz (e. g., Pedalgus), the larve of which have no exuda- toria but greatly developed salivary glands, though the latter are never ‘used for spinning cocoons in the prepupal stage. Thus in ants very much the same “cecotrophobiotic” relations exist between the adults and young as Roubaud! has so beautifully described for the wasps of the _ genera Belonogaster, Ropalidia (=Icaria), and Polistes. To these rela- tions, established by a mutual exchange of food-substances and which I Ppcige called “trophallactic,” the social life of ants in all probability owes its origin, development, and maintenance. Moreover, the exudates of ~ larval ants are strictly comparable with those of various castes of _ termites among themselves, of the queens of parasitic ants and even of _ workers (e. g., Crematogaster inflata of the East Indies), with the excre- q ment of coccids and aphids, the secretions of lycenid larve and the __ nectar of the extrafloral nectaries of plants. Thus trophallaxis, myrme- _ cophily, termitophily, trophobiosis, and the relations of ants with certain _ plants (myrmecophytes) are all seen to be merely so many particular _ manifestations of the same fundamental instinct of ants to foster and _ defend and, if possible, to feed and transport any small living object _ which can furnish droplets of agreeable secretion or exudates. _ _—sC The only account of the xthiops larva in the literature is by Emery.’ _ He describes the adult larva very briefly and figures its anterior end . a with some of the exudatoria but erroneously interprets the large pro- Pe _ thoracic pair as “‘ébauches de pattes,”’ or rudiments of the anterior pair of a In the same paper Emery created the subgenus Pachysima for the ~ accommodation of xthiops and latifrons, because those species have the i frontal carine of the worker and female much more widely separated _ than the numerous other species of Tetraponera (= Sima). I have raised J Pachysima to generic rank, because the larve of the two species are so .- nik different from those of Tetraponera. ont Se ya biologiques sur les guépes pdliaaieds ¢ et i caidas d’ Afrique.’ Ann. Se. Nat. Zool., pn 1 jon Soc. Ent. Belgique, LVI, p. 97. 120 Bulletin American Museum of Natural History [Vol. XLV Pachysima latifrons (Emery) Text Figures 28, 29, 30, and 31 Worker.—Length 7 to 8.5 mm. Similar to the worker of P. xthiops but smaller, smoother and more shining, and much more finely punctate, with the frontal carinze somewhat farther apart and more nearly parallel. The mandibles have less oblique apical borders and are smooth and shining and sparsely punctate, not coarsely striated as in xthiops; the mesonotum is shorter and semicircular; the epinotum in profile somewhat lower and more rounded; the petiole bears on its ventral surface a single large acute, backwardly directed spine, instead of two spines, and the postpetiole has in the same relative position a smaller spine of similar shape, representing the larger, blunter projection of zthiops. The erect hairs and pubescence on the body are dis- tinctly more abundant in latifrons, and the clypeus has a conspicuous fringe of yellow ciliary bristles, which are not developed in zthiops, and the antennal scapes have a row of long scattered hairs on their anterior surfaces. There is no difference in coloration. Fig. 28. Pachysima latifrons (Emery). a, body of worker in profile; b, head of worker from above; c, head of female. FemaLe.—Length nearly 12 mm. Closely resembling the worker and differing by the same characters from the female of zthiops. The head and thorax are more slender than in the latter species and the petiolar and postpetiolar nodes are narrower and less submarginate on the sides. The pilosity and pubescence are much less de- veloped on the body than in the worker, though the clypeus has conspicuous yellow ciliary bristles and the antennal scapes have a few long hairs along their anterior surfaces. The wings are blackened like those of xthiops. _ Described from numerous workers and a single female taken from a— colony at Niangara (Lang and Chapin), also in hollow twigs of a Barteria, presumably B. fistulosa. This species appears to be confined to western Africa; its distribution is still imperfectly known. The larval stages are quite as remarkable as those of P. xthiops and exhibit four stages as follows. 1922] Wheeler, Ants of the Belgian Congo 121 _ The trophidium, or first stage larva, shown in Fig. 29a-b, is very ____ hypocephalic, the prothoracic segment being greatly enlarged and pro- ___ jecting anteriorly. Stained preparations in toto and sections show that ___ the portion of the fat-body in this segment is heavily charged with urate __ erystals, so that it undoubtedly functions as a storage kidney till the __ Malpighian vessels are sufficiently developed to excrete. The first and __ second pairs of prothoracic appendages of the xthiops larva are absent, but the third pair is very large and embraces the sides of the head. The = -meso- and metathoracic segments each bear a pair of slender, pointed __ appendages, the first abdominal segment a huge leg-like pair which are a : Fig. 29. Pachysima latifrons (Emery). First larval stage or trophidium. a, ventral; 5, lateral view. swollen and fusiform at the base and running out into a slender process ___ which forms an obtuse angle with the basal portion. The sternal region between these appendages is protuberant and its cuticular covering, like that of the four pairs of appendages, is minutely prickly, unlike the smooth cuticle of the remainder of the body. Sections show that both ‘the four pairs of appendages and the sternal swelling are exudate organs, though the prothoracic and abdominal] pairs are evidently much more im- portant than the others. The prothoracic appendages are filled with blood and very little fat-tissue, but their hypodermis is much thickened 122 Bulletin American Museum of Natural History [Vol. XLV and consists of crowded cells arranged in peculiar clusters. In section, the abdominal appendages appear as in Fig. 30. The fusiform base is filled with large, clear trophocytes, or fat-cells, some of which in the middle of the swelling may be filled with urate crystals, like those in the prothoracic storage kidney, but the slender, tubular distal portion con- tains a granular liquid which can only be regarded as an exudate derived from the trophocytes in the basal enlargement. This exudate is evi- dently filtered through the thin cuticle covering the appendage by pres- sure, for there is a rather elaborate system of muscles, as in the xthiops — larva, surrounding the bases of the appendages and capable of subjecting their contents to pressure. The head is small and has soft, blunt, rudi- mentary and unchitinized mandibles and the labium bears a pair of long, Fig. 30. Longitudinal section through exudatorium of first abdominal segment of trophidium of Pachysima latifrons (Emery): t, trophocytes or fat-cells of the fusiform base, some of them in the middle with urate crystals (uv); z, granular liquid or exudate filling the distal portion; d, dermis; e, epidermis; h, hypodermis. palp-like appendages, which project forward in the deep depression between the head and the swollen sternal portion of the first abdominal segment. These are probably also exudatoria and seem roughly to correspond to the unpaired tentacle of the xthiops larva. The structure of the mouth-parts shows that the larva in this stage is fed with liquid food regurgitated by the workers. The convex dorsal surface is beset — with sparse, curved bristles of uniform thickness, with blunt tips. The _ segmentation of the body is indistinct and its posterior end curves for- — ward and terminates in a large tubercle with the anal orifice just anterior to its base. The Malpighian vessels have only just begun to develop at. the blind end of the proctenteron where it abuts on the posterior end of the large, elliptical mesenteron, or stomach, but no salivary glands can be detected. —- ‘ q he 4 a ee —-:1922] Wheeler, Ants of the Belgian Congo 123 _____ In the second stage larva (Fig. 3la) the body is more elongate and _ eylindrical and the four pairs of appendages can still be recognized _ though considerably smaller in proportion to the remainder of the body. _ The mandibles are becoming chitinized. Many of the long hairs on the ~ dorsal surface are still present, but a general covering of short, sparse _ hairs has made its appearance. The third stage larva (Fig. 315) is larger and still more elongate and ¢ _eylindrical and shows a further regressive development of the exudatoria. Those on the meso- and metathoracic segments have disappeared and Fig. 31. Pachysima latifrons (Emery). a, second stage larva; 6, third stage larva; c, anterior half of fourth, or adult, larval stage. the abdominal pair has short broad bases with the distal portions at- ___ tenuated to slender points. The labial appendages have also disappeared. __ The mandibles are well developed and chitinized, and the larva is now fed with pellets of crushed insects, like the xthiops larva in the corres- ponding stage. These pellets were found-still in situ in several of the alcoholic specimens as represented in the figure (Fig. 31b). The pellet lies in the deep pocket between the head and the sternal protuberance of the first abdominal segment and is, therefore, within easy reach of the 124 Bulletin American Museum of Natural History [Vol. XLV mandibles and labium of the larva. Cleared preparations show that the salivary glands have made their appearance, though they are small and slender. ; The anterior end of a fourth stage or adult larva is shown in Fig. 3lc. The exudatoria of the prothoracic segment now appear merely as a pair of welts or folds embracing the sides of the head and continuous with the more dorsal portions of their segment, which is relatively smaller and less projecting than in the preceding stages. The appendages of the first abdominal segment are still distinct but their distal portions are reduced to mere points, sometimes absent in larve just before pupation, and the sternal swelling is much less prominent. In this stage the larva resembles that of T'etraponera throughout its various stages. In the third and fourth stages of the latifrons larva, as in the corresponding stages of zthiops, the salivary glands probably furnish secretions which are useful both in the extra-intestinal digestion of the food-pellet and as exudates that can be imbibed by the workers. Myrmicine WorkER monomorphic, dimorphic, or polymorphic, often very strongly so; the soldier form having a very large head and strong mandibles. Frontal carine nearly always separated, rarely close together; divergent or slightly convergent behind and rarely lobed anteriorly; usually the clypeus is wedged in between the frontal carinw; in the Metaponini and a few other forms the clypeus is not prolonged back, its pos- terior margin being rounded. Antenne from 4- to 12-jointed, often with a distinct club. Ocelli frequently absent in the ordinary worker, though in strongly dimorphic species they may still be more or less distinct in the soldier. Pedicel formed by the petiole and the postpetiole; very rarely (Melissotarsus) the postpetiole is nearly as wide as the basal segment of the gaster. Stridulatory organ usually present at the base of the gaster. Sting developed. Spurs of the middle and hind tibiw in the majority of cases simple or absent; pectinate in the Metaponini and Myrmicini only. Gizzard simple and tubular in most genera and of a very primitive type com- pared with the conditions in the Dolichoderine, Camponotinz, and Pseudomyrminz. FEMALE usually winged and larger than the worker; in a few cases ergatoid; true dichthadiiform queens are not known, but in some parasitic genera (Anergates, Anergatides) the gaster of the fertile female becomes enormously distended. MALE usually with the copulatory armature partly exserted; entirely retractile in a few genera of the Solenopsidini only. Anal segment with cerci. In a few cases (as in certain species of Cardiocondyla) ergatoid, wingless males are known, sometimes together with winged individuals. Antenne almost always 13-jointed, even when the. worker and female have very few antennal joints (11-jointed in Stereomyrmez and Cataulacus; 12-jointed in Metapone, certain Attini, Meranoplini, etc.). The venation of the fore wing offers much diversity. In some genera the more primitive type is still retained, with a closed radial, two closed cubital cells, and a closed discoidal cell, but all degrees of reduction are met with. When there is only Wheeler, Ants of the Belgian Congo 125 vital ce mm may be united with the radius by means of a long inter- Fiieaiat Getonopets or the intercubitus may disappear, the cubitus and radius ised in a spot or for some distance (type of Formica). anv thick-bodied, orthocephalic, without exudatory papillz around the mouth. dy ; is, as a rule, abundantly covered with chitinous hairs of very different kinds; PHS never enclosed in a cocoon. ‘Myrmicinz is the largest subfamily of ants, containing over 120 and many thousands of described species, races, and varieties, as many as the other six subfamilies together. As would be ex- ected, the taxonomic arrangement of this maze is exceedingly difficult ad ‘it is no wonder that such keen myrmecologists as Forel and Emery ve not yet succeeded in reaching satisfactory results and are obliged modify their views at every turn of the road. For practical and other , have felt at liberty to change somewhat the classification sed by Emery,' though have followed him in the main. Have d the two tribes Solenopsidini and Pheidologetini, which pass tedly into each other and are merely separated by the shape of the cell (closed in the Pheidologetini; open in the Solenopsidini), a ster the value of which seems to have been overrated by Emery. also accepted Forel’s tribe Proattini and, furthermore, separated ex from the Dacetini as an independent tribe. The very iar genus Archzomyrmez, recently discovered by Mann in the Fiji s, must also constitute a distinct tribe, which I have provisionally $ are granivorous, the most prominent in this precio being the bers of Messor and allied genera (Novomessor, Veromessor, Oxyopo- t, Pogonomyrmex, many species of Pheidole, etc.). In these ants nest often contains spacious granaries full of seeds. Many myrmicine are attracted by sugary substances such as are furnished by the ti aries of flowers or various extrafloral plant organs. Often, also, y attend aphids, coccids, psyllids, or leafhoppers for the sake of the ney’ they excrete. The New World “leaf-cutting” or “fungus- ing’ ants of the tribe Attini feed exclusively on the food-bodies pmatia’’) produc’d by fungi cultivated in their nests. There are apn y apne ee Accad. Se, Bologna, 1914, pp. 29-42. tour genres wt de genres Propon pour Gores amille dee Myrmicinag; modifica A la 126 Bulletin American Museum of Natural History [Vol. XLV lead to the disappearance of the worker caste (Wheeleriella, Epixenus, Epipheidole, Sympheidole, Epecus, Anergates, Anergatides, and probably several other genera of which only males and females are known). Temporary social parasitism is probably the rule in some species of Aphznogaster and in the Malagasy and Indomalayan subgenus Oxygyne of Crematogaster. PHEIDOLE Westwood Small ants with the worker strongly dimorphic, the two forms being designated — as the worker and soldier. In a few species these phases are connected by inter- mediates (medi). So.prer with very large head, subrectangular or subcordate, more or less deeply notched or excised behind and with a distinct occipital furrow, on each side of which the occipital region is convex. Clypeus short, depressed, carinate or ecarinate but not elevated in the middle, the anterior border entire or notched in the middle, the posterior border extending back between the frontal carine, which vary in length, being short in some species and in others greatly prolonged backward and forming the inner borders of more or less distinct scrobes for the antennz. Frontal area usually distinct, deeply impressed. Mandibles large, convex, usually with two apical and two basal teeth, separated by a toothless diastema. Antenne 12-jointed; the funiculus with long first joint; joints 2 to 8 small and narrow; the three terminal joints forming a well-developed club. Thorax small, usually with distinct promesonotal and meso- épinotal sutures and pronounced mesoépinotal constriction; the pro- and mesonotum raised, more or less convex, the humeri sometimes prominent, the mesonotum often with a transverse welt or torus; the metanotum sometimes represented by a distinct sclerite; the epinotum armed with spines or teeth, in profile with distinct basal and declivous outline. Petiole small and narrow, pedunculate anteriorly, the node posterior, compressed anteroposteriorly, its superior border sometimes emarginate, the ventral surface unarmed. Postpetiole broader than the petiole, convex and rounded above, contracted behind, the sides often produced as angles or conules, more rarely as spines. Gaster rather small, broadly elliptical or subcireular. Femora more or less thickened in the middle; middle and hind tibie without spurs; tarsal claws simple. Worker smaller than the soldier but very similar in the structure of the thorax, pedicel, and gaster; the head, however, much smaller, not grooved nor deeply excised posteriorly; the antenne longer; the mandibles less convex, with evenly denticulate apical borders. The pro- and mesonotum are proportionally less convex, and the petiole and postpetiole are more slender. FEMALE resembling the soldier but larger; the head proportionally smaller and shorter, usually not longer than broad and not broader than the thorax; the occiput _ only broadly and feebly excised. Thorax broad and massive; the mesonotum flat, overarching the pronotum in front. Epinotal spines shorter and stouter; petiole and postpetiole more massive; gaster much larger and more elongate than in the soldier. Wings long, with a discoidal cell, two closed cubital cells, and an open radial cell. MALE decidedly smaller and more slender than the female, the head small, with large, convex eyes and ocelli; mandibles small but dentate. Clypeus longer than in pee ‘= = ; Wheeler, Ants of the Belgian Congo 127 ‘the soldier. “Aptenees-18-Jointed; the scapes very short, scarcely longer than the second funicular joint, first joint subglobular. Thorax broad; the mesonotum flat- ened, without Mayrian furrows, anteriorly overarching the small pronotum; epino- tum unarmed. Petiole and postpetiole slender, with low nodes. Gaster slender, te. Genital appendages small. Cerci present. Legs long and slender. Wing ic fab in the female. The species of this very ais and difficult genus are distributed over _ tropics and warmer temperate areas of both hemispheres (Map 20). pets earctic Region the northernmost range is southern New England 44 ae fol is . by am NE s Katy eet h {pm Aoth | et 2 Riba ais Ty : nt wy h Y A > -~ “| AAS a SEAN iA Pa z Ver yp 1 hes = ” nS > 7” - oo - ~ = Map 20. Distribution of the genus Pheidole. and iOregon; in the Palearctic, Japan and northern Italy; in the southern misphere it reaches Argentina and Tasmania. Emery has divided the jus into a number of subgenera and has rejected a couple of subgenera, and Cardiopheidole, described by Forel and myself. The groups have been characterized by Emery ina recently published n of the ‘Genera Insectorum’ on the Myrmicine. “ss Nearly all the species of Pheidole nest in the ground, either under stone and logs or in crater or small mound nests. Many species feed ex- sively on insects and often have a peculiar fecal odor precisely like it of the Dorylinz, which also have an insect diet; but many species harvesters and store the chambers of their nests with the seeds of herbaceous plants. This is especially true of the desert species of . In some species in Australia and the southern United States, the soldiers take on the function of repletes and store in their crops sweet Sy : eos, =J 128 Bulletin American Museum of Natural History [Vol. XLV liquid for the use of the colony during periods of food and water scarcity. One species, Pheidole megacephala, has been carried to all parts of the tropics and has become a great pest in and about dwellings and planta- tions as it assiduously cultivates coccids on many economic plants and ruthlessly destroys and replaces the native ant-faunas. This has been observed in the Madeira Islands, Hawaii, Australia, and the West Indies. In all probability P. megacephala is of Ethiopian or Malagasy origin, as it shows a great development of subspecies and varieties in these two regions and nowhere else. Pheidole batrachorum, new species SoLpier.— Length 4.5 to 5 mm. Allied to P. caffra Emery. Head a little longer than broad, scarcely narrowed in front, with straight sides and deeply excised posterior border, the vertex convex, the occipital region distinctly depressed, the occipital and frontal groove shallow. Eyes small, broadly elliptical, rather flat, at the anterior third of the sides of the head. Mandibles convex with bluntly bidentate tips. Clypeus flat, carinate, its anterior border notched in the middle. Frontal area small, subtriangular, deeply impressed, without median carinula. Frontal carine not strongly diverging behind, backward as a pair of ruge to the posterior fifth of the head and forming the inner borders of flat, serobe-like impressions for the antenne. The latter slender, their scapes distinctly flattened but not dilated at the base, extending to nearly half the distance between the eyes and the posterior corners of the head; club shorter than the remainder of the funiculus; joints 2 to 8 distinctly longer than broad. Pro- and mesonotum not separated by a suture, convex; humeri prominent; mesonotum with strong transverse torus; mesoépinotal constriction very sharp and deep; epinotum broader than long, its base straight and horizontal, as long as the declivity, dorsally with a broad longitudinal groove; the spines acute, stout at the base, as long as the base of the epinotum and as long as their distance apart, directed upward and some- what backward and distinctly curved downward Petiole twice as long as broad, searcely broader behind than in front, with nearly straight sides; in profile with long, feebly concave anterior and short, vertical posterior surface to the node, the superior border transverse, sharp and feebly emarginate. Postpetiole nearly three times as. broad as the petiole, broader than long, very convex and rounded above, the sides. bluntly angular in the middle. Gaster smaller than the head, subcircular, its anterior ; _ border slightly truncated, the dorsal surface somewhat depressed. Legs long, femora. thickened in the middle. Subopaque; mandibles, clypeus, frontal area, and posterior half of gaster smooth — and shining. Mandibles coarsely and sparsely punctate; coarsely rugose at the base. Clypeus very finely rugulose, especially on the sides. Head densely and finely, but not. deeply punctate, longitudinally rugose, the ruge being rather widely separated and subsiding on the posterior fifth of the head; the posterior fourth also with a few large, shallow, elongate foveole. Thorax, pedicel, and anterior half of gaster more opaque than the head, finely and densely punctate; the pronotum also finely and rather asymmetrically transversely rugulose. Mesoépinotal constriction with sharp- Wheeler, Ants of the Belgian Congo 129 Sil iiiadieatE comn;.declivity. of epinotutn tranavereely rogoee above. ! of gaster with sparse, elongate, piligerous elevations. Legs smooth and s coarse, pointed, fulvous, long, and erect, lacking on the thorax and sides of se on the pedicel and gaster and front of head; short and closely appressed -piceous, black; mandibles, clypeus, cheeks, and appendages ; the funiculi, tips of scapes, tibie, tarsi, and articulations of the legs paler gth 3 to 3.5 mm. d (without the mandibles) nearly circular, the occipital border strongly . Byes rather small but convex, just in front of the middle of the sides of ‘Mandibles long, deflected, their external borders concave, their tips with minent teeth, the remainder of the apical border finely denticulate. Antenne lon and slen der, the scapes extending fully one-third their length beyond the occipital be of the head. Clypeus rather flat in the middle, ecarinate, its anterior border » broadly rounded. Thorax resembling that of the soldier, but the humeri at, the torus of the mesonotum is feebler, the epinotal spines are more ‘distinctly shorter than the base of the epinotum and more curved than ier. Petiole more slender, the node lower, more conical, its superior border nate, scarcely more than twice as long as broad. Postpetiole campanulate, broad, broader behind than in front. Gaster elongate elliptical, with trun- rior border, its dorsal surface convex. Legs long and slender. g; mandibles very finely and densely striolate. Clypeus, head, thorax, and nsely punctate or reticulate; the head somewhat smoother and more shining dle anteriorly; the sides of the pronotum smooth and polished; cheeks and nt with a few longitudinal rugules. Base of first gastric segment sculptured as in the soldier. _ Hairs less coarse than in the soldier, present also on the thorax; hairs on the legs ‘antennz longer and more abundant, on the scapes abundant and oblique. jlor very much like that of the soldier. eribed from four soldiers and twenty-one workers from Akenge id Chapin), all taken from the stomachs of toads (Bufo polycercus reus) and frogs (Arthroleptis variabilis). utu, 2, 8, 9, “taken from a colony under bark of a fallen tree x (J. Bequaert). I refer numerous specimens from these localities tschi’s variety, because they are of very small size and dark color, diers measuring only 3.5 to 4 mm., the workers 2 to 2.5 mm. 130 Bulletin American Museum of Natural History [Vol. XLV The type of the species is considerably larger (soldier, 4.6 to 5 mm.; worker, 3 mm.). According to Santschi, the species varies much in stature and color. The females from Walikale measure 7 mm. and are dark brown, like the soldiers and workers, with dull yellowish brown wings. If I am correct in my interpretation, attenuata would more properly constitute a distinct subspecies. Pheidole caffra mery subspecies bayeri Forel variety thysvillensis, new variety So.prer.— Length 4 to 4.5 mm. Smaller than the typical bayeri, with the head of the same shape, but subopaque and with only the front and occiput somewhat shining. The occipital depression is less distinct than in the subspecies abyssinica Forel, and the rugw are anteriorly less numerous, coarser, and farther apart, but very fine and distinctly transverse on the occiput. The antennal scapes are shorter than in the typical bayeri, reaching only a little beyond the middle of the head. The suberect epinotal spines are not pointed as in abyssinica and bayeri but somewhat longer, of uniform thickness or even slightly enlarged at the tips, which are blunt. The base of the epinotum is not longer than broad. The postpetiole is somewhat narrower than in bayeri and abyssinica, with blunter lateral angles. Thorax, petiole, and postpetiole more finely rugulose-punctate than in abyssinica; gaster shining, with the base of the first segment subopaque and alutaceous. Color as in abyssinica, with the head and thorax ferruginous brown but varying in some specimens to pale ferruginous red, with the gaster black or brown and the base of the first segment and posterior borders of all the segments paler and more reddish or yellowish. Worxer.—Length 2 mm. Smaller than the worker of bayeri. Head elliptical, without posterior corners, longer than broad. Antennal scapes extending two-fifths their length beyond the occipital border, which is rather sharply marginate. Shining; head and thorax finely reticulate; mesonotum, epinotum, petiole, and ventral and lateral portions of the postpetiole opaque and densely punctate. Ferruginous brown; head castaneous; mandibles except their teeth, yellowish. Described from numerous specimens taken both by Lang and Bequaert at Thysville, apparently from the same colony, “ nesting in sandy soil in the savannah.” Pheidole caffra subspecies senilifrons, new subspecies Text Figure 32 Sotpier.—Length 4 mm. Differing from the typical form and the subspecies bayeri in the sculpture of the head, the sharp longitudinal ruge between the prolonged frontal carine being surrounded by the ruge from the sides of the head, which run up to the posterior corners, then turn at a right angle and run transversely on the occipital lobes to the occipital furrow. These ruge are quite as strong as those on the front, but denser. The head is a little longer and a little more depressed posteriorly than in the variety thysvillensis, the transverse welt of the mesonotum less pronounced; the blunt epinotal spines distinctly shorter. The sculpture of the thorax and pedicel and the color and pilosity are much as in that variety. - — [1922 Wheeler, Ants of the Belgian Congo 131 Fig. 32. Pheidole caffra subspecies senilifrons, new subspecies. Soldier. a, body in profile; 6, head from above. _ Worker.—Length 1.8 mm. Very similar to the worker thysvillensis but the pronotum is smooth and shining and the epinotal spines are shorter, less obtuse, and more erect. __ Four soldiers and a single worker from Yakuluku, where they were found “nesting in a small mushroom-shaped termitarium”’ (Lang and Chapin). Pheidole kohli Mayr A single soldier from Medje (Lang and Chapin), without further data, agrees very closely with Mayr’s description of this species. Pheidole kohli Mayr, variety A single imperfect soldier and five females, three of them winged, taken from the stomachs of a toad (Bufo regularis) and a frog (Rana ornatissima) from Garamba (Lang and Chapin), appear to represent an undescribed variety or subspecies of kohli, the soldier being darker and having a distinctly narrower head. The pedicel, gaster and funiculi are, however, lacking in the single specimen of the soldier. It seems to be undesirable to base a new name on such defective material. Pheidole megacephala (abricius) Niangara, 8; Akenge, 9; Stanleyville, 9; Banana, a, 8 (Lang rE and Chapin); Zambi, 2, 8, 2 (Bequaert and Lang); Matadi, 2, 8; Thysville, 8; Boma, 2, 8, 2; Malela, a, 8, 9 (J. Bequaert). All these specimens belong to the typical form of this well-known tropicopoli- tan pest. I have been unable to recognize among it Forel’s subspecies nkomoana, originally described from the vicinity of Stanleyville. In the 132 Bulletin American Museum of Natural History (Vol. XLV colony taken at Zambi by Lang and Bequaert there are several specimens of an interesting Microdon larva, which is figured and described in Part VI. The female specimens from Akenge and Stanleyville, five in number, were taken from the stomach of a toad (Bufo polycercus) and a frog (Rana mascareniensis). Pheidole megacephala subspecies ilgi (Forel) A soldier and several workers taken by Dr. Bequaert at Lesse from a colony nesting at the base of a papaya. It was on the head of one of the soldiers in this colony that he found a singular phorid fly, Plastophora aculeipes (Collin), subsequently referred to by H. Schmitz." Pheidole megacephala subspecies melancholica (Santschi) Six soldiers, five workers, and seven females, mostly winged, taken at Garamba (Lang and Chapin) from the stomachs of a toad (Bufo regu- laris) and two frogs (Rana ornatissima and Kassina senegalensis). The female is a little larger than the female of the typical megacephala, with the head and thorax more sharply sculptured and the color of the body, including the clypeus and mandibles, darker, almost black; the legs more yellowish, as in the worker. This is the host of the singular workerless parasitic ant, Anergatides kohli, recently described and figured by Wasmann from the vicinity of 4 Stanleyville* Pheidole megacephala subspecies punctulata (Mayr) Boma, 2,8, 9; Ngayu, 2%, 8; Avakubi, 2, 8; Stanleyville, a, 8, 29, @; Bolobo, a, 8; Faradje, a, 8; Zambi, 2, 8, 9; Niapu, 4,%; Garamba, 2,8; Banana 2, 8 (Lang and Chapin). A well-known and widely distributed Ethiopian form, apparently more abundant in the Belgian Congo than the typical P. megacephala. The specimens from various colonies show considerable variation in color, some being dark brown, others pale and more yellowish or reddish, especially those from Stanleyville and Banana. Mr. Lang gives the native name of the species as ‘‘tuegeke”’ and his notes give the nesting. sites as “under heaps of decomposed, moist grass,” “‘in fallen stems of Hyphezne,” “in mushroom-shaped termitaria in swamps,” and “in the tops of termite mounds.” 11916, Zoolog. Meded. Mus. Leiden, II, p. 28. 71915, Ent. Mitt. Deutsch. Ent. Mus. Berlin, IV, p. 281. 1922] Wheeler, Ants of the Belgian Congo 133 Pheidole minima Mayr subspecies matlelana, new subspecies Saath 2.3 men. Head shaped much as in P. megacephala, without the mandibles a little longer Sigg Pilchard och min P. mepncepale, without the mancibes Eyes small, flat, at the anterior third of the head. Clypeus flat, ecarinate. Frontal area small, impressed; frontal carine diverging, reaching to the posterior third of the head, = _ bounding distinct scrobes for the antennal scapes, which are half as long as the head. -__ Punicular joints 2 to 8 distinctly broader than long, club longer than the remainder of the funiculus. Mandibles large and convex, coarsely bidentate at the tip. Thorax ____ robust, pronotum very convex, with small but distinct humeral tubercles. Mesono- tum falling almost vertically to the pronounced mesoépinotal constriction, with a slight transverse convexity in the middle. Epinotum broader than long, concave and sloping in the middle, its spines rather erect, shorter than the interval between their bases, with pointed tips. Petiole with rather high, anteroposteriorly compressed, _ distinctly emarginate node. Postpetiole only one and one-half times as broad as the petiole, broader than long, with the sides angularly produced. Gaster much smaller than the head, elliptical, convex, with subtruncate anterior border. Legs stout, femora thickened in the middle. _ Shining; mandibles sparsely punctate; clypeus rather smooth in the middle, indistinctly rugulose on the sides; anterior two-thirds of head with sharp, but not coarse, longitudinal ruge; occipital lobes with small, sparse, piligerous punctures. Pronotum and gaster very smooth and shining; pedicel smooth but less polished; -meso- and epinotum opaque, densely punctate. Hairs yellow, sparse, suberect on the body, short and appressed on the legs and antennal scapes. Castaneous; pronotum, first gastric segment, borders of clypeus, and mandibles blackish; remainder of mandibles and clypeus, cheeks and anterior portion of front, petiole and postpetiole yellowish red; legs brownish yellow; terminal gastric seg- ments pale brown; posterior borders of all the gastric segments broadly yellowish. Worker.— Length 1.5 mm. Head subrectangular, as broad as long and as broad in front as behind, with very feebly convex sides and nearly straight posterior border. Eyes just in. front of the middle. Mandibles with the entire apical border finely denticulate. Clypeus convex, with rounded, entire anterior border. Antennal scapes reaching beyond the posterior corners of the head to a distance equal to twice their diameter. Thorax shaped much as in the soldier, but the pronotum narrower and longer. Epinotal spines reduced to minute slender teeth scarcely longer than broad at their bases. Superior border of petiolar node straight and entire; postpetiole small, a little broader than the petiole, subglobular. Pilosity, sculpture, and color as in the soldier, but the head smooth and shining, with only the cheeks delicately longitudinally rugulose. Described from a single soldier and three workers taken by Lang from a colany nesting in a stem of Hyphzne at Malela. This form agrees with the typical minima in size and in most of its characters but the color is very different, the postpetiole is much nar- 134 Bulletin American Museum of Natural History [Vol. XLV rower in proportion to the petiole in both soldier and worker, and the antennal scapes of the latter are decidedly longer. Santschi has de- scribed a variety, catella, from Nigeria and the Gold Coast, which is evi- dently colored like malelana but his description is too brief to enable me to judge of its other characters. He has also described a subspecies, corticicola, from the French Congo. The soldier of this form measures 3 mm., the worker 2.3mm. Both are red or yellow and in the soldier the frontal carine extend to the posterior quarter of the head. Pheidole mylognatha, new species Text Figure 33 So._pier.— Length 6 mm. Head large, subrectangular, 2 mm. broad and 2.3 mm. long, as broad in front as behind, with straight, parallel sides, deeply and angularly excised posterior border, with depressed occipital surface and faint depressions on the sides of the front for the antennal seapes. Occipital and frontal groove deep. Eyes small, flat, at the anterior third of the head. Mandibles very convex, probably bluntly bidentate at apex but the apical borders are worn away in the specimen. Clypeus — very short, concave and indistinctly carinate in the middle, swollen and convex on the sides; the anterior border rather deeply emarginate in the middle and sinuate on each side. Frontal carine short, diverging; frontal area indistinct. Antenne small and slender; scapes when bent outward not reaching to the eyes, terete and slightly curved at the base; joints 2 to 8 only slightly longer than broad; club distinctly Oe Be shorter than the remainder of the funiculus. Thorax small, uiiesadie aail-anmale much shorter than the head and less than half as wide Soldier; headfrom above. through the pronotum, which is bluntly tuberculate on the sides both above and below. Mesonotum short, rapidly sloping to the pronounced mesoépinotal constriction, anteriorly with a feeble trans- verse impression and a small, sharp transverse ridge behind it. Epinotum distinctly broader than long, broadly concave and sloping in the middle, the base shorter than the declivity, marginate on the sides, the marginations continued into the spines which are short, acute, and erect, a little longer than broad at their bases, less than half as long as their interval. Petiole small and short, less than twice as long as broad, broader behind than in front, the node blunt, transverse, and emarginate in the middle. Postpetiole broader than long, its sides produced as short, acute, backwardly directed spines, the distance between the tips of which is about three times the width of the petiole. Gaster smaller than the head, elliptical, flattened dorsoventrally. Femora only moderately thickened in the middle. Shining; mandibles sparsely punctate in the middle, coarsely striated at the base and along the apical margins. Clypeus rugulose, irregularly in the middle, longitudinally on the sides. Anterior half of head longitudinally rugose, with pune- tate interrugal spaces, the punctures becoming more numerous on the very feeble scrobe-like depressions; posterior half of head very smooth and shining, with a few 4922) Wieser, ‘Ante df the Belgtin Conge 135 _ sparse, piligerous punctures. Thorax loosely rugose and somewhat reticulate- punctate on the sides, concavity of epinotum finely transversely striated. Petiole and postpetiole indistinctly punctate-rugulose, the latter smoother and shining above. Gaster and legs smooth and shining, with sparse, piligerous punctures. _ Hairs whitish, delicate, sparse, erect or suberect on the body, shorter, "more abundant and appressed on the legs; almost absent on the scapes. ____ Rich castaneous brown; gaster, except the base of the first segment, darker, almost black; legs and funiculi a little more reddish, the femora infuseated in the Lf TiEEE « Kettle longes them broad, as broad in front as behind, with feebly convex sides and feebly concave posterior border. Eyes rather convex, just in front of the ___ middle of the sides. Mandibles with the whole apical border very finely denticulate. ____ Clypeus convex, its anterior border entire, broadly rounded. Antennal scapes extend- _ing fully one-fourth their length beyond the posterior border of the head. Thorax and petiole very similar to those of the soldier but the mesonotum more sloping and _ with much feebler transverse convexity. Postpetiole only one and one-half times as _ broad as the petiole, its sides produced as short angles or conules. Shining; mandibles finely and indistinctly striate; clypeus and cheeks longitu- dinally rugulose; area between the frontal carinz and the eyes reticulate, remainder of head very smooth and shining. Pronotum smooth and shining above, reticulate on the sides; meso- and epinotum subopaque, densely punctate; petiole and post- _ petiole more finely punctate, the nodes above smooth and shining like the gaster and Pilosity and color much as in the soldier, but the fine appressed hairs on the scapes as abundant as on the legs. Described from a single soldier and two workers taken at Banana _ by Lang and Chapin. This species is related to P. schultzei Forel from the Kalahari Desert, as I find by comparison with cotypes received from Prof. Forel. _ The head of the schultzei soldier, however, has more convex sides, more rounded posterior corners, a less deeply excised posterior margin, less deeply impressed occipital groove, longer antenne, and a very different color, being yellowish red, with the legs and base of gaster yellow. The ___ worker schultzei departs further from that of mylognatha in being more slender, with decidedly longer legs and antenne, in lacking spines on the __ epinotum and in having a longer postpetiole, which is scarcely angular onthe sides. It is sordid or brownish yellow, with the head darker behind ; and on the sides. 136 Bulletin American Museum of Natural History [Vol. XLV Pheidole niapuana, new species Text Figure 34 SOLDIER. — Length 5 to 5.5 mm. Head, excluding the mandibles, as broad as long (2.3 mm.), cordate, considerably broader behind than in front, and with the occipital border very deeply and arcu-- ately excised. Behind the eyes the sides are convex but in front feebly concave. Eyes small, moderately convex, situated just in front of the anterior third of the head. In profile the head is most convex in the middle both above and below, but depressed in the occipital region. Frontal and occipital groove distinct but rather shallow an- teriorly. Mandibles large and convex, with two blunt teeth at the apex. Clypeus flat, carinate, its anterior border emarginate in the middle, bluntly bidentate, sinuate on Fig. 34. Pheidole niapuana, new species. Soldier. a, body in profile; b, hcad from above. the sides. Frontal area large, subtriangular, without a median carinula; frontal carine short, diverging, continued back as delicate rugew bordering an indistinct scrobe-like depression for the antennal scapes. Antenne slender; scapes terete, curved at the base, reaching to the middle of the sides of the head; all the funicular joints longer than broad, club somewhat shorter than the remainder of the funiculus. Gula with a pair of very large, blunt teeth at the anterior margin. Thorax short and robust, shorter than the head without the mandibles. Pronotum with very distinct and moderately acute humeral tubercles, mesonotum sloping to a deep mesoépinotal constriction, with a sharp transverse welt or ridge; epinotum broader than long, con- cave and sloping in the middle; spines acute, somewhat shorter than the base, a little longer than their interval, directed upward and slightly outward and backward, with their tips distinctly curved backward. Petiole very small, narrow, fully twice as long as broad, with subparallel sides, the node short, with acute transverse superior border, distinctly notched in the middle. Postpetiole three times as broad as the petiole, subtriangular, broader than long and broader behind than in front, with prominent, bluntly angular sides, its ventral surface with a distinct tooth, its dorsal surface convex and rounded. Gaster broadly elliptical, smaller than the head. Legs rather slender, femora only moderately thickened in the middle. 1922) Wheeler, Ants of the Belgian Congo 137 Shining; mandibles sparsely punctate, striated at their bases. Head longitu- ee tae rege ary, widely separated and not very strong, the interrugal _ spaces with dense shallow punctures, most distinct on the space between two ruge _ representing a very feeble scrobe-like area. The ruge on the front diverge, passing to _ the summits of the occipital lobes. Sides of head with finer, denser ruge. Occipital es with large, scattered foveole. Thorax, petiole and postpetiole covered with fine _ shallow punctures, more pronounced on the mesopleure and extremely fine and dense Or the petiole and postpetiole which are opaque. Pronotum transversely rugulose. sal half of first gastric segment finely reticulate-punctate and less shining than the Sher! "Hairs reddish yellow, glistening, coarse, uneven, erect, and rather sparse on the "body; short, sparse, and appressed on the scapes and legs. Rich ferruginous red; clypeus and borders of mandibles black; legs and antenne paler and more yellowish red; gaster infuscated on the sides and behind the first - Worxer.— Length 3 to 3.5 mm. _-~—s«sHeadd nearly circular, scarcely longer than broad, without posterior corners, i a Gecipital border strongly marginate. Mandibles large, their apical borders long and F finely denticulate, with two larger terminal teeth. Clypeus convex, with rounded, entire anterior border. Eyes just in front of the middle of the head, moderately large and convex. Antenne slender, scapes extending about two-fifths their length beyond the occipital border. Thorax slender, the pronotum rather depressed above, bluntly 2 tuberculate on the sides near the middle. Mesonotum long and sloping, with a broad transverse impression in front and a transverse swelling behind it. Mesoépinotal constriction deep and broad. Epinotum as broad as long, with subequal base and declivity, not concave in the middle as in the soldier. Spines longer, as long as the base and more strongly curved backward. Petiole similar to that of the soldier, but with a lower, blunter node. Postpetiole scarcely twice as broad as the petiole, longer than broad, rounded above and on the sides. Gaster distinctly smaller than the head. Legs slender. Shining; finely reticulate; mandibles finely and densely striate, lustrous; gaster more shining than the head and thorax; meso- and epinotum and ventral and lateral portions of the petiole and postpetiole subopaque, densely punctate. Pilosity much like that of the soldier, sparser on the body but more abundant on the legs. Color much paler, of a more yellowish red, or reddish yellow, with paler legs and brown gaster, the latter in most specimens yellowish at the base. Described from numerous specimens of both phases taken by Lang and Chapin at Niapu “from nests in the rotten wood of fallen trees or in old roots.” _.__._ This species is evidently related to P. areniphila Forel of the Kala- hari Desert but is certainly distinct, being larger and differing in many details of structure and sculpture. 138 Bulletin American Museum of Natural History (Vol. XLV Pheidole saxicola, new species Plate VII; Text Figure 35 SoLpIeR.— Length 5.5 to 6 mm. Head subrectangular, nearly 3 mm. long and very nearly as broad, scarcely broader behind than in front, with straight subparallel sides, rectangular anterior corners, deeply and angularly excised posterior border, and deep occipital and frontal groove. In profile the occipital region is very feebly depressed and the eyes are small, feebly convex, and at the anterior third of the sides. Gula anteriorly with prominent, blunt teeth. Mandibles convex, with two large apical and two basal teeth and a few denticles along the intermediate border. Clypeus convex and carinate in the middle, its anterior border broadly and feebly excised in the middle and sinuate on each side. Frontal carine very short, diverging; frontal area distinct, with a median carinula. Antenne slender, scapes reaching the middle of the head; funicular joints all longer Fig. 35. Pheidole saxicola, new species. Soldier. a, body in profile; b, head from above. than broad; club shorter than the remainder of the funiculus. Thorax shorter than the head, robust, through the pronotum nearly half as broad as the head, with very blunt humeri, convex and rounded in profile. Mesonotum sloping to the deep meso- épinotal constriction with merely a trace of a transverse convexity in the middle. Epinotum broader than long, concave and sloping in the middle, in profile with the base distinctly shorter than the declivity; spines short, suberect, acute, less than half as long as the base and about half as long as their interval. Petiole about one and one-half times as long as broad, broader behind than in front, with concave sides; node transverse, its superior border sharp, feebly excised in the middle. Postpetiole broader than long, about two and one-half times as broad as the petiole, its sides produced as short, acute, slightly backwardly directed spines, its ventral surface with a small, acute tooth. Gaster smaller than the head, subcircular or very broadly ellip- tical, somewhat flattened above. Legs with moderately thickened femora. a eee Si tm ig i ; 1922} Wheeler, Ants of the Belgian Congo 139 Shining throughout; mandibles coarsely striate, smooth and coarsely punctate _ in the middle: Clypeus longitudinally rugulose, less distinctly in the middle than on the sides. Head rather finely and sharply longitudinally rugose, the ruge diverging on the front and continued to the posterior corners, where they meet the also slightly divergent ruge between the frontal carine and the eyes. The interrugal spaces are Joosely reticulate. There are no transverse ruge on the occiput but only a finer con- +3 tinuation of the more anterior sculpture. Thorax, petiole, and postpetiole indistinctly _____ and loosely punctate rugulose, the prothorax transversely; epinotum with fine, dense _ but shallow punctures, so that the surface is more opaque. Gaster with fine, sparse, punctures. ___ Hairs yellowish, partly coarse, sparse, uneven and suberect and partly short, _ much more abundant, softer and appressed or subappressed like long, coarse _ pubescence. Legs with ‘numerous short, oblique hairs; scapes with a few longer Dark ferruginous red; inamstithcs, sides and border of clypeus, and frontal carine, blackish; petiole, postpetiole, and gaster, except more or less of the base of the first segment, dark brown or blackish. Legs a little paler than the thorax. Worker.— Length 2.7 to 3 mm. Head subrectangular, as broad in front as behind, with straight, subparallel sides, rounded posterior corners and nearly straight posterior border. Eyes convex, at the middle of the sides. Mandibles rather large, deflected at the tip, with denticulate apical borders and two larger terminal teeth. Clypeus distinctly carinate, with the anterior border very feebly sinuate in the middle. Antennal scapes extending one- third their length beyond the posterior corners of the head. Thorax similar to that of the soldier, but more slender, especially through the pronotum. Base of epinotum a little longer than the declivity; spines slender, acute, erect, about half as long as their interval. Petiole slender, twice as long as broad, scarcely broader behind than in front, with the sides only very faintly concave; node transverse, its border distinctly notched in the middle. Postpetiole twice as broad as the petiole, as long as broad, sub- globose, not toothed on the ventral side. Gaster about as large as the head. Shining; mandibles subopaque, finely striatopunctate. Sides of head delicately longitudinally rugulose and reticulate. Thorax, petiole, and postpetiole finely and densely punctate, opaque; upper surface of pronotum and postpetiole smooth and ~shining. Gaster and legs shining, sparsely punctate. Pilosity like that of the soldier but less abundant. Antennal scapes, like the legs, 4 i. with numerous oblique hairs. Brown; head darker above and behind; gaster, except the edges of the segments, middle portions of legs, fore coxw, and usually also the pronotum and upper surfaces of the petiolar nodes, darker than the posterior portion of the thorax. Described from numerous specimens taken by Lang, Chapin, and ____ Bequaert at Zambi (type locality) and by the latter at Boma. This ant is certainly very closely related to P. sculpturata Mayr 4 and might be regarded as a subspecies, but it will fit neither Mayr’s _ deseription of the typical form from South Africa nor Santschi’s and ___ Forel’s descriptions of the various subspecies from East and West Africa. Mr. Lang’s note shows that it isa harvester. ‘The nests were found on a 140 Bulletin American Museum of Natural History [Vol. XLV dry hill at the Post of Zambi in rocky soil. One of the entrances, the largest of three, can be distinctly seen in the photograph (Plate VII). The ants excavate their nests in the small amount of soil between the rocks and all or nearly all of them remain under ground during the day. They work during the night up to about 8 A.M. Then the workers may be seen moving along in files, accompanied by the soldiers, and the latter — carry seeds for a distance of some fifteen yards. They come and go in different directions indicated by runways left between the accumulated masses of débris and distinctly visible in the photograph. The débris, consisting of seeds and chaff, lies about the nest to a depth of four centi- meters and over an area of some sixty: centimeters. It is very difficult to obtain a view of the interior of the nest on account of the rocky soil. Some of the kitchen-middens about the nest entrances contained the dried remains of various ants and Coleoptera. In another locality the same species of ant was seen to have collected seeds of entirely different plants but of about the same size.” Fig. 36. Pheidole speculifera Emery. Soldier. a, body in profile; b, head from above. Pheidole speculifera Emery Four soldiers from Faradje, without further data, and five workers from the stomach of a frog (Rana ornatissima) from Garamba agree very closely with Emery’s description of the types from Abyssinia, but the workers are darker. Forel has described a variety, cubangensis, from Mossamedes and records it also from the Belgian Congo, but this form seems to be very close to the type. My specimens are not as large, since none of the soldiers measures more than 6 mm., whereas Forel 1922) Wheeler, Ants of the Belgian Congo 141 i gives the length of cubangensis as7mm. He describes the whole head as - Opaque, whereas my specimens have a pair of elliptical, very smooth, and _ shining areas on the vertex in the midst of the opaque and finely pune- tate sculpture (Fig. 36a and 6). -_ Myemicaria W. Saunders 4 a "Small or medium-sized, coarsely hairy, brown or black ants, with monomorphic _ workers, which have 7-jointed antennz, the funiculus enlarged toward the tip but 7 -. not clavate and all the joints, except the first, considerably longer than wide. _ Mandibles moderately large, subtriangular, with coarsely dentate apical border. _ Clypeus broad and convex. Frontal area indistinct behind. Frontal carine short, rather far apart, not strongly diverging posteriorly. Eyes not very large, convex, ae behind the middle of the head; ocelli absent. Thorax with indistinct or obsolete suture; mesoépinotal suture deep, the mesoépinotal constriction pro- ; the sides of the mesonotum raised and subauriculate behind. Epinotum armed with a pair of long, acute spines, which are often lobate or expanded at the base; inferior corners of pronotum dentate or spined. Petiole with a long peduncle _ sharply marked off from the abrupt node, which is high and rounded, subconical, _ sometimes laterally compressed. Postpetiole shaped like the node of the petiole, strongly contracted posteriorly. Gaster subglobose, its basal segment somewhat truncate in front. Legs long; median and hind tibie with simple spurs; tarsal claws '-—s« Fema considerably larger than the worker. Head and antennz of very similar _ structure, the latter being 7-jointed. Thorax robust; mesonotum and scutellum very - convex, the pronotum vertical in front though well developed, the epinotum with _ stouter and broader spines than in the worker. Pedicel as in the worker. Gaster much more voluminous, longer than wide, convex above; the basal segment truncate anteriorly. Wings long, with strongly marked veins, the anterior pair with an open radial cell, a single cubital and a discoidal cell. MA Lz nearly as large as the female but more slender. Antenne 13-jointed, fili- form, the scape short, about as long as the second funicular joint, the first joint very short, not swollen, the remaining joints all much longer than broad. Eyes large but not very convex; ocelli rather small. Mandibles small and vestigial, sublinear, with rounded edentate tips, which do not meet. Frontal carine short. Mesonotum with Mayrian furrows; epinotum without spines. Petiole very long, its node low; that of _ the postpetiole of a similar shape, decidedly longer than broad. Gaster cordate, searcely longer than broad, convex above, concave below. External genital append- ages long and narrow, blade-like. Cerci present, but minute. Legs slender. Wings rather short, venation as in the female. This extraordinary genus may be pioniiniiall at once by the 7- jointed antennz of the worker and female and the unique structure of the abdomen in the male. The species are distributed over the Ethiopian, Indomalayan, and Papuan Regions but do not enter Australia (Map 21). The majority of the species and the largest are Ethiopian. The large species form crater nests in the soil; some of the smaller, both in Africa and in the Orient, make small carton nests on the under sides of leaves. 142 Bulletin American Museum of Natural History [Vol. XLV One of Mr. Lang’s photographs (Pl. VIII, fig. 1) of crater nests of — M. eumenoides is very suggestive in connection with some observations of Petch' on the Indian and Ceylonese M. brunnea Saunders. This ant, he says, “brings up from its nest underground grains of sand and particles of earth through a small hole about a centimeter in diameter; it is gen- erally observed on footpaths. These particles are at first arranged on one side of the hole in a crescentic mound about 3 centimeters high which curves round and slopes away to nothing on either side of the hole, the — distance between the vanishing horns on the crescent being about 12 F f I. THe a Ae Se ) ae ba 0 S CTR ) G. is Y) 1S » 4s oo 75 90 105 120 135 160 Map 21. Distribution of the genus Myrmicaria. centimeters. The ants run up the slope from the hole with their burden and drop it over the ridge down the steeper outer side. The most striking feature of this is that when the hole is situated in the middle of a path, away from any bank, the ridge is always on the windward side of the hole. A smaller ridge of the same shape and in the same position is constructed by Pheidole (? nietneri Emery). If undisturbed Myrmicaria eventually constructs a complete funnel around the hole.”’ It would seem that the craters of M. eumenoides photographed by Mr. Lang were con- structed in a spot protected from the wind or during a calm since they show no definite orientation of their steeper slopes. 1906, Ann. Roy. Botan. Gard. Peradenyia, III, p. 196. 1922} Wheeler, Ants of the Belgian Congo 143 3 ‘Myrmicaris-eunssnoides (Gerstecker) subspecies opaciventris (Emery) oe 4 Plate VIII, Figures 1 and 2 = Malela, 8; Thysville, 8; Stanleyville, 9, @; Avakubi, 8, 9°; a Medje, 8, 2, @; Akenge, 8; Bafwabaka, 8; Ngayu, 8; Faradje, 8, . 2 (Lang and Chapin) ; Walikale to Lubutu, 8, 9 (J. Bequaert) ; -Yakuluku, ¢° (J. Rodhain). Seventy-five workers and one female from _ Bafwabaka, Ngayu, Medje, Akenge, and Stanleyville were taken from 4 the stomachs of toads (Bufo regularis, B. funereus, and B. superciliaris) ; a single worker from Faradje was taken from the stomach of a frog (Rana occipitalis). - Neither Forel nor Santschi seems to me to have recognized this form 4 2s very explicitly. Several years ago I received from the former six work- ____ ers labelled “Benguela (Buchner)”’ and, as Emery’s ergatotypes bore _ the same label and were also received from Forel and as my specimens agree perfectly with Emery’s description, I feel confident that they are cotypes. Later I received a worker and three deiilated females from -Gaboon (Staudinger) and, as Emery mentions specimens from the same locality , I believe that I have before me also the female of the true : opaciventris. The workers measure about 5 to 6 mm. and are pale ferru- i brown, with the antennz, legs, and gaster more fuscous. The mandibles have oblique 5-toothed blades; the clypeus is carinate. The epinotal spines are rather slender and very slightly bent downward, the base of the epinotum is less concave than in the typical eumenoides, the peduncle of the petiole is distinctly shorter and not longer than the node. The petiolar and postpetiolar nodes are laterally compressed and of the same height, the ventral surface of the postpetiole, unlike that of eumenoides, is swollen, and projecting and angular in front. The surface of the head and thorax is somewhat less shining than in eumenoides, the ruge on the front, pleurz, pro-, meso- and base of epinotum more sharply and regularly longitudinal and not reticulate. The gaster has the basal half or, in some specimens, the whole surface opaque and densely __ punctate, whereas it is smooth and shining in typical eumenoides. The nodes of the petiole and postpetiole have shining summits and in some __ specimens the sides of the petiole are also smooth and shining, in others x like those of the postpetiole, finely punctate and even feebly longitudi- nally rugulose. In the female, which measures 13 mm., the petiole and postpetiole are sharply longitudinally rugose, the summit of the former concentrically rugose, the scutellum vermiculately rugose. Emery’s description of the male, which I have not seen, includes no mention of characters that would distinguish it from the male of the typical eumenoides. 144 Bulletin American Museum of Natural History [Vol. XLV Numerous specimens from the various Congo localities cited above seem to me to be referable to Emery’s subspecies, though they differ more or less in the sculpture of the petiole, postpetiole, and gaster and in being mostly of a darker color. They average larger than the speci- mens of variety congolensis and variety crucheti, the workers being 5 to 6.5 mm. The petiole and postpetiole, especially the latter, are nearly always more or less longitudinally rugulose on the sides, though some- times merely punctate, as Emery remarks in the original description. The specimens from Walikale have the entire gaster opaque and punc- tate, whereas in others it is punctate usually only on the anterior half of the first segment. This character, however, varies in individuals from the same colony. Santschi says that the gaster of the worker is “entiére- ment sculpté, mat, brun clair,’ but Emery describes the gaster as fuscescent, with the anterior half of the first segment opaque. Trigirdh' and Arnold* have described the nests of the typical eumenoides of East and South Africa. The latter’s account runs as fol- lows. The colonies of this species are usually very large, often comprising 1000 or more workers. The latter bite and sting fiercely, but the sting is rather blunt, and does not easily pierce the human skin. Although their gait is slow, they are nevertheless active insects, travelling over large areas in search of food, which seems to consist chiefly of other insects. They do not appear to be aphidicolous, nor to attend membracid or lepidopterous larve for their secretions, yet they are known to harbour in their nests many myrmecophilous insects. A nest examined by me contained the following species of beetles: Allodinarda myrmicariz Brauns; Ogmocerus raffrayanus Brauns and Batrisus myrmecariophilus Brauns. The Botanical Gardens in Durban are infested with this species, but the exam‘nation of a large number of nests revealed only one species of myrmecophile, Allodinarda kohli Wasm.; which, however, was plentiful, as many as three dozen being taken in one nest. The nest has numerous entrances, and is surrounded by large heaps of excavated material, often covering an area of several square feet. Arnold’ has also described and figured the puparium of a fly (pos- sibly a form allied to Microdon?), with a peculiar tray covered with trichomes at the posterior end of the body, as occurring in the nest of M. eumenoides with the myrmecophilous beetles cited in the foregoing quotation. The following is his account of the migration of the colony and its guests to a new nest. I left this nest without filling up the hole, so that in about a week’s time it was filled with rain after a heavy shower. The water must have filtered through the soil and almost saturated the nest, for it took nearly half an hour for all the water to dis- 11914, Med. Géteborgs Mus. Zool. Afd., ITT, p. 45. 21916, Ann. South African Mus., XIV, 'p. 71914, Proc. Rhodesia Sc. Assoc., XIII, p. 25. 1922] | Wheeler, Ants of the Belgian Congo 145 "appear from the hole, ‘This state of affairs had evidently made the nest so uncom- | fortable that the ants decided to move to new quarters about 9 feet away. They be- gan to do this about seven o'clock that evening, or perhaps a little earlier, for the ‘migration was in full swing when I came on the scene again at that hour. Remember- E sing the reputation which this ant has for harboring guests, and also the observations _ made by various entomologists on some European ants which, when moving to a new ah oy are in the habit of carrying their guests with them, I decided to watch this _ migration carefully. At first I could see no guests at all; the workers were carrying ; in their mandibles only their own larve, pupe or males. In fact I was looking at the _ workers so attentively that I failed to notice their smaller companions on the road, to | _ which my attention was directed by suddenly catching sight of a Lepismid running _ by. Going back then to the old nest, I saw at intervals various myrmecophiles crawl- _ ing out of the pit made by my former excavation, and following the tracks of their hosts, to which they were guided, of course, by the sense of smell. These parasites included three different species of beetles, viz. a staphylinid, and two species of psela- phids, together with the common lepismid found in the nests of nearly all our ants. _No time was wasted by any of these insects, for once over the brow of the pit, they continued straight along the narrow path leading to the new quarters. While on th» march they were utterly ignored by their hosts, but on arriving at the entrance of the _ new nest, it was noticed that some of the pselaphids were seized by the ants dawdling around, and taken down into the nest. This change of dwelling took some hours to complete, for at midnight it was still in progress. = Mr. Lang contributes the following note on the habits of the sub- _ species opaciventris at Avakubi: “These ants, called ‘dufluguntu’ by the natives, are very common and noticeable because they tend to con- gregate in great numbers about any piece of meat or a dead insect. On one occasion I saw them tear up and carry off a butterfly two inches in diameter in exactly two minutes and a half. They are harmless and therefore not feared by the natives. A young Manis, which I kept in captivity, enjoyed making a meal of them. The nests, as a rule built at the bases of trees or bushes, can be easily recognized by the mound of loose earth thrown up while the chambers are being excavated. The walls of the chambers are not hardened or smoothed as in the nests of some other ants. One nest which I examined extended seventeen inches below the surface. It had many ramifications, though most of the brood was found around the roots of the tree. The whole nest, when exposed, covered an area less than two feet in diameter. These ants build long tunnels open above or with small openings (one-eighth inch), surrounded by a heap of loose particles. One of these, more than an inch wide, crossed a certain road in several places. I have seen a number of these tunnels superimposed one above another so that I could drop a stick down thirteen inches. In these tunnels the ants travel back and forth in great numbers.” 146 Bulletin American Museum of Natural History (Vol. XLV Myrmicaria eumenoides subspecies opaciventris variety congolensis (Forel) This form is not represented among the material collected by Lang, Chapin, and Bequaert. Santschi regards it as an independent subspecies, but it seems to me to be merely a variety of opaciventris. Three cotypes of congolensis were given me by Forel. Comparison of these specimens, which were taken from the stomach of a scaly ant-eater (Manis tem- mincki) captured by Solon in the Lower Congo, with opaciventris show relatively slight differences. They are somewhat smaller, of a more sordid yellowish brown color (possibly due to the action of the gastric juices of the Manis), and with much the same sculpture and lower portion of the postpetiole. The epinotal spines, however, are decidedly more slender and more strongly deflected, a character not mentioned in Forel’s original description, though noted by Santschi; the head is proportionally smaller and narrower, with straight cheeks, and the gaster is opaque only at the base of the first segment, the remainder being rather shining. Myrmicaria eumenoides subspecies opaciventris variety crucheti (Santschi) Stanleyville, 8; Leopoldville, 8; Ngayu, 8; Avakubi, 8 (Lang and Chapin). The workers from Avakubi, 22 in number, were taken from the stomachs of toads (Bufo regularis and B. funereus). I refer numerous specimens from these localities to the variety crucheti since they agree with Santschi’s very brief description in size (5 to 5.5 mm.) and in having slender but straight epinotal spines. The petiolar node in my specimens is distinctly broader and less compressed laterally than in the typical ewmenoides and not shorter than the peduncle. The surface of the petiole is not so smooth, though it is not longitudinally rugulose. I have received this same form in all three phases from Rev. Geo. Schwab, who took it at Metit, Cameroon. The female is very similar to that of the typical eumenoides, but the head is somewhat smaller, with slightly — more prominent posterior corners and the gaster is entirely opaque and punctate, except the bases of the second and following segments. I am unable to detect any differences between the males of the two forms. Arnold describes the wings of the male ewmenoides as paler than those of the female. This is certainly not the case in crischeti. Wheeler, Ants of the Belgian Congo 147 =" Myrmicaria salambo, new species > ite Plate LX, Figures 1 and 2; Text Figure 37 _ WorKeR.— i a Length 6 to 7 mm. Ms Of rather uniform stature and closely resembling eumenoides but a little more é Head relatively smaller, as broad as long, excavated behind, convex above, flattened below. Mandibles 5-toothed. Clypeus ecarinate, with entire anterior border. Eyes somewhat larger and more convex than in ewmenoides. Thorax very _ similar but promesonotal suture very distinct, impressed, the mesonotal lobes less _ compressed, their posterior outline in profile less abrupt, more sloping so that the -mesoépinotal impression, though deep, i is shallower and less acute than in euwmenoides ‘and appears longer. Epinotal spines longer, slightly sinuous, with very feebly up- turned points, directed backward and slightly outward. Base of epinotum longitudi- Fig. 37. Myrmicaria salambo, new species. Worker in profile. nally concave. Peduncle of the petiole longer than the node, which is thick and evenly rounded, not compressed laterally above. The ventral surface of the petiole armed below with two long, delicate hyaline spines, which curve towards each other and enclose an elliptical space. Postpetiolar node of the same size and shape as that of the petiole, its ventral surface straight in profile, not bulging nor angulate in front. _ Gaster and legs of the usual shape, the former with a straight, anterior border. Shining; mandibles coarsely longitudinally striated; clypeus smooth in the _ middle, with a few rugules on the sides. Rugosity of head, thorax, and pedicel much a8 in ewmenoides, but the ruge on the dorsal surface of the head and thorax less numer- ous and less pronounced, without distinct anastomoses; sides of the head with finer __ and less distinct rugules, so that the surface ismore shining. Gaster opaque and very finely punctate only at the extreme base above, otherwise shining. Legs and scapes shining, finely striate. Hairs dark brown, in length and arrangement much like those of eumenoides. Reddish brown; gaster brownish yellow; legs, including the coxe and lower pleurm, darker than the thorax. Mandibular teeth and antenne blackish. Described from numerous specimens taken at Garamba (Lang and Chapin) attending scale insects on the buds of a Protea which is shown in Plate IX. 148 Bulletin American Museum of Natural History (Vol. XLV This form is so closely related to eumenoides that it might, perhaps, be regarded as a subspecies. It is easily recognized by the unique ventral appendages of the petiole. These are so brittle that they are easily broken off, but their basal insertions on the low hyaline lamella in the midventral liné of the petiole are usually discernible. Evidently salambo is also related to M. striata Stitz, specimens of which I have not seen. Myrmicaria exigua Ern. André subspecies kisangani, new subspecies Worker.— Length 3 to 3.5 mm. Head through the eyes scarcely longer than broad, evenly rounded behind. Mandibles 4-toothed. Clypeus ecarinate, convex, with entire, rounded anterior border. Frontal carine subparallel. Eyes convex, just behind the middle of the head. Antennal scapes extending about two-fifths their length beyond the posterior border of the head; apical funicular joint fusiform, enlarged as in the typical erigua. Pronotum more flattened above, though bluntly angular on the sides and without inferior teeth. Promesonotal suture distinct. Mesonotum with a small but distinet tooth on each side in front and the posterior lobes larger, erect, and rather acute. Mesoépinotal impression very distinct and rather long. Epinotum not longer than broad, scarcely narrowed in front, its base longitudinally grooved in the middle, marginate on each side and not longer than the declivity, which is also marginate laterally; spines not longer than their distance apart at the base, straight, directed backward, upward, and outward, their tips not bent inward as in the typical exigua. Petiolar peduncle as long as the node, swollen at the spiracles; node longer than broad, as high as long, laterally compressed, constricted behind. Postpetiole longer than broad, broader and higher behind than in front, its node distinctly lower than that of the petiole. Anterior border of gaster straight or even slightly concave, with prominent anterior corners. Shining; mandibles subopaque, longitudinally striate. Clypeus smooth in ‘the middle, delicately rugulose on the sides. Head smooth in the middle of the front, delicately and irregularly longitudinally rugulose on the sides, posteriorly reticulate- rugose, but much less sharply than in the typical erigua. Pronotum with a few longi- tudinal rug, sometimes absent in the middle line; in some specimens reticulately- rugose over the whole surface, with very large meshes as in erigua. Sides of pronotum smooth and shining; meso- and metapleure subopaque, longitudinally rugulose. Base of epinotum transversely rugulose, declivity smooth and shining. Pedicel, gaster, and legs smooth and shining, with very sparse and minute, piligerous punctures. Pilosity like that of the typical exigua, gray or whitish. Piceous, nearly black; tips of mandibles, peduncle of petiole, declivity of epinotum, base of postpetiole and in some specimens the whole gaster or only the base of the first segment brown. Described from numerous specimens taken at Stanleyville (Lang and Chapin) “crawling about the base of an orange tree.” I have compared this form with two cotypes from Sierra Leone (Mocquerys), received many years ago from André, and a worker from a sel gla amend 149 a" The new subspecies differs in its much darker color, feebler sculpture, laterally more compressed petiolar node and in the shape of the mesonotum, which in the typical form of the, species lacks 1 the anterior tooth on each side and has only feeble indications of the lobes. Forel has described a variety, rufiventris, from carton BR nests 3 to 4 em. in diameter on leaves at St. Gabriel, Lumaliza, and Bati- _ amponde (Kohl), all localities near Stanleyville. This form is larger a (3.8 to 4.6 mm. ) and, according to Forel, “differs from the type of André . only in its paler, reddish abdomen and in having the head more elongate a and narrower behind.”’ What Stitz has described as a distinct species, gracilis, is evidently nothing more than a subspecies of exigua, as is _ shown by a comparison of his and Forel’s descriptions with the cotypes. __ André failed to mention the enlarged apical antennal joint, but it is very se SS in his specimens. Stitz says of the petiole: ‘“Hinten eo sich von seiner Basis ein kleines, sekundiires Knétchen ab.” _ This seems to refer to the swelling of the peduncle at the spiracles, a ___ swelling which is visible, though less accentuated in other species of the genus, when the peduncle is viewed directly from above. Forel, however, a interprets Stitz’s ‘secondary node”’ to mean the constricted portion of 4 the segment behind the node. As neither Stitz nor Forel compared their q specimens with André’s cotypes, they were led to regard gracilis as a CarpioconpyLa Emery Worker minute, smooth, almost hairless. Clypeus projecting over the bases of the mandibles, steep in front, with rounded anterior border. Frontal area strongly g impressed. Frontal carine short and straight. Eyes well developed; ocelli lacking. _ Mandibles broad, triangular, dentate. Antenne 12-jointed, with long first funicular _ joint and 3-jointed club, the last joint very large. Promesonotal suture indistinct; 4 ‘mesoépinotal constriction well developed. Epinotum armed with spines or teeth. Petiole with long peduncle and small, rounded node. Postpetiole conspicuously _ large, cordate or transversely elliptical. Gaster formed in large part by the first ‘segment. _. Femae winged (except in C. emeryi Forel), somewhat larger than the worker; 4 head of the same shape but with ocelli. Pronotum not covered by the mesoscutum 3 in front. Petiole and postpetiole usually broader than in the worker. Wings with a venation; pterostigma near the middle of the costal border; one closed y " eubital cell; distal portions of radius and cubitus obsolete; brachius not developed beyond the nervulus but bending up into the submedius. According to Emery, the _ female of C. emeryi is wingless and has the posterior ocelli vestigial. ~Mate usually ergatomorphic but winged in C. emeryi. In this form the ‘ ~ antenne are 13-jointed but in ergatomorphic males they are 10- to 12-jointed; with q lena and more indistinct club. Petiole and postpetiole resembling the corre- 5 sponding segments of the female, in the male of emeryi much as in the worker. 150 Bulletin American Museum of Natural History (Vol. XLV Cardiocondyla emeryi Forel A single worker taken at Thysville by Bequaert. This minute ant is very-widely distributed through the tropics of both hemispheres. It was originally described from the island of St. Thomas in the West Indies, but was later recorded from Syria, Madeira, Madagascar, and the East Indies. Arnold records it from South Africa and my collection contains specimens from the Bahamas, Cuba, Porto Rico, Jamaica, Bermuda, Tepic in Western Mexico, and Miami, Florida. According to Arnold it ‘is usually found nesting in grassy soil; the entrance to the nest is a minute hole, not surrounded by earth or other substances.” CREMATOGASTER Lund Crematogaster is one of the largest and most sharply defined genera in the family Formicide. The species are all small, with monomorphic worker, decidedly larger female, and the male usually as small as the worker. The worker and female have 10- or 11-jointed antenne, those of the male are usually 12-jointed. All the phases can be readily recognized by the peculiar structure and articulation of the petiole and postpetiole. The former does not bear a node but is more or less flattened above, the latter is short and articulated to the anterodorsal surface of the gaster, instead of to its anterior end as in other ants. The gaster, moreover, is in the worker and male sub- triangular or subcordate, with pointed tip, and its upper surface is concave or more or less flattened, its ventral surface more convex and protuberant. These peculiarities in the structure of the abdomen enable the workers of many species to turn the gaster forward over the thorax and head, so that they are sometimes called “acrobat ants.” As a rule, the sting is feebly developed. The anterior wings of the male and female have a discoidal and a single closed cubital cell. The species of Crematogaster all form populous colonies which nest in the ground, under stones, in logs, the cavities of living plants, or in peculiar carton nests attached to the branches or trunks of trees. This habit of making carton nests is best seen in the tropical species, but traces of it survive even in the species inhabiting temperate regions, such as the North American C. lineolata (Say). Many of the species have rank and disagreeable odors. The genus is cosmopolitan (Map 22), though the species scarcely enter the colder portions of the north and south temperate zones. Our common C. lineolata (Say) of North America occurs, however, as far north as Nova Scotia. The vast majority of species are confined to the tropics, being particularly numerous in the Neotropical and Ethiopian Regions. The African forms are so numerous and so variable that they constitute a veritable welter of subspecies and varieties. Mayr, Forel, Arnold, and Santschi have all dispaired of reducing this chaos to order. Unfortunately the portion of Arnold’s work dealing with the South Wheeler Ants of the Belgian Congo 151 ___ African species has been postponed by the war. He has, however, kindly written me concerning certain necessary changes in the synonymy of several of the species and I have adopted his interpretations in the list of Ethiopian species (Part VIII). Dr. Santschi, who has given more attention to the African species of Crematogaster than any previous author, has generously examined and identified a series of all the Congo ___ forms collected by Lang, Chapin, and Bequaert and has written the descriptions of several new forms. In the meantime he has published a a i e # min 0 \ ~ NE BL aL PL | ¥ . ace sas i] bi Hh oy a ad TAL sa BagN a \ / ui . XN i an — ot | Map 22. Distribution of the genus Crematogaster. revision of the subgenera of Crematogaster.! Forel was the first to begin the splitting of the genus, but Santschi has added several new subgenera. A translation of his table has been included in the key to the genera and subgenera of Myrmicinew. Santschi has arranged these various subgenera according to their natural affinities in the following sequence: 1. Decacrema 5. Sphzrocrema 9. Xiphocrema 2. Orthocrema 6. Crematogaster, sensu stricto 10. Physocrema 3. Eucrema 7. Atopogyne 11. Oxygyne 4. Neocrema 8. Paracrema 12. Nematocrema Of these, at least seven, Decacrema, Orthocrema, Sphzrocrema, Crematogaster, Atopogyne, Oxygyne, and Nematocrema occur in the Ethiopian Region. In the Congo material before me only Sphzxrocrema, Crematogaster, Atopogyne, and Nematocrema are represented. 11918, Bull. Soc, Ent. France, pp. 183-154. 152 Bulletin American Museum of Natural History - [Vol. XLV Crematogaster brunneipennis (Ern. André) subspecies acacie (Forel) variety victoriosa (Santschi) Numerous workers from Zambi (Bequaert), “nesting in a tree trunk.” The typical C. acaciz was originally taken by Keller in Somali- land in the swollen spines of acacias. Concerning one of the other varieties (generosa Santschi), Santschi writes me as follows: ‘I received from Mr. G. Arnold of the Rhodesian Museum under the name of C. brunneipennis Ern. André variety omniparens Forel some workers which differ only in their deeper color from what I have called acaciz variety generosa. The female of the latter form is very close to that of brunnei- pennis Ern. André, but the wings are even darker. I believe that brun- neipennis should be regarded as a subspecies of C. acaciz.” That Santschi is correct in regarding both forms as cospecific is proved by a comparison of two cotype workers of brunneipennis from Sierra Leone (Mocquerys), sent me by André many years ago, with a cotype of acaciz received from Forel. André’s workers are smaller, with longer antennal scapes, smoother and more polished thorax, with somewhat more circular and less cordate petiole, smaller and more slender and more pointed epinotal spines, and darker gaster and head, but the resemblances are so close in other respects that I cannot regard the differences as more than subspecific. As brunneipennis has priority of publication, acaci# must be reduced in rank and not brunneipennis, as Santschi supposes. Whether omniparens is to be retained as a distinct subspecies or is to be attached as a variety to acaciz, I am unable to determine. Comair castanea I’. Smith subspecies inversa (Forel) variety analis (Santschi) Bafwasende to Avakubi, 8 (Lang and Chapin); Thysville, 8 (J. Bequaert). The specimens from the former locality were collected on the road, without further data; those from Thysville were found “nesting in dry, dead wood, on the soil in the rocky savannah.” This and the following are merely color varieties of an extremely variable and widely distributed African and Malagasy species formerly known as C. tricolor Gerstecker. Crematogaster castanea subspecies inversa variety flaviventris (Santschi) Many workers from Garamba (Lang and Chapin), without further data. Both this and the variety analis were originally described from the Belgian Congo. The variety flaviventris has also been taken in acon: (C. Alluaud). Wheeler, Ants of the Belgian Congo 153 pate - Orematogaster excisa (Mayr) - Zambi and Thysville, ¢ (J. Bequaert); near Lie, 8; Faradje, 8 oe “lke and Chapin). The specimens from Thysville were taken “from a nest in a tree-trunk in the rocky savannah; ” those from Faradje “in a hollow tree.” The single specimen from near Lie was taken from the a cae of a toad (Bufo regularis). Crematogaster excisa subspecies andrei (Forel) Numerous workers from the Oso River and Sitaweza (between - Walikale and Lubutu) (J. Bequaert). Dr. Bequaert took this subspecies at the former locality in the hollow stalks of a myrmecophytic creeper _ (Unearia africana variety myrmecophyta) growing along the shore of the Oso River between Walikale and Lubutu (Part IV), in the latter locality in the hollow stalks of another myrmecophyte (Cuviera angolensis) in the Rain Forest (Part IV). 3 The following new variety of the subspecies impressa, though not from the Belgian Congo, was described by Santschi in connection with 4 the forms of excisa which I sent him. | Crematogaster excisa subspecies impressa (Emery) variety aglea . Santschi, new variety “Workxer.—Black; mandibles, funiculi and tarsi reddish brown. Dorsum of pronotum very densely punctate as in the typical impressa (Emery), the longi- _ tudinal ruge being feebly or not at all indicated. Head and thorax narrower. Pro- mesonotal impression feebler as in euphrosyne, with a small carina on the front of the mesonotum, which is sharply marginate, less concave than in andrei (Forel) and more so than in impressa (Emery). Basal surface of the epinotum scarcely broader than the petiole. Spines almost as long as the interval between their bases. Anterior angles of petiole truncated as in andrei. Otherwise like impressa (Emery). “Dimbroko, Ivory Coast (Le Moult). a “In impressa the funiculi are brownish black and in andrei the mesonotal carina is lacking.’ (Santschi) 3 - Crematogaster excisa subspecies impressa variety euphrosyne Santschi, a. new variety a “Worxer.—Length 3.5 mm. More or less pale chestnut brown. Thorax nar- row. Pronotum reticulate-punctate in the spaces between the fine longitudinal ruge. _ Mesonotum feebly carinate in front. Resembles the variety brazzai Santschi,' but the latter has a broader thorax, without carina and the sculpture of the thorax is merely reticulate.” (Santschi) q described as a subspecies of C. impressa and given in our catalogue (Part VIIT) as C. 2 ea oe Gasdhnstable vortety tracer. 154 Bulletin American Museum of Natural History [Vol. XLV Faradje (type locality) and Thysville (Lang and Chapin). The specimens at Faradje were found “nesting in hollow twigs. Snails (Pachnodus herbigradus Pilsbry) were found estivating in the same twigs inhabited by the ants and often in such numbers as to clog the passages.” Camponotus foraminosus was found in similar hollow branches together with the same snails (see p. 248). Crematogaster excisa subspecies impressa variety sapora (Forel) Numerous workers from Yakuluku (Lang and Chapin) “found nesting in the cavities of small mushroom-shaped termitaria.” Ceeaniianaadae impressiceps (Mayr) Panga and Faradje, 8 (Lang and Chapin). The specimens from Panga were found inhabiting the hollow twigs of Barteria fistulosa (see Part IV), those from Faradje were associated with aphids. Crematogaster impressiceps variety frontalis Santschi, new variety “Worxer.—Length 3 to 3.5 mm. Pale brown; thorax less sculptured than in the typical impressiceps. Frontal groove deeply impressed. Stature less variable and smaller than in the typical form of the species and larger than in the variety longiscapa Stitz, but the scape also extends beyond the occiput as in that variety.” (Santschi) Numerous specimens from Malela (type locality) and Kunga (Lang, Chapin, and J. Bequaert); those at Kunga found nesting in the hollow internodes of the myrmecophyte Cuviera species (Part IV); the specimens from Malela “living in a small carton nest, about 9 em. long, fixed upon a stalk of Raphia.”’ Crematogaster menilekii (Forel) subspecies proserpina Santschi, new subspecies “Worker.—Length 3.2 to4.mm. Pale brownish yellow; head, gaster, and ap- pendages shining; thorax and petiole nearly opaque. Front and sides of head finely striate, the remainder with a few punctures. Anterior border of head and the corners obliquely truncated. Postpetiole narrower than in the typical menilekii, completely sulcate in the middle, forming two ovoidal eminences. Gaster broader than the head. Allied to C. alulai Emery and C. menilekii subspecies satan (Forel).” (Santschi) Numerous workers from Malela (Lang, Chapin, and J. Bequaert), with the following note: “ Ants living in the stalks of Papyrus and mak- ing carton nests in their crowns. The workers swarm out in great masses and let themselves drop on the intruder. They bite furiously and it is ‘Whack, Ant ofthe Bain Cong 155 g by preference the softer parts of the skin.’’ _ Crematogaster (Sphwrocrema) bequaerti (Forel) variety atraplex oaiialela Santschi, new variety = _ “Worxer.—Length 4mm. Rather dull yellow; gaster, postpetiole and femora ____ yellowish brown; tips of the epinotal spines brownish black. In other respects like the type of the species and the var. mutabilis (Santschi), but the median impression of the oe ngpreoneeel The dark tips of the spines contrast with the pale color of the (Santschi) Fig. 38. Crematogaster (Spherocrema) concara Emery. Worker from above. Crematogaster (Spherocrema) concava Mmery Text Figure 38 _ Akenge, ¢; Stanleyville, 8; Lukolela to Basoko, & (Lang and # iGhevin). The specimens from Stanleyville were taken in twigs of = jeohehmatenaes (Part IV); those from Lukolela were found running 18antschi has recently described a variety pluton of collected uaert from similar ian Gietein the sews of Pancras by ad ce re pate, Bes ts 156 Bulletin American Museum of Naiural History (Vol. XLV over fire-wood. Three specimens from Akenge were taken from the stomach of a toad (Bufo polycercus). Crematogaster (Sphwrocrema) pronotalis Santschi variety liebknechti (Forel) Text Figure 39 Numerous workers from Yakuluku and Garamba (Lang and Chapin). According to a note accompanying the specimens from the Fig. 39. Crematogaster (Sphwrocrema) pronotalis variety liebknechti (Fcrel). Worker from above. latter locality, this ant “builds small carton nests on the blades of grass. It is common in swamps, from three to five feet above water | Y Crematogaster (Spherocrema) rugosior (Santschi) “Fema.e (undescribed).—Length 8 mm. Thorax smooth and shining like the posterior half of the head and that of the worker, except its upper surface and the sides of the epinotum which have rug# as in the worker. Head rectangular, a little longer than broad, scarcely arcuate laterally. The eyes occupy nearly the middle third of the sides and the scapes barely extend beyond its posterior fourth. Clypeus with a strong median impression near its anterior border. Thorax as broad as the head. Epinotum nearly vertical, but the insertion of the spines is marked by an angular ridge which occupies nearly the upper half of the sides of the segment. Petiole ie 1922] Wheeler, Ants of the Belgian Congo 157 ~ MAM Worker, with = tooth henesth.” Wings 7 mm. long, hyaline, with brownish _ veins. Otherwise like the worker.” (Santschi) - Numerous workers and a few females from Stanleyville (Lang, ce: Gavin, and J. Bequaert), without further data. oe aia | Crematogaster (Spherocrema) striatula Emery variety obstinata - is i (Santschi) Sits workers taken by Dr. Boqueert at Leopoldville in the : i Seiaielacty inflated stipules of a species of Uragoga, a rubiaceous plant (Part IV). The spaces inhabited by the ants are not true nests but 3 ; _ merely kraals or stables for Coecide, as no larve or pupx were found in ae the structures. : et eaiiatheeater (Atopogyne) africana (Mayr) variety schumanni (Mayr) _ Anumber of workers taken by Dr. Bequaert at Leopoldville in the hollow stems of a Barteria Dewevrei (Part IV). Crematogaster (Atopogyne) africana subspecies laurenti (Forel) _ Numerous workers taken by Dr. Bequaert in the Rain Forest on the Tshopo River, near Rererrte,} in the hollow stems of Plectronia Lau- "ivenié (Part IV). Crematogaster (Atopogyne) africana subspecies laurenti variety zeta (Forel) Many workers and a few females taken by Dr. Bequaert at Pale (Niembo, between Walikale and Lubutu) from the myrmecodomatia of ___ ~ Plectronia Laurentii (Part IV) and at Leopoldville in the rudimen- tary leaf pouches of Randia physophylla (Part IV); also by Lang and Chapin at Stanleyville in the stem cavities of Cuviera angolensis (Part Eno) o The female of this form is black and striated as in the typical C. _ Crematogaster (Atopogyne) africana subspecies tibialis Santschi, new a subspecies “WorKER.— “Length 3.5 mm. “Pale castaneous. Epinotum, postpetiole, and posterior half of gaster of a _ deeper castaneous tint, passing to reddish brown. A spot on the vertex and the _ appendages dark brown, the tibie and metatarsi blackish, the tarsi and the ex- 158 Bulletin American Museum of Natural History [Vol. XLV tremity of the thorax reticulate, the epinotum more finely, with some fine longitu- dinal ruge on the whole basal surface. Sides of the mesonotum regularly reticulate- punctate. Sides of the pronotum more shining and of the epinotum longitudinally striate. Petiole finely reticulate; gaster finely shagreened, almost smooth. The pubescence is rather well developed on the head, the gaster, and the appendages, sparse on the thorax. The hairs are very sparse, except around the mouth and at the tip of the gaster. Head square, with rather convex sides and straight posterior border. Eyes at the middle of the sides. Frontal area short, feebly impressed behind. Frontal carine developed. Clypeus slightly convex, with rather arched anterior border. Mandibles striate-punctate, with four blackish teeth. The pronotum forms with the basal surface of the mesonotum a plane surface with a contour like that of C. castanea Smith. Sides of the basal surface of the mesonotum blunt, not marginate, with the anterior eminence scarcely indicated. Promesonotal suture little or not at all im- pressed. Sides of the pronotum marginate. Declivity of mesonotum oblique, feebly concave from right to left, above with marginate sides. Mesoépinotal furrow moderately deep. Basal surface of epinotum trapezoidal, its length equal to its width anteriorly in the small worker. It is convex in front, more feebly behind. The spines are as short as a fifth of the interval between their bases, which is concave. They are directed backward and slightly outward. Declivity as long as two-thirds of the basal surface and forming with it an angle of about 145°. Petiole trapezoidal, as broad as long, and as broad as the epinotum. Last antennal joint reddish. A fine and dense striation disposed as in africana (Mayr) but more or less effaced on the front, vertex and occiput, where the reflection is more shining than silky. Epinotum transversely striate-rugose. Petiole smooth, postpetiole and gaster very finely shagreened, almost shining. The head is, moreover, punctate as in africana and much less smooth in the individuals with large head. . “The head, which varies in size independently of the rest of the body, which is almost invariable, is sometimes longer than broad and scarcely emarginate behind, sometimes broader than long, strongly concave behind and with convex sides. Eyes more posterior than in africana. Frontal area narrow, strongly impressed and shining. Mandibles punctate, feebly striate. Mesoépinotal impression stronger than in africana, the pronotum less marginate anteriorly. Mesonotum carinate, more elongate and with the declivous surface much less abrupt than in africana, with longer epinotal spines, even longer than in the variety variegata (Mayr) and a little farther apart. Petiole and postpetiole as in africana.’ (Santschi) Numerous workers taken at the village of Mosekowa between Wali- kale and Lubutu by Dr. Bequaert from the peculiar pouches of Maca- ranga saccifera (Part IV) growing in the Rain Forest. As only adult ants and no brood were found in the pouches, Dr. Bequaert does not re- gard them as true nests. The openings of the pouches were not closed with fibrous carton. Crematogaster (Atopogyne) africana subspecies winkleri (Forel) variety fickendeyi (Forel) Numerous workers taken by Dr. Bequaert at Masongo, between Walikale and Lubutu, in the cavities of the branches of a species of Sarco- cephalus related to S. sambucinus (Part IV). 1922] Wheeler, Ants of the Belgian Congo 159 email cmaaeata. depressa (Latreille) variety fuscipennis Emery Plate X PBlatiey ville, 2; Medje, 8, 9; Niapu, 8; Ambelokudi, 8, 9; _ Nang, 8 (Lang and Chapin): Leopoldville, 8 (J. Bequaert). The beautiful carton nest of this ant is shown in Plate X, from a fine Gaigeaph taken by Mr. Lang at Ambelokudi. “It was built along the trunk of a tree near the ground. The ants, especially when squeezed, gave off a stench like certain bugs. They came out of the nest in great - numbers and let themselves drop to the ground.” The female C. depressa is very aberrant in the form of the head, which is large, flat, and rectangular, with peculiar mandibles. It has long been known and has been repeatedly renamed, but only recently has it been correlated with the cospecific worker. Crematogaster (Atopogyne) theta (Forel) "Plate XI, Figures 1 and 2; Plate XII, Figures 1 and 2; Plate XIII, Figure 1 Medje, 8; Avakubi, 8; Stanleyville, 8, 2, @ (Lang and Chapin). Risording to Santschi (in litt.), ‘‘this form represents the extreme Tieais of the subgenus Atopogyne. The worker has a feeble groove on the postpetiole, and the promesonotal impression is feeble. | Moreover, the _ female is brown, smooth, and shining, with spined epinotum, very _ different from the female of C. africana (Mayr) and the variety zeta (Forel).” ____ The specimens from Avakubi were collected by the natives, who call this ant “lona.” The carton nests are shown in PlateXI and XII. Con- _ ¢erning the specimens from Stanleyville, Mr. Lang writes: ‘These small black ants are very common. They build carton nests in trees, on the _ trunks of which they travel up and down in uninterrupted columns. At the slightest disturbance the nest is covered with workers. They appear and move so rapidly that it is very difficult to study them, especially as they sting disagreeably. Large numbers of nests may be found in the same tree, sometimes as low as ten feet from the ground, or even in bushes __ as well asin the tops of the tallest trees, living or dead. They have almost _ any shape, depending on their position, whether in forks of the branches _ rabout twigs. In the latter situations they resemble mere lumps. The more regular nests, however, are somewhat conical, like the tops of termite hills and are placed upright on the boughs. In color, the carton is grayish or dark brown. In size, the structures are rarely more than ‘two feet in height and about a foot in diameter. Their cells are irregular, re. = | \ — rot SA ee ' 160 Bulletin American Museum of Natural History [Vol. XLV the walls of the chambers being from 1 to 3 mm. thick, and there are many entrances and exits. Though very light, the nests are so tough that slices can be chopped off with a hatchet without breaking the remainder. The carton seems to be made from the fibres of rotten leaves worked up with secretions from the oral glands of the workers. The chambers are often full of brood, which is not confined to any particular part of the nest. The rufous females were present in such numbers that twenty or more could be lifted at a time clinging to one another on the points of the tweezers.” Crematogaster (Atopogyne) transiens (Forel) A few workers from Avakubi and a female from Stanleyville (Lang and Chapin). Crematogaster (Nematocrema) stadelmanni (Mayr) A single female from Stanleyville (Lang and Chapin), apparently taken at light, seems to be referable to this, the typical form of the species. Crematogaster (Nematocrema) stadelmanni variety dolichocephala (Santschi) Plate XIII, Figure 2 and Plate XIV Bengamisa, 8, 2 ; Manamana, 8, 9 ; Kwamouth, 8; Ngayu, 8, 2 (Lang and Chapin). Numerous specimens from all these localities. The specimens from Bengamisa were accompanied by the photograph of the nest shown in Plate XIV, and the following note: ‘‘ Ants from a pendent nest in very hard, woody carton. These nests are very common in the Rain Forest. They often fall to the ground but, in spite of the great moisture, resist disintegration fairly well. The ants leave as soon as the nest has dropped. The nests are precisely like those of some termites in shape and material, so that it is often impossible to decide from their external appearance which insect inhabits them. The internal cellular structure is very irregular and seems to follow no particular plan. The larve and pupe are found in any of the cavities. The nest represented in the photograph was fixed to several creepers and was practically swaying in the wind about twenty-five feet above the ground. Size and shape vary much according to the situation of the structure.” The fol- lowing note accompanies the specimens from Kwamouth, together with the photograph shown in Plate XTII, fig. 2: “Black ants taken from a — - ‘ 1922] . Wheeler, Ants of the Belgian Congo 161 nest hanging on a tree about nine feet from the ground. This nest was cone-shaped and was fastened to several small branches in such a manner as to sway when it was struck with a stick. The ants raise their -abdomens and sting quite furiously when annoyed. The nest is rough on the outside and very irregular, with a great many exits. The internal _ eellular structure resembles crumpled leaves overlapping one another tile. like the shingles covering a roof. The walls separating the chambers are very thin, only one-eighth to one-sixteenth of an inch in thickness. The whole of the nest that was photographed was about eighteen inches long and eleven inches broad on top. The brood was abundant in the lower- most chambers. The ants dropped by hundreds to the ground when the nest was hit.” Monomorium Mayr 7: The numerous species of this large and difficult genus are all small but form populous colonies, commonly with several fertile females. The worker is usually monomorphic, in the subgenera Parholcomyrmex and Holcomyrmex tending more or less to dimorphism. Clypeus abrupt, not sharply marked off from the frontal area, with two longitudinal welts or ridges often bordering an impressed median area and terminating anteriorly in projections or teeth. (These welts are fused in the subgenus Syllophopsis). Mandibles narrow, with few teeth. Maxillary palpi 1- to 2-jointed, labial palpi 2-jointed. Antennz 12-jointed, in a few subgenera 11-jointed, in one species (M. decamerum) 10-jointed, the club typically 3- jointed, but sometimes 4-jointed or indistinct. Promesonotal suture obsolete, the mesonotum more or less impressed at the mesoépinotal suture, the epinotum nearly always unarmed. Petiole pedunculate, with high node; postpetiole lower, rounded. Tibial spurs simple or lacking. The remace is always much larger than the worker, in some species wingless; in one Australian form (subapterum) with vestigial wings. Venation like that of Formica, with a discoidal cell, rarely without. . The maze is smaller than the female, always winged, with 13-jointed antennz. Mesonotum usually without Mayrian furrows, genital appendages completely retrac- The division of the genus was begun by Forel when he established the subgenus Martia. Emery' has recently revised the grouping of species and has established several additional subgenera. Viehmeyer has also proposed a subgenus Corynomyrmex, and Santschi has since added the subgenera Syllophopsis and Isoleomyrmex. In a more recent paper, Santschi proposes to give Syllophopsis generic rank. Geeae Ane. Boo. Ent. ue, LVITI, p. 261; and Bull. Soc. Ent. France, 1915, p. 190. 71921, Ann. Soc. Ent. ue, LXI, p. 120. 162 Bulletin American Museum of Natural History (Vol. XLV These subgenera (see the key, Part VII) may be arranged more or less according to their natural affinities in the following sequence: 1. Anillomyrma Emery 7. Monomorium, sensu stricto 2. Martia Forel 8. Notomyrmex Emery 3. Lampromyrmex Mayr (= Mitara 9. Xeromyrmex Emery Emery) 10. Parholcomyrmex Emery 4. Chelaner Emery 11. Isoleomyrmex Santschi 5. Adlerzia Emery 12. Holcomyrmex Mayr . 6. Syllophopsis Santschi 13. Corynomyrmex Viehmeyer The genus Monomorium, though cosmopolitan and of even wider distribution than Crematogaster since it occurs even in New Zealand and Patagonia, is represented by the great majority of species in the Old World. The Neotropical Region possesses only a few species of the typical subgenus Monomorium and the species of Martia, which are not known to occur elsewhere. The subgenera Notomyrmex, Adlerzia, and Chelaner are exclusively Australian. Anillomyrma is monotypic and known only from Ceylon. Isolcomyrmex and Syllophopsis are exclusively Ethiopian. Xeromyrmex is properly African but spreads into the Pale- arctic and Indian Regions. Holcomyrmex, Parholcomyrmex, and especially Monomorium, sensu stricto, are more widely distributed. Several of the species of Monomorium, sensu stricto, (minutum, floricola, pharaonis), Xeromyrmex (salomonis), and Parholcomyrmezx (gracillimum, destructor) have been widely désseminated by commerce. The species of Holcomyrmez are harvesting ants of dry regions and this is true of certain Australian species which are allied to Parholcomyrmez, though I assign them to a new subgenus Protholcomyrmex (with the type Monomorium rothsteini Forel) to be described in a later paper. , Monomorium pharaonis (Linnus) Numerous workers and females from Stanleyville and Thysville (Lang and Chapin). This is the well-known, little, red house ant, spread by commerce throughout the world. Monomorium (Xeromyrmex) bicolor Emery Several workers from Leopoldville (Lang and Chapin), found “living beneath a log,”’ and two from Garamba, taken from the stomach of a toad (Bufo regularis). This species is apparently widely distributed in the Ethiopian Region. Wheeler, Ants of the Belgian Congo 163 ~ SS Monomariam imadidiaatiie’) afrum Ern. André variety fultor Forel Many workers from Niapu and Garamba (Lang and Chapin). . ‘Those from Niapu “came in thousands to the body of a dead bird. They had their nest in a cleared place about thirty yards away. The following a, day they had moved their nest to the base of a decomposed root but evening had returned to their original nest. This extended about ____ two feet below the surface of the soil.” At Garamba the species was __ found “making crater nests about three inches high about the stalks of grasses in a dry plain (savannah) with few trees.” Thirteen specimens 5 from this locality were taken from the stomach of a toad (Bufo regularis). ! —— (Parholcomyrmex) gracillimum (I. Smith) subspecies robustius Forel , iuberal. workers from Yakuluku (Lang and Chapin); found living in small mushroom-shaped termitaria. The typical form of the species _is widely distributed in Asia Minor, Arabia, Central Asia, India, ete., and is evidently spreading to other parts of the Old World tropics (Africa, Java, Laysan, etc.). According to Emery, it occurs in the desert f Algiers, nesting under stones. The subspecies robustius was originally described from Somaliland. Yakuluku is in the dry portion of the Bel- gian Congo towards the type locality. 4 SoLenopsis Westwood A large and difficult genus of mostly hypogwic ants; usually with very small, pale workers and much larger and dark-colored females and males. _ The workers are usually monomorphic but in a few species, such as punctati- ceps Mayr, sxvissima (Smith) and geminata. (Fabricius), distinctly polymorphic. Antennz 10-jointed, first funicular joint large, club large, distinctly 2-jointed, the last joint very long. Mandibles narrow, with few (usually 4) teeth. Clypeus raised 4 _ in the middle and projecting anteriorly, with two diverging ridges, or carinw, each in all but a few species terminating anteriorly in a strong tooth flanked by a smaller tooth on the side. Frontal carine short, somewhat diverging behind. Eyes small, __ often minute or vestigial; ocelli very rarely present. Promesonotal suture indistinct, _ mesoé@pinotal suture well developed. Thorax more or less impressed at the latter. ) Epinotum always unarmed. Petiole with short peduncle and high, rounded node; _ postpetiole rounded, much lower than the petiolar node. The remae has 11-jointed (rarely 10-jointed) antennew and moderately large eyes and ocelli. Fore wings with one cubital and one discoidal cell; radial cell open. : The MALE is somewhat smaller than the female, with 12-jointed antennx. Seape very short, first funicular joint globular. Eyes and ocelli very large and promi- nent. Mesonotum without Mayrian furrows. pi sana campanulate; first gastric segment large; legs slender. 164 Bulletin American Museum of Natural History (Vol. XLV The genus Solenopsis is cosmopolitan, but represented by the great- est number of species in the Neotropical Region. There are a few forms even in Australia. The species with small, nearly blind, yellow workers, like S. fugax (Latreille) of Eurasia and S. molesta (Say) of North America, are hypogwic and usually live in the nests of other ants and termites, feeding on their brood (cleptobiosis). | Some species, however, (punc- taticeps, sevissima, geminata, gayi, etc.) live in large independent colonies. S. sevissima and geminata, the well-known “‘fire-ants”’ of the tropics, sting very severely. They have well-developed eyes and lead an epigzice life, not only feeding on insects and other animal food but also harvesting seeds or destroying the tender shoots or fruits of plants. Solenopsis punctaticeps Mayr subspecies kibaliensis, new subspecies WorkER.— Length 2 to 2.8 mm. Apparently less polymorphic than the typical punctaticeps and pr subspecies caffra Forel and therefore more like the subspecies erythrza Emery. Head in all the individuals rectangular, with straight sides, as broad in front as behind, not longer than broad in the largest, distinctly longer in the smallest individuals. Median teeth of the clypeus long and slender, lateral teeth obsolete or indicated only by feeble projections. Petiolar node broader than the petiole, its upper border straight and transverse. ; Sculpture much as in typical punctaticeps and the hairs almost as abundant as in that form, but much shorter and less erect, especially on the head. Color yellowish brown, legs and antenne yellow; mandibular teeth dark brown. Small workers searcely paler. MALE.— Length 4.3 mm. Head with very large eyes and ocelli, the latter extremely prominent; without the mandibles broader than long. Mandibles with 3 denticles. Antennal scapes nearly as long as the first two funicular joints together. Thorax broadly elliptical, slightly flattened above, only slightly longer than broad, much broader than the head. Epinotum bluntly subangular in profile, the base distinctly longer than the declivity. Nodes of petiole very low, rounded. Wings rather long; legs very slender. Smooth and shining; head subopaque and finely longitudinally striate behind. Hairs sparser and more reclinate than in the worker. Brown; head black around the ocelli; mandibles, antennz and legs yellowish. Wings rather opaque brownish hyaline, with very distinct brown veins and ptero- stigma. Described from twenty workers and a single male from Vankerek- hovenville (Lang and Chapin), on the Kibali River or Upper Uele. The specimens were living in small craters in the soil and were seen feeding on dead insects. 1922] Wheeler, Ants of the Belgian Congo 165 _ Emery! has recently revised the various subspecies and varieties of * ‘punctaticeps. The form described above is certainly distinct. I am ’ a _ not sure that I have seen the largest workers, although the series of Specimens is rather large. The single male is smaller and much paler s than that of the typical punctaticeps, which is described by Arnold as . “black” and as measuring 5mm. He found that the typical form of the _ species lives in large colonies, independent of other ants or termites, 2 though it is hypogzic, ‘rarely coming to the surface except in dull a) ne) f ; Azromyrma Forel sai this genus the worker phase is strongly dimorphic, being represented by a stibin worker proper and a much larger soldier, both with 10-jointed antenne and distinctly 2-jointed antennal club. The head of the soldier is large, suboblong and, in some species, furnished with a ridge with a slight tooth-like projection on each side near the occipital border. Maxillary and labial palpi 2-jointed. Mandibles 5- or 6- Clypeus without teeth and usually without carine. Eyes reduced to a few facets, the anterior ocellus well developed, the lateral ocelli absent. Pro- and mesono- tum high and convex; epinotum short, unarmed or with small teeth. Promeso- and mesoépinotal sutures distinct. Petiole with a short peduncle, its node rather low and transverse; postpetiole also transverse, somewhat broader than the petiole. Gaster large, elongate, as long as the remainder of the body. Legs short. In the worker the _ head is small, scarcely longer than broad, without ocelli and with the eyes even more reduced than in the soldier, the gaster smaller, not elongate. _ Femate larger than the soldier, but with shorter head. Antenne 11-jointed, but also with a 2-jointed club. Thorax elongate elliptical, mesonotum seen from above covering the pronotum. Wings long, with a closed radial cell, a discoidal and one cubital cell. Mate smaller than the female, not larger than the soldier, with long, filiform, 13-jointed antennz, the scape very short, the first funicular joint not swollen, not _ broader than the succeeding joints. Outer genital valve long, narrow, acuminately _ rounded at the tip; middle valve with a short, hollow, subtriangular, external ramus, and an extremely narrow, rather long internal ramus terminating in a hook; inner valve with three ridges, the mesial of which is strongly dentate and with its point _ directed obliquely to the base of the valve. _ «The genotype, A. nossindambo Forel, was described from males and _ females taken in Madagascar many years ago. Sikora later found the _ soldiers and workers in a termitarium at Amparafaravantsiv in the _ same island. Forel therefore expressed the opinion that the species of _ Aéromyrma must be cleptobiotic. The fact that Emery found a worker E attached to the tarsus of a female is suggestive in connection with con- _ ditions in Carebara (vide infra, P. 171). 41915, Rend. Accad. Sc. Bologna, N. 8., XIX, pp. 60-65. 166 Bulletin American Museum of Natural History [Vol. XLV For many years the genus was supposed to be monotypic and pecu- liar to Madagascar, but within recent years eight species and a variety have been described from the Ethiopian Region; Forel has also described a species from Sumatra (Map 23). A single soldier in the collection made by Lang and Chapin is certainly different from any of the species known in that phase. I describe it as new, although it may prove to be - soldier of one of a species based on workers. 45 is 0 is b+ ee eee ve -_ sosggtemees a e » x - - _, = a ? Zz is ° 1S » 45 a) 7 90 ws 120 135 160 Map 23. Distribution of the genus Aéromyrma. Aéromyrma petulca, new species Text Figure 40 Soiprer.— Length 2.5 mm. Head suboblong, nearly one and one-half times as long as broad, with feebly convex sides and rather deeply and angularly excised posterior border. Anterior ocellus well developed; eyes very small, consisting of about six ommatidia, situated at the anterior third of the head. Posterior corners of the latter with a low but distinct ridge produced on each side into a minute tooth. Mandibles convex, with 4 small, subequal, rather acute apical teeth, and a large blunt and flattened basal tooth. Clypeus flat, ecarinate, its anterior border feebly and sinuately excised in the middle, its posterior portion narrow, rectangular, extending back between the diverging frontal carine. Frontal groove distinct. Antennz 10-jointed; scapes rather slender and curved at the base, reaching to the middle of the sides of the head; joints 2 to 7 of the funiculus minute, subequal, nearly as broad as long (somewhat too long in the figure); club a little shorter than the remainder of the funiculus, with the basal joint longer than broad and about one-third as long as the terminal joint. Thorax decidedly Wheeler, Ants of the Belgian Congo 167 rt anc ‘narrower than the head; pro- and mesonotum convex, steep in front, d above; promesonotal suture distinct; mesonotum subcircular; metanotal net. In profile the dorsal outline of the mesonotum slopes backward con- the base of the epinotum without a distinct impression at the meso- - Epinotum with a small tooth on each side, its declivity longer than er steeply sloping. Petiolar node compressed anteroposteriorly, in | rather angular summit, from above transverse; postpetiole transversely | somewhat broader than the petiole, with a blunt ventral tooth. Gaster distended with a transparent liquid, elongate elliptical, longer than the ol the body, its anterior border straight in the middle. Legs short. opaque; mandibles, posterior portion of clypeus, , Mmesonotum, and gaster shining; mandibles id indistir punctate; head finely and regularly ull y ragulose; sparsely and rather coarsely punctate 7 & with fine, scattered, piligerous punctures. yellowish, moderately abundant, suberect, of un- h, most conspicuous on the dorsal surface; very se and appressed on the appendages. rinous red; legs and antennz paler and more yellow- dark brown above, with the venter and bases and of the segments broadly yellowish. sribed from a single specimen taken by Fig.4. Agromyrma : ai at Malela “from a small mush- petwen new species ped termitarium,” probably belonging to apove. wre of lies Putgifaber Sjéstedt. ‘ulca differs from africana Forel from the Kalahari in its slightly wee darker color, in having the postpetiole only slightly broader 2 (nearly twice as broad in africana), in possessing epinotal : id longitudinal ruge on the head. In africana the head is finely ite ste and the remainder of the body is evidently more shining than . In nossindambo the head is broader and less sharply rugulose, xrax is more deeply impressed at the mesoépinotal suture, the ‘scapes are much shorter, the anterior ocellus is smaller and the | 1 states that the gaster of the africana soldier is “transparent yellow,” which indicates that it was full of a clear liquid as in petulca. Th ee: is seen also in the soldiers of many species of Pheidole » Australia and in our Southern States and seems to indicate that this fe in the two genera mentioned often functions as replete or food- e individuals as in the honey ants (Myrmecocystus, Leptomyrmez, orus, Plagiolepis, and Prenolepis). 168 Bulletin American Museum of Natural History 3 {Vol. XLV Emery! believes that Aéromyrma should be reduced to the rank of a subgenus under Oligomyrmex “because in O. debilis Santschi the worker has 9-jointed, whereas the soldier (and probably also the female) has 10- jointed antenne, so that if one wished to distinguish the groups as hereto- fore, the worker of O. debilis would be classified in the genus Oligomyrmez, the soldier in the genus Aéromyrma.”’ While admitting that the two genera are very closely related, I prefer to retain Aéromyrma as an independent genus until the species are better known. Probably there are important differences in habit between the species of the two groups. At any rate, A. nossindambo and petulca are cleptobiotic with termites, whereas two or three species of Oligomyrmezx which I collected in Austra- lia were always found nesting in small cavities in rotten logs quite apart from termites. Aéromyrma species A single winged female from Akenge, taken from the stomach of a frog (Arthroleptis variabilis), cannot at the present time be referred to any of the described species, mostly known from soldiers and workers. CareBara I. Smith WorKER minute, monomorphic, yellow, without eyes or ocelli; antennz 9- jointed, joints 2 to 6 very small, the two terminal joints forming a large and distinct club, with very long last joint. Mandibles with oblique 3- or 4-toothed apical margins. Frontal carine short; frontal groove and frontal area absent. Clypeus simple, unarmed, without carine. Epinotum unarmed. Petiole with a short peduncle, its node higher and larger than that of the postpetiole; both nodes from above trans- verse, subelliptical. FEMALE enormously larger than the worker, dark-colored, with well-developed eyes and ocelli. Antenne short, 10-jointed, the funiculi without a distinct club, their joints 2 to 5 not much narrower than the remaining joints. Thorax large and robust, convex above, higher than the head, the mesonotum anteriorly more or less over- arching the small pronotum, with well-developed parapsidal furrows. Epinotum un- armed, or with low flattened lobes or protuberances on the sides. Tarsi densely clothed with short, stiff bristles. Wings large, the anterior pair rather pointed, with one cubital, a discoidal, and a closed radial cell and a well-developed pterostigma. MALE somewhat smaller than the female, but similarly colored, with long, 13- jointed antenn, scapes short, first funicular joint not swollen nor globular, remaining joints long and cylindrical. Mesonotum large, without Mayrian furrows. Nodes of petiole and postpetiole only feebly developed. The genus Carebara (Map 24) is represented by seven species in the Ethiopian and two in the Indochinese Region (C. lignata Westwood and . C. castanea F. Smith). Santschi described some females and males 11915, Rend. Accad. Sc. Bologna, N. 8., XTX, p. 59, footnote. damon Ants of the Belgian Congo 169 ee —_— i. ch etania as Carebara carinata.' The former measure 12 12.8 mm., the latter 9.3 mm. He is of the opinion that the species ert referred to the Neotropical genus Tranopelta, originally founded ayr on male specimens, are also to be referred to Carebara. Forel, yer, in his description of the workers of T. gilva Mayr variety | shows that Mayr’s genus is perfectly distinct. These workers somewhat dimorphic, have eyes, and both the workers and females y 1 L-jointed cman with a 3-jointed clava. The male alone is very 0 rr © 75 0s 20 a) < ¢ ch iin er Le iA QU NIC LA TRY | | Sy. Saal a 7 yj po MOS | AP, : ; \ = a \ b! An 4 ba ra ae i 5 2 Sei Aeee eee | c 12 t tae . Ua v ° is 0 45 © 15 90 105 120 135 Ce » 24. Distribution of the genus Carebara in the Old World. This genus also occurs in South rt / _ eotypes of brunnea in my collection. Emery* had previously based another Neotropical genus, Carebarella, on females and males of a species (C. bicolor) from Brazil and Peru. He also described a worker from -Ega, Brazil, under the name Oligomyrmex anophthalmus. _ At first sight the occurrence of species of Carebara and Oligomyrmex ‘South America seems very doubtful. During a recent trip to British Guiana I was able to secure all three phases of a new subspecies of ‘Santschi’s C. carinata and of the typical form of Tranopelta gilva. The Worker of the former shows that it is without a doubt a true Carebara, , 31912 Soc. Ent. France, , Hoon! Ball: Gen’ Ent. te 137. 170 Bulletin American Museum of Natural History [Vol. XLV and Prof. Emery, to whom I sent specimens for comparison with his Oligomyrmex anophthalmus, writes me that the latter, though specifically distinct, belongs to the same genus. It should therefore be known as Carebara anophthalma. The new subspecies of carinata was taken in a large termitarium of Syntermes dirus Klug, and it is interesting to note that of all the Neotropical termites this is most like the large Termes species with which the Ethiopian Carebarz live (vide infra). 1 took Tranopelta gilva, however, in the deeper parts of the nest of the large ponerine, Paraponera clavata (Fabricius), and also living indepeniemer with coccids under bark. Emery has placed Tranopelta and Carebarella with Diplomagiaaale Solenopsis in the tribe Solenopsidini and has made a tribe Pheidologetini for the genera Pheidologeton, Aneleus, Lecanomyrma, Oligomyrmex (in- cluding the subgenera Aéromyrma and Octella), Erebomyrma, Pedalgus, and Carebara. It would seem to be more natural to include all these forms in the single tribe Solenopsidini. Evidently Carebara, in the diminution of the antennal joints and the loss of the eyes in the worker, in the secondary reduction of this caste to monomorphism, and the secondary enormous enlargement of the females and males, represents the most extreme development of the whole series of genera, which prob- ably started from forms like the existing species of Pheidologeton. Since the volumes of bodies of the same shape vary as the cubes of their diam- eter, a female Carebara vidua measuring 24 mm. would be 4096 times as large as the cospecific worker, which measures only 1.5 mm., if the two insects were of the same shape. But the female is a much stouter insect in proportion to her length than the worker, so that she must be nearly 5000 times as large. And this disproportion occurs not only among in- dividuals of the same species but of the same sex and among the offspring of the same mother! The only other insects which exhibit a like disproportion are the workers and physogastric queens of the very termites with which Carebara lives as a predatory parasite. The extra- ordinary differences in stature between the workers and sexual phases of Carebara are undoubtedly correlated with interesting habits of the species. Haviland! was the first to show that C. vidua lives in the — masonry of the large nests of Termes natalensis in Natal. He discovered the minute workers but was unable to elucidate the relations of the ants to the termites. Forel (loco citato), inferring from analogy with our northern cleptobiotic species of Solenopsis (S. fugax, molesta, ete.) ad- vanced the hypothesis that the Carebara colonies live in cavities of their In Forel, 1901, Ann. Soc. Ent. Belgique, XLV, p. 392. Wheeler, Ants of the Belgian Congo 171 4 eens of the termitaria and that these cavities are con- fe galleries of the termites by means of very tenuous pas- h which the Carebara workers, but not the termites, can pass. a workers, probably remaining unnoticed on account of their ze, rey on the termites with impunity and are therefore able to e sexual forms. The larve of these are so voluminous that lc no’ ot be moved by the workers and are so soft and vulnerable y would have to be reared in chambers inaccessible to the a no detailed observations on the relations of the two es h ave been published, the subsequent accounts of observers in the : a onfirm Forel’s inferences. iE eq uaer eaibecened the marriage flight of Carebara junodi Forel. “This spe scies is remarkable on account of the extraordinary disproportion between le and the workers. In the Katanga it lives in the mound-shaped nests of spiniger. October 6, 1911, I witnessed at Sankisia a nuptial flight of ara very being te any ne oh ro ei rained abundantly. Toward noon numerous winged females were flying > in the savannah; they came from a certain number of termitaria, oh rcv wih ng te ey al i e species. I did not see copulation but, in the evening, I captured several ut but no females. The following days the phenomenon was not repeated. Carebara females are, among the aborigines of the Congo, a much- ia Seincy Hence they take advantage of the nuptial flight to collect a of individuals. The swollen portion of the abdomen alone is utilized. it either roasted or raw. iiecisert informs me that his attention was directed to the flight described above by the excitement of the congregated ho were actually filling pails with the torn-off gasters of the dee. Each Carebara colony gave off hundreds of females and the of workers that covered a termitarium during the flight must mn into the millions. The workers of Carebara, like those of other e ants (Erebomyrma, Acanthomyops, etc.), apparently come to e surface of the soil only while the nuptial flight is in progress. Be “Arnolé adds the following interesting note to his description of vidua. tis probable that the dense tufts of hairs on the tarsi of the female serve an porta t of enabling some of the minute workers to attach themselves body of the female when the latter is about to leave the parental nest. Several cimens of the female have been taken by me with one or more workers biting into ees Boel ae ieaGe XIV, p. 252. 172 Bulletin American Museum of Natural History (Vol. XLV the dorsal fimbriw. I am inclined to suspect that the young queen cannot start a new nest without the help of one or more of the workers from the old nest, on account of the size of her mouth-parts, which would probably be too large and clumsy to tend the tiny larve of her first brood, and that it is therefore essential that she should have with her some workers which are able to feed the larva by conveying to them the nourishment from the mouth of the queen. I find that the workers also attach themselves to the tarsi of the males. Two specimens of this sex referable to C. vidua, evidently taken at light and sent me by Mr. C. C. Gowdey from Kampala, Uganda, each bear two workers firmly attached by their mandibles to the tarsal hairs. Such workers must, of course, perish with their carriers, unless they can manage to pass over to the legs of the females during copulation. The workers and females of the African Carebare can be eta by means of the following keys. FEMALES 1. Large species, more than 20 mm. Jong....:.0.00.scs00s0+0neceteses cee eee Small species, not more than 15 mm. long... ..............000ecseeeseerces 4. 2. Mandibles with only 2 teeth and the remainder of their apical borders undulated, not.propesiy dentate.) jc: ssssctawkcsensiauseumeiael ampla Santschi. Mandibles with more than 2 teeth, entire apical border dentate............. 3. 3. Black; the gaster sometimes red; mesonotum about as broad as long; clypeal border not emarginate in the middle; hind metatarsi much shorter than Dull rusty red; mesonotum with three dark brown longitudinal stripes; thorax narrower; clypeal border broadly emarginate in the middle; hind meta- tarsi but little shorter than the hind tibie.................. junodi Forel. 4. Length 13 to 15 mm.; dark brown or castaneous..................0.-ee00s 5. Length only 8 mm.; ‘paler and more reddish brown... . .. osborni, new species. 5. Body covered with short hairs; clypeus merely coarsely punctate. .sicheli Mayr. Body almost hairless; clypeus tcaneverstly rugulose in the middle. langi, new species. WORKERS 1. Mandibles 3-toothed. Length 1.7 to 1.9 mm................. arnoldi (Forel). Mandibles 4-toothed......; ..:«:«:0:.0.0 000 spiaiegm apledtie s 6.0 bs 0.00 00's 4 0 \ = Me on ‘a N ‘ Frag ; J on tia — Te be me oe a a N ‘ Scant oie “SS eee ° 1S » 45 6 75 20 108 120 35 160 Map 26. Distribution of the genus Meranoplus. Meranoplus nanus Ern. André subspecies soriculus, new subspecies Text Figure 45 WorKER.— Length 1.8 to 2 mm. Head subtrapezoidal, as broad as long, rather convex and rounded above, trun- eated behind. Mandibles with oblique, 4-toothed apical borders. Clypeus rather flat, with a short median carina posteriorly. Frontal area transverse, crescentic. Scrobes deep, extending to the posterior corners of the head. Eyes rather large, con- vex. Antenne robust; club distinctly longer than the remainder of the funiculus. Pro- and mesonotum transversely rectangular, slightly broader than the head without the eyes, about one and one-half times as broad as long (somewhat too long in the 1922) Wheeler, Ants of the Belgian Congo 185 e figure), with sharply « dentate shteales corners, the sides distinctly emarginate at the ___ mesoépinotal suture, which is straight and very distinct. Mesonotum rounded on the __ sides and narrowed to the posterior border, which bears four short, blunt, flattened teeth, the median pair being smaller and more approximated. Epinotum vertical, unarmed, somewhat concave in the middle, with a longitudinal welt on each side __ representing the spines. Petiole cuneate in profile, the node much compressed . ety, much higher than the length of the segment, narrowed and bluntly Fig. 45, Meranoplus nanus subspecies soriculus, new subspecies. Worker. a, from above; 6,in profile. pointed above, higher than the postpetiolar node, which has a similar shape but is less compressed above and with broader, transverse border. Gaster large, convex above, pointed posteriorly, its anterior border excised in the middle. Legs rather stout. Shining; mandibles opaque, finely and indistinctly striatopunctate. Clypeus and upper surface of head longitudinally but not strongly rugulose, with indistinctly punctate-reticulate interrugal spaces. * Cheeks longitudinally rugose. Truncated posterior surface of head rather regularly reticulate rugose. Pro- and mesonotum with similar sculpture but the rugw are feebler, so that the surface is more shining; 186 Bulletin American Museum of Natural History {Vol. XLV sides of thorax and epinotum nearly smooth, as are also the petiole and postpetiole. First gastric segment evenly covered with shallow punctures interspersed with ex- tremely minute punctures. Hairs white, delicate, soft, and abundant, forming a uniform erect faite on 1 the upper surface of the body, more oblique on the appendages, on the legs interspersed with a few exceptionally long hairs. Brown; upper surface of head and first gastric segment, except at the base, dink brown; mandibles, except the teeth, legs, and antennw brownish yellow. FEMALE.— y Length 4.5 to 4.8 mm. 4 Head like that of the worker. Thorax broader than the head, about one and three-fourths times as long as broad; broadest through the pronotum, the sides of which are somewhat swollen, but have blunt, though distinct, teeth. Mesonotum somewhat broader than long. Petiole and postpetiole much as in the worker, but the postpetiolar node is thicker above in profile. Sculpture like that of the worker, but the mandibles coarsely striate and the sides of the thorax coarsely and irregularly reticulate rugose. Hairs yellow, coarser, and shorter, especially on the gaster, than in the worker. Color like that of the worker, but the mesonotum with three large, pata 9 dark brown patches. Wings yellowish hyaline, with pale yellow a MaLe.— Length 2.5 mm. Head, including the eyes, as broad as long, very convex behind. Eyes and ocelli large and convex; cheeks very short. Clypeus convex in the middle. Antennal scapes scarcely more than twice as long as broad; first funicular joint globose, second somewhat longer than the scape but distinctly more slender than the third joint. Thorax short, broader than the head including the eyes. Mesonotum convex, with distinct Mayrian furrows. Epinotum like that of the worker, but more sloping. Petiole longer than high or broad, the node low, angular in profile, with subequal anterior and posterior slopes, the former straight, the latter slightly concave. Post- petiole as long as high, somewhat depressed above, transverse, broader than the petiole. Clypeus smooth and shining in the middle. Head subopaque, reticulate-rugulose. Pronotum and epinotum indistinctly punctate-rugulose, subopaque; mesopleure smooth and shining; mesonotum and scutellum less smooth but shining, indistinctly punctate. Petiole longitudinally rugulose-punctate; postpetiole smoother. Gaster as in the worker but the large punctures are less distinct. Pilosity much as in the female, but the hairs on the body are even less even and on the legs are shorter and more appressed. Colored like the worker, but the antenn# and legs are yellow. The veins and pterostigma of the wings are distinctly paler than in the female. = Described from numerous workers, five females, and six males taken at Avakubi (type locality) and a number of workers from Medje (Lang — and Chapin). According to Mr. Lang, these ants “build small crater nests in the plantations. One crater was one and one-half inches high and four inches in diameter. The whole nest, three inches wide, ex- tended beneath the surface to a depth of only six inches. The workers Wheeler, Ants of the Belgian Congo 187 wo nove very slowly. The native name is ‘tungangele.’ Eight workers rom Medje were taken from the stomach of a toad (Bufo funereus). TI have described this form at length because it belongs to nanus André and is very closely related to Forel’s subspecies nanior and isc kiboshanus and to inermis Emery. The last I regard as a s of André’s species. All of these are known only from the 7 . -M. nanus measures 2.75 to 3.25 mm. and has two small, acute, ay m teeth on the epinotum. The subspecies nanior, though of the con e size as soriculus (1.9 mm.), is described as having the promesonotum re one and three-fourths times as broad as long, the variety kiboshanus ner s of the mesonotum rectangular and, judging from Emery’s figure, xs the mesoépinotal suture. The various forms mentioned are from idely separated localities, nanus from Gaboon, inermis from Transvaal i Rxitrea, nanior and kiboshanus from East Africa. ti oe Macromiscuorpes Wm. M. Wheeler is a small, monomorphic. Head subrectangular, with rounded posterior sorners, rather convex lateral borders, and convex, moderately large eyes at the middle of the sides. Ocelli absent. Mandibles triangular, their apical margins with numerous unequal teeth. Maxillary palpi 3-jointed; labial palpi 2-jointed. Clypeus convex; \ its anterior border entire or feebly notched in the middle; its posterior portion ex- ‘AS _ tending back between the frontal carinz; its sides not greatly narrowed and without a a trenchant ridge in front of the antennal fovea. Frontal carine short, rather far a _ apart, diverging behind, not prolonged as borders of scrobe-like depressions. Antenne - Cieto , 12-jointed, with a 3-jointed club, which is shorter than the remainder % funiculus, terminal joint somewhat enlarged, as long as the two preceding joints = Thorax rather long and slender, distinctly constricted in the mesoépinotal om with very long straight epinotal spines, but without metasternal spines. Pro- notum on each side above with a bluntly angular elevation, the inferior border broadly paapee Peduncle of petiole long and slender, the node compressed anteroposteriorly, -yery slightly squamiform. Postpetiole small, scarcely broader than the petiole, constricted behind. Gaster ovate, rather small. Legs long and slender; middle and pas ‘Fewate similar to the worker, but larger. Thorax not broader than the head _ including the eyes; pronotum not covered by the anterior portion of the mesonotum, which is short and convex. Epinotum sloping, with stout spines. Abdomen shaped _ much as in the worker. Fore wings with a single cubital, a discoidal and a closed cell. “ _ Mate nearly as large as the female. Head small, with prominent eyes and ocelli. well developed, with several teeth. Antenne 11-jointed, the second funi- cular joint representing three fused joints. Mesonotum without distinct Mayrian “vein Petiolar node very low. Cerci distinct; hypopygium with a bluntly rounded ; external genital valves short and stout, obtusely pointed. Legs very slender. Wings as in the female. = Genoryre.—Macromischa aculeata Mayr. I include in this genus also Mayr’s M. africana, which is hardly more | than a subspecies of aculeata. Emery placed both of these- forms in Tetramorium. Their habitus is certainly that of certain forms of Macro- mischa, as Mayr observed, but Emery was right in excluding them from that Neotropical genus. Both species are confined to the rain forests of West Africa (Map 27) and do not nest in the ground like the species of Tetramorium but build loose carton nests between leaves or on their under surfaces. Mayr claimed that the male aculeata has 11-jointed antenne, but Emery, after examination of six specimens, maintained that these appendages are 10-jointed and that Mayr’s specimens must — have been abnormal. There are four males in the Congo collection from two different localities and all of them have 11-jointed antenne. Emery probably overlooked the third funicular joint, which is rather rigidly articulated with the second joint so that the suture can be distinctly seen only in a favorable light. The number of joints in the male antennz, the shape of the clypeus in the worker and female, the absence of spurs F Piel, Ants of the Belgian Congo . 189 i the middle lo and ‘hind tibie, the long slender legs and antennz, the psence cl thie Mayrian furrows in the male, and the reduced number of p : joints are all characters which seem to me to justify a new generic _ The peculiar habits, too, are important in this connection, al- aa ‘h alone they would hardly justify a change in Emery’s allocation of “et he. species, since in a well-marked genus like Myrmicaria we have seen _ that some of the smaller species build carton nests on leaves whereas the larger species nest in the ground. The genus Tetramorium certainly mes more homogeneous by the removal of the two Mayrian species. Macromischoides aculeatus (Mayr) aks Plate XVII, Figure 1 . cs Stanleyville, 8; Avakubi, 8, 9; Bafuka, 8; Medje, 8, 2, 0; sangi, 2 ; Leopoldville, § (Lang and Chapin); Bumba, ¢* (J. Bequaert) Many Sioekere-and females and four males. [ae The following note by Mr. Lang accompanies the specimens from © ot Medje: “These ants build their nests by filling out interstices between oe eel leaves with a rough-looking, light mass of decomposed , matter. They prefer densely leaved trees and there are some- ____ times several hundred nests on the same plant. If one touches the tree, oN and at once rush out of their nests in great numbers and hurry along ___ the branches to reach the intruder. They cling to the human skin and _ double themselves up while biting and stinging. The result is rather painful and very annoying. There is no swelling but the pain endures for several minutes. All of the ants climb towards the head. The nests __ are often empty and contain only a few workers, but sometimes they are filled with brood and winged individuals. These ants have a strong odor, ___ especially when rubbed between the fingers.” In the plate (Pl. XVII, __ fig. 1) two of the nests are shown, one in situ, the other with one of the __ two thick leaves between which it was built removed. ____-M. aculeatus is so common in the Congo that its nests have been __ seen by several previous observers. Santschi' says of these structures: __ “Their nest consists of the leaf of a tree or shrub rolled up and lined with a _ felt-work of very fine vegetable débris and of a mycelium bearing fructi- _ fications. It would be interesting to study this fungus where it grows and to ascertain whether or not it is used habitually by the ants as food ___ and is cultivated for this purpose.” Commenting on the variety rubro- flava, Forel* remarks that it was “found in nests woven of silk, fixed to 11909, Ann. ey ey ty, fre Vaan S06. 71916, Rev. Suisse Zool +p. 421, 190 Bulletin American Museum of Natural History [Vol. XLV leaves, and, according to Mr. Kohl, similar to those of @cophylla and Polyrhachis. From this fact I conclude that the nest of 7’. aculeatum is probably only superposed on a woven tissue, i.e., it is a combination of carton and tissue, as I have proved to be the case in many species of Polyrhachis.” f Examination of a nest of aculeatus preserved in alcohol by Mr. Lang and conversation with Dr. Bequaert, who is well acquainted with the habits of the ant in the Congo, have convinced me that both Santschi and Forel labor under a misapprehension in regard to the structure of the nest. It consists of particles of the most diverse vegetable substances, bits of bark, dead leaves, trichomes, etc., loosely felted together and in- vaded by fungus mycelium, but the latter bears nothing resembling frue- tifications or ambrosial bodies such as are found in the gardens of fungus- eating ants. Dr. Bequaert informs me that aculeatus often nests in forests that are inundated during the rainy season and, as fungus hyphe in such situations in the tropics grow readily on any dead vegetable matter, it is not surprising that we should find them invading the loose carton of the aculeatus nests. These hyphe were interpreted as silk by Forel and suggested to Santschi the possibility of the ant being pepe phagous. Macromischoides aculeatus variety wasmanni Forel _ Numerous workers from Zambi (Lang and Chapin); one eae from Stanleyville. This variety is smaller than the typical aculeatus, with somewhat shorter epinotal spines, less regularly sculptured and somewhat paler. TETRAMORIUM Mayr Worker small, monomorphic. Antenne 12-jointed, with a 3-jointed club, Clypeus narrowed on the sides where its posterior margin is raised in the form of a short trenchant ridge or carina as the anterior border of the antennal socket. Frontal caring rather far apart, usually continued back some distance and often the full length of the head as subparallel ridges forming the inner borders of scrobes or demiserobes for the accommodation of the antennal scapes. Maxillary palpi 4-jointed; labial palpi 3-jointed. Eyes well developed; ocelli absent. Mandibles rather large, tri- angular, their apical border with a few large and several small teeth. Promesonotal suture indistinct, mesoépinotal suture more or less distinct; mesoépinotal constriction usually feeble; epinotum with two spines or teeth and episterna usually spined or dentate. Petiole with a short but distinct peduncle and the node large, subeuboidal, rounded above, rarely squamiform; the postpetiole usually broader than the petiole. Legs rather short, middle and hind tibie with small, simple spurs. Head, thorax, ane petiole sculptured, usually rugose or reticulate rugose. _ Wheeler, Ants of the Belgian Congo 191 MALE resembling the worker, but somewhat larger. Pronotum usually very posed above; mesonotum and scutellum raised above the level of the pro- uum, the latter with stouter and shorter spines than in the worker. Fore one cubital, one discoidal, and a closed radial cell. _ Mate slightly smaller than the female, with 10-jointed, very exceptionally with - oF 13-jointed antenne. Second funicular joint very long, representing a fusion of ints. Head small, ocelli and eyes large. Mandibles small but dentate. Pronotum rar d by the mesonotum, which has distinct Mayrian furrows. Epinotum trun- identate. Wings as in the female. This genus might be described as peculiar to the Old World, because ly all the few species occuring in America (7. cespitum, simil- um, and guineense) are known to have been introduced by commerce. e group reaches its greatest development in the Ethiopian Region so r as the number of species, subspecies, and varieties is concerned. 1 has included Triglyphothrix, Xiphomyrmex, and Decamorium ibgenera, but I have treated them as genera, though a few species simple Give may be assigned indifferently either to Tetramorium or ‘iglyphothriz. 1 have still further reduced the size of the genus Tetra- orium by establishing a new genus, Macromischoides, for T. africanum id aculeatum (vide supra). The species of Tetramorium form moderately ‘ge colonies and nest in the ground, usually under stones or logs. One e species, 7. czspitum, has a remarkable distribution, ranging from ‘itain to Japan, around the shores of the Mediterranean, and reappear- z at higher elevations on Mt. Kilimanjaro. Tetramorium sericeiventre Emery | ; Two workers from Thysville (J. Bequaert) and two others from wamba, taken from the stomach of a toad (Bufo regularis) by Lang and pin, are referable to this species, which is distributed over the whole Pon rican continent. ‘Tetramorium sericeiventre Emery subspecies continentis (Forel) ; Plate XVII, Figure 2 ____ Numerous workers from Zambi (Lang, Chapin, and Bequaert), found making small nests in sand (Pl. XVI, fig. 2). According to Mr. _ Lang’s notes, “the craters were often very regular, perfectly circular and 4 ‘composed of the excavated particles of white sand. The colony photo- graphed shows three entrances close together. The nest extended 4 about 50 em. below the surface to just above a moist layer of sand. The territory in which the ants nest is evidently inundated during the rainy season (at high water), but now (during the dry season) the water is a about four feet below the surface. One colony was seen covering small 192 Bulletin American Museum of Natural History [Vol. XLV areas about as large as the hand; the nest entrance was oblique, running under an overlapping thin layer of sand. The ants were working at noon in fairly bright sunshine. When disturbed, they all disappeared inside the nest. The craters consisted entirely of fine white sand-grains, with-- out admixture of food particles.” Tetramorium guineense (Fabricius) Two workers from Ngayu, taken by — and Chapin from the stomach of a toad (Bufo superciliaris). Tetramorium guineense subspecies medje, new subspecies Worxker.—Length nearly 4 mm. Decidedly larger than the typical guineense but of the same color, except that the head, thorax, petiole, and legs are concolorous and somewhat more brownish. Clypeal border distinctly emarginate in the middle; funicular joints 2 to 4 small, strongly transverse. There is a very distinct transverse crest to the pronotum like that described by Stitz for the subspecies cristatum. The epinotal spines are long, and stout, and curved forward as in the subspecies peulli Forel. The episternal spines are strongly curved upward and fully half as long as the epinotal spines. Petiolar node of the same shape in profile‘as in cristatum, with its anterior and posterior surfaces subequal, abrupt, distinctly concave and marginate above, but the node is much longer than in the typical guineense, broader behind than in front and with its dorsal surface roof-shaped as in peutli, Postpetiole robust, nearly as long as broad. Mandibles smooth and shining, with minute, scattered, indistinct punctures. Sculpture much coarser than in the typical guineense; clypeus with three prominent longitudinal carine or ruge; the ruge on the head and thorax longitudinal but connected by reticulations; the sculpture of both nodes equally coarse and as coarse as that of the thorax. Anterior fourth of first gastric segment sharply longi- tudinally striate. Pilosity yellow, decidedly longer and coarser than in the typical guineense. Described from two specimens, one taken from the stomach of a toad (Bufo regularis), from Medje (Lang and Chapin). This form is so strongly marked that it might be called a species, but, as many of its characters are those of described subspecies of guineense and as I have seen only two specimens, I prefer to attach it provisionally to that species. Tetramorium meressei Forel A single worker taken by Dr. Bequaert at Masaki (between Masisi and Walikale) agrees very closely with Forel’s description, except that — the erect hairs on the body are coarser and not “wooly” and the gaster is not darker in the middle but uniformly yellowish brown like the re- mainder of the body. Dr. Bequaert took his specimen from one of the domatia of a Cuviera (probably C. angolensis), the other swellings of which were occupied by Engramma denticulatum Wheeler. 1922] Wheeler, Ants of the Belgian Congo 193 Tetramorium pusillum Emery variety hemisi, new variety ‘Worxer.—Length 2.5 to 2.8mm. Agreeing closely with Emery’s description of 1¢ typical pusillum in size, sculpture, and coloration, but with the basal third or 1 of the first gastric segment densely punctate and nearly opaque, and with the teeth acute. The latter are distinctly larger than the metasternal teeth. - Deseribed from fourteen workers taken from the stomach of a frog - (Hemisus marmoratum) from Niangara (Lang and Chapin). The Abys- subspecies ghindanum Forel is slightly larger than this variety 4 (at least this is true of several cotypes sent me by Prof. K. Escherich many years ago) and the opaque basal portion of the gaster is more ex- tensive and finely striolate-punctate. Tetramorium setigerum Mayr subspecies querens Forel Plate XVIII, Figures 1 and 2 Numerous workers from Niapu (Lang and Chapin). The note accompanying the specimens states that they “form a ring of loose particles of soil about the entrance of their nests during the rainy season, each ant carrying the particle to a certain distance and then letting it drop and returning at once to the entrance. During the dry season they carry out the particles and food-remnants without attempting to con- struct a crater. The photographs (Pl. XVIII) show the difference in the. appearance of the nest during the wet and dry seasons. These ants are very common, as about a dozen colonies were observed about the village of Niapu. They were usually situated along the paths or in clearings and seem to prefer dry soil.” Tetramorium simillimum (F. Smith) A single worker from Stanleyville (Lang and Chapin). This is a common tropicopolitan ant, now widely distributed by commerce. Pppeeriam simillimum subspecies isipingense Forel variety dumezi Forel A single worker taken by Dr. Bequaert at Thysville. XrpHomyrmex [orel This genus is very closely related to Tetramorium, differing only in having the antennez of the worker and female 11- instead of 12-jointed. The scrobes of the antennz are well developed in all the species known to me. The genus is widely distributed, being represented by a number of species in tropical Africa, Madagascar, the Indomalayan and Australian Regions and by one species, X. spinosus (Pergande), with several sub- species, in the Sonoran Province of North America. The various species nest in the ground, like Tetramorium, often in very populous colonies. 194 Bulletin American Museum of Natural History [Vol. XLV _ Xiphomyrmex angulinodis Santschi Medje, 8, 2, co; Irumu, 9 (Lang and Chapin). Santschi has described all three phases of this species from the French Congo and has figured the worker and male. The specimens before me agree perfectly with his account. They bear no data beyond the localities. Xiphomyrmex occidentalis Santschi subspecies akengensis, new sub- species Worxer.—Length 1.8 to2 mm. Smaller than the typical form, which measures 3.5 mm., with the mandibles red, the tarsi, middle and hind coxe and tips of fore cox brownie yellow, and the remainder of the legs and the antennz reddish brown. The seventh funicular joint is as long as broad; the eyes smaller and more flattened than in the type, scarcely more than one-sixth as long &s the side of the eal ails anterior orbits somewhat narrowed and bluntly pointed. The postpetiole is twice as broad as long, its node somewhat transverse and compressed anteroposteriorly, the petiolar node also somewhat broader and more squamiform than in the type. In other respects agreeing very closely with Santschi’s figure and description. Described from numerous specimens taken at Akenge (Lang and Chapin) from a single colony in “a dark brown paper nest.’’ There is nothing to show that these specimens were not inhabiting the abandoned nest of some other ant. A single dedlated female from Liberia in my collection belongs, in all probability, to this subspecies. It measures nearly 2.5 mm. and is very much like the worker. The larger eyes are not bluntly pointed in front, though rather flat. The thorax issmall, with small mesonotum, bluntly pointed in front and not covering the pronotum, the epinotal spines are much stouter and further apart than in the worker, the petiolar node is broader, more squamiform and more transverse above, more sharply separated from the peduncle, and with its anterior surface decidedly concave. The color is the same as that of the worker, the body being brownish black with the appendages paler. as RHOPTROMYRMEX Mayr Worker small, allied to Tetramorium. Antenne 12-jointed, with 3-jointed club, as long as or slightly longer than the remainder of the funiculus. Maxillary palpi 3- jointed; labial palpi 2-jointed. Head broader behind than in front, with convex sides and small, moderately convex eyes at the middle of its transverse diameter. Ocelli absent. Clypeus flattened or moderately convex, ecarinate, its anterior border entire, a little produced, narrowed on the sides and bluntly ridged in front of the small an- tennal fovee. Frontal carine short and more or less diverging; frontal area large but not impressed. Serobes absent. Thorax short and stout, convex and rounded above, with feeble or obsolete promesonotal suture, somewhat constricted or impressed at the mesoépinotal suture, the epinotum unarmed. Petiole pedunculate, the node \ a ae Wheeler, Ants of the Belgian Congo 195 “a ea ee s = . T . \ 0 de ey \ \Y a OF) | y | : \\ Be eA { | | bs i% y k Py i e.'% S = of as ‘i 4 , rs ‘q c | a as , : 3 y Meh | } < - ie o \ > \¥ ~ eee i | 7 a = | > ~ 7 é “a5 : Map 28. Distribution of the genus Rhoptromyrmex. ed, narrower than the postpetiole, which is transversely elliptical and rounded : _ Gaster oval, formed very largely by the first segment. Legs moderately ” Jong, femora not inerassate in the middle, the middle and hind tibise with or without _ short simple spurs. __ Fare somewhat larger than the worker, with 12-jointed antennz, but differ- ing considerably in structural details in the various species. Fore wings with a cubital, eee ee * cpen zadlil eal Mate with 10-jointed antennx and elongate second funicular joint, as in Tetra- - morium, and closely resembling the males of this genus also in other respects. Wings | sin the female. © pale, eee this genus are confined to the Ethiopian Region (Map _A few Indian forms formerly referred to the genus have been re- placed by Emery in a new genus, Acidomyrmez, characterized by Pirin ry lot straight and diverging epinotal spines. q a : Rhoptromyrmex opacus [Forel _ Numerous workers taken at Thysville by Bequaert. These were found nesting in sandy sil in the savannah, 196 Bulletin American Museum of Natural History (Vol. XLV TRIGLYPHOTHRIx Forel Small ants closely allied to T'etramorium. The worker has 12-jointed antenne, the funiculus terminating in a 3-jointed club. Mandibles and clypeus as in Tetramorium. Head with distinct scrobes, often divided by a longitudinal carina for the reception of the folded scape and funiculus. Thorax short and stout, the promesonotal and mesoépinotal sutures nearly or quite obsolete. Epinotum and episterna armed with spines much as in Tetramorium. Pe- tiole pedunculate, its node and especially that of the postpetiole decidedly broader than long. Hairs on the body abundant, soft, erect, trifid or many-branched, cover- ing the surface like a delicate white mould. FeMA.e similar to the worker but larger; anterior wings with one closed cubital cell and an open radial cell. cs F q = afb is m Ye ‘ s a. ‘ ia ; ye Pe eet 4 = j NE 2 | it | | . . - = .. 2 -* > '_ - - ee ee we es ; Map 29. Distribution of the genus Triglyphothriz. The crosses indicate the localities wae z. striatidens (Emery) has been found outside of its range. pS MALE with 10-jointed antennw, the second funicular joint very long, the third shorter than the first. Mesonotum with Mayrian furrows. Petiolar and postpetiolar nodes narrower than in the worker and female, the petiole subpedunculate. This genus is paleotropical, ranging over the Ethiopian, Indomalayan, and Papuan Regions (Map 29). One Indian species, 7’. striatidens (Emery), is rapidly spreading to other parts of the world and has been taken in such widely separated localities as Queensland, Formosa, Tunis, Sierra Leone, Seychelles, Barbados, Mexico, Louisiana, and England, In the locality last mentioned it occurs in the hothouses of Kew Gardens. The species of Triglyphothrizx are all very timid, usually curling up and feigning death when touched. They live in the ground. One South African species, 7’. arnoldi Forel, according to Arnold, “is most fre- 'According to iciedt this species is a typical Tetramorium, in which genus I have placed it in the catalogue of Ethiopian ants. Wheeler, Ants of the Belgian Congo 197 ~ ly found in the nests of other ants, apparently in plesiobiotic or stobiotic association.”” He mentions its occurrence in the galleries vo large Ponerinz, Platythyrea lamellosa subspecies longinoda variety esi. a Forel and Ophthalmopone berthoudi Forel. | Triglyphothrix gabonensis Ern. André a 8, 9; Niapu,8; Ngayu, 8; Medje, 2 (Langand Chapin). ee eeihees and two deiilated females, all taken from the stomachs : ° = s x 45 a s = 105 a IY 160 * - t UH é FPS NI LES \ i \ ! / 1g ‘ aie _ “ ao ° ~ ‘a $ ° ey Aan ae . % < A . poe=-T--~~. AS ty eo! ee L 2 a << K ™ “~ ‘ Le 9 ‘ . v4 bad \ ar- -_ C= -----4 — Peat ay A a , yt “ PT att ~ » a ‘ Yl wey woe & 2 3 5 ° is » 45 6 15 90 105 120 Bs ay Map 30. Distribution of the genus Cataulacus. Triglyphothrix mucidus Fore] eS Medie, 8; Ngayu, 8; Boyulu, $ (Lang and Chapin). Four speci- _ mens from the stomachs at toads (Bufo funereus). CatTautacus I. Smith Worxer.—Small or medium-sized, rather flat, opaque or subopaque, black ants, __ with coarse sculpture and the head and thorax often dentate or spinulate on the sides. Antenne in all three phases 11-jointed with 3-jointed club and apically flattened or = dilated scape. Head on each side with a deep scrobe situated beneath and external to the eye and capable of accommodating the whole of the folded antenna. The g frontal carine are far apart and diverge, but border the scrobes only at the base. The clypeus is wedged in between the frontal carinw and is not sharply delimited posteriorly. Thoracic sutures often indistinct or obsolete. Epinotum armed with spines. Petiole and postpetiole stout. the former usually more or less cuboidal, with a 198 Bulletin American Museum of Natural History [Vol. XLV laminate process below, the latter subglobular. Gaster elliptical or suboblong, flat- tened, the first segment forming its whole dorsal surface. Legs rather short, the femora and tibia incrassated. The rema.e, though larger, closely resembles the worker. The pronotum is large and forms a considerable portion of the thoracic dcrsum. Wings without a dis- coidal cell, with a single cubital and a narrowly open radial cell. The MALE resembles the female and worker in the shape of the head but has larger and longer petiole and postpetiole. The mesonotum has well-developed Mayrian furrows. The ants of the genus Cataulacus bear a strange superficial resemblance, both in structure and habits, to those of the Neotropical genus Cryptocerus. The genus ranges over tropical Africa and eastward over Madagascar, India, Indonesia, and the Philippines, but is repre- sented by the greatest number of species in the Ethiopian Region (Map 30). Concerning the habits, Arnold says that “all the species of this genus are tree-ants, usually forming medium-sized nests in hollow twigs and stems, or more rarely under the bark. They are timid and slow- moving insects, often feigning death or dropping rapidly to the ground when disturbed.”’ He has seen them breaking open the earthen tunnels constructed by termites on the trunks of trees and attacking the inmates. Cataulacus erinaceus Stitz Text Figure 46 Stanleyville, 8; Risimu, 8 (Lang and Chapin). The collection contains many workers of this large and beautiful species, originally described and figured by Stitz from the Cameroon and Spanish Guinea. Forel many years ago gave me a specimen labelled “ Cataulacus princeps Emery” and has himself refer- red to it under that name, which seems to exist only in manuscript. Lang and Fig. 46. Cataulacus erinaceus Stitz. Chapin found this ant running up and Worker from above. down large trees. a _ and Chapin). Numerous specimens, all ap- a from Akenge were taken from the stomachs Wheeler, Ants of the Belgian Congo 199 Gataulacus guineensis F. Smith Text Figure 47 ___ $tanleyville, 8; Bolobo, 8; Leopold- 2 a aiile to Yumbi, 8, 9; Lukolela to Basoko, 8; Tsangi, 8; Medje, 8, 9; Akenge, @ (Lang F parently belonging to the typical form of the species. Many were taken on fire-wood. Nine workers from Medje and twelve males J of toads (Bufo polycercus, funereus, and _ tuberosus). Cataulacus egenus Santschi Medje, 8, 2 (Lang and Chapin). : 4 ‘Numerous specimens. The hitherto unde- PR 47; -Clitiaterns guinemeis scribed female measures 4 to 4.5 mm.,andis *: Smit. Worker from above. very similar to the worker except in the structure of the thorax. The mesonotum and sides of the pronotum are longitudinally rugulose. The wings are whitish hyaline, with the anterior border suffused with yel- low, the veins pale yellow, the pterostigma dark brown. Cataulacus pygmeus Ern. André subspecies luje (Forel) Five workers from Garamba (Lang and Chapin) without further data. Cataulacus tregaordhi Santschi variety plectroniw, new variety Worxer.—Length 2.8 to 3.2 mm. Smaller than the typical form of the species and the variety ugandensis Santschi. The ruge of the head have no longitudinal trend, those on the pronotum are very coarse and somewhat transverse and those on the mesonotum and base of the epinotum fine and indistinctly longitudinal The portions of the antenn» and legs, which in the typical form are red or yellowish red, are whitish yellow or white, the scapes, and tibie being paler than the tips of the femora, the tarsi brownish yellow, the funiculi reddish brown. Described from two dozen specimens taken at Stanleyville by Lang a and Chapin from the cavities of a species of Plectronia (Plectronia A, see Part IV). Dolichoderine A very homogeneous subfamily, comprising comparatively few genera. WorkKER monomorphic, very exceptionally (certain species of Azteca) more or less dimorphic. Clypeus protruding between the insertions of the antennw. Antenn» 12-jointed (except in Semonius). The metanotum contributes to the dorsal face of the thorax, being wedged in between the epinotum and mesonotum, and its stigmata are often 200 Bulletin American Museum of Natural History (Vol. XLV protuberant. Pedicel formed by the petiole alone, the postpetiole forming the basol segment of the gaster; the following segment without stridulating surface. Sting vestigial, except in Aneuretus, where it is well developed and can be protruded. Usually there is one pectinate spur on the middle and hind tibia, homologous with the median spur of the Ponerine; sometimes with a second, lateral spur which is much smaller and simple. FemaLe always winged; similar to the worker. Some genera still retain a more generalized wing venation with two closed cubital cells and one discoidal cell; but frequently the venation is more or less reduced, often considerably so in the mate. Antenne of the male 13-jointed, even in Semonius. The Dolichoderine: males with two closéd cubital cells can usually be separated by the well-developed mandibles from such Ponerine as have no constriction behind the postpetiole. The clypeus protruding between the frontal carinz is a good char- acter by which to separate them from the male Formicinz with a similar venation. Nympus never enclosed in a cocoon. The anatomy of the gizzard or proventriculus is very important for the taxonomy of this subfamily; for a description of this organ the student is referred to the writings of Forel' and Emery.* The larve are fed with liquid food, almost always of vegetable origin, regurgitated by the workers; the Aztece are mostly insectivorous. All the workers possess anal glands, the secretion of which hardens on ex- posure to the air, becomes sticky, and has a peculiar, often unpleasant odor like that of rotten cocoanuts or rancid butter; it is used as a means of protection against other insects. The habits are rather varied; many species are inconspicuous, shy, and live in small colonies under bark of trees or in dead wood. In the Australian Leptomyrmex the worker can store vegetable liquids in its much inflated crop (honey ants). Several species of Iridomyrmex, Azteca, and Engramma inhabit the cavities of various myrmecophytic plants, and are undoubtedly adapted to this peculiar form of symbiosis. Other species of Aztecx build carton nests, often of large size, which may be free, attached to branches or trunks of trees, or may be placed inside cavities; certain species are associated with epiphytes which cover their carton nests; according to Ule, these “gardening ants”’ carry soil and seeds of these epiphytes in the branches of the trees. The Argentine ant, Iridomyrmex humilis (Mayr), is one of the most troublesome pests of tropical and subtropical countries. Its original home was South America, whence it has recently spread through a large part of the globe. It is sometimes found in hothouses of temperate regions. In the Ethiopian Region it has thus far been recorded from ‘Forel, 1878, ‘ Anatomie du gésier des fourmis,’ Bull. Soc. Vaudoise Sc. Nat., XV, pp. 339-362, PI. 1. ?Emery, 1888, ‘Ueber den sogenannten Kaumagen einiger Ameisen,’ Zeitschr. Wiss. Zool., XLVI, pe oe Pls. xXviI-xxrx; 1912, ‘Genera Insectorum, Dolichodering,’ pp. 4-5. See also Wm. M: eeler, 1910, * Ants,” pp. 33-35. XXII or Wheeler, Ants of the Belgian Congo 201 | Afri ! its appearance is said to date from the time of Ss ast Boer War, when it was probably introduced with forage (Arnold). eee pone hoes near Cape Town; it is also very injurious 2 “i we be —"* a oe ~ Pe ‘ any ¥ \ - we 7 Ais Se %, . ; (es = a “4 a a) ee a)’ Sas ay eo on . Nites Bu ek . Fag % ae 2 } . = ea FP te j a ‘. ot ae “ co ll! 7 i * rk | oan —" 4 A Wm v “ : ae cl ee i : ? 6) : Meee * al Vc a . SS | ee ; a at jy gees ; - oe on? Oe — eee |) : coutiin = =) = ¥ = = < race af” A ‘at Ie eee me ; 3M > - L x _ ee, Map 31. Distribution of the genus Engramma. Encramma Fore! os BM, viksbad is Filion. _ Worker small, monomorphic, with the head more or less excised behind and the anterior border of the clypeus semicircularly notched in the middle and posteriorly __ extending back between the short but widely separated frontal carina. Maxillary oleae labial ra ards peat Antenne 12-jointed, with long first and last funicular joints. Gizzard with narrow, separated, anchor-like sepals. Gaster large, ip semen negment overlying the petiole; anus terminal or subterminal. Femaxe larger than the worker; its fore wings with a discoidal, one cubital and - w elosed radial cel, 3 5 ‘ ‘See the references given under this species in the catalogue of Ethiopian ants. an Oe 202 Bulletin American Museum of Natural History (Vol. XLV MALe as small as the worker, with 4-jointed labial and 5-jointed maxillary palpi. Antennsz long, filiform, 13-jointed, the scape as long as the first and second funicular joints together. Mandibles large, denticulate, decussating. Mesonotum not over- arching the pronotum. Wings as in the female. } This genus has been known only since 1910 and comprises six described species. It has a very narrow range, being confined to equa- torial Africa and in all probability to the forest regions (Map 31). Most of the species evidently live in the cavities of myrmecophytes. At least one, however, lives in the ground (wolfi) and another, zimmeri subspecies okiavoénsis of the Congo, is said to inhabit “a large pale gray nest, soft, woven and mixed with fine vegetable matter and applied to the trunk of a tree.”’ The workers of the previously known and of three new species described below may be separated by means of the following table. 1. Mesoépinotal constriction very deep and long, so that the thorax is halteriform ; epinotum with a pair of denticles above........ denticulatum, new species. Mesoépinotal constriction only moderately deep, acute; epinotum without Gomtiches. «0 dicais< «6s p'e oy «ao hat ee aelinn + 02s 0 v's bao .0cin denn 2. 2. Body long and slender; head and thorax opaque; antennal scapes extending at least one-fifth their length beyond the occipital border................ 3. Body more thickset; head and thorax shining; antennal scapes shorter... . . . 4. 3. Seapes surpassing the occiput by one-fifth their length; clypeal notch very pS and deep and the median border behind it with a small sion; all the funicular joints twice as long as broad; color black, with easel SEOMUAGAS |. < « v Kanon Rabe de Re hones tne tga cane -wolfi Forel. Scapes surpassing the occiput by about one-fourth their length; clypeal notch broad and shallow, the median border without an impression; funicular joints shorter; color yellowish" red, gaster black, its segments narrowly ordered. with Omow:. =.<<. 27s set eke ca oot zimmeri Forel. 4. Eyes very large, nearly one-third as long as the sides of the head.. .ilgi Forel. Byes much smaller... 1. ¢. cosdses wince eas ss ys 6as voce ee vss ewan 5, 5. Head, without the mandibles, as broad as long, deeply excavated behind. . . . . . 6. Head longer than broad, feebly excavated behind.......................- 8. 6. Antennal scapes slightly surpassing the occipital border; funicular joints 2 to 7 slightly longer than broad; base of epinotum nearly as long as the declivity, horizontal; pilosity well developed ai 5 a 5 5) i tae lujz Forel. Antennal scapes not reaching the occipital border; funicular joints 2 to 7 broader than long; base of epinotum much shorter than the declivity, sloping forward; pilosity less abundant. 6. .564.0 <0. c-s0s00+ cen Hae + 7.-:Beown; length 2.25.00 3: aaiiicisces suey sate bak ors de ks gnae laurenti Emery. Dull yellow, with brown gaster; smaller, length 1.8 to 2.2 mm.... .kohli Forel. 8. Antennal scapes reaching the occipital border; funicular joints 2 to 7 broader than long; epinotum evenly rounded, without distinct base and deelivity. griseopubens, new species. __Wheeler, Ants of the Belgian Congo 203 ! aie Nick eek Uhide length beyond the occipital order; funicular joints all distinctly longer than broad; epinotum with yaa ‘short base sloping forward and long, flat declivity, sloping backward. “ ss Be tly gowdeyi, new species. 6 Fig. 48. Engramma kohli Forel. Worker. a, head from above; 6, thorax and petiole in profile. — Engramma kohli Fore! Text Figure 48 Viapt 8, Q@ (Lang and Chapin); Lubutu to Kirundu, 8; Tshopo near Stanleyville, $ (J. Bequaert). The specimens ‘cin Niapu taken in the leaf ascidie of Scaphopetalum Thonneri De Wildeman rand (see Part IV); those from Lubutu to Kirundu in the ar structures of Cola Laurentii De Wildeman (see Part IV); those from the Tshopo River were found nesting in the stem swellings hairy Plectronia (species A, see Part IV). The type speci- is of the species were also taken by Father Kohl “in myrme- plants”’ at St. Gabriel, near Stanleyville. : ‘ ‘ b Fig. 49. Engramma lujx Forel. Worker. a, head from above; 6, thorax and petiole in profile. 4 Roe. Engramma luj# Forel pe ‘Text Figure 49 oe A single worker of this species was found among the numerous speci- mens of the preceding species and was taken in the leaf-pouches of um Thonneri at Niapu. I have compared it with a cotype of — from Prof. Forel and = in the accompanying Dae iN gm! 204 Bulletin American Museum oj Natural History {Vol. XLV Engramma wolfi Forel Text Figure 50 Akenge, 8, 2; Ngayu, 8; Medje, 8 (Lang and Chapin); Walikale to Lubutu, 9, o (J. Bouasit): Femae (undescribed) .— Length 4.6 to 5 mm. Very similar to the worker. Head scarcely excavated behind. Eyes about two- fifths as long as the sides of the head. Clypeal border each side of the notch flattened and angularly projecting. Head and thorax a little more finely punctate and therefore a little more shining than in the worker. Epinotum feebly convex, sloping, without Fig. 50. Engramma wolfi Forel. a, head of male; b, head of worker; c, thorax and petiole of worker in profile; d, fore wing of female; ¢, antenna of male. distinct base and declivity. Dark brown; mandibles, antenne and wing-insertions _ pale brown; legs, including the coxw, white, with a dark brown band around each femur and the tips of the hind cox of the same color. Wings grayish hyaline, with pale brown veins and pterostigma. Mate (undescribed).— Length nearly 3 mm. Head through the eyes as broad as long. Eyes and ocelli large. Mandibles well developed, decussating, with long, very finely and evenly denticulate apical borders. Clypeus short, with nearly straight, entire anterior border. Antenne long and slender; rs A nd er Wheeler, Ants of the Belgian Congo 205 wand al joints excep the frst funicular, cylindrica the latter as broad as long broader than the succeeding joints. Thorax short, not broader than the head; ; ~oenennewanl broader than long, not overhanging the pronotum. Epinotum slop- (lane papel Petiole with more distinct trace of the node ae anterior end than in the worker. Genitalia moderately large, exserted. Legs eee venetnon on int the female: 2 and pilosity much as in the female, the hairs and pubescence being very sparse and short, the former apparent only on the mouth-perts and tp of the = hace ¢ hileck brown; front of head and three large spots on the mesonotum pale rusty mn; mandibles pale yellowish; scapes, first funicular joint, and legs, including the , sordid white; the femora without brown bands. Wings and their veins a little he ate: id eeecimens am Akenge, Ngayu, and Medje (a female and four __ workers) were taken from the stomachs of toads (Bufo polycercus, super- ciliaris, and funereus), those from Walikale at lights. Kohl took the __ workers from which Forel described the species in the virgin forest in the mm ground among rotten leaves. This habit accounts for the occurrence of _ specimens in the toads’ stomachs. Engramma denticulatum, new species Text Figure 51 - Worxer.— q Length 2.6 mm. F Head subhexagonal, a little longer than broad and slightly broader behind _ than in front, with the sides subangulate in the middle and the posterior border _ feebly concave. Eyes moderately large, near the middle of the sides. Mandibles rather small, convex, with three large apical and several small basal teeth. Clypeal notch small, semicircular, less than one-fifth as long as the anterior border, with sharp __ gorners. Frontal area indistinct; frontal groove obsolete. Antennal scapes extend- __ ing somewhat farther than their greatest diameter beyond the posterior corners of the head; first funicular joint as long as the two succeeding joints together; joints _ - 2to 7 about one and one-half times as long as broad, joints 8 to 10 slightly longer than broad. Thorax long, with very deep and broad mesoépinotal constriction so that it is dumb-bell-shaped, the pronotum and mesonotum convex and hemispherical above, the impression bearing the prominent metathoracic spiracles, the epinotum high and 206 Bulletin American Museum of Natural History [Vol. XLV convex like the promesonotum, with two blunt denticles and prominent spiracles. Petiole stout, through the distinct node-like thickening at its anterior end nearly half as high as long. Gaster shaped as in the other species of the genus, with the first segment overlying the petiole; anus terminal. Shining; head and clypeus finely but distinctly longitudinally aciculate; man- dibles smooth, with coarse, scattered punctures; pronotum finely and indistinctly punctate; meso- and epinotum opaque, densely and rather coarsely punctate; gaster finely reticulate. Pilosity and pubescence very sparse, the latter distinct only on the appendages. Deep castaneous, nearly black; apical portions of mandibles, bases of anee, terminal tarsal joints and petiole yellowish. Described from two specimens taken by Lang and Chapin between Lukolela and Basoko on fire-wood. Two imperfectly preserved specimens were taken by Bequaert at Masaki, between Masisi and Walikale, from the caulinary swellings of a Cuviera (probably C. angolensis; Part TV). — This is a very strongly marked species on account of the peculiar shape of the thorax, the two denticles of the epinotum, and the ees sculpture of the head and thorax. Fig.52. Engrammagriseopubens, new species. Worker. a, head from above; b , thorax and petiole in profile. Engramma griseopubens, new species Text Figure 52 : WorKER.— Length 2.7 mm. Head without the mandibles slightly longer than broad, much broader: behind than in front, with somewhat angularly excised posterior border and feebly convex sides. Eyes small and flat, in front of the middle of the head. Mandibles rather large, convex, their long apical margins with numerous crowded denticles. Clypeal notch semicircular, about one-fifth as broad as the anterior margin. Frontal carine some+ what closer together than to the lateral margins of the head. Frontal area and groove obsolete. Antenn rather slender, scapes not reaching to the posterior corners of the head; first funicular joint twice as long as broad, remaining joints except the last, as broad as long. Thorax with sharply marked promesonotal and mesoépinotal sutures, the pro- and mesonotum forming a hemispherical mass, the latter circular, the humeri rounded; the mesoépinotal constriction moderately deep, acute; thé epinotum lower than the promesonotum, only a little longer than the mesonotum, broader than long, in profile rather convex, sloping, without distinct base and declivity. Wheeler, Ants of the Belgian Congo 207 , elliptical, with its anterior border thickened above as the First gastric segment overlying the petiole as in the other species of ; anus nearly terminal. Legs rather slender. _ Shining; whole body very finely and uniformly punctate. Hairs absent, except on the mandibles and tip of the gaster, where they are very | -Pubeacence gray, short and fine, rather abundant, uniformly covering the body, but not concealing the surface. sk; mandibles, sides of clypeus, cheeks and gula brown. pes “ ibed from a single specimen taken by Lang and Chapin on d between Lukolela and Basoko. This species is quite distinct e node. . species, though not from the (ideo: may be most con- tly described in this place. Fig. 53. Engramma gowdeyi, new species. Worker. a, head from above; 6, thorax and petiole in profile. eter etme ei ao i: 2 -, er; Engramma gowdeyi, new species ae Text Figure 53 — Worker.— ss Length 2.4 to 2.7 mm. - Head without the mandibles distinctly longer than broad, broader behind than in front, with feebly concave posterior border and feebly convex sides. Eyes flat, in front of the middle of the head, about one-fifth as long as its sides. Mandibles convex, with about a dozen even, crowded teeth. Clypeal notch about one-fourth the length of the anterior border, broader than deep, with sharp, slightly produced corners. Posterior clypeal border distinct; frontal area and groove obsolete; frontal carine nearer to the sides of the head than to each other. Antennal scapes extending about one-sixth their length beyond the occipital border; funicular joints 2 to 10 perceptibly longer than broad. Thorax short, seen from above with distinctly angular humeri; __ promesonotal and mesoépinotal sutures distinct; pro- and mesonotum moderately __ eonvex, the latter broadly elliptical, slightly broader than long; mesoépinotal con- striction rather deep, acute; epinotum as long as broad, broader behind than in front, in profile with a short base, rising rather steeply from the mesoépinotal suture, one- fourth as long as the flat, backwardly sloping declivity. Petiole elliptical, flat, its node obsolete. Gaster rather voluminous, its first segment overlying the petiole; anus terminal. Legs rather slender. Shining; very finely and uniformly punctate. 208 Bulletin American Museum of Natural History (Vol. XLV Hairs sparse, blackish, erect, rather coarse, present on the clypeus, vertex, pro-, meso-, and epinotum, and all the segments of the gaster. Pubescence grayish, short and fine, rather abundant, covering the whole body but not concealing the shining surface. Castaneous brown; thorax and anterior portion of head paler; mandibles, in- sertions of antenns, funiculi, tarsi, and articulations of legs yellowish brown. Described from numerous specimens taken by Mr. C. C. Gowdey at Kampala, Uganda. [| at first supposed this form to be Z. ilgi subspecies stygium Santschi, described fr6m British East Africa, but careful perusal of the description shows that it is quite distinct. — ieee | Ty DTT PF FR i Same i) Kober th, CP totet oa vine Y — Na a oe 4 eet \ \ \ é T : r * ft fo Map 32. Distribution of the genus Tapinoma. Tapinoma Firster WokrKER small, monomorphic, with 4-jointed labial and 6-jointed maxillary palpi, multidenticulate mandibles, 12-jointed antenne and entire or medially more or less emarginate clypeus. The node is reduced to an anterior thickening of the depressed or flattened petiole which is overlain by the first gastric segment; anus usually in- ferior. Gizzard short, calyx usually not divided into distinct sepals, feebly convex, covered with fine hairs, with the bulb almost exposed when viewed from the side. _ The FeMALe is usually considerably larger than the male. The anterior wings have a single cubital cell, rarely two, and the discoidal cell is often lacking. The MALE is commonly as small as the worker and has well-developed denticulate mandibles. Antenne filiform, with long scape, usually surpassing the posterior border of the head and as long as the three first funicular joints together. Thorax stout; mesonotum not overhanging the pronotum. Genital appendages voluminous, the stipes with a large squamula and its free portion of variable shape. Wings as in the female, but the discoidal cell is often lacking in the smaller species. —— : 1922] Wheeler, Ants of the Belgian Congo 209 —— " PGolouies of Tapinoma are usually populous and live in the ground or a ‘in the cavities of plants. The workers are timid and emit from their anal glands a strong odor like that of rancid butter (‘‘ Tapinoma-odor’’). a ‘The genus is cosmopolitan and in the Nearctic Region reaches to rather high latitudes and altitudes (Map 32). One of the species, Tapinoma __ melanocephalum, has been widely distributed by commerce throughout a the tropics of both hemispheres. It is often a pest in shops andis known _ in Cuba as the “hormiga bottegaria.” _ Tapinoma luridum Emery subspecies longiceps, new subspecies _ Worxer.—2.5 to 3 mm. Larger than the typical form and the subspecies connexum Santschi and differing also in the following characters. The head is longer, narrower behind and the posterior border is straight, not convex, as figured by Santschi for connerum, nor concave, as described by Emery for the type. The anterior elypeal border is straight in the middle, not feebly notched, and the scapes surpass the occipital border by nearly a third of their length. The eyes are decidedly smaller than in either of the other forms of the species. The thorax and petiole agree with Santschi’s figure of connerum. The body is uniformly lustrous or moderately shining, the pilosity as described by Emery for the type, the pubescence exceedingly fine so that it somewhat dims the shining surface. The color is uniformly brown, except the tarsi, which are pale brownish yellow. Described from numerous specimens found by Lang, Chapin, and Bequaert nesting in a deserted carton termitarium on a tree on the forested bank of the Congo River at Zambi. This form is so distinct that it may prove to be an independent species. TECHNOMYRMEX Mayr Allied to Tapinoma and distinguished by the peculiar structure of the gizzard, the calyx of which is covered with small clear spots apparently representing thin areas in the chitin. The anus is terminal in the worKER and FEMALE. The former is small and monomorphic, the latter but little larger. The anterior wings have two closed eubital cells and a discoidal cell. The mae has a short antennal scape, not longer than the two first joints of the funiculus. Wings like those of the female, but with the cubital vein more or less interrupted near the second cubital cell. In one species, 7’. albipes, both apterous and winged males are known to occur. The genus is confined to the Old World tropics, ranging over the Ethiopian, Indomalayan, Papuan, and Australian Regions (Map 33). Some of the species nest in the ground, others make small carton nests on the bark of trees. 7. albipes is being rapidly disseminated in the tropics by commerce and sometimes occurs in hothouses in temperate regions. 210 Bulletin Amerigdan Museum of Natural History (Vol. XLV Technomyrmex nigriventris Santschi Two workers taken by J. Bequaert at Thysville “beneath decaying leaves on the soil of a patch of forest.” Formicine (Camponotine of authors) Worker monomorphic or more or less polymorphic, only in a few cases with pronounced dimorphism (Dimorphomyrmer). Frontal caring often feebly developed and the clypeus is only exceptionally produced between them (Dimorphomyrmez, Gesomyrmex); even then, it is not properly wedged in. Antenne 8- to 12-jointed, usually long and filiform; the funiculus rarely with a feeble 4- or 5-jointed club. Abdominal pedicel always formed by one segment, the petiole, which is usually Is ° 1S 30 45 © 7S 90520 135 160 ual etch : VFI, \ 5 0 ol 1s ooo) - oo y 2 \ d C— “s 0 1S ° 1s » 45 6 15 20 105 120 135 6 Map 33. Distribution of the genus Technomyrmez. seale-like; there is never a trace of constriction between the second and third ab- dominal segments and the stridulatory organ is also lacking at the base of the third segment. Sting vestigial; the poison-glands are converted into a cushion of convolu- tions (Forel's pulviniferous vesicle) ; the sting forms merely the sustentacular apparatus for the orifice of this poison vesicle. The ejaculation of the poison can in certain genera ~ (Formica) be effected with great force. Orifice of the cloaca always circular and terminal, ciliated round the margin. FeMALE always winged and similar to the worker, though of much larger size. MALE winged, with thr genitalia always exserted. The venation of the wings is more or less reduced, often considerably so. In its most primitive stage there is still one cubital, a closed radial, and a closed discoidal cell; but there is no intercubitus, the radius and cubitus being confluent over 4 part of their course (Formica-type). Reduction has usually started by the disappearance of wie, Ants of the Belgian Congo 211 . the recurrent vein, ae Mists no discoidal cell (Camponotus, cophylla). An inter- - eubitus is only rarely present and then very short (Myrmoteras, which has the most ‘ Den venation of this subfamily). = Nympas usually enclosed in cocoons; but there are some exceptions ((co- ee - phll, Prenolepis). me bet The members of this subfamily are morphologically the most highly — develc of all ants; this is also true for their ethological peculiarities. o t only are their habits very diverse, but they show the most specialized _ form of mental and social behavior. The diet is in large part vege- P See and these ants show great predilection for sugary substances, _ which are sometimes stored in a special, replete form of worker (honey a ants: Melophorus, Myrmecocystus, certain Plagiolepis, etc.). The species s of Ecophylla and certain Polyrhachis and Camponotus build silk nests in leaves, using their larve as silk-producing shuttles. Moreover, the nest- co . ing habits i in this subfamily are very varied. Certain species of Formica and Polyergus are slave-makers; the species of Polyergus are true s social parasites of Formica, entirely dependent upon their slaves, but the __ worker caste is still present. ¢ PLaGioLeris Mayr Worker medium-sized to very small, monomorphic or feebly polymorphic. _ Mandibles rather narrow, with oblique, usually 5-toothed, apical borders. Clypeus large, convex, carinate or subcarinate, lozenge-shaped, its anterior border arched and somewhat over the bases of the mandibles. Maxillary palpi 6-jointed, labial palpi 4-jointed. Frontal carine short, subparallel, rather far apart. Frontal area poorly defined. Antenne 11-jointed, inserted very near the clypeal suture, the _ funiculi slender, gradually thickened towards their tips, the first joint long, the re- maining joints gradually lengthening distally, the terminal joint elongate. Eyes moderately large and flat, placed in front of the middle of the head. Ocelli usually _ absent. Thorax short, more or less constricted in the mesonotal region, the epinotum simple and unarmed. Petiole with its scale anteriorly inclined, its superior border _ entire. Gaster rather voluminous, elliptical. Legs slender. Gizzard with the calyx strongly reflexed, parasol-shaped. Femate much larger than the worker. Head small, thorax and gaster massive, ___ the mesonotum somewhat flattened above, the gaster elliptical. Antenne 11-jointed. Wings long, with one cubital cell and usually without a discoidal cell. a MAL somewhat smaller than the female. Mandibles acutely toothed. Frontal area large. Antenn# 12-jointed, with long scapes; funiculi with elongate first joint. Thorax voluminous, mesonotum large, flattened above, covering the small pronotum. _ Petiole as in the female. External genital valves large, rounded. Wings as in the female. 4 Pur enclosed in cocoons. : This genus is peculiar to the warmer parts of the Old World (Maps _ $4 and 35) and is represented by the largest and most numerous species in the Ethiopian Region. Two of the latter, P. custodiens and stein- Bulletin American Museum of Natural History [Vol. XLV 212 0 ss » 4s oo % 90 0s 120 a ws oe a “5 Pd / AE Pam! e j J dr ~~ j ” g = 2 4 4 ’ N ; SAN y ‘N “ - \ ‘ \ ial & - s --- \ * so soles Be Vet. a ee f » - —~ N ae wi Chere 4“ is ° 1S » “as a ti) 90 105 120 5 160 Map 34. Distribution of the genus Plagiolepis. is 0 1S 30 45 ao 1° Js} tae ¢ — , ” oy ae mi rr ' 4- —- s ] 4 = / ‘ 8/ 0} ’ to - ‘\ : r , / i) ~ 7 ’ £ ‘ Pt ee a ea ae a See SR Oe) a “a 3% .° - 4“ s ° is » 4s a 75 90 105 120 135 160 Map 35. Distribution of the subgenus Anacantholepis (crossed area) of Plagiolepie and of the allied genus Stigmacros (dotted area) Wheeler, Ants of the Belgian Congo 213 ol ‘ eee : . ri, resemble our northern species of Formica in stature and structure. ngle aaa species, P. longipes (Jerdon), has been widely. ibuted by commerce in the Old World tropics and has also gained a ng in Mexico. Another species, P. nuptialis Santschi, recently dis- 1 by Dr. Hans Brauns in the Cape Province, is parasitic on P. (vide infra). So far as known, the species of Plagiolepis nest exception of P. mediorufa, which inhabits plant-cavities. Santschi has recently separated the genus into three subgenera: : is, sensu stricto, ia eaaere and Anoplolepis, on the struc- we Plagiolepis mediorufa (Forel) ee aa workers from Stanleyville (Lang and Chapin), taken from : renal of Cola Laurentii. This form was originally described _ as a simple variety of the Palearctic P. pygmxa (Latreille), from speci- _ mens taken by Kohl “dans une plante myrmécophile,” near Stanley- ville. It should, in my opinion, be regarded as a distinct species on account of its peculiar habitat, for pygmza nests in the soil under stones. Moreover, the worker mediorufa is decidedly smaller, with much shorter _ antenne, the median funicular joints especially being distinctly shorter than long, whereas in pygmza they are longer than broad. The head is e- proportionally smaller and narrower, with more rounded sides and with q the occipital border straight or slightly convex, not concave as in pygmza. Plagiolepis (Anoplolepis) custodiens (I. Smith) Plate XIX, Figures 1 and 2 _ Banana, 8, 2, o’; San Antonio, § (Lang and Chapin). At Banana this species was found nesting in flat craters in the _ pure sand of the sea-beach (Pl. XTX, figs. 1 and 2). According to a note by Mr. Lang, “the ants were found very near the water, where the sand was moved by the wind or even inundated by the breakers. Only a slight _ excavation, marking the entrance of the nest, was visible, and it was difficult to trace out the galleries. These ants carry particles of sand _ considerable distances, sometimes two or three feet from the nest- _ entrances. They work during the day-time and retreat into their nests _ when disturbed.” ____ P.. custodiens has been previously taken in Banana by Busschodts and in Angola by Silvestri, and is well known from other parts of the Ethiopian Region as far north as Abyssinia and as far south as the Cape. 214 Bulletin American Museum of Natural History [Vol. XLV It is the host of P. nuptialis Santschi, which was discovered by Dr. Brauns at Willowmore, Cape Province. Up to the present time only males of this ant have been taken. Dr. Brauns, who sent me a series of them, writes me March 24, 1920, as follows: “I am well aware of the interest attaching to the parasitic habits of P. nuptialis. Hitherto I have been unable to discover the female, but hope to unearth it eventu- ally: The males always come out of the nests of P. custodiens and most | years are not uncommon at Willowmore. I also found the male flying in numerous swarms over the Keurbooms River on the coast, near | Plettenberg Bay, during a rain-storm, but could nowhere find them in copula with females. Perhaps the female is unable to fly! The males often remain for months at a time in the custodiens nests before swarm- ing, which occurs only during a shower. The nests of P. custodiens and steingréveri are frequently close together, but the latter does not harbor nuptialis, though both species usually have the same myrmecophiles. At Willowmore steingréveri is showing a tendency to displace custodiens.” It would seem from Dr. Brauns’ observations that nuptialis, like the North American species of Epecus, Sympheidole, and Epipheidole, must be a workerless parasite. Plagiolepis (Anoplolepis) tenella Santschi Niapu, 8; Bafwasende, 8; Garamba, 8, co; Akenge, 8; Medie, 8 (Lang and Chapin). The specimens from Akenge and Medje were taken from the stomachs of toads (Bufo funereus and polycercus) and two males from Garamba from the stomach of a Bufo regularis. The Niapu speci- mens were found running about on the ground in the clearing of a native village. The female of this species was mentioned by Forel from specimens 4 found in the stomach of a pangolin (Manis temmincki) from the Lower 4 Congo, but was not described. The hitherto undescribed male measures about 5mm. The wings are long (6mm.). The head is only about half as broad as the thorax, broader through the eyes than long, with small, acutely 5-toothed mandibles. Color, sculpture and pilosity as in the worker, but the head is dark brown behind and the thorax is more shin- _ ing, with three obscure, brownish, longitudinal blotches on the mesonotum. ee ACANTHOLEPIS Mayr Worker small, monomorphic. Head subquadrate, rounded laterally and posteriorly. Mandibles with oblique, dentate apical borders. Clypeus broad and high, carinate or subcarinate. Clypeal and antennary fovee confluent. Frontal area small but distinct, triangular. Frontal carine subparallel, short, rather far apart. . Maxillary palpi 6-jointed, labial palpi 4-jointed. Antenne 11-jointed, inserted close { Wheeler, Ants of the Belgian Congo 215 —— ne aoe c ‘a to the elypeal suture; scapes long, funiculi slender, not thickened distally. Eyes moderately large, ocelli distinct, rather far apart. Thorax constricted at the meso- notum, the pronotum broad and usually convex anteriorly, somewhat compressed posteriorly; promesonotal and mesoépinotal sutures distinct; epinotum more or less __ swollen and obtusely dentate on each side. Petiolar scale bidentate or more or less excised above. Gaster broadly oval, with rather pointed tip. Legs slender. Gizzard .. much like that of Plagiolepis. ____ Femaxe larger than the worker. Head resembling that of the worker but broad- . SD oeal Gibind. Thorax robust, mesonotum large, gibbous in front where it overhangs em cbecely longitudinally carinate in the middle as is also the scutellum. unarmed or bluntly dentate. Wings with a single cubital cell and usually a discoidal cell. Mate scarcely larger than the worker and resembling that caste in the shape of © aiead. Eyes large, cheeks ‘very short. Antenne 12-jointed; scapes long and em funiculi filiform, all the joints elongate, the first shorter than the two fol- : bape: together. Thorax massive, about as broad as high; epinotum oblique, un- armed; mesonotum slightly convex but not subcarinate. Petiolar scale inclined for- ward, its upper border entire. External genital valves small, elongate, triangular. - Like Plagiolepis, the genus Acantholepis is confined to the warm pase of the Old World, one species, A. frauenfeldi (Mayr), occurring as far north as southern Europe, Syria, and Persia. In Australia the genus is represented by a peculiar group of species, Stigmacros, which Forel regards as a subgenus but which, I am inclined to believe, should be raised to generic rank. The colonies of Acantholepis are moderately populous and usually nest in the ground, under stones, or in the fissures of rocks, rarely in the cavities of plants. Acantholepis capensis Mayr variety anceps Forel 5 Stanleyville, 8; Medje, 8 (Lang and Chapin). Numerous speci- mens. This variety is close to the subspecies depilis, having sparse, short, whitish pilosity. In shape the epinotum and scale, as Forel re- marks, approach those of the subspecies simplex Forel. The variety was _ originally described from specimens taken by Kohl in the Belgian Congo, _ probably near Stanleyville. According to a note by Mr. Lang, this ant _ makes tiny craters in the soil after the rain. The colonies seem to be _ rather small, judging from the few workers seen outside the nests. Acantholepis capensis variety guineensis Mayr | A single worker from Thysville (Lang and Chapin) appears to be- long to this variety, which is not black, like the other forms of the species, but reddish brown. The hairs are yellowish. It was originally described from the Gold Coast. q s 216 Bulletin American Museum of Natural History (Vol. XLV Acantholepis capensis variety validiuscula Emery Thysville, 8 (J. Bequaert, Lang and Chapin). Five opeavanian” This variety is decidedly larger and more robust than the typical capensis, with abundant, erect, dark brown pilosity. It seems to have a wide distribution, since it is known from Abyssinia, the Congo, Rhodesia, and Cape Province. o Acantholepis capensis subspecies canescens (Emery) — Thysville, 8 (J. Bequaert); Avakubi, 8 (Lang and Chapin). A form with long, white pilosity and abundant pubescence, distributed throughout the Ethiopian Region. A note by Mr. Lang states that “these small ants had their nest in the dirt which had accumulated at the bases of the cut leaves on the stem of an oil palm. They were numer- ous and travelled continually up and down, one by one, without forming a regular file. There were numerous nests along the trunk of the palm, but all of them were situated in the higher portion of the hollowed, partly decomposed stumps of the leaf-stalks, which had been cut off for some time. These hollows had evidently been made by the ants themselves.” Acantholepis capensis subspecies canescens variety cacozela Santschi- Faradje, 8 (Lang and Chapin). Four workers taken from the hol- low stems of an unidentified plant belonging to the family Melasto- macewe (Dissotis). This variety has longer hairs than the typical canescens and the petiolar scale is thickened at the summit, with scarcely excised border. Acantholepis carbonaria Emery Two workers from Banana (Lang and Chapin), without further data. This opaque species, originally described from Somaliland, hes also been previously taken in the Belgian Congo. PRENOLEPIS Mayr Worker small to very small, monomorphic, the body, legs, and scapes usually beset with sparse, coarse, erect, blunt hairs. Head rounded subrectangular or sub- elliptical, with rather narrow, dentate mandibles, their apical borders oblique. i peus large, convex, its anterior border entire or sinuately emarginate in the middle, not or scarcely produced over the bases of the mandibles. Frontal carine very short and straight; frontal area poorly defined. Antennary and clypeal fosse not confluent. Maxillary palpi 6-jointed; labial palpi 4-jointed. Antennz 12-jointed, inserted near the posterior angles of the clypeus; scapes elongate, funiculi filiform or slightly thiek- ened distally. Eyes moderately large; ocelli absent. Thorax short, more or less constricted in the mesonotal region. In some species the mesonotum is elongate and subcylindrical. Promesonotal and mesoépinotal sutures distinct. Epinotum more 0, s k ] Wheeler, Ants of the Belgian Congo 217 a convex above, unarmed. Petiole with an anteroposteriorly compressed scale, sh is inclined forward. Gaster oval, convex in front, where the first segment covers etiole, the tip pointed. Legsslender. Gizzard long and narrow, its calyx straight base, with the sepals reflected at their anterior tips. __ Femate decidedly larger than the worker. Head proportionally small; thorax and ‘massive; pronotum short, vertical; mesonotum broad, flattened, with distinct upsidal furrows; scutellum convex, often longitudinally impressed in the middle. with a single cubital cell; discoidal cell present or lacking. [aLe scarcely larger than the worker. Head resembling that of the worker and ale. Mandibles usually edentate. Antenne 13-jointed; scapes rather long; culi filiform. Petiolar node thick. Genital valves rather small and narrow, ‘ing conside ably in the details of their structure in different species. Wings asin - The Pure are not enclosed in cocoons. This genus is cosmopolitan, but most abundantly represented in the und in small craters or under stones and usually form only moderately lous colonies. They are timid, harmless ants of little or no economic npo! . Emery has divided the genus into three subgenera: Preno- epis, sensu stricto, Euprenolepis, and Nylanderia, the last containing the reat majority of the species. om Prenolepis (Nylanderia) longicornis (Latreille) $Stanleyville, 8, 9; Zambi, $, 9 (Lang and Chapin). Numerous specimens showing some variation in color. The forms with paler work- ers might be assigned to Forel’s variety hagemanni, originally described from Boma, in the Belgian Congo, but of the few distinguishing char- acters mentioned by the Swiss myrmecologist, the whiteness of the hairs s noticeable in all the longicornis workers I have seen from various parts the world and the body color varies even in the same colony. These facts and a study of a cotype of hagemanni received from Prof. Forel convince me that the name should be relegated to the synonymy. Forel believed that his hagemanni might be the worker of Emery’s P. wael- broecki, described from female and male specimens, but the females ac- eompanying pale longicornis workers from Stanleyville and Zambi are the same as those accompanying darker workers from other localities in ‘the East Indies and tropical America and do not agree with Emery’s description of the waelbroecki female, which is larger, ferruginous instead 218 Bulletin American Museum of Natural History (Vol. XLV of dark brown, more hairy, with a much broader head, larger eyes and shorter antennal scapes. Prenolepis (Nylanderia) vividula (Nylander) Niapu, 8, 9 (Lang and Chapin). Although this species is being rapidly disseminated by commerce throughout the tropics of both hemispheres and has long been known to occur in northern hothouses, it has not before been recorded from the Ethiopian Region. The workers before me are a little darker than typical specimens, but the differences are too insignificant to justify a new varietal name. is 0 Ss x» 45 © te] 90 105 120 35 160 : ral AUR »—7 ~ —{30] a St 4~ e a a 0 sat » © ° . eRoeies pee a - ry g | f N | eI oe ° 1s » 45 60 5 90 105 120 a5 6 Map 36. Distribution of the genus Pseudolasius. Psgupotasius Hmery Worker small, polymorphic, the head of the major being large and differently shaped from that of the minor., Mandibles well developed, with oblique apical borders furnished with 5 to 6, more rarely with 7 to 8 teeth of different sizes. Clypeus large, convex, and more or less carinate in the middle, its anterior border projecting some- what over the bases of the mandibles. Frontal area indistinct, triangular; frontal carine short, subparallel, rather widely separated; frontal groove indicated. Clypeal and antennary fosse confluent. Head of major worker cordate or subrectangular, deeply emarginate posteriorly; in the minor worker much less deeply concave behind. Eyes small to very small, rarely completely lacking; ocelli absent. Antenna» 12- jointed, inserted near the clypeal suture; funiculi filiform, slightly thickened towards their tips. Thorax short, stout; promesonotal and mesoépinotal sutures distinct; pro- and mesonotum convex above, mesonotum impressed; epinotum short, unarmed, with short base and long sloping declivity. Petiolar scale suberect or inclined forward, its apical border emarginate or entire. Gaster short, elliptical. Legs moderately long and stout. See se a ee oe 1922] — Wheeler, Ants of the Belgian Congo 219 ____ Femaue considerably larger than the worker. Head similar to that of the worker ____ major but broader behind, with well-developed eyes and antenne. Thorax broader than the head, the mesonotum flattened above, the pronotum short and vertical. Wings long and ample, with a single large cubital and no discoidal cell. Mate as small as the worker and of a similar color. Mandibles dentate. Eyes ___ and ocelli large. Antenne 13-jointed; scapes long, funiculi filiform, all their joints __ longer than broad. Thorax similar to that of the female; gaster more slender; ___ external genital appendages rather narrow, hairy. Wings long and broad; venation _ as in the female. _-_-__ Until recently these ants were supposed to be peculiar to the Indo- _ malayan Region, but Forel has described a species from Australia and Santschi has described one from the French Congo (Map 36). Emery! has keyed all the species known up to 1911, but several Indonesian forms have since been described. The African material before me comprises four species, one of which I refer to P. weissi Santschi, the other three being undescribed. Two of the latter were taken by Lang and Chapin in the Belgian Congo, one by Mr. Gowdey in Uganda. All these forms have very poorly developed eyes, compared with the majority of Indo- malayan species. Further search will probably reveal many additional _ species in the Ethiopian Region. The workers are hypogzic or nocturnal and are therefore rarely seen; the males and females, however, are not infrequently taken at lights. _ Pseudolasius weissi Santschi variety sordidus Santschi Text Figure 54 To this variety I refer a major and six minor workers and two partly deilated females taken from the stomachs of toads captured by Lang and Chapin at Akenge. Owing tothe fact that both females were taken from a Bufo polycercus, while the workers were taken from a B. funereus, I cannot be certain that the specimens belong to the same species. The females are of the same size as those of the typical weissi (6.5 mm., the fore wings nearly 7.5mm.). The eyes are elliptical and obliquely placed, but distinctly smaller than indicated in Santschi’s description; the wings a are paler, being rather uniformly brown, with dark brown veins and | - pte ig ¢ ss “s I have figured the head of the worker major and minor. The eyes, as Santschi says, are present only in the former and are very small and slightly elongate. In one of the media I find them reduced to a single ommatidium. The apical border of the petiole is slightly concave in larger, entire in smaller workers. The color seems to be somewhat darker than described by Santschi for his variety sordidus, but this may be due to the action of the gastric juices of the toads. 41911, Ann, Soc. Ent. Belgique, LV, p. 214. i 220 Bulletin American Museum of Natural History (Vol. XLV Pseudolasius bufonum, new species page, Text Figure 55 WORKER MAJOR.— ee Length 2.8 to 3 mm. 7 Head scarcely longer than broad, subrectangular, with nearly straight, ee sides and sinuately excised posterior border. Mandibles 5-toothed, the median tooth | small, the apical long and pointed, the others shorter and subequal. Clypeus convex, subcarinate in the middle, its anterior border entire, only slightly projecting over the bases of the mandibles. Eyes very small, consisting of only three or four ommatidia, situated a little in front of the median transverse diameter of the head. Antennal scapes not reaching to the posterior corners of the head; first funicular —— es b Fig. 54. Pseudolasius weissi variety sordidus Santschi. a, head of worker major; 6, thorax and petiole of same in profile; c, head of worker minor. than the two succeeding joints together; second joint as broad as long, joints 3 to 7 slightly longer than broad. Thorax short, stout; pronotum large and broad, longer than the mesonotum, which is as long as broad; epinotum broader than long. In profile the pro- and mesonotum form a large convexity with rather uneven outline, interrupted by the strong promesonotal suture. Mesoépinotal impression short and ~ not very deep, the stigmata prominent. Epinotum decidedly lower than the mesono- tum, in profile rounded and sloping, with very short base and long sloping declivity. Petiole small, rather strongly compressed antero-posteriorly, with entire superior border. Gaster elongate elliptical. Legs rather stout. Mandibles opaque, very finely and longitudinally striated. Remainder of body a shining, very finely and rather densely punctate, but not more coarsely on the head = and thorax than on the gaster. Clypeus smoother and more shining than the Te- mainder of the head. Hairs and pubescence yellowish, abundant; the former erect, longest on the thoracie dorsum, sparser and shorter on the scapes and legs; pubescence rather long. and dense over the whole body but only slightly obscuring the shining surface. } Yellowish brown; gaster and appendages paler and more yellow; mandibles castaneous, their teeth and a blotch on the vertex blackish. Wheeler, Ants of the Belgian Congo 221 ‘ER MINOR.— igth 2.5 to 3 mm. ) ee ree eres tae Nesad, Wileks is decidedly ir, ( ly longer than broad, with straight sides and only feebly excised Eyes reduced to a single ommatidium or absent. Antennal scapes te te poser sores th rd; first funicular joint broader than long, », pilosit'y, aiid 6olde ‘as ini. the tanjde worker, but the black spot on the oF altogether absent. deg soa Fig. 55. Pseudolasius bufonum, new species. a, head of worker major; 6, thorax and petiole of same in profile; c, head of worker minor; d, head of female. ay Antennal scapes extending nearly one-third their length beyond the erior corners of the head; all the funicular joints longer than broad. Thorax der than the head; the mesonotum and scutellum flattened. Apical border of ol blunt, straight, and transverse Gaster large, elliptical. Wings long. ‘Seu pilosity, and color much as in the worker, but the body darker brown, eile dees the theres, Anterior border of clypeus blackish. Hairs 1g (possibly rubbed off), pubescence shorter and more delicate, and the surface, _ oa lly of the head and thorax, somewhat more opaque than in the worker. Wings a eee remn Feng nnd ptercetigme, i taasth 3.5 cu. Sees 08 brood 90 long, somewhat narrowed behind and in front. Eyes convex, herical, somewhat in front of the middle of the sides, the posterior border nearly f z Mandibles denticulate, overlapping. Clypeus convex. Antennal scapes ending about one-fourth their length beyond the posterior border of the head; he funicular joints distinctly longer than broad, the first nearly as long as the two ling joints together. Thorax and petiole shaped somewhat as in the female. er and legs slender, external genital valves rather long and pointed. 222 Bulletin American Museum of Natural History [Vol. XLV Sculpture and pilosity much as in the worker. Color yellowish brown above, with brownish yellow appendages, genitalia, venter, and anterior portion of head. Ocellar triangle dark brown. Wings paler than in the female. Described from four major and eleven minor workers, three fesidins and eight males, all taken from the stomachs of toads (Bufo a and polycercus) captured at Medje (Lang and Chapin). ; This species differs from weiss? in the shape of the head of the major worker, the slightly larger eyes, more strongly striated and more opaque mandibles, shorter antenne, and much more abundant pilosity and pubescence, and especially in having erect hairson the scapes and legs. The female is smaller than that of weissi, with a differently shaped head, less excised behind, larger’and more nearly circular eyes and longer antenne. Pseudolasius bucculentus, new species Text Figure 56 WorKER MAJOR.— Length 3.2 mm. Head large, as broad as long, broader behind than in front, with convexly inflated sides and front and deeply and angularly excised posterior border, the posterior corners being somewhat conical. Mandibles apparently 5-toothed, folded under the ' clypeus, which is short and in the middle convex and obtusely carinate; its anterior border in the middle with a shallow excision. Eyes very small and indistinct, situated a little in front of the median transverse diameter of the head. Frontal groove rather distinct; frontal carine very short; frontal area transverse, triangular, not impressed. Antenn# rather slender, the scapes not reaching to the posterior corners of the head; first funicular joint as long as the two succeeding joints together; joints 2 to 7 of subequal length, all slightly longer than broad. Thorax robust, pronotum broad, in profile only feebly convex above, the mesonotum rising higher than the pronotum to its middle and then sloping and concave to the mesoépinotal suture. Epinotum with distinet base and declivity, the former short, sloping upward but not reaching the height of the mesonotum, the declivity flat and gradually sloping backward, more than twice as long as the base. Petiole small, with sharp, compressed, very distinctly notched superior border. Gaster voluminous, subelliptical, its anterior segment flattened in front and overlying the petiole. Legs long and stout. Whole body, including the mandibles, shining and very finely and uniformly punctate, except the mandibles, which are longitudinally striate. Pilosity and pubescence yellow, the former short, very sparse, absent except about the mouth and on the thoracic dorsum and as a single row of hairs along the posterior border of each gastric segment. Pubescence short and delicate but very dense, more conspicuous on the head and gaster than on the thorax, very fine and short on the appendages, the latter without erect hairs as in bufonum. Uniformly brownish yellow; mandibular teeth and eyes blackish. WorkER MINOR.— Length 2.2 to 2.5 mm. Differing from the major in the shape of the head, which is distinctly longer than broad, as broad in front as behind, with less convex, subparallel sides and less deeply Wheeler, Ants of the Belgian Congo 223 _ posterior border. Eyes extremely small, reduced to one or two ommatidia, Jess " Seapes extending a short distance beyond the posterior corners of the other respects like the major worker. th 3 mm. Stems ay tent Ds and more swollen behind. Body and wings much paler, brownish he posterior portion of the head dark brown. Wings opaque, grayish, with re ene pentane. The pilosity is also very different, the hairs being d to the mouth-parts and genital appendages. 4 from a single major worker, two mino~ workers, and a 1 at Medje (Lang and Chapin), without further data. This » distinct in the peculiar shape of the head and mesonotum 2 caine the Exteeued notched petiolar border and the very Poo é Fig. 56. Pseudolasius bucculentus, new species. a, head of worker major; 6, thorax and petiole of same in profile; c, head of worker minor. 2.5 mm. Ry Gened es eng, cabrestangiler, as broad in front as behind, with straight sand feebly but distinctly excised posterior border. Eyes absent. Mandibles five acute teeth on their oblique apical borders, the median tooth small, the apical X long as the other three. Clypeus convex but not carinate, its anterior border arly straight. Antenne slender, the scapes extending about one-fifth their length ) 1 the posterior corners of the head; the second funicular joint not longer than jl, the succeeding joints slightly longer than broad. Thorax short and robust, 224 Bulletin American Museum of Natural History (Vol. XLV the pro- and mesonotum forming together an evenly rounded convexity; mesopleure somewhat compressed; epinotum short, nearly horizontal, lower than the mesonotum, passing through a curve into the sloping, flat declivity. Petiolar scale narrowed above, its sides curved, its superior border rather blunt, truncated, entire. Gaster elliptical. Legs rather short. Whole body smooth and shining, except the mandibles, which are opaque and very finely and densely striated. Integument of the body and appendages spparently microscopically but not densely punctate. Hairs and pubescence white, the former sparse, conspicuous only on the clypeus, Be thorax, and gaster, the appendages being without erect hairs. Pubescence short, rather dense on the head and gaster, longer on the latter, slightly oblique on the scapes and legs. os Pale yellow, the head and thorax a little darker, mandibular teeth dark brown. . WORKER MINOR.— ny Length 1.8 to 2 mm. Differing from the major worker in its witli head, which is elongate and with very feeble a ipi isi Antennal seapes reaching occipital excision. nearly one-fourth their length beyond the pos- terior corners of the head; joints 2 to 6 of the funiculus as broad as long. Described from two major and six- teen minor workers taken by Mr. C. C. ; Gowdey at Entebbe, Uganda. They spa’ < bend of parr lagore po por », were found attending subterranean coc- wee r-yoha igi cig same in profile; ¢ gids (Pseudococcus citri Risso) about the roots of coffee. This is readily distinguished from all the preceding species by its smaller size, paler color, the complete absence of eyes even in the major workers, the shape of the head and thorax, and the pilosity. 6 G@copxyiia IF’. Smith Worker medium-sized, slender, slightly polymorphic. Head rather ray broader behind than in front, with rounded sides and posterior corners and semi- circularly excised occipital border, very convex above. Eyes large, convex, broadly elliptic l, situated in front of the middle of the head. Ozelliabsent. Palpi very short, maxillary pair 5-jointed, labial pair 4-jointed. Mandibles long and large, triangular, with nearly straight lateral borders, a very long curved apical tooth and numerous. short denticles along the straight apical border. Clypeus very large and convex, but not distinctly carinate, its anterior border entire or very feebly sinuate in the middle, depressed and projecting over the bases of the mandibles. Frontal area rather large, subtriangular; frontal carinw moderately long, subparallel. Antennz very long, 12- jointed, the scapes inserted some distance from the posterior corners of the elypeus, rather abruptly incrassated at their tips; the first funicular joint very long and slender,. longer than the second and third together, joints 2 to 5 much shorter, subequal, =a. f _ poe ? SS ee ee jena Ants of the Belgian Congo 225 enna : sr, the remaining joints, except the last, shorter and distinctly thicker. Thorax and narrow; pronotum longer than broad, evenly convex above, narrowed and 1 anteriorly; mesonotum anteriorly long and constricted, subcylindrical, y broadened behind where it joins the small, short, unarmed epinotum, which unded and convex above and without distinct base and declivity. Petiole long r : ‘diainder, much longer than broad, subcylindrical, with a very low rounded node seeing end, its ventral surface near the middle more or less convex, its border on each side with a small rounded, projecting lamella, appearing like acute toth when the segment is viewed from above. Gaster short, broadly ellip- ical, its first segment suddenly contracted to the petiole, the tip rather pointed. Legs ry long and slender; claws, pulvilli, and last tarsal joint enlarged. Gizzard with slender sepals, which are not reflected at their anterior ends. Femate much larger than the worker. Head broad, subtriangular; eyes not much larger than in the worker; ocelli well developed Thorax and gaster very broad ed massive, flattened above; thorax nearly as broad as long, pronotum small and i al, overhung by the large depressed mesonotum; epinotum nearly vertical. Reticle short and stout, broader than long, its node low and rounded, more or less impressed in the middle, obliquely truncated or concave behind. Gaster short, nearly _ asbroadaslong. Wings very long and ample, decidedly longer than the body, heavily veined, with a narrow closed radial, a large single cubital, and no discoidal cell. ‘a MALe somewhat smaller than the largest workers. Head small, broader than ‘ with very prominent, hemispherical eyes and moderately large ocelli. Mandibles very small, spatulate, with a few minute denticles. Antenne slender and rather short, 2 ; seapes elongate, their apical halves somewhat abruptly incrassated; first _— funier r joint clavate, enlarged at tip, slender at base; remaining joints much shorter, _ except the last, and slender. Thorax short and massive, the mesonotum broader 5 than the head, very convex and gibbous in front where it overhangs the small mesono- - tum. Petiole and gaster similar to those in the worker, but the former more flattened —_ § a yve and without a node. Genital appendages very small, narrow, linear; legs long _ and slender, tarsal claws obsolete, but pulvilli well-developed. Wings ample, dis- netly’ paler than in the female. Head, thorax and gaster much more pilose than in _ This interesting genus is confined to the Old World tropics and ges over the Indomalayan, Papuan, and Ethiopian Regions, but does -occur in Madagascar (Map 37). It comprises the famous and vicious ee-ants,” or “tailor ants,’ which make peculiar globular or elliptical ts of leaves on living trees. The leaves are spun together with films 5 ably Holland and Green, Wroughton, Rothney, Dodd, Saville Kent, flein, Bugnion, the Sarasin Brothers, Jacobson, Kohl, and myself, have described the extraordinary manner in which the workers use the - young larve as animated shuttles. We According to the majority of myrmecologists, the genus (cophylla 226 Bulletin American Museum of Natural History (Vol. XLV accumulating in my collection for the past twenty years, together with the fine series of specimens taken by Lang and Chapin in the Congo, has convinced me that there are really two distinct species: (2. smarag- dina (Fabricius) of the Indomalayan and Papuan Regions, with the varieties selebensis Emery, gracilior Forel, and gracillima Emery and the subspecies subnitida Emery and virescens (Fabricius); and (2. longinoda (Latreille) of the Ethiopian Region, with the varieties teztor Santschi, rubriceps Forel, annectans, new variety, and fusca Emery. Ern. André described a form brevinodis, from Sierra Leone, as a distinct species, and Stitz has recently cited it from Spanish Guinea, remarking that longinoda is 0 s 30 45 © 75 90 0S - 120 as 160 z a a Kc 2 | Kt A Ame Sy Be “abe ae ao X Ae Swe Le ‘s t : 6 1S | —— bt 40 Ln Be teas sO ry : s / < Pn > oo \ aj ‘ FES se Fane iy , es tp ? we ° 1s %0 45 6 75 90 105 120 135 160 Map 37. Distribution of the genus @cophylla. occurs on the coast, brevinodis in the hinterland, and that there are no transitions between the two. He implies also that brevinodis does not make silken nests like longinoda. The abundant Congo series from vari- ous nests shows, however, without the slightest doubt, that brevinodis is nothing but the worker minima of longinoda (see Fig. 58c), as Emery maintained as long ago as 1886, and the localities of the material before— me show that this species is not confined to the west coastal region. It occurs also in East Africa, Santschi’s variety textor being from Zanzibar. Several authors have cited the true smaragdina from East Africa. Un- fortunately I have little material from that region and what I have is certainly longinoda, presumably belonging to teztor, though this variety seems to me to be poorly characterized and possibly not distinet from 1922) »., Wheeler, Ants of the Belgian Congo 227 : typical form of the species. I am unable to say, therefore, whether . smaragdina actually occurs on the African continent. _ According to Emery, longinoda is the most primitive of the existing as of cophylla, because most closely allied to @. sicula, which he ibed from the Miocene amber of Sicily. In the Baltic amber I have sd two species of the genus, (2. brischkei Mayr and brevinodis heeler. As the latter name is preoccupied by brevinodis André, which basec , as T have shown, on the minima worker of longinoda, I suggest a petiole both of the Baltic amber forms, being of Lovee Oligocene ge and therefore older than sicula, are also more like longinoda, and cophylia longinoda (Latreille) Plate XX, Figures 1 and 2; Text Figures 58 and 59 _ Faradje, 8, 2, &%; Malela, 8; San Antonio, 8 (Lang and Chapin); _ The Eollowing flerences between this species and smaragdina may e noted. In the worker the polymorphism is greater, for not only do the individuals of the same colony show a greater range in size (from 3 to mm.) but the minimz differ more from the mediz# and maxime in the shape of the thorax and petiole. The head of the, worker longinoda is _ distinctly more triangular than that of smaragdina, being broader behind, with less convex sides; the eyes are distinctly larger, the mandibles peer, the clypeus more nearly subcarinate behind, its anterior border : - ) es feebly and sinuately emarginate in the middle, the pronotum “ ae the petiole decidedly stouter, more thickened behind, with ye stigmata much less prominent when the segment is viewed from abc »ve and its ventral surface much more convex anteriorly on the ventral ide, when viewed in profile. The sculpture, pilosity, and color are very : a in the two species, but in longinoda the integument is more decidedly opaque, the mandibles are somewhat more coarsely striated, ‘ ilways darker, being concolorous with the posterior portion of the head, at least i in the large workers and especially in the dark varieties. The tr » furrow on the second and succeeding gastric segments just ehi 4 the anterior border is more pronounced in longinoda. i is fernale of this species measures 12 to 14 mm. (wings 16 mm.) and _ is, therefore, distinctly smaller than the corresponding sex of smaragdina, 3 eam 15 to 17 mm. (wings 18 to 19 mm.). The body of the —_— species is much more opaque throughout, the wing-veins more ball 228 Bulletin American Museum of Natural History (Vol. XLV heavily bordered with dark brown, and the transverse bands at the bases of the second and following gastric segments are broader, darker, and more sharply marked off from the remainder of the segments. The green portions of the typical longinoda female are slightly more olivaceous and less pea-green, and the basal bands of the gaster are more exposed and brownish; the appendages are more brownish. | The inale longinoda is scarcely smaller than that of smaragdina ani . measures 6 to 6.5 mm., but the head, thorax, and petiole are darker and Fig. 58. (€£cophylla longinoda (Latreille): a, body of worker major in profile; b, head of same; c, body of worker minima in profile; d, head of same. more blackish; the head is decidedly broader, especially behind, the mandibles, petiole, antennal scapes, and wings are decidedly shorter and the integument is less shining. The workers of the various subspecies and varieties of the two species may be separated by means of the following key. 1. Petiole very slender, its stigmata seen from above very prominent, its ventral = surface nearly straight or very feebly convex in profile (smaragdina)... .2. Petiole stouter and higher, its stigmata seen from above not prominent, its ventral surface strongly convex in profile (longinoda).................. 7. 2. Body ferruginous or testaceous... ......-.+is- 006+ cle sess eens mus oe 3. Gaster and sometimes the head pea-green, head more rounded and less truncated behind; size smaller, petiole somewhat shorter (Queensland, New Guinea, Islands Aru and Key)) isi “Via ia subspecies virescens (Fabricius). ous (India, Ceylon, Cochin China, Indonesia). : Se smaragdina (typical). | more testaceous, seenotae and petiole a little narrower (Java). variety gracilior Forel. — slightly shining (Papua, Philippines, Melanesia). subspecies subnitida oe | dull red, Sdsstia clless taioiaials (Bolgiasi Congo)’. Seabees valicscis (Forel). 1 and gaster black or dark brown (Belgian Congo, Nigeria, Liberia, ‘Cameroon, a (SONS Sal ee a variety fusca (Emery). ps, which shows some color variation in the direction of the typical noda. The discovery of another variety, annectens described below, ecting rubriceps and longinoda is additional evidence that fusca ot be maintained as a species. In my opinion it is merely an extreme ¢ variety, for I am unable to detect in it any morphological char- of even subspecific value. All of the varieties of longinoda are ‘ polymorphic in the worker caste and the smallest individuals all with the description of André’s brevinodis, except in color. Eehe ethological observations of Chun! and Father Kohl? refer to this ‘uaa Mr. Lang’s photographs reproduced on Pl. XX, figs. 1 and 2, show t Sl the nests of the typical longinoda from Malela, consisting of the flets of a bush skillfully folded and united with the white silk spun by young larve. He found that the nests of longinoda and its varieties most often constructed on bushes and are sometimes only a few feet £1903, Tisten des Weltmoares, I Lp 71906, ‘Zur der spinnenden Am ~LA a Offenbarung, LI, pp. 166-169. 230 Bulletin American Museum of Natural History (Vol. XLV Fig. 59. Nest of (cophylla longinoda (Latreille) at Avakubi, October 27, 1909. This nest, 16 cm. long, was placed about four feet from the ground in one of the coffee trees of a deserted plantation. Photograph by H. Lang from the ground. Text Fig. 59 shows a nest of this ant placed in a coffee tree at Avakubi. The habits seem to be the same in all essential particu- lars as those of smaragdina. Ccophylla longinoda variety annectens, new variety WorKER very similar to the typical form but brown instead of ferruginous, the gaster sometimes slightly darker than the remainder of the body. Mandibles, except in the small workers, darker brown than the front, cheeks, and clypeus. Incrassated tips of antennal scapes with a dark brown spot; funiculi, knees, tarsi, and tips of tibiw pale yellow; pulvilli black. FEMALE brown, instead of green and brown like the typical longinoda, with darker brown markings on the thorax. Second and following gastric segments with the basal bands velvety black, so that the gaster is distinctly fasciate. Funiculi, tips of scapes, tibiee, tarsi, and vertex paler, more reddish brown. Wings slightly darker than in the typical form, with deeper brown margins to the veins. Wheeler, Ants of the Belgian Congo 231 than the worker. Mandibles, antennz, tarsi, and articula- prowniah yellow; last tarsal joint black. Wings distinctly paler than in ss 1 from long series of specimens from the following places: “eye locality), 8 2, &; Stanleyville, 8; Niangara, 8 (Lang pin); Malela, & (J. Bequaert). _ @cophylla longinoda variety rubriceps (Forel) eR black or dark brown, the head dull, blood red, often darker laterally and y, tips of antennal funiculi and second to fourth tarsal joints pale brownish . Gaster in specimens from some colonies brown, the posterior margins of the paler. ; dark brown, almost black, the gaster very little paler, the bands at the e segments velvety black; tarsi and tips of funiculi pale brown. Wings d from many specimens from two colonies taken at Stanley- le aie: iiiid Chapin). The workers of one colony agree closely with rel’s description of the types from the Belgian Congo in having the x nearly or quite concolorous with the thorax, and some of the specimens are scarcely distinguishable from the variety fusca; the of the other colony have the gaster rather pale brown and, _ therefore, connect the variety with annectens, which seems to be a more Sees cnn erictpe. cophylla longinoda variety fusca (Emery) b Worker differing from rubriceps only in having the head entirely black or dark brown, though sometimes with a reddish tinge above. Mandibles black, with dark brown teeth. Large workers have the clypeal border very feebly sinuate in the middle and the surface just behind it with a faint longitudinal impression. The smallest _ workers are a little paler, with paler mandibles, but in the structure of the thorax and _ petiole precisely like the corresponding phase of the other forms of the species. Fema.e like that of rubriceps, but perhaps a shade darker. Matz indistinguishable from the male of rubriceps, except that the erect white hairs on the dorsal surface of the head, thorax, and gaster are distinctly longer and ; g * & Redesoribed from specimens taken at Stanleyville and Garamba : Diane and Chapin). There is also a worker of this variety from Mon- 4 Tovia, Liberia, (J. Morris) in my collection. 232 Bulletin American Museum of Natural History (Vol. XLV . Camponotus Mayr Worker medium-sized to very large, polymorphic, rarely dimorphic, the worker maxima having a large, broad head, the minima a much smaller head and more slender body, the media being intermediate in structure. Head differing considerably in form in different species, usually broad and more or less excised behind, narrower in front, very convex above and flattened beneath. Mandibles powerful, short, triangular, with coarse teeth on their broad apical borders; external border and upper surface convex in large individuals. Palpi moderately long, the maxillary pair 6-, the labial pair 4-jointed. Clypeus large, trapezoidal or subrectangular, usually carinate or sub- carinate, often divided into a large, median, subhexagonal and two small, triangular, lateral divisions, which do not reach the lateral border of the cheeks, the anterior border entire or emarginate, often excised on each side, with a broad, more or less projecting median lobe. Frontal area small, triangular or lozenge-shaped; frontal groove distinct; frontal carinew long, prominent, marginate, and sinuate or S-shaped, rising from the posterior border of the clypeus. Eyes moderately large, broadly elliptical, not very prominent, situated behind the middle of the head; ocelli absent, the anterior ocellus sometimes indicated. Antenne 12-jointed; scapes thickened distally, inserted some distance behind the posterior border of the clypeus; funiculi long, filiform, not enlarged at their tips, all the joints longer than broad. Thorax differing greatly in shape in the various species, typically broadly and more or less evenly arcuate in profile, broad in front, laterally compressed behind, the epino- tum usually simple and unarmed. Rarely the mesonotum is impressed or sellate. Petiole surmounted by an erect scale, the upper border of which may be blunt or anteroposteriorly compressed, entire, subacuminate or more or less emarginate. Gaster rather large, broadly elliptical, its first segment forming less than half its surface. Legs long and well developed. Gizzard with a long slender calyx, — of which afe not reflected at their anterior ends. Fema.e larger than the worker maxima but usually with smaller head. The latter and the petiole much as in the worker. Ocelli present. Thorax elongate ellip- tical; pronotum short, its posterior margin arched, its posterior angles reaching back to the insertions of the wings, mesonotum and scutellum long, convex; metanotum depressed below the scutellum. Gaster elongate elliptical, massive. Wings long and ample, the anterior pair with a radial, one cubital, and no discoidal cell. Mate small and slender; head small, with very prominent eyes and ocelli. Mandibles small and narrow. Antenne 13-jointed, slender, scapes long. Petiolar node thick and blunt; gaster elongate, with small slender genital appendages. Legs very slender. Wing venation as in the female. Pur nearly always enclosed in cocoons. This huge cosmopolitan genus, comprising more than 1000 deseribed forms, has become so unmanageable that Forel and Emery have recently — _ split it up into some thirty-six subgenera. The frequent occurrence of species of Camponotus in all countries, except Great Britain and New Zealand, and the extraordinary variability of many of the species in re- sponse to slight differences of environment make the genus one of con- siderable interest to the student of geographical distribution. In the Ethiopian Region, it is represented by numerous species assignable to no an oe Wheeler, Ants of the Belgian Congo 233 _ a ss than eleven of the thirty-six subgenera recognized by Emery and ew: Myrmoturba, Dinomyrmex (Map 41), Myrmosericus, rr) (one species, probably introduced), Orthonotomyrmez, a (Map 38), Myrmopiromis, Myrmorhachis, Myrmopsamma, mn blys, and Colobopsis, and species of six others, Camponotus, ai , Myrmosaulus, Myrmosaga, Mayria, cc intenuen and a ‘ys | < | NX — i. er 35" be 1) | ea | 25 a = rs Map 38. Distribution of Myrmotrema, a subgenus of Camponotus of the Ethiopian and Malagasy ions. According to Emery (1920) one species occurs in India. Myrmopytia, occur in the Malagasy Region. A few of these subgenera, ysamma and Myrmopiromis, are peculiarly African, while others, Myrmosaga, Mayria, Myrmonesites, and Myrmopytia, are only found in Wed development of the subgenus Myrmoturba and of the species maculatus (Fabricius), the typical form of which s West African, is extraordinary, as will be seen by consulting the cata- ogue (Part VIII). C. (Myrmoturba) maculatus (Map 39) and two other C. (Myrmosericus) rufoglaucus a 42) and C. (Orthonotomyr- 9 — 234 Bulletin American Museum of Natural History (Vol. XLV occur in all the continents; rufoglaucus, with many varieties, ranges from southern China across India and equatorial and South Africa to the Gulf of Guinea; and sericeus occupies a similar range, though showing little tendency to produce subspecies and varieties. The species of Camponotus often form very populous colonies and * exhibit a great diversity of nesting habits. Many live in the ground, either under stones or in crater nests, others under bark, in dead wood, hollow twigs, and galls, and a few construct carton nests or employ their larve, after the manner of @cophylla, in spinning together particles of vegetable detritus with silk (C. senex and formiciformis). The foodofthe various species consists of miscellaneous insects, the excreta of aphids (honeydew), and nectar. Many of the smaller forms are stolid, apathetic, or timid, but the maxima workers of the large species belonging to the subgenera Dinomyrmex, Myrmoturba, Myrmothriz, and Myrmo- . piromis are very pugnacious and capable of inflicting painful wounds with their powerful mandibles. Camponotus (Myrmoturba) maculatus (Fabricius) Medje, 2, 8, 2; Yakuluku, 2%, 8; Garamba, a, 8; Vankerck- hovenville, 8; Faradje, @ (Lang and Chapin). Six of the workers from Garamba, all minors, were taken from the stomach of a toad (Bufo regularis). The major workers agree perfectly with Donisthorpe’s re- description! of the Fabrician type of this ant in the Banks Collection, presumably from Sierra Leone, except that they ie a few short, erect hairs on the gular surface of the head. The distribution of C. maculatus and its various forall is shown on Map 39. Camponotus (Myrmoturba) maculatus subspecies guttatus Emery I refer fourteen minor workers from Zambi (Lang, Chapin, and J. Bequaert) to this pale subspecies. The specimens were taken “only at night-fall, visiting the tables in the camp. They are shy and fast runners.” Camponotus (Myrmoturba) maculatus subspecies melanocnemis (Santschi) Faradje, 2, 8; Yakuluku, 9 (Lang and Chapin). Numerous specimens from several colonies. 11915, Ent. Record, XXVII, p. 221. * aa SS ee See - 1922] Wheeler, Ants of the Belgian Congo 235 a F tienes) maculatus subspecies congolensis Emery ~ Yakuluku, a, 8; Faradje, 2, 8; Medje, 8; Niangara, 8, @; ra a, 8, @ (Lang and Chapin). Numerous specimens. A r and two minor workers from Faradje are from the stomach of a ene occipitalis), one of the major workers from Garamba from the = of a toad (Bufo regularis). a Te Tl b s Bll la BAN CPT bl “s ; , RA °s > 4 4 AN. ee : ‘ : an a as 7 1% ; 2 am 4 Bs Ap ? 2 1 _—— i | ~ == ee |) a a ae a. ee ee — Je Map 39. Distribution of > (Myrmoturba) maculatus (Fabricius) and its forms. - Camponotus (Myrmoturba) sisi subspecies miserabilis Santschi 4 variety pessimus, new variety os _ ‘The major worker measures only 6 to 6.5 mm., the minor 5 to 5.5 mm. Both _ agree closely with Santschi’s description and figure of miserabilis, except in their con- - siderably smaller size. The head of the major is distinctly narrower anteriorly, the cheeks being less convex and the frontal carine are less approximated. Sculpture, a and color very much as in miserabilis. _ Four major and five minor workers from Yakuluku (Lang and in), without further data. 4 _ Camponotus (Myrmoturba) maculatus subspecies solon Forel _Bafwabaka, 2 8; Niangara, a; Akenge, 2; Medje, a (Lang and Chapin). All the specimens from the three former localities, twenty in number, were taken from the stomachs of toads (Bufo regularis, _ funereus, and polycercus), the single specimen from Medje from the stomach of a frog (Rana albolabris). 236 Bulletin American Museum of Natural History (Vol. XLV Camponotus (Myrmoturba) maculatus subspecies solon variety jugurtha, new variety Worker MAXxIMA.—Differing from the typical solon in its much paler color, the antennsw, head, and thorax being red; the mandibles, front, and a streak down the middle of the clypeus castaneous; the posterior corners of the head, the legs inelud-_ ing the coxw, the petiole, and the three basal gastric segments brownish yellow; the tip of the gaster more brownish. The mandibles are very finely striated and the petiolar scale is much compressed and prolonged above as in the typical solon and not blunt as in brutus. In the feebler punctuation of the head this variety is also like the typical solon. A single specimen from Batama (Lang and Chapin), without further data. Camponotus (Myrmoturba) maculatus subspecies brutus (Forel) Avakubi, 2, 8; Medje, a, 8; Faradje, a, 8, 2; Bafwasende, 4, 8; Stanleyville, 2, 8, 2; Batama, 9; Lukolela, $; Malela, a, 8; Isangi, 2,8; Nouvelle Anvers, 2,8; Zambi, 8; Poko, a; Akenge, a, 8; Niangara, a, 8 (Lang and Chapin); Malela, 2, 8 (J. Bequaert). The workers from Akenge and Niangara, ten in number, are from the stomachs of toads (Bufo funereus, polycercus, and regularis). To judge from the many series of specimens, this large red ant must be very com- mon in the Congo. Its native name, according to Mr. Lang, is ‘‘maola.” It nests in rotten wood. The specimens from Nouvelle Anvers were found nesting in an old oil palm trunk. Camponotus (Myrmoturba) maculatus subspecies brutus bess lycurgus Emery Two major and four minor workers, taken at Leopoldville (Lang and Chapin), may be referred to this variety, which has the dark head and thorax of the typical subspecies solon. Camponotus (Myrmoturba) acvapimensis Mayr Faradje, 2, 8; Garamba, 2, 8; Bolengi, near Coquilhatville, 2, 8; Stanleyville, 8; Thysville, 8; Vankerckhovenville, 4, 8; Niangara, 4, 8; Akenge, 2, 8 (Lang and Chapin); Zambi, 2, 8; Thysville, 8 (J. Bequaert). Of the numerous specimens of this small black species, thirty from Garamba and Niangara are from the stomachs of toads — (Bufo regularis and funereus). A single major worker from Faradje is from the stomach of a frog (Rana occipitalis). The specimens from Bolengi were found nesting in the trunk of an oil-palm; some of those from Faradje were captured while attending plant lice on the young leaves of orange trees. The distribution of this species is shown on Map 40. Wheeler, Ants of the Belgian Congo 237 Se Map 40. Distribution of Camponotus (Myrmoturba) acrapimensis Mayr. Camponotus (Myrmoturba) maguassa, new species _ Mason workKER.— Length 9 to 10 mm. ‘Head rather small, subrectangular, as long (1.3 mm. without the mandibles) _ a8 broad, a little narrower in front than behind, with straight posterior and very _ feebly convex lateral borders. Eyes rather large and convex, situated about their length from the posterior border when the head is seen from the front. Mandibles moderately convex, coarsely 6-toothed. Clypeus sharply carinate behind, rather _ deeply emarginate on each side of the median lobe, which is short, with straight _ border, distinctly dentate at the corners. Frontal area subtriangular, indistinct ‘behind; frontal groove pronounced; frontal carine approximated anteriorly. An- _ tenn slender, the scapes straight, terete, not enlarged at the tips, reaching about two- 3 _ fifths their length beyond the posterior border of the head. Pronotum flattened above, _ its sides distinctly marginate anteriorly; mesonotum evenly arched in profile; _ metanotum indistinct; epinotum with subequal base and declivity, both surfaces 4 straight and sloping, meeting at a rounded obtuse angle. Petiole rather high, oval __ when seen from behind, in profile with flattened anterior and posterior surfaces, its : aumerior border rather charp and entire. Gaster and legs as usual, hind tibia nearly __ eylindrical, only very slightly compressed, without a row of bristles along their flexor 238 Bulletin American Museum of Natural History (Vol. XLV Body subopaque, the petiole, gaster and legs more shining. Mandibles coarsely and sparsely punctate, their tips striated, their bases sharply shagreened. Head very densely, evenly and finely punctate, eo that it appears granular; the clypeus, cheeks, front, and vertex also with large, scattered, irregular, piligerous punctures. Sculp- ture of the thorax like that of the head but finer, especially on the pleure; the dorsal surface with coarse, sparse, piligerous punctures, Gaster finely, sharply and trans- versely shagreened, with coarse, sparse, transverse piligerous punctures. These have minutely papillate anterior borders so that the coarse hairs seem to rise from small projections. Legs finely shagreened or coriaceous. Hairs fulvous red, coarse, erect, rather abundant, long on the dorsal surface of the head, thorax, and gaster, somewhat shorter on the gula and petiolar border, still shorter but suberect on the cheeks, seapes and legs. Pleurw, anterior and posterior surfaces of petiole hairless. Pubescence rather coarse, very sparse, visible on the cheeks and gaster. Brownish black; funiculi, tips of scapes, legs, including the coxe, petiole, and gaster rich castaneous, the legs and funiculi slightly paler. WORKER MINOR.— Length 5 to 7.5 mm. Differing from the major worker in the shape of the head, which is longer than broad, with straight, parallel sides and broadly convex posterior border. The eyes are more convex, the antennal scapes longer, extending somewhat more than half their length beyond the posterior corners of the head. The clypeal lobe has more rounded corners. Described from numerous specimens from two colonies taken at Avakubi (Lang and Chapin). According to a note accompanying one lot, ‘‘these ants are said to be common in the forest in the decayed wood of large trees. Native name ‘maguassa.’”’ This species bears a striking resemblance to C. festai Emery from Asia Minor. The single worker major cotype of this insect in my collee- tion lacks the head, so that in making comparisons of this part of the body I have to rely entirely on Emery’s description. The head of the worker major of festai is evidently larger (3.53.5 mm.), more narrowed in front, with the postefior border slightly concave; the mandibles are 7-toothed, the scape is somewhat flattened, the declivity of the epinotum much shorter than the base, the petiole much broader above, with sharper border; the hind tibie are prismatic, with dorsal groove and their flexor border has a row of bristles; the hairs and pubescence are yellow, the latter much longer and more conspicuous on the gaster than in maguassa, and the hairs on the legs are distinctly longer; the head and gaster are black, the thorax, legs, and petiole deep brownish red. oe — = = —-1922) | Wheeler, Ants of the Belgian Congo 239 ——— 3 - Camponotus (Dinomyrmex) pompeius Forel subspecies cassius, new subspecies Text Figures 60 and 61 iiniek MAXIMA.—Differing from the maxima of the typical pompeius in having a the head distinctly smoother, more shining, and more superficially shagreened, the apical tooth of the mandibles much longer, the corners of the clypeal lobe much more acute, the superior border of the petiole somewhat more obtuse, the petiole and thorax Speeeeeem ted, except the pronotum and dorsum of the mesonotum, which are dark brown. The thorax and coxe are covered with much longer, denser, and more con- -_spicuous yellowish pubescence than in typical pompeius. Worker mintma.—Very similar to the typical form but the thorax and legs Sart and the head and thorax with longer pubescence. _ Described from a single maxima and seven minime from Yakuluku (Lang and Chapin). There is also a single mermithergate from Medje _ which I have figured (Fig. 61). It is 15 mm. long, the gaster measures -$ mm. and is enormously distended with nematode worms of the genus ___ Mermis, which are visible through the thinner portions of the lateral and _ ventral integument. The head and thorax are like the corresponding _ parts of the minima or small media and there are no traces of ocelli. ___ The petiole, however, has a somewhat more pointed node and therefore approaches slightly the condition in the female. _ Four males from Medje and Faradje and three females from Stan- leyville are probably referable to this or to one of the other forms of pompeius. They have the epinotum and legs more reddish than in the _ typicalform. The wings of both females and males are slightly yellowish, with resin-colored veins and dark brown pterostigma. Camponotus (Dinomyrmex)'pompeius subspecies marius Emery Medje, 2, 8; Akenge, 2, 8; Niapu, 8 (Lang and Chapin). Two __ maxima and twenty-nine minima workers all from the stomachs of toads (Bufo polycercus, funereus, and superciliaris) and one small worker from _ Niapu from the stomach of a frog (Xenopus tropicalis) seem to belong to this form. Though from different localities, the two maxime both have the head much smaller and narrower (without the mandibles, 4.5X 3.9 mm.) than in the typical pompeius or the preceding subspecies and _ agree very closely with Emery’s description. He believed that the speci- ; men he examined was not a maxima, but the two specimens from Medje _ and Akenge seem to indicate that the small narrow head may be char- _ acteristic of the largest worker of the subspecies. The petiolar scale in _ my specimens is also high and pointed, precisely as in Emery’s figure, the seapes are long (4.5 mm.), and the coloration and sculpture agree with his description. 240 Bulletin American Museum of Natural History {Vol. XLV. Fig. 60. Camponotus (Dinomyrmex) pompeius subspecies cassius, new subspecies. a, head of worker maxima; b, thorax and petiole of same in profile; c, head of worker minima. Fig.61. Camponotus (Dinomyrmex) pompeius subspecies cassius, new subspecies. Mermithergate from Medje. _ Wheeler, Ants of the Belgian Congo 241 aa — ; 16 Camponotus (Dinomyrmerx) langi, new species a Text Figure 62 ~ Worker maxima.— Length 12.5 to 14 mm. ie: Fe Head unusually small, decidedly longer than broad (4.1 X 3 mm., without the mandibles), slightly narrower in front than behind, with broadly ed wok deeply y excised posterior border and evenly, feebly convex sides. Eyes rather small and situated twice their length from the posterior border of the head. Mandibles a rather small, convex, with 7 short teeth. Clypeus carinate, its anterior border __ emarginate on each side, the median lobe very short, its border coarsely crenulate, z eet aaen cbstuse, Frontal area small, subtriangular; frontal carine closely approxi- __ mated, especially in front. Antenne long (4.5 mm.) and slender, not enlarged distally, their bases distinctly flattened but not dilated, reaching nearly half their length beyond the posterior border of the head; funiculi long, filiform. Thorax low and ‘narrow; metanotum distinct; epinotum long, its base nearly four times the length of the deelivity, with a distinct, transverse impression in the middle. Petiole very low, _ subquadrate, and as broad as long when seen from above, in profile scarcely higher _ than long, obliquely truncated anteriorly and posteriorly, with very blunt superior border. Gaster long and narrow. Legs very long and thin; tibie triangular in cross- ___ seetion, deeply channelled on all three surfaces, their flexor borders without row of bristles. ____ Mandibles, clypeus, legs, sides of thorax, and sides and venter of gaster somewhat shining, remainder of the body opaque. Mandibles more opaque at the base, where _ they are densely shagreened, smooth and coarsely punctate in the middle, coarsely _ striated towards the tip. Clypeus, head, and thorax very densely shagreened, the _ head more distinctly; clypeus, cheeks, and sides of head with small, scattered 5 a ress Gaster very finely and transversely shagreened,with a acme sind chart more abundant on @ the gula and top of the head, very short, sparse, and appressed on the appendages. _ Sides of head with short, sparse, stiff hairs. Pubescence very dilute, distinct on the ) gaster and all parts of the head, longest on the gula. a Head and gaster deep castaneous; mandibles dark red, with black borders; _ ¢lypeus and adjacent portions of cheeks often reddish; tips and insertions of antennal seapes, palpi, thorax, petiole, trochanters, and femora dull brownish yellow; upper _ surface of pronotum, mesonotum, and base of epinotum dark brown with paler _ sutures; tibia, femora, and tarsi dark brown, the latter somewhat paler at their _ tips; posterior borders of gastric segments rather broadly yellowish and shining. Worker MiInIMmA.— ~ Length 11 to 12 mm. Head very long (3.4 mm., without the mandibles) compared with its width _ {19 mm.), the portion in front of the eyes nearly as broad as long, a little broader in _ front, with straight sides; behind the eyes it narrows rapidly into a neck with concave __ 8ides, the occipital border being somewhat less than one-third of the anterior border. _ Eyes prominent, situated more than twice their length from the occipital border. _ Clypeus resembling that of the maxima. Antenn» longer, the scapes not flattened, straight, reaching fully three-fifths of their length beyond the occipital border. Thorax and petiole as in the maxima but lower, and the transverse impression on the base of the epinotum scarcely indicated. 242 Bulletin American Museum of Natural History (Vol. XLV Sculpture much finer, pilosity and pubescence even sparser than in the maxima. Color paler; clypeus, cheeks, funiculi, petiole, ventral portions of thorax, coxw, and femora yellow; mandibles, scapes, posterior portion of head, tibie, and dorsal surface of thorax and gaster brown. Fig. 62. Camponotus (Dinomyrmer) langi, new species. a, head of worker maxima; 6, thorax and petiole of same in profile; c, head of worker minima. Femae (deilated).— Length 21 mm. Head large, slightly longer than broad, broader behind than in front, with straight sides and feebly and broadly excised posterior border. Mandibles more con- vex than in the maxima, clypeus very similar. Antennal scapes very slightly flattened _ at the base, extending nearly one-third their length beyond the posterior corners of the head. Thorax through the wing-insertions not broader than the head; mesono- tum as long as broad. Petiole much higher than in the worker, nearly twice as high as long, elliptical from behind, its anterior and posterior surfaces convex, its border narrowed above and slightly notched in the middle, in profile rather acute. Mandibles shining, coarsely punctate, their bases opaque. Head and body more shining than in the maxima, but similarly sculptured. Pilosity like that of the maxima but the pubescence very long and abundant on the prosterna, fore coxs, and lower portions of the metapleure; as long but sparser on the gula and posterior surfaces of the head; short on the scapes, but longer and oblique towards their tips. Tibia and tarsi with short, stiff, oblique hairs. 4929) Wheeler, Ants of the Belgian Congo 243 as Head black; mandibles, sutures of thorax, upper portions of mesopleure, and pro- and mesonotum, scutellum, and gaster castaneous; remainder of thorax, petiole, middle and hind coxm, and trochanters yellowish red. Legs castaneous, tips of tarsi = Mars.— - Length 13 mm. Head twice as long as broad, the portion in front of the eyes long, with subparallel, slightly concave cheeks, the posterior portion rapidly narrowed to the occiput, the : ae sides and occipital border nearly straight. Eyes convex, at the middle of the sides of the head. Mandibles spatulate, bluntly pointed, edentate but with overlapping tips. earinate, without an anterior lobe, its border broadly rounded. Antennz ; _ very long and slender. Thorax and gaster long and narrow; epinotum elongate, evenly convex, sloping, without distinct base and declivity. Petiole much as in the worker minima. Legs very long. 1 Mandibles, head, thorax, and legs rather opaque; epinotum, petiole, and gaster shining, punctuation feeble and inconspicuous. Hairs yellow, short, and sparse as in the worker minima. _ Brownish yellow; head, mesonotum, scutellum, tibie, and tarsi brown; _ mandibles darker. Wings distinctly yellow, with yellowish brown veins and dark brown pterostigma. , Described from forty-one workers from Faradje (type locality), a female and worker minima from Garamba, and two males from Faradje (Lang and Chapin). The following note accompanies the specimens from Faradje: “‘These long-legged ants are very fond of sugar or any- thing sweet, such as fruits, etc. They are seldom seen during the day- time. The colony had made its nest between boxes that were piled up on the verandah of a house, and the ants were assembled in a hollow space about half an inch wide. A few fibrous particles of detritus were used in the construction of the nest.’”’ There are no data accompanying the two specimens from Garamba, so that I am not certain that the female is cospecific with the worker. C. langi is very peculiar in the small, narrow head of the maxima and the long neck-like occipital region of the minima. There can be no doubt that what I have described as the maxima is really the largest worker _ form. Fifteen specimens of the series all agree in the shape and size of J _ the head as represented in the figure; the remaining specimens are all _ minime. Mediz, apparently, do not exist. Camponotus (Dinomyrmex) cesar Forel A single imperfect worker minima from the stomach of a frog (Rana _ occipitalis) taken at Faradje (Lang and Chapin) seems to belong to this light-colored species. 244 Bulletin American Museum of Natural History [Vol. XLV Camponotus (Dinomyrmex) c#sar subspecies imperator Emery A single media from Isangi (Lang and Chapin), without further data. Bi Camponotus (Dinomyrmex) massinissa, new species FeMALE.— Length nearly 21 mm. Head as broad as long (4.5 mm., without the mandibles), subrectangular, slightly broader behind than in front, with straight sides, feebly but broadly concave posterior border, and rather acute posterior corners. Mandibles large and convex, with 6 flattened, acuminate teeth, the apical tooth very long and broad at the base. Clypeus carinate only at the base, its anterior border emarginate on each side, the median lobe piste TULL PTL aaa am i} BL aL let ‘ Pah haw { site ees < F . z eK ah S46 SEY kK Yeu tJ Saf nw ‘ y ~ So a Bh (a . q :t . hs a aes ot | 4. Map 41. Distribution of the subgenus Dinomyrmez of Camponotus. indistinct, somewhat crenate, without pronounced corners, with a small notch in the middle. Frontal area small, subtriangular, impressed. Frontal carine closely approximated. Antenne long; scapes measuring 6.2 mm., extending half their length beyond the posterior corners of the head, not flattened at the base nor enlarged distally. Thorax robust, through the wing-insertions broader than the head. Mesonotum broader than long, with a narrow, shining, median, longitudinal groove on its anterior half. Epinotum sloping, evenly convex. Petiole higher than long, rather pointed _ above, its anterior surface made of two subequal planes which meet at a very blunt — obtuse angle when seen in profile; the posterior surface flat, the superior border blunt. Hind tibize and metatarsi distinctly flattened and channelled, the flexor border of the former with a row of strong bristles, except on their basal fourth. Mandibles shining, coarsely punctate, striate near their apical borders, opaque, finely shagreened and less coarrely punctate at the base. Head and clypeus nearly opaque, densely and finely punctate, with coarser, sparse, piligerous punctures over the whole surface. Thorax and petiole with similar sculpture, but the piligerous 1922] Wheeler, Ants of the Belgian Congo 245 ———— ‘punctures less pronounced. Gaster more shining, densely, coarsely and transversely shagreened, with coarse, scattered, transverse, piligerous punctures. Antennal _ seapes and legs shining, rather strongly and unevenly punctate. Hairs fulvous, coarse, erect or suberect, long and abundant, especially on the head, gula, dorsal portion of the pronotum, mesonotum, epinotum, and fore cox. Antennal scapes also with long erect hairs; those on the tibie stiffer, much shorte and more oblique. _ Black; mandibles except their bases and teeth, deep red; insertions of antenne, funiculi beyond the tip of the first joint, thoracic articulations, trochanters, and tips of cox yellowish; gaster and legs castaneous. Wings heavily infuscated, blackish, with dark brown veins. is ° 5 30 45 6 75 90 105 120 135 160 c ¢ ae -- Sos tc mAs a ay ad et SLES Oh {(Ts_ ts mel € # :. x - d od / is x — A ‘5 Ay — = / as Leaver * bw jo : - - $ r) i f . a andi, “= . ae : | (fs rt IS om 182 VL hs ts iS 0 1S ) 45 60 15 90 105 120 135 ree Map 42. Distribution of Camponotus (Myrmosericus) rufoglaucus (Jerdon). A single specimen from Medje (Lang and Chapin), without further data. This female is so easily recognized and so peculiar in its characters that I do not hesitate to describe it as new. It certainly does not be- long to any of the workers in the collection and I am unable to regard it as the female of any of the Ethiopian species of Dinomyrmezx that have been described from workers only. Camponotus (Dinomyrmex) wellmani lore! variety rufipartis Forel Stanleyville, 8, 7; Niangara, $; Faradje, 8; Ngayu, 8 (Lang and Chapin). The specimens agree very closely with Forel’s description. Two workers from Ngayu were taken from the stomachs of toads (Bufo superciliaris and funereus) and one from Faradje from the stomach of a frog (Rana occipitalis). 246 Bulletin American Museum of Natural History (Vol. XLV Camponotus (Myrmosericus) rufoglaucus (Jerdon) subspecies cinetellus (Gerstecker) Five workers from Zambi (J. Bequaert). The distribution of C. rufoglaucus and its various forms is shown on Map 42. Camponotus (Myrmosericus) rufoglaucus subspecies cinctellus variety rufigenis Forel Faradje, 8; Niangara, 9; Garamba, $; Stanleyville, $; Medje, 8; Poko, 8; Akenge, 8 (Lang and Chapin). Six of the workers from Garamba are from the stomach of a Bufo regularis and a single worker from Akenge is from the stomach of a B. funereus. The specimens from Faradje were taken while they were attending plant-lice on young orange trees. Camponotus (Myrmosericus) rufoglaucus subspecies syphax, new subspecies Plate XXII, Figure 1 WorKER very similar to the subspecies zulu Emery from Natal and quite as large, the largest specimens measuring fully 9 mm., but not more slender than other forms of the species. The scapes and tibixe are distinctly compressed, the former as in C. eugeniz Forel, but not so broad. Epinotum evenly arcuate in profile, without distinct base and declivity. Pubescence dull yellowish, not very long, slightly golden on the gaster of large individuals, only feebly converging at the mid-dorsal line on the posterior portions of the second and third segments. Color brownish black, the legs a little paler, the funiculi, cheeks, clypeus, mandibles, and tarsi castaneous. Gastric segments with very narrow, dull-yellowish posterior margins. Numerous specimens from Zambi (type locality) and Boma (Lang, Chapin, and J. Bequaert). The Zambi specimens are from three colonies, two of which bear the following notes. ‘Ants forming numerous small craters in the white sand (Pl. XXII, fig. 1). Only a few individuals were seen outside the nest before noon. The nest extended to a depth of 50 cm. below the surface.”’ ‘Nest in the rotten base of a Hyphzxne. No larve nor pupe could be seen, though there were certainly as many as 1000 workers in the colony. The nest was loosely arranged in the soft, decomposing mass.”’ Bequaert says of the specimens from Boma that they “run very swiftly and were nesting in the road.” Workers of this ant were sent to Prof. Emery, who compared them with his cotypes of the subspecies zulu. He pronounced them to belong to a new subspecies “with the pubescence on the gaster much more parallel and less sinuous.” i Wheeler, Ants of the Belgian Congo ay aa : (M ; icus) rufoglaucus subspecies flavomarginatus z . (Mayr) , 8; Vankerckhovenville, 8; Garamba, 8; Faradje, 8 (Lang in); Thysville, 8 (J. Bequaert). A small number of speci- each locality, without further data. _ Camponotus (Orthonotomyrmerx) vividus (}'. Smith) _- Plate XXTI, Figures 1 and 2; Text Figure 63 erous workers of this shining black ant taken at Malela (Lang, in, and J Bequaert) and a single deilated female from Lukolela ng anc Chapin). At Malela, the ants had occupied the large nest of sal termite( Pl. XXI, figs. 1 and 2). “This consisted of strong, Fig. 63. Camponotus (Orthonotomyrmex) vividus (F. Smith). Worker major. a, body in profile; 5, head, dorsal view. fp carton and was built around the stem of a sapling, which grew in a ‘ove Swamp among raphia palms. When the nest was disturbed oi sr ants swarmed out and covered the nest in great numbers and m2 in up on our bodies and attacked us furiously. Only after we had the nest open did we notice that it had been originally built by mi Some dead specimens of these were found in one corner. As own in the photograph, the ants themselves had excavated the strong mn, making more spacious and more irregular cells. There were al large and many small entrances on the surface of the nest.”’ 248 Bulletin American Museum of Natural History (Vol. XLV Campon otus (Orthonotomyrmex) vividus variety semidepilis, new variety Worxer.—Exactly like the typical form, except that the erect hairs on the dorsal surface of the head and body are distinctly paler and only about half as numerous. The pubescence, too, is more dilute and shorter, especially on the gaster. Described from numerous workers from Medje (type locality) and Leopoldville (Lang and Chapin). The following note relates to the speci- mens from the former locality: “These ants were taken out of their nest in the rather rotten portions of a tree. Their galleries were often large enough to admit one’s finger. The workers, when disturbed, ran out and bit viciously. The specimens were taken about five miles south of the Nepoko while we were collecting accessories for the Museum group of okapis.”’ Fig. 64. Camponotus (Orthonotomyrmez) vividus subspecies cato (Forel). Worker major. a. hody in profile; 6, head, dorsal view. Camponotus (Orthonotomyrmex) vividus subspecies cato (Forel) Text Figure 64 Stanleyville, 2,8, 9; Garamba, 2,8; Medje, 2,8, 9; Avakubi, a,%; Akenge, 8; Thysville, 8; Bengamisa, 9, &; Niangara, 9, 7— (Lang and Chapin). The workers from Akenge, two in number, were taken from the stomach of a Bufo polycercus, a female from Medje was from the stomach of a B. funereus, and one from Stanleyville from the stomach of a frog (Rana mascareniensis ). Under separate numbers two different native names, “‘suma” and “likulu,” are given for this ant. The specimens from Stanleyville were Wheeler, Ants of the Belgian Congo 249 Cc framing up and down the trunks of big trees near the Tshopo tiver in great numbers;” those from Medje were found in similar situa- ions and also crawling over the tents. ‘When crushed, they gave off a enc ylike bugs.” s i _ Camponotus (Orthonotomyrmex) sericeus (Fabricius) = Figure 65 . Bequaert). Numerous specimens, without further data. Map 43 s the distribution of this species. iupcen: ae 2 C re ahh ini rpe Ee All gc eth awn ~, a a oo Fig. 65. Camponotus (Orthonotomyrmez) sericeus (Fabricius). Worker. a, body in profile; 5, head, dorsal view. Camponotus (Myrmotrema) foraminosus Forel, variety Numerous workers and four males from Faradje and Avakubi, and Ee probebly several females from Stanleyville and Bengamisa (Lang and _ Chapin), belong to a variety of this species, which I refrain from naming, _ owing to the small amount of material of this extremely variable species in my collection. Prof. Emery, to whom specimens were submitted, ues that ‘the pubescence is more abundant and more golden than in » specimens sent by Forel as corresponding to the type of the species The true type is a unique and is in the collection of the Museum _ of Geneva.” _ The specimens from Faradje were found “living in the hollow eavities of twigs and branches which they probably bored themselves. ‘ "The cavities also contained numerous estivating snails, which were evi- 250 Bulletin American Museum of Natural History {Vol. XLV dently not molested by the ants. The snails were so tightly attached té the surface that they were often broken when an attempt was made to remove them. About this time (the latter half of December and begin- ning of February) the grass is burned all over the country. The flames leap high and the heat is incredible, many of the branches of the trees being killed by the fire. This may be a reason for the snails’ seeking refuge in the cavities made by the ants.” The snails belonged to Pach- nodus herbigradus Pilsbry.' (See p. 154). 8 2 “ss “4 ei ey 4 7 ee ee eee Sia 2 neath | 4 ie GY Tf 45 . ° = Pm * Pm = 0 os “~ as a. Map 43. Distribution of Camponotus (Orthonotomyrmez) sericeus (Fabricius). Camponotus (Myrmotrema) foraminosus subspecies hereticus Santschi A single worker major from Lukolela (Lang and Chapin) seems to be referable to this subspecies. Camponotus (Myrmotrema) foraminosus subspecies auropubens Forel’ variety A single minor worker from Stanleyville (Lang and Chapin), which @ I am unable to assign with certainty to any of the described forms of this subspecies. 'Pilsbry, 1919, Bull. American Mus. Nat. Hist., XL, p. 308. Wheeler, Ants of the Belgian Congo 251 "Camponotus (Myrmotrema) perrisii Forel subspecies jucundus Santschi ee 2 Text Figure 66 _ Kwamouth, 2, %, 9, &; Niangara, 2, %; Faradje, 2, 8, 9; Garamba, 2, 8 (Lang and Chapin). Many specimens, some of __. which were identified by Prof. Emery as belonging to this subspecies. Those from Kwamouth were found with their pupx nesting in the gal- eries of a large, conical termitarium; those from Faradje were taken in mushroom-shaped termitaria. Those from Niangara, however Z on & “were nesting “in the hollow of a tree.” oe ee Fig. 66. Camponotus (Myrmotrema) perrisi: subspecies jucundus Santschi. Worker major. a, body in profile; 5, head, dorsal view. 4 The female of this subspecies measures 11 to 11.5 mm. (wings 12.5 mm.) and resembles the major worker very closely in sculpture, pilosity, _ and color, except that the erect whitish hairs are shorter and less numer- _ ous on the upper surface of the head and thorax. The antennal scapes are not so pale at their base. The wings are suffused with brown and ___ have dark brown veins and pterostigma. The male measures 7 to 8mm., is black throughout, with wings colored like those of the female, but _ paler. The scapes and hind tibiz are distinctly flattened, though much less so than in the worker and female, and the upper border of the petiole ___ isstraight and transverse, with a small elevation or tooth on each corner. __ The body is rather shining; the thorax without erect hairs above. 252 Bulletin American Museum of Natural History (Vol. XLV Camponotus (Myrmotrema) perrisii subspecies jucundus variety grandior (Forel) 2 Yakuluku, 2, 8; Garamba, 2, 8 (Lang and Chapin). Numerous specimens. Those from Yakuluku were found “nesting in small mush- room-shaped termitaria, which were only about five yards apart.” Camponotus (Myrmotrema) olivieri (Forel) variety sorptus (Forel) Seven minor workers taken at Kwamouth, Leopoldville, Lukolela, and Stanleyville (Lang and Chapin). The types were taken by Forel from the stomach of a pangolin (Manis temmincki). Camponotus (Myrmotrema) bayeri Forel Thirteen workers from Faradje (Lang and Chapin), without further data. Camponotus (Myrmotrema) micipsa, new species Text Figure 67 WorRKER MAJOR.— Length 9 to 10 mm. Head large, longer than broad (without the mandibles, 3.8 X 3 mm.), broader behind than in front, with excised posterior border and evenly and very feebly convex — sides. Mandibles very convex, with 6 short, subequal teeth. Clypeus rather flat, longer than broad, ecarinate and feebly longitudinally grooved in the middle, sub- hexagonal, narrower in front than behind, its anterior border somewhat truncated, straight. Frontal area impressed, lozenge-shaped; frontal carine widely separated, as far apart as their distance from the sider of the head. Antennal scapes distinctly flattened but not dilated, somewhat narrower at their tips than in perrisii, extending a little beyond the posterior corners of the head. Eyes rather small and flat. Prome- sonotal and mesoépinotal sutures more impressed than in perrisii; the epinotum some- what cuboidal, as long as broad, the base and declivity subequal, nearly rectangular in profile, the former flattened, the latter very feebly concave, both slightly submarginate on the sides. Petiole similar to that of perrisit but broader above, the upper margin feebly notched in the middle. Hind tibie somewhat flattened but neither prismatic nor channelled, their flexor borders without a row of bristles. Mandibles, clypeus, upper surface of head, thorax, and gaster opaque; mandi- bular teeth, frontal area, antennal scapes, gula, sides of thorax, posterior surface of petiole, legs, and venter shining. Mandibles finely punctate on a very finely and evenly shagreened ground. Head very finely, densely and evenly punctate; the cly- peus and cheeks with coarse, shallow, rather sparse, piligerous foveole, which are elongate and oblique, with their posterior edges more pronounced. Front and sides of head with similar but more scattered and less pronounced foveole. Antennal scapes covered with round punctures of very unequal size. Thorax and gaster very finely and densely punctate like the head, with small, rather sparse, piligerous punctures. Wheeler, Ants of the Belgian Congo 253 S Hic coarse, erect, rather long and abundant on the upper surface r Bs loc wat spp coh Hoven ars « chor, oat, tnt, wb tits ieiasarly-enianlitg the ania. iffering from the worker major only in the smaller and shorter head, which is r than wide behind. The foveol of the cheeks and clypeus are less distinct, >y, erect golden hairs arising from them are as striking as in the major. Fig. 67. Camponotus (Myrmotrema) micipsa, new species. a,hecd of worker maxima; }, clypeus of same; c, thorax and petiole of same in profile; d, head of worker media, zyand Chapin). This species is evidently allied to perrisit, oliviert, 7, and maynei Forel, but distinct from all of them in the structure of the oa sculpture, pilosity, etc., though apparently most closely re- to maynei. 254 Bulletin American Museum of Natural History (Vol. XLV Camponotus (Myrmorhachis) polyrhachioides Emery = Lukolela, 8, 9; Lie, 8 (Lang and Chapin). The workers from the latter locality, two in number, were taken from the stomach of a toad (Bufo regularis); the specimens from Lukolela, comprising two workers and three winged females, were found running on fire-wood. U c Fig. 68. Camponotus (Myrmamblys) chapini, new species. a, worker major, body in profile; b, head of same, dorsal view; c, head of worker minor. Camponotus (Myrmamblys) chapini, new species Text Figure 68 WoRKER MAJOR.— Length 5.5 to 6.5 mm. Head very large in proportion to the remainder of the body, longer than broad (without the mandibles, 2.4 X 2.2 mm.), broader behind than in front, with deeply excised posterior, rather convex lateral borders and prominent, rounded posterior corners. Mandibles stout, convex, coarsely 6-toothed. Clypeus flattened, strongly carinate, its anterior border notched on each side, with a short median lobe, angularly — emarginate in the middle and rounded at the corners. Frontal area obsolete; frontal groove distinct; frontal carine approximated in front, subparallel and widely sepa- rated behind, nearly as far apart as their distance from the lateral borders of the head. Eyes small and flat. Antenne short, seapes (1.2 mm.) curved, somewhat flattened basally and thickened at their tips, which extend only about three times their greatest diameter beyond the eyes. Thorax small, short, and robust, not longer than the head, very broad through the pronotum, which is as broad as long, very rapidly narrowed , aupiis) Wheeler, Ants of the Belgian Congo 255 — to the Interalty cor ied sities the meso- and epinotum together not longer : the y Promesonotal suture strongly impressed, metanotum very small or hut distinct. In profile the general dorsal outline of the thorax is arcuate, sega is somewhat raised in front at the suture above the pronotum; notum sloping, rounded, with indistinct, subequal base and declivity. Petiole its. ale elliptical from behind, evenly rounded above, with a slight angular on in the middle of the superior border; in profile scarcely thicker below than much compressed anteroposteriorly, about three times as high as thick, with superior border. Gaster much smaller than the head, the first segment rly truncated, the dorsal surface convex. Legs rather stout, tibie slightly flat- tarsal claws rather long. Shining throughout; mandibles coarsely cadre at their bases shagreened and yue. Clypeus and head sharply shagreened and covered with coarse, sparse s, which are very uniform on the clypeus and cheeks, somewhat shallower more seattered on the front and vertex. Posterior corners of head with a few ate foveole. Thorax and gaster more finely shagreened than the head, the . Ganeversely, and both with scattered piligerous punctures. ‘Hairs yellow, sparse, coarse, erect, and rather short. Petiolar border with four -gula with only a few short hairs; cheeks hairless. Scapes naked; tibia with amerous, very short subappressed hairs. Pubescence sparse, appressed, distinct, +t on the mandibles, clypeus, and cheeks, longer on the gaster. Head deep castaneous, almost black; mandibles and anterior portion of clypeus ; antennz, pronotum, cox#, and legs brownish yellow or testaceous; re- ‘Gf thorax, petiole, gaster, and an inverted V-shaped spot on the dorsal surface of er re erncens. “Length 3 to 4.5 mm. s broad as long, a little broader behind than in front, with straight sides pie feebly ead border. Clypeus strongly carinate as in the major, but its anterior 4 heal pilosity, and color much as in the major dations: but the thorax uni- " formly brown throughout, and the head paler, though darker than the thorax and : Described from five major and eleven minor workers from Garamba type locality), a major from Medje, and a minor from Faradje (Lang uM eenrg The specimen from the locality last mentioned is fre rom the 256 Bulletin American Museum of Natural History [Vol. XLV PHASMOMYRMEX Stitz Worxer.—Rather large, elongate, monomorphic, varying little in size. Head rectangular, with rounded posterior corners. Clypeus rather flat, indistinctly carinate, without an anterior lobe, its anterior border broadly and angularly excised. Thorax long, flattened above, obtusely marginate on the sides; anterior corners of pronotum angular; metanotum distinct, bounded by well-defined sutures anteriorly and pos- teriorly, its stigmata situated below its lateral marginations; mesometanotal suture impressed; epinotum subcuboidal, truncated behind. Petiolar node thick, with a distinct angle at the sides of its dorsal margin. Gaster small. Legs long, hind tibia three-sided. Fem aLe.—Head as in the worker. Thorax depressed, pronotum seen from above nearly as long as the mesonotum and overarched by the latter only very slightly. Scutellum not projecting over the postscutellum or epinotum. Wings as in Campo- notus. Mate unknown. A single species, originally described by Forel as Camponotus buch- neri and known only from the West African region, from Cameroon to Angola (Malange) and eastward to the Ituri forest. Phasmomyrmex buchneri (Iorel) Lukolela, 8; Avakubi, 8; Medje, § (Lang and Chapin); Teteee: 8 (J. Bequaert). Single specimens. Those from Avakubi and Lukolela were taken on fire-wood brought in from the forest. Potyruacuis |. Smith Large or medium-sized ants closely allied to Camponotus. WorkKER monomorphic. Head orbicular, oval or rounded subrectangular, very convex above, with very prominent, long and sinuate frontal carine. Palpi long, the maxillary pair 6-jointed, with the basal about half as long as the second joint, the labial pair 4-jointed. Clypeus well developed, usually convex or more or less carinate. Antennx long, 12-jointed, the scapes inserted some distance behind the posterior border of the clypeus, as in Camponotus; funicular joints considerably longer than broad. Thorax more or less arcuate above, often more or less carinate on the sides, and more or less dentate or spinose, but exhibiting great differences in conformation in different species. Usually either the pronotum or the epinotum or both are armed with teeth or spines, rarely the mesonotum. The petiole has a large scale, the superior border of which is nearly always armed with pairs of spines or teeth, more rarely also with a median, unpaired spine or tooth. Gaster large, broadly elliptical or sub- globular, very convex above, the first segment forming more than half of its surface and often more or less truncated or concave in front. Legs long and well developed, the tibiw often constricted at the base. Gizzard much as in Camponotus. Femate decidedly larger than the worker, with massive thorax. Spines and teeth on the thorax and petiole smaller. Wings long, the anterior pair with a radial and a single cubital cell; discoidal cell lacking and cubital vein usually reaching the outer margin of the wing. Gaster massive, its first segment often proportionally” shorter than in the worker. \ a Wheeler, Ants of the Belgian Congo 257 J kde sloscly ‘resembling the male of Camponotus, small and slender; the thorax and petiole quite unarmed, the latter with a low, thick scale. Frontal carine more front more convex, pronotum overarched by the mesonotum. External ital valves small and slender. Cerci distinct. Pup enclosed in cocoons. A large genus comprising several hundred species, many of which are 10ng the most beautiful of ants, confined to the tropics of the Old 4 World, though, like @cophylla, absent from Madagascar (Map 44). _ The species of Polyrhachis, however, have a wider range, since a small number of forms occur as far north as Syria in Asia and as far south as the eastern Cape Colony and Tasmania. The majority of the species are aggregated in the Indomalayan, Papuan, and Australian Regions. Forel and I have divided the genus into subgenera, eleven of which, based s 0 1S EY) 45 Cy 75 90 105 120 3S 160 vt ’ or Die wot ea fb = 7 5 a) WAYS Q ee, “a = ' “a Pes a = 3 + ae 6 , i -~S iS ‘ 6 |. NS 7 \ © x + r) 7 . eS _ \ “S ‘s\ ieee - a ’ . 7. a ij is . “4 ee F ’ bo wa / s y A “ 30 \ é ~~ ; “ea S f iS ° 1s » 45 60 5 90 105 120 a5 es Map 44. Distribution of the genus Polyrhachis. _ on peculiarities in the structure of the thorax and petiole, have been - recognized up to the present time, namely, Polyrhachis, sensu stricto, _ Campomyrma Wheeler, Hagiomyrma Wheeler, Myrma Billberg, Hedo- __ myrma Forel, Myrmhopla Forel, Chariomyrma Forel, Myrmatopa Forel, | Cyrtomyrma Forel, Myrmothrinax Forel, and Dolichorhachis Mann. In the Ethiopian Region only two of these, Cyrtomyrma and Myrma, are _ known to occur, the fcrmer represented by a very few aberrant species, _ the latter by a number of forms which show much greater diversity of _ structure than do the species of the same subgenus in the Indomalayan - and Papuan Regions. This fact, together with that of the wide distribu- 258 Bulletin American Museum of Natural History [Vol. XLV tion of Myrma, would seem to indicate that it is the most archaic of all the subgenera of Polyrhachis. The species of Polyrhachis form only moderately large colonies and none of them is sufficiently common to be of economic importance. Many of them are, in fact, rare and sporadic. They are very timid or pacific insects and are most frequently found singly walking up or down tree-trunks or on the foliage of trees or bushes. Their nesting habits are very diverse. According to my observations in Australia, the species of Campomyrma nest in the ground, under stones, or more rarely in crater nests. The same is true of the species of Hagiomyrma and Chariomyrma, though I have always found P. (Hagiomyrma) semiaurata Mayr in large logs and certain species of Chariomyrma in earthen termitaria. So far as known, none of the species of these three subgenera employs silk in the construction of the nest. The species of Hedomyrma, as Mann and I have observed, live in high trees, but we have been unable to find the nests. Several of the larger species of Myrma nest in the ground or in logs and some of them line their nests with silk spun by the larve. Many of the smaller species of this subgenus make carton and silken nests on or ‘between the living leaves of trees, and this is the general habit also of many species of the subgenera Myrmhopla, Myrmothrinax, Myrmatopa, and Cyrtomyrma. : + > oh . z rs in on o Fe Bou.erin A. M. NH, Vor. XLV, Pirate V Piate VI Fig. 1. Dorylus (Anomma) nigricans Mliger, at Amani, fon as Army crossing a path. The workers carrying the brood pass between so soldiers which, with their mandibles lifted and wide open, pee ee the army. Photograph by Dr. J. Vosseler Fig. 2. Megaponera jetens (Fabricius), at Avakubi, October z trance to a nest, surrounded by a small mound of excavated earth, deserted plantation. When dug up, five galleries were found to open ir aperture. On two occasions Mr. Lang observed from 30 to 40 pupa cas _ side in the sun, near the entrance, with a few ants in steady attendance. no true chambers in the nest, but the galleries for the pupe and gy wide. When touched, these insects sting before using the mandibles, wl pierce the thick skin of the hand. The nohoseie of Sere sea Sete individuals and, when closely approached, break up at once, the men nervously in all directions and making a stridulating noise. “After they reform the ranks and continue their mareh. They are great termite Mr. Lang counted as many as eight such insects held between the mandibles ant. They never opened the jays todrop their prey, even: when Salam D forceps. Beuietin A. M.N. H. Vor. XLV, Prate VI Pheidole saxicola W at Z bi, night and early morning, the . Near the entrances to the nests chiefly forage. are and chaff, and often also of dead ants and other i TTA 8491 ‘ATX “TOA ‘HN NY Smariog iF Piate VIII . Myrmicaria eumenoides subspecies opaciventris (Emery). Fig. 1. Crescent-shaped craters of excavated earth at the er level, hardened soil at Rungu, July 7; 1913. The ants usually burrow tl after a heavy rain, either by day or night. The workers then busily e: of soil which they drop near the edge of the crater. Often the moist « roll down but sticks to the upper margin which thus becomes an ove The mounds in the photograph are of typical form, but some of the twice as high (5 to 6 em.). It is said that these craters suggested the famous hairdresses of the Mangbetu tribe. ; Fig. 2. Crepeent-aiaped. crater: of sbk ieieiatetaal a mae ae 22, 1909. In this case it was not as true to form as those shown in Fig. entrances were placed near the base of a bush. The galleries showed n tions and extended 17 inches below the surface; ee did not cover an area more than two feet in diameter. Most of the pu about the roots of the bush. These harms and eommon ants also ranean tunnels in various directions from their nest and make t by their immediate appearance in = sumbepempond a eee meat or Beier A.M.N-H. Vou. XLV, Puate VIII Fig. 1. A flowering branch of the tree. _ Fig. 2. ‘The eatin sree in 101) Buiierm A. M. N. H Vou. XLV e IX tT > / U . * Pe n * us a 3 j f : P — t = % 3 5 ¢ Pe | ° i = a 4 a = bs ‘ * 1, . : c e. % t - A H To ae i ~ phe acm , ; aS 4 * ‘ 4 ~ . . hg n d 5 Mg , . « 6 ‘ d “> : . f +5 af f y * ne ne ¢ ¥ ¥ , : ; = # : ¢ Nis as P| . i 3 . a . “ ’ , ¥ i bam f os *. , 7 ; i r b “4 , > : 1 i . Ww ns . . ¥ -t : ; ‘ " ‘ i > . 24 Buteerrs A. M.N.H Vou. XLV, Pearse XI Crematogaster (Atopogyne) theta ), Lec Fig. 1. Oniaia, view of 6 tacton’ snk ainle ol vests light gray or brownish color. The caterpillasshows on Plate X11, uA over the surface of this nest. Pues 7. Fig. 2. Inside, toas-nection vise et MatNaine ih. The white brood (eggs, larve, and pupa). The structure was 10.4 em. broad and attached to a small tree in the forest, about 8 feet from the g turbed, the ants stream outside and let themselves drop upon the i sting is painful and can be felt for many miirtes afterwards. X11 PLAT? XLV Vor Piate XIII Fig. 1. Portion of the outer surface of the nest of Crematogaster (Atopogyne) theta (Forel) shown on Plate XII. In the upper right corner is seen a caterpillar that was found crawling over the surface, its segmentation being visible at the time; but when the creature stops and tightly adheres to the nest, its body becomes quite un- ; noticeable as it then resembles one of the numerous protuberances of the formicary. Fig. 2. Nest of Crematogaster (Nematocrema) stadelmanni variety dolicho- cephala (Santschi), at Kwamouth, July 14, 1914. This cone-shaped carton nest was hanging in a tree, about nine feet from the ground. It was fastened to several small branches in such a way that it moved about when the boughs were tapped with a stick. The outside surface was quite rough and simulated crumpled up leaves that cover one another like the shingles of a roof. The cellular structure inside was * irregular, with very thin walls, and a great many exits; larve were especially abund- ant in the lower portion. It measured about 18 inches in length and 11 inches in width at the top. a at i. rs Pirate XIV_ Crematogaster (Nematocrema) stadelmanni variety la Bengamisa, September 27, 1914. Pepilie nen ofeey Mies wane that of certain termites in shape as well as in material, a fact usu: possible to tell from the outside appearance which insect photographed was so fixed to several creepers that it anil sive of Uns carton esha wa?y gseatly Wialilack ta it structure is irregular, the galleries and cells seemingly arran ad pope ney ee ee 4 d ay al Bouierix A. M,N H ot. XLV, Pirate XIV XLV, Prate X\ Vor H N BetieTin A. M Muskinqou gersien of Acaniholertuee mdiitorts & 6, 1915. The minute ant, Pxdalgus termitolestes ' TAX 44V1q ‘ATX “TOA H NW ‘Y Nia T1IOg Beitetms A. M.N.H Vou. ALV, Prare XVII Tetramorium setigerum s 8 Fig. 1. Regular ring-shaped craters of | m of so the entrance of the nest during the rainy season. ‘These ants open places. < Fig. 2. Aspect of the entrance to the nest of the same ant At that tine te insects GS calif aaanana seme ing to construct acrater, i ae Bt tl eo ee . 5 pene os ‘ ee + fy n ae rs = Se * . 4 © bi . ~“e % . Bouieri~x A. M.N. H. Vout. XLV, Pirate XVILI Fig. 1. ee a the narrow beach of white sand in the upper part of which the are excavated, Photograph by J. Bequaert Fig. 2. Nest of P. euslodiens in the sandy beach of the August 1915. : Buuietin A. M. N. H ov. XLV, Prare XIX Ccophylla longinoda til) at Mal July 6, 915, consist of leaflets closely woven together with white silk. thorny bush about three feet from the ground, In order to. compound leaves of the plant were cut off and laid on the ground. Fig. 1. Six leaflets have been united intoone nest. Fig. 2. A closer view of another formicary of the same species. _ ys vi +s ‘ t A “ 7 yy vie 4 ‘ - ° * Bounetix A. M. N.H Vow. XLV, PLrare XX Fig. 1. Coin Lat et ecm ae ea sro builder and now occupied by a colony of Camponotus Ort Smith); near Malela, July 7, 1915. The structure was e: of a sapling in swampy woods. Fig. 2. Intro ofthese nt showing theca in the termitarium. ie s\ r i IXX 4£V1q 'A'LX “TOA IN CY Sieeriag Puate XXII Fig. 1. Craters of white sand at the entrances to the Camponotus (Myrmosericus) rufoglaucus mabnpgeics syphax V June 30, 1915. Fig. 2. Nest of Polyrhachis (Myrma) laborioea F. Smith, at Ni 1910. it was built ine fexk of s bach no aie vegetable fibres and leaves fastened together. ve atenty ae tered into its construction; dark gray outside, brown inside. ‘ somewhat damaged there were apparently many exits. ‘When Hu proabarred pear nd raecivlwenge sys at ‘sa they cusited considerate eae ee er for tween the legs. ms =tak Bouterin A. M.N.H. Vou. XLV, Pratre XXII Nests of Polyrhachis (Myrma) rset June = 1915. Buuterin A. M.N.H. Vor. XLV, Pratre XXIII Ill—THE PREDACEOUS ENEMIES GF ANTS By J. BEQUAERT _ The various means by which Nature prevents an excessive increase of “species not only forms in itself an interesting chapter of ecology, t its study is also of great importance in an understanding of the true ____ meaning of Natural Selection. In the case of ants it has been contended ‘that they are better defended than other insects against the attacks of _ predatory animals. Poulton! evidently takes this for granted when he considers that ants, together with wasps, are among the favorite models for “mimicking” insects and other arthropods. These ant-like arthro- j pods, having acquired by Natural Selection their resemblance “to the __ aggressive, abundant, and well-defended ants,” would according to this theory escape many of the attacks of their deceived and disgusted -predaceous enemies. Though the evidence presented in the following _ pages is still very fragmentary, I trust the reader may easily conclude for himself to what extent such resemblances, which, in some cases at least, _ ean hardly be doubted, have a real protective value. There is certainly little or no evidence to show that, as the theory-is often expressed, ants — are unpalatable to most insectivorous animals and are merely eaten accidentally or ‘during the time in which young birds or other animals _ are learning what to eat with impunity and what to reject.’”? _____ Another consideration of interest is the relative efficacy of parasitism Pr. : ‘and. predatism i in acting as a check on the reproductive power of the _ species. This point has been profusely discussed, and the argument has frequently been made that parasitism is in this respect of foremost importance. It must, however, be kept in mind, that, while we have been very completely and steadily informed of the activities of parasites, ____ predatism has been much less investigated. It is not my intention to go further into this question; but I think a rather conservative view will ____ be to consider that ecto- and endoparasites, while working all the time, though affecting only a small number of individuals at once, constitute a more regular check to the increase of the species. On the other hand, predatory enemies as a rule destroy large numbers of individuals at a Bees Fe Be caret at, Sep Dee & and experiments (1596-1901) on the bionomics of South African insects, chiefly to the investigation ing colors,’ Ent. Soc. pp. 287 Pis, rx-xx1u1; McAtee, W. L., 1912, ‘The experimental method of testing the efficiency of warning and colora from their enemies,’ Proc. Acad. Nat. Sci. Philadelphia, pp. wots rene Yay OM the Tg cong pe Tipe S080 with notes on the significance cussed the possibility of sn (Soo Cpr hE Sieoete vain of Venere Gelsction, ether to enable them to more readily obtain their food or to protect a, from natural enemies. 271 272 Bulletin American Museum of Natural History [Vol. XLV time, but only at intervals. They are also apt to make their influence more felt when their prey for some reason or other suddenly multiplies on an exceptional scale. Professor Forel’s aphorismic statement that “the most dangerous enemies of ants are always other ants, just as the worst enemies of man are other men,”’ may be true in a general way for temperate regions, where ants are not superabundant and lead a rather inconspicuous life, but it can hardly be applied to the tropics. Ants, it is true, attract comparatively few of the predaceous arthropods, against which they are very effectively armed. They form, however, a consider- able portion of the diet of many reptiles, amphibians, birds, and certain insect-eating mammals, some of these vertebrates being almost exclu- sively myrmecophagous. It may be further mentioned that many of these predaceous animals by no means confine their attacks to the smaller, more timid species of ants, but rather prefer the large-sized, powerfully defended members of the ponerine and doryline groups. The information contained in the following pages is based to a con- siderable extent upon examination of stomachs and pellets of predaceous animals in the wild state. I fully agree with Swynnerton that these sources of information are most valuable with regard to the general preferences of a predaceous animal, the insects it usually feeds upon and on which it for the most part “fills up.” But I also believe with the same author that a knowledge of its detailed preferences must come in the main from continuous observation of individual wild animals and from special experiments both in nature and in captivity. The experimental method has been used with much skill and care by Swynnerton! to test the palatability of butterflies and its bearing on the efficiency of eryptic form and coloration. Miss A. H. Pritchett® has also published the re- sults of a number of experiments with lizards and various insects, including ants, that possess protective, mimetic, and warning colors or that have some disagreeable characteristics which in a measure are sup- posed to prevent their being devoured by insect-eating animals. Such investigations with’ants and their natural enemies should be extended and could not fail to add considerably to a better understanding of predatory habits. iSwynnerton, C. F. M., 1919, ‘Experiments and observations bearing on the eae tion of form and colouring, 1908-1913,’ Journ.’ Linn. Soc. London, Zool., NNT No. 224, pp See Poulton, E. *The experimental as = of the 7 eee ‘value of colour and a4 Dahl, Fe io! in reference to their vertebrate enemies,’ 1. Soc. London, 1887, pp. 191-274; F., 1913, ‘Vi ergleichende Physiologie und } Soriglope or Spinnentiere unter besonderer Beriicksich Lebensweise,’ tml I, vi+113 pp.; Hei 1919, pe metéke Myrmekoidie. T: material zur Lésung des Mimikryproblems,’ Bi ienirabl XIX, 65-102: Dahl, F oy ; spemeonaindle eat eee mit Bienen, Wespen und Ameisen, \ Neto ochenschr., N. F., XX, pp ¢ 70-7: : 3 : \ Hen) wPritchett, A. H., 1903, ‘Some experiments in feeding lizards with protectively colored insects,” Biol. Bull., V, pp. 271-287. Wheeler, Ants of the Belgian Congo 273 1 ARTHROPODS In the following account I shall consider only the arthropods which ee on ants without entering their nests; the nidal synechthrans, or carnivorous inmates of ant nests, are better studied in connection with ‘ true ant guests, though they may in some cases have been derived from outside marauders. Neither have the predaceous activities of ants towards other ants of the same or of different species been considered _ Ants are comparatively immune from the attacks of predaceous arthropods, being themselves usually well provided against such enemies with offensive, defensive, or repulsive weapons. They nourish, however, a host of parasites and commensals belonging to almost every group of arachnids and insects, but these fall outside the scope of the present account. It must be admitted that, with the exception of certain of the ‘most striking cases, such as ant-lions, but little attention has been paid Arachnida _____ Ants do not often fall a prey to spiders and their relatives, except in the winged phases during the short period of the nuptial flight when large numbers of them perish in spider webs. The cautious ways of __ most worker ants make them a difficult game for terrestrial arachnids -__ and in the larger forms the sting is an effective weapon against the attack ____ of the soft-bodied spider. At one of the meetings of the Entomological __ Society of London, Poulton exhibited a spider and its prey taken at Itigi (former German East Africa) by Carpenter, the specimens being accom- panied by the note: “Spider seen coming out of a nest of Megaponera bearing one feebly struggling, upside down in its fangs. Caught in a box the spider settled down to feed on the ant.”? Poulton comments upon the remarkably small size of the spider as compared with its victim, which is one of the largest of African ponerine ants. Certain terrestrial spiders of the Old World genus Zodarion Wal- ckenar (=Enyo Audouin) are true ant hunters. “The Zodarion,”’ says E. Simon,? “which I have observed in southern Europe, live at the expense of the ants and settle in their vicinity. They make neither snare nor web to stop their prey, but during their hunting hours they roam about the formicaries and mix with the long rows of ants, going from one a 5, Trans. Ent. Soc. London, (1917), Proc. ‘Histoire naturelle des veer ak 2, BP. ont. Bee alee Hanon, B. 1874, de France,’ (Paria), I, p. 242: van Hassel Ent., XXXIV, . xxxiv-xxxvi; Krausse, A., 1913, ‘Eine Spinne yD any Fe a4 Be a By Abfallpliteen der teameisen auf Sardinien,’ Arch. f. Naturg., LX XIX, A, Heft 9, pp. 66-6 274 Bulletin American Museum of Natural History [Vol. XLV to another and unexpectedly seizing feeble individuals, or such as are hurt or hampered by too heavy a burden. When the spider has caught its prey, it drags it aside, near its own abode; this is always surrounded by remains which leave no doubt as to the nature of its diet. These observations relate to Z. elegans and nigriceps E. Simon which, in south- ern France, Sardinia, and Corsica, live at the expense of the ants of the genus Atta” (=Messor Forel), — Many other terrestrial spiders are probably to some extent myrme- cophags. Such is the case, for instance, with Celotes atropos Walckenzr, which was observed in the act of capturing ants by Wasmann' in south- ern Germany. According to H. Lebert,? Dysdera erythrina Latreille, in Switzerland, constructs its tubular silk tent near ant hills, or sometimes even in the middle of ant nests, and plays great havoc with these insects. E. Wasmann* and H. Schmitz‘ describe the skill with which the “‘gallows-spider ’’(Theridion triste Hahn) of western Europe preys upon the blood-red ant (Formica sanguinea Latreille) and related species. This spider spins no web, but lies in wait on a low plant for foraging worker ants: suddenly it drops from its lurking place on to an unsuspect- ing victim passing below. Then, quickly rendering the ant helpless by a few threads entwined around the body, the spider hoists its prey up to the plant as to a gallows and fastens it there. The sucked bodies of the ants are left hanging from the plant, either singly or in groups of two or three. Here again, there is a strange disproportion between the large and fierce worker ant and the small, soft-bodied, feebly armed spider. Another European species, Theridion riparium (Blackwall), was observed by Henking® feeding chiefly on the workers of Myrmica levi- nodis Nylander. This spider spins an irregular web between leaves and branches a short distance from the ground; in the middle of the web is woven a conical tent of silk, closed above, open below, and densely covered on the outside with bits of earth and remains of insects. A number of oblique or vertical sticky threads connect the whole structure with the ground and serve to entrap the ants. If a worker Myrmica happens to touch one of these snares with the antenne or legs, its frantic efforts to get loose attract the attention of the spider hidden in her tent; re oted ty van Hasselt, A., 1892, Tijdschr. v. Ent., XX XV, p. xxii. In the same periodical (1890, XXXIII, pp. 212-214), van Hasselt gives an rs gia of ‘European spiders associated with ants, includ- ing those that have been found inside formicari 21877, Neue Denkeelr: Schweiz. Naturf. Ges, .XXVII, Abth. 2, p. 33. 71898, ‘ Ameisenfang von Theridium triste Hahn,’ Zool. Anzeiger, X XI, PP. 230-232. r 41916, ‘De Nederlandsche mieren en haar gasten,’ Jaarb. 1915 Natuurh. Genootsch. Limburg, Maastricht, pp. Beet (of se te). “ SHenki 1886, ‘Nahrungserwerb und Nestbau von Theridium riparium (Blackw.) Thor., Kosmos, xvitt sg pp. 1-11. Cl —= © _ Wheeler, Ants of the Belgian Congo Fiibich sooner or later loses its arp and is then quickly dragged up a ‘and entangled in the irregular maze above. In his account of the agricultural ant of Texas, H. C. McCook! tion has yet determined, are the spiders. There is a large theridioid (Theridion linea- mentum McCook = T.. lineatum Hentz) who is especially destructive of these ants. I found her nest established upon the grass-grown disks in the followingmanner: several ks of the Aristida were bent over near the top, or midway of the spire, and firmly bound by silken cords. Within this tent and just below the apex, the strong snare of right lines (retitelarian) was fixed, in the midst of which the spider hung in the usual inverted position. The ants are constantly climbing the grass-stalks for purposes which I could not divine. . . . They thus become entangled in the snare and fall victims to the watchful aranean. It is not impossible that the spider, whose snare sometimes hung quite near the ground, swings down and seizes the ants as they pass through the tent. Their dry shells might be seen clinging to the threads, or the yet warm bodies trussed up and swathed for food. Under one of these tents I picked bed a small ball of six or ae ant skeletons rolled up and tied together just as they Coleoptera us i One might expect that certain of the predaceous members of this & 2 both larval and adult, occasionally capture ants, though this kind of prey is often carefully avoided. Adult tiger-beetles (Cicin- _ delid#) have been seen catching ants. Wasmann? mentions the fact _ that in the vicinity ot Para, Brazil, the columns of the leaf-cutting satiba- ant (Atta sexdens) are often attacked by Megacephala (Tetracha) rutilans J. Thomson. Chitty, in England, observed Cicindela campestris hold- inga Myrmicarubrainitsjaws: __ I thought the ant was struggling, for it was alternately right inside the mouth of the beetle and then nearly out, but I think this was really the mode adopted by the beetle in devouring its food. Finally the mesothorax and spiny metathorax were ejected from the mouth and also the shell of the abdomen, which had been sucked empty. The rest of the ant was apparently consumed, but possibly it was only the contents of the abdomen that were really eaten. The larve of the tiger-beetles are very voracious and fierce. They ____ live in deep, tube-like holes which they burrow more or less vertically 11879,‘ The \gotaral Cee? the agricultural ant of Texas,’ (Philadelphia), p. 203. ps eed *Genera Insectorum, Fam, Cara’ Subfamily Cicindelingw,” p. 10. tty, A. J, st id “Cheindele campestris feeding on Myrmica rubra,’ Ent. Record, XVI, p. 206. 276 Bulletin American Museum of Natural History [Vol. XLV into the ground; the hole is blocked, a short distance below the entrance, by the strongly chitinized, horizontal upper surface of the enlarged head and prothorax. If a spider or insect drops into the burrow and comes in contact with this plate, with a reflex motion the larva’s head automati- cally jerks back, throwing the prospective prey against the walls of the tube. Thus stunned the victim is easily seized by the larva’s long, sharp jaws, dragged to the bottom of the burrow and sucked out. From pub- lished data it would seem that the exact nature of the food of these larve has been but little investigated. In his interesting account of the life-histories and larval habits of Cicindela, V. E. Shelford! writes: “The food of the larve consists of land crustacea, centipedes, spiders, dragon- flies, butterflies, flies, beetles, and larve of all sorts, in fact any small animal that comes within reach.” Because of their inquisitiveness, terrestrial ants must frequently enter the burrows of cicindelid larve. In a recent publication, Stéger* concludes from his feeding experiments with Cicindela larve kept in glass tubes, that ants which drop into the burrows are merely stunned, killed and hurled out without being sucked dry, so that they can not be regarded as part of the diet of these larve but rather as their most dangerous enemies. Neuroptera Perhaps the best known ant enemies among insects are the ant-lions or certain members of the genus Myrmeleon. The larve of these Neurop- tera secure their prey by means of funnel-shaped pitfalls which they ex- cavate in sheltered places in dry, loose soil. The size of these funnels varies with that of the larve and the nature of the soil, and may be a few millimeters to 10 or 12 em. across, the depth being about half the diameter. The interesting habits of the common European species, Myrmeleon formicaleo (Linnwus) (=formicarius Linneus) were first accurately described by Réaumur and have since been frequently studied. The larva buries itself at the bottom of the pit, only the upper part of the head and the elongate, widely extended jaws projecting out of the dust. Thus ambuscaded, it remains motionless, sometimes for hours, until a wandering insect runs over the edge of the funnel and either tumbles down at once into the jaws of the waiting ant-lion or slides only a short way and then attempts to crawl up and out of the pit. In the latter case, however, the soft, loose soil on the slope readily yields beneath the legs Mp Journ. Linn. Soc. London, Zool., XXX, No. 197, p. 178. er, R., 1918, ; Sieee Beobachtungen an der Cicindelen-Larve,’ Mitt. Naturf. Ges. nae ‘de1y), pp. 22-44 | #22) "Wheeler, Ants of the Belgian Congo 277 i of thentruggling i insect and rolls down on the ant-lion larva, which at once _ foreefully throws dust with its head. At one time it was believed that these particles were aimed at the victim, but as a matter of fact they are flung out of the pit. In this way the ant-lion merely deepens its funnel, the steep walls then crumbling down under their own weight, carrying the unfortunate insect with them into the jaws of the larva. The man- dibles and maxille of the latter act together as sucking jaws; their _ tips are thrust into the body of the captive and do not loosen their grip until it has been emptied ot its liquid contents, when the corpse is hurled out of the hole. Any insect that happens to drop into the pitfalls _ is taken by the ant-lions, but ants are most likely to do so and many e sucked-out bodies of these insects are usually found near the pits.' ____ The other genera of the family Myrmeleonidz also have predaceous ei esi but, so far as known, they do not dig pits and apparently hunt in ___ the open, their prey consisting chiefly of plant-lice and other soft-bodied = insects. Behe Diptera ___ It is most interesting that, in the dipterous family Leptide, the larve ____ of certain genera have acquired the behavior and some of the structural ____ peculiarities of the ant-lions. These belong to the genera Vermileo and _ Lampromyia, while the other members of the family possess free-living ___ predaceous larve. The best studied case is that of Vermileo vermileo 4 (De Geer) (=V. degeeri Macquart), of southern Europe, the “ver-lion”’ _ of Réaumur, very completely described and figured by both this natural- ist and De Geer, about the middle of the eighteenth century.2. The larva of this fly hides at the bottom of a funnel-shaped pitfall after the manner of the ant-lion; it is a vermicular maggot, which buries and fixes itself in the loose sand by means of four digitate processes, armed with stiff, hooked bristles, at the end of its anal segment; and by means of supple- mentary stiff bristles on some of the posterior rings. The four anterior segments are slender and fimbriate on the sides; they can be curved against a ventral projection of the fifth segment so as to form a loop, with which the larva throws out the dust while burrowing its pitfall. When a small insect, usually an ant, drops into the pit it is seized and 1742, ‘ Mémoires raervir 4 l'histoire des Insectes,” VI, pp. 333-386, Pls. xxxir—-xxx1Vv. sea my ee oe F., 1916, ‘ Die n Ao bipat Hist RP 00 nerds, co Algu nas eoqpaniees hormigas y hormigaleones,’ ctas ist vat. iasner ’ * Ameisen und p her ome men Ein Beitrag zur A yey Ste re mag XXVI, pp. "De Geer, 1752, ig Ac. Sci. Suee., p. 180, Pt. v; 1776, ' Histoire des Insectes,’ VI, p. 168, Pl. x. Réaumur, ‘Histoire du Ver-Lion,’ Mém, Ac. Sei, Paris, 1753, pp. 402-419, Pi. xvi. Brauer (1883, Denkechr. Ak. Wiss. Wien, math. naturw. KI, XLVITI, Abt. 1, pp. 3-44, Pl. xv, figs. 84 and 85) gives an accurate description and drawing of this larva. The earliest reference to the ‘babits of Vermileo is by an anonymous writer in Histoire Ac, Sci. Paris (1706), 1731, pp. 7-8, where it is called Formica-rulpes. 278 Bulletin American Museum of Natural History [Vol. XLV firmly held by the loop around the thorax or behind the head, the loop thus taking the place of the ant-lion’s jaws. Many years ago a similar funnel-burrowing fly larva was discovered by Prot. J. H. Comstock in the Sierra Nevada, California, but could not be reared to the adult stage. Prof. W. M. Wheeler has recently been more successful in obtaining the flies of these larvee, thus adding a second, North American species to the genus Vermileo.! He states that the larva is in behavior and structure very similar to that of V. vermileo, and that it also traps in its pitfalls small insects, especially ants. The adults of the allied genus Lampromyia are very distinet in theiz greatly lengthened, slender, stiff proboscis, but the larve differ only in minor details from those of Vermileo. P. Marchal* has written an ihter- esting paper on the habits of Lampromyia pallida Macquart (=L. miki Marchal), of which he discovered the funnel-burrowing larve near Tunis. Three other species have been described in this genus: L. cylindrica (Fabricius) from Northern Africa and Spain, L. canarrensis Maequart from the Canary Islands, and L. sericea Westwood from Damaraland. During my stay at Algiers in June 1910, I had the good fortune to observe rather closely the larve of a species of this genus, probably L. pallida. They were found in numbers on the outskirts of Mustapha Supérieur, along the highway to Blidah, in the suburb of Colonne Voirol. Wherever the soft sandstones of the road banks happened to be excavated or weathered into miniature caves, one was sure to find the dry, powdery dust beneath the shelter ot the overhanging rock fairly dotted with the funneled pits of Lampromyia. At that season adult flies were fre- quently seen resting on the rocky ceilings of the excavations. 1 found that the most common victims of these larve were workers of the little Tapinoma erraticum (Latreille). Robber-flies (Asilide) are occasionally observed sucking the juices of winged ants, but I am not aware that they ever attack the workers. Certain tropical muscid flies of the genus Bengalia have developed predaceous habits quite unique among the calyptrate Muscoidea; they are frequently found on roads and in clearings hunting for soft-bodied insects after the well-known manner of robber-flies. Attention was first called to these peculiar habits by Nangle* in India and E. E. Green‘ in Ceylon; in both cases the flies, Bengalia obscurepennis (Bigot), were iWheeler, W. M., 1918, ‘ Vermileo eomalers new species, an interesting leptid fy from California,’ Proc. New England Zool. Club, VI, pp. 83-84. *Marchal, P., 1897, ‘ Notes d’ entomologie biologique sur une excursion en Al 9g et en Tunisie. L ampromyia Miki, nov. species; Cécidies,’ Mém. Soc Soe. ool. France, X, pp. 5-25, Pl. °1905, Journ. Bombay Nat. Hist. Soc., XVI, No. 4 , p. 747. #1906, Spone Zeylanica, III, p. 220; 1907, ibid., IV: pp. 183-184 (the fly is here called Ochromyia juna F.), oulton, Trans. Ent. Soc. London, 1906, p. 304 _, —_—— ae Wheeler, Ants of the Belgian Congo 279 aniline winged termites flying at night. J. W. Yerbury saw the same pecies “trying to take her burden from a large ant (Lobopelta species).”’ W. Thomson made the following observation with regard to the Indian . jejuna (Fabricius): “1 always noticed specimens of this species on the ground, or on a stone or leaf near an ant’s nest. On watching, I saw them swoop down on any ant carrying an ‘egg’ or larva, take it from the t, carry it away a short distance and proceed to suck it." Bengalia iro de Meijere, in Java, lurks in the neighborhood of the columns of Pheidologeton diversus (Jerdon); when a worker ant comes along carry- ing its prey, the fly dashes into the moving ant column, quickly steals the prey from the carrier, and returns to its perch where it devours its catch at leisure Lastly, G. R. Dutt, in his entertaining ‘Life Histories of Indian Insects,’* writes of Monomorium indicum Forel as follows: “‘One morning I observed the inmates of a nest marching out with young ones. Close to the nest was sitting a muscid fly (Ochromyia species) which ; attacked from time to time the larve and pupe that were being carried _ by the workers. The fly never snatched the victim from the grasp of _ the ant, but simply ‘licked’ it from its place with the proboscis, which when withdrawn left the larva or pupa quite shrivelled up.” ‘The African Bengaliz evidently have much the same habits as their 3 “Indian congeners. According to W. A. Lamborn,‘ Bengalia depressa - (Walker), in Southern Nigeria, regularly follows the marauding armies of _ Dorylus nigricans, to rob them of their prey. On one occasion the whole _ performance was closely watched and described as follows: | _ I soon saw three or four of the muscids flying about the moving column and occasionally settling near it, sometimes on the ground quite close to the ants, some- _ times on a blade of grass, stone or other raised object. Such as settled on the ground were extremely alert, and being able to run rapidly, never allowed any ants to ap- proach any nearer to them than about a quarter of an inch. When, as frequently happened, any ant made a little circuit away from the main body, a fly would generally pursue it at a distance of about half an inch, but backing away directly the ant turned towards it. Other flies, having rested motionless a few minutes, flew up and poised themselves on the wing over the ants, but, immediately the drivers realized their presence and stretched out towards them with widely opened mandibles, flew again to a place of rest. Eventually I saw a muscid stalking a minor ant which had strayed from the main body carrying a pupa in its jaws. Suddenly the fly rushed forward, and it must have driven its proboscis, which seems to me armed with strong bristles, into the pupa, for the ant was brought to a standstill with a sharp jerk. Then ensued ‘Observations recorded by Poulton, 1914, Trans. Ent. Soc. London, (1913), Proc., pp. exxvilimcxxix. Jacobson, E., 1910, * Pheidologeton diver versus sont und eine myrmecophile F liegenart,’ Tijdschr. v. Ent. = Lt DB. 328-335. Meijere, J.C. a4 de, 1900," Ueber drei von Jacobson auf Java bei Pheidologeten diveraus beobachtete Fliegen,’ pp. 336-340. : #1912, Mem. Dept. prety India, Ene Ser, nia No. 4, p. 251, *1914, Trans. Ent. Soc. London, (1913), Proc., pp. exxv-exxviii 280 Bulletin American Museum of Natural History [Vol. XLV a tug-of-war between ant and fly fastened on at opposite ends of the pupa, but neither had the advantage till, as it seemed to me, the ant must have got annoyed and loosen- ing its hold rushed towards the fly, which of course instantly flew off with the pupa, and this it proceeded to suck on the ground about a foot away from the ants. It allowed me to get quite close before taking to the wing with its prey, and it settled again two or three feet further off and became so preoccupied with its meal that it fell an easy victim to my net. I then carefully watched a fly hovering over the ant column. It suddenly swooped down and rose instantly with an ant pupa, with the driver that had been carrying it still hanging on, fixed to its proboscis. The fly carried this burden for about a foot, then dropped it and alighted on the ground near by. The ant started to run away with the pupa, but the fly pursued it, again impaled the pupa and started a tug-of-war with the ant. Neither side had any advantage, and then the fly rose again about three feet into the air with the pupa and ant and after a flight of about eighteen inches let them fall. The ant being diseomposed by this procedure let go of the pupa, and no sooner had it done so than the fly seized it and, flying off with it triumphantly, settled near by and proceeded as in the previous case to suck the prey. This one again fell easily to my net, so that the flies are evidently keenly alert only when in the immediate vicinity of the ants. I subsequently noticed that the Diptera seemed to have certain preferences in regard to their prey, for I repeatedly noticed one poised over the ant column make an unsuccessful swoop and then fly, keeping level with the ant carrying the particular object which it had missed, making oceasional rushes in an endeavor to secure it. Those I took had obtained ant pupx, but I am sure they take other things from the drivers, probably portions of dead insects. Further observations by Lamborn' in East Africa have shown that the Doryline are by no means the only species of ants favored by the attentions of the African Bengaliz. At Lindi, former German East Africa, a female B. peuhi Brauer and v. Bergenstamm was observed alighting near a column of Crematogaster castanea Smith which was passing up and down a baobab tree; the fly made various attempts to rob some of the ants of their food, tiny fragments of beetles; it was very alert, retiring immediately when any stray ant happened to come its way. Bengalia gaillardi Surcouf was seen in the same locality stealing food carried to the nest by workers of Pheidole liengmei Forel, Camponotus species, Leptogenys stuhlmanni Mayr, Prenolepis longicornis (Latreille), etc.; at Daressalaam this fly was watching for similar pur- poses the home-coming Plagiolepis custodiens (F. Smith). The genus Bengalia is restricted to the Old World tropics and be- longs in the Calliphorine. It differs conspicuously from the othermem- bers of this group in the structure of the proboscis, which is rigid, chitin- ized, strongly toothed at the apex, directed forward, and evidently 11920, Trans. Ent. Soc. London, (1919), Proc., pp. lii-lviii. 1922] Wheeler, Ants of the Belgian Congo 281 adapted to its predaceous habits.’ Bengalia gaillardi Surcouf was ob- ° served by Gaillard at Koulouba, French Sudan, preying upon termites in 7 \ tree stump which had been freshly dug up. G. D. se ate met = - Several species of Rencaica are commonly found in the Belgian go, along paths and roadsides, hunting for various insects. On June Yaa T caught a number of B. spurca Brauer and v. Bergenstamm and van der Wulp* near a column of driver ants in a forest gallery ‘silat for purposes very different from those of Bengalia. In his account of the foraging Eciton of the Amazon, H. W. Bates has this of UM Abtiies of all. Reitons are accourpanied by small swarms of a kind of two- wine fly, the females of which have a very long ovipositor, and which belongs to the é us Stylogaster (family Conopside). These swarms hover with rapidly vibrating at a height of a foot or less from the soil, and occasionally one of the flies darts __ with great quickness towards the ground. I found they were not occupied in trans- fixing ants, although they have a long needle-shaped proboscis, which suggests that conclusion, but most probably in depositing their eggs in the soft bodies of insects, _ which the ants were driving away from their hiding-places. These eggs would hatch ____ after the ants had placed their booty in their hive as food for their young. If this i 4 _ supposition be correct, the Stylogaster would offer a case of parasitism of quite a novel a As kind. ____ Similar observations were made some years later by C. H. T. Town- ___ send in the State of Vera Cruz, Mexico. Under the heading Stylogaster, this author writes? * See Green, E. E., Trans. Ent. Soc. London. 1908, Proe., . PP 2 xxvi-xxvii. Beazi, M. 1913, ‘Einige peoerenentionge 4 uchmeromyia und Bengalia,’ Ent. Mitt. Deutsch. Ent. II, pp: 70-78. 5 agg 1918, ‘The mouth-parts of Ochromyia jejuna, @ predaceous mouscid, ourn. Med. mise I Lop , Pl. xiv. Roubaud, rE. 1914, tudes faune mertsurcout J, Bul Oce, ky Pi ‘ in no whatever to show that the larva of the true B. depressa is a La ol og | Fy #1863, ‘The naturalist on the River Amazon,’ II, 365-366. Bates’ observations on this fly have been by later authors ) toes apparently first makes the statement tes, verfolgt mit ihrer Legerdhre Termiten’’ (1883, Denkschr. A Wien, math eh nr ay . XLV ip D> S4). oda vie ee be ecoouatinn & is unfortunately repeated chr., ¥ nt VI (1903), p. 151), and Fagan by Krober (1010 Are. f. Sosa, LX NUL Ptisih). Abt. A, Heft 8, p. 11). oh not the sl evidence at hand to show that Stylogaster is associated with 71897, Ann. Mag. Nat. Hist., (6) XIX, p. 23. 282 - Bulletin American Museum of Natural History (Vol. XLV Fifty-one specimens of this interesting genus were taken hovering over the front ranks of a moving army of ants, in a cafetal at Paso de Telayo, during the last hour or | two of daylight on March 29. In company with them were numerous.specimens of Hyalomyia and some other small tachinids. The gnts have been determined by Mr. Theo. Pergande as Eciton foreli, Mayr. . .' The column of ants was about 15 feet wide and 25 feet long, and moved slowly but surely in a straight line through the cafetal, swarming rapidly over the thick covering of dead leaves, branches, and other obstructions that strewed the ground under the coffee-trees. The specimens of Stylogaster hovered continually over the ants, now and again darting at them, with- out doubt for the purpose of ovipositing in their bodies. During the whole three months of my collecting in this locality, I saw not a single specimen of Stylogaster at any other time, but on this occasion, during the short time that I had before dark overtook me, I succeeded in capturing fifty-one specimens, by sweeping closely with the net over the front ranks of the ants. From the accounts quoted above it is evident that both Bates and Townsend base their conclusions on mere surmises, since neither of them has succeeded in finding the eggs. Their observations merely show that Stylogaster is in some way associated with the columns of driver ants, though it is by no means certain that this is true for all the members of the genus. Some of the North American species are found as far north as Illinois and New York, in regions where foraging ants are altogether absent. Yet it is possible that the African species of Stylogaster are associated with the columns of the Doryline.2 G. D. H. Carpenter, in Uganda, in his description of the frantic efforts made by cockroaches to escape from the columns of Dorylus, remarks: ‘I twice saw, hovering over these cockroaches, and occasionally suddenly pouncing down (apparently for the purpose of ovipositing) several of a small long-bodied insect—it might have been a dipteron or an ichneumon, but the hover- ing and darting flight suggested rather a syrphid. It was so extra- ordinarily active that I failed to catch it.’ In a recent account of his observations on army ants in British Guiana, Wheeler! observes that although he saw Stylogaster on several occasions accompanying the advancing armies of Eciton burchellii and darting at the ants or even at open spaces on the ground, there was nothing to convince him that these flies were ovipositing. Once he came upon a swarm of both sexes of Stylogaster hovering above a spot where 1According to Wheeler, a synonym of Eciton burchellii Westwood. _ _ *Dr. H. Brauns, of Willowmore, wrote me recently as follows: “ Stylogaster habe ich seiner Zeit einige Male in Westafrika (Cameroon, Gaboon) beobachtet. Die Thiere fielen mir dadurch anf, dass sie wie Falken iber Anomma Zogen schwebten. In welchen Zusammenhang sie mit den Doryliden stehen, weiss ich nicht. Dass sie ihre Eier auf dem Raube von Doryliden, den diese mit sich ablegen sollten ist kaum anzunehmen, da dieser alsbald verzehrt oder verfiittert wird.’’ June 5, 1920). I have collected one of the North American species (Stylogaster neglecta Wi ) on flowers of Clethra alnifolia, Monarda clinopodia, Helianthus strumosus and Eupatorium purpureum. *1915, Trans. Ent. Soc. London, (1914), Proc., pp. eviii-cix. #1921, Proc. American Ac. Arts Sci., Boston, ive. p. 296. a LT Sa ae . Se es ‘ co ¥ Wheeler, Ants of the Belgian Congo 283 there were no Ecitons, although a few workers of Gigantiops destructor _ ___ and Eectatomma ruidum were running about in the vicinity. ‘This ob- servation,” he says, “and the fact that some species of Stylogaster occur ____ in North America north of the range of Eciton, make it seem doubtful ___-whether these flies are as intimately attached to the ants as some authors : Save supposed. They are, perhaps, attracted by the rank odor of the be" a Hymenoptera The following four species of Sphegoidea are the only ones known to on! 4 - provision their nests with ants. It is somewhat surprising that so few ____ predaceous wasps have developed a liking for this kind of prey. _ 1—Tracheloides quinquenotatus (Jurine) (=Crossocerus luteicollis ~ Lepeletier and Brullé; Fertonius formicarius Ferton). This remarkable little wasp is apparently distributed over the entire Mediterranean subregion. Its curious habits were first observed by Ferton' in Algeria. It preys there on the workers of Tapinoma erraticum (Latreille), storing forty to fifty paralyzed ants in each cell; the nest is placed in crevices bs in walls or burrowed to a slight depth in sandy soil. I have frequently ___ observed its hunting behavior in the vicinity of Algiers (June 1910) in the same locality in which I found the pitfalls of Lampromyia described above. The females were hovering over the foraging files of Tapinoma erraticum and would suddenly pounce on one of the ants, seldom missing _ Similar habits were described for this species by Ferton* in Corsica, where also it preys on Tapinoma erraticum. Bignell* likewise found quinquenotatus there, taking small ants which were travelling in a continuous stream across the road. 2.—Tracheloides curvitarsis (Herrich-Scheffer) is only known from southern Germany, Italy and Austria. Emery‘ observed this wasp near Portici and Bologna, Italy, storing about forty partly paralyzed workers of Liomelopum microcephalum (Panzer) in each of its cells which were located in abandoned beetle borings in a tree. 1Under the designation Fertonius luteicollis: Ferton, C., 1890,‘Un Hyménoptére ravisseur de four- mis," Actes Soc. Linn. Bordeaux, XLIV, pp. 341-346 "Under the tion Fertonius formicarius: Ferton, C., 1895, ‘Nouveaux Ryste fouisseurs et observa sur l'instinct de quelques * Actes Soc. Linn. Bordeaux, XLVI L — 261-272. Under that of Brachymerus quinguenotatus: erton, C., 1901, ‘ Notes détachées sur |’ instinet Hyménoptéres melliféres et ravisecura (lére série),’ Ann. Soc. Ent. France, LXX, pp. 83-145, i C., 1900, ‘ Crabro 5-signatus, Jurine, carrying off ants in Corsica,’Ent. Monthly Mag., aoe, c “1891, ‘Zur Bi der Ameisen,’ Biol. Centralbl., XI, pp. 165-180; 1893, ‘Sur un Crabronide chasseur de fourmis,’ Ann. Soc. Ent. France, LXII, Bull., pp. lxiii-txiv. 284 Bulletin American Museum of Natural History [Vol. XLV Tracheloides Aug. Morawitz (of which Brachymerus Dahlbom and Fertonius Pérez are synonyms) is regarded by Kohl in his able Mono- graph of the Palearctic Crabroninz' as a species-group or subgenus of Crabro Fabricius. Only the two species mentioned above are known; they possess a large, much thickened head, with the face strikingly broad below, a peculiarity evidently adapted to their ant-hunting habits, since it makes the jaws with which they seize the ants much more power- ful than is usual among species of Crabro. Indeed, most other members of this extensive genus prey on rather soft-bodied and harmless insects, - chiefly Diptera. 3.—Aphilanthops taurulus Cockerell. Ainslie* found this philanthid wasp preying on the workers of Pogonomyrmex barbatus subspecies rugosus Emery in New Mexico. 4.—Aphilanthops frigidus (F. Smith). This interesting species of eastern North America has been very completely investigated by Wheeler near Boston. Curiously enough, it selects only fertile females, or queens, of ants to provision its nests and seems to restrict its attacks to various species of the genus Formica (Formica fusca Linnzus and its variety subsericea Say; F. pallidefulva Latreille subspecies nitidiventris Emery; and F. neogagates Emery). It forms colonies of from thirty to sixty nests, located in open patches, roads or clearings in woods. The burrow descends with a very steep slope to a depth of six to eight inches, where it terminates in a small cell, there being two or three other cells on the sides. The Formica queens are captured during the short time of their nuptial flight, before they have lost their wings, and are merely stung and paralyzed. The wasp does not mutilate or malaxate her victims, which still move their palpi, legs, and antennz either spontane- ously or when touched, for several hours or even for a few days after they have been captured and placed in the nest. The wasp carries the ant under her body, supporting it by means of her middle and hind legs and holding its antenne in her mandibles. Having dragged the ant a few inches into the burrow, she proceeds to cut off its wings, usually very neatly, although the stubs she leaves attached to the body are a little longer than in queen ants that have dedilated themselves; more rarely the wasp simply gnaws off the tips or apical halves of the wings. Wheeler believes that each female Aphilanthops secures several queen ants, usually five to seven, often belonging to more than one species, and iKohl, F. F., 1915, ‘Die Crabronen der paléiarktischen Region,” Ann. Naturh. Hofmus. Wien, XXIX, pp. 1-453, Pls. 1—xrv. Ainslie, C. N., 1909, ‘A note on the habits of Aphilanthops,’ Canadian Ent., XLI, pp. 99-100. *Wheeler, W. M., 1913, ‘A solitary was: Ae frigidus F. Smith) that provisions its nest with queen ants,’ Journ. Animal Behavior, III, pp. 374-487. Wheeler, Ants of the Belgian Congo 285 _ stores them in two or three cells, from which they are taken as needed to feed a single larva. “The egg is evidently laid on an isolated ant which the mother wasp cuts in two in order that the larva may gain access to _ the nutritious contents of the thorax and gaster. Then the other ants are fk taken from storage and brought to the larva one by one as they are re- = aes till all are consumed and the larva is ready to pupate.”’ _____ Aphilanthops Patton is a strictly Nearctic genus of fossorial wasps, =a of which eleven species have been described, mostly of the western United States. It is highly probable that all will prove to be ant hunters, and + aninteresting field of study is here open to the myrmecologist ! _--‘The prey of Polybia scutellaris (White), a social wasp of southern Brazil, consists mainly of winged termites, which are stored whole in the nest, often by the hundreds; but occasionally this wasp collects winged male ants too. In one case about a hundred males of Dorymyrmex __ pyramicus (Roger) and a few other male ants were found in its nest.’ Ty 2. ; AMPHIBIANS® The diet of many amphibians consists almost exclusively of various arthropods. Only living and moving prey is devoured; dead or motion- less food has little or no attraction for them. In the frogs and toads the tongue, attached in front and free behind, is often the chief organ used in ___ seizing the food, being thrown out with lightning-like rapidity; it is soft, extensile, coated with a glutinous secretion, and adheres firmly to the prey, which is swallowed whole. The teeth, when present, are used only for catching and holding the prey; they are absent in many genera. Digestion is very rapid. The American toad, Bufo americanus Hol- brook, for instance, feeds continuously throughout the night, except when food is unusually plentiful; in twenty-four hours it consumes a quantity of insects equal to about four times its stomach capacity. In other words, the toad’s stomach is practically filled and emptied four times in each twenty-four hours.* Se ee ea eS ee) eae ae ee eee 'The of the genus Microbembez store dead insects in their nests, a very unusual procedure atte wasps; can occasionally be seen collecting dead ants ‘that have been thrown out entrance of ant nests. Parker, J. B., 1917, Proc. U.8. "Nat. Mus., LI, pp. 134-141. ‘asmann, E., 1897. *Beutethiere von Polybia scutellaris (White) Sauss.,’ Zool. Anzeiger, XX, pp. 41 am under great obligation to Mr. C. L. Camp for valuable suggestions on the subject of ants as food | batrachians and 4 ys wi = 1808. “Usefulness of the American toad,’ U. 8. ve ta Agric. Farmers’ Bull. No. 196, p-6 merece th changes from Bull. 46, Hatch Exp. Stat. A t, Mass., 1897, pp. 1-30, tee tchis A. 5., 1869, ‘The toad as an entomologist,’ Canadian Naturalist, N. 8., IV, pp 74-178: Garman t. 1901, “The food of the toad,’ pang Agric. Exp. Stat. Bull. 91; Hodge, ioe, The sermon (Worester, Mama) torr 7 1 gous, the, colori oad Ae niv ourn ic. 80-01; Munz, P. A., 1920, ‘A St a see hg ple species of Anura during foonalocuetion,’ Pomona Journ. Ent. Zool., it pp. 33-56. 286 Bulletin American Museum of Natural History [Vol. XLV Toads and frogs being more often seen while in search of good, the stomach contents of specimens in collections are frequently little or not at all digested and can then be easily identified; many insects with hidden habits may thus be obtained. Amphibians are in this respect of very great help to the-collector of ants. Numerical data relating to the food of these animals has not often been published, even for the species of temperate regions. Perhaps the most complete records of the kind are those in H. A. Surface’s ‘Report on the economic features of the amphibians of Pennsylvania.’ From this paper it may be seen that, while almost all salamanders, toads, tree- - frogs, and frogs occasionally eat ants, these insects constitute an aS ant item in the diet of certain species. Food of Certain Amphibians in Pennsylvania (Surface) \Number of Stomachs Number of Niansban cail iatuate with (Specimens A E ae Recogniz- | Eaten nts Ants able Food Plethodon cinereus (Green) 260 583 182 30 = glutinosus (Green) 125 367 63 17 Desmognathus fusca (Rafinesque) 235 378 33 8 Bufo americanus Holbrook 52 150 20 13 Kirkland’s paper referred to above contains the result of an examina- tion of 149 stomachs of toads (Bufo americanus Holbrook) in Massachu- setts; in this case 19 per cent of the total contents were ants; the percentage was higher in May, when ants formed 23 per cent of the food and were present in 70 per cent of the stomachs. The Texan robber frog, Eleutherodactylus latrans (Cope), a land animal of secretive and nocturnal habits, probably feeds extensively on ants. J. K. Strecker® mentions that “‘the stomach of one example con- tained the elytra of a ground beetle and the remains of many spiders and = ants.” True frogs of the genus Rana take very few or no ants, at least in North America, though, as may be seen below, the stomachs of certain * of the African species contain a fair proportion of these insects, mostly in the winged phases. Surface, in Pennsylvania, found few or no ants in- Rana, and this result is confirmed by C. J. Drake’s very extensive study 11913, Zool. Bull. Div. Zool. Pennsylvania Dept. Agric., III, Nos. 3 and 4, pp. 134-147. 21910, Trans. Ac. Sci. St. Louis, XIX, No. 5, p. 82. — Wheeler, Ants of the Belgian Congo 287 iachs examined only 19 contained one or twospecimens of ants ny Ra ‘cent of the total animal tood). J.C. Needham obtained similar : s with the bullfrog (Rana catesbiana Shaw) in New York State; mal s of Camponotus pennsylvanicus, which evidently had dropped on t e surface of the water, where they were taken by the frogs.’ In his paper on Nicaraguan amphibians, G. K. Noble? mentions ants among the stomach contents of the following species. __ Dendrobates tinctorius (Schneider). The stomachs of two speci- mens “contained mostly ants, although a few beetles and other insects : were present. There were about fifty antsin each stomach. Dr. Wheeler _ has identified most of these as Wasmannia auropunctata Roger. Seven <) other genera were represented, but each by only a few workers: Strumi- -genys, 2 species; Rhopalothriz, new species; Leptogenys (Lobopelta), _ species; Trachymyrmex, species; Ponera, species; Pheidole, 2 species; aan See Belenopeis, species.”” (op. cit., p. 322). ae _Dendrobates typographus Keferstein. The stomachs examined con- “ae tained “mostly small red ants.”’ (op. cit., p. 323). _-—s- Eleutherodactylus polyptychus (Cope). The stomachs “contained : only insets and mostly large ants.”’ (op. cit., p. 329). _——— Eleutherodactylus rugosus (Peters). “One medium-sized individual a ‘contained in its stomach two beetles and a large ant, Neoponera obscuri- ____ eornis (Emery).” (op. cit., p. 331). Bufo hematiticus Cope. “Tt had been feasting on ants. Itsstomach ___ contained a great many large red and black ones. The following species __-_ were represented in the contents of this single stomach: ‘Pachycondyla __ harpazx F. (4), Ectatomma ruidum Rog. (1), Eciton hamatum F. (3), _ Alta cephalotes (4), Apterostigma species (1).’’ (op. cit., p. 333). . Bufo coniferus Cope. “A few ants contained in one stomach belonged to Paraponera clavata F., Neoponera obscuricornis Em., A pterostigma species and Hylomyrma species.” (op. cit., p. 334). | Hyla quinquevittata Cope. ‘The largest specimen contained in its stomach over a dozen termites and one ant (Tetramorium guineense Fabr.)”’ (op. cit., p. 341). It is evident that many more observations are needed before we can fully realize the part amphibians play as predaceous enemies of ants. ‘Needham, J2 (2, 1008, The summer foo ofthe bull the bullfreg (Rana calesbiana Shaw) at Seranae Inn,’ New York State Mans . 86, pp. 9-15. 71018, ‘The amphibians collected by the American Museum Expedition to Nicaragua in 1916,’ Ball. American Mus Net. Hit, XXXVill. pp siL.o,fe ae 288 Bulletin American Museum of Natural History (Vol. XLV Enough i is known, however, to make it certain that these animals are of prime importance in this respect. While studying the amphibians of the Lang and Chapin collection, . Mr. G. K. Noble, Assistant Curator of Herpetology at the American Museum, dissected the stomachs of a large number of specimens and has turned their contents over to me for identification. The results of these examinations will be published in detail in Mr. Noble’s report. From the point of view of the myrmecologist they were of great interest, yielding a large number of remarkable forms; eighty different species, subspecies, and varieties were obtained in this way and, of these, forty werenotother- wise represented in the collection upon which Prof. Wheeler’s report is based; seventeen of these forms were new to science. Many of the ants found in the stomachs of amphibians are in an excellent state of preserva- tion; others are considerably improved by a thorough cleansing with caustic potash. Future collectors in tropical countries are urged never to neglect this novel manner of increasing their material. In the table below, I have condensed the results of the examina- tion of 308 stomachs of the eleven species of Congo frogs and toads which apparently show a decided preference for ants; for five of these species ants constitute about 50 per cent or more of the total stomach contents. In addition, several species of Congo frogs had eaten isolated specimens of ants, which may, in some cases, have been swallowed accidentally together with mud, dead leaves, or vegetable matter, an abundance of Stomach Contents of Congo Amphibians Number of Stomachs Total with |Number of Number of} Per Cent Recogniz- | Insects Ants of Ants able Food| Eaten Bufo polycercus Werner 53 759 406 66.66 “ tuberosus Giinther 5 160 38 23.73 “* funereus Bocage 55 1292 7 54.56 “ regularis Reuss 31 963 484 50.25 * superciliaris Boulenger 50 746 182 24.39 Rana occipitalis Giinther 25 55. ° | 14 25.45 * ornatissima Bocage 14 30 ) 6 20. “* albolabris Hallowell 19 oe. 25 59°28 “ mascareniensis Duméril and | | Bibron Ss 40 il 27.50 Hemisus marmoratum (Peters) | 22 ) 1006 96 9.54 Phrynobatrachus natalensis(A Smith), 10 ) 47 20 42.55 : \ Wheeler, Ants of the Belgian Congo 289 “which is often found in the stomach. For other species, however, the } ‘h b number of stomachs examined was too small to furnish reliable data; when more completely investigated, some of these may prove to be true ant-feeders. A number of amphibians collected by the American Museum Congo “oe Expedi ition and the forms of ants which could be identified by Prof. Wheeler among their stomach contents are listed below. Such recerds _ give an insight into the great variety of ants eaten by some of these _ animals and also, to a certain extent, into the preferences shown by in- _ dividual species. I must, however, point out that much of the ant _ débris found in the stomachs was too poorly preserved to permit correct identification, at least with our present knowledge of African myrme- cology. These lists could, therefore, be considerably lengthened. ___ Nevertheless, in the case of the toads a sufficient number of specimens have been examined to show that ants are a very important article in their diet; a total of 1815 ants was found in 194 stomachs of the five species of Congo toads; these ants belong to 72 forms, six (or 8 per cent) of which are Dorylinz, thirty (or 42 per cent) Ponerine and Cera- pachyine, sixteen (or 22 per cent) Myrmicine, and nineteen (or 27 per cent) Formicine. Terrestrial ants seem to be taken almost exclu- sively and this fact undoubtedly accounts for the high proportion of the Ponerine represented. Xenopus miilleri (Peters) A common frog of the Sudanese and East African savannahs. Of ten stomachs examined, only one contained a single ant: Polyrhachis aérope Wheeler. Xenopus tropicalis (Gray) A frog confined to the Rain Forest. Of eleven stomachs examined, two together contained five ants: Camponotus pompeius subsp. marius Emery. Bufo regularis Reuss This widely distributed African toad occurs in the forest and in the savannah as well. Of thirty-eight stomachs examined, thirty-one showed recognizable food and nineteen of these contained ants: Dorie nigricans subsp. burmeisteri (Shuckard). ¥ subsp. sjestedti Emery. Platythyrea gracillina Wheeler. Megaponera fatens (Fabricius). Bothroponera soror (Emery). 290 Bulletin American Museum of Natural History [Vol. XLV , : Anochetus bequaerti Forel. Phéidole kohli Mayr, var. - megacephala subsp. melancholica (Santschi). Myrmicorta eumenoides subsp. opaciventris (Emery). var. crucheti (Gantechi),. Crematogaster excisa (Mayr). Monomorium bicolor Emery. e afrum var. fultor Forel. Tetramorium guineense subsp. medje Wheeler. Plagiolepis tenella Santschi. Camponotus maculatus (Fabricius). ~ subsp. congolensis Emery. . “ subsp. solon Forel. - - subsp. brutus (Forel). ey acvapimensis Mayr. <3 rufoglaucus subsp. cinctellus var. rufigenis Forel. e chapini Wheeler. * polyrhachioides Emery. Polyrhachis militaris (Fabricius). Bufo funereus Bocage This toad is commonly found in the Rain Forest and the pore forest galleries. Of sixty-three stomachs examined, aed contained — recognizable food and forty-three of these ants: Dorylus emeryi subsp. opacus Forel. “ kohli Wasmann “nigricans subsp. arcens (Westwood). “ _ wilverthi Emery. Cerapachys cribrinodis Emery. Paltothyreus tarsatus (Fabricius). Megaponera fatens (Fabricius). Bothroponera talpa Ern. André. ig pachyderma (Emery). nd soror (Emery). Phrynoponers gabonensis (Ern. André). ’ var, esta Wheeler. as “ var. fecunda Wheeler. - i var. striatidens (Santschi). Euponera ingesta Wheeler. “ sennaarensis (Mayr). Plectroctena cristata Emery. Psalidomyrmez procerus Emery. Leptogenys stuhlmanni subsp. camerunensis var. opalescens Wheeler. Anochetus estus Wheeler. " opaciventris Wheeler. Odontomachus assiniensis Emery. “ S var. furvior Wheeler. ¥ hematoda (Linnzus). ' 291 Bufo tuberosus Giinther forest toad, much less common than the other species. Only chs a be examined and each contained a number of worker tareatus (Fabricius). : Bufo polycercus Werner er One of the three common forest toads of the Congo. Of the fifty- al i cnienate/tidems (Fabricius). ~ Bethroponera pachyderma (Emery). : var. funerea Wheeler. y talpa Ern. André. * soror (Emery). Bulletin American Museum of Natural History (Vol. XLV Phrynoponera gabonensis (Ern. André). me var. esta Wheeler. “ ad var. fecunda Wheeler. “ var. umbrosa Wheeler. a6 S var. striatidens (Santschi). Euponera ingesta Wheeler. e subiridescens Wheeler. pS minor Emery. . Psalidomyrmezx procerus Emery. * reichenspergeri Santschi. Leptogenys stuhlmanni subsp. camerunensis var. opalescens Wheeler. S ergatogyna Wheeler. Anochetus opaciventris Wheeler. Odontomachus assiniensis Emery. = ns var. furvior Wheeler. Pheidole batrachorum Wheeler. “« megacephala (Fabricius). Crematogaster concava Emery. Triglyphothriz gabonensis Ern. André. Cataulacus guineensis F. Smith. Engramma wolfi Forel. _ Plagiolepis tenella Santschi. Pseudolasius weissi var. sordidus Santschi. Ly bufonum Wheeler. Camponotus maculatus subsp. solon Forel. " subsp. brutus (Forel). pompeius subsp. marius Emery. _ vividus subsp. cato (Forel). Polyrhachis militaris subsp. cupreopubescens var. nkomoénsis Forel. - nigrita Mayr. “ Bufo superciliaris Boulenger A common toad of the Rain Forest in Cameroon, Gaboon and the Congo. Of fifty-six specimens examined in this respect, fifty showed — recognizable remains of food in the stomach and thirty-five ot these con- tained ants: Dorylus nigricans subsp. sja@stedti Emery. Paltothyreus tarsatus (Fabricius). Megaponera fetens (Fabricius). Bothroponera talpa Ern. André. ne pachyderma (Emery). ‘4 soror (Emery). Paryeouoners gabonensis var. esta Wheeler. Cy> eae fecunda’, Wheeler. ” bequaerti Wheeler. Euponera subiridescens Wheeler. — — 1922} —S~S*é«SwWiheeder, Ants of the Belgian Congo 293 _ Pleetroctena cristata Emery. ay Psalidomyrmex procerus Emery. e ve obesus Wheeler. een Odontomachus assiniensis Emery. igs “var, furvior Wheeler. ¥ Myrmicaria eumenoides subsp. opaciventris Emery. _ Tetramorium guineense (Fabricius). "ee : Camponotus pompeius subsp. marius Emery. oe wellmani var. rufipartis Forel. Phrynobatrachus perpalmatus Boulenger Hc .% water frog of the forest region. Only eight of the stomachs ex- : ‘enkiod contained recognizable remains of food and ants were found in is one of these: _ Polyrhachis alluaudi var. anteplana Forel. Parts of many more ants were seen in the stomachs of the related __ savannah species Phrynobatrachus natalensis (A. Smith), but too poorly ‘Sue + calla for correct identification. . Arthroleptis variabilis Matschie . - This is one of the typical frogs of the Cameroon, Gaboon and Congo ; Rain Forest. Seventeen of the stomachs examined contained recogniza- ble food and two of these included ants: __ Pheidole batrachorum Wheeler. Aéromyrma sp. Rana occipitalis Giinther A large-sized frog, common in the vicinity of streams, ponds, and swamps throughout the forest and savannah, from Senegambia to Angola, ‘Uganda and East Africa. All of the twenty-five stomachs examined, contained recognizable food and ants were present in ten cases: Dorylus nigricans subsp. sjastedti Emery. Paltothyreus tarsatus (Fabricius). Megaponera fortens (Fabricius). Bothroponera soror (Emery). Odontomachus hematoda (Linnzeus). Myrmicaria eumenoides subsp. opaciventris (Emery). Camponotus maculatus subsp. congolensis Emery. ¥ acvapimensis Mayr. $ cxsar Forel. ¥ wellmani var. rufipartis Forel. by chapini Wheeler. Rana albolabris Hallowell A characteristic frog of the Rain Forest, extending a little beyond the limits of the forest in swamps and along forest galleries. The 294 Bulletin American Museum of Natural History [Vol. XLV stomachs of twenty-three individuals were dissected and nineteen of these showed recognizable remains of food; ants were present in — cases only: . Bothroponera pachyderma (Emery). Odontomachus assiniensis var. aterrimus Wheeler. Camponotus maculatus subsp. solon Forel. Rana mascareniensis Duméril and Bibron + Perhaps the most common frog throughout the larger part of the African continent. Of the thirty-nine stomachs examined, twenty-four contained recognizable remains, and a small number of ants, all of the winged phases, were found in five of them: Pheidole megacephala (Fabricius). Camponotus vividus subsp. cato (Forel). Rana ornatissima Bocage This frog is much rarer than the three preceding species; it is known from the savannahs south of the Rain Forest, from Angola to Southern Rhodesia and also from the northeastern Uele, where Lang and Chapin collected a number of specimens at Garamba. Of these, fifteen were examined for their food contents and fourteen contained Bina: remains; a few ants were found in a single stomach: Pheidole kohli Mayr, var. “ megacephala subsp. melancholica (Santschi). - speculifera Emery. In addition, twenty stomachs of two other common Congo frogs (Rana oxyrhynchus A. Smith and R. christyi Boulenger) were dissected, but only a single winged ant was found. The pronounced aquatic habits of all species of Rana, which keep them in or near the water, evidently prevent them from feeding to any large extent on ants, except on in-. dividuals that accidentally drop into the ween as for instance, during — their nuptial flights. Kassina senegalensis (Duméril and Bibron) A small frog occurring throughout the savannah country of Africa, with rather terrestrial habits and also said occasionally to ascend trees. A few ants were found in two of the nineteen stomachs dissected. The occurrence in one stomach of a number of workers of the hypogeic ant Dorylus kohli is interesting in connection with the burrowing habits of this frog. Dorylus kohli Wasmann. Pheidole megacephala subsp. melancholica (Santschi). Wheeler, Ants of the Belgian Congo 295 Hemisus marmoratum (Peters) This little burrowing frog, of pronounced terrestrial habits, is found savannah country of a large part of Africa, north, south, and east » Rain Forest. It lives mostly underground, and, according to Mr. s observations, comes out of its burrows only after heavy rains. It le most typical “ant-eater” of all Congo amphibians; twenty-two machs examined contained no other food than termites and worker en termites were by far more abundant. True ants were found Ni Bisa often chew or lacerate their food to such an extent that the examination of their stomach contents gives but very general indications with regard to their diet. There can hardly be any doubt, however, that : are part of the bill of fare of many of these reptiles. In Miss ne. HH. Pritchett’s careful experiments,' ants, Pogonomyrmex barbatus a subspecies molefaciens (Buckley) and Pachycondyla harpax (Fabricius), were eaten readily by Sceloporus spinosus floridanus (Baird), a common = lizard of Texas. Another species, Gerrhonotus infernalis Baird, refused to eat Camponotus maculatus subspecies sansabeanus (Buckley) and C. ___ fumidus variety festinatus (Buckley), but the author suggests that these ants were possibly too small to be noticed, as insects below a certain tr size are apparently not perceived by the large species of lizards. Con- a eerning Phrynosoma cornutum (Harlan), Miss Pritchett writes: “The ___ *horned toads’ were kept in cages with other lizards and also separately ¥ _ and were never seen to eat anything but ants. They are especially fond aes the large agricultural ant, Pogonomyrmex barbatus Smith variety ___-molefaciens Buckley” (p. 284). ____Inhis paper on ‘The horned lizards of California and Nevada of the genera Phrynosoma and Anota,’ H. C. Bryant* says that ants, flies, and ___ other insects constitute the principal diet of these genera and remarks: _ “Why the animal is never bothered by being stung internally by the ants it eats, seems hard to explain. Certainly the lining of the mouth and stomach must be particularly adapted to withstand the poisonous sting of insects, for when stung externally, the lizard shows no little dis- comfiture”’ (p. 17). Unlike most other reptiles, the horned toad catches 11903, Biol. Bull., Vv, pp. 271-287. 1911, Univ. of California Publ. Zool., IX, No. 1, pp. 1-S4, Pls, t-rx. 296 Bulletin American Museum of Natural History [Vol. XLV the insects on the end of its viscid tongue and swallows them alive, its feeding habits being indeed very similar to those of true toads. C, L. Camp! has published more detailed observations on the food of many California lizards. He found remains of ants in the stomach contents of the following species: Uma notata Baird, Callisaurus ventralis ventralis (Hallowell), Uta stansburiana elegans (Yarrow), Sceloporus magister Halle- well, Phrynosoma platyrhinos Girard, and Cnemidophorus tigris tigris Baird and Girard. In the case of one of the horned toads (Phrynosoma) examined, the contents of the stomach were: “fifteen parasitic nema- todes, six Coleoptera, one orthopter, 145 red-headed ants, all apparently of the same species and swallowed whole, and one pebble” (p. 528). These ants belonged in all probability to one of the’seed-storing species of Pogonomyrmex, for Mitchell and Pierce? also note that in Texas re- mains of P. barbatus (F. Smith) subspecies molefaciens (Buckley) were found several times in the excrement of the horned toad, Phrynosoma cornutum (Harlan), and “one colony was absolutely exterminated before the enemy left it.” The Australian horned dragon or moloch (Moloch horridus) i is said by Saville Kent to feed exclusively on ants of the minutest size. The small black evil-odored species [of ant], common in both South and Western Australia, was always a prime favorite with the specimens kept by the author, and wherever these ants abounded, in conjunction with a sufficiently warm temperature, no difficulty was experienced in maintaining these lizards in perfect health. . . . . They would soon settle down to feeding in a row, and the number of ants an individual lizard would assimilate was something astonishing On several occasions experimental reckoning elicited the fact that no less than from one thousand to fifteen hundred ants were taken in successive order at a single meal, each ant being separately picked up by a flashlike protrusion of the slender adhesive tongue.’ On examination of the stomachs of the lizards and chameleons col- lected by the American Museum Congo Expedition, Mr. K. P. Schmidt* found remains of ants, usually in a condition preventing any further identification, in the following species: Lygodactylus picturatus gutturalis (Bocage), Agama colonorum Daudin, Bedriagaia tropidopholis Boul- enger, Algiroides africanus Boulenger, Holaspis guentheri Gray, Ger- 11916, ‘ Notes on the local distribution and habits of the amphibians and reptiles of Southeastern California in the vicinity of the Turtle Mountains,’ Univ. of California Publ. Zool., XII, No. 17, pp. 503- 544, Pls. x1ix-xxu. Through the co of Mr. Camp I have been able to examine the sto: - tents of a number of reptiles collected by him near the ‘Turtle Mountains, Riverside Co., California. In the case of Une netate the momachs were alsiost entirely Bied with heads and parts of the body of bn ies ex, while in those of Phrynosoma platyrhinos there were heads of ants and also pieces of 21912, Proc. Ent. Soc. Washington, XIV, p ‘Saville Kent, W., 1897, ‘The naturalist in n Australi,” (London), pp. 85-86. 41919, {Contributions to the herpetology of the Belgian Congo based on the collection of the American ‘Museum Co: 7 oe a. 1909-1915. Part I. Turtles, ¢ Crocodiles, Lizards, and Chame- leons,’ Bull. American \ at. Hist., XX XIX, pp. 385-624. _ Wheeler, Ants of the Belgian Congo 297 Satie favigularis nigrolineatus (Hallowell), Mabuya polytropis Boul- , Chamezleon gracilis (Hallowell), and C. ituriensis K. Schmidt. The African skinks of the genus Mabuya feed on a great variety of ects, but certain of the forest species often follow the columns of the ver ants (Doryline). Sjéstedt has observed this in Cameroon with uya raddoni (Gray). He says: _ This lizard is one of the most diligent persecutors of driver ants, and wherever of their columns was seen on the move in or at the margin of the forest, especially ” ‘the ants had scattered in search of food, one could be sure to find one or more of a animals preparing for an excellent catch. . It was a delight to observe how adroitly. the agile lizards would plunge into the crawling swarm, fill their mouth with ants and then retire to a place of safety to devour their booty. Busily engaged in their hunt, they would fearlessly run about the motionless observer and not even ) to climb - vteais always twinkling their lively little eyes, on the lookout Birps ca con aet Te ca ratively few birds of temperate regions have acquired a fond- ; - _ ness for ants; for instance, of fifty species in Tyrol whose stomach con- __ tents were examined by Dalla Torre,’ only the following five contained remains of ants to any extent: Cypselus melba (Linneus), Anthus arboreus Bechstein, Tetrao medius Meyer, Colymbus cristatus Linnzus, and Picus __ viridis Linnzus. In the case of the green woodpecker, the stomach was erammed full of Formica rufa and Lasius niger. Newstead* found that _ the European song-thrush, Turdus musicus Linneus, and the blackbird, _ Merula merula (Linnzus), occasionally eat worker ants. Many insectiv- orous and omnivorous birds undoubtedly have similar habits.‘ a A great fund of accurate information concerning the food of Hun- __ garian birds is contained in a series of articles by E. Csiki.’ The identifi- cation of the insects found in each bird stomach examined is given and also the number of specimens of each species. I have been able to con- sult only the first eight papers of the series (1904-1912), in which fifty- six species of birds are studied in this respect. All of them, however, are _ insectivorous or at least carnivorous, with the exception of the common Sees rene Gromake Vet. Ak. Handl., XXIII, pt. 4, No. 2, pp. 14-15 K. W. v., 1889, ‘U ntersuchungen ber den ageninhalt verschiedener Végel,’ Biol. Centralbl., i Vt, pp. 4 473-480. *Newstead, R ‘The food of some birds inhabiting an orchard,’ Gardeners’ Chronicle, (3) 197 and 217-218. MANE Nictok (1000; ‘American epiders and the Ay’ Th pp. 201-308) gives some sdaitionl evidence of birds fantinu on ant. See Bee also Cockerell, £'D. A., 1800, ‘What are the uses of aes py menopters . 65-68. *Caiki, E. ‘ Positive ee it beh g her rgh ag Agni, Pudapest 1904, XT, pp. S- 317; 1905, XU, pp. 312-330; oe oat RP. 48-161; 1907, XI 188 1 908, Vp, 183- ;,1909, XVI, pp. 130-144; pp. 205-218; 1013 XX, pp. 875-306; 1014, XX1. 220. Vir, too 14a; i Rebhuhne (Perdis perdis L), ibid i912, XIX, pp. 202-208 298 Bulletin American Museum of Natural History [Vol. XLV gray partridge, Perdix perdix (Linnzus), which is chiefly a grain-feeder. Of the fifty-six birds, forty-nine showed ants in their stomach contents, but in the majority of cases these insects were present in isolated speci- mens only. The following eight birds, alone, evidently exhibited a true myrmecophagous .propensity : Dryobates major (Linneus). The greater spotted wooded’ is a typical ant-feeder; of twenty-three stomachs examined, fifteen contained ants, often in large numbers, belonging to the following six species: Lasius flavus, L. niger, L. fuliginosus, Camponotus ligniperdus, Formica rufa, and Dolichoderus 4-punctatus. . Dryobates medius (Linneus). The middle spotted woodpecker feeds also largely on ants; of nine stomachs, six contained such insects, also often in abundance. The following species were recognized: Lasius fuli- ginosus, L. alienus, Formica rufibarbis, F. rufa, and Myrmica levinodis. Dryobates minor (Linneus). Ants are also readily eaten by the lesser spotted woodpecker; five of the eight stomachs examined con- tained specimens, often in great numbers, of the following species: Lasius alienus, L. fuliginosus, Camponotus sylvaticus, and Dolichoderus 4- punctatus. Picus viridis Linneus. The main food of the green woodpecker consists of ants, which were present in all of the twenty stomachs analyzed, often the only contents and in considerable quantities (as many as 500 or 600 specimens in a single stomach). Ten species of ants were recognized: Lasius alienus, L. flavus, L. fuliginosus, L. niger, Formica pratensis, F. rufa, F. rufibarbis, Camponotus vagus, Myrmica levinodis, and Aphzxnogaster structor. Picus canus Gmelin. Only ants were found in the stomachs of the three specimens of the gray-headed green woodpecker examined; they belonged to five species: Lasius alienus, L. flavus, Formica rufa, F. rufi- barbis, and Camponotus vagus. Dryocopus martius (Linneus). There were ants in five of the six stomachs examined of the great black woodpecker; often in abundance and of three species: Lasius alienus, Camponotus ligniperdus, and C. vagus. Jynz torquilla (Linneus). The wryneck subsists chiefly on ants; all the eighteen stomachs examined contained these insects, often in large numbers, six species being represented: Lasius niger, L. alienus, Formica rufa, Camponotus sylvaticus, Myrmica levinodis, and Tetramorium cespitum. Wheeler, Ants of the Belgian Congo 299 seeds and other vegetable substances, but it frequently picks up nals of various kinds. Of the 285 stomachs examined by Csiki, 177 r61.1%) also contained insects. The bulk of this insect food seems to have consisted of ants, which were found in 134 stomachs (or 47%), often in great quantities. Lasius alienus was present in 72 cases; . L. niger in 57 cases; Formica rufain 11 cases; F. pratensis in 2 cases. All European observers agree that the green woodpecker, Picus iridis Linnzus, is one of the foremost ant-feeders. According to Was- mann’s! observations in the Netherlands, this bird does not merely limit its myrmecophagous appetite to wood-boring ants (Camponotus), but frequently burrows into the nests of certain terrestrial species. In the _ spring and fall the excrement contains remains of many kinds, such as — Myrmica rubra, M. scabrinodis, Lasius niger, L. flavus, L. fuliginosus, = Formica pratensis, F. rufa, F. rufibarbis, and F. sanguinea, while in severe __ winters this woodpecker seems to feed almost exclusively on Formica rufa and F. pratensis, inserting its bill into their mound-shaped nests. ae We +30 Angus? also found that the stomach of one of these woodpeckers, shot in January in North Wales, contained Myrmica scabrinodis, “a common ant which nests on ground-hillocks, but never in trees.’”* ‘The very complete inventory of the food of the woodpeckers and their allies (Picid) in the United States published by Beal has led to the interesting results contained in the table below, in which the species __ are arranged in the order of their importance as ant-eaters. It may be seen that, for these birds, “ants constitute the largest item of animal - food—28.41 per cent, considering the whole 16 species collectively—and are actually the largest item in the stomachs of 8 species. The William- son sapsucker, the red-cockaded woodpecker, and the two flickers take __ the highest rank in this respect. Beetles stand next in importance, and amount to 20.42 per cent. These two items together form nearly half the _ food. The remainder of the animal food is composed of insects, with a few spiders, millepeds, and sowbugs, and occasionally a salamander, tree frog, lizard, or snail.’’* tWasmann, E., 1905, ‘Zur Myrepecephapie des Gromepednie: Tijdschr. v. Ent., XLVIII, pp. 214- 220. Wasmann likewise observed the h, Fringilla calebs rf bares boring into a smal) nest of a of which it ag? : raena coovone aad workers ae wel well, Proc. Nat. 7° oe tz, DOG, Uekor dis wi state hitches Bedeutung unserer Spechte,' Verh. Ornithol. Ges. Bayern, (1004 ag ». pi et #6. E.L., 191 “Food of the woodpeckers of the United States,’ U. 8. Dept. of Agric, Biol Surv., 37, p. 10. Seealso ¥. Riley s.cccount of Insects that woodpeckers eat’ in Warren, B. H., 1890, * Soper ce ho bade of Peceeyi 2d. Ed., (Harrisburg), pp. 176-178. 300 Bulletin American Museum of Natural History [Vol. XLV | Food of North American Picidse Number of} Per cent | Name of Species Stomachs | of Animal | Per cent — Examined| Food of Ants Williamson sapsucker, Sphyrapicus thyroideus pha (Cassin) 17 86.67 85.94 Red-cockaded woodpecker, Dryobates borealis by ie (Vieillot) 76 81.06 56.75 | Red-shafted flicker, Colaptes cafer (Gmelin) 183 67.74 53.82 Flicker, Colaptes auratus (Linneeus) 684 60.92 | 49.75 Red-breasted sapsucker, Sphyrapicus ruber ny ae (Gmelin) 34 68.92 42.49 Pileated woodpecker, Phlaotomus pileatus Se. (Linneus) 80 72.88 39.91 Yellow-bellied sapsucker, Sphyrapicus varius , ag (Linnzeus) 313 49.31 34.31 Downy woodpecker, Dryobates pubescens (Lin- ‘ nzeus) 723 76.05 21.36 Hairy woodpecker, Dryobates villosus (Linnzeus) 382 77.67 17.10 Lewis woodpecker, Asyndesmus lewisi Riley 59 37.48 11.87 Three-toed woodpecker, Picoides americanus Brehm 23 94.06 8.29 Nuttall woodpecker, Dryobates nuttallii te (Gambel) 53 79.41 8.19 California woodpecker, Melanerpes formicivorus bairdi Ridgway S4 22.59 8.09 Red-bellied woodpecker, Centurus carolinus (Linnzus) 271 30.94 6.45 Arctic three-toed woodpecker, Picoides arcticus ; (Swainson) ‘ 28 88.69 6.35 — Red-headed woodpecker, Melanerpes erythro- cephalus (Linnzeus) 443 33.83 5.17 Total 3453 Average 64.26 28.49 It would be worth while to consider in more detail the choice of food — made by these ant-eating woodpeckers. Unfortunately, I have not found the needed information for some of the species included in the above list, such as, for example, the Williamson sapsucker; many of the other woodpeckers, especially those of the genera Dryobates, Phleotomus, and Melanerpes, merely eat ants which they find in wood or underneath bark (Camponotus and Crematogaster). The flickers (Colaptes), however, are the ant-eaters par excellence among North American birds, for they have made ants their favorite food; they are also more terrestrial in - 1922) Wheeler, Ants of the Belgian Congo 301 , Prenolepis, ete.). | ants (Crematogaster species). The whole mass was divided with care into 16 arts as nearly equal as possible, and in one part 315 ants were counted, giving 5,040 n one meal of one flicker. In addition there were at least 100 pupe. Two other " } and crops examined in the same way each gave a little over 3,000 ants. ably each of the 100 stomachs in the collection contained nearly as much ant oe ates dom, but the number of ants was less because they were of larger species. A ae e proportion of the ants eaten are of species that live in the earth, and these appear ; to be the principal food the flicker obtains on the ground. In every case where the stomach held a quantity of these small ants, a lot of fine sand revealed their source.' ___ In his study on ‘The tongues of woodpeckers,’ F. A. Lucas has the ; nuloed interesting remarks which may be quoted in connection with a Wiclansidering the tongues in relation to food, we find that those of the various i Sicias of flickers (Colaptes) have the fewest terminal barbs and the longest dorsal tract of fine points; they are also among the longest. The members of the genus are particularly fond of ants, and the tongue seems especially adapted for probing ant hills. oo ‘The function of the fine points on the upper part of the tongue seems to be to form a rough surface to which the sticky saliva will readily adhere and to which in turn the ants will be stuck. In this genus the submaxillary salivary glands reach the maximum size in the group.* In North America the western meadowlark, Sturnella magna neglecta (Audubon), and the roadrunner, Geococcyx californianus (Lesson), may __ be taken as typical illustrations of occasional ant-feeders. The food of _these birds has been investigated in California by H. C. Bryant.’ About 2000 stomachs of the western meadowlark were examined, and 16.7 per _ _ ¢ent of these contained remains of ants, which amounted to 3 per cent (volume) of the total food of all the specimens studied. Ants appear to _ be taken by this bird irrespective of size or kind. Of species identified, I may mention Tapinoma sessile, Messor andrei, Pogonomyrmex cali- fornicus, and species of Camponotus and Formica. In the case of the roadrunner, of which 84 stomachs were examined, a little over 4 per cent of the total food was made up of ants, bees, and wasps; one of these stomachs contained over 250 red ants (Pogonomyrmex californicus), along with a quantity of caterpillars, crickets, beetles, and peers \Beal, y E. 11905, U. 8, Dept of Rane. Biol. Surv., Bull. 7, p. 38. *1014, * A yp bh of the economic status of the western meadowlark (Sturnella neglecta) in ’ Univ. of California Publ. Zool., XI, No. 14, pp. 21-24; 1916, ‘Habits and food of the road- runner in ornia,’ op. cit., XVII, No. 5, pp. 21-58, Pls. 1-1. > 302 Bulletin American Museum of Natural History {Vol. XLV another bird had eaten ten carpenter ants (Camponotus species). Accord- — ing to records in the United States Biological Survey, published by W. D. Hunter,' the following Texas birds are known to prey upon the agri- cultural ant, Pogonomyrmex barbatus subspecies molefaciens (Buckley): great-tailed grackle, Megaquiscalus major macrourus (Swainson); up- land plover, Bartramia longicauda (Bechstein); burrowing owl, Speotyto cunicularia hypogea (Bonaparte); Texas nighthawk, Chordeiles acuti- pennis texensis Lawrence; scissor-tailed flycatcher, Muscivora forficata (Gmelin); kingbird, Tyrannus tyrannus (Linneus); redbird, Cardinalis cardinalis (Linnzus); and mockingbird, Mimus polyglottos (Linngeus). Cleland’s recent account of the food of Australian birds,? makes it clear that the rich ant fauna of that continent is preyed upon by a great many birds of different families. Of a total of 224 species examined with regard to their stomach contents, 73 were found to contain ants, though as a rule these insects were present in small quantities only. The follow- ing list contains such species as seem to show a preference for ants. Black-breasted plover.—Zonifer tricolor (Vieillot). Lesser golden plover.—Charadrius dominicus (P. Miiller). Brown flycatcher.—Micreca fascinans (Latham). Flame-breasted robin.—Petreca phenicea Gould. Scrub robin.—Drymacedus brunneopygius Gould. Coach-whip bird.—Psophodes crepitans (Latham). Blue wren.—Malurus cyanochlamys Sharpe. Grey shrike-thrush.—Collyriocichla harmonica (Latham). Black-backed magpie.—Gymunorhina tibicen (Latham). White-backed magpie—Gymnorhina leuconota Gray. White-throated thickhead.—Pachycephala pectoralis (Gould). Yellow-breasted shrike-robin.—Eopsaltria australis (White). White-throated tree-creeper.—Climacteris picumna (Temminck). Brown tree-creeper.—Climacteris scandens Temminck. Noisy minah.—Myzantha garrula (Latham). Yellow-throated minah.—Myzantha flavigula Gould. Most of the ants found in these stomachs were not identified. In the case of the Micreca and the species of Myzantha, remains of Campono- tus nigriceps (Smith) and of a Polyrhachis were recognized. Two of the stomachs of Psophodes crepitans contained a large quantity of the heads and legs of ants, chiefly the “green-head ant” [Rhytidoponera metallica (F. Smith)]; some of the Malurus cyanochlamys, Gymnorhina tibicen, Eopsaltria australis, Climacteris picumna, and C. scandens had also fed on this or allied Ponerine. Bulldog ants (Myrmecia species) 11912, U.S. Dept. Agric. Bur. a. ‘ye No. 148, p. 6. Cleland, J. B., 1918, ‘Thé food of Australian birds. An investigation into the character of the stomach and crop contents,” Science Bull. No. 15, Dept. of Agric. New South Wales, 112 pp. ___ Wheeler, Ants of the Belgian Congo 303 a iwere found i in large numbers in the stomachs of Collyriocichla harmonica, _ Gymmorhina tibicen, G. leuconota, Pachycephala pectoralis, Eopsaltria australis {in this case the ant being identified as Myrmecia gulosa conan and Myzantha garrula. I am informed by Prof. Wheeler _ that the two ponerine genera Myrmecia and Rhytidoponera contain some of the largest and most conspicuous members of the Australian ant- - funn, and all of them sting or both sting and bite severely. ‘The myrmecophagous habit is perhaps most highly developed ~ among birds of tropical regions, many of which are entirely or almost restricted to this kind of diet. Thus, F. Dahl concludes that, in the Bis- of + ramanele Archipelago, insectivorous birds are the most dangerous enemies of ants. Of about ninety species of terrestrial birds examined by Dahl _ in that region, twenty-eight were found to be more or less ant-eating; _ some fifteen of these had captured the winged sexual phases only, at the cS : time of the nuptial flights; twelve others had also fed on worker ants _ picked up outside their nests, the list of these including flycatchers _ (Pecilodryas, Monarcha, Rhipidura), thickheads (Pachycephala), drongos (Dicrurus), honeyeaters (Myzomela), timeliids (Ortygocichla), and warblers (Cisticola). The stomach of one of the warblers, Megalurus macrurus (Salvadori), was filled with the workers and sexual phases of a species of Polyrhachis (near schenki),' of which it apparently destroys the nests, though it feeds on many other insects too.* _ In their discussion of the food-habits of Indian birds, Mason and -Maxwell-Lefroy’? summarize the evidence concerning the Formicide as follows: The ants, like the grasshoppers, are exceedingly abundant insects and form a very large proportion of the insect food of birds in India. They are perhaps the favorite food of the woodpeckers, wrynecks, rollers, and some of the pheasants. Most. birds that eat insects of any kind will almost certainly be found to take ants of one species or another. The following species occur in this paper as taken by birds: Acantholepis frauenfeldi variety bipartita, Camponotus compressus, Calaulacus (apro- banz, Crematogaster sulmuda, Dorylus species, Meranoplus bicolor, Myrmecocystus selipes, Ecophylla smaragdina, Pheidole malinsi, and Polyrhachis simpler. Of 109 species of birds examined by these authors in the plains of northeastern India, near Pusa, forty-eight showed remains of ants in their stomach. In most cases these insects were — in small ‘Forel remarks in this connection: * yelds bauaposted at the podnin og of the species of Polyr- hachis may have the pp ce Ele ipo | mals from using them as food. But this is a mere tion. As sny pean Be S ap of Br rof. Dahl show that they are eaten by birds, which seems nat considering that they so freuen ‘Dahl, F. 1901, "Das Leben der Ameisen im "Bismnarek- “Archipel,” Mitt. Zool. Mus. Berlin, I, noted that the siosendh wy hep egalurus macrurus Was examined rot ene Sm aa sea H., 1912, othe Be ob tosh India,’ Mem. Dept. Agric. India, pp. 304 Bulletin American Museum of Natural History [Vol. XLV numbers only. Certain Indian birds, however, feed entirely on ants and foremost among these are, again, the woodpeckers. Three stomachs of the northern rufous woodpecker, Micropternus phxoceps Blyth, contained exclusively ants: 1459 Crematogaster subnuda in the first; 2600 of the same ant in the second; 725 of this Crematogaster, 304 Pheidole malinsi, and 27 pup and larve of @cophylla smaragdina in the third. Of 3921 insects taken by 16 specimens of another woodpecker, Brachypternus aurantius (Linneus), 1738, or 44 per cent, were ants (Camponotus com- pressus, (Ecophylla smaragdina, Meranoplus bicolor, Myrmecocystus setipes, and Crematogaster subnuda), and in several instances the bird’s stomach contained nothing else. An interesting result was obtained with the wryneck, Jynz torquilla (Linnzus): seven stomachs contained 1540 insects, all but eight of which were ants, mostly of the species Pheidole malinsi. Another prominent ant-feeder in India is the brown shrike, Lanius cristatus Linneus; of 111 insects taken by seven birds, 41, or 36 per cent, were ants (Hcophylla smaragdina and Crematogaster subnuda). Similar observations on South and Central American birds would be extremely valuable, for it is surprising how few accurate data have been published, as yet, with regard to the food habits of most tropical birds. For this reason, I include a list of the Nicaraguan birds in the stomach of which Mr. W. De Witt Miller has found remains of ants." Geococcyx velox (A. Wagner). One stomach contained a mass of insects, including three fairly large ants; several other birds of this species showed no ants. Chloronerpes rubiginosus yucatanensis (Cabot). Fragmentary re- mains of many ants were found in one stomach. The proventriculus and stomach of another individual were filled with ants, some of these being mostly yellowish and 10 mm. long; there was also one beetle. In a third case the stomach contained a large number of ants of at least two kinds, by far the majority belonging to a small yellowish species; also at least one small beetle. Many ants were present in the stomach of a fourth bird. Ceophleus lineatus similis (Lesson). Two stomachs examined con- tained numerous ants and bits of other insects. Centurus hoffmanni Cabanis. In one case the stomach showed no other food than many ants of various kinds, while that of another bird was filled with fruit of a Cecropia. 1] am greatly indebted to Mr. W. DeWitt Miller, Associate Curator of Ornithology in The American Museum of Natural History, for permission to use this information, and also for many valuable sug- gestions and criticisms on my account of birds as predaceous enemies of ants. 1922] __ Wheeler, Ants of the Belgian Congo 305 J an _ Xiphocolaptes emigrans emigrans Sclater. Insects, including many ants, in one stomach; three other stomachs contained no ants. Saltator magnoides medianus Ridgway. One stomach contained, among other things, a number of myrmecine ants. Amblycercus holosericeus (Lichtenstein). One stomach was filled ee with insects, including at least one small black ant. _ Thamnophilus doliatus (Linneus). One stomach contained two small black ants among other insects; that of another individual was filled with fair-sized ants of at least two kinds, some black, some yellow; no ants were found in three other stomachs of this species. Pachysylvia decurtata (Bonaparte). Two stomachs contained in one rs _ ease one, and in the other four ants, among other insects. _ Pachyrhamphus cinnamomeus Lawrence. One stomach was com- pletely filled with insects, including two ants. _ Syrtallaxis pudica nigrifumosa (Lawrence). One ant among many other insects, in one stomach. _ Hylocichla ustulata swainsoni (Cabanis). Insects, including one ant head, were found in one stomach. Euthlypis lachrymosa (Cabanis). The contents of one stomach con- sisted of insects, including ants. Cyclarhis flaviventris Lafresnaye. Insects, including a few ants, were found in one stomach. Another bird contained no ants. Myjiopagis placens accola Bangs. One stomach showed a few insects, including one ant head. Salpinctes fasciatus Salvin and Godman. In two stomachs examined a number of ants were found, together with other insects. As Mr. Miller points out, it would seem that, except in the case of certain woodpeckers (Chloronerpes, Ceophleus, and Centurus), ants are an exception in the food of the insectivorous birds of Central America. Perhaps the most pronounced ant-feeders of all Neotropical birds are the curious woodpeckers of the genus Celeus. G. F. Gaumer describes the habits of the common species, Celeus castaneus (Wagler), in Yucatan, as follows: “ This bird has a very strong and peculiar odour, derived from its food, which consists exclusively of a small hymenopterous insect called the Uss. It is solitary, and lives in the deepest part of the forest. The specimens obtained were very tame and were watched for some hours before being shot; they jump nimbly about the trees, and are constantly catching the small insects which seem to be attracted to them by their odour.’"' I am informed by Mr. Miller that, according to Mr. 1Quoted by A. Boucard, Proc. Zool. Soc. London, 1883, pp. 452-453. 306 Bulletin American Museum of Natural History [Vol. XLV G. K. Cherrie’s observations made in Central America, the ‘‘hymenop- terous insects” in question are ants.’ Phenicothraupis rubicoides (Lafresnaye), one of the tanagers, is often credited with following swarms of ants in search of its food, as, for instance, by G. F. Gaumer® from observations made in Yucatan and by C. C. Nutting’ in Nicaragua. The latter remarks: ‘Curiously enough, although a tanager, this bird is usually seen clinging to the tree-trunks, like the Dendrocolaptide, and hops about the ground like the Formicarii- dx. Indeed it seemed to be living almost entirely upon ants. There were many places where the ground was actually swarming with these insects, and there P. rubicoides would congregate in large numbers, either picking up the ants from the ground, or climbing about the trunks of trees in pursuit of the same insect.” Eucometis spodocephala (Bonaparte), another tanager, and Dendro- cincla sancti-thome (Lafresnaye), one of the Dendrocolaptide, were also seen by C. C. Nutting in Nicaragua, feeding largely upon ants (op. cit., pp. 382 and 385).‘ It is especially the Neotropical ant-thrushes, or Formicariide, that have been credited with habitually following the columns of the foraging ants (Ecitonini) in much the same manner as will later be described for the African ant-thrushes and doryline ants. R. Schomburgk® mentions that, in British Guiana, the moving armies of Eciton are always accom- panied by large numbers of several species of birds, the most common of these being Formicarius colma (Boddaert) and Pithys albifrons (Linnzus). This traveller evidently believed that the birds were feeding on the ants themselves. H. W. Bates, speaking of his experiences with the foraging ants in Brazil, also writes that ‘when the pedestrian falls in with a train of these ants, the first signal given him is a twittering and restless move- ment of small flocks of plain-colored birds (ant-thrushes) in the jungle.” Belt’s observations in Nicaragua are somewhat similar: ‘The numerous birds that accompany the army ants (Eciton predator) are ever on the lookout for any insect that may fly up, and the heavy flying locusts, grasshoppers, and cockroaches have no chance of escape. Several species of ant-thrushes always accompany the army ants intheforest. They do iLaderwaldt (1909, Zeitschr. Wiss. Insektenbiol., V, p. 312) tells that the Campos woodpecker (Canave ? campestris (Vieillot)} ransacks the ground nests of ¢ Camponotus rufipes (Fabricius) in southern pes ig by A. Boucard, Proc. Zool. Soc. oy 1883, p. 443. 883, Proc. U. S. Nat. Mus., VI, No. 24, p ‘Noms of the few stomachs of Phenicothraupis oii Dendrocincla from Nicaragua examined by Mr. Miller contained ants. ‘Schomb urgk, R., 1848, ‘ Reisen in Britisch-Guiana,’ (Leipzig), it ‘Pp. 287-288 and 421. £1863, ‘The naturalist on the River Amazon,’ (London), pre) _ "Wheeler, Ants of the Belgian Congo 307 not, however, feed on the ants, but on the insects they disturb. Besides _ the ant-thrushes, trogons, creepers, and a variety of other birds, are __ often seen on the branches of trees above where an ant army is foraging below, pursuing and catching the insects that fly up.”'! It does not appear, however, that the food of the Formicariide has often been determined from actual examination of the stomach contents of these During his sojourn in Africa with the American Museum Congo _ Expedition, my friend Mr. J. P. Chapin made accurate investigations as to the food of birds, examining the stomach and crop contents of most of the specimens collected by him. He has kindly allowed me to use his - observations, and some of his field notes are quoted in full below. Of. about 6000 Congo birds examined by him in this respect, some 200, be- | longing to about 85 or 90 species, had included ants in their diet. In the following account I have grouped the African ant-eating birds according to the interest they show in this kind of food and the manner in which they procure it. Data heretofore published bearing on the subject have been referred to, in so far as I have been able to ascer- tain in the extensive literature on African ornithology; in this, too, I have keen very effectively aided by Mr. Chapin. 1.—In a first group may be placed birds that feed occasionally or accidentally on ants, without, however, showing much preference for this kind of diet. A great number, if not all, of the African insectivorous and omnivorous species should perhaps be included here; for most of them available records merely give “‘insects’’ in general-as food. The following are the only species for which ants have been expressly men- tioned as part of the diet. Glareola fusca (Linneus), according to v. Heuglin,? in Nubia pursues swarms of winged ants in the evening, as do other species of Glareola. Sarciophorus superciliosus (Reichenow). Zech, in Togo, found ants in the stomachs.* (Edicnemus adicnemus (Linveus), according to v. Heuglin,‘ feeds partly on ants in Nubia. Abdimia abdimi (Lichtenstein) eats even ants, according to Hart- mann,’ and G. K. Marshall* in Rhodesia found, in the stomach of this 1874, ‘The naturalist in Nicaragua,” (London), p. 20 *Heuglin, T. v., 1873, ‘Ornithologie Nordost-Afrika’s,’ (Cassel), L1, p. 982. by Reichenow, A., 1901, ‘ Die Vogel Afrikas,’ (Neudamm), I, p. 191 ‘Heuglin, T. v., 1873, op. cit., 11, p. 988. “Quoted by Heuglin, T. v., 1873, op. cit., 11, p. 1104 #1902, Trans, Ent. Soc. London, p. 350 308 Bulletin American Museum of Natural History [Vol. XLV stork, beetles and “ants of the genus Carebara” (probably of the winged, sexual phases.) Melierax canorus (Rislach). The stomachs examined by Oates, in Transvaal, contained large ants, rats and lizards.’ Falco concolor Temminck. Antinori observed in Eritrea flocks of this bird hunting winged ants (perhaps termites?).* Pogoniulus pusillus uropygialis (Heuglin). v. Heuglin*® found some ants among the stomach contents of this bird in Nubia. Trrisor senegalensis (Vieillot), according to v. Heuglin,* eats ant among other insects in Nubia. Batis orientalis (Heuglin). G. W. Bury noted, for a specimen col- lected in Northern Somaliland, that “the stomach was found to contain a large number of ants.’’® Batis molitor (Hahn and Kiister). Insects of various kinds, also ants, in the stomachs of Gazaland specimens.® Laniarius erythrogaster (Cretzschmar), according to v. Heuglin, 7 eats ants among other insects in Nubia. Ploceus aureoflavus A. Smith. Fischer* found in the stomach of this weaver-bird, in British East Africa, seeds and sometimes also ants and caterpillars. me Nectarinia arturi Sclater. The crops examined by Swynnerton,® in Gazaland, contained flying ants, small flies and several large gnats. Chalcomitra kirki (Shelley). The crop of one bird examined in Gazaland by Swynnerton” contained beetles and ants. Tarsiger stellatus (Vieillot). The crops of two specimens examined in Gazaland by Swynnerton"™ contained berries, various insects and ants. Muscicapa cerulescens (Hartlaub). Large black ants and beetles were found in the stomach ot a specimen taken in Gazaland by Swyn- nerton."! At Salisbury, Rhodesia, G. K. Marshall” found remains of ants in the stomachs of the following birds: Bradornis mariquensis (A. Smith). Pratincola torquata (Linnzeus). ‘Oates, F., 1881,‘ Matabele Land and the Victoria Falls,’ (London), .P. 298. 2Antinori, O. and Salvadori, T., 1873, * eres Mus. Civ. Genova, IV, p . 389. *Heuglin, T. v., “ae ap-ae- I, p. 762 . 41869, op. cit., 1p. 3 ‘Bannerman, 191 Ibis. (9) af p. 312. *Swynnerton, 1907, Ibis, (9) I apr *Heuglin, T. v., 1871, op. cit., Ir, aa SQuoted by Reichenow, A., 1904, ‘Die Vogel Afrikas,’ (Neudamm), III, p. 92. #1907, Ibis, (9) 1, p. "Op. cit., p. 43. NOp. cit., p. 67. 1902, Trans. Ent. Soc. London, p. 351. Ea a a re — oe j — Wheeler, Ants of the Belgian Congo 309 _ Monticola angelensia niasse Reichenow. Sazicola pileata (Gmelin). ; ae ‘ Dicrurus adsimilis divaricatus (Lichtenstein). 7. Campothera bennetti (A. Smith). Crecopsis egregia (Peters). From Mr. Chapin’s observations it appears that in the Belgian Congo swallows [Riparia riparia (Linneus) and Psalidoprocne nitens __ centralis Neumann],' Coracina pectoralis (Jard. Selby), shrikes [Nilaus afer (Latham), Corvinella corvina (Shaw)], and certain kingfishers | Gialeyon pallidiventris Cabanis) are very fond of catching sexual winged ants together with other flying insécts, while francolins (Francolinus ___ lathami Hartlaub, FP. squamatus Cassin, F. icterorhynchus Heuglin) and Guinea fowl (Guttera plumifera schubotzi Reichenow, G. pallasi Stone, __ ~Phasidus niger Cassin, Numida ptilorhyncha Lesson) often indiscrimi- nately pick up worker ants from the ground withsnails, beetles, seeds, and even pebbles. With regard to the two species of forest Guinea fowl, Mr. Chapin remarks that their flesh “is rather dry eating and has, in addi- tion, a peculiar strong taste, due probably to something they eat, pos- sibly the ants usually found in their crop.” At Ngayu the crop and stomach of a black forest Guinea fowl, Phasidus niger, were filled with thick green leaves and driver ants. In the case of the savannah Guinea fowl, Numida ptilorhyncha major Hartlaub, ants were frequently found in the crop; usually, as in a specimen examined at Faradje in September 1911, these ants belonged to the large, black, termite-hunting species, Megaponera fetens (Fabricius). The following list contains the birds from the Belgian Congo which showed remains of ants in their stomachs. In most cases these insects were present only in small numbers, or the individuals taken belonged ta the winged phases. In some of the pipits (Anthus) and thrushes (Thamnolza and Monticola), however, worker ants and even their larve were sometimes present in large quantities; it is possible that these birds, and perhaps others in the list, may prove on further observation to be rather regular ant-feeders. Glareola melanoptera Nordmann. Himantornis hematopus whitesidei Sharpe. Galachrysia nuchalis emini (Shelley). Ciconia ciconia (Linneus). Neotis denhami (Children). Coturniz coturnix (Linnzus). ~o wee we tee ee - = —— ae meen of irundo nigrita G. R. Gray, from Gamangui, Mr. Chapin found that “the right metatarsus had been bitten by a pve ant, whose head still adhered to it, an all the lower part of ‘oot had died and dried up, without falling off.” z 310 Bulletin American Museum of Natural History (Vol. XLV Coturniz delegorguei Delegorgue. Muscicapa striata (Pallas). Milvus xgyptius parasitus (Daudin). Stizorhina vulpina Reichenow. Centropus grillii Hartlaub. Oriolus larvatus letior Sharpe. Cuculus solitarius Stephens. Onychognathus hartlaubi Hartlaub. Cuculus jacksoni Sharpe. Hyphantornis cucullata feminina Grant. Lybius guifsobalito Hermann. Malimbus nitens microrhynchus Coracias abyssinicus Hermann. Reichenow. Eurystomus afer (Latham). _ Pyromelana crassirostris Grant. Eurystomus gularis Vieillot. Estrilda atricapilla Verreaux. Bucorvus abyssinicus (Boddaert). Melanopteryx nigerrimus (Vieillot). Ceratogymna atrata (Temminck). Parmoptila jamesoni (Shelley). Bycanistes sharpei (Elliot). Anthus leucophrys gouldi Fraser. Horizoceros granti (Hartlaub). Pycnonotus tricolor (Hartlaub). Irrisor erythrorynchos (Latham). Cinnyris superbus (Shaw). Scoptelus adolfi-friederici Reichenow. + Cinnyris splendidus Shaw. Caprimulgus inornatus Heuglin. Hedydipna platura (Vieillot). Caprimulgus batesi Sharpe. Eremomela badiceps (Fraser). Macrodipteryx longipennis (Shaw). Monticola saxatilis (Linnzus). Cosmetornis verillarius (J. Gould). Thamnolxa nigra (Vieillot). In the case of the weaver-birds included in the above list, it is evi- dent that some of the species (Parmoptila, Malimbus, Estrilda, etc.) have a marked predilection for ants, since the crop and stomach very _ often contained their larva, pupw, and workers. Mr. Chapin’s note concerning a Melanopteryx nigerrimus obtained at Avakubi is worthy of quotation: “its stomach contained many of the large light brown ants' that bind together the leaves of mango, as well as those of other trees, with silky fibers produced by their larve.” 2.—Birds that feed chiefly or to a very large extent on ants are of more interest to the myrmecologist, and some of them have developed peculiar habits in connection with this kind of diet. a.—Swifts and bee-eaters seem to show, in tropical Atrica, a marked preference for ants in the winged phases, which they catch in flight. The stomach of one of the most common swifts, Micropus apus (Lin- nus), was frequently found by Mr. Chapin to be filled with brownish- black winged ants; many other species, such as Micropus streubeli (Hartlaub), M. affinis (Gray and Hardwicke), Tachornis parvus (Lich- tenstein), Chetura cassini Sclater, C. ussheri sharpei Neumann, C. sabini Gray, and C. melanopygia Chapin, have similar habits. Mr. Chapin observes that swifts feed mainly on winged ants, while swallows catch them only occasionally. The red-breasted bee-eater, Merops “UEcophylla longinoda (Latreille). [J.B.] - a i ~~ Wheeler, Ants of the Belgian Congo 311 -malimbicus Shaw,' also shows a great predilection for winged ants; in the _ eight specimens shot near Monsembe,on the Congo River, from a flock of 175 to 200 which was resting in the top of a dead tree, the gizzard was well filled with such insects. Similar observations were made on related species at Avakubi (Merops albicollis Vieillot) and Bafwabaka [Melit- a tophagus miilleri (Cassin) }. _ b.—A rather small group of insectivorous birds attack the nests @ _of ants and feed on the workers as well as on the brood. This is a very common habit with many species of woodpeckers. Sjéstedt? relates that some of the Cameroon species seem to live chiefly on ants, which were the only insects he found in the stomach of Campethera permista (Reiche- now). Kersting* found ants in the stomach of Campethera nivosa (Swain- son) and v. Heuglin‘ in that of Mesopicus schoensis (Riippell) and Dendro- -picos obsoletus (Wagler), brood as well as worker ants being present. 4 : Similar observations were made by Mr. Chapin on the following Congo species: Campethera caroli (Malherbe), C. permista (Reichenow), C. balia Heuglin, C. abingoni chrysura (Swainson), and C. nivosa (Swainson). ‘His following note relates to a specimen of the last-named species from Avakubi: “the stomach contained larve and pup2 of a very small black B ant that builds large brown nests in the trees.’ From this it would seem that this woodpecker had been pecking holes in a nest.” A specimen of Campethera abingoni (A. Smith) obtained by Swyn- nerton® in Gazaland had its stomach filled with hundreds of a small black tree-ant in all stages of development. -c.—Some birds of the African forests have developed the curious habit of following the columns of doryline driver ants, much as do the South American Formicariide I have previously mentioned. The earliest observations of the kind were made by Du Chaillu’ in the Gaboon: “Hunting in the rear of the village (of Obindji) on the 15th [of April 1858], I shot a curious bird, the Alethe castanea—a new species. . . They fly in a small flock, and follow industriously the bashikoway ants {driver ants] in their marches about the country. The bird is insectivo- rous; and when the bashikoway army routs before it the frightened grass- hoppers and beetles, the bird, like a regular camp-follower, pounces on the prey and carries it off. I think it does not eat the bashikoway.” 'Hartert found winged ants, together with other insects, in ies -eemaae of Merops malimbicus examined — in Nigeria (1886, ayers. {. Ornithol., Iv, 21895, 1 Svenska Vetensk. Ak. Handl., N.8., XXVIII, Ne. on “4 54 and 56 nT Reichenow, A., 1903, * Die Vigel Afrikas,’ (Neudamnin), Il, p. 170 ¥., 1871, ‘Ornithologie Nordost-Afrikas,’ (Cassel), I, p. 804 and 810 of C rematogaster. B.) 1007, atly 9) L, p. 200. Du Chait, . Be 1861, ‘Explorations and adventures in equatorial Africa,’ (New York), p. 319 312 Bulletin American Museum of Natural History [Vol. XLV Reichenow' made similar observations on the same species of bird in Cameroon, but he found the stomachs of specimens examined by him filled with driver ants. He also claims that Turdinus fulvescens (Cas- sin) has similar habits.? According to Sjéstedt,’ the following birds are found near the moving columns of Dorylus (Anomma) nigricans sub- species arcens in Cameroon: Bleda notata (Cassin), B. syndactyla (Swainson), Alethe castanea Cassin, Criniger calurus (Cassin), and Neocossyphus poensis (Strickland). This observer notes that Bleda notata on such occasions does not remain on the ground, but rather on the lower branches of trees and shrubs, whence it jumps down to the ants and returns at once to its perch. The stomach of Neocossyphus poensis was found to contain ants only, while that of Bleda notata contained ants and beetles. At Efulen, Cameroon, G. L. Bates‘ also saw Alethe castanea “in thickets where an army of driver ants covers the ground and bushes, as they are very fond of feeding on these ants, though they do not come into open places to do so.”” In another paper, Bates’ writes: ‘‘ Whenever you see a number of birds of different kinds flitting about near the ground in one place and twittering excitedly, you may be pretty sure there is an army of ‘driver ants’ at hand. Many different kinds of birds join in the chase of driver ants. I have even seen the small white- crested hornbill (Lophoceros hartlaubi) engaged in it.’’ Another hornbill, Ortholophus cassini Finsch, was once seen by Bates® to join with smaller birds in pecking at a swarm of driver ants on the ground. On Mount Ruwenzori, between 6500 and 9000 feet, R. B. Woosnam found Alethe poliophrys Sharpe “frequenting the forest zone and the lower edge of the bamboo. It appeared to be particularly fond of the soldier ants and might often be seen attacking a column of theseinsectsas they crossed a path or open spot. Whether it really ate the ants or merely snatched away the eggs they were carrying, was a point we could never decide; probably the eggs were the attraction, for it seems difficult to imagine anything more unsatisfactory than a meal of angry soldier ants.’”? 11875, Journ. f. Ornithol., XXIII, p. 29. ; ‘ : 2Mr. Chapin did not find this to be the case with 7. fulrescens in the ed Cone 38) t, Y., 1895, ‘Zur Ornithologie Kameruns,’ Kongl. Svenska Vetensk. Ak. Handl., N. 8. XXVIII, No. 1, pp. 1-120, Pls. 1-x. In a later paper Sjéstedt further mentions certain the genus Campethera and Stiphrornis gabonensis 8 as occasionally foll the columns of the dain in Cameroon, though not so regularly as the Criniger and Alethe (‘Exped. Kilimandjaro, Meru, etc.,’ II, 8, 1908, p. 111). ted by R. B. Sharpe, 1908, Ibis, (9) II, p. 128. é oted by R. B. Sharpe, 1904, Ibis, (8) IV, p. 92. : ° #1905, Ibis, (8) V, p. 89. Under the name Csinclaaiion albocristatus (Cassin). *7Woosnam, weal O. Grant, 1910, Tratis. Zool. Soc. London, XIX, p. 374 1922) Wheeler, Ants of the Belgian Congo 313 In his account of the columns of Dorylus in Gazaland, South East Rhodesia, Swynnerton!' has the following remarks: ‘I have also on a occasions watched birds attending Dorylus, to rob stragglers of their , and for the sake of the flying and hopping insects flushed by the s. Some of the birds on occasion eat the ants themselves. In my erin on many species of insectivorous birds, I found that some ate ants generally, including Dorylus, far more readily than others. Of these others some showed a strong repugnance to them, and it is oubtless in relation to this latter class of enemy, that ant-mimicry finds its main use. Yet even the birds that prey on ants show caution in attacking Dorylus in column, merely (in my observations) dropping down to stragglers and hastily returning to their perch.” It would be interesting to know which species of birds in South Africa have acquired these habits, since most of the true ant-thrushes are more at home in the ‘ West African forest region. tah In the forest of the Belgian Cais; Mr. Chapin found the following birds associated with the armies of Doryline: Alethe castanea woosnami _ Grant, A. poliocephala carruthersi Grant, Bleda eximia ugande van 3 Someren, B. syndactyla woosnami Grant, Neocossyphus rufus gabunensis Neumann, N. poensis prepectoralis Jackson? Several of these species ___ occur together, as indicated in the following field-note written at Avakubi, __ April 16, 1914: “We came to a spot in the forest this morning where a _ great number of driver ants were crossing the road in several columns; ___ and, noticing that there were also birds on hand, we stopped for some ____ time to watch the proceedings. Besides a half-dozen small brown thrushes _ (Alethe, mostly A. c. woosnami, but also one or two A. p. carruthersi), there were tworufous thrushes (Neocossyphus rufus gabunensis) , at least one with ___ white patches in the tail (Neocossyphus p. prepectoralis) and one Bleda _-&. woosnami. It was quite evident that all these birds were attracted by the ants, and they seemed especially interested in a spot where these ___ irritable insects had spread out widely over the path. The Alethe were, of course, most in evidence, fluttering back and forth across the road, - occasionally darting down right among the ants, and perching in the bushes bordering the way. Alethe c. woosnami has a habit of fluttering its wings slightly, like a bluebird, while perching. From time to time one of the larger rufous thrushes would fly out of the undergrowth, sometimes even alighting on the ground amid the ants, but, as usual, 11916, Trans. Ent. Soc. London, eel: p. 318. 21t may be remarked that these African “‘ant-thrushes"’ are not all related forms, for, while Alethe Neocossyphus belong to the Turdide, Beda i is placed among the Pyenonotida. 314 Bulletin American Museum of Natural History . [Vol. XLV these birds were very shy and it was only after long waiting that I could shoot one. “Now, what are the birds after? It is not, as a rule, the adult ants, for these are generally only eaten, if at all, in very small numbers. Nor is it their young, for they frequently do not carry any, and this cireum- stance has no relation to the presence or absence of birds. Is it the victims—other insects and the like—being carried by the ants? Surely there ought to be easier ways than this to procure the same food. Yet the three ant heads in the stomach of one of the Neocossyphus rufus might have come there in that way. Seizing some coveted morsel, the bird found, perhaps, that several ants had buried their jaws in it, but plucked off their bodies at, any rate, before eating it.” On another occasion, at Bafwabaka, the stomach of an immature Neocossyphus p. prepectoralis was found filled with driver ants; but in most of the other “ant-thrushes”’ examined for this purpose the food consisted mostly of small insects, with occasionally a driver ant. A number of stomachs of Alethe also contained the bones of small frogs. Plate I (frontispiece) represents a typical association of three driver ant birds commonly found in the Ituri Forest following the columns of the dorylines: Alethe c. woosnami Grant, Neocossyphus rufus gabunensis Neumann, and Bleda eximia ugandx van Someren. ‘ MAMMALS That many insectivorous and omnivorous mammals, such as moles, shrews, monkeys, and the like, will at times feed on ants can be expected after what we have learned above of the feeding habits of insectivorous birds; we know, however, but little about this from actual observation. We have the authority of John Muir that certain North American black bears are very fond of carpenter ants (Camponotus); they “tear and gnaw their home logs to pieces, and roughly devour the eggs, larve, parent ants, and the rotten or sound wood of the cells, all in one spicy acid hash.’ Mr. C. L. Camp has kindly informed me that he once saw _ in the Yosemite National Park, California, bear-droppings containing — masses of the chitinous remains of ants. Moles, too, must devour large numbers of worker ants and their pupx, though I have found no a ee 1Mr. Chapin also notes that, at Faradje, he once watched a chicken eating army ants. *Muir, J., 1916, ‘ My first summer in the Sierra,’ (Boston and New York), p. 46. C. H. Merriam (1884, ‘The mammals of the Adirondack Region,’ New York, p. 95) writes that the American black , bear (Ursus americanus Pallas) ‘‘is par excellence an omnivorous beast, and his larder consists notonly of mice and other small mammals, turtles, frogs, and fish; but also, and largely, of ants and their y bees and their honey, cherries, blackberries, raspberries, blueberries and various other fruits, venetablen, and roots. . . . He delights in tearing open old stumps and logs in search of the ants that make their homes in such situations.” a 1922), —— Wheeler, Ants of the Belgian Congo 315 definite records thereof, except in the case of the American mole, Scalopus aquaticus (Linneus). Scheffer' has examined one hundred stomachs of this animal in Kansas and found remains of ants in nineteen of them, these insects being then, as a rule, present in large numbers; one of the stomachs, for instance, contained 205 ants and 44 other insects; another, “e hans puparia and 6 other arthropods. _—s So -eallled “‘ant-eaters”’ are found in practically all tropical regions, . but the confusion so commonly made by casual observers between — the true ants (Formicide) and the “white ants,’ or termites, in many cases makes it hard to decide from published accounts which of these mammals are truly myrmecophagous and to what extent. Moreover, but little information based on actual study has been published con- cerning their feeding habits and stomach contents. White ants, or termites, constitute, of course, an attractive food for almost every insectivorous animal, while true ants, as Beebe remarks, “are all flavored more or less strongly with formic acid, and must be an acquired taste.” Further interesting questions which cannot be answered at this time are whether the various ant-eaters prefer ants to other insects and whether they can make a selection between different species of ants. These points would be of importance in considering the possible use of these animals to combat the leaf-cutting ants of tropical America, as suggested by certain observers. _ The echidnas, or spiny ant-eaters (Echidna aculeata Shaw and allies), of New Guinea, Tasmania, and Australia belong to the order Mono- tremata and are among the most primitive and odd-looking of present- day mammals. The Australian species, at least, is said by most obser- vers to feed on “ants,’’ though from the descriptions of G. Bennett* and Saville Kent it would appear that by this termites are meant as well as Formicide. Saville Kent, for instance, writes that when the echidnas are placed in contiguity to a teeming ant track, they take no notice of it, “appreciating the insects only under the conditions obtaining in the nests or hillocks. These edifices they would soon tear open with their powerful claws, exposing to view the white succulent nymphs, larve, and pup, or so-called eggs, upon which alone they concentrated their attention.’”* Scheffer, T. H., 1910, ‘The common mole,’ Bull. 168 Kanens Brose gre ‘ vol. Exper. Stat., “RB: 1-36; 1914, ’The common mole of the Eastern United States Farmers Bull. 1-10. Bee also Weat, J. A., 1910, ‘A Study of the food of ie in Ii Rooks Bull. Illinois St. Lab. ‘at. Hist., IX, 2, pp. 14-22. 21918, ‘Ball, RAO Soc. New York, XXI, No. 1, p. 1561. *Bennett, G., 1860, ‘Gatherings of a naturalist in Australasia,’ (London), pp. 147-150. 41807, 'The naturalist in Australia,’ (London), p. 19 316 Bulletin American Museum of Natural History {Vol. XLV The echidna is chiefly nocturnal and shows many remarkable adap- ~ tations to its habit of feeding on subterranean insects. The face is drawn out into a long, tapering, cylindrical snout, terminating in a very small mouth. The tongue is elongated, very slender, and capable of being protruded for a considerable distance. The jaws are slender and entirely destitute of any kind of teeth, of which, moreover, no trace has been found in the young. The palate, however, and the back of the tongue are rough with small spines, presumably to hold the living prey. “For ants and their eggs form the staple food, and these the Echidna obtains by digging up the ant or termites’ nests with its powerful limbs. Then the tongue covered with a sticky saliva is protruded; it becomes covered with ants, and is then quickly drawn back into the mouth.”! - More circumstantial evidence concerning the food of the echidnas in Queensland is to be found in a short note by Bennett’s son: They are particularly partial to the white ants, which erect small mounds of clay about 18 inches in height. These they attack in a most systematic way, by working round the nest, by clearing away the earth and forming a trench where the nest joins the earth, and devouring all before them; and then they make a hole in the center and clear out the whole nest, leaving none behind to tell the tale of their visit. The soldier ant (a large stinging ant) they do not touch; their nests were close to the white ant mounds, but were untouched. The larger sugar ants, which raise mounds of sand about 16 inches high and 4 feet in diameter, they attack first, by lying on the mound with their tongue out and drawing in the ants that cross it; there they remain sometimes for hours. This, I have no doubt, is the time that they get the sand found in their stomach. They then make a hole from one side to the other, and devour the most delicate morsels coming in their way. In the daytime they do not move about much, beginning their search about a couple of hours before sundown.” K. Dahl also states that the Australian Echidna aculeata depends upon termites for its food. Among the extensive order of marsupials, many of the insectivorous species must occasionally eat ants. One of them, the banded Australian ant-eater, Myrmecobius fasciatus Waterhouse, is often considered as belonging to a peculiar subfamily, the Myrmecobiine, and is said to feed on “ants’’ and perhaps also on other inseets. This interesting animal offers, among the marsupials, all the adaptive characters of the South — American ant-eaters: the elongate and pointed muzzle, the slender and extensive tongue, the stout fore limbs, and the long, curved, digging ‘Lucas, A. H. 8. and Le Souef, W. H. D., 1909, ‘The animals of Australia,’ ( mg ok i Bennett, G. J., ‘Observations on the habits of the Echidna hystriz of Nahe? Proce — ag Foon 737-739. pe ae A, t, (4) I, p. 200. See also Kinig, C., 1911, ‘Der Ameisenegel,’ Aus der Natur, VIT, pp. ~ Wheeler, Ants of the Belgian Congo 317 Eee. In Western Australia, the stomach of one example proved upon i ‘dissection to be full of white ants, most of which had evidently been order” Edentata which are said to feed chiefly on “ants.” In the case of the exclusively African aardvarks (Orycteropus; Pl. XXIV, fig. 1), Tam assured by Mr. H. Lang, from his examination of stomach contents, that the regular food consists of termites, while true ants are only taken ____ when they happen to occur in the termite mounds, as is frequently the _ ease, and are then unintentionally swallowed together with the white ants. How far this is true also of the South American ant-eaters (Myr- __ mecophagid, with the genera Myrmecophaga, Tamandua, and Cyclopes) ___ remains uncertain; the available information does not go beyond the Be: general statement that they feed on ants, termites, and their larve. Concerning the great ant-eater or ant-bear, Myrmecophaga tri- : dactyla Linnzus, Flower and Lydekker? say: “Its food consists mainly ___ of termites, to obtain which it opens their nests with its powerful anterior _ claws, and as the insects swarm to the damaged part of their dwelling, it draws them into its mouth by means of its long, flexible, rapidly-moving tongue covered with glutinous saliva.” That Myrmecophaga feeds on termites, and not on true ants, would also appear from the accounts ° given by H. W. Bates* and others. On the other hand, Hensel‘ maintains that ft the tamandua (Tamandua tetradactyla Linneus) does not feed on termites, but that in all the specimens of that species examined by him the stomach was filled with true ants, even in localities where termite mounds were very common. His statements are, however, contradicted by A. Zietz® who fed the tamandua in captivity with termites, while ants were obstinately refused. The little, or two-toed, ant-eater (Cyclopes didactylus Linnzus) is an arboreal species which seems to feed chiefly on true ants. At least, Miss Snethlage® was unable to feed it in captivity with termites; she says that not all ants are to its taste; the pup of a species which lives in dry imbauba trunks are eaten with predilection, also the pup# and workers of another small, black ant with triangular abdomen, found chiefly in inga trees. 1G. C. Shortridge, uoted 0. Thomas, Proc. Zool. Soe. London, 1905, II, p Flower, W. H. be, gee yO: R., 1891, ‘An introduction to the study of re living and , London 191. aise, "the aturallat on the River Amazon,’ II, be. 178- a7. am (1907, Satine, Boobechter, ’ yrmecophaga jubata in tivity on ante (Pormica rufa), which the galmal woold take La ‘Der Zoolog. Garten, urta.M., 1872, p. 177. *Der Zoolog. Garten, Frankfurt a.M. 1872, *Quoted by Strassen in Brehm's * Tierleben,’ ‘PR Ese *SAugetiere,’ I, 1912, p. 544. ) ea es oe 318 Bulletin American Museum of Natural History {Vol. XLV The armadillos (Dasypodid), which range in many species over the tropical and temperate parts of South America, one of them even reach- ing Texas, are said to be omnivorous, feeding on roots, insects, worms, reptiles, and carrion; in how far this diet may include true ants is by no means easy to gather from the very scanty descriptions of the habits of these animals; in many cases termites are in all probability the chief food.!’ The snout of the armadillos is moderately elongate, and the tongue is long, pointed, extensile, though less so than in the Myrme- cophagide. It would thus appear that the pangolins or scaly ant-eaters (Mani- de; Pl. XXIV, fig. 2; Pl. XXYV, figs. 1 and 2) of the Old World tropics are the only edentates whose myrmecophagous propensities are beyond doubt. These animals are at once recognizable by the large overlapping scales which cover the whole of the upper surface of the head, the upper surface and sides of the body, the whole of the tail, and the outer sides of the limbs; the legs are short and end in curved claws, those of the fore limbs being especially powerful. The snout is pointed and conical; teeth are entirely absent; the long, vermiform, protractile tongue is flattened toward the tip and kept sticky with saliva abundantly produced by enormous submaxillary glands. The structure of the stomach shows further curious adaptations to their ant diet; in Manis javanica, for instance, most of the mucous membrane is transformed into a pavement epithelium of horny texture, raised into folds in the cardiae region near the cesophagus, while it forms horny teeth in the pyloric part, at theend of the great curve; opposite these pyloric teeth, at the end of the small curve, the middle line is swollen into an organ of trituration, covered with numerous horny teeth and moved by powerful underlying muscles. The gastric glands are united into a few voluminous glandular bodies which pour their abundant secretion into the stomach by way of wide glandular ducts. The insects are swallowed whole and reach the stomach together with saliva, sand, and small pebbles often as large as a pea; this mixture is then ground up by the peristaltic movements of the stomach, whose inner walls are effectively protected by the horny pave- ment epithelium; gastric juice is profusely poured over the stomach contents, which undergo a final grinding by the organ of trituration in the pyloric region.” 'Laderwaldt (1909, Zeitschr. Wiss. Insektenbiol., V, p. 312) mentions incidentally that the armadil- los in J pi woes to grub about i in the earthen mounds of the stinging Solenopsis geminata. r, M., 1904, * Die Siugetiere,’ (Jena), pp. 426-427. — "Wheeler, Ants of the Belgian Congo 319 2 Seven species are now generally recognized in this family and are all cluded in the genus Manis: four of these occur on the African conti- nt, while the remaining three are found in the Oriental Region (Ceylon, i toe southern China as far as Kianghsi, Formosa, and Sunda - The ant-eating habit is common to all, though it has been ‘igated in only a very general way. I have been able to find but one ‘ “e the complete analysis of the stomach contents of one of these ils. It was made from a specimen of Manis (said to have been ncki, but probably gigantea Illiger), the stomach of which was sent on from the Lower Congo to Forel, who extracted from it the fol- nts,' several of which were at that time new to science. Anomma) emeryi subsp. opacus Forel. Numerous workers. dole porns Mayr. Several workers and two soldiers. tog. Sesgasares Emery. Very numerous workers and several males. mis s aculeata (Mayr). A few workers. icar voesadagiren var. congolensis Forel. Very numerous workers. mex opacus var. estus Forel. Very numerous workers, a number of males 4 manidis Forel. A small number of workers. Camponotus foraminosus subsp. delagoensis var. sorptus Forel. A very large number of 3 _ workers, a goodly number of females and several males. . _ Biittikofer,? in Liberia, fed the smaller, arboreal species, Manis Tisisesdats, with lary taken from mushroom-shaped termite nests. Of the large, terrestrial Manis gigantea, he says that the anterior portion __ of the stomach of a specimen contained about six liters of termites, while the posterior portion was filled with an equal amount of driver ants. ~ Vosseler* found that the excrement of a Manis temmincki Smuts which he a observed alive at Amani in Usambara consisted entirely of the chitinous s remains of driver ants. a The habits of the oriental species of the genus should not materially | differ from those of their African relatives. Kreyenberg,‘ who observed __ Manis javanica in China, states that all stomachs examined by him j 4 contained large numbers of ants and theirlarve exclusively. And speak- _ ing of the same species in Borneo, Beebe® writes: “Ants, both stinging _ and harmless, form the entire food, although we must extend this general * . ss: 84909, Ann. Soc. Ent. ue, LITT, pp. 58-63. ss tBattikofer, J S00, Meenas aus Liberia,’ (Leiden), 11, pp. 393-395. - iso eke “doektanen SEV p. 197. ‘ oy HS We .° et a VIII, p. 154. a 1914, * ‘pangolin or scaly-anteater,’ Zool. Soc. Bull. New York, XVII, pp. 1141- 320 Bulletin American Museum of Natural History (Vol. XLV term to include the neuropterous white ants or termites. Ihave counted five hundred fire ants in the gizzard of a pangolin, their bites and stings powerless against the sticky, merciless tongue which played and played again among them, each time sweeping away scores. Lacking teeth, the — creature swallows tiny pebbles which, as in a chicken, aid in crushing the hard bodies of the ants.” The following notes on ant-eating mammals in the Belgian Congo have been contributed by Mr. Herbert Lang. “The scaly ant-eaters (Manide), or pangolins, are distributed over southern Asia, part of the Malay Archipelago, and Africa. Those of the Ethiopian Region frequent chiefly the wooded portions where hiding is rendered easy and in the Savannah Province their distribution coincides with the forest galleries. During the day they rest, slightly rolled-up and concealed in any suitable shelter, thus escaping observation. The terrestrial forms usually dig their own retreat. The signs of their fossorial practice are as often a cause of their discovery as is the strong odor they emit, and dogs of native hunters never fail to challenge their presence. Various highly valued talismans, which their captors obtain from the claws, scales, hairs, and other parts of some of the scaly ant-eaters, suffice to make them an always welcome prize and their meat is an additional incentive for their destruction. “The giants among living Manidz are found only in Africa and are represented by two closely related forms. Of these, M. temmincki Smuts is apparently restricted to the southern and eastern portions of. the Ethiopian Region and M. gigantea Illiger (Pl. XXIV, fig. 2; Pl. XXV, fig. 2) to the western parts. The two other, much smaller, species occur only in the West African Forest Province, though the rarer, long-tailed M. tetradactyla Linneus (Pl. XXV, fig. 1) alone is truly arboreal. The most common and smallest of these ant-eaters, M. tri- cuspis Rafinesque, is also an excellent climber, but more frequently re- mains near the ground. It reaches a length of about three feet, less of two-thirds of which are taken by the tail. In M. tetradactyla, the tail is proportionally longer. “The African scaly ant-eaters generally appear so sluggish as to detract much from the interest they otherwise might arouse. Being timid, they readily make use of their natural safeguard and, when even slightly annoyed, roll up in a ball cinched by the grip of the strong, muscular tail. M. tricuspis and M. tetradactyla often hook the tip over the reclined dorsal seales, thus closing up very tightly and sometimes so fast that one has to be careful not to have a finger caught between the 1922} Wheeler, Ants of the Belgian Congo _ 321 geales. When forcibly unrolled, they may succeed in driving off their . -_ tormentors by well directed jets of an ill-smelling, acrid liquidfrom the anal region; native dogs suffer for a considerable time from the effect __ of this substance, which greatly irritates their mucous membranes. The a es sharp claws, however, are not used at all for defense, though in a struggle they may inflict severe wounds. _ “In spite of their timid ways, these animals are not really shy, for, if “a re unmolested and placed near their favored prey, they uncoil readily and, ‘not minding the presence of man, surprise by their agility even more than __ by their cleverness. One soon realizes how thoroughly they are special- ized as ant-eaters, for their methods of attack and disposal of ants are as _ effective as their ways of guarding themselves against the defensive means of their prey. In the regions we visited, the pangolins preferred true ants, as stomach contents clearly showed, though many of our captives would plunder termitaria with great eagerness. After opening the galleries of ants’ nests they watched for a moment the infuriated masses swarming outside to defend their home, adjusted their position, and commenced feeding. , “The feeding process is assisted by many interesting adaptations. The strong, muscular fore limbs readily break into and tear apart the ‘structures built by the tiny insects. Most of the ants that attempt to attack the pangolins are readily shaken off by a shivering movement of the scales. Other protective features are the moist snout; the easily closed, narrow nostrils; the thick, swollen-looking eyelids, acting like heavy pads over the small, globular eyes; and the practical absence of external ears, represented merely by muscular folds which shut the ear- ' opening at will. The mouth, even when fully opened, is but a nar- row tube. The slender, slimy tongue shoots out and, well loaded, slips quickly back into its furrow. The food adhering to it is thus automati- cally pushed off and slides down into the stomach. Immediately the tongue, newly charged with slime, is thrown forward again and the performance continues with great rapidity. A huge gland, providing a steady supply of viscous matter, lies on either side of the throat and in M. gigantea attains the size of a goose egg. The furrow which accom- modates the tongue and is so essential in removing the food and renew- ing its viscosity reaches far back into the thoracic cavity and carries with it the hyoid muscles. Their increased and altered function gives great importance to the sternoglossal muscles fastened upon the xiphi- sternal cartilage, which varies in extension in the different species. In the African and especially in the smaller arboreal forms it is more highly 322 Bulletin American Museum of Natural History [Vol. XLV specialized. In M. gigantea it has the form of a loop, consisting of two, broad, band-like projections distally united and reaching back half the distance from sternum to pelvis. In M. tricuspis and M., tetradactyla, however, two rod-like, cartilaginous projections extend outside the peri- toneum much farther back and, turning upward to the right, are loosely fastened to the last ribs. The prehensile tongue also acts as an organ of - touch and, due to its shape, can follow the intricate turns and windings of the galleries in ant and termite structures. This explains why the in- mates and their larve are cleaned out as by enchantment. An adult male M. tricuspis pushed its tongue into the galleries of a sectioned ant nest for a distance of four inches, moving it just as easily sideways as up or down. After making room to insert its tiny snout, it sniffed into the tunnel, thereby still more inciting the inhabitants that, hurrying to the place of disturbance, were then lapped up so rapidly that it was eifficult to see how well loaded the tongue was as it shot back and forth. “The prey is disposed of so instantaneously that neither the ejec- tion of formic acid, the powerful, pinching mandibles, the armature of spines, nor even the stings of the ants are of much avail. The giant ant- eaters, with their broader, more ribbon-like tongue, are more deliberate than the smaller species in feeding, but their methods are equally effi- cient. From the behavior of various forms observed it appears that they are not affected by the defensive weapons of any of the ants they feed upon. Probably these insects have little chance to make effective use of them before they are enveloped in slime, and later the gastric juices and the triturating action of the stomach render any further efforts impossible. “The variety of ants taken by these pangolins proves that taste alone does not guide them in their choice, and I have already mentioned that our captives fed on termites with the same eagerness. Furthermore, the food, covered with slime, passes through the ecmpletely toothless mouth and throat so quickly that flavor is perhaps of little or no import- ance. In fact, the passage from mouth to stomach might be compared to a chute, and a process replacing mastication begins only after the food reaches the stomach. ‘In spite of this apparent immunity of the scaly ant-eaters, we found that certain kinds of ants are evidently not preyed upon by at least some of the pangolins. Near one of our camps at Avakubi there was anest of robber ants (Megaponera fetens). When we inserted a grass-stalk into its entrance, the owners hurried out to attack the intruder. In a very few moments a Manis tricuspis lying rolled-up nearby was overrun by the on as 1922} —~S™S:—*~S~S*SWheeler, Ants of the Belgian Congo 323 ants, which belabored it with their mandibles as well as with their pain- ful stings.) The pangolin became restless, unrolled by fits and starts, got to , erected its scales, and hurried off to some distance. Then, again and again hooking its fore limbs into the ground, it dragged itself from spot to spot, at every pull exercising considerable pressure against the grass, thus endeavoring to free itself of the tormentors. Rolling up and unrolling and scratching with its claws exhausted its means of defense. “Experiments with other captives of this species showed consider- able variation in individual behavior. One taken near a column of army ants (Dorylus) merely made good its escape, another quickly broke up the well-ordered line. Sitting on its hind limbs and with its tail steady- ‘ing its movements, the fore part of the body was swung about freely. The claws of the fore limbs were kept busy removing those of the fierce assailants that, in spite of the oft repeated shivering movements of the scaly armor, succeeded in gaining a hold. Lashing its sticky tongue through the confused crowds, the ant-eater lost no time in moving back and forth along the ant column as quickly as the dense clusters vanished into its mouth. Its hunger satisfied, it at once retreated, freeing itself of the few army ants that had managed to dig their mandibles into the soft parts of its hide. M. tricuspis fed freely on many other kinds of ants. Those we had alive at Avakubi, Medje, and Niapu were particu- larly fond of ants of the genus Myrmicaria. Brought within reach of such colonies, the pangolitis always turned their attention to the deeper, open tunnels these ants construct across cleared spaces and trails. Here the steady stream of tiny travelers made their meals doubly easy. Curious was the habit of the ant-eaters, especially when sitting partly erect, of turning the outer edge of the tail down and suddenly sweeping into a heap all the fragments of ant or termite structures they had scattered _ about. This gave them a new chance of disposing of their victims that emerged again in numbers from the débris. Though undoubtedly noc- turnal in habits, our captives had no objection to feeding during the day and only the direct rays of the sun interfered with the chances thus offered. “While African pangolins have helped to enrich the stores of witch- craft both helpful and injurious, those of some parts of Japan, China, and Malaysia have furnished the folklore with a curious tale, slightly differing in details in the various regions, on their supposed feeding habits. According to the legend, the pangolin, after tearing open an ants’ nest, erects its body scales and waits until as many ants as possible have 324 Bulletin American Museum of Natural History (Vol. XLV crawled beneath them. Suddenly the scales are pressed down hard, crushing the tiny prey to death; the ant-eater then goes into the water, erects its scales, and proceeds to enjoy a meal of dead ants floating on the surface. c “However great their reputation for slowness, under certain condi- tions the African species seen could proceed in a shuffling manner for a short distance at the rate of eighty yards a minute, the giant species being slightly faster. All four limbs and the tail take an active part, but walk- ing on hind limbs or leaping was not observed. They can sit erect, steadied by their strong tail and pillar-like hind limbs, thus enabling them to carry out any movement with the fore part of the body and greatly increasing their ability to dig and feed. In walking, M. tri- cuspis and M. tetradactyla held their claws in a normal position, the tip of the claws striking the ground. The giant pangolins, however, walk on the ‘‘knuckles” of the fore limbs, so to speak, the claws being folded beneath and slightly turned inward so that only the longer, outer curve of the claw touches the ground. “The strong, prehensile tail of the smaller pangolins, M. tetradactyla and M. tricuspis, is provided on the lower surface of the tip with a rough skin pad of great tactile sensibility. By means of this the long tail can rapidly explore the neighborhood for possible means of progress. It can grasp firmly even the slightest projection, thus enabling these ant-eaters hanging upside down to plunder ants’ nests even more easily than when sitting on a branch, for-at any moment they can pull themselves out of reach of the attacking ants. By forcing the head up over the breast and belly they can hook the claws into their tail as into any nearby branch. These pangolins readily carry out a three-quarter twist with the forward part of the body, or turn back at a right angle to the surface on which they are climbing, and descend any slender tree or branch head downward by quickly shifting the grip of the prehensile tail. “The smaller species, when suddenly frightened while climbing, may let themselves drop from any height, landing uninjured in a rolled-up condition, the flexible scales, backed by the resilient, strong panniculus carnosus, evidently absorb the shock. In the arboreal M. tetradactyla, the long tail, with its sharp-ridged and pointed scales on the under side, is dexterously used in getting about and often serves as stabilizer. As soon as the claws of the hind feet have gained firm hold in the bark the security of the position is greatly increased thereby. The body can then be bent even backward and the free fore limbs are put into action to widen the breach in the ant galleries as fast as the sticky tongue can empty them. ; 1922) —— Wheeler, Ants of the Belgian Congo 325 “The two giant species are terrestrial and fossorial in habits but are rather scarce. They alone have succeeded in holding their own over most of the Ethiopian Region. The Vaal-Orange River in the south and Abyssinia in the north are probably the limits of distribution for M. tem- mincki, and M. gigantea is known from the West African Rain Forest and the adjoining wooded galleries. The latter is the only large species we met in the Belgian Congo; specimens were taken at Bafuka, Niangara, Poko, and Niapu, the largest attaining five feet in length, the tail being Teas than half of this. _ “Near the last-named place various burrows from which Pygmies had secured giant pangolins, both dead and alive, showed that the tun- nels attain a length of fifteen feet and reach about five feet below the surface. In these forests the ant-eaters seemed to prefer the higher- lying, sloping sites for their permanent homes, evidently a safeguard against being drowned in a country with such a heavy rainfall. The heap of excavated soil near the open entrance seldom offers a clue to the real size of the irregular, winding burrow, as the weather rapidly effaces the traces of diggings carried on from time to time. Pygmy boys, with one - end of a strand of rattan fastened to the waist and the other held by friends waiting outside, entered the burrows without hesitation and stated that there is a more spacious resting place at the very end of the tunnel shared often by an adult pair and their young. These boys, armed -with only a knife, merely fastened the rattan around the live pangolin, which they prodded from behind while their companions pulled it slowly out of the hole. These otherwise harmless beasts, when touched while rolled up, suddenly switch their tail sidewise with such force that, if one’s hand is caught between the rough body scales and the tail, it is seriously mutilated by the shearing action. Natives of the Ituriand UVele districts claim that the giant pangolins stay for weeks at a time in their burrows, but it is certain that at times they leave them several nights in succession to feed. One trailed to its underground home after a heavy rain was caught in nooses eighteen days later when trying to escape. “One might think that animals so large and muscular would need great quantities of food, but this is only relatively true, for their sluggish habits considerably reduce the demands for nourishment. An adult male from Niapu measured 1530 mm. from snout to tip of tail, the latter ac- counting for 690 mm. The capacity of their stomach is relatively small, hardly more than two quarts (about two liters). In an adult female the stomach measured antero-posteriorly only 170 mm. and dorso-ventrally 70 mm. Biittikofer’s remark, cited by Dr. Bequaert (p. 319), about a 326 Bulletin American Museum of Natural History [Vol. XLV stomach of M. gigantea containing six liters of ants is evidently due to a slip of the pen. Ants and other food arrive intact in the stomach of the pangolin, but soon afterwards appear well disintegrated. ‘In general outline and arrangement, the stomach of these giant ant- eaters is similar to that of M. javanica which is so well deseribed by Weber (cited p. 318), though without the horny, tooth-like structures in the pyloric part. Nor does the large gland situated near the middle of the great curvature terminate in a common orifice but it presents an even surface, the individual follicles of the oval patch secreting directly into — the stomach. In the absence of teeth, the stomach, with its highly special- ized grinding mechanism, has become an organ replacing mastication. Half a handful of pebbles, the largest not exceeding five millimeters, usually found in the cardiac section, and the wall-muscles assist the trituration of food. These, and especially the more heavily muscled and distinct pyloric section, remind one of the gizzard of gallinaceous birds. The larger of the little stones are probably selected for this very purpose and are not incidentally introduced with the food as may happen with fine grit and other débris. “The stomach is divided into two distinct aris: a larger cardiac * and a smaller pyloric section. Both have somewhat the function of a gizzard. The cardiac section is lined with pavement epithelium and irregularly folded except for the large, well-defined gland patch. Here the processes of assimilation are greatly advanced. The food, mixed- with the excretion of the glands, is easily ground to pieces between the pebbles and fine grit; usually only the hard, chitinous covering of the head and the strong mandibles of the ants escape being crushed in this section. The milling process is carried much further between the power- ful muscles of the pyloric section, which is well set off from the other. This portion is lined with an epithelium similar to that in the cardiac section. Near and along the pyloric orifice there is, however, a well- defined smooth glandular area. The semicircular muscular mass, opposed on either side by, or rather fitted between, two other strong muscular pads isimportant. With the assistance ot fine grit, this arrange- ment works much like a mill and the food before reaching the pylorus is transformed into a finely ground mass from which nourishment can easily be assimilated. The intestine is without cecum and rather long, measuring in an adult female 10.8 meters. A large amount of the black- ish, hard excrement consists of the glossy particles of chitin of ants. —_——_—--_ | Pe. ¢ a oT aaa ‘Audead te Belgian: Congo 327 ago ‘ago for the latter by Klinekowstiten,' are aléo divided into two parts. The cardiac section is lined with horny Sn OTe eT the mucous cides, the muscular portion of the pyloric section in both is much like __ that of M. gigantea. A mass of fine grit also helps pulverize the ants Bee ager the extended milling process. “The numerous forms of the aardvark, Orycteropus afer (Pallas) “ a. XXIV, fig. 1), are distributed over most of the Ethiopian Region “ay and are equally common in the Savannah and Western Forest Provinces. ; Their food consists of white ants (termites), and true ants are only in- ite eidentally taken, as they often inhabit termitaria. In external char- _ acteristics the aardvark resembles a pig, about six and one-half feet in total length, with a slender head, long ears, and a heavy, tapering tail Pe about two and one-third feet long. Its very muscular limbs with their 3 enormous claws denote fossorial abilities. The mouth is small, the snout . slightly protruding and rather easily moved. The nostrils can be opened and shut at will and the edges are set with a dense border of short, stiff _ bristles turned outward in such fashion as to prevent insects from enter- __ ingthenose. The long, extensile tongue is of relatively normal shape and _ the rather flat-crowned, peculiar cheek-teeth are capable of crushing food. __The stomach lacks the highly specialized triturating organs of the _ Manidx, though strong, muscular walls are present in the pyloric sec- tion. The absence of stone and grit also indicates that the gastric juices play the most important réle in the disintegration and digestion of food and are sufficient to assimilate the soft-bodied termites but not the well- chitinized ants. Numerous parasitic worms are thus enabled to live in the stomach. _ “Of the many aardvark burrows seen near Faradje, those with one entrance were scarcer than those with two, but three and even as many _ as eight openings to a single retreat were recorded. In one case the three entrances to a burrow were as much as fifty feet apart. Many of the tun- nels, which reached about five feet below the surface, were deserted; those inhabited seemed to indicate that the aardvarks occupy them at intervals and occasionally dig holes merely for shelter. At times these inoffensive animals are driven out of their lodgings by warthogs and py- wie, i 11895, Zool. Jahrb. Abt. Anat. Ontog., VIII, p. 495. 328 Bulletin American Museum of Natural History [Vol. XLV thons, the latter being known to feed upon their young and to estivate in their burrows as well as in the cavities of termite hills. “The extensive tunnels of the burrows are large enough to admit a small man, who, among the Logo at Faradje, is armed with a short spear but trusts far more to his talismans for protection. The beast usually tries to save itself by. digging, throwing the excavated earth into the face of its pursuer. Should the aardvark succeed in walling itself off, the undertaking is generally given up. If, however, the native is able to kill it, he indicates his position to an eagerly listening friend by tapping against the upper wall of the burrow. As rapidly as possible a shaft is sunk in his direction and the valiant hunter and the aardvark are lifted out. The meat is highly prized and in many regions the body has to be presented to the chief before any of the parts containing powerful medi- cines are yemoved. “One of the African mongooses, Bdeogale nigripes Pucheran (PI. XXV, fig. 3), not only satisfies its regular carnivorous instincts but, as stomach contents proved, feeds also on ants and termites. This mon- goose attains the size of an otter, which it resembles in general appear- ance. It has large, dagger-like canines and an otherwise strong denti- tion; the palate is relatively wide and, especially in the young, has the general shape of that of the termite-feeding Proteles cristatusSparrman. “Of nine specimens collected, the stomach of one contained ter- mites and those of three, driver: ants which filled two of them to capacity; their possible incidental introduction with other food need, therefore, not be considered. The ants in the stomach were only slightly chewed and some of them were completely intact. This carnivore seems to have no adaptations that would allow it to devour with impunity an insect so dreaded by most other animals. It may be that these driver ants are swallowed dead, since they are often killed in masses when their droves are unexpectedly exposed to the deadly effect of the direct rays of the sun, as it may happen after a shower, when they are still on the march or feeding in great numbers on carrion. ‘“‘ Among other mammals, some of the insectivores, especially certain Macroscelidide (Rhynchocyon, Elephantulus, and Macroscelides), are credited with occasionally feeding on ants, probably those emerging from the ground in masses during their nuptial flights and which are easily taken. “Chimpanzees (Pan schweinfurthii Giglioli) are well known to he omnivorous and, in addition to their regular vegetarian diet, feed on many insects and their larve. Nevertheless, I was surprised to see to > ihaaies dels af he Balelad Cones 329 cs, oe it laaseble they would go merely to relish half a handful of cocoons that . a nest of robber ants (Megaponera fatens) might contain, leaving un- touched all the dead ants they might kill in the process. These prjmates are evidently not deterred in their raids by the painful sting and strong, Be "pinching mandibles of the ants. They seldom pass one of the rather in- ae , temporary ant nests, which are marked only by a small 4 heap of iphenvated particles of earth near the open entrance, without digging it out. After uprooting the plants they sometimes scoop out a hole one or two feet deep: In the Rain Forest near Niapu, I saw about e. seven nests of these robber ants destroyed in this manner. a “Colonies of certain large Camponotus are also looted by the chim- Ss panzees, which, in this case, are fond of the ants themselves. Hollow or decayed trees are torn apart and the galleries searched for these ants, _ which, when attacked, do not swarm out but retreat speedily into the pa _ deeper recesses they excavate. In the forest about Niapu and on the a road to Medje there were several such instances. In one case a troop of ___ five or six of these anthropomorphs must have spent considerable time in trying to tear open a hollow portion of a log.” Man Certain species of ants constitute an important article of food with many uncivilized peoples, especially of the warmer parts of the globe.' In tropical Africa the large, winged queens of Carebara vidua F. Smith, which at certain seasons emerge in great numbers from termite mounds, are often highly prized as delicacies; they are eagerly gathered for their ___ swollen gaster, which is eaten raw or roasted.?_ In Kanara and other parts of India, and throughout Burma and Siam, a paste of the green weaver ant, @eophylla smaragdina (Fabricius), is eaten as a condiment with eurry.* Beccari* also records that the Dyaks of Borneo “eat this ant, eet ene tic aeeece, <2, DD. 20-51) mentions thet the Comedies the cits a n nicus a CE Sa J., 1913, Rey. Zool. Afr., II, pt. 3, p. 429. Mr. Lang, who observed the nuptial flight of Cc. vidua at in Mare 1915, also notes that these ants are comestible; “‘ only the eaten by the natives, sometimes raw, sometimes fried, also crushed.’ ., 1903, ‘Fauna of British India. Hymenoptera,’ (London), II, p. 311. (an Murries of a southernmont Native State in ae aeong 2 Provinces—use the red cophylla smaragdina non superar ticle . Throughou ear eat meveemonntye uring ge Teg a oes Berrie —conert nests of the red ants, and after tearing open, shake out contents mae a IX, soe Yom beat the atecasmiees and immature—into a mass enclose them in a packet, about the size of a goose's egg made of sal leaves. Ay Rees od the article is taken to the bazaar and sold, about 16 being sold for a pice, or 4 cowries each. To the squashed ants for food, they are mixed with salt, turmeric and chillies, and ground down cen stones, and are then eaten raw with boiled rice. They eeenraes cocked up with rice flour, salt, chillies, etc., into a thick paste, and in this condition the food is said to ge the eater of it great power of resistance against fatigue and the sun's heat” 1, ‘Red ants as an article of food.’ Journ. Bombay Nat. Hist. Soc., XIII, No. 3, p. 536). 1004, * ip the great forests of Borneo,’ (London), Pp. 161, (1903. ‘Nelle 330 Bulletin American Museum of Natural History [Vol. XLV or rather they mix it with their rice as a condiment; it has a pungent acetic taste and smell which they evidently like.” The same ant is used by the natives of North Queensland mashed up in water, like lemon squash, and forms the basis of a pleasant acid drink appreciated even by many European palates.’ Moreover, it is generally known that certain American Indians eat ants, as well as other insects, freely. This is especially true of the tribes that are but little inclined toward agriculture, periods of famine with them being rather frequent, due to the absence of permanent vegetable staples.? In his delightful book, ‘My first summer in the Sierra,’ John Muir* tells how the Digger Indians of California are fond of the larve and even of the adults of the large jet-black, wood-boring ants (Campono-_- tus), of which “they bite off and reject the head, and eat the tickly acid body with keen relish.”’ In his account of the honey ants of North Ameri- ca, MeCook‘ remarks that the uses to which the Mexicans and the Indians ot the southwestern United States put the replete of Myrme- cocystus are various. ‘That they eat it freely, and regard it as a delicate morsel is beyond doubt. Prof. Cope, when in New Mexico, had the ants offered to him upon a dish as a dainty relish. The Mexicans (Loew) press the insects, and use the gathered honey at their meals. They also are said to prepare from it by fermentation an alcoholic liquor. Again they are said (Edwards) to apply the honey to bruised and swollen limbs, ascribing to it great healing properties.” One finds in the narratives of Barrére,’ de Azara,® Humboldt,’ Rengger,® Richard Schomburgk,’ and other travelers” frequent allusions to the fondness of many South American tribes for the large males and ‘Saville Kent, W., 1897. ‘The naturalist in Australia,’ (London), p. 253 *See Skinner, 'A., 1910. ‘The use of insects and other invertebrates as food by the North American Indians.’ Journ. New York Ent. Soc., XVIII, pp. 264-267. *Muir, J., 1916, * My first summer in the Sierra,’ (Boston and New York), ip. & McCook, H. C., 1882. ‘The beg: ants of the Garden of the Gods and Occident ants of the American lains.’ (Philadelp’ phia), p re, P., 1741, ‘Essai sur l'histoire naturelle dela France Equinoxiale,’ (Paris), p. 198. Speaking of an a of British Guiana which he calls Formica major, volans, edulis, this traveller writes: “Cette fourmi est passagére et parait en grand nombre au commencement des pluies. a négres et les créoles — mangent le derriére de cet insecte, qui est une sorte de petit ge de chiche, rempli d’une ee blanchatre, miellée, qui ne parait étre autre chose que les ceufs oe dépose dans ce tem 1h00. tde Azara, F., 1 * Voyages dans 1’ Amérique a, Pei B Bai 199: * ‘ Les habitants ville de Santa Fé, qui est de ces c6tés-ld, vont A la chasse de ces fourmis ailées: on = Set panes qui est fort grasse, on la fait frire, et on la nent SS ou bien hen suaneeee on les au sirop et on les mange Vit» .” After quoting thie pe Se (1916, An. Mus. Nac. Buenos Aires, XXVIII, — that the" gaster of of Atta sexdens (Linnseus), called tanajGra, is still eaten ton the THumboldt, Al. de, 1822, °V ti aux Oy # équinoxiales du Nouveau Continent, fait par A. de Humboldt et A. Bonpland,’ (Paris), ie pp. 4 a et A., 1835, ‘ Reise nach jae + Gea deleen 1818 bis 1826 von Dr. J. R. Rengger,' arau), Pp *Schomburgk, Rich., 1848, ‘ Reisen in British Guiana,’ (Lei II n, J., 1876, ‘The Andes and the Amazon,’ 3d ea Sew ow ork), p. 301; Spruce, Rich., 1908, ‘ Notes a botanist on the Amazon and Andes,’ (London), te AIXX #4970 ‘A'TX “TOA ‘HN ACY SiatTiog Fig. 2. Manis visits Illiger. Live ace Typical pose of the animal while in search of its food. in Fig. 3. Bdeogale nigripes Pucheran. Freshly killed male, at 1913. Bouceti~ A. M. N. H. Vou. XLV, Pirate XXV ‘ ; . . i : a ‘ - ’ ___IV.—ANTS IN THEIR DIVERSE RELATIONS TO THE PLANT WORLD are By J. BeQquarERT The following contribution is an attempt to summarize what is Tea known at present of the widely varied and often intricate relations which __ exist in nature between ants and vegetation. It has primarily grown out ___ of an examination of certain so-called “‘myrmecophytes,” or ant-plants, which I frequently came across during my travels in the Belgian Congo ; ; _ in quest of zoological and botanical specimens. Prof. Wheeler’s study _ of the feeding habits of the larve of certain plant-inhabiting ants col- _ leeted by Messrs. H. Lang and J. P. Chapin and myself in the Congo, - and Prof. I. W. Bailey’s investigation of the anatomy of myrmecophytes show that the whole subject of the mutual adaptations of plants and ants is in need of a thorough revision. It thus seemed appropriate that the present opportunity be used to bring together the many isolated and ‘seattered observations which have been made on the interrelations of these organisms. Indeed, the problem of myrmecophytism is dominated by the feeding habits of the ants and their young and, until these are perfectly understood, we can scarcely hope to grasp the true ecological meaning and the probable origin of the extreme cases of apparent or true symbiosis between certain ants and certain species of plants. It is, therefore, of the utmost importance to inquire carefully into the various ways in which ants are wont to benefit by the vegetation. Although much time has been spent in consulting botanical and entomological papers for information bearing on the subject, undoubted- ly a number of interesting observations have escaped my notice. In keeping with the general purpose of the present contribution to African Myrmecology, especial attention has been paid to work accomplished in the tropics, primarily in Africa. The bibliography appended to this part is as complete as possible. In it are included many papers which may seem to have but remote connection with the subject—such as, for instance, those on fungus-growing termites, intracellular symbionts, and the like. I believe, however, that they are indispensable in reference to the study of certain activities of ants. Some students may find the botanical side rather too fully treated, but this seemed unavoidable in providing the necessary background for future field work, especially to the myrmecologist. Moreover, I am convinced that the ultimate solution of many of the problems involved can only come from a close eoéperation between botanical and entomological experts, and this, under present conditions of specialized training, is not so easily realized. 333 334 Bulletin American Museum of Natural History {Vol. XLV Many ot the data here presented were brought together during the several weeks I had the pleasure of spending at the Bussey Institution for Applied Biology of Harvard University. I wish to thank President Henry Fairfield Osborn and the authorities of The American Museum of Natural History for the liberal manner in which I have been able to carry on this work. I am also under great obligations to Professors Wm. M. Wheeler and I. W. Bailey, of Harvard University, for their many suggestions and criticisms during my stay at the Bussey Institution. The interest they have shown in the work has been a steady encourage- ment and their advice invaluable. 1. Vartous RELATIONS BETWEEN ANTS AND VEGETATION Economic Importance of Ants The question whether ants are, broadly speaking, noxious or bene- ficial insects is still debated by agriculturists and economic entomologists. While it is believed on the one hand that ants attack and mine only sick and de- caying plants, especially decaying roots, on the other hand it is claimed that healthy plants, which show no trace of disease, are also assailed by ants. In any case further exact observations concerning the relation of ants with plants will be needed in order to clear up this problem. The elucidation of the question of the direct noxi- ousness of ants to plants is the more desirable, since we possess in the ants partly a — welcome help against other animal enemies of culture-plants, which they pursue and | destroy. It is therefore necessary that we learn more in detail whether their harm- fulness outweighs their utility or vice versa. In general one can perhaps say that, judging from statements which have been made thus far, their noxiousness to plants, by attacking roots, stems or branches, is not very great. (G. Aulmann and W. La Baume, 1912, p. 61.) In their recent study on the feeding habits of ants, Wheeler and Bailey (1920, p. 236) have pointed out that one reason why the economic importance of many common ants remains so dubious or ambiguous is the lack of precise information with regard to the quality and quantity of their food, especially in the larval stage. These authors have shown, for instance, that ants carry on their bodies and in the food-pellets of their infrabuccal pockets an extraordinary number and variety of fungus spores and bacteria. It is, therefore, quite possible that these insects have a great but hitherto only vaguely apprehended importance as carriers of the germs of certain plant, animal and human diseases. That ants are active carriers of pathogenic micro-organisms has been further suggested by Darling (1913), Wheeler (1914), Studhalter and Ruggles (1915), Grabham (1918), and Bailey (1920). . ll “Wheeler, Ants of the Belgian Congo . 335 The leaf-cutting ants of the tribe Attini, so abundant in tropical __ and subtropical America, are decidedly destructive to the vegetation and are rightly considered one of the worst pests to South Americkn agricul- ture. Accounts of their depredations are found in practically all narra- _ tives of South American travellers. Though they attack many of the _ native herbs, shrubs, and trees, they often show a predilection for cul- tivated plants.. It is no uncommon thing to find the sativas, Atta __ cephalotes (Linnzeus), so numerous in certain spots that the planters are _ forced to abandon their fields. Speaking of the ants in the Brazilian ___ eoffee districts, Van Delden (1885, pp. 297-298) writes: “The enemy most dreaded in the fazendas (plantations) is indubitably the sativa, or ___ tana-jura, a dark-brown ant, two centimeters long, which undermines the ____ ground by digging extensive passages and dens in all directions. It __ attacks all sorts of trees, the coffee-shrub among others, but has a _ decided preference for the orange and citron trees in the coffee gardens.’’ ____ H.W. Bates (1863) and others have noted that these ants often become ___ troublesome to the inhabitants because of their habit of plundering the 3 stores of provisions in houses at night. The Attini are not represented in the Old World tropics, but pos- _ sibly ants of other groups have developed similar habits there, though on a smaller scale. G. Aulmann (1912, p. 156) and Moorstatt (1914) mention that a leaf-cutting ant was observed in German East Africa at times causing considerable damage to cotton plants. The specific iden- tity of this ant has not been ascertained, but it probably belonged to the genus Messor, which is known to collect pieces of grass in addition to seeds and grain (see Sjéstedt’s observation quoted below, p. 359). King (1911) also notes that Messor barbarus (Linnzus), at Khartum, damages garden plants by biting off and carrying away the leaves, and adds that in cotton fields the sites of their nests are marked by bare patches devoid of vegetation. What use these ants make of the vege- _ table matter thus carried into their nests has not been investigated. There are a few other cases on record of ants directly destroying living parts of plants. It is generally known that certain ants will injure buds and fruit in order to feed on the exuding sap (see Miller-Thurgau, 1892, pp. 134-135). Forel (1885, p. 338) mentions instances of Tetra- morium cespitum (Linnwus) attacking young roots of healthy sugar- beets at Vaux, Switzerland, many of the plants dying from the injuries received. J. Pérez (1906, pp. xxxii-xxxiv) records the havoc played by the same ant on the tubers of potato, near Bordeaux, more or less deep cavities being excavated and many young plants killed; 7. c#spitum 336 Bulletin American Museum of Natural History [Vol. XLV was also found burrowing superficial galleries in the stems of living potato plants and attacking the roots of young cabbage and carrot.? In North America, Solenopsis geminata (Fabricius) and S. molesta (Say) often do injury to the soft parts of planted seeds, and the former also to strawberries (Webster, 1890) and other fruit. S. molesta has proved very injurious in gardens and fields; the chief damage is done to seeds of sorghum and corn, which are hollowed out undoubtedly for the purpose of extracting the oils (McColloch and Hayes, 1916; Hayes, 1920). According to Green (1900a) and G. R. Dutt (1912, p. 247), the Indian Dorylus orientalis Westwood is mainly or exclusively herbivorous, feed- ing on the bark of trees and the healthy tubers of plants, a habit the more remarkable since the majority of Doryline are highly carnivorous. In Cameroon, certain ants have been seen attacking the fruits of cacao- trees: Camponotus maculatus subspecies brutus (Forel) gnaws the base of fruit-stalks where they are inserted into the trunk, licking up the sap at the wound, causing the fruits to drop off or dry; Crematogaster africana variety winkleri (Forel) gnaws away the skin of the cacao-fruit, often almost completely; while Camponotus acvapimensis Mayr and @cophylla longinoda (Latreille) are accused of the same evil, though they cause but little damage (H. Winkler, 1905, pp. 129-137). The greatest harm to the vegetation is undoubtedly done indirectly, both in tropical and temperate regions, by a host of species of ants that have a pronounced fondness for pasturing and guarding plant lice, scale insects, tree-hoppers, and other plant bugs on roots, stems, and foliage; all these Hemiptera suck the juices of plants, and their protection by the ants must, therefore, be regarded as pernicious. The “milking” habit among ants seems to be of very frequent occurrence, evidently because it offers so many advantages over direct feeding on plant-juices. Not only is the food supply much more abundant at any one time and within easier reach, but, in addition, the plant saps undergo chemical changes in the digestive tract of the Hemiptera, whose anal secretion, on which the ants feed, therefore contains a great amount of invert-sugar, instead of the much diluted cane-sugar of the plant. Many of the aphids at- tended by ants have undergone adaptive modifications of structure and behavior which show that their relations with ants have become of a mutualistic nature, and it is probable that the same will be found true for some of the ant-attended coccids and membracids of the tropics. 1This habit of Tetramorium cespitum in attacking subterranean of plants was known to Lin- nzeus, since he adds to the original description of this ant (‘Syst. Nat.,’ Ed. 10, 1, 1758, p. 581): “ Habitat in Europe tuberibus.”’ It is rather surprising that injuries by this ant have been so little noticed in —- Laces Concerning ants noxious in gardens, see also F. Heim (1894), Andersson (1901), and Cooley ( ix ie 1922] Wheeler, Ants of the Belgian Congo 337 % Indeed, the association between phytophagous Hemiptera and ants Offers a typical illustration of symbiosis in the strict sense, advantageous _ to both parties. The benefit that accrues to the ants has been explained __ above and needs no further comment; that derived by the Hemiptera, however, is of a more complex‘nature. It is obvious that the ants pro- tect the plant bugs by driving away coccinellid beetles, ichneumon _ flies, and other enemies. In the case of aphids and coccids the ants frequently build tents or cowsheds over these insects, which thus con- _ tinue to suck the juices of the plant while being “milked” by the ants. and are, at the same time, protected from their enemies, from alien ants, and intemperies, and prevented from escaping to other plants. ____ The tent-building habit was discovered by P. Huber (1810, pp. 198-201) for Lasius niger (Linneus) in Europe, and Forel (1874, pp. 204-205 and 420-422) gives an interesting account of it in his classical ‘Ants of Switzerland.’ Lasius niger has similar habits in North Ameri- ea (Wheeler, 1911b) and Japan (Stopes and Hewitt, 1909, pp. 1-6). _ This ant builds its tents of detritus or wood-fibres; while, according to Forel, certain species of Myrmica enclose their aphids in earthen cells, ___ which communicate with the ground nest by means of covered galleries. _____ Wheeler (loc. cit.) has described in detail the tent-building of the North American Crematogaster lineolata (Say) and I have.found that several __ African members of this genus which attend coccids have similar habits. Certain North American species of Lasius (L. flavus, L. niger, and the species of the subgenus Acanthomyops) which live to a very large extent or exclusively on the excrement ot root-aphids and coccids, remain throughout the year the constant companions of the lice, even hoarding in their nests during winter the eggs or the wingless, agamic form of the aphids and the fertile females of the scale insects. Forbes (1896), Webster (1907), and others have shown that the common North Ameri- can Lasius niger variety americanus Emery guards the eggs of the corn root aphid (Aphis maidi-radicis Forbes) throughout the winter, shifting them about, as it does its own young, to accommodate them to changes of weather and moisture. In spring, the young lice, on hatching from these eggs, are conveyed by the ants during fair weather to the roots of various weeds, being taken back to the burrows in bad weather or on cold nights. After the corn plants have started to grow, the young root lice, all of which belong to the wingless, agamic form, are transferred from the weeds to the roots of young corn, where they are tended throughout the spring and summer. It would thus appear that, without the aid of the little brown ant, this aphid is unable to reach the corn plants. 338 Bulletin American Museum of Natural History [Vol. XLV Still more surprising is Lubbock’s observation that Lasius flavus cares in a similar manner for the eggs of certain aphids on the aérial portions of plants. The eggs are laid early in October on the food-plant of the insect. They are of no direet use to the ants, yet they are not left where they are laid, where they would be exposed to the severity of the weather and to innumerable dangers, but brought into their nests by the ants, and tended by them with the utmost care through the long winter months until the following March, when the young ones are brought out and again placed on the young shoots of the daisy. This seems to me a most remarkable case of prudence. Our ants may not perhaps lay up food for the winter, but they do ‘more, for they keep during six months the eggs which will enable them to procure food during the following summer. (Lubbock, 1880, p. 184.) In temperate regions the honeydew (or sugary: excrement) secreted by aphids from the posterior end of the alimentary canal is eagerly sought for by many of the common Myrmicine, Dolichoderinez, and Formicine,' those attending root lice being especially harmful to the vegetation for the reasons mentioned above.? Certain tropical ants also nurse root-aphids. In Java, Acropyga acutiventris Roger may thus be- come a serious pest to coffee plantations, and, according to Forel, vari- ous species of Rhizomyrma attend root lice in South America and New Guinea (K. Escherich, 19116, p. 227, footnote). In the tropics, however, aphids are far less common than in colder climes and are there replaced - as ant “cows” by various Coccide, Membracide, Fulgoride, Cicadel- lide (Jasside), and Psyllide, certain members of these families being occasionally attended by ants even in North America and Europe. The relations between various species of tree-hoppers and certain ants have been recently reviewed by Funkhouser in his ‘ Biology of the Membracide of the Cayuga Lake Basin’ (1917, pp. 399-404), to, which the student is referred for further details. Funkhouser comments on the number of unsolved problems in connection with this subject. One of the first of these questions is suggested by the fact that some of the species are attended by ants while others are unattended although there are apparently no physiolcgical or anatomical differences to cause the distinction. Another question arises from the fact that certain species attended locally have never been reported as being attended in other parts of the country, while on the other hand some of the species that are never attended in this basin are always attended in other localities. Again, certain species that the ants ignore in this basin are represented by closely related species in other regions and these exotic forms—often of the same genus and very near systematically—are well attended. 1It would appear that these trophobiotic habits are of great apnenty aesee ants, dating as far back as the Tertiary. Wheeler (1914, p. 21) found a block of Baltic amber containing a number of workers of Iridomyrmex gépperti (Mayr) together with a lot of their aphid wards. *See the publications of 8. A. Forbes on the corn root aphid, | in the bibliography; also Gar- man’s (1895) account of the bean root louse. 4 Wheeler, Ants of the Belgian Congo 339 ar a He also notes that certain common species which, in the nymphs at ist, appear to exude the characteristic anal fluid when disturbed, never- less are not attractive to ants. He found the following species of ib attended by ants in the vicinity of Ithaca, New York: ia bimaculata (Fabricius), Telamona ampelopsidis (Harris), T. unicolor Fitch, Cyrtolobus vau (Say), Atymna castanee (Fitch), Ophiderma sscens (Emmons), Vanduzea arquata (Say), Entylia bactriana Germar, Publilia concava (Say). The following ants were actually observed by Funkhouser taking ‘the ‘secretion from the membracids: Formica truncicola subspecies curiventris (Mayr), Formica exsectoides Forel, Camponotus pennsyl- anicus (DeGeer), Crematogaster lineolata (Say), and Prenolepis imparis Say). All these ants seemed to make no distinction between the various species of tree-hoppers listed above and the mutual behavior of these insects was much the same in all the cases studied: ‘‘The ants stroke their charges with their antenne, whereupon the membracids give off from the anal tube a liquid that issues in bubbles in considerable quan- tity. ‘The anal tube of the membracid is capable of great evagination especially i in the nymphs, in which it is long and cylindrical and usually _ tipped with a fringe of fine hairs. The honeydew is eagerly taken from _ the end of this tube by the ants. In many species the adults as well as ___ the nymphs are sought, and the ants seem to be as attentive to one as to e the other but the adults have not been observed to excrete the liquid to the same extent as the nymphs.” (Funkhouser, 1917, p. 403.) The liquid sought by the ants “‘is colorless and transparent, rather heavy and somewhat sticky. When first exuded it is inclined to be frothy, due no ____ doubt to bubbles of air which emerge with it, but it quickly clears on __ settling. Itis practically tasteless even in comparatively large quantities, and many attempts to distinguish a sweet taste have proved unsuccess- _ ful. The term honeydew, therefore, commonly applied to the fluid, is hardly a descriptive one. It is very likely, of course, that the liquid may contain sugars not detected by the human tongue, and this would _ seem to be indicated by the fact that fermentation appears to begin if the substance is left exposed. No chemical analysis of honeydew has been made.”’ (Op. cit., p. 404.) Miss Branch (1913, pp. 84-85) states that young Entylia sinuata seemed unable to molt successfully without the presence of ants. This fact led her to believe that the ants are necessary factors in the life of an individual membracid. Funkhouser’s experiments, however, gave no support to this theory. Tree-hoppers of many species were reared 340 Bulletin American Museum of Natural History [Vol. XLV in the field and in the insectary, with and without ants, and no differ- ence was noted in the length of the instars or success of the molting process. Kornhauser (1919, p. 546) gives the following account of the manner in which Thelia bimaculata (Fabricius) is attended by ants. This mem- bracid feeds on the sap of the common North American locust tree, Robinia Pseudo-acacia Linneus. It deposits its eggs in slits in the bark, where they remain during the winter, hatching in early June. The first, second, and third instars occur on the branches, constantly attended by ants: In my principal collecting fields [at Cold Spring Harbor, New York], Formica truncicola Nylander subspecies obscuriventris and Cremastogaster lineolata Say were the chief ants associated with Thelia. When tapped by the antennz of the ants, the Thelia nymph or adult exudes from the anal tube a drop of clear fluid which is taken by the ant with great alacrity. | Toward the middle of June, the ants build collars about the bases of the locust trees, and inside these collars in the cracks of the bark are to be found hundreds of Thelia nymphs of third to fifth instar, quietly feeding and undisturbed by the numerous ants in attendance. In this moist situation, protected from many of their enemies, the nymphs thrive. Formica builds the protecting collar of leaves, twigs, and bits of wood; Cremastogaster builds of sand grains cemented together. When one breaks the collar, many ants swarm out and attack the intruder, Formica biting one’s fingers ferociously, while others grab the Thelias and drag them into underground passages. These pugnacious ants seem to have complete mastery of the Thelia nymphs. Membracide are sometimes carried by ants into their formicaries (Enslin, 1911, pp. 19-21; W. M. Mann, 1915, p. 162), but they usually die soon, probably due to lack of food.' Lamborn (1914) has described in detail several cases of trophobiosis between ants and coccids, membracids, jassids, and psyllids in Southern Nigeria. Regarding Leptocentrus altifrons Walker, a tree-hopper which is invariably ant-attended in its mature and larval stages, he writes as follows: ‘‘The solicitude of ants for the larve has a very definite object, for they are extremely partial to the fluid excreted at the anal extremity, and I remember seeing a Camponotus akwapimensis variety poultoni with the caudal] whip of a membracid larva actually in its mouth.”’ (Lamborn, 1914, p. 495.) I have on several occasions, in the Belgian Congo, col- lected ants which were in the act of attending tree-hoppers: so, for in- stance, in April 1912, at Elisabethville, Katanga, a number of workers of the common Pheidole megacephala subspecies punctulata (Mayr) were 1Additional information concerning the relations between Membracid and ants is given pf Leo (1874), Mrs. Rice (1893), Green (1900c), Froggatt (1902, p. 717), Baer (1903), Buckton (1903, p Poulton (in Buckton, 1903), Distant (1908, p. 209), Enslin (1911), Miss Branch (1913), Keesha att Lamborn (1914), and others. ~ Wheeler, Anis of the Belgian Congo 341 a iarigaged i in licking the sweet excretions of some of these hemipter- ous insects feeding on a bush; again, at Welgelegen, Katanga, | Myrmicaria eumenoides subspecics opaciventris variety congolensis (Forel) was found attending membracids fixed on the calyx of a -malvaceous plant (Bequaert, 1913, pp. 427 and 428). Bell-Marley at Durban, Natal, observed that the common South African tree-hopper, 2 : Oxyrhachis tarandus (Fabricius), attracts great numbers of “small red = Sram (Distant, 1908, p. 209.) _ The nursing of scale insects by ants has repeatedly been noticed s ee te Cockerell, Newstead, King, and others.! A rather interesting phase is offered in the case of various ants which keep coccids inside the ___ swellings of myrmecophytes. Zimmermann found Lecanium tenebri- __ cophilum Green at Buitenzorg, Java, together with ants in living branches Of Erythrina lithosperma Blume (Green, 1904, p. 204). In southern __ Europe, Crematogaster scutellaris (Olivier) and Camponotus pubescens (Mayr) often become harmful to olive trees by the care they bestow __ upon scale insects (Peragallo, 1882). Keuchenius (1914a and 6) holds the view that @cophylla smaragdina is very noxious to coffee plantations through its habit of keeping and protecting in its nests the green coffee seale, Lecanium viride, one of the most serious pests to the coffee tree. Gowdey (1917) also mentions that the root form of Pseudococcus citri, a parasite of coffee, orange, lemon, and cacao in Uganda, is attended by the ant Pseudolasius gowdeyi Wheeler. Most of the wood-boring ants either accommodate themselves to pre-existent galleries made by other insects or attack dead wood only. Occasionally they find their way into houses. Forel (1874) and R. Brun (1913) have described cases in which populous colonies of the European Camponotus ligniperdus and C. herculeanus had excavated the beams, ____window-sills, and other wooden parts of buildings. Certain carpenter ants of temperate regions (Camponotus ligniperdus, C. herculeanus, C. pubescens, and others) extend their burrows into healthy wood (Forel, 1874); they may thus become very destructive in forests, the more so since they attract woodpeckers, which bore large access-holes through the perfectly healthy outside layers of the tree in order to feed on the carpen- _ ter ants and their brood. 8S. A. Graham (1918) describes how carpenter ___ ants of an unidentified species are responsible for great damage to stand- -, India the lac-producing coccid, or lac insect, Tachardia lacca (Kerr), is frequently attended by oe te ater submuda (Mayr) and C ‘amponotua compreasus (Fabricius), which may become a annoyance to the lac grower. In their eagerness to obtain * ‘honey dew" the workers pe ge on the white filaments, the two anterior of which are connected with the respiratory apparatus of the lac insect, the coccid being killed consequently (G. R. Dutt, 1912). ae | 342 Bulletin American Museum of Natural History [Vol. XLV ing white cedar in Minnesota, at least twenty per cent of the trees cut showing ant injury to the stump. In this case, so far as observed, the ant never attacks a sound tree, but always gains entrance through a wound or decayed spot. When a colony has been established in a tree, the ants usually work well above the rotten area into the sound heart- wood, honey-combing the tree with longitudinal galleries until there is often only a thin outer shell of solid wood. From the main nest they cut openings to the outside, frequently following a knot, through which the sawdust can be cast and through which the inhabitants may pass to and fro. Ants which make their galleries in the bark (such as many species of Leptothorax) usually do not burrow beyond the external dead layers and occasion little or no damage, except in cases where the bark itself is of economic value: Camponotus herculeanus vagus (Scopoli) and Cremato- gaster scutellaris (Olivier) are credited with destroying the bark of cork- oaks in southern Europe and North Africa (Maceira, 1904; Emery, 1908; Seurat, 1901; A. Krausse, 1913 and 1919). » Harvesting ants have often been accused of depredations in cereal fields, but these charges are apparently much exaggerated. Emery (1891, pp. 176-177), it is true, has observed in Italy that species of Messor actively engage in carrying off grain during the harvest. It does not seem, however, that the damage thus done could be very serious, since harvester ants collect mainly seeds of weeds and wild grasses. Yet in certain regions of North Africa, where colonies of Messor are very numer- ous, the grain these ants store away may amount to an appreciable portion of the harvest. Ducellier (1912) estimates that, in Algeria, Messor barbarus collects 50 to 100 liters of wheat from each hectare. J. Pérez (1905, pp. xxxiv—xxxv) has recorded cases in which Messor bar- barus stole freshly sown carrot-seeds and also the ripe seeds of coriander in a vegetable garden near Bordeaux. Similarly, Koningsberger (1908, p. 99), in Java, blames Plagiolepis longipes (Jerdon) with stealing planted seeds of tobacco. A few species of ants are commonly found in houses, boats, and ships; they are spread by commerce to considerable distances, and rapidly become cosmopolitan. Such domestic species in the Belgian Congo include, among others, Monomorium pharaonis (Linnzus), Tetramorium simillimum (F. Smith), and especially the many forms of Pheidole megacephala (Fabricius); the last-named is the famous house ant of Madeira (O. Heer, 1852 and 1856), which has now established it- ‘Donisthorpe (1915, pp. 334-350) has given an interesting account of the exotic ants which have been ‘ateadaund fate Britain. His list includes fifty-one species, but only a small number of these have established themselves there; they are most commonly found in hothouses. _ Wheeler, Ants of the Belgian Congo 343 here in the tropics and subtropics.' In the Congo, the large rs of a form of Camponotus maculatus can also frequently be seen night in houses i in search of food. _They are particularly fond of tion, destroying a large number of phytophagous insects.’ Fore- | this respect are the driver ants (Ecitonini in America, ee hi - Richard Spruce’s account (1908, II, pp. 370-373) gives the st idea of the usefulness of their operations and it is interesting th to be quoted at length: seen carrying pupe, 1 niin hl of their own species; but they sojourn sometimes for pee days whenever they come upon suitable food and lodging. . The first time 1 saw a house invaded by Cazadoras was in Noweaber 1855, on the forest slope of Mount Campana, in the Eastern Peruvian Andes. I had taken up my abode in a solitary Indian hut, at a height of 3,000 feet, for the sake of devoting a month to the exploration of that interesting mountain. The walls of the hut were _ merely a single row of strips of palm trees, with spaces between them wide enough to admit larger animals than ants. One morning soon after sunrise the hut was suddenly - filled with large blackish ants, which ran nimbly about and tried their teeth on every- see My charqui proved too tough for them; but they made short work of a bunch In : Pit, has of late been replaced as house ant in Madeira by the Argentine ant, _ Ants more rower sy pA found in or near houses in India are, — to Asamuth (1907, p. 302), ; activity pins rt op ng noxious insects was discussed my detail by H. Stitz (1917) ina recent ef MR a Delpino (2 expressed the view ao} *the ants are the chief equilibrating and moderating factors ccting pb insects,”” perhaps a somewhat overdrawn statement. ‘Some of the Anos nw are almost exclusively termitophagous. See, for instance, the ac- counts of the habits of Megaponera fatens (p. 65), Paltothyreua tareatus (p. 62), and others. 344 Bulletin American Museum of Natural History (Vol. XLV of ripe plantains, and rooted out cockroaches, spiders, and other suchlike denizens of a forest hut. So long as they were left unmolested, they avoided the human inhabitants; but when I attempted to brush them away they fell upon me by hundreds and bit and stung fiercely. I asked the Indian’s wife if we had not better turn out awhile and leave them to their diversions. “Do they annoy you?” said she. “Why, you see it is impossible for one to work with the ants running over everything,” replied I. Where- upon she filled a calabash with cold water, and going to the corner of the hut where the ants still continued to stream in, she devoutly crossed herself, mut- tered some invocation or exorcism, and sprinkled the water gently over them. Then walking quietly round and round the hut, she continued her aspersion on the maraud- ers, and thereby literally so damped their ardour that they began to beat a retreat, and in ten minutes not an ant was to be seen. Some years afterwards I was residing in a farm-house on the river Daule, near Guayaquil, when I witnessed a similar invasion. The house was large, of two stories, and built chiefly of bamboo-cane—the walls being merely an outer and an inner layer of cane, without plaster inside or out, so that they harboured vast numbers of cockroaches, scorpions, rats, mice, bats, and even snakes, although the latter abode chiefly in the roof. Notwithstanding the size of the house, every room was specdily filled with the ants. The good lady hastened to fasten up her fresh meat, fish, sugar, etc., in safes inaccessible even to the ants; and I was prompt to impart my experience of the efficacy of baptism by water in ridding a house of such pests. “Oh,” said she laughingly, ‘we know all that; but let them first have time to clear the house of vermin; for if even a rat or a snake be caught napping, they will soon pick his bones.”” They had been in the house but a very little while when we heard a great commotion inside the walls, chiefly of mice careering madly about and uttering terri- fied squeals; and the ants were allowed to remain thus, and hunt over the house at will, for three days and nights, when, having exhausted their legitimate game, they began to be troublesome in the kitchen and on the dinner-table. ‘ Now,” said Dofia Juanita, “is the time for the water cure’; and she set her maids to sprinkle water over the visitors, who at once took the hint, gathered up their scattered squadrons, reformed in column, and resumed their march. Whenever their inquisitions became troublesome to myself during the three days, I took the liberty to scatter a few sug- gestive drops among them, and it always sufficed to make them turn aside; but any attempt at a forcible ejectment they were sure to resent with tooth and tail; and their bite and sting were rather formidable, for they were large and lusty ants. For weeks afterwards the squeaking of a mouse and the whirring of a cockroach were sounds un- heard in that house. In their general economy and behavior, the African Dorylini differ but little from the Ecitonini, as can be seen from various descriptions of their marauding columns quoted in Prof. Wheeler’s Report of the Congo ants (pp. 46-49). It may, however, be noted that their armies are apparently much more populous than those of the ecitons and also more troublesome when invading human dwellings. A rather successful method of keeping them away from inhabited places consists in making a barrage of hot ashes across their highways. = ____- Wheeler, Ants of the Belgian Congo 345 t armies of the tropics can have no doubts as to the benefit they afford t he vegetation by destroying caterpillars and other noxious insects. nee it tis evidently the general i impression that driver ants indiscrimi- After giving an impressive account of the columns of driver ants us nigricans variety molestus Gerstecker) which “seize on any prey, from a minute beetle to a cow, that is so foolhardy as to proach them,” Swynnerton describes with much detail his experi- = ts to ascertain whether any non-flying insects are safe from these Aral The unexpected conclusion was reached that these ants n a in insects left unharmed by the ants of one of the columns were iin beetles (Mylabris, Epilachna) and caterpillars (Amauris, Acrza). ii Il sciarid fly hae thoracica Macq.) had been settled 2 amongst them, but tho numbers, I might say hundreds, inspected it, often passing their antennz over it, all moved on and no attack whatso- ever was made.” The eggs and very young larve of most Rhopalocera 4 experimented with were found to be quite unacceptable to driver ants.’ ____ Swynnerton’s experiments, however, do not materially detract from P i total of the highly beneficial activity of the driver ants which, in- deed, are a blessing to all tropical cultures. As Vosseler (1905, p. 298) states, “in a given time they destroy more insect vermin than all other - insect-eating animals (birds, lizards, turtles, frogs, spiders, etc.) together, _ sinee they clean out to a certain depth the entire fieldinvaded by them.”’ _ The invasions of these Huns of the insect-world should be welcomed by all agriculturists in tropical regions, even if their pugnacious character and great numbers make them troublesome at times to human beings - and domestic animals. In Europe, foresters generally believe, apparently with good reason, that trees which attract ants or are surrounded by ant nests are less *Mesars. Lang and Chapin inform me that, according to their observations, driver ants are unable to g as tn the chogndems com of hairs and also to the manner in which these larva: can bend their body. Sreqnentiy witnessed tke unsuccessful attermpta of one 7” or even several driver ants to grasp a Dermestes 346 Bulletin American Museum of Natural History [Vol. XLV subject to the attacks of caterpillars and other noxious insects.!. The very populous colonies of certain species of Formica prove most valuable in this respect. Forel has calculated that a large colony of the European Formica rufa daily destroys at least 100,000 insects. Certain plants possess various organs, such as nectaries and myrmecodomatia, which are often utilized by the ants.» Whether these structures are intended merely to allure the ants which would thus form a body-guard to the plant, as Delpino and other botanists have believed, is a much dis- cussed problem and will be considered more in detail elsewhere. The protection afforded to the vegetation by many ants is so evi- dent that it has been employed by some of the most progressive agricul- tural people, such as the Chinese and the Malays.* In Southern China and Indo-China it is an ancient custom to place the nests of certain in- sectivorous ants in the trees; in this way orange and mandarine trees are said to be kept free from caterpillars (McCook, 1882). Such use was recorded as early as 1640, and Emery identified the ant in question as (Ecophylla smaragdina, the common silk ant or red tree-ant of the Old World tropics.* The Javanese of certain districts use ant nests, again probably those of @cophylla, to protect their mango-trees from fruit- boring weevils, Cryptorhynchus mangifere (Fabricius), and, in order to give the ants a broader field for their activities, the various trees of a plantation are connected by means of bamboos (Vorderman, 1895).* The benefit derived from the presence of the predaceous @cophylla is, however, partly offset by the fact that these ants usually keep coccids and peculiar caterpillars within their own nests, as shown by many observers (F. P. Dedd, 1902; Maxwell-Lefroy and Howlett, 1909, pp. 230-231; G. R. Dutt, 1912; Keuchenius, 1914a and 1914)). . Various attempts by agriculturists to make a more direct use of protection by ants have not thus far proved very successful. Perhaps many of these experiments have failed from lack of proper knowledge of 1Ratzebur *Popenoe fig21) has recently called attention to the use of certain unidentified ants. Yemen to combat insects noxious to date-palms. He quotes P. E, Botta {1841, ‘ Relation d'un voyage dans l'Yémen,’ (Paris), p. 155] who says he verified the fact and who credits Forsk4l with having first observed it about 1764. In Forskal’s posthumous work, however, edited by Niebuhr [1775, ‘Descrip- tiones animalium qu in itinere orientali observavit, "(Copenbague), p. 8.5], under the name Formica animosa the iobowing rather obscure statement appears: “ Welcome to eners because of the useful animosity with which it pursues the ‘Dharr’ ants perniciously infesting Phenix dactylifer. To this war it is led by heaping up ‘Heml’ (camel excrement) as its imperial reward."’ I have been unable to additional information on this subject in Niebuhr’s account of his travels with Forskl in Arabia fin pe tor 2 ager general collection of the best and most interesting voyages and travels.’ on), X, pp. 1— s *Emery, C., 1889, p. 15 of separate. Emery received his specimens from Bangkok.. Dr. C. W. Howard recently sent Prof. Wheeler ants used for similar purposes by the Chinese near Canton; they also belong to @cophylla smaragdina. ? , ‘ ‘In the Congo the silk nests of cophylla are very frequently found in fruit-trees and in coffee and rubber plantations. , 1844, III, p. 42; 1866, 1, p. 143; 1868, II, p.429. Judeich and pee ee st ___ Wheeler, Ants of the Belgian Congo 347 oats —— ant behavior. The Guatemalan kelep-ant, Ectatomma tuberculatum (Olivier , introduced some years ago into Texas for the purpose’ of exterminating the cotton boll weevil (Anthonomus grandis Boheman), _ apparently has not in any way helped control this ill-reputed pest.! Solenopsis geminata (Fabricius), the “fire-ant’”’ of the warmer regions of the world, apparently is a much more powerful enemy of the boll weevil (W. D. Hunter, 1907; W. E. Hinds, 1907). In certain parts of _ Brazil, the “formigas cuyabanas,” Prenolepis fulva Mayr,?are considered _ very effective in fighting the leaf-cutting ants (“sativas” or Attini), though there seems to be but little foundation for this belief (H. v. __ Thering, 1905 and 1917; A. da Costa Lima, 1916). F. v. Faber (1909) es claims that in Java ‘‘a black ant, 3to4 mm. long,” but not otherwise identified, successfully controls the bugs of the genus Helopeltis in cacao __ plantations. Perhaps this is Dolichoderus bituberculatus Mayr, an ant oh which, according to de Lange (1910) and Moorstatt (1912), is used in Java to combat these same Helopeltis of cacao. According to Rothney (1889, p. 355), two ants, Monomorium ditions (Linnzus) and Solenopsis geminata (Fabricius), are deliber- _ately introduced into warehouses in Madras to check the depredations of white ants. “This practice is not uncommon in Northern India and _ the natives of India are familiar with the kind of ant which should be brought in” (Maxwell-Lefroy and Howlett, 1909, p. 226). _ Another service of ants which should not be overlooked by ecolo- ‘gists is their ceaseless activity in excavating, transporting soil particles, and hastening the decay of organic substances. Their multiple burrows, extending in all directions underground, bring about a very thorough ventilation of the soil and an easy and even distribution of moisture. They comminute and bring to the surface a large quantity of soil and subsoil, often from a considerable depth, and leave it exposed to the weathering action of the meteoric agents. Furthermore, they introduce into their subterranean excavations much organic matter which thus more readily decays and in turn yields acids that act upon the soil. Owing to the hidden habits and minute size of most ants, their importance as geologic agents may be easily lost sight, of, especially in temperate regions. In tropical and subtropical countries the result of their toil is often much more apparent, though it rarely approximates oO and 1800) by O. F. Cook (1904, 1905, and 1906) and their criticism by Wm. M. Wheeler A list of ants known.to prey on the cotton boll weevil is given by W. Pierce "fo mi aa cear enses’ or “for migas ragua aa’ * var ious ot he rT spec ie . of ants are pa uv ie aN og Shy or “ cuyabanas.” Oe 348 Bulletin American Museum of Natural History [Vol. XLV that produced by termites. Only certain species of Formica in temperate Europe and North America construct mound or hill nests of sufficient size or number to attract much attention; with them, the accumulations consist of a small part of excavated soil, most of the material being gathered in the vicinity by the workers. The conical mounds of the North American Formica exsectoides sometimes reach a meter in height and two to three meters in diameter at the base, while those of the European F. rufa often are much larger (over two meters high and eight to ten meters in diameter). The crater-shaped or conical mounds of certain North Avibebeds harvesting ants are partly made of earth brought from underground excavations. Those of Pogonomyrmex barbatus subspecies molefagiens may attain one to two meters in diameter and fifty centimeters in height, while in the common P. occidentalis they are but little smaller and often form extensive colonies (Headlee and Dean, 1908; Wheeler, 1910). Ischnomyrmex cockerelli, of the southwestern United States, surrounds _ the entrance to its nests with huge craters, from sixty centimeters to two meters in diameter and from 0.2 to 0.5 centimeters in height, built of coarse desert soil intermingled with pebbles sometimes two contametens in diameter (Wheeler, 1910, p. 281). The volume of material moved by some of the leaf-cutting ants (Attini) of tropical America is much greater than in any of the cases mentioned above. H. v. Ihering (1882), Gounelle (1896), and Branner (1896, 1900, 1910, 1912) have called attention to the importance of these insects as geologic factors. In certain parts of Brazil the ant hills of the sativa (Alta species, probably cephalotes) are so large and numerous that they become a remarkable feature of the landscape. At one place in the Rio Utinga region, in the interior of Bahia, where the forest had been cleared away so that the mounds were visible, Branner counted fifty- three of them within an area of 10,000 square meters. Their bases cov- ered close to one-fifth of the total space under consideration and their volume was estimated at 2225 cubie meters. The cubical contents of the mounds, if evenly distributed over the entire 10,000 square meters, would _ have been 22.25 centimeters thick. In this case, the height of the ant hills varied from 1.2 to 4.5 meters, with an average of 2.5 meters. These were not the largest seen, for on the upper drainage of the Rio Utinga, Branner measured mounds of leaf-cutters five meters high and sixteen or seventeen meters in diameter at the base, each containing about 340 cubic meters of earth. The illustrations in Branner’s latest papers (1910, 1912) remind one of strikingly similar landscapes with scattered termite Wheeler, Ants of the Belgian Congo 349 is ix many parts of tropical Africa (see Pl. XV). A considerable it of living vegetable matter is carried by the leaf-cutting ants mner chambers of their nests, where it is cut up and worked in room-beds; vegetable substance is thus rapidly transformed al matter iid Hindered. available to new plant-growth. mound- or hill-building ants are not found in tropical Africa; pecies, however, build small crater-shaped accumulations of the entrance to their nest. Those of the seed-storing Messor a very conspicuous in the arid parts of the continent; their 's sometimes measure a meter or more across and the earthen walls ch twenty-five centimeters in height (Passarge, 1904, pp. 290- . also the photograph of a nest of Messor species taken by Mr. on the Athi Plains, British East Africa, Pl. XX VI, fig. 1). its, when establishing their temporary abodes, often excavate able quantities of soil, as is shown by Mr. H. Lang’s photograph sting site of Dorylus (Anomma) wilverthi Emery (PI. II). following chapters deal with many other activities by which 1e into direct contact with plants. They will further emphasize ortance of ants in the economy of nature, in which they must itedly be regarded as the dominant insects (Wheeler). From the r point of view of human interests, by far the greatest number of ei , re indifferent or negligible organisms, either because of their small ct pa “a consideration of all the facts forces us to mit, with Forel, that as a group ants are agg! beneficial and that Bee receoductives! functions to one or a few queens s that are able to resist | ape mn = in the inaccessible penetralia of their nests” Ants as Agents in the Pollination of Flowers Ih Knuth’s celebrated ‘Handbook of ete Pollination” ants are i nee 350 Bulletin American Museum of Natural History [Vol. XLV Perhaps even in temperate zones this is not entirely true, and it is difficult to believe that, in Umbellifere and other flower associations with freely exposed nectar on which ants are most commonly met with, these insects are not at least effective agents of geitonogamy. In the tropics, more- over, ants are so abundant everywhere that very likely they are of even greater importance as carriers of pollen, the more so since many trees and shrubs of tropical forests bear flowers on their old wood on the very highways of the ants, so to speak. One might even venture to suppose that cauliflory is mainly of use to the plant in that the flowers are thus placed within easy reach of pollinating ants. Indeed, the question as to the origin and significance of cauliflory in tropical trees and shrubs has not thus far been satisfactorily answered. Wallace’ regards it as an adaptation to pollination by butterflies, which, he says, keep to the - undergrowth of the forest and rarely ascend to the crown of the trees. Haberlandt (1893, p. 132) argues that many of the caulinary flowers are dull colored and also otherwise but little adapted to Lepidoptera, and, from my personal experience in the Ituri forest, I must agree with him. I cannot recall a single instance in which I saw caulinary flowers visited by butterflies and I greatly doubt whether Wallace’s explanation was founded on actual observation. In Haberlandt’s opinion, cauliflory is merely the result of a tendency to a more complete division of labor, resulting in a sharper differentiation between the assimilating and the reproductive parts of the plant. Evidently A. F. W. Schimper (1903, p. 338) is also satisfied with a mere physiological solution when he supposes that the frequent occurrence of cauliflory among tropical trees is due to a weaker development or slighter degree of roughness of the bark. The foregoing remarks will suffice to show that the relations between ants and cauliflorous plants are worthy of further attention. In his biological studies of tropical flowers, H. Winkler (1906) enumerates a number of plants in Cameroon which he asserts are pollinated by ants, though he does not enter into details nor describe any adaptations of the flowers to this peculiar mode of fecundation. It is interesting to note that most of the species thus mentioned by Winkler are cauliflorous trees or shrubs. The cacao tree (Theobroma Cacao) affords a classical illustration of cauliflory, its flowers being borne on both stem and main branches; in this case G. A. Jones (1912), from his experiments carried on in Dominica, West Indies, has reached the conclusion that ants are in all probability the chief agents of pollination. 11891, ‘Natural selection and tropical nature,’ (London), p. 244. 7 ¥ (Wheeler, Ants of the Belgian Congo 351 i. N. Ridley (1910, pp. 461-462) has made some sinteweetifig ob- ‘e ns in Singapore on certain species of the anonaceous genus tothe s, notably G. Ridleyi King, which produce their flowers in vasses at the base of the tree.! ; ‘The flowers are of large size and dull reddish in color. They are almost invariably ed by a nest of very small black ants, which pile up powdery soil all over them, it they are often quite concealed. It would, I think, be difficult for a bee or er insect to get to the honey of these flowers through the nest, yet I think no species ye genus fruits so regularly or heavily as does Goniothalamus Ridleyi. That the are distinctly attracted by the flowers, is clear from the fact that the flowers m the trunk which are too high up for the ants to cover with the nest are generally isely covered by a swarm of the insects. Owing, however, to the minuteness of the t and the difficulty of making observations in such a mass of them, I have been able to definitely decide whether the ants do actually fertilize the flowers by con- ring the pollen from one to the other, but I can not see any other way in which the fertilization can be effected. The ants generally throw up the mounds over the flowers e the buds open, as if in anticipation of the honey within the flowers. In most of the genus the flowers are borne on the branches or upper part of the stem, and are brighter in color, white or orange, and these are not haunted by ants, but doubtless fertilized by hymenopterous or dipterous insects. If the flowers of G. Ridleyi are, as I believe, fertilized by ants, their position at the base of the stem may _ be taken as a modification to that end. This, however, could not be classed as symbiosis, but rather as a modification for fertilization, as the main nest of the ants is apparently always underground near the tree. Ants and Extrafloral Nectaries ie Mader the term “extrafloral nectaries” botanists include all glands - secreting saccharine substances located on the vegetative organs of _ plants, while the “floral nectaries” are similar nectar-secreting glands _ found on parts of the flower or of the inflorescence.? There is still con- siderable discussion as to the true significance of nectaries. In this con- nection it is rather interesting to observe that all earlier botanists re- garded even the floral nectaries as having a physiological function. Some believed that the saccharine secretion accumulated in the flowers q served to feed the embryo; others considered the nectaries as excretory _ organs, eliminating waste substances of no further use or perhaps even - noxious to the plant. In later years the majority of naturalists have | tecepted none but an ecological explanation. That the nectar panes of eee ag ond 8. Evans (978) has described cross-pollination by means of ants in an unnamed rubiaceous is 2 Dsipino (1874 aad 18750) o) praggeed to replace these terms with ‘ ‘extranuptial” and “nuptial” nectaries respectively. A “nectary"’ wa» opgsna'ly {one by Linnaus (1751, *Philosophia botanica,’ a that wa da th ye the ‘Sonu which produces the mente *nectarium, pare parse tat flori propria.” 7 mesma of the word to apply to all — of the plant producing sweet excre- tions. Tesnae tte (1848, ‘ De nevtariis.’ Eberfeld) apparent! t made the distinction between floral and _ extrafloral nectaries. The historical side of the question has been fully treated by G. Bonnier (1879), 352 © Bulletin American Museum of Natural History [Vol. XLV flowers attract pollinating insects, which in turn assure or greatly facili- tate cross-fertilization, is too well established a fact to be doubted. It is, however, by no means certain that these floral nectaries are not at the same time more directly useful to the plant in a physiological way." Ants are frequently seen busily visiting the extrafloral nectaries of certain plants. They are, for instance, seldom absent from the large stipular glands of certain species of Vicia (V. sepium, V. sativa, and V. Faba) in Europe (see Rathy, 1882, pp. 29-36; Hetschko, 1908). In North America the stipules of some species of Cassia are especially at- tractive to these insects. In the Belgian Congo, I have taken numbers of ants, together with many other Hymenoptera and Diptera, as they were sucking up the sweetish fluid secreted at the base of the leaf-blade of Urena lobata variety reticulata Guerke, a very common weed in native villages and cultures.? The foliar nectaries of several Javanese species of Hibiscus are also very inviting to ants (Keernicke, 1918). It is on similar observations that Delpino (1874, 1875, 1879), A. F. W. Schimper (1888), and Kerner von Marilaun (1876) based their ecological inter- pretation of extrafloral nectar glands. The following passage from Del- pino’s earliest paper (1874, pp. 237-238) may be reproduced in full, as it sums up his views: What then is the function of the extranuptial nectaries, which are found on the caulinary leaves, on the bracts, and on the calyx? Though I reserve for another paper the publication of my studies of such and other extradichogamie relations between plants and insects, I do not hesitate to announce now that the chief function of these nectaries is to place the ants, wasps, and Polistes in the position of sentries and guards, to prevent the tender parts of the plant from being destroyed by larvae. Where ants and wasps are present, larvae cannot exist because they will be devoured. Thus certain plants have adopted the same means of defense and bait that we see used by the tribe of aphids, coccids, Tettigometra, and other cicadellids, which spon- taneously place themselves under the powerful protection of ants. Still another fune- tion, though a subordinate one, can sometimes be carried on by the above-mentioned nectaries. namely that of keeping the ants from the nuptial nectaries by detaining them at the extranuptial nectar glands. Indeed we can ascertain the noxious effects of ants when they succeed in infesting the flowers. In the first place, ants have seden- tary habits, remaining motionless for whole hours on the same flower: therefore, they are of no use in dichogamy. Secondly, ants are objects of fright and aversion to the natural pollinating insects of the plant, as for instance, flies, butterflies, and bees; hence, their presence on the flowers renders useless the dichogamic devices of these plants. I have repeatedly observed bees and bumble-bees avoid visiting flowers when they saw ants there. Which all makes it clear how plants under given circum- 1G. Bonnier (1879, p. 200) after 9 ox! a critical study of the subject, from an anatomical and a pen cal viewpoint, concludes: ‘* The nectariferous tissues, whether floral or extrafloral, whether or ducing a liquid externally, represent special food reserves directly connected with the life of hook plant.” ed *See my notes on this plant in Rev. Zool. Afr., III, fase. 1, 1913, p. 3. _ Wheeler, Ants of the Belgian Congo 353 r } may find great profit in producing extranuptial ‘nectaries, either to secure omy ee ne ere om ta lore the ants ener ai not been accepted by many other natuwaliota: probably because aot extend the concept of myrmecophytism to include a very con- shen of the world’s flora? a has a clever explanation of how the involucral nectar glands certain Composite attract ants which defend the capitula against uel He claims that ants climbing the plant thus find on their “up an ample and readily accessible es of honey, and conse- rme will not readily admit Kerner’s supposition as : to the __ limitation of the ants’ feeding propensities. As a matter of fact, these __ insects are sometimes found inside flowers of various types, and fre- Pe _ quently so on those with freely exposed nectar, such as the Umbelli- ___ ferme. In tropical regions at least, as I have suggested above, they 3 should not be wholly disregarded as pollen carriers. Secondly, observa- ___ tion shows that the extraforal nectaries, while present in a great number ___ of species, are in many of them seldom if ever visited by ants. Thirdly, the visitors of extrafloral nectar glands especially attractive to insects = Frequently do not consist of ants only, but include various other Hymenop- tera, Diptera, and Coleoptera, which are by no means deterred by the ig And lastly, it has not been sufficiently well established that the ‘a 10. F. F. Cook's papers on “kelep” ant sn According to this of the le to which “ myrme- . " may be cmed spent naturalists. According to author (1 ¢, p. 666) the cotton- of eastern Guatemala has, through its extensive system of extrafloral nectaries, secured the active oh Oe helen os veer eating ont, t, Ectatomma roe sca ar (Olivier), against the boll weevil! 385) and at the invol bracts of certain Som jp nen have been Fe further investigated by We ge pes 1858) and _Rateee G97). _ The last-named observer found that the sweet excretin ee are rained not 7 by anta (Myrmica levinodis, Hymenoptera (A ), Diptera, and Coleoptera. *Hetachko (1908) gives a list of the visitors he ueeves s sks stipulay nectaries of Vicia sativa Rinniwes. It includes, in addition to four ants Prermien sius niger, and Myrmica levinodis), 24 species of 6 Ameo. $F 2 Sphegoiden 10 Ichneumonidw, and 2 ),21 of Diptera & Spry, Muscide, and i Bibionidw), § of Coleoptera (3 Canthari- dm, 2 Elaterids, 1 Coccinellide), and 1 of Hemiptera. 354 Bulletin American Museum of Natural History [Vol. XLV presence of the “body-guard” of ants actually favors the species or individual plant on which they are found, though it cannot be denied fhat, when present in large numbers, they give to the plant a certain amount of protection.' The so-called “food-bodies” of the myrmecophytes Acacia sphzro- cephala (Beltian bodies) and Cecropia adenopus (Miillerian bodies) are probably also of glandular origin (F. Darwin, 1876); they are described in my synopsis of the myrmecophytes (p. 503). Such structures are by no means restricted to certain typical ant-plants. Leea hirsuta Blume, a common Javanese bush of the family Ampelidacez, produces spherical excrescences on the tender parts of the plant, in abundance on the young petioles, also on the young leaf-blades near the midrib and on the stem of young shoots. These glandular bodies, about 0.7 mm. long, consist of an outer layer of small cells enclosing much larger cells filled with oil drops and albuminoid granules. They are eagerly collected by ants and consequently often difficult to find on the plant (Raciborski, 1898). A similar case is that of Plerospermum javanicum Junghuhn, one of the Sterculiacew in Java, which bears in its funnel-shaped stipules minute food-bodies also collected and carried away by ants (Raciborski, 1900). In both these cases other species of the same genus lack these food-bodies completely. Since neither Leea hirsuta nor Pterospermum javanicum possesses myrmecodomatia, they could not well be regarded as true myrmecophytes, no more than the many plants which are merely provided with extrafloral nectaries. Their case offers a suggestive comparison with the Miillerian and Beltian bodies and weakens the argument that the last-named growths are myrmecophilous organs con- ~ nected with the presence of ant-dwellings in Cecropia adenopus and Acacia spherocephala. It thus seems that, from the point of view of the myrmecologist, extrafloral nectaries and ‘“food-bodies”’ are little more than additional sources of food which ants are so keen in detecting and in exploiting to the very limit. All ants are fond of sweets and this is especially notice- able in species with a vegetarian or semi-vegetarian diet. In many cases” the sugary juices are absorbed so eagerly by the workers that their crop distends considerably and the gaster is temporarily inflated to a size entirely out of proportion to the rest of the body. Extreme instances of the kind are the so-called ‘honey ants”’ of the arid plains and deserts of North America, South Africa, and Australia. In certain ants of these 1A comprehensive criticism of De iecte® ~ sony of extrafloral nectaries has been given by Mrs. M. Nieuwenhuis von U xkGll-Géldenbran t (1907) __-Wheeler, Ants of the Belgian Cong> 355 ns some individuals of the worker caste have developed into a special “repletes,” which act as living reservoirs of liquid food for the : n s of different plant organs and even of certain galls. A few ago Wheeler published a complete account of the honey ants inc lamine of the meneats appear like Sue: The anterior * n of the first segment has a hollow depression into which the | van be fits. With the aid of a lens it is possible to distinguish, these Sten i is, according to the same post slay as fully dis- ded as that of the North American Myrmecocystus melliger, and locomotion must be almost impossible for this insect. ____ The habit of using some of the members of the colony as honey pots will probably be discovered in certain other ants of the African deserts. Among other species it may be still in an incipient stage, as, for instance, in the case of Acantholepis arnoldi Forel in Southern Rhodesia. The __ nests of this ant are found in loose, sandy soil in the hottest places. at They sometimes contain workers with gaster considerably swollen, as long as the head and thorax together, but not so rotund as in the re- - pee ot Myrmecocystus or Plagiolepis trimenii (Arnold, 1920, p. 564). ee Dispersal of Seeds by Ants > = That certain ants gather seeds and preserve them in special granaries ____ in their nests has been known since very ancient times. There are frequent allusions to harvesting ants, and even more or less accurate accounts of their activities, in the writings which have come down to us _ from the older civilizations along the shores of the Mediterranean.* _____ Yet such keen myrmecologists of western Europe as Latreille and P. 1Plagiolepia decolor Emery, a very closely allied South African species, is, according to Forel, also a old accounts are given in the works of Moggridge (1873, pp. 5-11) and MeCook (1879a, pp. ae 42460). ~ > 356 Bulletin American Museum of Natural History (Vol. XLV Huber, unacquainted with the spectacular seed-storing habits of certain southern ants, discredited the assertions of the ancient writers. Though Sykes (1835) and Jerdon (1851) in India and Buckley (1861a) and Lincecum (1862) in North America had actually observed certain ants collecting large quantities of seeds, it needed the careful investiga- tions of Moggridge (1873) in southern France and of MeCook (1877 and 1879a) in Texas to dispel the skepticism of modern entomologists. It is only more recently, however, that naturalists have come to appreciate the general importance of ants as seed distributors. Their réle in this respect seems to have been first realized by Kerner von Marilaun (1895, pp. 866-867). Later F. Ludwig (1899, p. 38) definitely asserted that ‘‘ants do not only aid in scattering plant seeds, but that they play a prominent part in the dispersal of the indigenous (European) vegetation.”” In Sernander’s comprehensive ‘Monograph of European Myrmecochores’ (1906b) one finds a detailed and critical history of the subject, together with an immense array of new and interesting observa- tions. His conclusions show that in Europe a great many grasses and herbaceous plants rely almost exclusively, or at least to a large extent, on certain species of ants for the successful scattering of their seeds. Many of the more common ants, belonging to such ubiquitous genera as Formica, Lasius, Tetramorium, and Myrmica, gather seeds of various plants more or less consistently. To the phytecologist these widely distributed ants are perhaps factors of greater importance than the true harvesters. The latter, to be sure, are more spectacular in their per- formances, but they are restricted to certain desert or semi-arid regions and are evidently extreme cases, remarkable for the huge quantities of seeds stored in their granaries. The ecological significance of seed-transporting ants can only be adequately, realized upon closer scrutiny of the actual results of their activity in thisline. Sernander’s calculations, though based on moderate figures, show that the amount of seeds carried about by ants must be considerable. He found, for instance, that a single colony of Formica rufa transports during one season about 37,000 seeds and fruits. Obser- vation also discloses that the seeds are in this way conveyed appreciable distances (100 to 200 feet) from the mother-plant. On their foraging excursions ants frequently drop or lose seeds along the road. Further- more, many of the seeds finally stored in the recesses of the nest are sooner or later cast out near the entrance along with chaff and other débris from the ants’ household, and a number of them are still able to germinate. Finally, with further investigation, the number of myrme- Wheeler, Ants of the Belgian Congo 357 ochor : or species of plants whose seeds are garnered by ants, increases 3 4! ne might reasonably surmise that in tropical countries too ants will inc Ye to be efficient agents in the dispersal of the seeds and fruits of sp ; but, as yet, this side of tropical ant behavior has been iigesthed upon. O. Kuntze (1877, p. 24) mentions incidentally at in South America he saw ants carry off the seeds of papaw-trees (Carica Papaya Linneus). R. H. Lock (1904) gives a short account of jp diopereal i in Ceylon of Turnera ulmifolia Linnzus by ants (Pheidole pathife Forel) which are apparently attracted by the arillus of the ed. More recently, W. and J. Docters van Leeuwen-Reynvaan (1912) e carefully investigated the scattering of the seeds of Dischidia esiana Wallich and D. nummularia R. Brown, which are common es in Java. The pappiferous seeds of these Asclepiadacee beara r, white caruncle of thin-walled cells filed with fatty and albumi- eh i oO , ¥7 8 rminate when sufficiently eilitioed. but such seedlings do not develop — inte adult plants. Plenty of healthy seedlings can, however, be found s ir n the galleries of Iridomyrmex myrmecodiz Emery, an ant which builds s nest on and in the bark of trees. Moreover, this ant has been seen it the act of transporting Dischidia seeds, to which it was probably d by the caruncles. These minute ants, being unable to grasp oe seed itself, pull off the longer, fragile hairs of the pappus and by a of the shorter, stronger hairs, drag the seed into a slit in the bark or = 5 the roots and stalks of other Dischidiz. It may be noted that in Z = the pitcher-shaped leaves of these species of Dischidia? are usually _ inhabited by the same Iridomyrmez, so that this is perhaps one of the clearest examples of true symbiosis between ants and plants. It would be important to investigate further whether the ants actually feed on the E page the seeds. The case of Dischidia also suggests comparison _ with the “ant gardens” of the Amazon, which are considered in more _ detail elsewhere (p. 365). Ul (1900, p. 123) records finding the pea-sized seeds of Ipomea q ore Linnzus lying in long rows on the sandy sea-shore at Copaca- writes: it one would draw un barbarus in Arbe, an island in the Adriatic Sea, F. Neg ete emerng "If one would draw up a list of all the plants whose seeds or { fruits are carried 2 oe. | this list would almost be equivalent to an enumeration of the flowering me A oF (1810, pp ; 462-465) concludes that D. Raflesiartia cannot be regarded as a true myrme- ee the seeds of this plant escaped his notice, so that the foal iees oll fentios coed 358 Bulletin American Museum of Natural History [Vol. XLV bana, near Rio de Janeiro; he saw leaf-cutting ants (Attini) moving along, each carrying one of the seedsintoa hole. It would thus seem that the Attini also store seeds in their nests or perhaps use them in their fungus gardens. H. Winkler’s (1906, pp. 236-237) statements concerning the dis- persal of seeds by ants in Cameroon do not enter into much detail and merely show that the réle played by ants in this respect in tropical Africa should not be disregarded. He says that in the dispersal of “num- erous dry fruits with small seeds, ants are undoubtedly also ot significance, since no spot in the tropical Rain Forest is free from these insects. I have almost always found that the arilli on dropped seeds of Blighia and other Sapindacew had been eaten away by ants. I have, however, never seen flower-gardens (due to ants) in Cameroon.” Harvesting Ants The reader will find a complete review of this fascinating subject in the chapter devoted to harvesting ants in Prof. Wheeler’s ant-book (1910b, pp. 267-293). The following account, therefore, will deal with what little is known at present of the seed-storing ants in the Ethiopian Region.! The typical Old World harvesters of the genus Messor are at home in the desert and semi-arid parts of the southern Palearctic, of the Ethiopian, and of the Indian Regions (Map 45). It is noteworthy that in Africa these ants, though widely distributed over the dry parts of the continent, avoid the moist West African Region (Engler’s Western Forest Province), where seed-storing on a large scale is rendered practi- cally impossible by the great moisture which prevails throughout the year, or at least for long periods, and would soon cause the stored seeds to sprout. Though Messor occurs as far north as Mossamedes and Bulawayo, as far west as the Great Rift Valley, and has recently been taken at Fort Crampel, French Congo, it has not been recorded from any- where within the Congo Basin; yet it is not impossible that some of its forms might be found in Katanga. In East Africa this genus has the same general habits as in the Mediterranean Region (see Moggridge, 1873), as far as can be gathered from Sjéstedt’s account of Messor cephalotes Emery, observed by him at the northern foot of Mt. Kiliman- jaro: ‘Additional observations on the harvestiog ant Messor barharus subspecies meridionalie (Ern. André), in Macedonia, have recently been published by F. D. Doflein (1920). ; 1922) Wheeler, Ants of the Belgian Congo 359 ,_ At Feuveral ‘eaotn one could see cleared spaces amidst the dry grass where every - had been removed and the red-brown soil lay open to view, plane and clean as a well-attended garden plot. Such places were somewhat variable in size, mostly _ up to about 6 paces across and nearly circular. Heaps of fine grass-stalks cut to pieces (one to several liters; often 2 to 3 em. long), together with grass panicles, were lying around. Scattered ants were wandering all over the place, the soldiers being especially striking on account of their big heads. The cleared place showed a large entrance, often more than finger-wide, into which the ants were dragging the stalk cuttings; ‘more in particular I saw soldiers disappear with such cuttings through tHe entrance. a The largest space I saw was 8 paces in diameter, with 4 or 5 separate entrances, one _ of which was larger than the others (as is the rule when there are many) and sur- __ rounded by an irregular, funnel-shaped depression, 15 to 20 cm. deep. The heaps ____ of stalks, elsewhere clean and free from earth, were in this case mixed with soil and did not look as clean as usual. There were also holes in places along the path, into which the ants were dragging grass-stalks cut to pieces.” (Mayr, 1907, pp. 14-15.) si s ° iS 30 45 Ca % 90 105 120 3s 160 3 . [>> ect ie BG A : TRI Pa ; as - — ns 30 = - My Fi Ss ° bi > t) y m = ay s f A rs | 'S AN | Te i Sh be ~ Ms 4 a} 30 0 Ss ° 1s » 45 0 5 20 105 120 BS ig Map 45. Distribution of Messor, a genus of harvesting ants. A subspecies of M. barbarus has recently been described from Fort Archambault, at about 9° 5’ N., 18° 35’ E. Figure 1 on Plate XXVI represents one of these nesting sites of East African Messor from a photograph taken by Mr. H. Lang in the Athi Plains, British East Africa, during the R. Tjader Expedition (July 1906). It is interesting to learn from Sjéstedt’s experience with Messor cephalotes in East Africa and that of Neger (1910a) with M. barbarus in southern Europe that certain species of the genus Messor are leaf-cutting. What use these ants may have for the plant cuttings in their nests is as yet unknown. 360 Bulletin American Museum of Natural History (Vol. XLV According to K. Escherich (191la, pp. 48-51) Messor barbarus in Eritrea stays within its nests in the daytime, coming out in numerous columns after sunset to cut off panicles of grass and collect seeds, which are taken home; often pellets of earth or little stones are carried away by mistake. A few workers were also dragging bulblets of a Cyperus, probably C. bulbosus, which are in this way effectively scattered.' The genus Pheidole, abundantly distributed over all tropical and warm temperate regions of the world, is so closely allied to Messor that both have been included by Emery in one tribe, the Pheidolini, which also contains many other harvesting genera (Oxyopomyrmez, Goniomma, Novomessor, Veromessor, etc.); thus the seed-storing he- | havior is to some extent rooted into the phylogeny of the group. Certain Indian Pheidole, such as Pheidole providens (Sykes), are famous as harves- ters, and many other members of the genus are also more or less granivo-— rous. Mr. H. Lang and I discovered a typical seed-storing species, Pheidole saxicola Wheeler, in the Lower Congo. At Zambi the nests of this ant were placed in the interstices of stones on a rocky hill as shown on Plate VII and described by Mr. Lang in his field-notes (p. 139). From the débris, heaps of chaff, and rejected seeds thrown out by the ants and accumulated near the entrances of the nest, it was seen that the seeds gathered by this Pheidole belong chiefly to a few common grasses, such as Chloris polydactyla Swartz and various species of Andro- pogon. Concerning the genus Pheidole, Arnold (1920, p. 416) remarks that the South African species “are omnivorous, with a marked fondness for sugary substances, but some species, e. g. excellens, crassinoda and arnoldi, are mainly graminivorous, harvesting the seeds of grass in the same way as the species of Messor.”’ P. xocensis Forel and its variety bulawayensis Forel are also mentioned by Arnold (1920, p. 445) as being ‘‘at least partly graminivorous, as the nests contained accumulations of grass seeds.” According to Arnold’s observations, the commonest harvesters in South Africa are various forms of the genus Messor and certain species of Tetramorium. In his ‘Monograph of the Formicide of South Africa’ (1920, pp. 409-410), he writes of Messor capensis subspecies pseudo- zgyptiacus (Emery) as follows: “This variety is very common in the neighborhood of Bulawayo. It is eminently a harvesting ant, usually collecting the seeds of one particular kind of grass. The rejected husks of these seeds are deposited in a circle all around the entrance of the nest, 1The bulblets of Cyperus bulbosus Vahl and C. esculentus Linnwus are also occasionally eaten by certain African natives. Wheeler, Ants of the Belgian Congo 361 half of the circle being generally deeper than the other, which may be ay the prevailing winds. These rubbish heaps when made by a yulous colony sometimes reach very large dimensions, covering as ic! h as one square foot of ground, and from one to three inches deep. ‘ site of such a nest is very plainly indicated by these accumulations, ice the husks are bleached almost white by the action of the sun. e nests of this ant appear to be very free of myrmecophilous insects d even the ubiquitous thysanuran is rarely to be found in them. The ants appear to have definite foraging grounds, to which access is obtained y well-marked and smooth paths leading from the nest in various direc- ns.” Tetramorium setuliferum Emery he describes (1917, p. 291) as ‘a harvesting and graminivorous species. The entrances to the nests are often surrounded by small accumulations of husks of a grass seed. . These heaps are smaller than those of Messor, and much less tidily dis- a _ The ponerine ants are well known for their predaceous habits and Fag a carnivorous diet. Yet one at least of these ants, the common : 3 _ African Euponera sennaarensis (Mayr), is to a large extent granivorous. ama (1913, p. 13; 1915, p. 7) found that the nest of this species in Rhodesia “often contains considerable accumulations of grass seed which may be used as food,” though this ant is also a keen hunter of termites. Similar observations have been made on this species by K. _ Escherich in Abyssinia (Forel, 1910, p. 245) and by myself in Katanga (Bequaert, 1913, p. 421). ‘= _ There is little doubt that certain ants derive at least part of their ___ 8ustenance from the seeds which they carry into their nests. Yet it is ____by no means clear how they manage to utilize the various amylaceous, nitrogenous, and oily substances contained in the seeds, either for their Own nourishment or as food for their brood. In the case of the many widespread species which use seeds only in small quantities, as an addi- tional food supply, it would seem that the caruncle alone is bitten off, _ __ neither the coats of the seeds nor their contents being touched. This is, however, not the case with true harvesters, some of which have become almost purely granivorous and, as a rule, remove the entire kernel of the seed. In his experiments with a colony of Messor structor kept in an artificial nest, Emery (1899 and 1912b) found that this ant would more or less readily accept cooked or dried meat, various fresh mushrooms, husked rice, a variety of ripe and unripe seeds, plant buds, bread, and dry vermicelli. These substances would all be to a certain extent triturated 362 Bulletin American Museum of Natural History [Vol. XLV between the mandibles, and finally a large or small quantity of residue would be dumped out of the formicary; but the ants steadily hares raw starch. Emery also made some feeding experiments with a colony of Misi barbarus minor kept in an artificial nest of the Janet pattern. He found that this ant is less omnivorous than M. structor. It shows a predilection for dead insects; seeds rank only second in its choice, though they often constitute its principal food. When a ripe, dry, and unsprouted grain of wheat is offered to this species, the ants carry it into their nest and sooner or later gnaw off the embryo, always beginning to eat the grain at that end. This curious habit was even known to the ancient writers (Plutarch and others) who consequently attributed to the harvester ants a most wonderful instinct of preventing the sprouting of the grain by removing the germ. Emery, however, has shown experimentally that this is due merely to a matter of taste or gluttony manifested by the ants for this daintiest part of the grain. He believes that the ants mutilate the radicle of sprouted seeds for a similar reason, though he admits that this behavior may be of a more complicated nature. Harvester ants can thrive perfectly on unsprouted grain, as shown by Emery’s experiments, but in most cases they allow a partial germina- tion of the seeds before using them as.food. Neger (1910a) found that most of the seeds which Messor barbarus places in the sun near the en- trance to its nest are already partly sprouted; these sprouted seeds are carefully removed from their envelopes and are only carried back into the formicary when thoroughly dry; under such conditions the germ plants are evidently killed. It has been supposed (Moggridge) that the ants allow the seeds to germinate in their nests so the starch will be con- verted into grape sugar, the whole procedure being somewhat compar- able to the malting of grain. Neger, however, discards this explanation because he found that in the sprouted seeds which are placed to dry in the sun the process of germination was not sufficiently advanced to convert any large quantity of starch. He believes, therefore, that the ‘practice of allowing them to sprout has no further purpose than te facili- tate the removal of the coatings, which are sometimes very hard to detach from ripe seeds; on sprouted seeds, these envelopes split open and are then easily pealed off by the ants. Neger has also investigated what happens to the germinated seeds after they have been taken back into the formicaries. He found that, at certain hours of the day, the ants carry out of the nests small, shapeless, pasty masses of a brownish-pink color, which are lett to dry in the sun. . _ Wheeler, Ants of the Belgian Congo 363 ——, n carried out these masses are soft, damp, and bitter to the taste; Size varies from that of the head of a pin to that of a grain of er. Microscopic examination shows that they consist of com- ted parts of seeds, plant hairs, fibres, pollen, ete.' Neger calls these y masses “‘ant-bread-crumbs”’ and, although he never saw them being 1s , he supposes that they are eventually carried back into the st by the ants. In a number of these crumbs he found spores and celium of a mould which he identifies with Aspergillus niger, having obtained this fungus in a number ot cultures made with fresh “ant- “a * taken from worker ants. He formulates the hypothesis ut the amylolytic and proteolytic action of this mould may help to Tender: the crumbs more readily digestible so that they can be fed to the young as “larva-bread.” ___ Emery (1912) completely rejects Neger’s supposition that the e and aleurone of the seeds need to be prepared by a ferment before : being fed to the larve. He offered his colony of Messor barbarus minor wheat-paste made up in the form of small rings and found that this sub- - stance was readily accepted by the workers, who carried it into the = r part of the nest. There the rings were malaxated for some time . and divided into small, twisted pieces, more or less irregular in shape, ae which were finally dumped into the drier chamber of the nest and never a touched again by the ants. Fragments of this paste were also presented by the workers to the larve, the largest of which applied their mouth- parts to it just as to other food. Emery determined the weight and starch contents of fifty of these paste rings before anc after malaxation by the ants. He infers from his figures that the workers either digested or fed to their larve at least 7.3 per cent of the starch and that they consumed 2 also an unknown quantity of nonamylaceous substances, probably pro- _ teids; the latter he regards asa much more important aliment than the a This brief consideration of the feeding habits of harvester ants may _ be properly concluded with Emery’s remarks concerning the ethological _ significance of granivorous behavior among the Formicide: ; The granivorous ants are derived from insectivorous ants. They represent an adaptation to the climatic conditions of dry prairies, steppes and deserts. When, owing to the summer droughts, insects become scarce and are no longer sufficiently numerous to satisfy the needs of the ants, the granivorous species substitute the living but dried seeds of plants, but at least the species I have observed will not refuse any a 'The composition of these pasty masses suggests great similarity with the pellets found in the infra- ' _ buecal pockets of many ants. Part of Neger’s “ant-bread-crumbs" may well have consisted of such _ infrabuceal pellets, which, after being regurgitated by the ants, were merely discarded outside the nest 364 Bulletin American Museum of Natural History {Vol. XLV insects that may be obtainable. The seeds, however, have the very great advantage that they keep for a long time; they can be accumulated in granaries, thus providing abundant provisions, not exactly for the winter, as the ancient sages maintained, but in general for any periods of scarcity. Ants and Epiphytes Wherever in tropical and warm temperate regions the continued dampness of the air allows plants to thrive without being dependent on the soil for their water supply, epiphytes or air plants become an im- portant and often very striking feature of the vegetation. They are especially abundant in the humid rain forest and are at their best in the mountain cloud forests of the tropics. The roots of these plants, boring into the many crevices on the tree’s surface and retaining in their net- work decaying vegetable matter, rapidly loosen the outer layers of the bark and accumulate a cover of humus, affording favorable ecological conditions for a great variety of animals. Ants have not failed to recog- nize the nesting facilities here offered them by the many nooks and the uniform moisture and ventilation of this aérial root system. Indeed, the botanical collector in the tropics soon learns of the partiality of ants to the cover of humus on tree bark among and beneath the epiphytes. Though the ethology of the various ants that live with epiphytes has been but little studied, there are a number of observations to show that theinterrelations of these organisms are not always merely accidental but have in some cases produced reciprocal adaptation. The reader is referred to the Synopsis of Myrmecophytes (p. 494) for an account of the epiphytic Myrmecodia, Hydnophytum, and related rubiaceous genera which habitually harbor ants in the tubers of their rhizomes; similar pseudobulbs, inhabited by ants, are also known for a number of epi- phytic ferns (p. 497.) Ridley (1910, pp. 466-470), from observations in Singapore, has called attention to the tact that ants, mainly of the genus Dolichoderus, seem to be of considerable importance to the growth of certain epiphytic orchids. As soon as these plants start to grow, the ants bring up soil trom the foot of the tree and fill the spaces between the roots, thus con- structing shelters in which they raise their brood. This soil supplies nutritive substances to the roots and also keeps them cool and moist. From a comparison with young plants grown under different conditions, it would appear that the presence of ants among the roots is distinctly advantageous to the epiphyte, since seedlings not infested by ants are much weaker and suffer more from the drought. Though certain epi- phytes, such as the orchid Dendrobium crumenatum Swartz and the ferns __Wheeler, Ants of the Belgian Congo 365 Asplenium nidus Linnzus and Platycerium biforme Blume, are apparently re attractive than others to ants, Ridley does not mention that any of se plants in Singapore grow only on arboreal ant nests.'_ According to a number of species of Brazilian ants have acquired the habit of selecting seeds of certain epiphytes, which they carry up trees and shrubs into the crevices on the bark and into the axils of the branches, where they cover them with soil. As the plants grow their entangled roots pro- duce sponge-like ant nests with epiphytic shoots growing out on all _ sides, the whole resembling “witch-brooms”’ or bird nests. In certain _ parts of the Amazonian Rain Forest these aérial agglomerations of plants __ are so abundant as to form one of the striking features of the scenery. = (Ule, 1901, 1905a, 1905d, 1906a, and 1908, pp. 435-436.) * ~—__ Ule has described two main types of these so-called seteprdane.” ‘The largest are made by Camponotus femoratus (Fabricius) and placed high in the trees of the inundated forest; they consist of the following _ plants: Philodendron myrmecophilum Engler, Anthurium scolopendrinum Kunth variety Poiteauanum Engler, Streptocalyx angustifolius Mez, _ ~Achnea spicata Martius, Peperomia nematostachya Link, Codonanthe ____Uleana Fritsch, and Phyllocactus phyllanthus Link. The smaller gardens ___ are more elegantly constructed and inhabited by species of Azteca (A. ____ traili Emery, A. ulei Forel, and A. olitriz Forel) ; they are preferably placed in the lower trees and show the following flora: Philodendron myrme- __ cophilum Engler, Nidularium myrmecophilum Engler, Ficus paraénsts Link, Marckea formicarum U. Dammer, Ectozoma Ulei U. Dammer, _ Codonanthe formicarum Fritsch, and two Gesneriacee. Ule claims that, ___ with the exception of Anthurium scolopendrinum and Phyllocactus phyl- lanthus, these “ant epiphytes”’ are so intimately connected with the ants that they are not found in the Amazon Basin in any other station. If Ule’s conclusion be true, we have here a most remarkable instance of __ *selection’’ practiced by ants. As pointed out by’ Massart (1906), the ____ results in this case show a striking parallelism with the effects of cultiva- tion by man of crops and vegetables. By persistently caring through countless generations for the cultivated plants, man has grddually de- prived them of most of their means of defense in competition with other plants and against the hardships of environment. Crops and vegetables, when left to themselves, are no longer able to hold their own in the wae (1895) in Gh ny me i tee oe noticed the necessity for a presence of ~ in the epiphytic certain assure the healthy growth of these plants. J. Rodway ae hg and 139) Inentions tat, ie British pod ne ge of the epiphytic orchids ( of th the gpa oO ium) shelter large aa ccambing ol cate in thaon woven y Be the ants sate dling u yl interestices to a a waterproof nest, so that the collector finds it very difficult to he plant without being severely bitten. 366 Bulletin American Museum of Natural History [Vol. XLV struggle with wild plants. Similarly, in the case of the plants domesti- cated by the ants in their “gardens,” though it is certain that the seeds of these epiphytes are occasionally dropped elsewhere in the forest, they have lost the devices which allowed them to fight their rivals and are at present doomed unless cared for by the ants. The partiality of certain ants to the clusters of Tillandsia and other epiphytic bromeliads was first noted by Wheeler (1901la, pp. 526-528, and 1901b) in Mexico. He relates his experiences as follows: On accidentally pulling to pieces one of the large bud-like epiphytic tillandsias (probably Tillandsia Benthamiana Klotzsch), very common both in this and other localities about Cuernavaca, I was surprised to find it containing whole nests of ants, with their larve and pup snugly packed away like so many anchovies in the spaces between the moist overlapping leaves. A closer inspection showed that the ants had gnawed little holes through the leaves to serve as entrances to their chambers. These holes occasionally perforated a single leaf, but quite as often they threaded several leaves and extended to the very core of the bud. Sometimes a single colony of ants was divided up into companies, each occupying the space under a single leaf. But the most remarkable fact concerning these nests was the frequent occurrence of two or even three flourishing colonies belonging to different species in a single tillandsia, the whole habitable basal portion of which was rarely more than two to three inches long by one and one-half inches in diameter. Often these colonies were curiously inter- mingled in such a manner that there was no actual blending and the space under a single leaf was always occupied by ants of the same species, still, whole colonies or portions of a single colony were often completely surrounded by leaf spaces occupied by another colony. Wheeler collected the following ants from these Mexican tillandsias: Pseudomyrma gracilis (Fabricius) variety mexicana Emery, Cremato- gaster brevispinosa (Mayr) variety minutior (Forel), Leptothorax petio- latus Forel, Cryptocerus aztecus Forel, C. wheeleri Forel, Camponotus reclangularis Emery variety rubroniger Forel, and C. abdominalis F. Smith variety. Though the tillandsias appear to suffer no injury from their tenants, Wheeler is not inclined to regard this association of plants and ants as a case of symbiosis, because at least four of the seven species enumerated above occur also under other conditions in the neighborhood of Cuernavaca. Wasmann (1905a, p. 210, Pl. vin, fig. 1) also describes and figures an interesting carton nest of Crematogaster sulcata (Mayr), from Rio Grande do Sul, Brazil, which was interwoven in a pensile cluster of epiphytic tillandsias. Calvert (1911), in Costa Rica, found the clumps of epi- phytic bromeliads frequently inhabited by ants, especially by the large black species Odontomachus hastatus (Fabricius) “with enormously developed jaws, bent near the tip, which are carried wide open and measure one-quarter inch from tip to tip; occasionally they would be _Whéeler, Ants of the Belgian Congo 367 a ——" ee s Or ed shut with a very audible click.”"' A species of A pterostigma, A curious case of parabiosis between Odontomachus affinis Guérin species mayi Mann and Dolichoderus debilis Emery variety rufescens These two species of ants were nesting agether i in an earthy nerietiiee 3 ii the fork of the branches of a tree about 40 feet above the ground: a ait recently Wheeler (1921) has published much additional information with regard to similar “ant-gardens” or “flower-gardens” vomit Laboratory of the New York Zoological Society at Kartabo, _ British Guiana. These gardens agreed very closely with Ule’s descrip- _ tion even in their floral make-up. Among the plants growing out of the a a. or elliptical lumps of black earth, which vary from the size eb a establish flourishing pikictise in the gardens, namely, Camponotus _ (Myrmothriz) femoratus (Fabricius), Crematogaster limata F. Smith _ subspecies parabiotica Forel, Anochetus (Stenomyrmex) emarginatus (Fabricius), and one or more small, black species of Azteca very closely _ related to, if not the same as, the species taken by Ule in Brazil. The _ Camponotus and Crematogaster are by far the most frequent, occurring in fully 90 per cent of the gardens; the Aztecx are rather sporadic and the q Anochetus even less numerous. In more than 80 per cent of the gardens Camponotus and Crematogaster nest together in friendly parabiosis. The former, large and aggressive, and the latter, tiny and timid, mingle in the same long files that continually ascend and descend the trees, traverse __ the soil and explore the foliage. Their main occupation is to herd the jheseting > ape nother ponerine, weet ooh i Forel, of the dense Rain Forest of qestieatnese on the trees in ‘the Sone o da Platycerium, Its sting is much Sriksnstas Fost 1915, Arkiv. {. Zool., IX, No. 16. p p.7 in a recent on the fauna of "Costa Rican necanatiads (1913, p . 273), evidently has misunderstood Calvert, for he writes: “ Parmi les animaux bromélicoles ay cophages, on peut citer Champignone "0 Fourmis os ER Odontomachus, dont la nourriture habituelle est constituée par des Cpentemasios B ponstine ant y known to _ on on pane a0 ~ Condyeope subunilateralis P. anv D. Saccarvo, 1905, ‘Syll. Fungorum,’ XVII, p. 3 826. 3 From the confluence of the Paraand Surinam Rivers, Dutch Guiana, __ on the thorax of an unidentified ant (J. Michaélis Coll.). . Cordyceps unilateralis (L. and C. Tulasne) 3 Torrubia unilateralis L. anv C. TuLasne, 1865, ‘Selecta Fung. Carpologia,’ III, p. 18, PI. I, figs. 3-4. Berxetey anv Cooke, 1876, Journ. Linn. Soc. London, ; _ Botany, XV, p. 394. = arene unilateralis Saccarpo, 1883, ‘Syll. Fungorum, Pyrenomyc.,’ II, p. 570. = Fawcett, 1886, Ann. Mag. Nat. Hist., (5) XVIII, p. 317. Tuaxrer, 1891, Botan. Gazette, XVI, p. 203. M. C. Cooxe, 1892, ‘Vegetable Wasps and Plant Worms,’ p. 33, fig. 7. Masser, 1895, Ann. of Botany, IX, p. 15; 1898, Revue Mycologique, XX, p. 86. Specazztnt, 1912, An. Mus. Nac. Hist. Nat. Buenos Aires. XXIII, p. 76. This is a rather generally distributed parasite of ants, andit attacks other insects too. Originally described from Brazil on Atta cephalotes (Linnzeus), it was again found there on the same ant by Traille.' Fawcett records it on Camponotus abdominalis (Fabricius), =atriceps (Smith), also from Brazil, and on Echinopla melanarctos Smith and Polyrhachis merops Smith, both collected by A. R. Wallace at Tondano, a village in the island of Celebes. Thaxter found it in North America on an ant which was not further specified at the time, but is, according to Prof. Wheeler’s identification, Camponotus herculeanus (Linnzus) subspecies pennsylvanicus (De Geer) from North Carolina? Finally, Spegazzini mentions it from an unidentified ant found at Puerto Le6én, Misiones, Argentina. * The external part of this Cordyceps consists of a black, very slender, thread-like stroma, 13 to 20 mm. long and ‘d to 4 mm. thick at the base, onaepen gh ce gg AL be pte dl R. Inda (1907, p. 4, fig. 2) on a leaf-cutting attine ant ‘An form Cuarnava hAeeording to information given by Prof. Wheeler, there are also in Prof. Thaxter's collec- Cordycepa on Camponotus herculeanus subspecies pen penneyleanicus variety noveboracen- ais (Fitch) from Maine, ae C. abdominalis (Fabricius) from Trinidad. An unidentified Cordyceps has also been mentioned on peter ae sexquttatus from Brazil by Fawcett (1886, p. 317). 392 Bulletin American Museum of Natural History (Vol. XLV Fig. 78. Cordyceps unilateralis variety javanica F. v. Hohnel, growing on Camponotus ies, near Batavia, Java: a, ant with complete fungus, X 5.5; _}, a of asci, X 7; ¢, longi’ sec- tion of capitulum showing perithecia, X 30 (after v. Héhnel, 1909.) feebly bent about or above the middle of its length, where it bears on one side the perithecia fused into a subglobose or hemispherical head, 1 to 2mm. in diameter, with rosette-like protuberances. Cordyceps unilateralis variety javanica F. v. Héhnel Cordyceps unilateralis variety javanica F. v. HOunex, 1909, Sitzb. Ak. Wiss. Wien, math. naturw. KI., CX VIII, Abt. 1, p. 305, PL, figs. a-h. Saccarpo AnD TROTTER, 1913, ‘Syll. Fungorum,’ XXII, p. 512. The stroma of this fungus (Fig. 78 a-c) was growing out between the head and thorax of an unidentified ant, probably, to judge from the drawing, a species of Camponotus, collected near Batavia, Java, by van Rozenburg. ers ___Wheeler, Ants of the Belgian Congo 393 Fig. 79. Isaria myrmicide C. G. Lloyd, growing on Pachycondyla striata F. Smith, at Tijuca, in the vicinity of Rio de Janeiro, Brazil: a, natural size; b, X 7. 394 Bulletin American Museum of Natural History [Vol. XLV ~ Isaria myrmicide C. G. Lloyd Tsaria myrmicide C. G, Luoyp, 1920, Mycological Notes, No. 62, p. 915, Pl. CXLIT, figs. 1636 and -1637. The above name is given by C. G. Lloyd to a parasite found in Brazil by J. Rick on an unidentified ant; his figure evidently represents the petiole and gaster of a ponerine, perhaps of the same speciesof Pachy- condyla mentioned below as host of this fungus. The brief description reads as follows: “This is not a Cordyceps as would appear from the photograph but an Jsaria which is only named for convenience in the museum. Our figure (Fig. 1636) enlarged six-fold tells all to be told about it. Fig. 1637 is natural size. The stem is about a centimeter long, slender and black. The head is cylindrical and white. Spores are ‘pip-shape,’ 3x4" Fig. 79 shows a parasite of Pachycondyla striata F. Smith,' which is evidently Lloyd’s “‘Isaria myrmicide.”’ The drawing was made from a specimen in the Herbarium of the New York Botanical Garden kindly loaned to me by Dr. N. L. Britton and Dr. F. J. Seaver. It was obtained by J. N. Rose and P. G. Russell in 1915 at Pei in the vicinity of Rio de Janeiro, Brazil. Stilbum formicarum Cooke and Massee Stilbum formicarum Cooke AND Masser, 1889, Grevillea, XVIII, No. 85, p. 8. M. C. Cooke, 1892, ‘Vegetable Wasps and Plant Worms,’ p. 38, PI. 1, fig. 12. An undetermined species of ant, sent from Cheltenham, Victoria, Australia, was bearing upon its body a little Stilbum, with elongated slender stems, from five to eight millimeters in length, black, and flexuous, slightly thickened towards the base, and bearing at the apex an obovate, pink-colored capitulum or head, with elliptical conidia (104 long and 3u broad). Several of these fungi occurred on the body of each dead insect. The genus Stilbum comprises imperfect fungi, usually placed in a family StrLBace&. Most of the species are saprophytic and only a few have been found on insects. It is quite possible that the Australian form mentioned here represents the conidial form of some ant-attacking Cordyceps, and I have, therefore, thought it convenient to mention it in connection with the Hypocreacee. 'The ant was identified by Prof. Wheeler. Wheeler, Anis of the Belgian Congo 395 ghly specialized and most interesting of fungoid parasites of insects. ill are found growing on living arthropods exclusively. The family is ually included among the Ascomycetes and, even in his most recent ers, Thaxter sees no sufficient reason why it should not be placed in the ‘Pyrenomycetes. Because of their combining in some respect rities of the true Ascomycetes with others shown by certain Alge ‘of the class Floridex, certain mycologists suggest that these fungi be _ considered as a class of themselves, for which the names Phycascomycetes : or Laboulbeniomycetes have been used. The following brief account of the Laboulbeniacez is adapted from R Thaxter’s admirable monographic studies of these plants and will, ‘it is hoped, enable entomologists to recognize them without difficulty. Unlike the Cordyceps described above, they are inconspicuous and, when - examined i in situ on the host insect, appear in general like minute, usually _ dark-colored or yellowish bristles or bushy hairs, projecting from its __ ehitinous integument either singly or in pairs, more commonly scattered, but often densely crowded over certain areas on which they form a furry ee: coating. When studied with a proper magnification, the structure of a _ fully developed parasite corresponds to the following general scheme. ire (polycellular) main body, or receptacle, is fixed by means of a blackened base, or foot, to the integument of the host, and consists in most cases of a very small _ number of cells differently arranged in different genera. This receptacle gives rise above to certain peculiar appendages of very variable form, commonly connected with the production of the male sexual organs; while from the same individual, with few exceptions in which the plants are diacious, female organs are also variously __ produced from which perithecia are eventually developed. In the perithecia, which may arise singly or in considerable numbers from a given individual, and which are quite remarkable in structure, are produced the reproductive bodies or ascospores that are formed in asci identical in all respects with the organs thus named in other members of the great group of ascomycetous fungi. The ascospores thus formed germinate on the surface of the host to which they become attached by a blackened modification of their basal extremity, and, without the formation of any hyph. grow directly to new individuals by means of successive cell divisions, (Thaxter, 1896, p. 198.) Perhaps the mot remarkable peculiarity of the Laboulbeniacee is their ability to thrive freely on their host without interfering much with its activity, inflicting little if any appreciable injury. The parasit- ism is external and, except in rare instances in which the foot sends into the body a rhizoid-like haustorium, the parasite derives its nourishment through a at most slight perforation of the host’s integument. Indeed, so feeble are the ill-effects of their parasitism that the idea has at one ee ‘The LABOULBENIACE or Laboulbeniales constitute by far the most 396 Bulletin American Museum of Natural History {Vol. XLV time been advanced these fungi be mere saprophytes, not feeding on their host but absorbing from the surrounding humid air such elements as are needed for their development (Cavara, 1899). Rick (1903), commenting upon the abundance of Rickia Wasmannii in some ant colonies, goes even a step further. ‘The animals,” he writes, ‘apparently suffer but little or almost not from the fungus; one finds decidedly populous colonies which are much attacked. Possibly the animals may even derive some benefit from the fungus. It is not much out of the question to think of a kind of symbiosis, though I cannot for the present give any further in- dication concerning this point. Perhaps the fungus could be of advantage to the ants in providing them with sugar.” There is, however, not the slightest proof for Rick’s surmise that the fungus is of any real use to its host, while there is plenty of evidence that the Laboulbeniacez are true parasites. ‘‘The rigid limitation of species of Laboulbeniales to single genera or even species, of insects, which holds in general throughout the group, could hardly, it would seem, be ex- plained on the basis of pure saprophytism; and-although, as previously stated, the growth of these plants is not associated with any appreciable injury to the host, it is nevertheless a true parasitism of a typically obligate type.’”’ (Thaxter, 1908, p. 223.) Moreover, the exact manner in which the fungus derives its food from its host is still not quite clear. The occurrence of a number of rhizoidal forms seems to render it certain that all Laboulbeniaceew feed on the juices of the insect; in the ant parasites, as in a majority of cases, these nutritive elements are absorbed, without penetration, through the sucker-like foot (Thaxter, 1908, p. 248). According to Cépéde (1914, p. 396), the fungus takes from the superficial layers of chitin certain carbohydrates which are localized there (glucose and glycogene). The greater number of Laboulbeniacee attack beetles, especially of the family Carabide; they are much rarer on other insects and only the three following species have hitherto been recorded from ants. They are among the smallest members of the family, not exceeding one-tenth of a millimeter in total length in the North American Laboulbenia formi- carum; the two other forms being slightly larger. Rickia Wasmannii Cavara Rickia Wasmannii Cavara, 1899, Malpighia, XIII, p. 182, Pl. vi. Riex, 1903, (Esterreich. Bot. Zeitschr., LILI, p. 163, fig. THaxtrer, 1902, Proc. American Ac. Arts Sc. Boston, XX XVIII, p.39; 1908, Mem. American Ac. Arts Sci. Boston, XIII, pt. 6, p. 248, Pl. xxxiv, figs. 1-13. Cirztpe anp Picarp, 1909, Bull. Scientif. France et Belgique, XLII, p. 252. Prcarp, 1913, Bull. Soc. __ «Wheeler, Ants of the Belgian Congo 397 sal found it on Myrmica levinodis Nylander, it was observed Rick on the same ant at several other localities in Luxemburg (Belle Vue), Germany (Berncastel on the Moselle), and Austria (Feldkirch and Garina in the Vorarlberg). Spegazzini mentions it from Italy on Myrmica scabrinodis Nylander (Fig. 80a-b). Donisthorpe (1912, p. 5; 1913, p. 96; 1915, p. 154) mentions the discovery at Ran- noch, England, of a nest of Leptothorax acer- vorum. (Fabricius), all the ants of which were covered with a fungus, though quite alive. The specimens, unfortunately, were lost, but the author thinks that the fungus was probably a species of Laboulbeniacez. Fig. 81. Rickia formicicola Spegazzini, a parasitic fungus of Prenolepis silvestrii Emery in Argentina. Two mature individuals, X 300 (after Spegazzini, 1917). . Rickia formicicola Spegazzini Rickia formicicola, 1917, SpeGazzint, An. Mus. Nac. Hist. Nat. Buenos Aires, XXIX, p. 665, fig. nim. 195. This species (Fig. 81) was found in the island of Santiago, La Plata, Argentina, growing on Prenolepis silvestrii Emery. Laboulbenia formicarum Thaxter Laboulbenia formicarum, 1902, Tuaxter, Proc. American Ac. Arts Sci. Boston, XXXVIII, p. 39; 1908, Mem. American Ac. Arts Sci, Boston, XIII, No. 6, p. 359, Pl. vm, figs. 14 and 15. Wueeter, 1910, Psyche, XVII, pp. 83-86. J. Bequaert, 1920, Bull. Brooklyn Ent. Soc., XV, p. 71. 398 Bulletin American Museum of Natural History [Vol. XLV This parasite (Fig. 82a-b) attacks various species of North American ants; strange to say, it has only been recorded thus far from the vicinity of Boston, where it appears to be rather common. Thaxter discovered it at Cambridge, Massachusetts, on Lasius niger variety americanus. Emery and Formica subpolita variety neogagates Emery. Wheeler found the same fungus infesting nearly all the nests of Lasius niger variety neoniger Emery, on the seashore at Ellisville, Massachusetts, and gives scme interesting details with regard to the ecology of the infested colonies. On the beach itself, which consists of a deep layer of pure sand, there are colonies of Formica fusca variety argentata Wheeler, Myrmica scabrinodis Nylander variety sabuleti Meinert, Tapinoma sessile Say and Lasius neoniger. The last is far and away the most abundant and its workers are of large size. None of the ants in this locality, including the neo- niger, was found to be infested with Laboulbeniacex. On the border of the salt meadow, however, immed- iately adjoining the beach, where the soil is moist, consisting of a mixture of rather sour, decomposing humus mixed with sand, and probably not infre- quently wetted by the spray and occasionally even submerged at very high water, the only ant is L. neo- niger, but its colonies are less populous than those on the beach, the workers are distinctly smaller and are practically all infested with the Laboulbenia. Passing’ over from this zone of infestation to the pasture land adjoining the salt meadow, the variety neoniger is replaced by L. niger L. variety americanus Fig. 82. Laboulbenia formi- Emery which is the form of the species eommonly carum Thaxter, a fungous parasite occurring in higher and dryer pastures and fields. - of various North American ants None of the workers of this form, which lacks on the posi: mag eg ie scapes and legs the erect hairs so conspicuous in the variety neoniger, was found to be infested with the fungus. It would seem, therefore, that while neoniger, unlike any of the other ants, is able to exist in a depauperate condition in the damp, sour soil at the edges of salt meadows, it does so only at the risk of becoming infested with Laboulbenia formicarum. Indeed, the infestation of the ants in this strip of littoral at Ellisville is often so execes- sive that they resemble hedgehogs, fairly bristling with tufts of the fungus. (Wheeler, 1910a, p. 84.) Though Laboulbenia formicarum may occur on all parts of its host, it appears from Wheeler’s observations that it grows most abundantly on the abdomen, middle and hind femora and tibia, and posterior portions of the head. The thorax and cox, as a rule, are entirely free from the — fungus; the clypeus and gula are generally free, and this seems to be in- variably the case with the mandibles, antennal funiculi, palpi, labium, Wheeler, Ants of the Belgian Congo heen Saexille, anc and eyes. In a very few specimens, one or two of the little - plants were s were seen on the antenna! scapes, but, 2s a rule, these organs are perfectly clean. In August 1919, I took a worker of Formica pallide-fusca subspecies _ schaufussi Mayr infested with Labeulbenia formicarum at Forest Hills near Boston. I was, however, unable to locate the nest to which this individual belonged, but this observation shows once more that this _ fungus, though restricted to ants, attacks indifferently many species _ (Bequaert, 1920). Prof. Thaxter has also informed Prof. Wheeler that he has taken this Laboulbenia on various species of Formica, at Cam- _ bridge, Massachusetts. __ Several so-called “imperfect fungi” —incompletely developed, coni- _dia-bearing or sterile stages of various Ascomycetes—are known to _ attack insects, and some of these have been seen on ants. I have men- tioned above Stilbum formicarum Cooke and Massee and have also alluded to the Isaria stage of Cordyceps, which may be expected on ants, since so many species of the latter genus have been found in the asci- _ bearing stage on these insects. H. Bischoff (1912) has mentioned the finding by Quiel, at Potsdam, Germany, of two nests of Formica rufa heavily infested with fungous growths, about the size of a pin-head and attached mainly to the thorax, more rarely to other parts of the body. __ The ants were apparently but little hampered by their parasites. From _ cultures obtained with these fungi, Bischoff concludes that they belonged to several species, among them a Mucor (of the spinulosus group), a _ Penicillium, and a yeast with sexual reproduction; characteristic brown _ hyphz present in the tufts on the ants, were not obtained in the cultures. More recently, Thaxter (1914, p. 239) found in the vicinity of _ Cambridge, Massachusetts, a fungus forming blackish incrustations on _ various parts of ants, and giving rise to a few short, colorless, erect _ branches; the exact nature of this plant has not been determined, nor is - the name of its host mentioned.' Thaxter (1891, p. 203, Pl. xx, figs. 1-9) has described, under the - name Desmidiospora myrmecophila, a new genus and species of fungus __ which was growing luxuriantly on a large black ant fastened to the under side of a rotting log in Connecticut.2. The hyphx, much branched and septate, covered the host in a white flocculent mass; they emerged a 1J. Leidy (1884) has recorded finding in Pennsylvania a Ca under the bark = Je ofa | eae a dap any was infected with a Hee gus which spread spread through e gh een al the boil. : This an ve a saprophytic f nv ant after ; may This ant io ale oped Le ogee tant Wheeler in Prof. ie bce’ 's collection as Camponotus herculeanus 6ul neylranicus ( om ‘Prof. Thaxter ter also possesses the same fungus from New Raagehiee ged meagan subspecies pennsyleanicus and its variety aanshoracenots (Fitch). 400 Bulletin American Museum of Natural History [Vol. XLV especially from between the abdominal segments, enveloping the insect more or less completely and extending a short distance over the sub- stratum. The spores are of two kinds, the microconidia being minute — (12 2u), hyaline, subfusiform, and produced at the apex of subulate lateral basidia; while the macroconidia are much larger, terminal, brown, flat, multilocular, irregularly lobed, up to %>o mm. broad (80-100 X 68-904 and 12-14 thick). Thaxter remarks that it is not impossible that this fungus is an imperfect form of some Cordyceps or possibly parasitic on an immature Isaria or Cordyceps previously developed within the insect. As it is, Thaxter places Desmidiospora among the Hyphomycetes and Lindau! regards it as genus of the Mucedinacez. Some years ago Patouil- lard (1892) described, under the name Hirsutella entomophila, a curious fungus found growing on a beetle in Ecuador. At first sight it resembled an Isaria, but Patouillard thought he had observed that the spores were borne on basidia; he, therefore, included this parasite among the Basi- diomycetes, placing it in the Clavarie. Recently, however, Speare (1920) has shown that this, as well as similar fungi, do not produce true basidia and must be removed from the Basidiomycetes. It is rather a definite form of imperfect fungi, probably a stage of one or more species of Cordyceps or related genera. In the same paper, Speare remarks apro- pos of Desmidiospora myrmecophila Thaxter (p. 65): ‘ While its resting spores are anomalous in character, and although no structures analogous to the synnemata of Hirsutella were described, its subulate sporophores and fusoid spores are of the same type as the corresponding organs of the form under consideration.” A snowy white mould, Sporotrichum minimum Spegazzini (1881, pt. 4, p. 123 of reprint; Saceardo, 1886, p. 101; M. C. Cooke, 1892, p. 37), also one of the Mucedinacee, was found in Argentina upon the putrescent body of Acromyrmex lundii (Guérin), in a rotten trunk. It was diffused over the insect, at first in a powdery and then a cottony white stratum, forming minute tufts. The threads were creeping and densely interwoven, branched, very slender (scarcely 2u in diameter), sparingly septate, hyaline, with conidia scattered here and there. It is possible that both this Sporotrichum and Thaxter’s Desmidios- pora myrmecophila are mere saprophytes, which have grown over the ant after the death of the insect. . Hormiscium myrmecophilum Thaxter, another imperfect fungus found on ants, is described by Thaxter (1914, p. 238, Pl. xrx, figs. 22-25) as follows: ‘Filaments nearly hyaline, becoming brownish, darker near 1In Engler and Prantl, 1900, Die Natirl. Pflanzenfam., I, Abt. 1**, p. 454. 1922] Wheeler, Ants of the Belgian Congo 401 __ the base, closely septate, the cells often as broad as, or broader than long, undifferentiated, distally bluntly rounded, erect or curved upward, rigid, ‘simple, less frequently sparingly branched, tapering but slightly if at all, __ one to several arising from a deeply blackened foot of variable size and shape. Maximum length about 2804 by 7-8u in width.’ It was found on various parts of a species of Pseudomyrma collected by W. M. Mann along the Amazon River, Brazil; the majority of the individuals taken from a nest were infected by the fungus, which is sufficiently large to be readily visible as it projects from the surface of the host. It produces no _ differentiated cell-groups or definite spores, as far as has been observed, and appears to propagate itself by fragmentation only, the filaments _ proliferating after a terminal portion has been broken off. The opaque and somewhat variable foot, by which the individuals are attached to the _ surface of the host, appears to correspond to such a small fragment brok- __ en from a hypha, which, adhering laterally, becomes blackened and in- ____ durated, and gives rise to new filaments, while at the same time it serves the office of attachment as well as of food absorption. The other mem- bers of the genus Hormiscium are saprophytic, being mostly found on __ decaying vegetable matter; it is placed in the family Dematiacez among _ the Hyphomycetes. Finally, I must mention that Donisthorpe, in his treatise of British ‘ants (1915, p. 235, fig. 86; see also Donisthorpe, 1913, pp. 96-97), figures a worker of Lasius umbratus variety mizxto-umbratus Forel with patches of alge on body and legs. Concerning this parasite, he _ expresses himself as follows: ; On August 11th, 1912, when at Weybridge in company with Professor Wheeler, _ we found two colonies of this variety, very many of the ants of both being infested _ with a curious dark brown warty growth in patches on parts of the body and legs— this Wheeler thought might be a fungus which was unknown to him. I kept a number ___ Of these ants in captivity, and added uninfected workers of umbrata from other locali- _____ ties; the growth however did not increase nor spread to the new ants, but rather __ seemed to decrease. I sent some of the infested ants alive and others in spirit, to Dr. Baylis Elliott, and she considered the patches were colonies of unicellular organisms growing on the outside of the ants; eventually she came to the conclusion that they _ were not fungoid growths, but probably colonies of an alga. 3 : Intracellular Bacteria of Ants In various groups of insects unicellular organisms of a fungous or ____ bacterial nature have been discovered inside certain cells of the body. 4 They are apparently not parasites, but must rather be considered as 'Hormiscium pithyophilum variety myrmecophilum J. B, Elliott (1915, pp. 139-142) is an entirely different fungus, which was found growing on the carton walls of certain nests of Lasius umbratus in England (p. 379). 402 Bulletin American Museumoj Natural History (Vol. XLV living in symbiosis with their host. Special devices, often of a compli- cated nature, assure their transmission within the insect egg from one generation to another, so that they have become normal constituents of all the members of certain species, genera, families, or even higher groups of insects. Some of these micro-organisms float freely in the lymph, and in many cases great numbers of them also fill the plasma of certain fat-cells, thus forming so-called “mycetocytes.”” The mycetocytes may occur isolated in various parts of the body, or they may be grouped together and even more or less fused into special fungous organs, the ““mycetoms.’’ Sometimes micro-organisms of two or even three different kinds live within separate cells of the same host, either quite apart from one another or in compound mycetoms. Typical illustrations of intra- cellular mycoses are presented by certain hemipterous insects. In particular, all the Homoptera possess such intracellular, hereditary symbionts; their mycetoms are often enclosed within a pigmented epithe- lium and connected with special branches of the tracheal system of the insect (P. Buchner, 1913). Intracellular symbionts also occur in certain ants, and it is probable that their presence in these insects is more frequent than is known at present. They were first seen by Blochmann (1884, 1888) in Camponotus herculeanus subspecies ligniperdus, densely filling cells which this investi- gator regarded as belonging to the epithelium of the intestine. According to Buchner’s recent researches (1919) the intracellular organisms of that ant are really contained in special mycetocytes placed in a continuous layer beneath and between the true epithelial cells of the mid-gut. They are present in all individuals in the form of tiny, thread-like bodies, 10 or 12 uw long, generally regarded as bacteria. At the beginning of the sexual maturity of the insect, some of the symbionts leave their mycetocytes, in the worker as well as in the queen. They invade the egg-follicles from all sides and penetrate the egg, the entire plasma of which at first becomes densely filled with bundles of bacteria placed parallel to one another; but, as the egg grows, these organisms are pushed to its posterior pole. Blochmann found similar, but smaller (4 to 5 wu), organisms in Formica fusca, where they occupy two groups of cellsin the adipose tissue. According to Buchner (1918, p,77, footnote), intracellular bacteria live in many species of Camponotus, such as C. senex (F. Smith), C. maculatus subspecies congolensis Emery, C. maculatus subspecies brutus (Forel), C. maculatus subspecies atramentarius Forel, C. rectangularis subspecies rubroniger Forel, and perhaps in all the mem- bers of that genus. But they are absent in many other ants, as, for instance, in Myrmecina latreillei Curtis. Wheeler, Ants of the Belgian Congo 403 ) 2. A A Bavesiw OF AvEtiay MyYRMECOPHYTES y defined as plants which Sirians life are continuously inhabited by species of ants. This definition, however, calls for certain ex- atory remarks which will be found in the introduction to the general vy of recorded ant-plants (p. 494). What is known of the ecology er, in addition to my own field observations. For the convenience he entomologist, I have compiled from the taxonomic literature the hnical descriptions of these plants. To most students they will, I y seems able to offer nothing better. In themselves, they afford ient apology for the fact that in so many cases a correct identifica- f the plant in question can not be made.' It is to be hoped that the ice of a specific name will not render the observations recorded rely valueless, since more often than not future field workers will be e to recognize the plants by some of the peculiarities shown in the drawings or mentioned in the text. _ Being more familiar with the African flora, I may be permitted to ll attention to a few general features of myrmecophytism as suggested ibed thus far from the Ethiopian Region, only 42 of them can be arded as more or less well-defined or probable myrmecophytes. In nore than 20 of these cases have the relations to ants been established * In other tropical parts of the world, the number of plants with cial accommodations for sheltering ants is somewhat higher (about | species in the Neotropical and 109 species in the Indomalayan, Papuan, and Australian Regions), but it must be remembered that their loras are much richer than that of the Ethiopian Region, so that their roportion of myrmecophytes is but little if any higher. The compara- e 1In the case of plants numbers are given potent tom um imens which Siew tn the hands of Prot, fe Doe Wikdenae, Directs 0 the Tasca Bee tn Candas Dome ot have been identified by that authority =e te study eet eeetedie be completed re myrmecoph are at present known from the Belgian Congo. According to the latest : mat the ef hava emake cf temas , belonging to 1261 genera, had been described from that rie nd of 1918 (Geren 101 "A de nos connaissances actuelles sur la flore du Con- * Bull. Agricole Xx, pp. 154-161). 404 Bulletin American Museum of Natural History [Vol. XLV tively small number of myrmecophytes is rather surprising considering the abundance and variety of tropical plant life and the many opportuni- ties which ants must have had to become acquainted with it. (b) The African myrmecophytes belong to a few taxonomic types, represented by 7 families and 12 genera, as follows: ; Leguminose Schotia, with 1 species. Euphorbiacee §Macaranga, with 2 species. Sterculiacese Cola, with 3 species. Scaphopetalum, with 2 species. Flacourtiacese Barteria, with 5 species. Apocynacese Epitaberna, with 1 species. Verbenacew Vitex, with 2 species. Rubiaceze Uncaria, with 1 species. Sarcocephalus, with 1 species. Randia, with 3 species. Plectronia, with 6 species. Cuviera, with 15 species. Schotia, Cola, Scaphopetalum, Barteria, Epitaberna, and Cuviera are precinctive Ethiopian genera, while the others are either also repre- sented in the Oriental and Indomalayan Regions (Macaranga, Sarco- cephalus, Plectronia) or tropicopolitan (Vitex, Uncaria, Randia). The family Rubiacez leads the list with the largest number of myrmecophil- ous species (26, belonging to 5 genera), whiclris true also in other tropical regions. For Barteria, Epitaberna, and Cuviera, myrmecophytism is to all appearances one of the generic peculiarities, probably being present in all the members. It is acurious fact that in the Ethiopian Region and elsewhere some of the largest families of the vegetable kingdom, in which differentiation into species has been most active, show very few (Leguminose, Orchida- cee) or no cases of myrmecophily. As illustrations of the latter may be mentioned the Graminez, Cyperacee, Liliacexw, Labiate, and Composite. (c) True myrmecophytes are restricted to the sections of the earth situated between the tropics, a fact easily accounted for by the uniform temperature which prevails there and permits ants to establish their perennial abodes within the rather thin walls of plant tissues. I already have shown (p. 371) that the so-called ant acacias of the dry East African . plains and Clerodendron formicarum of the savannah country are by no means myrmecophytes. When these cases are eliminated, all African ant- plants known at present occur only in the permanently moist and ever- green Rain Forest of the western and equatorial parts of the continent. All the Oriental and Indomalayan and the vast majority of the Neo- tropical myrmecophytes grow similarly in the moist tropical forest areas. Wheeler, Ants of the Belgian Congo 405 , could not be provided by annual or biennal plants. In addition, oody texture of the walls adds considerably to the solidity of the a and to the protection of the formicaries. In a number of cases d by ants are abundantly covered with long, stiff hairs. e) The structures offered as myrmecodomatia by the ant-plants but little diversity, are usually of a very simple type, and affect ns of the plant. There is nothing here comparable with the in- rved margins enclosing a pouch-like cavity: Macaranga saccifera. ore primitive condition of stipular myrmecodomatia is illustrated by Jragoga described on p. 453. —The leaves produce pouches at the base of the blade: species of d Scaphopetalum. y because they do not offer enough solidity and permanency as rs for formicaries. In the few cases in which I observed ants using ollen stipules of Macaranga saccifera and the foliar pouches of ii rand Scaphopetalum Thonneri, the colonies were very smal] stems of the plant are externally normal, but hollowed out sally their entire length: Vitex Staudtii and Barteria Dewevrei. es (Unearia, Sarcocephalus, Plectronia, and Cuviera), or at the base of ain branches (Barteria fistulosa). 406 Bulletin American Museum of Natural History [Vol. XLV In other tropical regions there are a number of additional types not yet recognized in Africa, such as stipular thorns (Acacia), swollen petioles (Tachigalia, Nepenthes bicalcarata), pitcher-shaped leaves (Dis- chidia), inflated leaf-sheaths (Korthalsia), hollowed pseudobulbs (Schom- burgkia), and fistulose rhizomes (Polypodium sinuosum, Lecanopteris carnosa, Myrmecodia, etc.) In the case of stipular or leaf pouches, the slit which leads into the © cavity is a natural result of the production of the pouch. In all other African myrmecophytes, there is no preformed entrance to the domatia and the apertures are gnawed by the ants. (f) A very small number of African ants have become adapted to nesting in the domatia of ant-plants. A distinction should be drawn here between OBLIGATORY plant ants, that live exclusively in myrmecophytes, and species which are only occasionally or accidentally associated with these plants and may therefore be designated as FACULTATIVE (Wheeler, 1913, p. 115). Most of the African plant ants fall in the second group; they belong to such genera as Crematogaster, Tetraponera, Monomorium, Leptothorax, Telramorium, Cataulacus, Technomyrmex, and Prenolepis, which are abundant in the forest, usually leading an arboreal or semi- arboreal life; many of the species make no distinction between cavities of dead or living plants wherein to shelter their formicaries. Viticicola tessmanni and the two species of Pachysima (P. zxthiops and P. latifrons) are the only African obligatory plant ants. They have never been found away from their hosts, Vitex Staudtii in the case of Viticicola and various species of Barteria and Epitaberna myrmecia in the case of the Pachy- simez. It is possible that certain African species of Engramma and Plagio- lepis, which have been collected only in plant domatia, are also of the obligatory type, but their case calls for further investigation. There are a number of doubtful cases of myrmecophily among African plants and also others that are based on erroneous or incomplete observations. Some of these have been dealt with in the present paper under their respective families or genera, but a few others must be briefly mentioned here for the sake of completeness. Stereospermum dentatum Richard (Bignoniacez), of Abyssinia and Kordofan. According to Penzig (1894) the pith in the upper part of a flowering branch is excavated for a space of one or two internodes and the cavity is inhabited by Tetraponera penzigi (Emery), its offspring, and also some coccids. The aperture is found at the tip of what appears to be an aborted limb in the bifurcation of the flowering branch. There Wheeler, Ants of the Belgian Congo 407 the normal stems are filled with pith. Penzig be- ‘the ants trim the growing upper end of the branch in order to e pith and are thus responsible for the dichotomous inflorescence pecies. I am rather inclined to think that the galleries are bored ct larva and are only settled by ants after being left by their innibale (19072) mentions two other African Bignoniacee, Kigelia icani (Lamarck) and Newbouldia levis (P. de Beauvois), as “myrme- ;” because he found nectaria on the under side of the leaves. In itior , herbarium specimens of Newbouldia levis examined by him oo: in the upper part. of the flowering branches, the cavities yin ig one or two apertures at the base. The author assumes that these are natural formations of the plant and are settled by the ants, ch pierce the exit holes. He does not state that these insects were tually y found in the branches, and the explanation offered above for ‘cavities i in Stereospermum is probably also true here. _Grumilea venosa Hiern (Rubiacexe). Belgian Congo. “ Bush of about ., always inhabited by numerous black ants” (Dewévre; see De Ideman and Durand, 1901, p. 130). _ Microdesmis puberula J.D.Hooker (Euphorbiacew). Belgian Congo. n. Laurent vegarded this plant as a myrmecophyte; indeed some of branches on specimens collected at Bombaie and provided with ‘ e rooms, are excavated with galleries; but the myrmecophytic racter is not much pronounced.” (De Wildeman, 1910, ‘ Etudes ore fe Bas- et Moyen-Congo,’ III, 2, p. 250.) n addition to the ants indicated i in the general account of African rophytes which follows, Father Kohl collected at Stanleyville in nearby localities a number of species ‘‘in myrmecophilous plants” h have not been identified thus far in the literature. I subjoin a isos, compiled from Forel’s recent paper (1916) on the ants | le sted in the Belgian Congo by Kohl: _ Crematogaster ruspolii variety atriscapis (Forel). __ €. sjdstedti subspecies kohliella (Forel). nigeriensis variety wilniger (Forel). _ kasaiensis (Forel). kohli (Forel). ) pete solenopsides subspecies flavida variety convexiclypea (Forel). ‘onomorium oscaris subspecies springvalense variety paternum Forel. |. exiguum subspecies flavescens Forel. evelynx Forel. Tetramorium simillimum subspecies isipingense variety dumezi Forel. 408 Bulletin American Museum of Natural History (Vol. XLV Engramma laurenti variety congolense Forel. Prenolepis grisoni Forel. Leguminosez Though this is one of the four or five largest families of plants and contains many of the more common bushes and trees of the tropics, only very few of its members are known to be myrmecophytes. After the elimination of the East African so-called “ant acacias,” which, as I have shown elsewhere, do not possess true myrmecodomatia, there remains in Africa only one genus that possibly presents bioccenotic associations with ants. Scuotia Jacquin Scholia Jacquin, 1786, ‘Collectanea Austriaca ad Botanicam Chemiam et His- toriam Naturalem Spectantia,’ I, p. 93. Oxtver, 1871, ‘Flora of Tropical Africa,” II, p. 309. Harms, in Engler and Prantl, 1897, ‘Die Nattirl. Pflanzenfam.,’ Nachtriige zu III, pt. 3, p. 196. Theodora Mevixus, 1786, ‘Theodora speciosa, ein neues Pflanzengeschlecht,’ p. 16. TauBert, 1894, in Engler and Prantl, ‘Die Natiirl. Pflanzenfam.,’ III, pt. 3, p. 138. “Unarmed trees or shrubs. Leaves abruptly pinnate, with coriaceous often small leaflets; stipules small. Flowers red or purple, clustered in short often dense panicles, heads or racemes. Bracts and bracteoles caducous or subpersistent. Calyx-tube turbinate, campanulate or narrowly infundibuliform; segments 4, much imbricate. Petals 5, slightly unequal, clawed or subsessile, longer or shorter than the calyx, imbricate. Stamens 10, free or shortly coherent below; anthers uniform, dehiscing longitudinally. Ovary stipitate with elongate style and small terminal stigma; ovules 4to8 or 10, or more. Legume oblong, often falcate, compressed, coriaceous, dehiscent or subdehiscent. Seeds exalbuminous” (Oliver, 1871). This genus belongs to the subfamily Czsalpinioidex, in which the flowers are not of the papilionaceous type usual in the family, but possess a rather spreading, zygomorphous corolla; in the bud the upper sepals and petals are covered by the lower. Schotia is restricted to tropical and southern Africa and contains twelve species, one of which is supposed to be myrmecophytic. . Schotia africana (Baillon) Humboldtia africana BaiLLon, 1870, ‘ Histoire des Plantes,’ II, p. 99, footnote (Tropi- cal West Africa). Schotia humboldtioides Ot1ver, 1871, ‘Flora of Tropical Africa,’ II, p. 310. Harms, 1915, in Engler, ‘Die Pflanzenwelt Afrikas,’ III, pt. 1, p. 454, fig. 249. Theodora africana (Baillon) Tausert, in Engler and Prantl, 1894, ‘Die Natiirl. Pflanzenfam.,’ IIT, pt. 3, p. 138. “A glabrous tree of 25 to 30 feet; extremities (in our specimens) tumid im- mediately under each node, narrowing gradually nearly to the middle of the inter- node. Leaves \ to 1 ft. long, 2~ to 4-jugate, glabrous; leaflets thinly coriaceous, the Wheeler, Ants of the Belgian Congo 409 es broadly ovate, about } line ols Flowers patent, on pedicels of about i 3 to % in. long, puberulous or glabrate, the tube but slightly exceeding node and hollow inside. The wall is pierced with a hole through ants gain access to the inner cavity. This supposed myrmeco- : bited by ants after being left by their makers, as is so often in the tropics. Euphorbiacez i MacaranGa DuPetit-Thouars 2 DuPerit-Trovars, 1806, ‘Gen. Nov. meee rage p- 26. Pax, Se eticesterech, IV, AR 147, VII, p. 298. Trees or shrubs. Leaves alternate, petiolate, simple or lobed; their base often rved and sometimes peltate, occasionally penninerved. Spikes or racemes ary or lateral or sometimes forming a terminal panicle. Flowers dicecious, rarely apetalous. Male flowers small, clustered. Female flowers solitary. or minute, entire or lobed or fimbriate. Male flower: calyx globose, n bud, splitting into 3- to 4-valvate lobes; stamens sometimes few (1 to 3), ‘numetous (10 to 30); filaments short, free, very rarely united or as if branched; ers short, terminal, usually 4-celled, 4-valved, sometimes 3-celled, 3-valved, rarely dd; no rudimentary ovary. Female flower: calyx truncate or shortly toothed, iately wide-cupular or obliquely spathaceous; ovary 2- to 3- (rarely 4- to 6-) ed; styles short, stout, entire, free or slightly united at the base, rarely long, der or united in a globose mass; ovules in each cell solitary. No disk. Capsule ing up into 2-valved cocci or occasionally, when 1-chambered, almost in- scent. Seeds glebose; testa crustaceous; albumen fleshy; cotyledgns broad, (After Prain, 1912.) 410 Bulletin American Museum of Natural History [Vol. XLV The genus Macaranga includes over 170 species of trees and shrubs distributed in Africa and its islands, Indomalaya, and the Australian and Polynesian Regions. Some forty species have been described from Tropical and South Africa, fourteen of which are recorded from the Belgian Congo. It is probable that a number of the African species are more or less associated with ants, since several of the Indomalayan forms exhibit various mutualistic relations with these insects. Ridley is even inclined to believe that in some species of this genus symbiosis of the ants and the plant appears to be as complete as possible (see p. 516). Two of the African species, M. saccifera Pax and M. Schweinfurthii Pax, have persistent pouch-like stipules which are occasionally occupied by ants. Em. Laurent noticed that in another species, M. dibeleensis £. De Wildeman,} the leaves attract ants, probably by means of the nec- taries at the base of the blade; the stipules are more or less concave, not at all pouch-like, and soon deciduous, so that this species probably is not a true myrmecophyte. Macaranga saccifera Pax Macaranga saccifera Pax, 1894, Engler’s Bot. Jahrb., XIX, p. 93, Pl.1. Ta. Duranp © anp H. Scutnz, 1896, Mém. Couronnés Ac. Roy. Belgique, tru, 4, p. 246. EB. De WILDEMAN AND Tu. Duranp, 1900, ‘Contributions Flore du'Congo,’ I, p. 57; 1901, ‘Reliquie Dewevreanz,’ p. 212. E£. De WitpeMan, 1905, ‘Mission Emile Laurent,’ fase. 2, p. 130, Pls. xxx1x-xu1; 1908, ‘ Etudes Flore Bas- et Moyen- Congo,’ II, 3, p. 283. Tu. ann H. Duranp, 1909, ‘Sylloge Flor. Congol.,’ p. 496. H. Kout, 1909, Natur u. Offenbarung, LV, p. 149. &. De Witpeman, 1910, ‘Compagnie Kasai, Miss. Perm. Et. Scient.,’ p. 330. Eneuer, 1910, ‘Die Pflanzenwelt Afrikas,’ I, p. 644, fig. 555. Prarn, 1912, in Thiselton-Dyer, ‘Flora of Tropical Africa,’ VI, 1, p. 934. Pax, 1914, ‘Das Pflanzenreich,’ IV, pt. 147, VII, p. 312, fig. 51. E. De Witpeman, 1919, C. R. Ac. Sci. Paris, CLXTX, p. 394. “A shrub or tree; branches armed with spines, densely tawny-pubescent. Leaves long-petiolate, rounded-ovate, deeply 3-lobed; lobes obovate-oblong or tri- angular, acute; margin repand or toothed; base narrowly deep-cordate; 8 to 10 in. long, nearly as wide, subcoriaceous, gland-dotted beneath, with a pair of marginal glandular processes at the junction with the petiole; petiole 6 in. long; stipules converted into large coriaceous acute flask-shaped sacs. Male flowers in axillary panicles; bracts ovate, acute, subtending several flowers, buds globose. Female flowers unknown”’ (Prain, 1912). Pax (1914) distinguishes two forms: Variety a. genuina Pax and K. Hoffmann, 1914, ‘ Das Pflanzenreich,’ IV, pt. 147, VII, p. 312, fig. 51. Described by De Wildeman, 1908, in ‘Etudes Flore Bas- et Moyen-Congo,’ II, pt. 3, p. 281. See also H. Kohl, 1909, p. 150. Wheeler, Ants of the Belgian Congo 411 y glandular underneath. Rachis sparsely pilose; young : Lower Congo: in the Cataract District between Matadi and e (Laurent). Kwango: Madibi (Sapin). Kasai: Mukenge (Pogge); : Eala (Pynaert; M. Laurent); Bokakata (Dewavre); Buinba (Seret) ; M. Laurent). Eastern nite Forms: re ges near Yambuye (M. riety as dentifera Pax and K. Hoffmann, 1914,‘ Das Pflanzen- , pt. 147, VU, p. 313. ves sparsely glandular underneath. Rachis pubescent; bracts more loosely denticulate.”’ ik: Roente (village of Bemba); Molusda (Mildbrecd). Macaranga Schweinfurthii Pax Schweinfurthii Pax, 1894, Engler’s Bot. Jahrb., XIX, p.92. Ta. Duranp ip H. Scurnz, 1896, Mém. Couronnés Ac. Roy. Belgique, LIII, 4, p. 246. Tx. p H. Duranp, 1909, ‘Sylloge Flor. Congol.,’ p. 496. Pratn, 1912, in Thiselton- yer, ‘Flora of Tropical Africa,’ VI, 1, p. 935. Pax, 1914, ‘Das Pflanzenreich,’ pt 147, VII, p. 313. nga rosea Pax, 1899, Engler’s Bot. Jahrb., XXVI, p. 328. £. De Witpeman, “Btudes Flore Bas- et Moyen-Congo,’ Il, 3, p. 283. Ta. anp H. Duranp, kp Flor. Congol.,’ p. 496. Prawn, 1912, in Thiselton-Dyer, ‘Flora of Africa,’ VI, 1, P. 935. Muitpprarp, 1912, ‘Wiss. Ergebn. Deutsch. ta Lacomtei Burizx, 1908, Bull. Soe. Bot. France, LV, Mém., 8, p. 78. ga calophylla Pax, 1909, Engler’s Bot. Jahrb., XLII, p- 221. shallowly 3-lobed; lobes a. or triangular, sate: i Giaielie’ irveenutasty _ base narrowly deep-cordate; 6 to 18 in. long, nearly as Boris membranous, g firmer with age, gland-dotted beneath, with a pair of glandular processes in. long. Male flowers in lateral panicles fascicled in the axils of fallen leaven: “ts concave, thick, entire, subtending many flowers. Stamens 2 to 5 (usually 3). > flowers i in short lateral racemes; eerd short, stout. a ovate, obtuse. French Sudan: Desbasda i in the Boro Valley (Chevalier). . Southern Nigeria: Oban (Talbot). Cameroon: Tibati; Songalong (Ledermann); Bipindi (Zenker); Johann- ; Epprenbtehbhe (Biisgen). _ ¥Dhe synonymy accepted here is from Pax's recent monograph of the Euphorbiacew (1914). 412 Bulletin American Museum of Natural History {Vol. XLV French Congo: Brazzaville (Chevalier). Belgian Congo: Upper Congo: Eala (Pynaert); Ikenge (Huyghe). Kasai: Mukenge (Pogge). North-eastern Congo: Beni (Mildbraed); Mangbetu Country at Munza’s (Schweinfurth). Angola: Bamba (Monteiro); Ambriz (Welwitsch). Anglo-Egyptian Sudan: Niam Niam Country near the river Diagbe and near the river Djur (Schweinfurth). Uganda: very common (Scott-Elliot; Bagshawe). Prain still thinks that M. rosea differs from M. Schweinfurthii in the shape of the basal sinus of the leaf, which is open in the former and narrow in the latter; Pax (1914), however, believes that both are forms of the same species. ; According to Pax (1914), the stipules of M. Schweinfurthii are per- sistent, 3 to 5 em. long, 2 to 3 em. broad, slightly saccate at the base, obliquely inserted, acuminate, membranous, glabrous, shortly connate at the base. It is quite possible that, although much less pouch-like than in M. saccifera, they may occasionally be occupied by ants, though this has never been observed. ; Eco.ocy or Macaranga saccifera This species is one of the cémmon elements of the undergrowth in the Rain Forest of the Congo, in the eastern district of which I frequently observed it. Growing preferably in low-lying, rather swampy portions of primary forest, it is often found along the banks of rivers or at the edge of brooks. All the specimens I saw were low bushes, rarely over three feet high and generally smaller. Since, however, none of them were in flower or with fruit, they may have been juvenile or dwarfed. The very con- spicuous, persistent stipules, placed in pairs at the base of the petiole, are always much swollen, saccate or flask-shaped, about 1 to 3 em. long, and end in a curved, acuminate apex (Fig. 83). Their texture is more or less coriaceous and hispid hairs are scattered over the outer surface. In each stipule the free, lateral margins are curved close to each other, leaving a very narrow, upper slit as entrance to the pouch. At the foot of the leaf-blade occur two folds, one on either side of the petiole, covering nectaries which I have seen visited by ants. De Wildeman formerly supposed that these folds might shelter mites or even be myrmecodo- matia, but I doubt whether such is the case. That the stipular pouches of Macaranga saccifera were occasionally used by ants was merely surmised by Pax on account of the analogy of these organs with similar structures of other myrmecophytes. E. Laurent, however, found ants inside the pouches of the specimens which he collected in the Kasai District, Belgian Congo, and this observation Wheeler, Ants of the Belgian Congo 413 I : $3. _ Macaranga saccifera Pax: a, portion of branch with pouch-like stipules; b, extremity of ach with stipules and a leaf seen from above showing the two folds at base of blade. About one- at size (after De Wildeman, 1905). _M. saccifera has been cultivated for several years at the Brussels Garden, where it still continues to produce its saccate pouches gh these are never utilized by ants. n only one occasion have I found ants inside the stipular pouches is plant. Near the village of Mosekowa, between Walikale and utu, in January 1915, a few specimens of Crematogaster (Atopogyne) ana subspecies tibialis Santschi occupied some of the stipules. In | case the upper slit leading inside the pouches was not closed with s or carton, and no coccids were found with the ants. Since no ung or pup# were present, these pouches can not be regarded as the nest of the ants, but merely as temporary shelters or annexes. In opinion, this plant belongs to a very primitive stage of myrmecophily, n compared with some of the other African ant-plants. For this very ason, however, its relations to the ants deserve to be more fully 414 Bulletin American Museum of Natural History (Vol. XLV Sterculiacezs Cota Schott and Endlicher Cola Scuorr aANp Enpiicuer, 1832, ‘Meletemata Botanica,’ p. 33. Masters, 1868, in Oliver, ‘Flora of Tropical Africa,’ I, p. 220. K. Scuumann, 1895, in Engler and Prantl, ‘Die Natiirl. Pflanzenfam.,’ III, pt. 6, p. 99; 1900, ‘Sterculiaces Africane,’ in Engler, ‘Monogr. Afrik. Pflanzenfam.,’ V, p. 110.1 Edwardia Rarixesqur, 1814, Spechio delle Scienze, Palermo, I, p. 158 (not Edwardsia Salisbury, 1808). Lunanea A. pp CANDOLLE, 1825, ‘Prodromus Regn. Veget.,’ IL, p. 92. Trees, shrubs or bushes. Leaves entire or lobed, often polymorph, rarely digitate; glabrous, hairy or sealy. Flowers in axillary panicles or clusters, sometimes out of the old wood. Flowers through abortion unisexual or polygamous. Calyx cup-shaped or campanulate, 4- or 5-cleft. Petals absent. Staminal column sometimes very short, bearing at the top a ring of 10 to 12 anthers, disposed in one or two, regular rows; anther-cells (theew) parallel or superposed. Ovary 3- to 10-celled, with as many styles as cells. Ovules numerous in each cell. Fruit of 4 or 5 leathery or woody oblong carpels, ultimately splitting lengthwise. Seeds numerous, obovoid, exalbuminous; cotyledons‘thick, sometimes deeply bifid; radicle next to the hilum. (After K. Schu- mann.) This large genus is restricted to the continental part of the Ethiopian Region. About one hundred species have been described, most of which grow in that portion of Africa defined by Engler as the “ Western Forest Province’ and twenty-five of them occur in the Belgian Congo. A few very closely allied forms possess at the base of the leaf-blade a pair of small pouches which are occasionally inhabited by ants. In addition, these myrmecophilous species differ from their relatives in having branches and leaves covered with numerous long, stiff, erect hairs of a brown or brownish red color; the other members of the genus being glabrous. Cola Deweviei Do Wikleman and Dorand Cola Dewevrei E. De WitpeMaN anv Tx. Duranp, 1899, Bull. Soc. Bot. Belgique, XXXVIII, 2, p. 184; 1901, ‘Reliquiwe Dewevreane,’ p. 24. E. De Witpeman, 1907, ‘Mission Emile Laurent,’ fase. 4, p. 406, Pl. cxxvit. Tx. anp H. Duranp, 1909, ‘Sylloge Flor. Congol.,’ p. 62. E. De Wiiveman, 1920, ‘Mission de Briey au Mayumbe,’ p. 191. “A low shrub, 1 to2 m. high. Branches hollow, terete, with long pilosity. Leaves trilobate, deeply cordate at the base, abruptly acuminate at the apex, shiny on upper and under sides, pilose, especially on the veins. Secondary veins arcuately anastomos- ing toward the margin and uniting with the reticulate finer venation, a little promi- nent above, more strongly so below. Petiole very long, more or less grooved above, with long pilosity, 6.to 32 cm. long. Leaves 13 to 25 cm. long and 15 to 24 em. broad. Stipules linear-lanceolate, dropping, pilose, about 6 to 11 mm. long and 1 to 1.5 mm. broad. Flowers yellow, fasciculate, axillary, subsessile, bracteate. Calyx 1According to the rules of botanical nomenclature the name Cola should be replaced, it seems, either by Edwardia or by Lunanea, since the latter two are not preoccupied and evidently have priority. Wheeler, Ants of the Belgian Congo 415 lobe tomentose externally, brown and less pilose inside, about 11 mm. long; its lobes 2 to 3 mm. long, acute, with more or less reflexed tips. Androecium of the male flowers stipitate, smaller than the calyx, with subglabrous stipe, 4 mm. ong; the anthers placed close together, parallel and united into a ring which is about ‘15 mm. high” (De Wildeman and Durand, 1899). Belgian Congo: Mayombe: Lemba River (Dewévre). » - It would seem from the descriptions that the later C. Laurentii De i. ‘Wideman and the earlier C. marsupium K. Schumann are not ae 4 oa basal lobes almost touching each other. In these Cole, however, _ the form of the leaves varies to such an extent even on the same plant that this character is by itself unsatisfactory for the distinction of the ‘species. The existence of foliar pouches is not mentioned in the original description of C. Dewevrei, but De Wildeman’s figures of that species ‘published i in 1907 show them distinctly. Cola Laurentii De Wildeman Cola Laurent &. De Witpeman, 1907, ‘Mission Emile Laurent,’ fase. 4, p. 403, fig. ss 68, Pls. cxxxv, cxxxvi, and cxxxvu; 1908, ‘Etudes Flore Bas- et Moyen- Congo,’ II, 3, p.304. Tu. anp H. Duranp, 1909, ‘Sylloge Flor. Congol.,’ p. 63. _ H. Kont, 1909, Natur u. Offenbarung, LV, p. 148. A. EnGixr, 1912, ‘Wiss. __ Ergebn. Deutsch. Zentr. Afr. Exp. (1907-08),’ II, p. 506. £. De Witpeman, 1919, Bull. Jard. Bot. Bruxelles, V, p. 358. “A small tree with cylindric branches densely eee with brownish, elongate, a. iinading hairs which drop late. Leaves with more or less lengthened petioles, which are cylindrical, hispid with spreading hairs, 5 to 35 em. long. Leaf-blade 3-lobed or nearly 5-lobed, cordate at the base; the midlobe about two thirds the length of the leaf, which varies from 11 to 36 cm.; the midlobe is oblong, rather suddenly acuminate at the tip, acute; lateral lobes about of the same shape, a little shorter and narrower than the terminal lobe, which reaches a length of 23 em. and a width of 13.5em. Leaf- blade paler on the under than the upper side or about the same color, with 7 basal veins, the lateral ones often united at the base. Leaf-blade coriaceous, glabrous, ex- _eept on the veins of both sides, especially on those of the under side which are very _ prominent and bristling with stiff hairs. Between the midrib and the first lateral vein on each side of it there is a small pouch strongly projecting on the upper side; the two veins between which this pouch is formed are united at the base by a plate of tissue. Stipules filiform, hispid, rather dropping, about 2 em. long, acute. Flowers __ faseiculate at the axils of the leaves; the rachis about 1 em. long; the bracts linear, __ aeurninate, hispid, about 2 mm. long; the pedicels villose, a little over 1 mm. long. Calyx campanulate, about 8 mm. long; with 4 to 5 lobes one-third the length of the tube; calyx densely villose, brownish on the outer side, with more scattered hairs internally. Male flowers with an uniseriate andraecium, composed of theciw a little over 2 mm. long, borne on a slender, feebly elevated androgynophore which is 3 to 4 416 Bulletin American Museum of Natural History [Vol. XLV mm. long and shorter than the calyx-tube. Female flower with a densely villose, ovoid ovary; the style shorter than the ovary, with spreading stigmata which are as long as the calyx-tube. Staminodes reduced, surrounding the base of the ovary. Fruits red, 5 to 6 em. long including the acumen, with 4 to 5 seeds’? (De Wildeman, 1907). Belgian Congo: Lower Congo: Sabuka (M. Laurent); between Boma and Yanga (R. Verschueren). Kasai: Dibele; Kondué (Em. and M. Laurent) ; forest of the Sankuru (Luja). Upper Congo: Eala; Yakusu (Em. and M. Laurent); Yam- binga (M. Laurent); Dundusana (F. Reygaert); Barumbu (November 3, 1913; - J. Bequaert; Coll. No. 1081). Eastern Congo Forest: Yambuya (M. Laurent); Basoko (Em. and M. Laurent); Fariala between Mawambi and Avakubi (f. integri- folia; Mildbraed); between Lubutu and Kirundu (village of Uchibango, February 1, 1915; J. Bequaert; Coll. No. 6790); Stanleyville (March 1915; H. Lang). De Wildeman classified as ‘‘form intermedia” plants of this species in which entire and trilobed leaves are found on the same branch together with all intermediate shapes; his “form integrifolia’’ includes speci- mens in which all the leaves are entire, ovate-cordate at the base and as much as 35 em. long and 18 em. broad; in this last form pouches are also feebly developed along the midrib in the axils of the first or first and second lateral veins, above the large basal pouches. In recording the form integrifolia, Engler (1912, p. 506) also men- tions that in his opinion C. Laurentzi is not sPponnaty. distinet from C. marsupium, ‘ Cola marsupium K. Schumann Cola marsupium K. ScH#uMann, 1891, Ber. Deutsch. Bot. Ges., LX, pp. 68-70; 1900, ‘Sterculiacew Africans,’ in Engler, ‘Monogr. Afrik. Pflanzenfam.,’ V, p. 113, Pl. XII, figs. A-D. H. Konx, 1909, Natur u. Offenbarung, LV, p. 148. “A shrub or tree,.with slender, terete branches, the younger ones flattened and strongly hispid, later glabrescent. Leaves with long, terete, hispid petioles; oblong or obovate-oblong, shortly and very sharply acuminate, cordate at the base; with 7 or even 9 veins; provided with a pair of pouches forming basal swollen domatia between the midrib and the lateral veins; covered with rather long hairs on the veins on both sides, rather rigidly herbaceous. Stipules filiform, hispid, persistent for a long time. Flowers short pedicellate, axillary, fasciculate, placed either at the extremity of branches which are rather sparsely leaved below or on leafless branches. Bracts and bracteoles linear, acuminate, hispid. Calyx campanulate; its upper third split into 4 or 5 ovate, acute lobes; tomentose outside, papillate inside. Male flower: andra- cium uniseriate of 16 to 20 thece, raised on a gracile, glabrous column. Female flower: ovary subglobose, pentamerous,tomentose; the style glabrous, straight, 5-lobed; 8 ovules in each cell; follicles short stipitate, fusiform. “The shrub reaches a height of 1 to 2.5 m.; the tree as much as 10 m. The foliate, flowering branches are 3 to 3.5 mm. thick at the base and 20 to 25 em. long: they are rough, being covered with simple, spreading, brown red or brown hairs, which are thickened into a tubercle at the base. The petiole is 1.5 to 15 em. long and covered with the same pile. The blade has a length of 6 to 30 cm. and a width of 3 to 13 em. ie ie ii ee Wheeler, Ants of the Belgian Congo 417 above the in addition to the basal veins, it is crossed on each side of the mid- ib’ Hin 6 stxmger veins, which ai0s little more promincat on the under side, as is also the reticulate venation; sometimes the blade is somewhat gibbous; in life it is dark green, brownish green when dry. The basal pouches can be entered from the under side; they are not always present, but usually found on the larger leaves. The stipules are 1 to 1.5 cm. long and covered with brown hair. The bracts of the flowers are usually somewhat broader than the stipules, but otherwise similar. The # eee wove calyx i 5 t0.7 mm. long. Male flower: andrcecium 1.5 mm. long, as well as the androgynophore. Female flower: calyx slightly larger; ovary 6.5 mm. long, surrounded at the base by a ring of staminodes 2 mm. high. The fruit is red, but _ perhaps not entirely ripe” (K. Schumann, 1900). Cameroon: Abo (Buchholz); Johann-Albrechtshéhe; in the Senge Mountains “tia > i Congo: Maveli Mountains near the Sibange Farm (Dinklage). ___ It seems probable that the three forms described above, C. Dewevre, _C. Laurentii, and C. marsupium, all belong to one species, for which the “name C. marsupium K. Schumann should be retained. This is, however, a question to be decided by botanists and, in order to avoid any possible confusion, I have here used the name C. Laurentii for the plants observed by me in the Belgian Congo, because the description of that species fits them most nearly. < Eco.oey or Cola Laurentii This plant is rather common in the Congo Basin, where it prefers the drier, more elevated parts of the primitive Rain Forest. It usually _ grows as a shrub of moderate size (1 to 2.5 m. high), more rarely as a small tree (as much as 10 m. high) and flowers in both forms. The leaves are, as mentioned above, of variable size and shape, usually elongate- oval, with cordate base; the margin may be entire, or slightly or deeply lobate. The pair of basal, elongate-oval pouches on the leaves are more or less developed; wholly absent in certain cases, in others they may attain 15 mm. in length and 5 mm. in width; on the average they are 4.5 to 9 mm. long, 1.5 to 4 mm. broad and 6 mm. high. Placed at the base of the blade close to the midrib, they project on the upper side of the leaf and on the under side have a narrow slit their entire length. The general aspect of C. Laurentii is illustrated on Plate X XVII, Figure 2, by a photograph of a branch, with flowers and fruit, made by Mr. H. Lang at Stanleyville, while the shape of the myrmecodomatia is ___ seen in Text Figure 84. As mentioned by Em. Laurent (De Wildeman, 1907, p. 405), the pouches are only occasionally occupied by ants. They ____ were empty on most of the many plants which I examined. On one ocea- sion, near the village of Uchibango, between Lubutu and Kirundu (February 1915), ants belonging to the dolichoderine Engramma kohli 418 Bulletin American Museum of Natural History [Vol. XLV Forel were found inside the pouches; they had closed the slit at the under side with vegetable detritus. Unidentified ants were also found in such swellings at Barumbu (November 1913). Some of the plants col- lected by Mr. H. Lang along the Tshopo River near Stanleyville, in March 1915, were inhabited by Plagiolepis mediorufa (Forel), an ant originally described from specimens taken by Father Kohl in a nearby locality from an unidentified myrmecophilous plant. Engramma kohili, Fig. 84. Cola Laurentii De Wildeman. Lower part of a leaf seen from above, with the two pouches at base of blade. Drawn from life at Barumbu, November 1913. About natural size. like certain other members of the genus, is a frequent inhabitant of various myrmecophytes. Both Engramma and Plagiolepis are so timid and small that they could not well act as body-guards to their host plant. Although K. Schumann (1891, pp. 68-70) describes the ascidia of Cola marsupium very fully and regards them as myrmecodomatia, he was unable to find ants on his herbarium specimens from Gaboon and Cameroon. H. Kohl (1909, p. 148) is inclined to believe that, on account of the small size, these pouches are not adapted to the use of ants but serve better as shelters for coccids or plant lice. Wheeler, Ants of the Belgian Congo 419 ~~ ScapHoreTatum Masters phopet. Masters, 1867, Journ. Linn. Soc. London, Botany, X, p. 27; 1868, : in Oliver, ‘Flora of Tropical Africa,’ I, P. 236. K. Scuumann, 1895, in Engler and Prantl, ‘Die Natiirl. Pflanzenfam.,’ IIT, pt. 6, p. 86; 1900, ‘Sterculiacer Africans,’ in Engler, ‘Monogr. Afrik. Pflanstnfam: ,” V, p. 90. Shrubs with petiolate, oblong, entire, unicostate, gious leaves. Flowers yellow re pedunculate, placed in cymes or clusters, axillary or emerging from @ trunk or older branches. Calyx of 5 valvate sepals, more or less coherent, some- nes forming a 2-valved calyx. Petals 5, hooded, nervose-striate, without append- sorlaminz. Filaments united into an angular, funnel-shaped, membranous tube, _ which bears at the upper margin 5 roundish reflexed staminodes alternating with the sepals; anthers in phalanges of 3 between the staminodes, more or less concealed within the concavity of the petals, 2-lobed and 5-celled; ovules either numerous, : uged in two rows on the inner angle of each cell, or few, two or four above one another in one or two rows. Styles connate. Stigma obsoletely 5-lobed, capitate. ‘Fruita capsule with feebly fleshy walls, loculicidal. Seed, as far as known, with a curled % arillus. Cotyledons flattened, foliaceous in the albumen. This small genus contains eight described species and is peculiar to the Western Forest Province of the Ethiopian Region, from Cameroon and Gaboon to the Upper Congo. The two myrmecophytic species are the only members of the genus which have thus far been recorded from the Belgian Congo. They have been placed by K. Schumann in a sec- tion of their own, whose characters are as follows. Section Physcophyllum K. Schumann. Ovules few in number, from 2 to 4 in each _ cell of the ovary, placed in one or two rows. Upper side of the leaves with an elongate, spindle-shaped pouch at the base of the blade on one side of the midrib. Flowers small; calyx-lobes with soft hair (after De Wildeman and Dutand). Contains only two species which have been separated thus: Leaves oblong-lanceolate. Fruit not as high as broad, with 5 very distinct cells, which __ are rounded at the back and end in a pointed tip; two seeds in each cell. S. Thonneri De Wildeman and Durand. ee ch owate, narrowed towards the base. Fruit higher than broad, with 5 very distinct cells, which are subangulate at the back and taper gradually towards an erect, feebly pointed apex; four seeds in each cell. S. Dewevrei De Wildeman and Durand. Scaphopetalum Dewevrei De Wildeman and Durand Seaphopetalum Dewevrei . De WitpeMaN anv Tu. Duranp, 1901, Bull. Soc. Bot. Belgique, XXXIX, 2, p. 97; 1901, ‘Reliquie Dewevreanz,’ p. 26. BE. Dr _- Wiipeman, 1905, ‘Etudes Flore Bas- et Moyen-Congo,’ I, p. 167. Tx. ANp H. _ Duranp, 1909, ‘Sylloge Flor. Congol.,’ p. 66. H. Kont, 1909, Natur u. Offen- __ barung, LV, p. 111. Scaphopelalum Thonneri E. De Witveman anv Tu. Duranp, 1898, ‘Illustrations Flore du Congo,’ I, pp. 13-14 (pro parte), Pl. vu; 1900, ‘Plante Thonneriane Congolenses,’ p. 26 (pro parte), Pl. x1x. 420 Bulletin American Museum of Natural History {Vol. XLV “Shrub about 2 m. high, covered with brown pile towards the apex of the branches, the stem otherwise glabrous. Leaves alternate, obovate, subcordate at the base, abruptly and sharply acuminate at the apex, entire; greenish-gray above, greenish-brown below, subcoriaceous, not shiny above; glabrous or sparsely pilose near the veins; 14 to 20 em. long, 4.5 to 6.5 em. broad beneath the apex, about 2 em. broad near the base; petiolate, the petiole 5 to 6 mm. long, thick, silky. Leaves asymmetric, unilaterally constricted towards the base, which bears on the upper side a small pouch acuminate towards its tip and opening below in the axil of the penulti- mate lateral vein. On each side of the midrib there are about 8 lateral veins, project- ing slightly on the upper, more strongly on the under side and arcuately anastomosing before the margin; « conspicuous, dense network of anastomosing venules. The basal lateral vein and the midrib nearly meet on one side of the leaf, and unite by a secon- dary vein, enclosing thus the opening of the pouch. Stipules subulate, more or less persisting, 5 to 12 mm. long, fasciculate, each cluster 10 to 20 mm. long, branched, axillary, pedicellate, the pedicel 5 to 7 mm. long, bracteate; the bracts subulate, ciliate. Sepals 5, free almost to the base, oblong, velutinous externally, more or less keeled, with three veins. Petals 5, subequal with the sepals, oblong, with recurved apex, hood-shaped, longitudinally striate. Tube of the stamens membranous, penta- gonal, salver-shaped, with 5 fertile edges, the intervening lobes sterile; the fertile lobes opposite the petals. Petals covering the stamens in the bud; theew six for each phalange; the thecw subsessile; sterile lobes brietly tridentate, the median tooth obtuse, the lateral teeth narrow. Ovary oblong, 5-celled. Style entire, erect or slightly curved at the apex. Fruit red, stellate, 7 mm. long and about 3.5 mm. broad, with a prominent, horned apex; it is divided into 5 distinct cells, each of which contains 4 seeds, inserted on a central placenta. “ Differs from S. Thonneri in the leaves and fruit, and in the number of ovules or seeds contained in each of the cells of the ovary or fruit’’ (De Wildeman and Durand, 1901). Belgian Congo: Eastern Congo Forest: forest at Matchacha (Dewévre).' Kasai: Kondué (Luja). Deweévre wrote the following field-notes for his specimens: “calyx green; corolla orange-yellow with red stripes; coronula (or stamen- tube) with red edges; anthers brown; the leaves have at the base a fold inhabited by numerous red-brown ants with black abdomen.’ Luja found Engramma luje Forel in the pouches of S. Dewevrei at Kondué. Scaphopetalum Thonneri De Wildeman and Durand Scaphopetalum Thonneri E. De WitpemMan anv Tu. Duranp, 1897, Bulletin de Herbier Boissier, V, June, p. 521, Pl. xx1; 1901, Bull. Soc. Bot. Belgique, XXXIX, 2, p. 96. BE. De Wiipeman, 1907, ‘Mission Emile Laurent,’ fase. 4, pp. 400-402, figs. 66 and 67, Pls. cxii and cxii; 1907, ibid., fase. 5, p. cexxiv, fig. xu. Tu. ann H. Duranp, 1909, ‘Sylloge Flor. Congol.,’ p. 66. H. Kout, 1909, Natur u. Offenbarung, LV, p. 110, Pl. 1. A. ENeuer, 1912, ‘Wiss. Ergebn. Deutsch. Zentr. Afr. Exp. (1907-08),’ Il, p. 503. E. De Wiroeman, 1919, Bull. Jard. Bot. Bruxelles, V, p. 356. judge from Dewévre's itinerary, this locality is situated on the banks of the Congo River (L icehes between Ponthierville and Nyangwe; I have not found it on any map. Wheeler, ante of the Belgian Congo 421 petal physea K. ScuUMANN, 1897, in Engler and Prantl, ‘Die Natirl. ofa. Nachtriige 2. I-IV, October, p. 247, without description ; 1900, ; Africans,’ in Engler, ‘Monogr. Afrik. Pflanzenfam.,’ V, p. 93, vodlgeg A, a-i. ~ Stipules subulate, with well-marked, parallel vein, sometimes deciduous. ities of the branches covered with erect brown hairs. Leaves alternate, oblong, ite, entire, green, shiny on the upper face, coriaceous, glabrous except along , on the upper face, where the pilosity of the petiole extends upwards, but is listinct. Leaves 9 to 23 cm. long, 2.5 to 7.5 em. wide, narrowed towards the base, = +; the right half of the leaves placed on the right side of the branch, and the f on the left side. Lateral veins pennate. Upper face of the leaves showing on »west half a basal pouch-shaped fold which opens on the under side near the “* Veins of the under side asymmetric; the first lateral vein on either the right 1 aol half, instead of leaving the midrib almost at a right angle, makes a very an jeand, at 20 to 25 mm. from its base, unites with the second lateral vein close ib by means of a veinlet. The opening of the pouch is placed between the biand this first lateral vein, and is thus not a pore. Flowers small, about 5 mm. in branched, more or less dichotomous clusters which are erect, 10 to 20 mm. long, «din the axils of leaves, which usually hide them, the leaves often covering part of the branch with their broadened base. Flower peduncles 5 to 6 mm. long, with small, subulate bracts. Buds elliptic-ovoidal, with 5 distinct ribs ending at the apex obtuse tips and narrowing towards the base. Sepals 5, free almost to the base, ong, villose externally, more or less keeled, with 3 well-marked veins. Petals 5, en, about as long as the sepals, oblong, obtuse, recurved and hood-shaped towards heir apex, distinctly striate. Tube of the stamens membranous, pentagonal, divided 9 10 segments; the 5 edges are fertile, the intermediate segments sterile. Fertile lobes covered in the bud by the petals, bearing outside the apex 6 thece. Sterile lobes somewhat recurved towards the apex and with a small horn on each side of their ‘median portion, near the anthers. Ovary oblong, with 5 feebly marked lobes, 5- ‘lied. Style simple, straight or feebly recurved towards the apex. Fruit red” (De deman and Durand, 1897). Additional characters of importance are found in the fruit. which in this species about 1 em. long and distinctly broader than high; its 5 lobes are rounded on the back and distinctly apiculate at the apex: each of the 5 cells contains two seeds. _ Cameroon: Bipindi; Undua (Zenker); Lolodorf (Staudt); between Kribi and gumba (Dinklage). Belgian Congo: Upper Congo: Bobi near Gali, type locality (Thonner); Kapin- ; Thaka; Yakusu (Emile Laurent); Barumbu (November 1, 1913; J. Bequaert; oll. No. 1058); Dundusana (Mortehan); Mobwasa (H. Lemaire). Eastern Congo : Yambuya (November 26, 1913; J. Bequaert); Avakubi (January 1, 1914; Bequaert; Coll. No. 1919); Niapu (Lang and Chapin); Ihulu between Mawambi id Irumu (Mildbraed); near Walikale (January 1915; J. Bequaert); Mission St. Gabriel near Stanleyville (H. Kohl); Lesse (June 15, 1914; J. Bequaert; Coll. No. ___ All the specimens I have seen in the Belgian Congo agreed with the descriptions of S. Thonneri. It must, however, be noted that the shape and size of the leaves are extremely variable; it is not rare to see them a 422 Bulletin American Museum of Natural History [Vol. XLV length of 30 to 40 cm. and a width of 12 to 15 cm.; they may be gradually tapering at the apex, abruptly constricted into an acumen, or even sub- obtuse. Eco.oay or Scaphopetalum Thonneri This species is in many places a common bush of the undergrowth in the primitive, rather dry, and often very shady Rain Forest. Its stems are irregularly branched and never grow very high, usually reaching 1 to 2m., more rarely as much as 4m. The young branches are densely cov- ered with stiff, erect, brownish-red hairs, a peculiarity which is often found among myrmecophytes, though far from being the rule. While I have observed this plant in many places and at various seasons, I have but seldom seen it with flowers. These are inconspicuous, yellowish green, with the petals carmine red on the inner side. The peculiar pouch at the base of the leaf-blade is shown on Plate XXVII, Figure 1, from a photograph taken by Mr. Lang. Such an ascidium is present on all the leaves of the plant, though its size is vari- able. As a rule, it consists of a very elongate, club-shaped evagination of the blade on the upper side, laterally near the midrib, and opens on the under side by a narrow slit its entire length. This pouch may be 25. to 50 mm. long and is very narrow in the distal half or two-thirds; nearer the base of the leaf it swells rather suddenly and reaches a width of 6 to8mm. The slit on the lower surface of the pouch is placed between the midrib and the first lateral vein, which, on that side of the leaf, is de- flected from its normal, oblique course and runs close to and parallel with the midrib the whole length of the slit. Furthermore, at the distal end of the opening the deflected lateral vein is connected with the midrib. by means of a short cross-vein. As a result of this peculiar structure, the base of the leaf becomes asymmetric, the pouch-bearing side being usually much narrower and tapering more gradually towards the petiole, while the opposite side expands into a broad, semi-cordate lobe which covers. the branch. The leaves are apparently arranged alternately in two rows and are more or less horizontal, nearly in one plane. When a branch is: seen from above with the extremity farthest from the observer, all the leaves to the right have the pouch on their right half, while those to the left have the pouch on the left half. This arrangement of the leaves and ascidia, more or less distichous in appearance, is well illustrated on the plate. In most cases the pouches of this plant are empty, but on two occa- sions, at Barumbu and Yambuya, in November 1913, I found unidenti- fied ants in them. These insects had established regular formicaries. Wheeler, Ants of the Belgian Congo 423 a, with a queen, larve, and pupe, and had even brought coccids 0 the cavities; furthermore, they had closed the slit almost completely t] Saat of brownish vegetable fibres. At Niapu, in January 1914, Mr. Lang collected two species of Engramma, E. kohli Forel and E. luje rel, from the ascidia of this Scaphopetalum. So far as recorded, the is which inhabit these pouches are small and timid; they do not emerge from their retreats when the plant is disturbed and contribute ; e or nothing to the protection of their host. The leaves of Scaphope- talum Thonneri are frequently injured by phytophagous insects, even n their pouches are occupied by ants. While drawing up the original description, De Wildeman and ‘and found a few ants in the pouches of the specimens collected Dik tainee and thus recognized the myrmecophily of this species. Emile Laurent’s short field-notes are to be found in the account of the ints he collected (De Wildeman, 1907). s Flacourtiacerz _ Only one genus of this family, Barteria, is definitely known to con- | true myrmecophytes. Certain species of other genera have been . ind in association with ants, but there is reason to believe that they been settled only by accident. The best-known cf these is the Afri- cal hema specicsum Guerke,' a common bush or small tree of the primary Rain Forest, also found in forest galleries along streams in the Savannah. On a specimen observed at Romée, near Stanleyville, 4B 11 ohl (1909, pp. 109-110) found that ‘the branches, 1 m. in length, were all hollow to within 5 cm. of their tips and inhabited by small black ants, Crematogaster excisa Mayr. Two or three apertures led into th “ ‘eavity. I did not find coccids on the inner walls of these branches, everal of which I cut open, though such were seen in the axils of the 2s where they were actively attended by the ants.’’ Kohl, however, s that this plant was only accidentally occupied by ants, an opinion a , which I am in complete agreement. I have repeatedly found this uchnerodendron growing under a variety of conditions and, though my attention was especially directed to its possible relations with ants, I never saw any of these insects inside its branches. in Engler’s Bot. Jahrb., XVIII, 1894, p. 161, Pl. rv . p. 69 and 1916, p. 408) identi this ant acl by Kohl inside Buchnerodendron 424 Bulletin American Museum of Natural History [Vol. XLV In the original description of Caloncoba Laurentii (De Wildeman and Durand)! the branches of this tree are said to be fistulose and the ' following notes are given: “C. Laurentii is myrmecophilous; the stem is hollow for a long distance and pierced with exit holes at various levels, either at the cicatrice of a leaf base or at any other point along the inter- node. There were several ants inside the specimens we saw.” Gilg, who, it seems, examined some of the type material, did not find the stems hollow nor pierced with orifices, and concluded that one of the branches had been accidentally settled by ants, probably in a former burrow of some wood-boring larva. Barteria J. D. Hooker Barteria J. D. Hooker, 1860, Journ. Linn. Soc. London, Botany, V, p. 14. Masters, 1871, in Oliver, ‘Flora of Tropical Africa,’ V, p. 510. Tree or shrubs, rarely over 20 m. high, usually much lower, with thick, horizontal branches. Leaves large, leathery, alternate, oblong or oval, subacuminate, almost entire, with short, thickened petioles. No stipules, the decurring base of the leaf forming a raised line on both sides of the stem. Flowers dichlamydeous, herma- phrodite, subsessile, arranged in dense axillary or supra-axillary tufts or rows, rarely solitary; surrounded by overlapping bracts which completely enclose the flower-bud. Calyx-tube short, deeply divided into 5 oblong-lanceolate, overlapping, white sepals, which are silky at the outer side. Petals 5, inserted on the inner edge of the calyx- tube, similar to the sepals, white. Corona duplicate, emerging from the throat of the calyx-tube; outer row membranous, jagged at the edge, about half the length of — the petals; inner row much smaller, consisting of a ring of thick, fleshy tubercles. Stamens numerous, monadelphous at the base, emerging from the base of the calyx- tube; filaments in two rows; anthers linear-oblong, introrse. Ovary sessile, globose, surmounted by a single, thick style, which terminates in a large, mushroom-shaped stigma. Ovules numerous, inserted on 3 or 4 parietal placentas. Fruit a coriaceous, ovoid, indehiscent berry; seeds ovoid, compressed, with a crustaceous, coarsely pitted testa. : The genotype, B. nigritana J. D. Hooker, was discovered by Barter at the mouth of the Niger, during the Baikie Niger Expedition (1859). The genus is strictly Ethiopian with a small number of species peculiar to the Rain Forest and extending but little beyond it into the forest galleries of the neighboring grass-lands. The area of its distribution, indicated by the interrupted line on Map, 19, falls entirely within the limits of the “Western Forest Province”’ as defined by Engler. That Barteria is thus far unknown from the forests of Upper Guinea, west of ‘Contributions Flore du Congo,’ I, 1899, p. 8 ey; meet ilg (Engler’s Bot. Jahrb., XL, 1908, p. 463), this is merely a synonym of C cinschi (live), a common bush or low tree in the forests of Cameroon, G ge Be ae Omer oo The synonymy of Laurentii and C. Welwitschii is accepted by Th. and Durand in the el Sylloge Flore Congolane,’ 1909, p. 37. I have very often observed C. Welwitechii and never seen it associated with ants, Wheeler, Ants of the Belgian Congo 425 a, is remarkable, and can hardly be ascribed to insufficient investi- 1. Nor has its guest ant, Pachysima, been recorded there, which is sting in view of the fact that Map 19 shows the known distribution e two species of that ant genus to be included within the area occu- by Barteria. This genus of pseudomyrmine ants is, indeed, almost : r c 1 to the hollow stems and swellings of various Barteriz, its only a ! known habitat being the caulinary swellings of Epitaberna myrmecia Cal mi on (Stitz, 1910, p. 131; see p. 442). It would not be amiss in whether the Harter of Uganda are also inhabited by these The species of Barteria are, together with those of Scaphopetalum, commonest and most widely distributed of African myrmecophytes. y are erect bushes or small trees with a very characteristic habitus. r all the branches are uniformly hollow throughout or some of them we hollow swellings at their base. The flowers are large and showy, th white calyx and corolla, numerous stamens and a single, entire ending in a mushroom-shaped stigma; they are enclosed in over- ag bracts and placed in oblique rows, in loose tufts, or singly, in the of the leaf or along the decurrent leaf-bases. There are undoubtedly a number of different species in the genus; how many is hard to say at present, since the published diagnoses so incomplete as hardly to permit the correct identification of speci- ns. _ Gilg (1908) recognizes four species in his recent revision of Afri- 1 Flacourtiace, but he has evidently overlooked the description of B. E. G. Baker, which is possibly identical with B. Stuhlmannii Barteria acuminata Baker | Barteria acuminata EF. G. Baxer, 1905, Journ. Linn. Soc. London, Botany, XXXVI, a “Low ‘oe or bush. Branches striate, with fine rufous pubescence, or later on b Leaves oblong or oblong-elliptic, coriaceous, almost glabrous, acuminate i the apex, attenuate at the base into the petiole. Petiole very short, thick, not - stipu decurrent. Leaf-blade with about 16 to 19 lateral veins visible on both le ond uniting arcuately before the margin; also with a reticulate venation rather _ ~pre nt on the upper face. Flowers: 1 or 2 in the axils, sessile, bracteate at the base; Tiel detnencie, closely imbricate, cupuliform, brown, shiny, ciliate along _ the ymargin. Sepals 5, ovate-oblong, acuminate, longer than the petals, coalescent at _ the base. Petals white, oblong, mucronate at the apex. Stamens numerous. Stigma very jaree, conical-globose, yellow. Fruit globose. “Species related to B. nigritana Hook. fil., but differing in the leaves being gradu- jp ally acuminate at the apex. 426 Bulletin American Museum of Natural History (Vol. XLV “Leaves 22 to 24 cm. long, 6 to 7 em. broad. Petiole about 6 to 8 mm. long, canaliculate above. Sepals 2.8 to 3 cm. long, 10 to 11 mm. broad. Anthers about 3 mm. long’ (E. G. Baker, 1905). Uganda: Musozi on the shore of Lake Victoria, type locality (Bagshawe). This is very close to Bukoba, the type locality of B. Stuhlmannii which perhaps is merely a synonym of B. acuminata. Barteria Dewevrei De Wildeman and Durand Barteria Dewevrei E. De Wu.peMAN AND Tu. Duranp, 1899, ‘Contributions Flore du Congo,’ I, p. 8; 1901, ‘Reliquie Dewevreane,’ p. 97. E. De Witpeman, 1906, ‘Mission Emile Laurent,’ fasc. 3, pp. 247-249, Pl. xxx; 1908, ‘Etudes Flore Bas- et Moyen-Congo,’ IT, p. 316. Guia, 1908, Engler’s Bot. Jahrb., XL, p. 480. H. Kout, 1909, Natur u. Offenbarung, LV, p. 108. Ts. anp H. Duranp, 1909, ‘Sylloge Flor. Congol.,’ p. 223. “Tree 5 to 6 m. high, branched, glabrous. Leaves oblong-elliptic, green above, paler underneath, brown when dried, acuminate, attenuate at the base into the petiole, which is very short, thick, blackish, not stipulate, decurrent; the blade 27 to 34 em. long and about 11 cm. broad; with about 14 nerves below and above on each side, uniting before the margin; the under side with a feebly prominent, reticulate nervation. Flowers 2 to 4 togethér, axillary, sessile, bracteate at the base, the bracts numerous and closely imbricate, cupuliform, brown, smooth. Sepals 5, white on the inner side, rufous-velutinous on the outer side, oblong, united at the base, acuminate, about 3.5 em. long and 1.5 em. broad. Petals little longer than the calyx, white, oblong-obtuse, about 3.5 em. long and 1.4 em. broad. Stamens inferior, numerous, in several rows, coalescent at the base, with white filaments and vellow anthers, about 3 mm. long. Ovary globose, green, glabrous, with a heavy style and a very large, conico-globose, 5-lobed, yellow stigma (according to Dewévre). Fruit globose, 2.5 em. broad, with three parietal placentas’’ (De Wildeman and Durand, 1899). Judging from the descriptions, this species is a near relative of B. nigritana. De Wildeman and Durand compare it with that species, and in a later publication De Wildeman (1908, p. 248) writes that B. Dewevrei is “perhaps only a variety” of B. nigritana. Gilg (1908, loc. cit.), however, says: “this species is very closely allied to B. fistulosa Mast., yet, I presume, distinct from it. The broad, thick, leathery leaves are different, as also the larger flowers, and above all is the fact that the flowers are inserted as a rule 3 or 4, rarely 5, together in the axils of the leaves.” Only known thus far from the Belgian Congo: Lower Congo: Sabuka (Em. and M. Laurent); Leopoldville (March 26 and May 19, 1915; J. Bequaert; Coll. Nos. 7173 and 7663). Kasai: Dima; cliffs of Batempa; along the Sankuru; Kondué; Bena Dibele; Olombo (Em. and M. Laurent); Bena Makima; Bombaie (Leserau- waet). Middle and Upper Congo: Bolombo; Inongo (Em. and M. Laurent); Ban- gala, type locality (Dewévre; Hens). Eastern Congo Forest; Yalutcha; Yanonge (H. Kohl). ee er Wheeler, Ants of the Belgian Congo 427 ___.- Barteria fistulosa Masters osa Masters, 1871, in Oliver, ‘Flora of Tropical Africa,’ V, p. 511. K. [ANN, 1890, Verh. Bot. Ver. Brandenburg, XX XI, 2, p. 121, footnote. A. sR, 1892, Engler’s Bot. Jahrb., XIV, p. 392. Warsura, 1894, in Engler anv Tu. Duranp, 1901, ‘Reliquiz Dewevreanz,’ p. 98 (type and variety phylla). H. Wixxier, 1906, Engler’s Bot. Jahrb., XX XVIII, pp. 259- BE. De Witveman, 1906, ‘Mission Emile Laurent,’ fase. 3, pp. 250-258, cr pand xen; 1907, ‘fitudes Flore Bas- et ig age II, 1, p. 57; {GLE zn 1910, ‘Die Pflanzenwelt Afrikas,’ La ~ 642, fig. 553B. a Giuuar Asie __E. Pique, 1910, ‘Plantes Printipalés Kisanta,’ p.81. Gre, 1913, ‘Wiss. Ergebn. peteh. Zentr. Afr. Exp. (1907-08),’ II, p. 568. &. De Witpeman, 1920, fission de Briey au Mayumbe,’ pp. 104, 203, and 255. , each encircled at the base by numerous overlapping leathery shiny chest- p oblong obtuse or boat-shaped bracts, increasing in size from below up- Flowers smaller than those of B. nigritana. Sepals and petals downy on the Baospolats, wavy at the margins. Corona and inner organs of the flower as in : d species, but smaller. Anthers apiculate. ally adnate to the branch for some distance. The manner in which the emerge from between the sides of the base of the leaf and the stem is very us” (Masters, 1871). . De Wildeman and Durand’s variety macrophylla (1901) was based on specimens larger leaves (25 to 35 cm. long; 14 to 15 cm. broad); but, as De Wildeman ob- d later, this variety cannot stand, because the shape and size of the leaves in this are extremely variable: “the normal obovate-elliptic shape, rather broadly 2 at the base, may change in terminal leaves into elongate obovate-lanceolate, long-cuneate at the base and reaching a length of 27 em. by a width of 7 em. her forms. . . broadly obovate, shortly attenuate leaves reach a length of 38 id a width of 16cm.” (‘Mission Emile Laurent,’ p. 249.) Lee to H. Winkler (op. cit., p. 260, footnote) there are two forms of B. ulosa in Cameroon: “In one of them the lateral hollow branches inhabited by the s are longer, the leaves are larger and inserted on the branch by a broader base. this form the fruits are mostly divided into four, while in the other form they often st of 5, or even 6, carpels. There was also a clear and characteristic difference the shape of the seeds; while in the first variety they are 6 to 7 mm. long, 3.5 to 4 wide and 2 mm. thick, the seeds of the other which were the same length mens- only 3 mm. in width or even less, being thus much more slender.” Fernando Po, type locality (Mann). Cameroon: Victoria (Wederbauer; Winkler); Barombi (Preuss; Staudt); di (Zenker). 428 Bulletin American Museum of Natural History [Vol. XLV Belgian Congo: Lower Congo: Tumba (Em. and M. Laurent); Kisantu o Thysville (June 4, 1915; J. Bequaert). Kwango: Madibi (Lescrauwaet). Dima; Manghe; Lomkala; Olombo (Em. and M. Laurent); Bachi-Shombe grt crauwaet). Middle and Upper Congo: Ibali; Inongo; Eala; Botuma; Bolombo (Em. and M. Laurent); Coquilhatville (Dewévre); Lake Leopold II (Body); Betutu (Bruneel); Barumbu (October 28 and November 17, 1913; J. Bequaert; Coll. Nos. 1003 and 1209). Eastern Congo forest: Stanleyville (Dewévre; February 1915, J. Bequaert and H. Lang); Romée; Yangandi; Yalutcha; Yanonge (H. Kohl); Avakubi (January 17, 1914; J. Bequaert); Penge and at many places in the forest between Penge and Irumu (February 1914; J. Bequaert; Coll. No. 2339); Moera near Beni; between Mawambi and Avakubi (Mildbraed); in the forest between Wali- kale and Lubutu (January 1915; J. Bequaert). Mr. H. Lang also photographed at Medje what is evidently this species. Mayombe: Ganda Sundi (de Briey). Winkler (1906, pp. 259-260) has published some interesting morpho- logical and ethological notes on Barteria fistulosa studied by him at the Botanical Garden of Victoria, Cameroon. One of the flowering periods,—if there be more than one—starts in March. The large white flowers are crowded together side by side on the broad base of the leaves. They seem to open with dawn and the anthesis apparently lasts a few hours only. I have not found nectar in them and never observed pollinating insects; bugs and little beetles which are often found in the flowers, have, I presume, hardly to be considered as such; nor, as it seems to me, the ants whtch inhabit the tree. The fruits ripen about 3 months after the flowering. They have the shape of a walnut, and are 3.5 to 4 em. long with a diameter of 27 to 30 mm. They are flattened on two sides at the base by pressure against one another. They have four distinct protuberances at the apex, the stump of the style being placed between the four grooves. The fruits which I picked were covered at their base by the brown, closely appressed calyx;! however, the latter apparently remains on the tree when the ripe fruit drops. The consistency of the fruits can best be compared with that of a celluloid ball. The numerous, parietal placentas are arranged on four longitudinal bands. Each seed is enveloped by an arillus-like pulp, which has an agreeable, sweet-sour flavor; the pulp of the various seeds fills the fruit with a slimy mass. The seed is flattened, of rounded-rhomboid shape, with a small umbilicus and a network of dimples on the surface. To be sure the seeds are scattered by animals, which trace the pulp. The genets which I kept in captivity preferred these to almost all fruits. I have found, on fruits still adhering to the tree, holes the size of a hazelnut or an entire half of the pericarp lacking; the seeds together with the arillus had disappeared. Traces of bites could be distinctly recognized on a number of fruit envelopes which I found at some distance from one of the trees; they certainly were not from a bird’s bill, but from teeth, probably of fruit-eating bats.2, When compressed, and consequently also when bitten, the fruits split open at the top in the form of a cross between the grooves; but they open by themselves only when rather intensively drying. The author evidently means the involucrum of bracts, not the true calyx. , 2In the African Rain Forest fruit bats undoubtedly me im t agents in scattering the seeds of many fruit-bearing trees. See the remarks on this — df - Winkler (1906, p. 236) and H. Lang and J. P. Chapin (1917, Bull. Amer. Mus. Nat. Hist., X p. 484). _C Wheeler, Ants of the Belgian Congo 429 Barteria nigritana J. D. Hooker leria nigritana J. D. Hooxer, 1860, Journ. Linn. Soc. London, Botany, V, p. 15, ‘Pl. I, figs. 1-5. : eria nigritiana Masters, 1871, in Oliver, ‘Flora of Tropical Africa,’ II, p. 510. |. Winker, 1906, Engler’s Bot. Jahrb., XXXVIII, p. 260, footnote. Gri, 1908, Engler’s Bot. Jahrb., XL, 1908, p. 479; 1913, ‘Wiss. Ergebn. Deutsch. Zentr. Afr, Exp. (1907-08),’ II, p. 568. ia Braunii Encier, 1892, Engler’s Bot. Jahrb., XIV, p. 392. Warsure, 1894, in Engler and Prantl, ‘Die Natiirl. Pflanzenfam.,’ III, pt. 6a, p. 27, fig. 2A. Eneter, 1910, ‘Die Pflanzenwelt Afrikas,’ I, 2, p. 642, fig. 553A. A small tree or shrub with stout branches, covered with rusty down, and marked ther side with a raised line continuous with the base of the leaves. Leaves coriace- is glabrous, 6 to 10 in. long, 2 to 3 in. wide, oblong, subacute, crenulate or entire; “unicostate, rounded at the base or tapering into a short, thick leaf-stalk. Stipules deciduous. Flowers large, 1 to 1}4 in. in diameter, sessile or subsessile in axillary tufts, each tuft consisting of 2 to 4 flowers, each of which is invested in a series of over- _ lapping coriaceous chestnut-colored acute or cuspidate bracts. Flower-tube very 1ort, glabrous. Sepals 5, somewhat coriaceous, oblong-lanceolate or obtuse, downy nd golden brown on the outer side, smooth and whitish within. Petals oblong, wider an the sepals and about equal to them in length, white. Stamens hypogynous or htly perigynous; filaments slender. Ovary smooth; style simple, as long as the ments and terminated by a large conical or cushion-shaped stigma. Fruit ovoid, bout the size of a pigeon’s egg, coriaceous, reddish, 1-celled, with numerous compressed ed seeds attached to parietal placentas” (Masters, 1871). _ Southern Nigeria: Nun River, type locality (Barter); Bonny River (Mann); Old Calabar (Thomson). Cameroon: Batanga (Dinklage); Kribi (Zenker). Spanish Guinea: on the coast of Bata near Campo (Busse). French Congo: on the Gaboon River near Libreville (Mildbraed). _As pointed out by Gilg this species seems to be restricted to the coastal forest belt (‘eine echte Seestrandsptlanze”) which grows inland of the ‘mangrove formation along the Gulf of Guinea. Similar patches of dense ~~ forest are to be found immediately landward to the mangroves in the estuary of the Congo, but I have never seen any Barteria there. Barteria nigritana variety uniflora De Wildeman and Durand Barteria nigritana variety uniflora £. De WitpeMaN anv Tx. Duranp, 1900, ‘Con- tributions Flore du Congo,’ II, p. 24; 1900, Bull. Herbier Boissier, (2) I, p. 22. PB. De Wireman, 1904, ‘Etudes Flore Bas- et Moyen-Congo,’ I, p. 169. H. Kout, 1909, Natur u. Offenbarung, LV, p. 109. Tu. ann H. Duranp, 1909, ‘Sylloge Flor. Congol.,’ p. 224. velantes, late ovata cuspida ter Potala cal: ulolongiora. Stami luri e © oon, ° . glaberrimum. Fructus magnit. ovi columbm."’ = pee 430 Bulletin American Museum of Natural History [Vol. XLV “A high tree, with thick branches, which are striate, ferruginous-pubescent, marked on either side with a raised line connecting the leaf-bases. Leaves alternate, shortly petiolate; the petiole 5 to 6 mm. long and 3 mm. broad; oblong, subacuminate, 11 to 17 cm. long and 3.7 to 5.5 em. wide, entire, shiny above and below, dark above, paler below, the upper side subglabrous or with a few scattered hairs; the under side with short, sparse, brown pilosity, especially on the veins; lateral nerves a little promi- nent above and below, arcuately anastomosing towards the margin and prominent in the more or less recurved margin. Flowers sessile, solitary in the axils of the leaves; at the base with closely imbricate bracts, which are scarious, brown, pilose externally, embracing. Calyx with ovate-lanceolate lobes, ferruginous pilose on the outer side, glabrescent on the inner side, acuminate, 3 cm. long and 12 mm. wide. Petals equal to the sepals but completely glabrous. Corona erect, membranaceous, fimbriate- lacerate at the apex. Stamens numerous, with connate filaments. Ovary globose, with parietal placentas, and numerous ovules; style solid; stigma very large, 6 to7 mm. broad, conico-globose’’ (De Wildeman and Durand). Belgian Congo: Lower Congo: Forest of Talavanje, type locality (Cabra); Kisantu (J. Gillet). It seems doubtful whether this form is really a variety of B. nigritana in view of its occurrence inland far from the coastal belt. It may possibly be specifically distinct or constitute a form of B. Dewevrei, a species commonly found in the Lower Congo. From the description, it appears very similar to B. Stuhlmannii. Barteria Stuhlmannii Engler and Gilg Barteria Stuhlmannii ENGLER AND Gia, 1908, Engler’s Bot. Jahrb., XL, p. 479. Barteria nigritiana W arBurG, 1895, in Engler, ‘Pflanzenwelt Ost-Afrikas,’ C, p. 278 (not nigritana Hooker). “Shrub or tree with fistulose branches, which are densely and very shortly fulvo- pilose when young. Leaves ovate or ovate-oblong, very seldom oblong; acute or often shortly and broadly acute-acuminate at the apex, rounded toward the base, though narrowed at the very base into a 6 to 8 mm. long petiole, on both sides of which there is a 3 to 4mm. wide wing; leaves obsoletely sinuate-denticulate, or more often sub- entire, with cartilaginous margin, glabrous above (except on the median nerve), - very sparsely and shortly pilose below, leathery, with 13 to 15 pairs of lateral nerves which run almost straight to near the margin where they unite by curves; with num- erous transverse nerves running parallel to each other and strongly prominent on both sides; other reticulate veins almost absent. Flowers solitary or occasionally by twos in the axils of the leaves. Bracts enclosing the flower in an involucrum, coriaceous, with scattered fulvous pilosity on the back. Outer sepals entirely covered on their outer side with dense fulvous pile, which on the back of the inner sepals forms only a median vertical line; otherwise glabrous, oblong, with very acute apex. Petals a little shorter than the sepals, but similar in shape, very tender; glabrous. Outer corona membranous, glabrous, about half the length of the petals, unevenly incised and fimbriate; inner corona much shorter, thickened, forming a raised ring which is dis- tinctly but feebly emarginate and furrowed. Stamens numerous, coalescent at the base into a tube. Ovary short, turbinate, glabrous, with 4 parietal placentas. Style elongate, reaching the anthers, thick, gradually thinner upwards, ending in a very 1922] Wheeler, Ants of the Belgian Congo 431 thick, head-shaped stigma. Fruit subglobose; the pericarp parchment-like or charta- ceous, fragile; seeds numerous, inserted on 4 parietal placentas, oblong, yellowish, “The winged petiole is 6 to 8 mm. long and, with both wings spread, 7 to 9 mm. wide. The swollen, hollow stalk is 6 to 10 mm. thick. The blade of the leaf is 16 to 19 em. long, 7 to 9 cm. broad. The bracts which enclose the base of the flowers are 7 to 9mm. longand equally wide. The outer sepals are about 2.5 cm. long, 1 em. wide; Fig. 85. a, Barteria fistulosa Masters: portion of branch with fruits along decurrent leaf base; b, Barteria Dewevrei De Wildeman and Durand: portion of branch with fruits clustered in axils of leaves (after De Wildeman, slightly modified). the inner ones decrease gradually. The petals are about 2.2 cm. long, 8 to 9 mm. broad. The style is about 1.5 cm. long, the stigma 4 mm. long and 3 mm. thick at the base. The fruit has a diameter of about 2.5 cm. The seeds are 5 mm. long, 3 mm. wide and 1.5 mm. thick” (Engler and Gilg, 1908). German East Africa: Bukoba, type locality (Stuhlmann). This plant will, I believe, prove to be identical with B. acuminata E. G. Baker (see above, p. 425), described from Musozi on Lake Victoria, which is practically the same locality as Bukoba. During my travels in the Belgian Congo, I came across two species of Barteria, B. Dewevrei and B. fistulosa. The latter is by far the more common and can be best recognized by the very peculiar way in which it 432 Bulletin American Museum of Natural History [Vol. XLV grows, by the basal swellings on some of its horizontal branches, and especially by its inflorescence. In this species a number of flowers or fruits are placed close together in a row on either side of the petiole from the axil along the decurrent base of the leaf (Fig. 85a). In Barteria Dewevrei, however, the flowers or fruits occur either singly in the axils of the leaves or two to four together in loose axillary clusters (Fig. 85d). The anatomical structure of the two species also shows certain differ- ences, which have been pointed out by Prof. Bailey (Part V, p. 599). Eco.oey or Barteria fistulosa Perhaps the most striking’ of Congo myrmecophytes, this plant occurs throughout the entire Rain Forest belt, where it favors the higher, more open spots, being rarely met within damp soil. In the Savannah of the Lower Congo and Kasai District it follows the forested banks of water courses. The natives of the forest are familiar with the plant and aware of its being inhabited by very aggressive ants. Whenclearing the underbrush to establish a road or plantations, they usually avoid the un- pleasant task of cutting this small tree. Thus it happens that Barteria is frequently found standing by itself in the center or near the sides of forest paths (Pl. XXIX). For the same reason, it is often met with in secondary forest growth. Among the Wangata, at Barumbu and else- where, its vernacular name is “ Bakokombo,”’ and other Bantu tribes use similar sounding designations (Monkukono, Makonkomo, Okakumbu, etc.)! This species is a typical element of the undergrowth of the forest. Under favorable conditions it may become a small tree, reachirty a height of from 6 to 10 meters, or in exceptional cases 20 meters or more, but it is frequently much lower, 3 to 4 meters being a common average; its trunk is, as arule, 10cm. in diameter, though there are occasional records of over 40cm. The trunk is simple or very little ramified, and bears long lateral branches, usually also unramified and spreading almost hori- zontally in all directions. The broad, alternate leaves are placed to the right and left of the branch, more or less horizontally or slightly curved upward. Due to this arrangement, the plant has a very peculiar appear- ance, well illustrated in the photographs taken by Mr. H. Lang (Pl. XXVIII, Fig. 1 and Pl. XXIX). Another unusual feature is the fact that most of the branches are deciduous. On reaching a certain length 1According to J. Gillet, the natives at Kisantu, in the Lower Congo, call Barteria fistulosa either _ “Sakala’’ or ‘ Nsakala,” and de Briey gives ‘‘Zinzi” as its vernacular name in the region of Ganda Sundi. 1922) Wheeler, Ants of the Belgian Congo 433 they stop growing, lose their foliage, and gradually dry up; finally, these dead members are dropped by a histological process similar to that caus- ing the leaves to fall. One always finds, therefore, a number of dead branches scattered over the ground at the base of this Barteria. Whether ___ there is a law of periodicity or other rule governing this peculiarity can- not be decided at present, but so much is sure: the few flowering branches remain on the stem until after the fruits are ripe. _ The lateral branches of Barteria fistulosa are of two kinds. The ____ Sterile branches—and, as noted, these are in the great majority—present at a short distance from their base an abrupt and conspicuous swelling ___ which continues almost uniformly to near the apex with only slight con- ___ strictionsat thenodes (Plate XXVIII, Fig. 2; Text Fig. 86). Except in Fig. 86. Barteria fistulosa Masters: a, lateral branch suddenly swollen beyond its base, where it idiows the circular opening (¢) gnawed by the ants and leading into the domatium; about one-half natural size (after De Wildeman); 6, part of longitudinal section of very young swollen lateral branch, - showing gradual drying up of pith before ants gain access to the cavity; drawn from life at Berumbu, October 1913; natural size. very young plants, these swellings are nearly always hollow and inhab- ited by ants. The flowering branches appear only at certain seasons, and on older trees; they are normal, not swollen, yet frequently hol- lowed out and also occupied by ants. From an examination of very young specimens and others not in- habited by ants, I found that the trunk and normal, flowering branches are filled with pith and remain so unless excavated by the ants. The swollen branches (Fig. 86), on the contrary, become hollow naturally. 434 Bulletin American Museum of Natural History [Vol. XLV When young, their various internodes are at first only slightly swollen and entirely filled with soft, greenish, parenchyma; soon, however, the enlargement becomes more pronounced; the pith turns pale brownish, gradually dries and what remains finally forms brownish membranes on the inner walls or irregular partitions in a spacious cavity (Fig. 86b). The ant-chamber is thus ready for occupancy before the insects touch the branch. On uninhabited plants the sterile branches show no orifice, nor any depression or scar on their outer surface that might mark the spot where the entrance to the cavity will later be pierced by the ants. Moreover, the walls of the limb are soft and easily pressed down with the fingers, so that they must offer but little resistance to the powerful mandibles of the Pachysime. The larger specimens of Barteria fistulosa that one commonly meets 7 in the forest are, as a rule, settled by a populous colony of the large, black Pachysima xthiops (Emery),' the true body-guard of the tree. As soon as any portion of their host plant is disturbed, they rush out in numbers and hastily explore the trunk, branches, and leaves. Some of the workers usually also run over the ground about the base of the tree and attack any nearby intruder, be it animal or man. All observers agree that the sting of the Pachysima is exceedingly painful and is felt for several hours. Its effects can best be compared with those produced by female velvet ants (Mutillide; see Kohl’s remarks reproduced in Prof. Wheeler’s Report, vp. 115). Consequently these ants are greatly dreaded by the natives and there remains little doubt that they afford a most effective protection to their host plant. Trees inhabited by Pachysima are generally healthy and free from the attacks of most phytophagous insects. On specimens untenanted by ants, however, the leaves are often badly eaten by caterpillars, as I observed in two instances at Barumbu in October, 1913. On both of these trees there were also several nests of the weaver-ants, @cophylla longinoda (Latreille), and numerous workers of a small Crematogaster running over the branches and leaves. At Penge, in February, 1914, another uninhabited B. fistulosa showed the live wood of its trunk badly bored by adult bostrychid beetles. On the other hand, the Pachystme are not always successful in keeping smaller parasites from their host. At Barumbu a tree occupied by a populous colony of P. zthiops showed numerous cecidomyid galls on its leaves. They were small fleshy swel- lings of the parenchyma, about equally protruding on both sides of the 1Probably sometimes also by Pachysima latifrons (Emery). es ae ee, en Tee ere = 1922} Wheeler, Ants of the Belgian Congo 435 leaf, and irregularly scattered. Inside of them was a single chamber con- taining one gall-midge larva and surrounded by a wall of coarser tissue in the center of a solid, juicy, parenchymatous mass.! P An older, inhabited Barteria fistulosa may be regarded as the home of a single colony of Pachysima which has resulted either from the gradual growth of a small nest founded by one female, or from fusion of several ___ nests started independently by a number of females. Both modes are possible, but the second is probably the more common. At Avakubi, in January, 1914, 1 had an opportunity to examine a very young Barteria __ fistulosa not over one meter high, with but six short, horizontal branches, all of which were swollen beyond the base in the usual way. Only a few of the distended internodes were settled by ants and each was a closed, separate cavity containing one deiilated Pachysima queen; no workers, larve, or eggs were present. After the nuptial flight the Pachysima females had evidently entered the hollow internodes by gnawing through the wall. They had not again left the cavity, for the entrance was partly plugged up by callus growth. When disturbed, these gravid queen ants made no attempt to defend themselves, behaving in this respect very differently from workers. It is also interesting to note that some of the Pachysima females were dead and that in one such case another minute ant, of an unidentified species, had established its nest in the same in- ternode with the remains of a dead Pachysima queen. A colony of Pachy- sima xthiops in a somewhat more advanced stage was found in a young Barteria fistulosa,at Barumbu in November, 1913.. A queen ant, sur- rounded by an abundance of eggs and young larve, was found inside each of a series of swollen internodes, all still separated by the nodal partitions. Here, too, a growth of callus had partly closed the entrance which had been further plugged with dried particles of pith evidently brought there by the female. Since the older Barteria is finally occupied by one single colony, all the members of which live and work peacefully together and enter indifferently the various domatia, the initial formi- caries in all probability fuse into one. The workers in such a formicary not only enlarge the exit holes, which are usually placed at the base of the swelling toward the upper face of the branch, but also clean the cavi- ties of the remains of dried pith and pierce the partitions between the various internodes. Each lateral branch finally forms one continuous gallery. a ‘Lamborn (1914, p. 493) notes that he once found larva and pul of Tinthia lambornella Durant, an moth, in an internode of a Barteria rteria in Southern Nigeria: ity was separated f m the : an eperiid internodes, both of which were inhabited by Pachysima I dag 5 ’ ¥3 436 Bulletin American Museum of Natural History (Vol. XLV The origin and growth of new colonies of Pachysima in Barteria deserves to be further investigated in the field. Perhaps such a study will show us typical examples of secondary pleometrosis, or founding of an insect society through fusion of a number of colonies each started in- dependently by a fertile female. H. v. Ihering (1907) believes Cecropia adenopus is settled in this manner by Azteca miilleri. Furthermore, in his opinion, all but one of the fertile queens inhabiting the same tree are eventually killed by the workers, a conclusion drawn from the presence of a sole queen in each adult Cecropia. It will be important to look into conditions in this respect in the Pachysima formicaries of Barteria. — The Pachysime undoubtedly derive certain advantages from living inside Barteria. The hollow, nearly horizontal branches provide very convenient nesting chambers, where the brood is kept in safety under almost ideal conditions of aération, temperature, and humidity. Whether the ants also procure part or all of their food from the host is still doubt- ful. Kohl has often seen the workers actively licking nectaries at the insertion of the leaves, and also gnawing the young bark and the epider- mis on the upper and under sidesof the blades; they are particularly fond of the very young flowers, which they frequently destroy almost com- pletely. Certain other insects live in the domatia with the ants, the most common of these companions being coccids (Pseudococcus citri variety congoénsis Newstead) which, I am inclined to think, are not brought in by the ants, but migrate inside the swellings of their own accord. I have found this to be also the case with scale insects living in the myrmecodomatia of Cuviera. Even in very young Barteriz, of which only a few internodes are occupied by queen ants and their brood; one discovers coccids in the cavities. Another interesting inquiline of Barteria is a minute phorid fly, Hypocera tristis H. Schmitz, noticed by Father Kohl near Stanleyville in swellings of Barteria fistulosa occupied by Pachysima xthiops (Wasmann, 1915a, p. 320, footnote). . Whether the coccids of Barteria are really attended by Pachysima for the sake of their excretions remains uncertain. Wheeler and Bailey (1920, pp. 261-262) have dissected the pellets contained in the infra- buccal pockets of workers and the trophothylaces of larve of Pachysima zthiops and P. latifrons. They were much the same in both species and consisted of pieces of coccids or whole, crumpled-up bodies of young scale insects, fungus spores, bits of mycelium, portions of plant-tissue evidently gnawed from the walls of the cavities, pollen-grains, ete. In a few of the pellets Prof. Bailey found small nematodes resembling the species of Pelodera described by Janet as living both as parasites in | Tate Wheeler, Ants of the Belgian Congo 437 en once of certain European ants and as free organisms the itus of the nest. _ A thorough investigation of the feeding habits of both adult and val Pachysimz in Barteria will be the most important problem to be idied in the future. In this connection, it may be well to note a pecu- _ diar ty to which my attention was directed by my friend, Mr. J. P. Ch: apin, during our stay at Avakubi in January, 1914. When Barteria ulosa inhabited by Pachysima occurs in rather dense forest, one fre- : r ly notes about its base an open patch, fifteen to twenty feet in neter, where most of the heavy undergrowth has been cleared away. of the common marantaceous forest reed, Sarcophrynium Arnoldianum Ye Wildeman, are left standing. The ground at the foot of the tree is rtly covered with fallen leaves and dead branches of the Barteria. can always find a few Pachysima workers running over this open te, for a purpose unknown to me, perhaps in search of insects which y form part of their diet.!. I merely venture the supposition that the 's themselves are instrumental in preventing the growth of heavy ion near the base of their shelter, perhaps by nipping the tender s of the young plants.2. One can readily imagine that such a ing would be of use to the ants in their hunts for other insects, ‘ing the capture of their prey so much easier and quicker. Incident- , Barteria too may be benefited, since it is saved competition with vigorous species of trees or shrubs, which, if allowed to thrive near ts trunk, would soon interfere with its growth. The shade given by Barteria fistulosa i is so slight that this factor alone could net account for the absence of woody vegetation within a radius ot six to eight feet from me, which pay no further attention to them, one proof more of the strict, obligatory relations existing between these ants and the host plant. The iMy observations do not agree on this one point with those of Father Kohl, who believes that . 9 never pares Barteria “ since i te = vans hunting aed tag ob for Kohls. insects.’ sees): ae regard to this interesting - uote owin m Kohl's paper _ “Some natives assured me that vag Hos ew) om occasicnaly tri trim to half its bel Tat the low : which alg ry their h a I once observ = A, partly cut low plants in the of et od I did onan id not putas the ble agent St this.” ( also remarks yrma dendroica ror and P. triplaridis Forel, which inside stems and branches of Triplaris Schomburghiana alon Jurud River, Brazil, run down to or ecderoet a is allowed to grow. 438 Bulletin American Museum of Natural History (Vol. XLV species of ants. At Barumbu, in October, 1913, I came across a Barteria fistulosa whose living branches were inhabited by Pachysima zxthiops, while the dead twigs, still attached to the tree, contained small colonies of a Crematogaster. These little ants were apparently not molested by their large neighbors, but, when the tree was disturbed, they remained safely inside, while the Pachysimz rushed forth and ran feverishly over the plant. The myrmecophytic nature of Barteria fistulosa was first recognized by K. Schumann (1890, p. 121, footnote) on herbarium specimens collected in Cameroon. Some of the swellings cut open by him still con- tained a few Crematogasters. Its relations with ants were studied in the field by A. Dewévre (De Wildeman and Durand, 1901, p. 98), Emile Laurent (De Wildeman, 1906, pp. 250-258), H. Winkler (1906, p. 59), and H. Kohl (1909, pp. 97-108). Mention is made in Prof. Wheeler’s Report (p. 114) of some of these earlier observations which agree in most details with my own. The following ants have been found thus far in the swellings of Barteria fistulosa, but the two species of Pachysima alone can be regarded as obligatory guests of the plant. The others are all accidental tenants which nidify in other places also; they are usually met only on plants or — in branches which for some reason or other have been left by the Pachy- — sime. Pachysima xthiops (F. Smith). The large, black ant which is the regular inmate of Barteria fistulosa, was first collected in this plant by Father Kohl, near Stanleyville (1909, p. 106), and sent by him for identification to Forel (1916, p. 403). Both Mr. Lang and I commonly found the same species at Medje, Ambelokudi, Barumbu, Avakubi, etc.1 The scale insect Stictococcus formicarius Newstead was found by Kohl near Stanleyville with these ants (Newstead, 1910, p. 19). P. latifrons (Emery). Specimens ot this species obtained by Mr. H. Lang at Niangara were probably taken from Barteria fistulosa. Tetraponera anthracina (Santschi). Near Stanleyville (H. Kohl; see Forel, 1916, p. 403). I found several workers of this species at Thys- ville (June 1915) running over the leaves and twigs of a Barteria fistulosa whose swellings were free of ants; I did not find their nest. 1Pachysima wxthiops was originally described from South Africa, without indication of collector No species of Pachysima has since definitely been recorded from that part of the continent. Since the genus is restricted to Barteria and Myrmacia, which are not known to occur south of 7° 8. lat., there is a question whether Smith's type was wrongly labeled. It is, however, not so clear how he could have received West African specimens of P. #thiops at a time (1877) when hardly any m tions had been made in Equatorial Africa. I am rather inclined to believe that Smith's type was ob- tained in the forests of Natal from a myrmecophyte which has since escaped notice. Wheeler, Ants of the Belgian Congo 439 . i (Ern. André). Near Stanleyville (H. Kohl; see 1916, p. 403). | T. oberbecki (Forel). Leopoldville (H. Kohl; see Forel, 1916, p. 403). T. ophthalmica (Emery). Stanleyville and Bengamisa (H. Kohl; rel 1916, p. 403). I collected workers of this ant at Thysville, ogether with 7’. anthracina, as mentioned above. ga ¥: prelli (Forel) variety odiosa (Forel). Belgian Congo (H. Kohl; el, 1916, p. 403). e) excisa subspecies impressa (Emery). Discovered by Bt (1900, p. 103, footnote) in branches of Barteria fistulosa collected . Laurent at Isangi. Also near Stanleyville (H. Kohl; see Forel, , Pp. 69). a impressiceps (Mayr). Taken from hollow twigs of Barteria a by Mr. H. Lang at Panga. .striatulaEmery. Romée near Stanleyville (H. Kohl, 1909, p. 167). . Eco.oey or Barteria Dewevrei ‘This species is less common than the preceding and has been but little studied so far. The following notes were made on specimens I found near Leopoldville, in one of the small patches of forest which ‘are scattered through the savannah of that region (March, 1915) and also in a forest gallery along one of the small affluents of the Congo (May, 5). In that locality, it is a low tree, rarely over 6 meters high, with a tht, simple, or feebly ramified trunk, 20 em. thick at the base. ie alternate leaves are more elongate-elliptic than in B. fistulosa, be = g as much as 40 cm. long and 9 cm. wide. The lateral branches on m ly specimens were all alike, feebly branching and irregularly spreading, giving the tree a very different appearance from that’ of the species just fioned. The specimen collected in May was in flower. ‘There were no swellings on any of these plants from Leopoldville, all the lateral branches (Fig. 87a-b) were hollowed out almost their » length, each with one continuous cavity. Only the upper ex- us to dry up very soon, as the pith channel is hollow 6 cm. from the ‘ip. A few entrances to the inner cavities had been pierced, mostly on ie upper side of the branch, at intervals of about 6 to 14 cm. from one nother, usually a short distance above the insertion of a leaf. 7 a lOne of the specimens of B. Dewevrei from ville apewered 4 well the description given by hm. Rear enacts me aces cutee ants Cave chins B.flerciona)s on tetmches do's 'd.con, Ge dia sont hablices des fourmis et leur canal mndauilaire de min de diamétre resterait apres la de a He seen ey le limbe mesure de sur A FA aa Wuaies ds corvune inte térales" (De Wildeman, 1906, p. 250). pata mophy Ierahing but this was doubted by De Wildeman. | i | Z F Fig.87. Barteria Dewevrei De Wildeman and Durand: a, external view of portion of lateral branch inhabited by ants; 5, longitudinal section of this branch, showing a coccid (c) fixed on inner wall and small depressions (d) in which scale insects are often found; ¢, orifice leading intothedcmaticm. Drawn from life at Leopoldville, March, 1915; natural size. 440 Wheeler, Ants of the Belgian Congo 441 eubivel tha bivapediemne ct B. Dewevreit examined mntaine iindine of Crematogaster africana variety schumanni (Mayr), hs queen, workers, and brood; also some coccids which were usually a small, scar-like depression in the wall. In one tree some of the ches contained insect larve, a lepidopterous pupa, and an adult 2 but these were in cavities quite separate from those inhabited by EDewevre, who discovered this species in the Bangala region, on the yer Congo, mentions finding ants in its hollow branches (De Wilde- 1 and Durand, 1901, pp. 97-98). A few notes on its relations with ts \ re sls made by Em. Laurent (De Wildeman, 1906, pp. 247-250) nM id E Kohl (1909, pp. 108-109). The following ants have been found ts myrmecodomatia: Pe Pachysi a ethiops (F. Smith). Dima (Em. Laurent; see H. Kohl, . 108); Yalutcha and Yanonge (H. Kohl, 1909, p. 108). : Cre watogaster africana (Mayr) variety. Dima (Em. Laurent; see '. africana easiey schumanni (Mayr). Leopoldville (J. Bequaert). Apocynacesz EriTaBerna K. Schumann K. Scuumann, 1903, Engler’s Bot. Jahrb., XX XIII, 2, p. 316. Neitiedhi with branches thickened and hollow below. the nodes. Leaves large, t petiolate, lanceolate, short acuminate. Flowers diclinous, showy, axillary; r pedicel with a pair of lower bracteoles, simulating interpetiolar stipules, and also 1 a second pair of bracts below the ovary. Sepals large, foliaceous, subinequal, rnating with very large, linear, solitary glands. Corolla very large, infundibuli- n; its lobes ample, curled along the margin, their sides in the bud inflexed and _ covering each other dextrorsely; its throat with variegated hairs. Stamens inserted eal the throat, without any stiff appendage at the base, acute and not appendiculate the gag Ovary perfectly inferior, pentapterous, 2-celled; with numerous ovules 1 on a thickened placenta; dise annular; style thickened and bilobed at the “Only one species is known. “The genus is a relative of Tabernamontana, from which it differs in its com- etely inferior ovary and in the large sepals” (K. Schumann, 1903). _ This is the only genus of the large family Apocynacee which has is far been recognized as a myrmecophyte and the true nature of its ‘lations with the ants has apparently not been further investigated on living specimens. It contains only one species. 442 Bulletin American Museum of Natural History [Vol. XLV Epitaberna myrmecia K. Schumann Epitaberna myrmacia K, ScuuMann, 1903, Engler’s Bot. Jahrb., XX XIII, 2, p. 317. “Branches thickened at the nodes, quadrangular, glabrous. Leaves short petiolate, lanceolate, ample; short and sharply acuminate, acute at the base, glabrous above; slightly hairy on the under side in youth and later on with scattered pile on the midrib. Flowers short pedunculate. Ovary glabrous. Sepals lanceolate, acumi- nate, large, glabrous. Corolla with a tube extending hardly beyond the calyx; glabrous outside; densely villose at the throat on the inner side; the lobes broadly elliptic, acuminate, curled along the margin, twice the length of the tube, lanceolate in the bud. Stamens linear, enclosed. Style glabrous, filiform, thickened at the apex. “The flowering branches, 15 cm. long, are 2 mm. thick in the middle of the inter- nodes; the upper part of the internodes is swollen into a spindle-shaped cavity with thin walls, which is as much as 5 em. long and 9 mm. in diameter and serves as a myrmecodomatium. The heavy, glabrous petiole is grooved on the upper side and at most 5 mm. long. The blade has a length of 11 to 28 em. and a greater width of 5.5 to 11 em. in the middle; it is crossed on each side of the midrib by 6 to 10 stronger veins, which are prominent on both sides; in dried condition it is dark green above, pale green below. The flowers do not always present an ovary, there being male and - female flowers; but otherwise they do not differ from each other. The peduncle is 5 mm. long, and the inferior ovary about the same length. The green sepals reach a length of 2.5 cm. The tube of the white corolla, with its chrome-yellow throat, is 2.2 cm., and its lobes 5.5 cm. long. The stamens are inserted at 15 mm. above the base of the corolla; the anthers are 7 mm. long. The style measures 1.3 cm. “The plant is remarkable, representing a new case of myrmecophily. I myself have collected the ants from the wool of the throat of the corolla. This is the first case of completely epigynous flowers among the Apocynacee; accordingly the fruit is probably also syncarpous” (K. Schumann). Cameroon: Bipindi (Zenker). Epitaberna myrmecia probably occurs throughout the forest of southern Cameroon and Spanish Guinea. According to Stitz (1910, p. 131), Tessmann found inside the caulinary swellings of this plant, the large Pachysima xthiops (F. Smith) (=spininoda André) which the Pangwe call ‘‘engunkun,’’ much fearing its sting in the belief that it causes fever. Tessmann, in his account of the Pangwe of southern Cameroon and Spanish Guinea, describes how the tribe uses this myrmecophyte in one of their religious ceremonies. During the initiation to the “Sso-cult” ot the Yaunde, the candidates are obliged to pass for several days through a succession of tests, one of which is as follows. Nests of stinging ants, especially those of Plagiolepis carinata Emery, and branches of Epita- berna myrmecia inhabited by Pachysima xthiops, are hung or placed in a low hut built for that purpose near the village. This place soon swarms with ants; pods of Mucuna pruriens covered with dangerously itching hairs are also thrown inside. The neophytes are then brought there and, Wheeler, Ants of the Belgian Congo 443 Verbenacez CLERODENDRON Linnzus om Linn.=vus, 1753, ‘Species Plantarum,’ Ed. 1, I, p. 637. Briquert, 1895, in Engler and Prantl, ‘Die Natiirl. Pflanzenfam.,’ IV, pt. 3a, p. 174. J. G. Baker, 1900, in Thiselton-Dyer, ‘Flora of Tropical Africa,’ V, 2, p. 292. _ “Trees or shrubs, sometimes scandent. Leaves opposite, rarely ternately verticillate, entire or toothed. Cymes axillary or terminal, lax or dense. Flowers li or large, various in color. Calyx not accrescent; tube campanulate; lobes 5, equal, longer or shorter than the tube. Corolla-tube cylindrical; lobes 5, obovate, spreading or slightly reflexed, subequal or unequal. Stamens 4, inserted below the throat of the corolla-tube; filaments long, filiform, involute in bud; anthers ovoid ‘oblong, with parallel cells. Ovary imperfectly 4-celled; cells l-ovuled; style long, bifid at the apex. Fruit a globose drupe with a fleshy pericarp and 4 smooth or ‘Tugose pyrenes. Seed oblong, exalbuminous”’ (J. G. Baker, 1900). "This is a very large genus, numbering some 200 species and distrib- fated between the tropics in the Old World; over 150 have been described from Africa, 35 of which have been recorded from the Belgian Congo. _ They are very common at the edges of the forest and along rivers, where the creeping species often are one of the striking elements in the land- “seape, on account of their beautiful, showy flowers. The species of the savannah are most frequently low shrubs or erect or trailing herbs. A number of species of Clerodendron have been found associated a with ants, but the few published observations are too fragmentary ____ to show whether any of the forms are true myrmecophytes. Among the __ African representatives, Clerodendron excavatum E. De Wildeman? is ___ myrmecophilous according to certain observers, while others assert that _ its hollow stems are merely filled with water. At all events, ants were never found inside the stems of that plant. s At Penge, in January, 1914 (Coll. No. 2205), I collected on the bank __ of the Ituri River in the dense undergrowth of the forest a low bushy ___ Clerodendron which may possibly be C. excavatum E. De Wildeman. The plant was 3 to 4 m. high and divested of leaves at that season of the year. Some of the branches, however, were covered with numerous, white, _ showy flowers, obliquely directed downward. No swellings nor domatia could be found, but the internodes of stem and branches were normally 4 1913, ‘ II, F a ‘Deseribed’ 1900: is ' totes ooh ytd 5 IIT, 1, pp. 132-134, Pl. xx; 1912, ibid., 444 Bulletin American Museum of Natural History [Vol. XLV hollow, due to the early resorption and drying up of the pith. Many of the hollow internodes contained nests, with a fertile queen, workers, brood, and newly hatched winged sexual forms of a small, unidentified ant. The insects entered and left by a circular entrance pierced through the wall about half-way between two nodes. In certain cases the parti- tion at the nodes had not been removed, whereas in others the entire limb formed one continuous nesting cavity. An internode of one of the living branches was occupied by a nest of a small solitary bee belonging to the genus Allodape. In a recent note De Wildeman (1920) directs attention to several African Clerodendrons with fistulose stems, such as C. excavatum De Wildeman, C. angolense Guerke, and C. cavum De Wildeman. The last named was described from specimens which I collected in the Savannah country of the northeastern Belgian Congo, near Boga (July 12, 1914; Coll. No. 5002), between Beni and Kasindi (August 9, 1914; Coll. No. 5205), and near Rutshuru (September 4, 1914; Coll. No. 5534). Itisa low bush of the open grass-land, with white flowers; I never observed ants living in or on it. Following the description of his Clerodendron formicarum, Guerke mentions-that he saw a specimen obtained by Stuhlmann near Bukoba. Ants of the genus Crematogaster were living in its hollow stem, the walls being pierced by a circular hole. Guerke, however, was doubtful as to the specific identity of this Uganda specimen and the typical C. formi- carum from Angola and the Kasai. The latter is a low, semi-herbaceous plant, 25 to 30 cm. high, which, as I have shown elsewhere, is not the myrmecophyte its name would imply. Stuhlmann’s specimen from Bukoba was a rather high, much-branched shrub, with smaller flowers and there is a possibility that it belonged to C. cavum De Wildeman, col- lected by me in several near-by localities. Virex Linneus Viter (Tournefort) Linnazus, 1753, ‘Species Plantarum,’ Ed. 1, II, p. 638. J. Briquet, 1895, in Engler and Prantl, ‘Die Natiirl. Pflanzenfam.,’ IV, pt. 3a, p. 170. J. G. Baker, 1900, in Thiselton-Dyer, ‘Flora of Tropical Africa,’ V, p. 315. “Trees or shrubs, with glabrous or hairy branches. Leaves opposite, usually com pound, digitate, rarely simple. Cymes dichotomous, axillary or forming a ter- minal panicle. Flowers whitish, yellowish, lilac, or blue. Calyx campanulate or funnel- shaped, 5-toothed or nearly truncate, accrescent. Corolla-tube short or long, sub- cylindric or funnel-shaped, straight or slightly curved; limb obliquely patent, sub- bilabiate. Stamens 4, didynamous, inserted in the corolla-tube and usually ex- serted from it; anther-cells nearly parallel or divergent. Ovary at first imperfectly | Wheeler, Ants of the Belgian Congo 445 _ ly finally * Leelled; ovules solitary, laterally attached; style filiform, bi sai Drupe with a more or less fleshy mesocarp and a hard, 4-celled endo- rp. Se cierstoee chioce, calearssoe” (J. G. Baker, 1900). s diagnosis should be amended to include creepers also. Apert e myrmecophilous species of the Ituri Forest described below, sr form was apparently thus far unknown in the genus. J. 2 ne ‘genus Vitex contains over one hundred species in the tropical tr ical parts of both hemispheres. A large number of these are te endocarp of the drupe forms a single 4-celled anklet. in Clero- Ga m each fruit contains two 2-celled or four 1-celled nutlets. In n, all known forms of Clerodondron have simple ain either pbat probably other tropical members of the genus also have s with ants. 4 Vitex Staudtii Guerke , Staudtii Guerke, 1903, Engler’s Bot. Jahrb., XX XIII, 2, p. 299. _ “Tree or shrub? with quadrangular branches. Leaves 5-foliolate, with very long tioles; the leaflets petiolulate, obovate, attenuate at the base, with entire margin, nding in a very long apex, rough above, glandular below. Inflorescences terminal, pidal, loose; peduncles puberulent. Calyx cupuliform, with truncate or tely margin. _ “The branches are sharply quadrangular, entirely glabrous, hollow. The posite leaves are 5-foliolate, with a petiole 10 to 18 em. long. The leaflets have a ole of 5 to 20 mm. ; that of the median leaflet longer than the others; they are obovate, twice as long as broad on the average, without the apex 10 to 14 cm. long d 5 to 7 em. wide; narrowed at the base into the petiole; with entire margins; ed into a tip which is suddenly constricted at the base and 1 to 3 cm. long; » upper side with very short, scattered, coarse hairs; under side glabrous, but ‘ly covered with minute, golden yellow glands. The thyrsoid inflorescences are as much as 30 em. long, very loose with far spreading branches, which are r like the petioles, and glabrous or with feeble downy hairs toward the ‘In Engler and Prantl, 1895, ‘Die Natérl. Pflanzenfam.,’ 1V, roy 133. en mee ot rere the dex Remenn nortan tiie Supplemen - _ *" Arbor vel frutex."" This should be amended to “ Maseiaeeie es iaih ot denstotlen. aoe. es i - i 446 Bulletin American Museum of Natural History (Vol. XLV apex; the subdivisions of the inflorescence are pseudo-umbels of 6 to 20 flowers. The peduncles are 2 to 4 mm. long, covered with fine downy hair and bear about the middle of their length 2 lanceolate, easily dropped, downy bracts, 2 to 4mm. long. The calyx is broadly cupuliform, 3 mm. long, with a truncate or very indistinctly 5-toothed margin. The corolla is greenish-white, covered with yellow glands outside, with curved tube. “The species belongs in the section Agnus Castus and more definitely in Briquet’s — Terminales-group. Among related forms, V. Buchanani Baker differs in the smaller, hairy leaves; V. quadrangula Guerke also is more strongly pilose. V. thyrsiflora Baker too belongs in this group, but is known to me only by the description according to which the leaves are pubescent on the under side also and the calyx apparently is more distinctly toothed. The present species is furthermore characterized by being inhabited by red ants; the hollow branches usually show at the nodes the almost circular orifices which are characteristic of so many ant-plants’’ (Guerke, 1903). Togo: not rare in the forest (Baumann). Cameroon: Yaunde (Zenker and Staudt). Belgian Congo: Northeastern Congo Forest: Avakubi (January 1914; Lang, Chapin, and J. Bequaert; Coll. No. 1803); Medje (July 1914; Lang and Chapin); Penge (January 31, 1914; J. Bequaert; Coll. No. 2216); between Penge and Irumu (village of Nduye, February 20, 1914; J. Bequaert); Kilo (June 30, 1914; J. Be- quaert; Coll. No. 4894). V. Staudtii must also occur in Spanish Guinea, since its peculiar host, Viticicola tessmanni (Stitz), was originally found at Alen, Spanish Guinea, by Tessmann. With the exception of the indication “tree or shrub,” Guerke’s diagnosis of V. Staudtii agrees perfectly with a myrmecophilous creeper obtained by me in the Ituri Forest and of which dried branches were also brought back by Messrs. Lang and Chapin. In the hope of identifying this plant, I have carefully read the numerous published descriptions of African Vitex and there is a reasonable certainty that the Ituri creeper is either identical with or very closely allied to Vitex Staudtii.1 The designation “tree or shrub” is, I believe, due to the fact that Guerke based his description on a few herbarium specimens, which gave not the slightest indication that the species was a creeper; moreover, all other members of the genus known thus far are either trees or erect shrubs. Vitex yaundensis Guerke Vitex yaundensis GuErke, 1903, Engler’s Bot. Jahrb., X XXIII, 2, p. 296. “Tree, with very long petiolate, 5-foliolate leaves. Leaflets short petiolulate, oblong-ovate, cuneate at the base, with entire margin, very glabrous on both sides. Flower-cymes axillary, with very long peduncles. Bracts linear. Calyx turbinate, 5- toothed, with deltoid teeth. Tube of the corolla hardly raised above the calyx. 1Mr. Chapin informs me that he saw this myrmecophilous creeper also near Ngayu. *The first indications as to the taxonomic position of this curious zayrmecnns were given by Prof. I. W. Bailey, who, from histological examination of the stems, concluded that it fo bedie to the Verbena- ce, most probably in the genus Vier. Wheeler, Ants of the Belgian Congo 447 ‘ a, “high, with quadrangular, glabrous branches. The leaves are 5- foliolate, borne on a petiole 15 to 22 em. long, which is glabrous with a flattened ove above. Leaflets with a petiole 1 to 2 cm. long; elongate-ovate, narrowed at e base into the petiole; with entire margin; long acuminate, herbaceous, entirely rous on both sides. Parallel lateral veins very numerous, as many as 25 on the median leaflet. The median leaflet reaches a length of 24 em. and a width of 9 cm.; the lateral leaflets nearest it are a little smaller, reaching a length of about 20 cm.; ey ae 0. leaflets are only 14 em. long and 7 em. wide, being in proportion vader than the two lateral leaflets. The inflorescences are placed in the axils of leaves and borne on peduncles 16 to 20 cm. long; they are loose, com- ?p double cymes (dichasia) with strikingly long ramifications. The bracts are aioe, linear, long acuminate, with fine downy hair, as much as 15 mm. long on the _ lower Tamifications; shorter on the upper ramifications. The peduncles are 2 to 3 _ mm, long and covered with fine downy hair. The calyx when expanded is top-shaped, = downy, 3 mm. long, 5-toothed; the teeth are triangular with even sides, rather acute, a. 1 mm. long and about as wide at the base. The corolla has a very thick, glabrous tube, which is only 4 mm. long; the limb is distinctly bilabiate and 5-lobed; the two a lobes are ovate, obtuse, downy, 1 mm. long; the two lateral ones have a - similar shape and pilosity, but are 2mm. long; the anterior one is spatulate, some- what emarginate, 4 mm. long, pilose at the base and on the middle line, otherwise The flower is greenish-yellow; the anterior lobe violet. etek “The species belongs near the very large leaved V. grandifolia Guerke and V. bipindensis Guerke, but differs in the squarrose, very loose inflorescences, and also _ in the remarkably numerous lateral veins of the leaves. The plant is certainly in- habited by ants, as one can conclude from the characteristic cireular openings at the nodes of the branches” (Guerke, 1903). Cameroon: Yaunde (Zenker). Ecoxrocy or Vitex Staudtii “My attention was first called to this remarkable myrmecophyte by my friend, Mr. J. P. Chapin, at Avakubi, in January, 1914. Knowing ee Cctereet in ant-inhabited plants, he directed me to a swampy, wooded spot on the banks of the Ituri River, about five miles upstream from that locality, where there were many specimens of a creeper in the under- growth of the forest. When the stems of this plant were slightly touched. or otherwise disturbed, large numbers of slender, reddish ants rushed out of the hollow stalks ready to attack. I later came across the same _ ereeper on several occasions during my travels in the Ituri Forest, and it appears to be fairly common throughout that region. On the other hand, I never saw it along the Semliki River or in the primitive forest between Lake Kivu and the Lualaba. All the specimens observed by me were growing in very moist places, usually in parts of the forest flooded after heavy rains. The older plants consist of a long, flexible, woody main stalk, about 15 to 20 mm. thick at the base, or occasionally more. This stem begins to 448 Bulletin American Museum of Natural History ]Vol. XLV branch feebly and irregularly a short distance from the ground and climbs freely among bushes and low trees, sometimes to a height of 8 to 10 meters. Its upper part is much more abundantly ramified and spreads leaves and flowers over the crown of the supporting vegetation. The compound, digitate leaves, of three to five nearly sessile leaflets, are borne on long petioles and placed opposite each other in decussate rows. Young branches and those on the upper part of the plant are quad- rangular their whole length, with four slightly convex or nearly flat sides and more or less winged angles. These four winged ribs are continuous along the limb, at the nodes running on both sides of the petioles. Older branches show the ribs much less pronounced, the surfaces between becoming more convex, but often they still possess fairly pronounced wings, which can even be traced along the main stalk. Stem and branches show no sign of swellings. I have never seen the flowers, but the fruit is small, spheroidal, dry, hard, and of a pale orange-yellow colo when ripe. Adult plants were always inhabited by ants, invariably ot the species Viticicola tessmanni (Stitz). The insects enter and leave their nests through a few orifices arranged in pairs at the nodes, nearly opposite cach other and between the points of insertion of the leaves (Fig. 88a). The aperture, usually more or less crater-shaped, is placed at the top of a slight elevation which is produced by a peculiar ring of sclerenchyma, as shown by Prof. Bailey (see Part V, p. 591). On examining a very young specimen of this Vitex still free of ants, I was unable to find a depression, elevation, or scar on the surface to indicate the points where the insects would later gnaw entrance holes. Prof. Bailey’s histological study shows that the most favorable situation for the nodal apertures is midway between the points of attachment of the leaves (see Part V, p. 592). The location of exits in Vitex Staudtii compares to a limited extent with that in Cecropia adenopus, in which, however, the entrances are always pierced above the axils of the leaves but in a section of the stem which is practically devoid of tough tissues (Schimper, 1888). In Cecropia the location of this diaphragm of softer tissues is marked ex- ternally by a roundish depression or prostoma, at the upper end of a shallow groove running upward from the insertion of the petiole; the — ants of Cecropia always locate the entrance to the hollow stems in the depressed prostoma. How in Cecropia, Vitex Staudtvi, and other similar cases the ants discover the spots particularly favorable for apertures and why they practically restrict their attacks to these parts of the stem are questions which cannot be satisfactorily answered at present. It 449 ‘Fig. 88. Vitex Staudtii Guerke: a, external view of of stalk inhabited by the ant Viticicola mar H nait yee of ceefret pe Ps ahere © the bark ot naan: 1, leteres oe eg ed pode ine ones blindly en e bar accessory ¢ wed through the rhe tints ds A and B-B indicate the are figured in Pl. X ae ata 1 (A-A) and fig. 2 eB) . Drawn from life 450 Bulletin American Museum of Natural History [Vol. XLV has been suggested (Wheeler, 1913, p. 136) that ants may be able, through their extremely delicate tactile (or rather chordonotal) sense- organs, to select the thinnest spot in the wall of a cavity for perforation. Their sense of smell may also warn them against gnawing parts of the stem containing certain distasteful substances. A longitudinal section of the stalks (Fig. 88b) discloses many features of further interest. In the first place, adult plants occupied by an ant’ colony are hollowed out nearly from top to bottom, all the internodes and various branches freely communicating with one another. The entire plant shelters one ant community, containing, in addition to one or more deiilated queens, a number of fertile, ergatoid, wingless females. The formicaries of Viticicola tessmanni in the stems of Vitex Staudtit are thus splendid examples of polygynous insect societies. As in the case of Pachysima colonies in Barteria, they probably originate through secondary pleometrosis, or subsequent fusion of several isolated colonies, each started by a fecundated queen in the various limbs. A young specimen of Vitex Staudtii, scarcely 1 m. high, growing near the village of ' Nduye, between Penge and Irumu, was particularly instructive in this connection. Each of the lower internodes on the side-branches was occu- pied by a fertile, deiilated female of Viticicola tessmanni, together with brood at various stages of development; no workers were present. The ants clean out most of the medullary tissue nearly the entire length of the internodes, leaving only a peripheral layer of it for a short distance a little above the node. This remaining pith partly constricts the cavity and is probably left to keep the brood of the ants from drop- ping below the node, thus helping to distribute it regularly over the various internodes of the vertical stems and also preventing it from obstructing the apertures at the nodes. On a level with this inner circle of tissue the walls of the stem are also slightly thicker than in the other parts of the internode. At Kilo, in June, 1914, I saw a very young Vitex Staudtii composed of an unbranched, leaved, erect, thin stem about two. feet high and unoccupied by ants. The central cylinder of the whole plant was filled with soft medullary tissue. It is possible that this sub- stance dries up by itself, causing the stems to be hollow without the intervention of ants. In nature, however, this must be rarely the case, for, in adult plants housing a colony of Viticicola, pith is found only in the topmost internode of very young branches which are still green and soft; the ants steadily work upward through the nodes and excavate the interior before it has begun to dry. ———— ee a ee: # eS ee ee eae ee ee Wale; Ants of the Belgian Congo 451 Tho inn inner avelii of the hollow stalks also show a peculiar series of ons or narrow channels, the like of which is not known for any sssonkiyte. These lateral cavities perforate the xylem and pendly just under the cambium; they are arranged at irregular is, one above the other, in two longitudinal rows. The rows are ite each other and their position shifts at every node, so that they run on the sides corresponding with the upper pair of apertures y internode. The number of channels in a row varies with the th of the internode, in some cases there are fifty or more, but often It occasionally happens that one of these lateral galleries per- es the bark, and this supplementary exit hole then produces the icola. They are not used by the inmates for their eggs or very young ». Coccids are not found in these channels and, furthermore, are nt from the hollow stalks of Vitex Staudtii. It was at first believed p oved by Prof. Bailey (see Part V, p. 586). He found that the bark tside the depressions presents no lenticels or patches of aérenchyma act and, in old stems, there are disks of impervious sclerenchyma led just opposite the blind ends of the cavities. Moreover, Prof. iley discovered that the channels are not natural gaps in the woody n of the wall, but are excavated by the ants in peculiar cores of icate, unlignified cells, that are symmetrically distributed in certain di a the stem and surrounded by abnormal tissues similar to those ssented by heteroplasmatic zoocecidia. The arrangement of the gal- in two rows below the apertures of the upper node results from fact that in Vitex Staudtii the principal water-conducting passage- ys in each internode are largely confined to those sides of the stele th pass out to the leaves at the next (higher) node. The lateral _ avities are excavated in the sides of the stele poorly supplied with _ vessels and, furthermore, located in those portions of the xylem which are _ devoid even of a narrow fringe of small primary trachee. _ Prof. Wheeler has given a detailed description of Viticicola tess- _ manni (Stitz), the obligatory guest of Vitex Staudtii, in its various adult _ phases and larval stages. My observations in the field furnish no clue . as to the possible food of these insects, but the ants are evidently adapted 452 Bulletin American Museum of Natural History [Vol. XLV to their life within the cavities. Owing to the fact that the plant grows in swampy places, I am inclined to believe that the ants seldom, or never, leave their host. Wheeler and Bailey’s examination of food- pellets dissected from the infrabuccal pockets of the adults and the trophothylaces of the larve failed to reveal traces of food from an out- side source. The insect substances in the pellets of the larvae resembled the yolk of ants’ eggs and the fat-body of the larva themselves, suggest- ing that some of the brood had been used as nourishment for the more vigorous progeny. In one instance pieces of the skin of a Viticicola larva could be clearly recognized. There were also spores and bits of hyphw in many cases and particles that seemed to be pith and callus tissue (Wheeler and Bailey, 1920, p. 261). Bailey thinks that the prin- cipal food of Viticicola tessmanni is provided by the medullary tissue of young twigs and the ‘nutritive layer” which is produced in the lateral galleries of the domatia (see Part V, p. 606). . Viticicola tessmanni is exceedingly vicious and alert. When its host plant is ever so slightly disturbed, the workers rush out of the hol- low stalks in large numbers and actively explore the plant. Their sting is extremely painful and sometimes produces vesicles on the skin. It is certain that they constitute a very efficient body-guard of their host. Yet, on one occasion, I observed galls on the leaves of a Viter oceupied by the ants. Rubiacez In Africa, as elsewhere, this family is the richest in myrmecophytes, and without doubt the list of its species which form ccenobiotic associa- tions with ants will be considerably increased by future investigation. Unfortunately many of the genera contain a large number of closely allied forms and even the generic distinctions are often unsatisfactory. It is, therefore, urgent that field-observations on these plants be accom- panied by complete and abundant herbarium specimens for later identifi- cation by botanical experts. ; So little is known about the two following cases that I have not treated them in the same detail as true myrmecophytes. Grumilea venosa Hiern : Grumilea venosa Hiern, in Oliver, 1877, ‘Flora of Tropical Africa,’ III, p. 217. De WILpEMAN AND Duranp, 1901, ‘Reliquie Dewevreanz,’ p. 130. Tu. ano H. DvrRanp, 1909, ‘Sylloge Flor. Congol.,’ p. 280. H. Konx, 1909, Natur u. Offen- barung, LV, p. 167. Wheeler, Ants of the Belgian Congo 453 Specimens collected re ebeieh in the Belgian Congo (Leopoldville; Bokakata) ‘the following note: “Arbuste de 2 m. environ, toujours habité par de nom- ses fourmis noires."” (De Wildeman and Durand, 1901, p. 130). i Uragoga species ? D the forest bordering one of the affluents of the Congo near Leo- _poldville, I came across a semiherbaceous, low bush, which I provision- alk refer to the genus Uragoga (May 18, 1915; Coll. No. 7656). The _ flowers are white, with greenish spots on the teeth of the corolla; the lit is ared berry. At each node, between P e points of attachment of the leaves, there » two curious, persistent stipules, occupy- the entire width of the stem (Fig. 89). sy are convexly swollen to the upper . side and the free margin is recurved down- : rd, the whole forming an inverted cup ouch broadly open below. Coccids were lly found inside this cavity and the _ Fig.89. Uragoga species? Swollen , Crematogaster striatula variety obsti- “tPules st base of tenting nata (Santschi), had built a tent of vegetable — May, 1915; twice natural size. material over the inferior opening of the . stipules. I did not find eggs, larve, or pupx of these ants inside the stipules, which I therefore regard not as myrmecodomatia but merely as _ “kraals” to shelter the scale insects. Yet this case suggests useful comparison with the stipular pouches of Macaranga saccifera and _ more typical ant structures of plants. Uncaria Schreber ee 1789, in Linnzus, ‘Gen. Plant.,” Ed. 8, I, p. 125. Hern, 1877, in Oliver, ‘Flora of Tropical Africa,’ III, p. 41. Havitanp, 1897, Journ. Linn. y Soc. London, Botany, XX XIII, p. 73. Ourouparia Ave.et, 1775, ‘Histoire des Plantes de la Guiane frangaise,’ I, p. 177, q _ *PL. uxvur. K. Scuumann, 1891, in Engler and Prantl, ‘Die Natiirl. Pflanzen- a fam.,’ IV, pt. 4, p. 57. _ Agylophora Necker, 1790, ‘Elementa Botanica,’ I, p. 145. , a _ Climbing shrubs with opposite, interpetiolar, fugacious stipules; the lower part _ of the terminal branches with axillary recurved hooks, often spirally rolled up and 4 _ placed opposite each other; in some cases these hooks still bear a few aborted, oppo- site leaves. On older branches the recurved hooks are often replaced by heavy, _ woody thorns. Leaves usually leathery, rarely herbaceous; the stipules entire or _ bifid. Flowers pedicelled or sessile, crowded into loose, globose heads, without inter- ) E vening bracteoles. Flower heads placed in the axils of the upper leaves, either singly _ or in decussate panicles. Calyx salver-, or bell-, or funnel-shaped; the calyx-tubes 454 Bulletin American Museum of Natural History (Vol. XLV not cohering, finally fusiform; the limb campanulate, 5-lobed or 5-partite. Corolla often silky outside, funnel-shaped with an elongated tube; the lobes overlapping one another in the bud; throat of the tube bare. Stamens 5, inserted at the throat of the corolla; filaments short; anthers short oblong, the thee more or less extended or in some species setose at the base. Disk inconspicuous. Ovary fusiform, 2-celled; style exserted far beyond the tube of the corolla, ending in a clavate or capitate stigma, Ovules numerous, ascending, inserted on a placenta which is borne by the middle partition of the ovary. Fruit a septicidal, many-seeded capsule; testa of the seeds. extended at both ends into capillary appendages. Uncaria is a close relative of the East Indian Nauclea, differing mainly in the characters of the fruit, which in the last-named genus is a capsule, not septicidal, but merely breaking up into two cocci or lobes. All the species of Uncaria are climbing shrubs found in the forested areas of the tropics. Over thirty species are known, most of them from tropical Asia; two occur in South America and one in Africa. Uncaria africana G. Don Uncaria africana G. Don, 1834, ‘General System of Gardening and Botany,’ ITI, p. 471. Hooxer, 1848, ‘Icones Plantarum,’ VIII, Pl. pecuxxxt. Bentuam, 1849, in Hooker, ‘Niger Flora,’ p. 381, Pl. xu. Hern, 1877, in Oliver, ‘Flora of Tropical Africa,’ III, p. 41. Havitanp, 1897, Journ, Linn. Soc. London, Botany, XXXII, p. 76. E. De Witpeman ANnv T. Duranp, 1900, Bull. Herbier Boissier, (2) I, p. 25; 1901, ‘Reliquie Dewevreane,’ p. 106. E. De Witpeman, 1903, ‘Etudes Flore Bas- et Moyen-Congo,’ I, p. 76; 1907, ibid., II, pp. 71 and 192. Tu. anv H. Duranp, 1909, ‘Sylloge Flor. Congol.,’ p. 241. E. De Witpe- MAN, 1910, ‘Etudes Flore Bas- et Moyen-Congo,’ III, pt. 2, p. 283; 1912, ibid., III, pt. 3, p. 484. Nauclea africana WALPERS, 1843, ‘Repertorium Bot. Syst.,’ II, p. 512 (nec Willdenow). Ourouparia africana BAILLON, 1879, Bull. Soc. Linn. Paris, I, p. 228. K. ScHumann, 1891, in Engler and Prantl, ‘Die Natiirl. Pflanzenfam.,’ IV, pt. 4, p. 57. Hiern, 1898, ‘Catalogue Afr. Plants Welwitsch,’ IT, p. 435. Uruparia africana K. ScuuMANN, in Engler, 1895, ‘Pflanzenwelt Ost-Afrikas,’ C, p. 378. : “A glabrous or sparingly pubescent shrub 4 to 40 ft. high. Leaves ovate-oval or lanceolate, acuminate, rounded at the base or nearly so, thinly coriaceous, with 5. to 7 lateral veins on each side of the midrib, shortly petiolate, 2 to 6 by 1 to 4 in.; stipules 4 to 4 in. long, usually bipartite with narrow partitions. Spines at first nearly straight, afterwards crooked. Flowering leads 1 }4 to 2 in. in diameter. Calyx tawny, as well as the corolla shortly and appressedly pilose-tomentose; limb shortly lobed. Corolla about \% in. long, greenish yellow. Stamens glabrous; anther-cells obtusely produced at base. Fruiting heads 3 to 3% in. in diameter; pedicels % in. long; capsules % in. long. Tails of the seeds linear-setaceous, undivided at one end, bipartite at the other” (Hiern, 1877). . 1Haviland’s diagnosis reads: ‘‘Frutex 1-15-metr. Ramuli glabri vel pubescentes; internodi 7-9 cm. Folia 13 em. longa, 6 cm. lata, elliptico-lanceolata, longe acuminata, supra glabra, pabten eave scentia, nervis 7-8. Petioli7 mm. Stipule bifide, lobis acutis. Pedunculi 7 cm., maxime Flores pedicellati vel subsessiles. Corolla sericea, mellea; tubus 1 em., lobi oblongi. Calyx sericeus; tubi pars —— 4mm.; lobi 1 mm., obtuse triangulares. Stylus 18 mm.; stigma 4 mm., elongato-~ clavatum. Capsule 2 cm., cum pedicellis equilonge. Bractew nullx.” : \ Whesler, Ante of the Belgian Congo 455 ; .) Re ef Fig.90. Uncaria africana G. Don. Extremity of branch with capitulum of fruits (after Bentham, 1849). ___ Sierra Leone, type locality (G. Don; Afzelius; Barter; Scott Elliot; Johnston). Nigeria (Vogel). __ Cameroon. _ §Spanish Guinea: Rio Muni (Mann). Belgian Congo: Lower Congo; banks of the Lukungu River (Dewévre); Kisantu (Gil et); Inkisi River (Vanderyst). Kasai: Linkanda (Gentil). Upper Congo: _Mondombe (Jespersen). Northeastern Congo forest: Mangbetu Country (Schwein- furth); Uele region (Seret); Barumbu (November 3, 1913; J. Bequaert; Col. No. 1069); Penge (January 27, 1914; J. Bequaert; Coll. No. 2136); between Penge and 456 Bulletin American Museum of Natural History {Vol. XLV Irumu (village of Tete, February 22, 1914; J. Bequaert; Coll. No. 2658); between Walikale and Lubutu on the Oso River (village of Mandimbo, January 18, 1915; J. Bequaert; Coll. No. 6664). Uganda. Angola: Golungo Alto—‘‘in the primitive forests of Sobato de Mussengue’’ and “in the very dense, primitive forest of Quibanga’’ (Welwitsch). Also known from Madagascar and the Comoros. Haviland distinguishes several varieties: Variety (1). Flowers subsessile. Upper part of the calyx-tube 4 mm. long. Sierra Leone, Niger, Mangbetu. Variety (2) madagascariensis (Ourouparia madagascariensis Baillon, 1879, Bull. Soc. Linn. Paris, I, p. 228). Flowers subsessile. Upper part of the calyx-tube 2 mm. long. Malagasy Region. Variety (3) angolensis Haviland. Flowers pedicellate. Upper part of the calyx- tube 4 mm. long. Angola. Variety (4). Flowers pedicellate. Upper part of the calyx-tube 2 mm. long. Cameroon. The variety angolensis Haviland is described more in detail by Hiern (1898) as follows: An arborescent shrub, glabrous except the inflorescence. Trunk in some cases more than 100 ft. long and 6 in. in the lower part, climbing to a very great height and then hanging down; branches patent, fuscous, rather glossy, tetragonal. Leaves opposite, elliptical, narrowly acuminate at the apex, obtusely narrowed or nearly rounded at the base, thinly coriaceous, glossy, dark green above, paler beneath, 2 to 4 in. long, 4 to 1 % in. broad; lateral veins about 6 on each side of the midrib, slender; petiole 4 to 4 in. long. Stipules ovate, small, somewhat hairy on the inner face, nearly deciduous. Spines axillary, mostly crooked, 4 to % in. long. Flower heads terminating the branches, shortly pedunculate, globose, about 2 in. in diameter. Flowers golden-tawny, about % to % in. long (including the exserted style), very numerous, crowded. Pedicels about to % in. long in flower, % in. long in fruit, tomentellous. Bracts 0. Calyx silky-tomentellous with short upward hairs, some- what constricted above the ovary, greenish-fuscous; tube broader than the ovary, campanulate, funnel-shaped, \% to ¥ in. long, shortly 5-cleft, lobes thickly lanceolate. Corolla 4 to % in. long; tube slender, except the base clothed outside with downward tawny short silky-tomentpee hairs, 4 in. long; limb much broader than the tube, hemispherical, 4 to \% in. in Giatneter, golden-tawny tomentose outside, glabrous in- side, deeply 5-lobed; segments about gin. long, obovate-oblong, rounded at the apex with an apiculus. Stamens 5, about half as long as the corolla-lobes, glabrous, introrse, inserted on short, flattened filaments at the base of the corolla-limb. Ovary ellipsoidal, tomentose, rather thicker than the base of the calyx-limb, much thinner than the top of the calyx-limb. Style filiform, exserted about \ in. beyond the corolla, glabrous below, stigmatose and rather thickened in the upper part towards the clavate stigma. Young fruit subglabrous, about % in. long, % in. thick, narrowed at both ends especially towards the base. This species probably occurs throughout the African Rain Forest. In a recent note, De Wildeman (1919) calls attention to the myrme- Wheeler, Ants of the Belgian Congo 457 Fig. 91. Uncaria africana G. Don. Longitudinal section of myrmecodomatium at a node, showing cavities communicating with one another; the aperture is not figured. Drawn from life at Barumbu, mber, 1913; natural size. ophytism of certain African plants of this genus. He proposes, provi- ‘sionally, the varietal name myrmecophyta De Wildeman for specimens n I collected in the Ituri Forest at Penge and between Penge and 1u, without, however, giving characters by which this new variety i be differentiated from the typical form. I am inclined to believe ‘myrmecophytism is normal for Uncaria africana throughout its ze and has merely been overlooked thus far. When terminal branches ne are collected, there may be no indication of the peculiar myrme- odomatia in herbarium specimens, even should such have been present m lower parts of the plant. Ant-inhabited parts of plants are also fre- _ quently avoided by botanical collectors. Moreover, it is possible that _ the myrmecodomatia are absent or but little pronounced in certain _ individuals or at certain stages of growth. 458 Bulletin American Museum of Natural History (Vol. XLV Eco.ocy or Uncaria africana Here (Figs. 90 and 91) we have one of the many climbing bushes or *‘scramblers’’—as Schimper proposed calling them—which frequently form tangles of vegetation over the low trees at the edge of clearings and along river banks. While the trunk and main branches are straight and stiff as in ordinary bushes, all or part of the lateral branches are limp and pliable. The latter either hang down freely or work their way upward, keeping hold of the other trees by means of the spirally curved hooks and woody thorns, which are placed in pairs above the nodes and are evidently modified branches. The leaves are glabrous, as well as the branches at the extremity of some of which the flowers or fruits form head-clusters. The myrmecodomatium of this Uncaria consists of the enlarged and hollow basal internodes of two opposite, lateral branches, the cavi- ties in this pair of swellings communicating with the hollow, very slightly swollen node of the main branch (Fig. 91). The middle chamber is more or less club-shaped, 5 to 6 em. long and 10 to 20 mm. wide in the upper half; it is dug farther into the pith below than above. The two lateral cavities are 3 to 6 cm. long and 6 to 10 mm. broad. All the specimens I had opportunity to study in the field were in- habited by ants of the genus Crematogaster, which were identified as C. excisa subspecies andrei (Forel) in the case of the plants found near the Oso River, between Walikale and Lubutu in January, 1915. The myrme- codomatia contained not only the queen, workers, and brood of the ants, but also numerous coccids. These scale insects were invariably located in the lateral swellings and fixed at the bottom of two deep, opposite, longitudinal grooves in the inner wall. One or more circular exit holes are pierced by the ants through the sides of the lateral cavities. Often the depressions occupied by the coccids are open to the exterior by means of irregular slits through which the ants enter and leave. It would thus seem that these grooves are gnawed by the ants, probably on account of some special hyperplasias formed in that region of the walls. The coccids merely select the grooves for nutritive, juicy tissue to be found there and continually renovated by the attacks of the ants. On the plants examined by me at Barumbu in November, 1913, there were a number of young branches whose basal internodes, though dis- tinctly swollen, were still filled with juicy pith tissue. In another in- stance, between Walikale and Lubutu, the basal swellings of many older branches were not yet inhabited by ants, presenting no exit holes; nevertheless, they were entirely hollow inside, so that the cavities of Wheeler, Ants of the Belgian Congo 459 cat ia_originnte ‘enoudhs the drying up of the pith and without the enc} pet Eeneet have shown to be the case with the myrmecodomatia SarcocepHatus Afzelius cocephalus Arzetius, 1818, in R. Brown, Tuckey’s ‘Narrative Exp. Zaire,’ _ Appendix No. 5, p. 467. Hiern, 1877, in Oliver, ‘Flora of Tropical Africa,’ II, p. 38. K. Scuumann, 1891, in Engler and Prantl, ‘Die Natiirl. Pflanzen- _fam.,’ IV, pt. 4, p. 58. : ‘Trees or shrubs with subterete or obtusely quadrangular branches and opposite, C0 petiolate leaves (rarely in whorls of three). Stipules interpetiolar 2 deciduous in the African species, often very large. Flowers whitish, pale pink or yellowish, crowded on globose, common receptacles, forming compact, pedunculate, globose, terminal and axillary heads, without intervening bracteoles. Peduncles of flower heads with small bracts below the middle. Calyx-tubes cohering, truncate or 4 or 5 teeth; these teeth hairy, jagged at the tip or in some species with as many appendages. Corolla narrowly funnel-shaped, with glabrous throat, rather fleshy, glabrous or pubescent, 4- or 5-lobed, imbricated in prefloration, decidu- ous. Anthers 5, subsessile, inserted at the mouth or throat of the corolla, ovate- oblong. Disk inconspicuous. Ovaries grown together, 1- or 2-celled. Style filiform, _exserted far outside the corolla, caducous. Stigma oblong or spindle-shaped, thicker . the style, glabrous, entire, emarginate, or bifid. Placentas centrally attached - (in the African species). Ovules numerous, anatropous. Syncarpium fleshy, globose, pitted or uneven, its cells with thin walls and divided by membranaceous septa. Seeds small, not winged, ovoid, and placed above one another (in the African species) ; funicles spongy; testa crustaceous; albumen fleshy. _ The members of this genus are usually easy to recognize in the field, especially when in fruit. Haviland in his monograph of the Nau- __ ele (1897) recognizes thirteen species in the tropical regions of the Old - World. The number of described forms has been considerably increased _ in later years and must now approximate thirty. Of these six are _ recorded from Africa and four from the Belgian Congo. _ Aceording to K. Schumann (18915, p. 59), Sarcocephalus macro- cephalus, from the Philippines, has swellings inhabited by ants. Haviland is inclined to believe that the plant in question was a Nauclea, a genus known to contain several myrmecophytic species in the Philippines. In ___ the Belgian Congo, however, there is a true Sarcocephalus with myrmeco- _ domatia. At present I can not give its specific identity, but it is cer- _ tainly very distinct from the common S. sambucinus (Winterbottom), _ =. esculentus Afzelius, which, according to my observations, is never _ inhabited by ants. Most probably it represents a distinct, undescribed form. 460 Bulletin American Museum of Natural History [Vol. XLV EcoLocy or Sarcocephalus species This myrmecophyte was first met with in the Ituri Forest, near the village of Banana between Penge and Irumu (February 24, 1914; Coll. No. 2605) and was again seen near the village of Masongo, between Walikale and Lubutu \ fe | (January 15, 1915; Coll. No. 6629). It is a low, erect e~--- ) tree or shrub, rarely over 8 meters high, usually much 2 jj) smaller (3 to4m.). The straight trunk bears, from its foot on, regularly spreading, opposite, decussate branches. The leaves are opposite, large and very broad, usually purplish-red on the under side, espe- cially when young. The terminal part of the branch bears, between the points of attachment of the leaves, striking, broad stipules which, however, are early deciduous.' It never happened that I saw flowering plants, but the fruit is a spheroidal, solid ball, 9 to 10 em. in diameter, placed at the extremity of a side branch, on a short, recurved pedicel. All the specimens observed grew on swampy, rather open spots of the primitive forest, either at the edge of a brook or in the water. It is quite possible that this species has been described before, perhaps under a related genus of Rubiacez, but it agrees with none of the diagnoses seen sofar. Its relation with ants would easily escape notice, for the myrmecodomatia are incon- Fig. 92. Sarco- spicuous and, when not actually oceu- Paulus species* pied by insects, could often be discov- jebiniees ag hoas ered only upon sectioning the branches. tudinally in the EXternally, they consist (Fig. 92) of sot of os of a very slight, often imperceptible swel- aperture. Draw, ling on the upper half or two-thirds of from life at Ma- the internode. | Inside, the central Wattale att. — eylinder is hollowed out into a spacious butu, January, cavity, 6 to 8 em. long and 5 to 7 mm. 115;naturalsize. Vide at the top. Domatia inhabited by ants have a circular aperture a short distance below the node. 1In the common African Sarcocephalus sambucinus the stipules are small (4 mm. long) and persistent; but they are large and caducous in many other species of the genus. Ree arene Oe Page Verve 461 ps Seetions made ‘al a number of young specimens of this myrme- ‘upper portion of the internodes becomes hollow of its own accord through _ the drying up of part of the medullary tissue; such cavities have no exit oles. In this species the lower internodes of the main trunk and side branches are neither transformed into domatia nor in the least swollen and remain completely filled with pith. Very young plants show no trace whatsoever of ant-chambers and on an adult tree the size and shape of the m#rmecodomatia becomes more pronounced toward the upper branches. - In both localities where I observed this Sarcocephalus a number of _ specimens were inhabited by small ants of the genus Crematogaster. _ Those taken from the domatia of the plants between Walikale and Lubutu were identified by Santschi as C. africana subspecies winkleri ' variety fickendeyi (Forel), a form commonly found nesting in other _ places. These insects had established regular colonies in the cavities, __ with a queen, workers, and brood; coccids were also among them, fixed on the inner walls. | Randi (Houston) Lixnus, 1753, ‘Species Plantarum,’ Ed. 1, II, p. 1192. Hrern, - _ 1877, in Oliver, ‘Flora of Tropical Africa,’ III, p. 93. K. Scaumann, 1891, in Engler and Prantl, ‘Die Natiirl. Pflanzenfam.,’ IV, pt. 4, p. 75. Erect or scandent, spinous or unarmed shrubs or trees, with opposite or verticil- late, often leathery leaves; stipules rather short, solitary, entire, more or less heath- like. Flowers large or medium-sized; solitary, few together, or corymbose; as a rule apparently axillary or terminating short lateral branches, or terminal. Calyx-tube ovoid or turbinate, ribbed or cylindrical; the limb usually tubular, truncate, toothed, lobed, or spathaceous; lobes sometimes foliaceous. Corolla white or yellowish, or ‘more greenish; campanulate, funnel-shaped, or salver-shaped ;- tube in some species _ much elongated; limb spreading or reflexed, dextrorsely contorted in the bud. ___ Anthers sessile or subsessile, narrowly linear, inserted at or near the throat or mouth _ of the corollar tube, included or exserted. Disk annular or cushion-shaped. Ovary _ 2celled. Ovules very numerous, immersed in the fleshy placentas. Style strong, glabrous or hairy; stigma club- or clapper-shaped, entire, bidentate or bilobed, sulcate. Berry 2-celled, usually many-seeded; the testa of the seed leathery or membranous. This genus is close to Pouchetia A. Richard and Oryanthus de Can- dolle; still more so to Gardenia Ellis, which it often resembles in general habitus. Gardenia has the ovary completely one-celled for the whole length; this character, however, is not always easy to decide upon ,be- cause in certain species of Randia, as, for instance, R. physophylla, the ovary is incompletely divided into two cells. AS e 462 Bulletin American Museum of Natural History [Vol. XLV About 150 species have been described, by far the majority being found in the Oriental and Ethiopian Regions and a few in Tropical America. In the Belgian Congo the genus is well represented by some twenty-five species. They are trees or bushes with large, showy flowers, growing mostly in the Rain Forest or in the forest galleries along the streams of the Savannah. Three of the African species are associated with ants; they all belong to that section of the genus in which the lobes of the calyx are elongate, slender, subulate, and not leaf-like. R. physophyll@ K. Schu- mann is characterized by the presence of glandular cavities at the base of the leaf-blade (Fig. 94). The two others, R. myrmecophyla BE. De Wildeman and R. Luje E. De Wildeman, possess caulinary myrmeco- domatia and, in addition, agree in the following characters: Trees or shrubs with glabrous branches, feebly flattened at the nodes; the inter- nodes often swollen, spindle-shaped; the swellings being hollow, usually pierced by one or more orifices and inhabited by ants; the leaves are opposite, or apparently verticillate, three of them being placed at about the same level; blade obovate, acuminate, constricted at the base into a rather thick, short petiole. They can be separated as follows on characters mentioned in their descriptions: Flowers erect, placed by. twos or fours in the axils of the leaves, about 22 em. long; tube of the corolla glabrous externally. Leaves smaller, with acarodomatia in the axils .of the -lateral : velte. 255) 2200). eae RR R. Luje De Wildeman. Flowers pendent, solitary, terminal, much larger; the corolla alone 22 to 25 em. long, shortly tomentose externally. Leaves larger, the blade as much as 30 em. long and 15 em. broad, without acarodomatia. .R. myrmecophyla De Wildeman. Randia Luje De Wildeman Randia Lujz E.. De Witpeman, 1904, C. R. Ac. Sci. Paris, CXX XVIII, p. 914; 1904, ‘Notices sur des Plantes utiles ou intér. Flore du Congo,’ I, pt. 2, pp. 282-284; 1907, ‘Etudes Flore Bas- et Moyen-Congo,’ II, p. 159; 1910, ibid., III, pt. 2, p. 286; 1912, ibid., III, pt. 3, p. 487. Tu. anp H. Duranp, 1909, ‘Sylloge Flor. Congol.,’ p. 259. H. Kont, 1909, Natur u. Offenbarung, LV, pp. 155 and 158. f. De Witpeman, 1920, ‘Mission de Briey au Mayumbe,’ pp. 43, 88, 222, and 264. “Large tree with glabrous branches, the internodes often thickened toward the base and pierced by one or two orifices leading into a cavity inhabited by ants. Leaves obovate, acuminate, narrowed at the base into a short and rather thick petiole; blade rather coriaceous, darker colored above than below, 20 to 25 em. long, 5 to 12 em. broad, glabrous on both sides, with an acumen of 15 mm. Lateral veins numbering about 9 on each side of the midrib, anastomosing into a curve before reaching the margin, little or not prominent above, prominent on the under side; in the axils of the origin of the lateral veins there are acarodomatia excavated in the tissue of the nervure and opening by a pore at the under side of the blade, more or less visible on Ee ae ee ee a ee Wheeler, Ants of the Belgian Congo 463 ide as fécblo owellings. Flowers by twos, erect, sessile or subsessile, about em. em calyx about 17 mm. long, glabrous, with 5 ribs ending in 5 subulate, rregular teeth; corolla with a long linear, glabrous tube, rather abruptly widening upper part, the broadened portion about 22 mm. long; ending in 4 ovate- late, acute lobes of about 3 mm.; glabrous externally, sparsely villous inter- 'Gtemens partly exserted, extending beyond the broadened funnel of the corolla about 11 mm. Fruit globose, voluminous, over 15 cm. in diameter” (E. De aan, 1904). Belgian Congo: Kasai: forest along the Sankuru River, type locality (Luja). ide _ Upper Congo: Lukolela (Claessens); Lokelenge (Bruneel); Bianga lef . Mayombe: Ganda Sundi (de Briey). Tiaenling to De Wildeman (1910, p. 286) the leaves are often placed verticils of three; the flowers frequently by fours; the fruit is grayish, ibspherical, with 5 more or less conspicuous ribs. This species is close to ndii lapeulats de Candolle, =R. longiflora (Salisbury), but differs the game of acarodomatia in the axils of the lateral veins and the " ellings of the internodes. Randia myrmecophyla De Wildeman ah: 2 myrmecophyla ©. De WitpeMan, 1907, ‘Etudes Flore Bas- et Moyen-Congo,’ II, pt. 2, p. 160, figs. 5-8, Pls. xxxvii-xxxrx; 1908, ibid., IT, pt. 3, p. 346; "1907, ‘Mission Emile Laurent,’ V, pp. cexxiii-cexxiv, figs. IX, XI. iia tiermecophita Tu. anv H. Duranp, 1909, ‘Sylloge Flor. Congol.,’ p. 260. - Re | myrmecophila H. Koux, 1909, Natur u. Offenbarung, LV, pp. 158-160, fig. 5. _ “Shrub with glabrous branches, which are flattened at the nodes, and swollen “Variety typica De Wildeman (1907, p. 160). “Petiole 15 to 25 mm. long, short tomentose, flat above. Blade of the leaves glabrous with age, densely villose and silky inside; its tube, including the ovary, about .5 em. long, with conspicuous ribs ending beyond the truncate margin in 5 linear teeth, 5to 13mm. long. Corolla with its tube 22 to 25 em. long, shortly tomentose exter- nally; more heavily villose inside, except in its widened, glabrous part which is 9 cm. long; lobes villose on both faces, 5.5 em. long and of about the same width, partly - oe in the bud. Fruit ovoid, 10 em. long, 8 em. in diameter, with 5 feeble a subglabra De Wildeman (1907, p. 163). ‘ _. “Petiole 8 to 15 mm. long, sparsely and short tomentose, flat on the upper side. __ Leaf-blade rather broadly cuneate at the base, glabrous and shiny above, glabrous and ___ dull on the under side, except on the lateral veins of which there are 12 or 13 on each 464 Bulletin American Museum of Natural History (Vol. XLV side of the midrib; 18 to 23 cm. long and 7.5 to 12.5 em. broad. Calyx short tomen- tose externally, becoming glabrous with age; densely silky-villose inside; its tube including the ovary about 2.5 to 2.8 em. long, often split on one side; ribs conspicu- ous, ending beyond the truncate margin into 5 linear teeth, 16 mm. long. Corolla with a tube of 21 to 22 cm., the lobes about 4m. by 4 em.; the villosity as in the form lypica. “Variety glabra De Wildeman (1907, p. 163). “Petiole 15 to 30 mm. long, glabrous. Leaf-blade long cuneate at the base, glabrous on both faces, shiny above, dull below; with about 14 lateral veins on each side of the midrib; 18 to 26 cm. long and 6 to 10.5 em. broad. Calyx glabrous external- ly; the tube including the ovary about 2 em. long (in the bud), the teeth 6 to 15 mm. long’ (De Wildeman, 1907). . Belgian Congo: Kasai: Bombaie (E. and M. Laurent). Middle and Upper Congo: Eala, type locality (Pynaert; M. Laurent; variety typica); Coquilhatville (M. Laurent; variety subglabra). Eastern Congo Forest: Yambuya (M. Laurent; Solheid; variety subglabra and variety glabra); Avakubi (January 13, 1914; J. Bequaert; Coll. No. 1917). De Wildeman’ s figure of a flowering live plant (1907, p. 160, fig. 5) shows that the very large, solitary, terminal flowers are pendent. Ac- cording to the same author, it belongs to the group of R. malleifera (Hooker), which species, however, differs in the absence of ant-swellings, the smaller corolla with much denser and longer tomentum on the tube, and the villosity of the stem. Randia physophylla K, Schumann Randia physophylla K. ScuuMann, 1899, Engler’s Bot. Jahrb., XXVIII, pt. 1, p. 64. FE. De Witpeman, 1903, ‘Etudes Flore Bas- et Moyen-Congo,’ I, p. 81; 1907, ibid., II, pp. 74 and 164; 1912, ibid., III, pt. 3, p. 487. Ta. anp H. Duranp, 1909, ‘Sylloge Flor. Congol.,’ p. 260. “Leaves very short petiolate or subsessile, oblong, short and sharply acuminate, broadly cuneate at the base, subcordate and auriculate below, the earlets excavated and glandular; leaves very glabrous on both sides, resinous and very shiny. Ovary, to judge from the fruit, globose, glabrous and scabrous. Calyx tubular and irregularly 5-lobed, the lobes costate and scabrous. Corolla pentamerous, each of the 5 lobes divided in the upper part into obovate, obtuse, carnose lacinie. Anthers curved, broad. Style exserted for a long distance out of the corollar tube; subclavate and . sulcate in its upper part. Berry globose, crowned by the calyx. “The petiole is hardly 3 to 4mm. long. The blade has a length of 30 to 35 em. and a width in the middle of 12 to 14 em.; it is crossed on each side of the midrib by 23 to 25 heavy lateral veins, which are visible on both upper and under surfaces; the leaf is shiny chestnut-brown in dried condition. The two semiglobose glandular cups at the base of the leaf are 5mm. deep. The glands of the stipules must secrete an abundance of resin, for it fairly drenches the leaves and forms a crust at the base of the petiole. The calyx has a length of 3 to 3.5cm. The corolla is very fleshy, 18 to 19 em. long, of which 15 cm. is the length of the tube. The stamens are 1.3 em. long. The style exceeds the corollar tube by about 3cm. The berry has a diameter of 2.5cem.” (K. Schumann, 1899). pe ere oenenn roe 465 SS Batangs (Dinklage); Bipindi (Zenker). Belgian Congo: Lower Congo: Kisantu; Lukaya (Gillet); Sanda (Oddon); ¢ 0 Region (Butaye); Leopoldville (April 15, 1915; J. Bequaert; Coll. No. ssa and Upper Congo: Ikela (Jespersen); Eala (M. Laurent); Lubi ." ee? Eco.iocy or Randia Luje 4s Fhe eesocies was discovered by Luja in 1903 along the Sankuru River oo apa Congo. According to De Wildeman (1904a, pp. 282-284; )4b) its myrmecodomatia are very similar to those of R. myrmecophyla ribed in detail below. They consist of spindle-shaped swellings of the rnodes, about 2 to 3 cm. thick and hollow; one or two apertures, in widest part of the swelling, lead into the cavity. There are, in addi- on the leaves, in the axils of the nerves acarodomatia in the form of ui Pouches. The ants found by De Wildeman in the domatia of om have not been identified. Ecotocy or Randia myrmecophyla On only one occasion did I observe this species in the field. At abi, in January, 1914, a specimen was found in the primitive Rain est, in a rather dry place. It grew as a bush with very broad and glabrous, smooth leaves; the blades were as much as 30 cm. long d 15 cm. wide and borne on a petiole sometimes 4cm. long. The large, ite, pendent flowers were very striking. The plant agreed perfectly 1 all particulars with De Wildeman’s descriptions, photographs, and drawings of R. myrmecophyla. _ The myrmecodomatia of this Randia (Fig. 93) are quite peculiar, ing elongate, regular, spindle-shaped swellings on the middle portion the internode, and extending about half its length. These expansions n to occur in all the internodes of the various branches, usually resent one circular aperture, occasionally two or three, near the middle one side, and are inhabited by ants. A longitudinal section shows the ior of the enlarged part of the internode to consist of a spacious eavity, 10 to 12 em. long and 6 to 7 mm. wide, which stops a long distance ‘Artem the nodes. Hf as ants I found in the myrmecodomatia of the specimens at Ava- kubi belonged to a small species of Crematogaster recently identified by arlieg as C. rugosa (André). Each cavity apparently contained its m formicary with brood, and in many instances was divided into a series of chambers by transverse walls of brown, malaxated pith débris. Sometimes one hollow would thus be separated into four successive compartments communicating by one or two holes pierced through the 466 Bulletin American Museum of Natural History [Vol. XLV Fig. 93. Randia myrmecophyla De Wildeman: a, portion of branch showing swollen internode inhabited by ants; 6, longitudinal section of this internode; e, entrance to cavity; p, partitions built by ants. Drawn from life at Avakubi, January, 1914; one-half natural size. partitions; even then, there would usually be only one external aperture to the domatium. By means of these dividing walls the ants undoubtedly make a much more efficient use of the hollow internodes, for it has been observed that in such cases the larve and pup are kept toward the OE Wieser; Ants of the Belgian Congo 467 up Sits dl the antain thie Randa. Crematogaster rugosa is a small and timid ant and probably does not > its pe much protection. Even when the branches containing i s are shaken, the inmates do not leave their retreats. The ci on near Avakubi, though settled by ants, had its leaves badly Em. Laurent, the distoveree of this Randia, recognized its myrme- : — cophily i in the field. He found an unidentified ant and coccids in the “ swollen internodes. I am not aware that additional information on this } plant has been published since, but Kohl in later years has collected from 5 its domatia specimens of Camponotus foraminosus Forel and Cataulacus __ weissi Santschi (Forel, 1916, pp. 427 and 443). Eco.iocy or Randia physophylla E I found a specimen of this species in a forest gallery near Leopold- ville, in April, 1915. It was a small tree, with very large leaves, about 46 em. long and 27 cm. wide, on short petioles (1em.). The young leaves, before complete expansion, are viscose, being covered with a resinous, “sticky substance. The large, showy flowers are erect; their calyx ends in broad lobes; the corolla, about 26 cm. long, is dirty white in its upper part and greenish white in the tubular, lower portion. The egg-shaped fruit is 6 cm. long without the persistent calyx, 4 em. thick, and de- prived of ribs. ______This species has no swellings on its branches and the stem is never . omer nor inhabited by ants. At the base of the leaf-blade (Fig. 94), _ on both sides of the midrib, there is an evagination of variable size, con- ~ vex on the upper surface of the leaf, broadly open below. On some blades re consists of a mere inflation of the leaf-base, whereas in others it may be 4 to 6 mm. deep and pouch-like, 5 to 8 mm. long and 6 to 7 mm. __ broad. In all cases, however, on looking into it from the under side, _ one finds in the bottom, close to the midrib, a conspicuous pale brown tn which secretes a sweet substance. On some of the leaves of the _ Specimen I examined near Leopoldville, a number of ants, Crematogaster africana subspecies laurenti variety zeta (Forel), had taken possession _ of these distended nectaria, closing the opening on the under side with a 3 _ tent of fine, agglutinated, dark brown vegetable fibres. Frequently they __ were accompanied by coccids. Never having seen queens or brood of the _ ants in the leaf swellings, I can not regard these structures as forming _ part of the nest. Ants of the same variety occasionally build fibrous shelters over coccids which are fixed on the fruits of this Randia. 468 Bulletin American Museum of Natural History [Vol. XLV From the foregoing it is evident that Randia physophylla is not a true myrmecophyte in the sense generally meant by thisterm. Yet its rela- tions with ants are not without interest, for here we have a primitive stage leading to the production of true ant-pouches such as those of Scaphopetalum Thonneri, Cola Laurentii, and certain South — Melastomacee. Fig.94. Randia slgvapiulie K. Schumann: a, base ofleaf-blade with thetwo swellings, seen from above, natural size; b, cross section of this base, one and one-half natural size; the nectarium is placed inn. Drawn from life at Leopoldville, April, 1915. Piectronia Linn2us Plectronia Linn.xvus, 1767, ‘Mantissa Plant.,’ I, p. 52. K.Scaumann, in Engler and Prantl, 1891, ‘Die Natiirl. Pflanzenfam.,’ IV, pt. 4, p. 91. Canthium Lanaatecx, 1783, ‘Encyclop. Méthod.,’ I, p. 602. Hiern, 187, in Oliver, ‘Flora of Tropical Africa,’ IIT, p. 132. Shrubs or trees, often climbing or clambering bushes, occasionally spinous, with opposite leaves and branches, and acuminate stipules from a broad, often sheath- like base. Frequently some of the branches are sarmentose, hooked or winding; or the plant emits whip-like shoots, often many meters long, somewhat compressed, leaf- less or with small leaves, furnished with heavy, more or less recurved spines; these shoots trail along the ground or work their way up the trees. Flowers small, axillary, in dense cymes or umbels, or short panicles or clusters. Calyx-tube short, turbinate, campanulate, or hemispherical; limb short, 4- or 5-toothed or cleft, or subtruncate, deciduous. Corolla coriaceous; tube rather short, exceeding the calyx, glabrous out- side, hairy with a ring of deflexed pilose hairs or rarely glabrous inside; throat rather constricted or dilated, often bearded; lobes 4 or 5, rarely 6, ovate or lanceolate, re- flexed, usually glabrous, valvate in the bud (toward the apex sometimes induplicate- valvate). Stamens 4 or 5, rarely 6, exserted, inserted at the mouth of the corolla; filaments short; anthers ovate, or oval, or lanceolate, acute or obtuse, usually sub- sagittate at the base, as a rule glabrous, fixed at the back. Ovary 2-celled, fleshy. Style flexuous, filiform or thickened, exserted or equalling the corolla, usually gla- brous. Stigma capitate, calyptriform or mitre-shaped, sometimes bifid at the tip, , often suleate. Ovules solitary, pendulous, orthotropous, the micropyle directed upward. Fruit a drupe, didymous, subdimidiate, or globose, 2-celled or by abor- tion 1-celled; stones 2 or 1, sometimes subrugose. Seeds pendulous, solitary, nearly straight or curved, sometimes bent into the form of a horseshoe round the placenta; « testa membranous; albumen fleshy, sometimes ruminated; embryo cylindrical, nearly straight or curved, axile; cotyledons short, radicle superior. Wheeler, Ants of the Belgian Congo 469 se- to five-celled ovary and a drupe containing three to five one- d stones or consisting of one three- to five-celled stone. Many s of the genus Psychotria Linneus, too, assume appearance and er of growth of certain Plectronix, but differ in the ovule being -anatropous, with the micropyle opening downward, and in the or radicle of the embryo; on this account Psychotria is placed in a Plectronia connata De Wildeman and Durand nia connata &. De WiLpEMAN AND Tu. Duranp, 1899, Bull. Soc. Bot. Belgique, XXXVIII, 2, p. 201; 1901, ‘Reliquie Dewevreanw,’ p. 122. E. De Wiive- MAN, 1904, ‘ Etudes Flore Bas- et Moyen-Congo,’ I, pt. 2, p. 204; 1912, ibid., III, pt. 3, p. 488. H. Kou, 1909, Natur u. Offenbarung, LV, pp. 161-162. a “Tyree orshrub. Branches more or less terete, glabrous, incrm. Leaves opposite; e, the petiole 1 to 1.5m. long; ovate-elliptic; dark green and sparsely pilose »ecially on the veins on the upper side, on the under side brown and paler in dried mdition, sparsely pilose especially on the veins: the blade more or less decurrent ong the petiole; abruptly and short acuminate at the apex, the acumen about 5 m. long; rounded at the base; 7.5 to 10 em. long and 4 to 5 cm. broad; on each side th about 7 lateral veins, which anastomose in curves before the margin and are united th the smaller veins. Stipules soon deciduous. Inflorescences axillary, 3 to 4 em. ng and about 3.5 cm. broad, opposite, dichotomous, bracteate at the base of the dichotomies; the bracts more or less broadly connate at the base, ovate-acute, 3 to _ 4mm.long. Flowers pedicellate, the pedicel about 3 mm. long. Calyx campanulate, ' ‘S+oothed, the teeth short. Corolla 5-lobed; the tube 2.5 mm. long and about the ddle 1.5 mm. broad: the lobes reflexed, ‘about 2 mm. long and 1 mm. broad, ovate- ate. Stamens 5, not exserted; the filaments short. Style filiform, exserted over a ; ng aistadieo, about 1 em. long, slabacien: the stigma capitate, short lobulate at the _ apex, about 0.5 mm. thick” (De Wildeman and Durand, 1899). ___— Belgian Congo: Lower Congo: Sele River (Butaye). Middle and Upper Congo: | Bolengi (M. Laurent); Likimi (Malchair); Lomami River, type locality (Dewévre). _ This species is easily recognized by the united bracts which form a F iheat at the base of the ramifications in the flower panicles. According to Marcel Laurent, the natives at Bolengi call this plant “Boka na ~ pombo” which means “ant-village.’’ It is possibly one of the myrme- a members of the genus, and has therefore been included here. “¥ 470 Bulletin American Museum of Natural History [Vol. XLV Plectronia glabriflora (Hiern) Plectronia glabriflora (Hiern) K. Scuumann, 1895, in Engler, ‘Pflanzenwelt Ost- afrikas,’ C, p. 386. H, Krause, 1911, ‘Wiss. Ergebn. Deutsch. Zentr. Afr. Exp. (1907-1908),’ II, p. 326. Canthium glabriflorum Hrern, 1877, in Oliver, ‘Flora of Tropical Africa,’ IIT, p. 140; 1898, ‘Catalogue Afr. Plants Welwitsch,’ II, p. 474. K. Scuumann, 1891, Ber. Deutsch. Bot. Ges., IX, pp. 61-62. H. Kon, 1909, Natur u. Offenbarung, LY, p. 162. Canthium polycarpum Scuwetnrurtn Mss., 1877, ex Hiern, in Oliver, ‘Flora of Tropical Africa,’ III, p. 139. “An unarmed tree, 40 to 50 feet high, with palm-like habitus; branches eect patent, obtusely angular, glabrous or somewhat hispid. Leaves oval, shortly and abruptly acuminate, with a broad somewhat excavated base, thinly coriaceous, scabrous-hispid or glabrate above, turning reddish when dry, more or less hispid on the veins beneath, 3 to 5 by 1% to 2% in.; lateral veins about 7 to 8 pairs; petiole 4% to % in., hispid or glabrate; stipules ovate, 4 to %in. long. Flowers \ in. long (exclu- sive of the style), on short puberulous or glabrate pedicels, many together, in dense dichotomous globose panicles of 1 to 1}4 in. diameter; common peduncle glabrate or puberulous, short or ranging up to 4 in., spreading, sometimes unilateral. Calyx- tube glabrous; limb truncate or obscurely toothed, glabrous or ciliolate. Corolla glabrous outside, bearded inside; lobes 5, subobtuse. Disk glabrous. Stigma elon- gate-calyptriform, much exserted ” (Hiern, 1877). San Thomé: at 1000 feet (Mann: Welwitsch). Southern Nigeria: Old Calabar (W. C. Thomson). Cameroon: Barombi (Preuss). Belgian Congo: Kwidjwi Island near Mgaturo in the forest (Mildbraed). North- eastern Region: Nabambisso River in the Niam-Niam Country (Schweinfurth). Angola: Malange (Buchner). ; Preuss, who observed this species in Cameroon, calls it an “ant- plant.’’ According to Schumann (1891), the ants live inside the hollow stem and probably also in the horizontal branches. No other observa- tions have been made on this form and its description is reproduced here chiefly on account of its possible identity with P. Laurentit. Plectronia Laurentii De Wildeman Plectronia Laurentii &. De WitpeMan, 1906, ‘Mission Emile Laurent,’ III, pp.. 294-296, Pls. xcvi—xcrx; 1907, ‘Etudes Flore Bas- et Moyen-Congo,’ II, pt. 2, p. 174; 1908, ibid., II, pt. 3, p. 348; 1910, ibid., III, pt. 2, p. 294. Tu. anp H. Duranp, 1909, ‘Sylloge Flor. Congol.,’ p. 268. H. Kou, 1909, Natur u. Offenbarung, LV, pp. 160-161. “Shrub reaching a height of about 2.25 m., with quadrangular stems showing opposite the leaves a groove pierced with openings which allow ants to enter the inter- nodal cavity. Branches spreading, glabrous when full-grown. Leaves oppposite, petiolate; the petiole reaching a length of 2 to 3 cm., ciliate on the sides; the blade wedge-shaped, rounded or almost subcordate at the base, very broadly cuneate or acuminate at the apex, more or less coriaceous, 7 to 28 em. long and 6 to 16 em. broad, Wheeler, Ants of the Belgian Congo 471 8 to 12 laterat-veinion each side of the midrib. Leaf-blade with scattered hairs, e ress Jon the upper side, somewhat more abundant on the under side, especially on 4 ‘the vein , which are villose, scabrous, and ciliate on the margins. Stipules triangular. rulate, about 1 cm. long. Inflorescences axillary, opposite, reaching a length | “of 5 to Gem. and about equally broad, Common peduncle short, glabrous, 3 to 8 mm. long dichotomous ramifications which bear below each bifurcation a more or less T ‘ring of bracteoles. Flowers fasciculate at the end of the ramifications; th e Pp edicel short, slender, accrescent on the fruit and sometimes reaching a length of Calyx with feebly widened limb, superficially denticulate, glabrous. Corolla Pee, long, glabrous externally, with 5 lobes. Style unknown in adult condition. te ceteces oe high, 9 mm. broad, and 4 mm. thick, sometimes one-celled by abortion” (De Wildeman, 1906). oe I eidlcien Congo: Middle and Upper Congo: Bokala; Irebu; Chumbiri; Bolengi; Eala (M. Laurent); Lukolela (Pynaert); Bolombo; Nouvelle-Anvers; Malema (BE. an M. oe. Eastern Congo Forest: Romée (H. Kohl); Tshopo River near itanleyville (March 6, 1915; J. Bequaert; Coll. No. 7042); between Walikale and t (village of Pale, January 12, 1915; J. Bequaert; Coll. No. 6585); Paku It would seem from the description that Plectronia glabriflora rn) is rather closely allied to, if not identical with, P. Laurentit; it is hardly to be expected that a plant so commonly found throughout _ the Congo Basin is absent from Cameroon and Angola. a Eco.oay or Plectronia Laurentii a ‘ The following notes were made on specimens in the forest region 2 Walikale and Lubutu (near the village of Pale, January, 1915; ol No. 6585) and along the Tshopo River near Stanleyville (March, ‘191 5; Coll. No. 7042). This plant is a bush or small tree, about 4 to 7 ‘meters high, with an erect, straight trunk, bearing from a short distance ve the ground regularly opposite, nearly horizontal branches. The st striking feature is the squareness of the limbs which, on the younger ts of the plant, show four very pronounced longitudinal grooves inter- J at the nodes only. Above the nodes, where the myrmecodomatia } dadated, the depressions expand into four broad, flat sides, the stem sing almost regularly square on a cross-section. Older branches often ae more cylindrical, only slight traces of the longitudinal furrows being left. The leaves are short petiolate, large and broad, as much as em. long and 16 em. wide. Thestipules drop off early. While the stalk and limbs are glabrous and smooth, the leaves are slightly hairy and & “Both the trunk and lateral branches of P. Laurentii were inhabited _ by ants, of the form Crematogaster africana subspecies laurenti (Forel) _ in the case of the specimens from the Tshopo River, and of the variety > 472 Bulletin American Museum of Natural History [Vol. XLV ~ Gow. Ya! pI, ert b Fig. 95. Plectronia Laurentii De Wildeman: a, portion of branch with swelling above node in- habited by ants, showing apertures (e) to the domatium; b, longitudinal section of this myrmeccdoma- tium. Drawn from life at Pale, between Walikale and Lubutu, January, 1915; natural size. zeta (Forel) of that race in those found between Walikale and Lubutu. The older stalks of the plants are not much swollen, but the medullary cylinder is almost completely excavated, even the partitions at the nodes being occasionally perforated. In younger branches the various myrme- codomatia are more distinct; they are then moderately pronounced, quadrangular swellings, with the flat sides separated by slightly raised, obtuse ridges (Fig. 95a and b). They usually extend the basal two-thirds Wheeler, Ants of the Belgian Congo 473 2 of each internode and very gradually disappear in the upper part toward the node. The internal cavity is quite spacious, 6 to 7 cm. long and 10 = to 15mm. wide. An examination of very young shoots shows that the __ swellings are normal productions of the plant and that the cavities & originate through the drying of the pith before the ants gnaw apertures. _ Hollows inhabited by these insects present a number of small, circular exit holes, which in my specimens were commonly located on any one of _ the sides. According to Kohl (1909, p. 161), they are placed on the surfaces facing the lower leaf pair, but this is far from being the rule. _ Many swellings, especially on the younger branches, have only one - aperture; more commonly there are 2 to 4 entrances to each cavity, and in some cases as many as 12 to 15. _ At least on the younger portions of the plant, every domatium con- tains a complete ant colony, with a queen, workers, and brood. Fre- quently coccids also are present and those found by Kohl near Stanley- ville, together with Crematogaster, in the swellings of P. Laurentii have been described by Newstead (1910, p. 18) as Hemilecanium recurvatum. A number of such scale insects were also fixed on the outer surface of the stem, especially near the nodes, within tents of plant-fibres built by the ants and often communicating with their cavities. Kohl (1909, p. 161) further mentions that some of the internodes of a Plectronia in that locality were occupied by small, white caterpillars, while others were inhabited by ants. Plectronia Laurentii was discovered at various places along the banks of the Middle and Upper Congo by Em. Laurent, who has given in his field-notes a good account of its relations with ants (De Wildeman, 1906, pp. 294-296). Much additional information on this species has been published by H. Kohl (1909, pp. 160-161). These observations agree in most details with mine. The ants, all of the genus Crematogaster, found associated with P. Laurentii are evidently facultative inhabitants of these plants. The fol- lowing forms have been recorded thus far: Crematogaster africana (Mayr), variety. Belgian Congo; found by Em. Laurent (Kohl, 1909, p. 161). C. africana subspecies laurenti (Forel). Found by Laurent at Bokala (Kohl, 1909, p. 160), by Kohl at Isangi and Stanleyville (Forel, 1909, p. 60), and by myself near the Tshopo River. C. africana subspecies laurenti variety zeta (Forel). Between Wali- kale and Lubutu (J. Bequaert) and in the Congo (Kohl; see Forel, 1909, p. 70). 474 Bulletin American Museum of Natural History (Vol. XLV C. africana subspecies winkleri (Forel). Belgian Congo (Kohl; see Forel, 1909, p. 69). C. africana subspecies winkleri variety fickendeyi (Forel). With re- gard to this variety Forel (1916, pp. 409-410) writes: Kohl has collected various forms transitional between the race winkleri and the variety fickendeyi, on one occasion in a nest, probably usurped, of T'etramorium aculea- tum, also in myrmecophilous plants or in termitaria. His No. 68 bears the following interesting remark: ‘Ant from plants. Lives in and on the myrmecophyte Plec- tronia Laurentii De Wildeman. Five meters above the ground the trunk bore a carton nest, 40 to 50 em. high, of this ant. But it inhabits at the same time all the hollow branches of the plant. Makanga on the Okiavo River.” One may conclude from this that there is no absolute contrast between the carton nest of buchnerit and the habit of living in hollow stalks. EcoLocy or Unipentirrep Arrican Spectres or Plectronia In addition to the species just studied, I have found caulinary swellings inhabited by ants on a number of rubiaceous plants which are provisionally regarded as belonging to the genus Plectronia. It is pos- sible, however, that one or more may be species of related genera, such as Vangueria, Grumilea, or Psychotria. At any rate, I have been unable to identify them with any of the described African Rubiacex and they may even represent forms new to science. Their correct identification will undoubtedly be made later when the study of my herbarium, now in the hands of Mr. De Wildeman, Director ot the Brussels Botanic Garden, is more advanced. Plectronia species A.—This species was first observed on the forested banks of the Aruwimi River near the village of Bafwalipa, between Bomili and Avakubi (December 29, 1913; Coll. No. 1696). It also occurred in the Ituri Forest, near the village of Tete, between Penge and Irumu (February 22, 1914; Coll. No. 2567), and, in company with Mr. Lang, I came across it again along the Tshopo River near Stanleyville (March 6, 1915; Coll. No. 7043). It is a climbing, much-branched bush of the forest, with simple, opposite, short petiolate or subsessile leaves, which are asymmetric and cordate at the base. There were no thorns or spines on the specimens I examined. The entire plant—leaves and stems —is abundantly covered with long, erect, brownish hairs. The flowers are small and clustered in corymbs in the axils of the leaves. Myrmecodomatia (Fig. 96) are found on some of the branches only. They consist of spindle-shaped swellings on the lower third of an inter- node, are about 30 mm. long and 8 mm. thick, and placed immediately above the node. The domatium is a spacious, rather thin-walled cavity. When inhabited, it is almost wholly cleaned of medullary tissue and com- Wheeler, Ants of the Belgian Congo 475 municates with the outside by means of a broad, irregular aperture, a _ placed about the middle of the swelling. It was noticeable that leaves at the base of the expanded area are shorter and more heart-shaped than 4 7 tregaordhi variety plectronie Wheeler in domatia of this species along the Tshopo River, while I a _ elsewhere on the plant. In this case, too, the enlargements are normal productions and their inner cavity originates through the drying of the pith and without the agency of ants. Mr. Lang collected specimens of Cataulacus found a few workers of Engramma kohli Forel in other swellings of the same plant. Plectronia species B.—-I obtained this Plec- tronia in the Rain Forest near Avakubi (January 10, 1914; Coll. No. 1871) and Penge (February 14, 1914; Coll. No. 2478), in both cases on the of the Ituri River. It is a creeper whose main stem, about 20 mm. thick near the ground, hangs freely in true liana-fashion between the bushes, while the branched upper part spreads its leaves over the crowns of low trees and under- growth. Evidently closely related to the pre- , fy oe sh eon ceding form (species A), it differs chiefly in being tongitudinal section; showing more sparsely hairy, and in having myrmeco- Scien Maier ta the domatia of another shape. Furthermore, the main cavity. Drawn from life at stalk bears at the nodes strong thorns placed in B*fwalipa, between Bomili pairs and formed by the hardened bases of satura on aborted branches. In this species ants inhabit the thickened main stem, as well as the branches. On the latter the myrmecodomatia (Fig. 97a) are elongate, spindle-shaped swellings of the nodal region, extending about as far below as above the node. They are almost completely excavated and rather thin-walled; their internal cavity is 9 to 11 cm. long, 8 to 10 mm. wide, and even extends a short distance into the slightly swollen bases of the opposite branches. In this case too the expansions are normally present on the plant and their medullary tissue soon dries up, the ants merely piercing the orifices and removing the remains of pith. When occupied by these insects, the domatia usually havea number of apertures, placed above the node in an irregular, longitudinal row; there is often an exit hole also at the enlarged base of the side twigs. Even the nodes of the main stalk (Fig. 96b) are tenanted but, owing to the thickening of 476 Bulletin American Museum of Natural History (Vol. XLV \ Fig.97. Plectronia species B; a, portion of younger branch in longitudinal section, showing myrme- codomatium at the node which also extends into the base of the lateral ramification; 6, portion of main stalk, showing shape of domatium in older parts of plant; e¢, apertures leading into the cavity. Drawn from life at Avakubi, January, 1914; natural size. the woody cylinder, are but slightly or not at all swollen and their inner cavity is much reduced (3 to 4 em. long, 5 to 7 mm. wide); they usually present two openings placed on a crateriform elevation, one above each of the nodal thorns. Frequently there are scars of other perforations which have been closed by callus growth. The ants found inside this Plectronia belonged in both localities to a small, unidentified species of Crematogaster which can hardly give pro- Forest, near Lesse (June 15, 1914; Coll. ___No. 4753), I came across a creeper whose - many hanging branches had covered the bushes at the edge of a clearing. It is per- - form (species B), possessing most of its gen- i , — — eS se _ 7 ' Wheeler, Ants of the Belgian Congo 477 tection to its host:—Even though most of the domatia were inhabited, _ the leaves had been eaten by caterpillars and both young branches and leaves bore numerous insect galls—elongate, pear-shaped swellings end- ing in a recurved tail-like apex and on one side of the tail with a small exit hole leading into a central chamber; their outer surface covered q . a with many erect, brownish-red hairs; all the galls seen were ai Plectronia species C.—-In the Semliki haps specifically identical with the preceding eral charaters. Yet the domatia are suffi- ciently different in shape to deserve separate _ The ant-swellings (Fig. 98) are short € and broadly spindle-shaped, and occupy the ~ lower part of the internodes of most of the branches. The inner cavity is very spacious, HY 6 to 8 cm. long and 15 to 20 mm. wide, con- Fig. 98. Plectronia species C. tinues a little below the node, and extends Lonsitudinal section of swollen node also into the slightly swollen bases of the rome po br rege side branches. A peculiarity of this Plectronia peed “me Sages is that the domatia lack circular apertures, Lesse, June 1914; natural size. but communicate with the outside by means of two long slits, placed opposite each other in the upper part of the swelling, above each of the side branches. Often these openings are partly closed by callus growth. The plant at Lesse was inhabited by populous colonies of a small Crematogaster with a queen, numerous work- ers, and brood at various stages; also coccids which were fixed on the cal- lus tissue near the inner margin of the slits. Cuviera de Candolle Cuwiera A. pe Canporie, 1807, Ann. Mus. Paris, IX, p. 222; 1830, ‘Prodromus Regn. Veget.,’ IV, p. 468. Hrern, 1877, in Oliver, ‘Flora of Tropical Africa,’ Ill, p. 156. K. Scrumann, 1891, in Engler and Prantl, ‘Die Natirl. Pflanzen- fam.,’ IV, pt. 4, p. 94. Glabrous shrubs or small trees, rarely with puberulent young branches. Leaves usually large, broadly ovate, entire, opposite, coriaceous or leathery; stipules apicu- late, united into a short sheeth between the bases of the petioles. Flowers polyga- 478 Bulletin American Museum of Natural History (Vol. XLV mous, with large foliaceous bracteoles, in many-flowered, axillary panicles. Sepals 3 to 5, almost free or shortly united at the base, foliaceous, spreading, often unequal, persistent, much longer than the petals. Corolla hypocrateriform, with a short, straight tube furnished inside with a ring of deflexed hairs, and large, fleshy lobes. Stamens 5, exserted, placed on the mouth of the corollar tube. Ovary 3- or 5-celled, each cell with one ovule. Ovule suspended, with upper micropyle and flattened funiculus. Style with a semiglobose, cap-shaped or mushroom-shaped, suleate stig- ma. Fruit an obovate drupe, often oblique or faleate, distinctly ribbed, with 3 to 5 : hah ae ty ; BS ara [ | Y f LS ee vt Is; t Bee. } ey vi . we OO ei Gay & SHU ; care | 7 ‘ 32 ~€ %. T VS pr , Ea ' mw i Ge SSS ch Sams : R } ial | 4% Map 46. Distribution of Curera, a genus of myrmecophytic plants. Cuviera is a strictly African genus, of which fourteen species have been described. Its general distribution is shown on Map 46. The geno- type, C. acutiflora de Candolle, is found in Upper Guinea. Only one form, C. australis K. Schumann, has been described from South Africa. All the others occur within the limits of Engler’s Western Forest Province, either in the Rain Forest proper or on the forested river banks of the adjoining Savannah, below 3000 feet. With the possible excep- tion of C. australis, all the members of the genus may be myrmecophytes Wheeler, Ants of the Belgian Congo 479 and their descriptions have, therefore, been reproduced here. Some of these eed species are perhaps mere synonyms. Cuviera acutifiora de Candolle a pe Canpo.ue, 1807, Ann. Mus. Paris, IX, p. 222; 1830, ‘Pro- _ dromus Regn. Veget.,’ IV, p. 468. Benraam anp Hooker, 1849, ‘ Niger Flora,’ -p. 407, Hiern, 1877, in Oliver, ‘Flora of Tropical Africa,’ III, p. 156. - Cwoiera africana SPRENGEL, 1825, ‘Syst. Veget.,’ I, p. 760. “A glabrous shrub, 15 to 20 ft. high. Branches terete, divaricate, supra-axil- : “lary. Leaves oval-oblong, acuminate, subequal and rounded or somewhat narrowed __ at the base, coriaceous, glossy, rather or scarcely paler beneath, 4 to 10 by 1% to 4 in.; some 4 to 6 in. wide (Bentham); lateral veins about 6 to 10 pairs, not conspicuous; _ petiole to 4 in.; stipules ovate, apiculate, }; in. long, connate and sheathing below, keeled, hairy within. Flowers greenish, % to % in. long in bud, on short, slender os pedicels, very numerous, in ample, divaricately branched, rather lax, axillary and _ terminal, shortly pedunculate, dichotomous panicles of 2 to 6 in. diameter; brac- teoles elliptic-linear, *{ to 1 in. long, accrescent. Calyx green; segments % to } in. long, linear-oblong, spreading, persistent. Corolla green and orange; segments lanceolate, caudate-acute, *{ in. long, spreading. Ovary 5-celled; style glabrous. Fruit _ obliquely egg-shaped, *; to % in. long, obtusely 5-sided; pyrenes 5 or fewer” (Hiern, Sierra Leone, type locality (Smeathman). | Ivory Coast: Grand Bassam (Th. Vogel). _ Cameroon: Ambas Bay (Mann). 4 Cuviera angolensis Hiern : Cuviera angolensis Hiern, 1898, ‘Catalogue Afr. Plants Welwitsch,’ II, p. 483. E. De Witpeman anv Tu. Duranp, 1901, Bull. Herbier Boissier, (2) I, p. 826; 1901, ‘Reliquia Dewevreanz,’ p. 124. fh. De Witpeman, 1904, ‘tudes Flore Se Bas- et Moyen-Congo,’ I, p. 205; 1907, ibid., II, pp. 78 and 173; 1908, ibid., IT, 8, p. 348; 1910, ibid., ITI, 2, p. 295; 1912, ibid., IIT, 3, p. 489; 1906, ‘ Mission Emile Laurent,’ II], pp. 296-299, Pl. cv1. H. Kont, 1909, Natur u. Offenbarung, LV, . spp. 163-166. Tu. anp H. Duranp, 1909, ‘Sylloge Flor. Congol.,’ p. 271. (K. _ ~~ Sehumann, in Engler and Prantl, 1891, ‘Die Natiirl. Pflanzenfam.,’ IV, pt. 4, p. 94, fig. 33, without description; also mentioned without description by K. Schumann, 1890, Verh. Bot. Ver. Brandenburg, XX XI, 2, p. 121, and 1891, Ber. . Deutsch. Bot. Ges., IX, p. 56). a “A small glabrous pyramidal tree, 12 to 20 ft. high, or in cultivated fields (arimos) usually only 8 to 12 ft. Sap milky. Trunk slender, straight, destitute of branches 2 below, but densely armed with opposite, decussate, strong, very acute, quite patent a opines of 1 to 2 in. in length. Branches and branchlets green, the latter swelled at the nodes. Leaves long, opposite, usually cuspidate at the apex, oblique and rounded at the base, papery, smooth, 4 to 9 in. long by 144 to 4 in. broad, dull-green above, paler beneath, those on the older branches pendulous; petiole % to in. long; lateral veins about 8 on each aide of the midrib, rather slender and beneath conspicu- ous. Stipules sheathing, keeled, acuminate, about % to \ in. long. Inflorescence axillary, branched, 2 to 4 in. long, pale yellow-greenish outside throughout except a bright rosy stellate patch about the naked throat of the corolla; pedicels very short; 480 Bulletin American Museum of Natural History [Vol. XLV common peduncle } to 1 in. long; bracteoles sub-linear, ranging up to 1 in. in length. Calyx including and adnate to the ovary; tube short, campanulate-ventricose, obtusely 3- to 4-angular, deeply 3- to 5-lobed; the segments elongate-lanceolate, un- equal in length, bract-like, exceeding the corolla, herbaceous-green, 4 to 2 or 3 in. long. Corolla shortly salver-shaped, fleshy-coriaceous, deep herbaceous-green out- side; tube short, bright-red inside, at the base inside with a ring of shiny silvery hairs directed downward; limb 5-cleft, shortly rotate; segments lanceolate or ovate-acumi- nate, rigid, green, expanded in a stellate manner in full flower, valvate at the base in estivation; the tips long, acuminate or subulate, contorted in the bud. Stamens 5, inserted in the sinuses of the corolla-lobes around the ring of hairs; rigid, exserted; filaments compressed-cylindrical, fleshy, curved-patent at the time of flowering: anthers oyate, cordate, introrse, 2-celled, obtuse at the apex, basifixed; cells sepa- rate at the base, cohering at the apex longitudinally, yellow. Ovary adnate to the calyx-tube, 5-celled; cells 1-ovuled; disk a little elevated, flat; style thick, columnar, rosy, densely pilose; stigma mitriform, large, obtuse, stigmatose and cleft at the apex. Fruit oblique, deeply furrowed, about 1 in. long, crowned with the more or less per- sistent calyx-limb or with its remains’ (Hiern, 1898). Angola: Golungo Alto: “among the mountainous forests of Alto Queta,”’ type locality (Welwitsch). : Belgian Congo: Kisantu (Gillet). Kwango: Kikwit (Lescrauwaet). Middle and Upper Congo: Lukolela (Dewévre); Likimi (Malchair); on the left bank of the Congo below Bolombo; Malema; Lie (Em. Laurent); Irebu (Pynaert); Eala (M. Laurent). Northeastern Congo forest: Isangi; Tshopo River near Stanleyville (Em. Laurent); Romée (H. Kohl); Nala; Lifungula (Seret); Manyema (Berger). Cuviera australis K. Schumann Cuviera australis K. ScHuMANN, 1899, Engler’s Bot. Jahrb., XXVIII, 1, p. 78. “Shrubby, with rigid, divaricate, terete, glabrous branches: the young branches flattened and puberulent. Leaves with short petioles, oblong, ovate, or oblong- lanceolate, obtuse, rounded or acute at the base, glabrous above; on the under side softly puberulent on the primary veins, otherwise glabrescent; discolored, herbaceous. Stipules subulate or filiform from a broad base, not setose inside. Cyme twice, more rarely three times trichotomous, axillary, appearing below the leaves, peduncu- late, minutely puberulent, with very slender branches. Flowers pentamerous, pedicel- late. Ovary sub-semiglobose, slightly hairy, 5-celled. Calyx divided to near the base into foliaceous, subspatulate, elongate lobes. Corolla divided beyond its middle into five lobes, which are lengthened subtriangular and hirsute externally; tube glabrous on the outer side. Style exserted for twice the length of the tube, with 5-toothed stigma. “The flowering branch at hand is 30 em. long and 2 to 2.5 mm. thick at the base where it is covered with gray bark. The petiole is 3 to 6 em. long and very finely pilose; the blade is 3 to 5 cm. long, 1.1 to 2.7 em. broad in the middle, traversed on each side of the midrib by 5 or 6 stronger veins which are slightly prominent on both sides, black above, gray below. The stipules are 2 to 3 mm. long. The flowers are borne on finely pilose pedicels, 5 to 9 mm. long. The ovary is 2 mm. long, the calyx 7 to 8 mm.; its lobes are very obtuse and reach a width of 2mm. above. The corolla- tube is 3 to 4 mm. long; its lobes are 6 to 7 mm. long and are very finely pilose out- side. The anthers are a little over 1 mm. long and inserted on a filament of 0.5 mm. The style is exserted for 6 to 7 mm. out of the corollar tube. Sm igh en ini 5 i ah ls _- Wheeler, Ants of the Belgian Congo 481 by the much smaller flowers and leaves. I believe I should have distinguished two forms, one of the specimens is more hairy and has much smaller flowers. Schlechter thinks, however, that both specimens come from one and the same bush” (K. Schu- mann, 1899). te — Delagoa Bay, at 30 m. (Schlechter). Cuviera calycosa Wernham Ts eaiycens Wernuam, 1914, Journ. of Botany, London, LIT, p. 7. “Tyee 90 feet high, glabrous, nigrescent in dried condition, with terete branches Sei isks covered with grayish bark. Leaves parchment-like, elliptic or objong, small for the genus, shortly and narrowly acuminate, obtuse, acute at the base, glabrous; petiole very short. Stipules small, lanceolate, acuminate, caducous except for the broad base. Inflorescences having few flowers, dichotomous, rather loose; bracts oblong-lanceolate, obtuse. Calyx large, much exceeding the corolla; with uneven, _ ovate-lanceolate, acuminate and very acute lobes. Corolla with a broadly funnel- 8k to cylindric, rather short tube; its 5 lobes acuminate with long appendages and a few scattered, rather long hairs. Drupe very glabrous, crowned by the per- sistent limb of the calyx. _ “A remarkable species, the nearest affinity being clearly C. nigrescens (Scott- _ Elliot); the present species is distinct, especially in the very large calyx and small corolla. The leaves measure 10 to 11 cm. X 4 to 4.5 em., with petiole about 1 em. long; secondary veins 5 to 6 pairs; stipules 6 to 8 cm. long. Peduncle 3 em.; cyme 11 to 12 em. wide, 5 to 6 cm. long. Pedicel 5 mm.; calyx-tube minute, lobes 3 to 3.5 cm. X 4to7 mm. Corolla-tube barely 5 mm. long, and nearly as much in average breadth; lobes, flat part 4 to 5 mm., sete over twice that length. Berry 14cm. X l.lem.” (Wernham, 1914). ‘ _ “Youngest flowers white, older ones cream, oldest thin orange. Centre of flower _ greenish. Calyx-lobes bright green, with margin and sete white. Sete of corolla- _ lobes white; anthers dark-purplish brown; style white, stigma pale green” (Mrs. Talbot). 4 - Southern Nigeria: near Esuk Ekkpo Abassi in the Eket District (Mr. and Mrs. eh P. A. Talbot). Cuviera latior Wernham _ Cuviera latior Wexnxuam, 1918, Journ. of Botany, London, LVI, p. 311. . “A very glabrous shrub, with very smooth, subterete, moderately robust, striate _ branches, swollen and excavated at the nodes (apparently with a myrmecodomatium). Leaves large, parchment-like, broad, oblong, but little acuminate, cordate and very unevenly oblique at the base; petiole short, though distinct; primary veins conspicu- ous, 10 to 12 on each side. Stipules connate into a broad sheath, which is very short, arcuate above, and obscurely spiculate between the petioles. Flowers large for the genus, placed in loose, few-flowered, forked cymes; common peduncle much flat- tened; pedicels very short. Calyx with 3 lobes which are full of veins, broadly lanceolate, long acuminate, large and leaf-like. Tube of the corolla broad and very _ short, its 5 lobes oblong, very acuminate, subsetaceous and cucullate at the apex Ovary deeply suleate; style thick, densely and finely hispid. , 482 Bulletin American Museum of Natural History {Vol. XLV “Notable for the broad calyx-lobes and the large flowers. Leaves 20 to 26 em. < 8 to 9 em., with petiole 6 to 8 mm. at longest; stipule-sheath 2.5 mm. deep. Peduncle 2 cm. long, forking at the tip into two floriferous branches about 10 em. long. Calyx-lohes 3 to 3.5 em. long, and 1 em. or more broad. Corolla-tube barely 4 mm. long; lobes 1.6em. X 4mm. Anthers 2 mm. long. Style 1 em. long’ (Wernham, 1918). Belgian Congo: north of Boyeka (Nannan). Cuviera Ledermannii Krause Cuviera Ledermannii Krause, 1912, Engler’s Bot. Jahrb., XLVIII, p. 418. “Erect shrub or small tree, with slender, strong, glabrous branches and branch- lets, which are swollen, a little flattened and hollow at the nodes; bark smooth, dark brown or almost black in spots. Leaves large, short-petiolate; stipules broadly ovate, minutely acuminate at the apex, soon dropping, connate at their base into a short sheath which persists longer; petiole short, strong, grooved above to near its base; blade thick, coriaceous, very glabrous on both sides, oblong or elliptic-oblong, rather long acuminate at the apex, obtuse at the base or even shortly decurrent along the petiole; primary veins 9 to 12, slightly prominent above, more distinctly so below, running in an obtuse angle from the costa. Inflorescence axillary, short, with few flowers; bracts large, narrowly oblong, obtuse. Ovary semiglobose; lobes of the calyx large, narrowly oblong, acute, 2 to 3 times longer than the ovary; tube of the corolla cylindrical, scarcely broadened above, the lobes lanceolate-oblong, acute, as long as or longer than the tube; stamens with very short filaments, the anthers small, oval-oblong; style rather highly exserted above the tube of the corolla, crowned with a rather large, mitriform stigma. “The plant is a shrub or small tree; the branches which I have before me are covered with dark brown or almost black bark; they are 2 to 3 dm. long and 5 mm. thick at their base; the thickened, hollow swellings, above the nodes, are 7 to 9 mm. in diameter; they undoubtedly are inhabited by ants. The stipules are 8 to 10 mm. long, the petioles 1.2 to 1.6 cm. The blades in a dried state are brownish-green to gray- green and, including their apex of 1.2 to 1.6 em., are 1.8 to 2.5 dm. long, 7 to 11 em. broad. The inflorescences attain a length of 7em. The large bracts, which may reach a length of 1.8 em., in drying take on a leather-brown color, as do also the sepals. The ovary has a diameter of 2.5mm. The sepals are 7 to8 mm.long. The corolla, white in life, turns dark brown in drying; its tube is 4 to 5 mm. long, its lobes 5 to 6 mm. The filaments are about 0.8 mm. long, the anthers 1.2mm. The style, includ- ing a stigma of about 1.5 mm., measures 8 mm.” (Krause, 1912). Cameroon: near Nkolebunde on the Nanga-slopes in a rather sparsely wooded place, at about 200 m.; also near Malende in the vicinity of Nkolebunde in dense, high forest with little underwood, at 150 m. (Ledermann; in flower during October). The species agrees in most respects with C. physinodes K. Schumann, from which it is said to differ in “the branches, which are less strongly flattened and broadened at the nodes, and also in the darker, partly almost black bark.” an oval base. Flowers fasciculate in small waiaiete § in the axils of the leaves, shortly pedunculate. Ovary subglobose, glabrous. Calyx very large, divided to eyond the middle into 5 acute lobes, membranaceous. Corolla twice as long as the x, divided beyond its upper third into apiculate, narrowly lanceolate lobes, with iry ring above its base. Anthers comparatively small. Style pilose at the skened base; stigma cap-shaped, 5-toothed. = bash 4 to 5 m. high, whose flowering branches of 12 to 15 cm. are scarcely NE its cae prrsoinet taken, The stipules are hardly 5 mm. long. _ The ovary, black when dry, is 1.5 to 2 mm. long and up to 3 mm. thick. The white ealy is 13 to 15 mm. long and membranaceous. The ochre-yellow corolla is 3 cm. long, of which 2 em. is the tube; a ring of white hairs hangs down, 3 to 4 mm. above ‘the base inside. The stamens are included and scarcely 3 mm. long. The style is “The species can not be conkused on account of its ek, gageiform, white wage calyx and the style which is hairy at the base’’ (K. Schumann, 1899). = of Cameroon: Bipindi (Zenker). va Cuviera longiflora Hiern aera longiflora Hern, 1877, in Oliver, ‘Flora of Tropical Africa,’ III, p. 157. K. _ Scuumann, 1891, Ber. Deutsch. Bot. Ges., [X, p. 56; 1891, in Engler und Prantl, ‘Die Natirl. Pflanzenfam.,’ IV, pt. 4, p. 12, fig. 5. H. Kont, 1909, Natur #. ____ Offenbarung, LY, p. 166. P i “A glossy shrub of 25 feet or a small tree of 20 to 25 feet, glabrous or nearly so. _ Branches subterete; smooth. Leaves oval-oblong, cuspidate, oblique and hollowed at the base, spreading, thinly coriaceous, paler beneath, 6 to 12 by 2 to 4 in.; lateral - veins about 10 to 12 pairs, inconspicuous; petiole \ to % in.; stipules deltoid, keeled at the apex, hairy inside, 4 in. long. Flowers 1%; in. in diameter when expanded, on short pedicels, several together, a js ogeracarnpt aah ter diameter; bracteoles nar- ____ rowly elliptical, 3; to 1) in. long; peduncle about 1 in. Calyx-segments lanceolate, s % to %; in. long. Corolle-segments 3, in. long, lanceolate, acute. Anthers drooping. Ovary 5-celled; style pilose-hirsute below; stigma cernuous’’ (Hiern, 1877). . Cameroon: Mt. Cameroon, at 2000 to 3000 ft., type locality (Mann). Also found in Cameroon by Preuss. " The presence of myrmecodomatia is not mentioned in the original diagnosis of this species, but Schumann found conical swellings on the ___ basal part of the internodes of specimens collected by Preuss in Cameroon. Two longitudinal rows of three or four superposed orifices, often sur- ___ rounded by a thickened ring, led into a cavity containing small, black ants of the genus Crematogaster. - — 484 Bulletin American Museum of Natural History [Vol. XLV C. longiflora is so closely related to C. angolensis that the differences are not clear from the descriptions. Some of the plants which I observed in the Belgian Congo agreed equally well with the descriptions of each and it seems possible that future investigation will result in the synony- mizing of C. angolensis with the earlier described C. longiflora. H. Kohl (1909, p. 166) states that C. longiflora differs specifically from C. physi- nodes and C. angolensis “in the sparse, short pilosity of the style, which is very strikingly narrowed toward the apex.” The style of C. physinodes is described as glabrous, which is also the case with C. acutiflora and C. subuliflora. C. angolensis, however, agrees with C. longiflorainhavingthe | style pilose, as was mentioned in the original description and as I have observed in my Congo specimens. Cuviera macroura K. Schumann Cuviera macroura K. ScHuMANN, 1903, Engler’s Bot. Jahrb., XX XTIIT, p. 352. “Branches slender, not fistulose nor swollen, cylindrical, even the young ones flattened and glabrous. Leaves short petiolate, lanceolate or suboblong-lanceolate, subacuminate, acute at the base, margined, glabrous on both sides, somehow folded by drying. Stipules tubulose-connate, bidentate, accrescent, finally pierced by the inflorescences and withering away, villose inside. Flower panicle tripartite from the base, with many or dense flowers, glabrous, Bracts linear, acuminate. Ovary 5- celled. Sepals linear, acuminate, glabrous, united at the base into a cupule. Lobes of the corolla with very long appendages. Style hirsute. “The flowering branches are 30 cm. long and only 3 mm. thick at the base; they are covered with brownish-black bark. The petiole is 3 to 5 mm. long and flattened above; the blade has a length of 9 to 15 cm. and in the middle is 2.5 to 4 cm. wide; it is crossed on each side of the midrib by 6 to 7 stronger veins which are prominent on both surfaces, but almost more so on the upper side; in dried condition it is black green above, leather-yellow below. The stipules are 7 mm. long. The 3 bracteoles are about 1.5 cm. long. The calyx has a total length of 1.6 cm., of which 1.3 em. is to be allowed for the lobes. The corolla is 2 em. long, half of this belonging to the appendages of the lobes. The stamens and anthers measure 1.5 mm. “The species strikingly differs from all the West African ones in its small leaves and the long appendages of the corollar lobes’ (K. Schumann, 1903). Southern Nigeria: Lagos (Millen). Cuviera minor Wright Cuviera minor C. H. Wricut, 1906, Bull. Mise. Inform. Bot. Gard. Kew, p. 105. “Differs from the other species in its smaller, membranous leaves. “A small tree. Branches ash-colored. Leaves ovate or oblong-ovate, acuminate, slightly uneven-sided, rounded or short cuneate at the base, glabrous; with about 6 lateral veins on each side of the midrib; 11.5 em. long, 4.5cem. broad. Petiole grooved above, slender. Stipules broadly triangular, dropping. Inflorescences axillary, with many flowers, 4 cm. long. Bracts oblong, narrowed at the base and at the apex, 8 mm. long. Lobes of the calyx 5, subfoliaceous, lanceolate, 8 mm. long. Tube of the Wheeler, Ants of the Belgian Congo 485 el i can iis ot halen hank downed. Lobes of ‘corolla alee. acuminate-caudate, 1 cm. long, pilose externally, yellowish. 5, inserted between the lobes of the corolla; anthers sagittate, twice as long : filaments. Ovary 5-celled, each with one ovule. Style 8 mm. long; stigma naped”’ (C. H. Wright, 1906). fee Coast: Kimaha (Johnson). Cuviera nigrescens (Scott-Elliot) migrescens (Scott-Elliot) Wernuam, 1911, Journ. of Botany, London, XLIX, e + nigrescens Scotr-Ex.ior, 1894, Journ. Linn. Soc. London, Botany, XXX, : 206, p. 81. Oxtver, 1894, in Hooker’s ‘Ieones Plantarum,’ XXIII, pt. 4, MMCCLXXXIII. Sees iss» K. guiness 1897, Engler’s Bot. Jahrb., XXIII, 4, p. 461. ) a veins where they are hirsute), oblong-ovate or obovate, obtuscly acuminate, subeuspidate, with coriaceous margin, narrowed at the base; 5 to 8 cm. long and 2 to a 3 em. broad; 5 or 6 pairs of lateral veins; petiole 6 to 8 mm. long. Stipules hirsute oral gts ane pga gg aap 3 to 5 mm. long. ts ovate, obtuse, with reticulate venation, 8 to 9 mm. long and 4 mm. iwond . lyx with 5 large lobes, which are lanceolate, subacute, 8 to 9 mm. long and 2 mm. broad. Lobes of the corolla caudate-acuminate, 15 to 17 mm. long (the acumen _ 8 to 4 mm.), sparsely hirsute on the outside with white hairs 1 mm. long, internally h a ring of reflexed pile. Filaments 2 mm., anthers 1 to 2 mm. long. Stigma large, 1 to 2 mm. long and 1 mm. bend: Ovary 5-celled’’ (Scott-Elliot, Sierra Leone: in the forest between 1000 and 3600 feet; near Kafogo in Limba ’ sn sear Fala (C. F. Scott-Elliot). Liberia: Golah Forest (Bunting). ah % The Liberia specimens differ from those of Siérra Leone only in the - length of the caudex of the corolla-lobes, which in the former appear to _be longer and more setaceous in character (Wernham). Both Scott-Elliot and Oliver compare this species with the two other Vangueriz with caudate corolla-lobes: V. velutina Hiern, which has densely tomentose leaves and inflorescences; and V. pauciflora Schwein- furth, with solitary or geminate flowers and truncate calyx. This species was evidently redescribed by K. Schumann, in 1897, as - Cuviera trichostephana, on part of the material collected in Sierra Leone _by Scott-Elliot. For the sake of completeness, Schumann’s description of C. trichostephana is translated here: A woody plant with slender, terete or subtetragonal branches, very glabrous even in youth. Leaves on the specimen examined not completely developed, petiolate, oblong, shortly and obtusely acuminate, acute at the base and often suboblique, a a “1 a oa a os “a 486 Bulletin American Museum of Natural History [Vol. XLV glabrous on both sides, but the axils of the veins with minute hairy domatia; stipules lineate-subulate, with triangular base. Axillary cyme with few flowers, glabrous; ovary 5-celled, glabrous; calyx with foliaceous or membranaceous, oblong, sharp lobes. Corolla divided to beyond its middle, with a corona of decumbent hairs me pilose at the outer side, with very long, caudate, linear-lanceolate lobes. The branch at hand is 15 em. long and at most 2 mm. thick at the eles The petiole reaches a length of 1 em. and is slightly excavated above. The blade is 4 to 9 em. long and 2 to 4 em. broad in the middle; traversed by 5 stronger veins on each side of the midrib; black when dried; herbaceous in the specimen studied, but the leaves are apparently not yet fully developed. The stipules reach a length of 7 to 8 mm. The entire inflorescence is about 3 cm. long. The pedicels of the flowers reach alength of 5mm. The ovary is semiglobose and 1.5 mm. long. The lobes of the calyx reach 10 mm. in length and 3 mm. in width. The corolla is 2.2 to 2.5 em. long, of which the tube takes 9 to 10 mm. only. The anthers are 2 mm. long, placed on fila- ments 3 to 4 mm. long, exserted from the tube and curved. The style is 1.7 em. long. This species is easily separated from all others by the corolla covered with hairs, the smaller leaves, and the short inflorescences. It has more the appearance of the genus Vangueria, so that it makes the generic limits less distinct. Sierra Leone (C. F. Scott-Elliot). Cuviera physinodes K. Schumann Cuviera physinodes K. ScuuMANN, 1891, Pringsheim’s Jahrb. Wiss. Bot., XIX, pp- 55-56; 1891, in Engler and Prantl, ‘Die Natdrl. Pflanzenfam.,’ IV, pt. 4, p- 12, fig. 5A. Cuviera idvectialle H. Kou, 1909, Natur u. Offenbarung, LV, pp. 162-163. “Leaves large, 20 to 30 cm. long, 7.5 to 11 em. broad, with thick petiole, ovate- oblong or oblong, shortly and obtusely acuminate, equilateral at the base, coriaceous, glabrous on both sides. Ovary 5-celled; stigma glabrous. Drupe oblong, 3 em. long, about 1 em. in diameter, acute at the apex, acuminate at the base, without ribs. “Jt is a tree-like shrub about 3 m. high, with large, leathery leaves. The cymes are axillary, with many flowers, short, ramified; only a few of the greenish white flowers produce fruit, though all seem to possess well-developed ovaries. The cylin- drical internodes, covered with gray bark, are regularly thickened in their upper part, - but do not develop swellings there. The swellings are situated rather above, and close to the nodes” (K. Schumann, 1888). Gaboon: Sibange farm, type locality (Soyaux). The myrmecodomatia of this species have been briefly described by K. Schumann from dried specimens. One of the hollow, nodal swellings had a length of 3.5 em. and greatest diameter of about 1 em., the wall being about 1.5 mm. thick. The inner cavity was nearly spindle-shaped and ended slightly below the node; three openings led into the cavity; one of these, 2 mm. long and 1.5 mm. broad, was probably alone used as entrance, while the two others were reduced to mere slits, 1 mm. long and hardly 0.5 mm. wide. Traces of former holes, evidently closed by callus growth, could be seen on two other spots. A few remains of ants were found inside the swellings. Wheeler, Ants of the Belgian Congo 487 aid Cuviera plagiophylla K. Schumann ra plagiophylla K. Scuumaxn, 1903, Engler’s Bot. Jahrb., XX XIII, p. 353. *A shrub with thick, fistulose-inflate branches, which, even when young, are 3s. Leaves strictly sessile, linear-oblong, short acuminate, rounded at the , strongly inequilateral, glabrous on both surfaces. Stipules tubular, villose ernally. Flower panicle axillary, with many flowers. Bracts very long, linear, m rate. Ovary 5-celled. Sepals free almost to their base, linear, acute. Corolla very short tube; the lobes acuminate, moderately appendiculate, cristate dor- - Style glabrous. ‘The bush reaches a height of 5 m. The leaves are 28 to 30 cm. long and 8 to 9 . broad; they are crossed on each side of the midrib by about 16 stronger veins, lich are more prominent on the under side, as are also the reticulate veins; they are ¢ when diy. The stipules are 9 mm. long. The lobes of the calyx are 11 mm. long and somewhat obtuse. The corolla is greenish-white, 15 mm. long, of which 2 mm. ; to be allowed for the tube; the appendages measure 5 mm.; the keels on the dorsal face of the lobes make the bud sharply 5-ribbed. “y “The species is very distinct by the strictly sessile, very oblique leaves and the acutely keeled corolla-lobes”’ (K. Schumann, 1903). “4 iy _ Camneroon: Bipindi, near Lokundje (Zenker). Cuviera subulifiora Bentham Cwira subuliflora Benruam, 1849, in Bentham and Hooker, ‘Niger Flora,’ p. 407. _____Hrery, 1877, in Oliver, ‘Flora of Tropical Africa,’ III, p. 157. ___ “An arborescent shrub or small tree of 15 feet, glabrous. Branches subterete, smooth, opposite. Leaves oblong, shortly acuminate, oblique and hollowed or p rounded or somewhat narrowed at the base, chartaceous, rather paler beneath, 6 to 15 - by 2 to 4% in.; lateral veins about 12 to 14 pairs, slender; petiole 4 in. long; stipules deltoid, aammate at the base, keeled near the apiculate apex, 4 to \ in. long, hairy ___ within. Flowers numerous, on short pedicels, in divaricately branched axillary and lateral, subsessile, dichotomous panicles of 2 to 3 in. diameter; bracteoles linear, nar- rowed at both ends, 4 to 144 in. long, accrescent as well as the ealyx-segments. Calyx _ whitish; segments narrowly or at length broadly linear, ranging up to 1 in. long. Corolla green; segments about }4 in. long, lanceolate, caudate-acuminate. Style glabrous. Ovary 5-celled. Fruit 1 in. long, obliquely egg-shaped” (Hiern, 1877). a Fernando Po; pn the sea shore (Vogel). . _ Southern Nigeria: Abo (Vogel). Cuviera trilocularis ican Se aoratants Hreen, 1877, in Oliver, ‘Flora of Tropical Africa,’ IIT, p. 157. “A small glabrous tree. Branches subterete, smooth. Leaves ovate-oval, acumi- ____ nate, rounded and suboblique at the base, thinly coriaceous, glossy, of nearly the same ___ eolor on both sides, 4 to 5 by 1% to 2 in.; lateral veins about 8 to 10 pairs; petiole \ to 4 in.; stipules apiculate, ovate, keeled, 14 in. Flowers on short pedicels, several together, in the terminal or subterminal axils; panicles 114 to 2 in. diameter; com- } mon peduncle about ‘4 in., bracteoles lanceolate, 4 to %4 in. long. Calyx-segments _ greenish white, }4 in. long or rather more, linear-elliptical, acute, narrowed toward the base. Corolla shorter than the calyx; lobes lanceolate, caudate-acute. Style glab- rous; stigma 10-suleate. Ovary 3-celled"’ (Hiern, 1877). Southern Nigeria: Old Calabar (W. C. Thomson). ——— 488 Bulletin American Museum of Natural History [Vol. XLV Eco.tocy or Cuviera in THE BerGian Conco The representatives of this genus observed by me mostly occurred in low-lying or moist places, though not in those apt to be frequently flooded; raised river banks are favorite sites. Usually growing as shrubs or bushes, 2 to 4 meters high, under favorable conditions they may be- come small, pyramidal, bushy trees of 5 to 7 meters. The trunk is slender, erect, and destitute of branches below where it often, but not always, bears opposite, decussate, very sharp spines, 2 to 5 em. long. The long, slender branches spread more or less horizontally and their tips hang down somewhat. In accordance with the decussation of the leaves, they are placed opposite each other in four regular rows. Asarule the upper part of the plant is unarmed, though in some specimens one finds in the axils of the leaves heavy, straight spines, evidently modified, aborted branches.' The leaves are very large, 10 to 25 em. long, 5 to 11cm. broad, borne on a short petiole (of about 1 cm.), entire and simple, thinly coriaceous, smooth and glabrous on both sides, dull green above, paler below; usually cuspidate or more or less acuminate at the apex, oblique and rounded or slightly heart-shaped at the base. The lateral nervures are rather thin, more conspicuous on the under side of the leaf, and number 8 to 10 on each side of the midrib. The stipules are connate into a short, loose sheath, which is keeled, acuminate, and about 0.5 to 1.5 em. long. The base of this stipular sheath persists on older branches. The plant is not often seen blossoming. Welwitsch, in Angola, found flowers in April and May and fruits in August; while in the Belgian Congo, flowering specimens were seen by Dewévre in March (Lukolela) and by me in February (Penge), July (Kunga), and December (between Masisiand Walikale) ; fruits were found in January, 1915, between Wali- kale and Lubutu on a plant not in flower. From these very incomplete data, which may relate to different species, it would appear that Cuviera blossoms from December to July, yet it is quite possible that there is no definite flowering season, as is so often the case with bushes and trees of tropical rain forests. Dewévre in his field-notes accurately describes the flowers of C. angolensis. They are large, conspicuously colored, and placed as many as a dozen together in axillary, polygamous panicles, toward the upper end of the younger branches. The common peduncle is 0.5 to 4 em. long, while the pedicels are very short, the flowers being subsessile in the axils Kohl (1909, p. 164) and De Wildeman (1906, p. 297) also note that the branches » a ee — ana — i EE a od te Wheeler, Ants of the Belgian Congo Fig. 100. Cuviera species? a, portion of branch giving external view of one of the domatia and the longitudinal section of another; 6, longitudinal section of one of the domatia; e, ants; d, pits often occupied by coccids. Drawn from herbarium specimens obtained at Kunga, near Malela; natural size. 494 Bulletin American Museum of Natural History (Vol. XLV pith cells containing it are aggregated in the center of the more deeply lobed medulla (Bailey). The plant from this locality is referred to as ‘unidentified Cuviera”’ in Prof. Bailey’s contribution ra V, p. 593). 3. Synopsis or RecorpED MyYRMECOPHYTES The study of ant-inhabited plants is in such an incomplete state that no adequate or standard definition of the term ‘‘myrmecophyte’’ has so far been formulated. The student must therefore be prepared, in reading the present synopsis, to meet with cases of very unequal value. Warburg (1892, p. 130) has proposed to classify plants according to the nature of their relations with ants into the following three groups: a. Myrmecorropuic plants provide only food to the ants, either in the form of sugary exudates (nectaries,) special food-bodies (bromatia of the fungi), seeds or fruits of the myrmecochores, and the like. b. Myraecopomic plants furnish only shelter to the ants’ nests, either in normal cavities, such as hollow stalks, or in special swellings or myrmecodomatia. c. Myrmecoxenic plants act as true hosts, offering to their ant guests both shelter and food. Typical cases of the kind are Cecropia adenopus (with the Miillerian bodies) and Acacia cornigera (with the Beltian bodies). The term “myrmecophyte” is here used to include Warburg’s ‘“‘myrmecodomic” and “‘myrmecoxenic” plants. A further distinction of these two categories seems very unwise at present, because we are, it appears, just beginning to understand the true relations existing between ants and the plants they inhabit. My definition of ‘“‘myrmecophytes”’ is based on practical considerations and is thus merely provisional. In the main, however, I agree with Ule (1906), p. 335), who proposes to designate as ant-plants all plants which are steadily inhabited by certain species of ants, excluding only cases where the ants occasionally settle in normal leaf-sheaths, slits in the bark, dead branches, ete. Schumann’s (1888) definition, on the other hand, is quite teleological and therefore of little use under present circumstances, since he wishes to restrict the term “‘myrmecophyte” to those plants “that are not merely visited by ants, but are purposely inhabited by them, and that therefore have probably entered with them into a true symbiotic relation.” The exquisite manner in which many ants have come “to know plants” (Michael Gehlerus, 1619) must indeed astonish the botanist who is but little acquainted with the psychic activities of these tiny insects. In his search fora much-needed explanation he naturally turns to the magic action of “‘ Natural Selection,” following in this the general trend of present ecological botany. Various theories of myrmecophytism are fully exposed and critically discussed by Prof. Bailey in part V of this 1922) Wheeler, Ants of the Belgian Congo 495 Report (pp. 610-614), « so that a further consideration of this interesting = topic is unnecessary here. ____ The origin of the various plant structures used by ants for nesting __ orfeeding purposes isa purely botanical problem.' To the myrmecologist, on the other hand, belongs the task of tracing the various modifications of ant behavior which have gradually led to the close, obligatory cceno- biotic associations of certain Formicide with certain plants. It is not difficult to show that here, as elsewhere, the specialization in the habits _ of ants has followed its own course, quite independently of any simultane- ous changes in the structure of plants. Numerous ants belonging to many genera of the higher groups, viz., the Pseudomyrmine, Myrmicine, Dolichoderine, and Formicine, establish temporary shelters or permanent colonies in dead branches, stumps of trees, dry stalks of herbs, and like places. In what perhaps may be regarded as the most primitive stage of this behavior, the ants _ merely appropriate existing cavities, such as old burrows of wood-boring _ larve, empty galls, and hollow pith channels.- Dry stalks of grasses, reeds, and other herbaceous plants are also great favorites as nesting sites with many tropical ants (Forel, 1896a-d). At Luali, Belgian Congo, I found in August, 1913, a beautiful and populous nest of a Camponotus established in a dry stalk of papyrus on the bank of the Shiloango. Tucker (1911, pp. 24 and 26) mentions finding nests of the North Ameri- ean Crematogaster lineolata Say subspecies leviuscula Mayr variety clara Mayr, at Alexandria, Louisiana, in corn-stalk cavities formed by a borer and also in hanging “bolls” or fruits of cotton. The common Lasius niger (Linnzeus) was observed in Europe in fallen apples, temporarily occupying the empty galleries made by the apple-moth (Ruzsky, 1913, pp. 61-63). In many cases ants excavate new galleries in dead or decaying vegetable tissues or transform the fibres into “carton” used as parti- tions or plugs. One of the most typical of these borers in dead wood is * salen sear se on Mpa poms mag pwnd SS ee wow a on 4 AY Th are mere su uently fen ey sare cei ean development inside the 7 c. vib cg am arious other question but that ‘di ant and that insects take no part in their production. Ri stotedy ses eecic inves ee the Central American bull-horn acacias, which reach their c even under cultivation. It must further be pointed out that the mere presence of inacet larva feeding nside plant swellings does not necessarily —_ that these are = : ~ insects, ne eovurrence thaws may Se be purely acciden . Many years ago 1 noticed that caterpi requently destroy of the swollen thorns of Acacia cavenia. oreover, that insect galls have been and ee ‘for true myrmecodomatia has been shown B wees wiry (pp. 371-375), so that each particular case must examined with the utmost care, in order to ascertain ite true stan ing. 496 Bulletin American Museum of Natural History (Vol. XLV the common Camponotus carye (Fitch), several forms of which occur throughout the Palearctic and Nearctic Regions (Wheeler, 1910c, pp. 219-220). Many species of Crematogaster and Leptothorax remove the pith from dead twigs of trees, briar and rose bushes, etc., to make homes for themselves (Forel, 1903b; Staiger, 1917 and 1919). A peculiar ccenobiotic association was described by Wheeler (1912a) in the case of a mistletoe, Phoradendron flavescens variety villosum Nuttall, which grows on live oaks (Quercus emoryi Porter and Coulter) in the Huachuca Mts., Arizona. The branches of this mistletoe are very frequently hollowed out for some distance by a curculionid larva; the beetle makes its exit through a round hole at the side of the twig and the deserted gallery is . then usually occupied by a colony of Crematogaster arizonensis Wheeler. Furthermore, the walls of these formicaries are invariably covered with reddish coccids, Pseudococcus phoradendri Cockerell. In the tropics of both hemispheres, many species of Cataulacus (Paleotropical) and Cryp- tocerus (Neotropical) are true wood-boring ants. Similarity in habits has gradually resulted in a remarkable resemblance in the shape of the head and the flattened body of these two genera, though they are not closely related to each other. The keenest carpenter ants, such as the holarctic Camponotus hercu- leanus (Linnzeus), with its various races and varieties, and the European C. vagus (Scopoli), frequently extend their burrows into the live, healthy wood of standing timber. It is, however, among the tropical and sub- tropical Pseudomyrminz that we find all transitional stages between the common wood-boring habit and the more specialized behavior of nesting inside living, normal organs of plants and myrmecodomatia. The impulse to gnaw through living vegetable tissues not only presupposes a greater inquisitiveness on the part of the ants, but it is undoubtedly also influenced by the anatomical structure and chemical composition of the plant, as is clearly shown by Prof. Bailey’s histological study of myrmecophytes (See Part V, p. 585-621).2_ From the habit of boring into normally existing cavities of plants it is only a step to the excavating 1In January, 1910, I found several nests, with queens and workers, of Leptothoraza us (Nylander’ variety bruneus Santschi inside dead, hollowed stalks of wild roses near Algiers. comnom Nauk American L. curvisninovus Mayr usually nestsin hollow twigsorstalkes. The larve of many Hymenoptera, such as the Tenthredinide and Siricidw, most Cynipidw, and certain Chalcididw, are phytophagous, feeding on the living tissues of healthy, growing plants. It may not be so commonly known that as adults, too, some of them attark living parts of p Certain of the larger saw-flies are known to injure twigs of bushes by ling them with their mandibles. The large hornet, Vespa crabro, gnaws the new bark of trees in order to get building material for its nests. A number of tropical and subtropial bees and fossorial wasps are known to excavate galleries in the n, juicy pith of living plants. According to Brauns, this is one of the the behavior of certain South Afri an Xylocopa, Ceratina, and Di octus. I have observed similar habits in species of Allodape and Dasyprortus in the Belgien Congo. Bertoni, in P: » found Xylo- crabro umbrosus S-hrottky nesting in the green stalks of radish which ripens its seed t the time the young wasps are hatching. > 1922) —— 497 of the pith of living ideaials which brings us then to the settling of the so-called myrmecodomatia. That pith-nests, such as those described for Endospermum formicarum Beccari by Dahl (1901) in New Guinea, _ have been so seldom noticed is probably merely due to a lack of proper investigation. In the following pages an attempt is made to review the various eases of true myrmecophytism which have thus far been recorded, as the pertinent observations are quite scattered in entomological and botanical publications. In this list the plants are arranged according to their _ systematic sequence. I have added a few remarks on distribution, a short description of the myrmecodomatia, and a record of the ants found therein. The available information is, however, often very scanty. So _ far as possible, doubtful or erroneous observations have been excluded or expressly questioned, while the recording authors are given in each case. The dates refer to the appended bibliography. PTERIDOPHYTA Polypodiacez A cosmopolitan family, containing some 100 genera and 2800 species. Polypedium Linnzus. Cosmopolitan, with about 200 species; some of the Oriental species are well-known myrmecophytes. P. sinuosum Wallich. Malay Region from Malacca to the Solomon Islands. Inhabited by Technomyrmezx albipes (Smith), an ubiquitous ant (Yapp, 1902; Ridley, 1910; Shelford, 1916); also by Iridomyrmex myrmecodiz Emery in Borneo (Wheeler, 1919, p. 100) and in Java (Miehe, 1911b), and by J. cordata (Smith) in New Guinea (Beccari, 1884). P. lomarioides Kuntze. Malay Region (Yapp, 1902). ' P. sarcopus De Vriese and Teysmann. Celebes (Yapp, 1902). P. imbricatum Karsten. Amboina. P. leiorhizon Wallich. Eastern Himalaya, Western China. These five species constitute the subgenus Aspidopodium Diels (=Myrmecophila Christ). They are epiphytic ferns, with creeping, semicylindric rhizomes, which are fleshy and much swollen on the upper side where the leaves are inserted on mammate protuberances; the flattened under side is pressed against the support. Originally the swelling is filled with an abundant aquiferous tissue, which in drying up causes the rhizomes to be tunnelled almost their whole length. The resulting cavities are, as a rule, inhabited by ants which pierce the en- trances (Gaebel, 1888; G. Karsten, 1895). 498 Bulletin American Museum of Natural History [Vol. XLV A species of Polypodium (?P. megalophyllum Desvaux =P. Schom- burgkianum Kuntze) of South America (Rio Negro; Rio Napo) is said to have rhizomes similarly swollen and occupied by ants. A Costa Rican species, Polypodium Brunei Werckle, possesses small bulbs, about 2 to 2.5 cm. in diameter, fixed by short peduncles at the sides of the rhizome; these bulbs are hollow, provided with an orifice, and divided by parti- tions into four or five spacious chambers. G. Senn (1910) regards them as water reservoirs; whether they are occasionally inhabited by ants is not known. Polypodium bifrons Hooker, of Brazil, has similar swellings which, according to Ule (19065), act also as water reservoirs and are not occupied by ants. | Lecanopteris Blume. Malay Region. Represented by four or five closely allied species, all epiphytes, with swollen, tuberiform rhi- zomes, traversed by a system of galleries inhabited by ants. The genus is doubtfully distinct from Polypodium. L. deparioides (Cesati). Borneo (Shelford, 1916). L. carnosa Blume (=Polypodium patelliferum Burck). Perak, Borneo, the Moluccas, Philippines, Celebes, Java (Yapp, 1902; Ridley, 1910; Shelford, 1916). Inhabited by Crematogaster yappii (Forel) and _ C. difformis F. Smith. Hooker believed that L. carnosa represented a teratological condition of Polypodium lomarioides, but this view has been discarded following Burck’s (1884a) observations of this plant. L. Curtisii Baker. Sumatra. L. Macleayiit Baker. Java. Some of the Old World epiphytic ferns of the genus Drynaria Bory have been improperly included among the myrmecophytes. They are remarkable in having, in addition to the normal, fern-like leaves, others which are sessile, broad, superficially divided, and pressed against the support and the rhizome. Humus accumulates underneath the cover of these appressed leaves and is soon invaded by roots. Frequently ants nest in this humus, but their presence there is merely accidental and I agree with Goebel (1888) that these cover-leaves (‘‘Nischenblitter”) can by no means be considered as myrmecodomatia. Drynaria Laurentit Christ is one of the commonest epiphytic ferns of the Congo Basin and shows all the peculiarities of the genus beautifully. D. quercifolia (Lin- nus) is abundant in the Oriental Region, from India to Polynesia. Pheidole javana Mayr subspecies jacobsoni Forel variety tatpingensis Forel was found by v. Buttel-Reepen forming small colonies in the cavi- ties of the irregularly thickened root of an epiphytic fern in Malacca 1922} Wheeler, Ants of the Belgian Congo 499 (Forel, 1913, p. 28). More details concerning this plant will probably be given in vy. Buttel-Reepen’s forthcoming paper on the biology of East Indian social insects. MONOCOTYLEDONEZ Palmez An abundant family in tropical and subtropical regions, especially in South America and the Malay Region. Approximately 170 genera, with 1200 species, have been described. A small number of species have been found associated with ants but, with the exception of certain Korthalsiz, they can hardly be called myrmecophytes. _ Korthalsia Blume. Oriental Region. With twenty species, all of which are rattan-palms. While the ligule of the leaf-base usually forms a close, tightly fitting sheath, in a few species which constitute a special section, this organ is dilated into a rounded or oblong, bulky sheath or ocrea of a stiff papery texture, frequently perforated and occupied by ants of the genus Camponotus. Emery has expressed the opinion that these Camponoti belong to a special group of the genus, adapted to living in the ocrea of these palms. K. scaphigera Martius. Malay Region (Beccari, 1884; Ridley, 1910; Shelford, 1916). In Sumatra Beccari found the ocrea of the leaves perforated on the sides and inhabited by Camponotus hospes Emery; in Borneo a related Camponotus was found in this palm. Accord- ing to Ridley (1907, p. 216) the natives of the Malayan Peninsula call this palm ‘‘ Rotan semut” or ant-rattan. K. echinometra Beccari. Malay Region (Beccari, 1884; Ridley, 1910; Shelford, 1916). The ant found in Sarawak in this palm by Bec- cari was Camponotus contractus Mayr, which had cut an entrance. On passing near the plant one may hear the ants running along the walls of the ocrea, which acts as a resonator. Emery (1888, p. 529, footnote) described C. contractus variety scortechinii from specimens taken in the ocrea of K. echinometra in Perak. Crematogaster difformis F. Smith had settled in the ocree of specimens of this palm cultivated at Buitenzorg; it had not pierced an orifice, as did the Camponotus mentioned above, but merely made its way along the slight depression near the upper margin of the ocrea. Camponotus contractus variety buttesi Forel, from Kwala Lumpur, Selangor (Malacca), was found in the hollow swellings of a plant called by the natives “Rotan udang,” in which the workers make a peculiar noise at night (Forel, 1902, p. 463). According to Ridley (1907, p. 216), this is the Malayan name of Korthalsia echinometra. 500 Bulletin American Museum of Natural History (Vol. XLV K. angustifolia Blume. Malay Region. In Sumatra, Beccari found the ocrew pierced with a hole and inhabited by Camponotus korthalsiz Emery (Beccari, 1884). K. horrida Beccari, K. Scorlechinii Beccari, and K. cheb Beccari, all from the Malay Region, have a similarly constructed ocrea, with an orifice undoubtedly pierced by ants which have not been identified (Beccari, 1884). Calamus Linnzeus. About 150 species in the Oriental Region from India to tropical Australia and Polynesia; one species in tropical Africa. C. amplectens Beccari. Borneo. The two lower segments of the leaves are folded back and embrace the stem so as to enclose it, the re- sulting cavity being inhabited by ants (Beccari, 1884; Shelford, 1916). Demonorops Blume. Oriental Region. Represented by seventy species, all rattans. In several of them ants habitually make nests in the: large, stiff flower-spathes, which often quite cover the flower-panicles. The genus is closely allied to Calamus. D. Jenkinsianus Martius. Malay Region. Flower-spathes in- habited by a Camponotus allied to C. mitis (F. Smith) (Ridley, 1910). Orchidacez One of the largest families of plants, containing 500 genera and over 15,000 species. Cosmopolitan, but chiefly in warm and humid regions. The following cases of myrmecophytism are still doubtful and need closer investigation. Diacrium Lindley. Epiphytes of the Neotropical Region; four species. D. bicornutum (Hooker), of Trinidad and Guiana, has a swollen, spindle-shaped stem, which is normally hollow and perhaps regularly in- habited by ants (Rodway, 1911, p. 111). Schlechter' claims that even under cultivation the pseudobulbs form at their base a slit through which the ants gain access into the cavity. Schomburgkia Lindley. Epiphytes of the Neotropical Region. Represented by thirteen or fourteen species, from Mexico to Guiana and Peru, several of which have hollow pseudobulbs. S. tibicinis (Bateman), in Central America (from Mexico to Vene- zuela), has voluminous, elongated pseudobulbs, which are hollow, with a smooth inner lining and usually inhabited by ants; these go and come through a small opening pierced at the base of the pseudobulb (Ross, Die Orchideen,’ (Berlin), 1915, p. 214. Wheeler, Ants of the Belgian Congo 501 “ has recorded Neoponera villosa (Fabricius) from the pseudobulbs of this r oe at Vera Cruz, Mexico. as _ Grammatophyllum Blume. Epiphytes of the Malay Region; or whe speciosum Blume is one of the largest orchids known; the stem reaches a height of 4m. and is thickened, especially towards the base; . a Ee ll > it shows galleries occupied by ants. voagheda DICOTYLEDONEZ Moracez oe ~ Cosmopolitan family, though chiefly tropical, with 70 genera and ea about 1000 species. The only myrmecophytic members known with _ certainty belong to Cecropia. Schimper has described and figured Ficus _ inequalis with swellings, supposed to bemyrmecodomatia, on the branches, a _ but Ridley (1910, p. 458) has shown that these swellings are accidental, - pathological productions. _-—- Pourouma guianensis Aublet, of South America, which is related to Cecropia, according to Rettig (1904), possesses trichilia at the base of the ____ petiole which produce food-bodies similar to the “ Millerian bodies” of __ Ceeropia adenopus; whether they are collected by ants is not known. Forel (1904b) mentions Azteca duroix Forel as having been found by Ule _ in the twigs of an unidentified Pourouma in Brazil. Cecropia Linnzus. This genus occurs throughout tropical America from Mexico to Brazil. There are thirty to forty species, apparently very few of which (subgenus Aztecopia H. v. Ihering) are myrmeco- _ phytes. These latter shelter nests of various species of Azteca inside their hollow stems and also produce food for the ants in the torm of so- called “Miillerian bodies.” Many other species of this dolichoderine genus of ants nest in various locations or even build free carton nests in trees. It seems, however, that the species which inhabit the Cecropiz are obligatory plant ants, being met with only inside these plants; the colonies perish when the trees die or are cut down. Alfaro found inside the stems of an unidentified Cecropia in Costa Rica the following ants: Azteca ceruleipennis Emery, A. alfaroi Emery, A. zanthochroa (Roger), and A. constructor Emery (Emery, 18968). ____Ule collected in Brazil A. a!faroi subspecies cecropix Forel from another __ species of Cecropia (Stitz, 1913a). Warming (1894) studied in Venezuela a species of Cecropia which he found inhabited by Azteca instabilis (F. Smith). 502 Bulletin American Museum of Natural History (Vol. XLV Cecropia adenopus Miquel (=C. peltata Vellozo, nec Linneeus). A common species on the east coast of Brazil between 28° S. lat. and the Equator (H. v. Ihering, 1907). The best account of this celebrated plant is given by Wheeler (19106, pp. 305-310): The tree known as “imbauba” or “‘imbauva”’ is very slender and candelabra- shaped, growing to a height of 12-15 m. The trunk and branches afe hollow except at the nodes, where there are thin transverse septa. The sap is colorless, not milky nor rubber-containing, as stated by some authors. The crown of foliage is meagre and consists of large, palmately lobed leaves. At some time of its life each node bears a leaf, the long petiole of which has at its base a hairy cushion, known as the trichilium, in which the yellow Miillerian bodies are imbedded. The cavities of older and larger trees are almost without exception tenanted by Azteca muelleri Emery, which perforates the septa and thus causes all the internodal cavities to communicate with one another, both in the trunk and branches. The ants do not, however, live in the smallest, still . actively growing twigs. The just-fecundated queen enters the branches while the tree is still young (50 cm. to 2 m. high) at a particular point, a small depression at the upper end of a furrow at the top of the internode, where, as Schimper has shown, the wall lacks the fibrovascular bundles and is most easily perforated. Von Ihering calls the depression the ‘‘prostoma,”’ the perforation which is formed in it the “stoma.” The queen thus enters the internode by making a stoma and feeds on the tissue (“stomatome”’) which, according to von Thering, soon proliferates over and closes the opening from the inside. In the small internodal cavity the first workers, six to eight in number, are reared, and these restore communication with the outside world by again opening the stoma. Several females may each start a colony in one of the internodes of the same tree. Since later only one colony is found in a tree, v. Ihering supposes that the various primary colonies fuse to form one large com- munity, after all except one of the queens have been killed. Such a fusion of workers from different colonies is, however, doubted by Wheeler. After the single community has grown and has perforated the septa, it starts a spindle-shaped carton nest in the bole, a little distance above ny ground. This so-called “metropolitan nest,’’ which was Susiaeeds by von Ihering, re- sembles the carton nests built by other species of the genus on the branches of Cecropia and other trees. Where the nest occurs the bole of the Cecropia presents a spindle- shaped enlargement, which von Ihering regards as a gall—“the largest known gall,” but his figures and several of these nests recently acquired by the American Museum of Natural History prove conclusively that such an interpretation is erroneous. The wall of the hollow trunk where it encloses the nest, shows no structural modification except a bending outward of the woody fibers. About half the thickness of this wall is gnawed away by the ants from the inside, leaving a thin zone encircling the trunk, which naturally bulges out under the weight of the superposed trunk and crown of foliage. As there is no hypertrophy of the tissues in the spindle-shaped deformation, the term gall, as applied to a structure of such simple mechanical origin, is a mis- nomer. When the metropolitan nest is established the ants make a large entrance in Wheeler, Ants of the Belgian Congo 503 the adjacent wall of the trunk and through this and the other openings in the branches pass to and from the foliage (Wheeler). _ ~~ Maregravius in 1648 (p. 91)! first mentions the constant occurrence a of ants in the cavities of the stem of a Cecropia. Belt (1874, p. 222) leads the series of modern writers with his studies of Nicaraguan Cecropiz. _ _ Inhis opinion the ants protect the tree; he found the stems of Cecropiz _ inhabited by three species of ants and also by coccids attended by these __ insects. Fritz Miiller (1876 and 1880) described the origin of the colonies = of Azteca in Cecropia adenopus in Southern Brazil and called attention to the oval depression or prostoma by which the ants always enter the hol- low internodes. He also discovered the food-bodies produced between the hairs of the trichilium at the base of the petiole and saw that the ants ‘earried them off to their abodes. Schimper’s (1888) careful investiga- tions brought to light additional facts; he proposed the term “ Miillerian bodies”’ for the food-bodies produced by the trichilium. They are white, pear-shaped or oval bodies composed of cells rich in proteids and fatty oils, so they can not be regarded as excreta. Since they are of no use to the plant save to attract ants, Schimper believes that they were origi- nally mucus- or resin-glands which have become highly modified through Japtation to the ants. A similar adaptation is found, he thinks, in the prostoma and the peculiar structure of the stem at that particular spot where only soft parenchym and mucus-vessels are present. Both F. Miiller and Schimper consider the leaf-cutting ants (Attini) the chief enemies of the Cecropia against which the protection by A2teca is devised. These authors call the ant which they observed on C. adenopus, “ Azteca instabilis,” but, as shown by Emery, this is A. muelleri Emery and not Smith’s A. instabilis. Later observers, such as Ule (1897, 1905b, 1906b), Rettig (1904), H. v. Ihering (1907), K. Fiebrig (1909), and Wheeler (1908a, 1913), have offered many objections to the Belt-Schimper hypo- thesis of symbiosis between Azteca and Cecropia adenopus. Among other points, it is very doubtful whether the leaves of Cecropia are particularly attractive to leaf-cutting ants; moreover, the foliage of older trees which are occupied by Azteca is often much eaten by sloths, caterpillars, and other insects. Rettig calls attention to the presence on the leaves of Cecropia adenopus of bead-like glands containing proteids and fatty oils and which are also collected and used as food by the ants. H. v. Ihering found Azteca nigella Emery nesting in the internodes of younger plants, 2 to 3 m. high, while older plants with a “metropolitan ‘John Ray (1688, p. 1373) reproduces Marcgravius’ observations. 504 Bulletin American Museum of Natural History (Vol. XLV nest” in a spindle-shaped swelling of the bole were inhabited exclusively by A. muelleri Emery; he believes that this is a case of dimorphism between the younger and older generations of workers in the same colony. He has, however, not given any conclusive evidence that such is the case, since he has not observed transitional colonies of these two forms." H. v. Ihering also found A. lanuginosa Emery in Cecropia adenopus; he mentions the frequent occurrence of coccids (Lachnodiella cecropiz H. v. Thering) in the nests. In his opinion, the main food of the adult ants consists in the soft pith-parenchyma of the upper, still growing internodes, also in the Miillerian bodies. He was unable to find how the larve are fed. Fiebrig’s (1909) observations were made in Paraguay on what he calls ‘‘Cecropia peltata L.,’’ but what is evidently not the Central Ameri- can C, peltata but C. adenopus of Brazil.2. The internodes were practically always inhabited by Azteca alfaroi variety mixta Forel. The ants go only short distances from their exit holes, unless disturbed, when they become very aggressive. Fiebrig thinks that the main food of the ants is the Miillerian bodies, on which the larve are probably fed ex- clusively, while the workers may also eat soft pith tissues and feed on the sweet fruits of the tree. There is little doubt that A. alfarot is wholly vegetarian, while most other species of Azteca are carnivorous. In Paraguay the internodes of Cecropia are very often invaded by caterpillars (Heliothis species). The very young larve of this moth were repeatedly observed in internodes where a queen ant had just started a new colony; later on the caterpillars crowd the ants out and finally occupy the entire branch and destroy even the septa. Wheeler has called attention to the occurrence in Cuba and Porto Rico of species of Cecropia fully equipped with prostoma and Miillerian bodies, though never tenanted by Azteca, since this genus of ants is lacking on all the larger Antilles. Cecropia lyratiloba Miquel. Under this name a swamp Cecropia of southern Brazil was studied by H. v. Ihering (1907). It possesses the same so-called myrmecophilous structures as C. adenopus and is also inhabited by a species of Azteca. Cecropia sciadophylla Martius. Brazil. Inhabited by Azteca emeryi Forel (Ule). *Emery (1912, ‘Gen. Insect., Dolichoderine,’ p. 34) still regards nigella as a distinct variety of A. , *Chodat and Vischer (1920, p. 235) assert that A. adenopus is the only species of the genus found in araguay. Wheeler, Ants of the Belgian Congo 505 . Polygonacese 7 cosmopolitan family of 800 species, belonging to 34 genera. The ~ myrmecop tic forms are trees or bushes of South and Central America. ‘Triplaris C. A. Meyer. Tropical South and Central America. Re- c " aeented by ten species, all of which have apparently hollow internodes, but the branches are not inflated, though they are usually inhabited by Pee ants.! Tt has been claimed that in some cases the entrance to the cavity segs © Rmery (1894a) described Pseudomyrma arboris-sanctz from Bolivia, ___ in stems of a Triplaris (collected by Balzan); Ule found Pseudomyrma ___—s sericea variety rubiginosa Stitz (Stitz, 1913a) inside the stems of an un- identified Triplaris of Brazil. Of his Pseudomyrma arboris-sancte subspecies cuabionie, Forel (1004, pp. 39-40) has this to say: I discovered this race in March 1896 at Dibulla, at the foot of the Sierra Nevada de Sta-Marta, Columbia, in the following manner. Having laid my hand on the _ trunk of a young, green tree, about 4 meters high, and with large leaves, I was stung, and discovered on the trunk this Pseudomyrma, the cause of the sting. Noting the agressive behavior of these ants, I suspected a symbiotic relation between the tree and them, for other Pseudomyrmz which run on trees take to flight instead of attacking. Finding, however, no dry branch and no aperture, I was at first puzzled. On noticing _ some passing Indians, I had the tree cut down with their machetes. I then broke the flexible, fresh branches of the tree and found them all provided with a very narrow pith channel. These channels constituted, from one end to the other of all branches and twigs of the tree, the nest of the Pseudomyrmz, which were occupying them in a file, with their males, their larve, and their nymphs, having just room to cross over one another notwithstanding the slenderness of their body. This curious habitation perplexed me much and I was wondering where the female foundress of the formicary might have entered this perfectly green tree, without any dry branch and apparently Pao of the earliest records Sap ree ne er eb pear is that by oe (1849, pp. 262-263 This publication nl not yrmecologists, I hav reprodu ‘Weddell’ etie ae nt elng aces description ona an any which has eoobber tn been iced] Wedde i's subsequent authors Capers papel mate a Pseudomyrma and perhaps the one known as P. arboris- “Le trone, les branches et mine lo les plus petite rassen cmap des x den expices de ce genre sont fistuleux et lorsqu "elle est excitée, une odeur oy Brees comparable a celle pom lent les sees aisce i l'on vient accidentellement A trone d'un Triplaris et surtout A lui imprimer un choc, on voit les fourmis surgir centaines Saeaere par de petits canaux qui font comsmnuniquer avec l'extérieur son ¢ médullaire, au plus douloureuse en ap ty les piqdres d’aucun autre insecte que fe connaisse. chose ue, A quelque ue de leur vie que l'on examine les Triplaris dans leurs foréts, on soit toujours sar d'y rencontrer ces fourmis. I) eat encore bien curieux que, dans les Ruprechtia que quelques auteurs réunissent encore aux Triplaris, on n'en trouve jamais. Je ne crois pas que cet insecte ait été observé dans d'autres conditions que celles que j'ai notées; po boede depeche particuliérement adaptée a son genre de vie. J'ai eu occasion de l'examiner et méme atteintes dans bien des parties du Brésil, en Bolivie et au Pérou; et partout il m'a paru identique ntique, Uh plunour v ont signalé une partie des faits dont il vient d'étre questicn, et ils la fourmi du T'ré, yi an pare te cag de Latreille; mais je ne sache yas qu'cn lui ait ophaoe celui de Murmica triplarina. Elle est ordinaire- ment d‘un brun clair. Sa longueur est de 6 millimétres, et sa largeur de 1 millimétre; I’ atdomen est cylindrique et un peu attenué whe gt AAs geeees Soe apdin pop or gh 506 Bulletin American Museum of Natural History [Vol. XLV without any exit hole. After long, unsuccessful investigation of ‘all the branches, I inspected the lower portion of the trunk and finally discovered there the remains of an early branchlet, dried and broken off, but with a pith cavity communicating with the central cavity of the very trunk. It is by this old branch that the Pseudomyrmex came and went. Warming (1894) has published some interesting information on a Triplaris of Venezuela, which he doubtfully identifies as 7’. americana. The ants found in this plant belonged to a species of Pseudomyrma which he calls ‘‘P. mordax Meinert, ’’ a name not backed by any description in the literature. T. americana Linneus. South America. The earliest accurate account of myrmecophilism in the genus T’riplaris was published, it appears, by Robert Schomburgk (1838, pp. 264-267) for the species under discussion. After a description of this tree, which he found com- mon on the sandy banks of the inland rivers in Guiana, and often over- towering the other vegetation, he continues: The uncautious botanist, who, allured by the deceptive appearance, should approach the tree to pluck the blossoms, would bitterly rue his attempt. The trunk and branches of the tree are hollow, like those of the trumpet tree (Cecropia), and provided with partitions, which answer to the position of the leaves on the outside. These hollows are inhabited by a light brownish ant, about two- to three-tenths of an inch long, which inflicts the most painful bites. Its antennw are placed near the middle of the anterior portion of the head; mandibles triangular; peduncle of the abdomen with two rings; the anus hairy and provided with a sting or piercer. They fall upon their prey with the greatest virulence, and insert their mandibles almost instantly, as soon as they come in contact with any soft substance, emitting a whitish fluid; their bite causes swelling and itching for several days. If they find themselves captured, they attack and kill one another like the scorpions. The Arawak Indians call the tree Jacuna, and the ant Jacuna sae; the Warrows Epouahari, the literal translation being ant tree; the Caribis Itassi; the colonists, from its growth, “long John.” Richard Schomburgk (1848, II, pp. 449-450) also records his pain- ful experience with the same tree, which he found growing on the banks of the Barima and Barama Rivers, British Guiana. Penzig found the caulinary cavities of 7. americana, cultivated at Buitenzorg, Java, occupied by Dolichoderus bituberculatus (Morteo, 1904). T. Cumingiana Fischer and Meyer. Central America. Wheeler (1913) observed this species in Panama and writes about it as follows: These trees were 15 to 20 ft. high, with very slender trunk, smooth, light gray bark, and long, narrow, lanceolate leaves. When the trunk was cut down and split longitudinally, it was seen to have a very slender cavity in the centre and extending its full length, and communicating with a similar slender cavity in the centre of each branch. ‘This continuous system of cavities communicated with the surface by numerous slender galleries, excavated by the ants, and ‘terminating in small round orifices, which served as exits and entrances. Wheeler, Ants of the Belgian Congo 507 el mis: Each tree was occupied by a single large colony of Pseudomyrma arboris-sancte Emery. Wheeler adds: ‘‘as the Triplaris trees were isolated and as their bases must stand in the water during the rainy _ season, it is difficult to understand how the ants manage to exist, unless _ they remain rather dormant this season or find some hitherto unknown _ food supply on the foliage.” Recent, unpublished observations of Prof. 1. W. Bailey on 7. surinamensis, in British Guiana, however, show that the eavities of Triplaris contain great numbers of coccids from which the Pseudomyrmz obtain atleast much of their food. ____-*T. caracasana Chamisso and Schlechtendal. Venezuela. Trunk _ inhabited by ants (Karsten in Huth, 1887). Schimper (1888) examined branches sent to him by Ernst and curiously enough states that they ___ Presented no adaptations to ants: ‘the branches possess an inner cavity which is only 5-8 mm. wide and interrupted by diaphragms; round apertures, pierced by the ants, lead into the cavity.”’ He does not believe that there is any true symbiosis in this case. T. nolitangere Weddell. Brazil. Stem inhabited by ants (Huth, 1. surinamensis Chamisso and Schlechtendal. Brazil, Guiana. ‘Myrmecophytic (Spruce, 1908). T. Macombii Don. Smith. Guatemala. Wheeler (1913) says: This is a larger tree (than 7’. Cumingiana), often attaining a height of 30 to 40 ft., with more diffuse branches and large, coarse, ovate leaves. Early in January it began to put forth bunches of long, yellowish flower-spikes, which were covered with a deciduous sheath. The branches have much larger cavities than in 7. Cumingiana and the septa at the nodes are not broken through. On examining the surfaces of the branches, each internode is seen to be surrounded near its distal end by a circle of lenticels, and one of these, for some unknown reason, often becomes considerably en- larged and bears a long slit-shaped impression. It is in this impression that the queen ant makes the circular perforation that permits her to enter and take possession of the internodal cavity. The same observer found the cavities of this species occupied by several species of ants belonging to the genera Crematogaster, Pheidole, Tapinoma, and Iridomyrmex, but two species were especially common, a small, black, narrow-headed Azteca and the black Pseudomyrma sericea _ Mayr. None of these, however, are obligatory plant ants. T. Schomburgkiana Bentham. Brazil. Inhabited by ants (Spruce, 1908). Ule (1917) found in this species Pseudomyrma dendroica Forel and P. triplaris Forel. Pseudomyrma dendroica was originally described from specimens found by A. Geeldi in the pith channel of young, unidentified Triplaris on the Rio Purus, Brazil. Some of these plants having been introduced 508 Bulletin American Museum of Natural History . [Vol, XLV into the Botanical Garden at Par&, Geeldi observed that this ant soon invaded one the T'riplaris of the Garden which thus far had not been inhabited (Forel, 1904, p. 41). Ruprechtia C. A. Meyer. Tropical and subtropical South America. There are twenty species, most of which are said to possess solid branches; the following is perhaps an exception. R. Jamesoni Meisner. Brazil. The stem and branches are hollow and inhabited by ants (Spruce, 1908). Symmeria Bentham. This genus contains two species; one has been described from Senegambia; the other, S. paniculata Bentham, according to Spruce (1908), is an ant plant; it occurs in Guiana, northern Brazil, and curiously enough also in Sierra Leone. Coccoloba Jacquin (including Campderia Bentham). Tropical and subtropical America. A large genus, with about 125 species; only one of them has been mentioned as a myrmecophyte, but the others should also be studied in this respect. The common sea-side grape, Coccoloba uvifera Linnzeus, in Porto Rico, sometimes has ants nesting in some of the internodes; but these are facultative forms, such as Camponotus sexguttatus (Fabricius), more common elsewhere. This species, at least, cannot be regarded as a myrmecophyte (Wheeler, 1908a, p. 157). C. parimensis Bentham. British Guiana, Brazil. The stem and branches are hollow, but not inflated, and are inhabited by ants (Spruce, 1908). Myristicacese A small, exclusively tropical family, which, according to Warburg’s monograph (1897), contains 15 genera with about 240 species. Myristica Linnwus. Indomalayan Region; eighty species. In two related species from New Guinea, the internodes are in places swollen and hollow; these swellings are irregularly scattered along the branches, and their inner cavities do not communicate with one another; they are inhabited by ants, which pierce the entrances, often slit-like and placed on the side facing the leaf of the lower node. Warburg (1897), who has studied their histology, concludes that these swellings are prob- ably not hereditary, but produced by the irritation of the ants; he considers them true ant galls, not myrmecodomatia. There is, however, no experimental proof that ants can produce such swellings. M. subalulata Miquel (=M. myrmecophila Beccari). This species has been studied by Beccari (1884) and Warburg (1892; 1897); the latter figures (1897, Pl. x1) coccids on the inner walls of the swellings. Wheeler, Ants of the Belgian Congo 509 _ Mvhélerophylla K. Schumann. Swellings on the branches inhabited so ants (Schumann, 1890; Warburg, 1897). pt —M.z euryocarpa Warburg, of New Guinea, is perhaps also inhabited ae Iti is still somewhat doubtful whether these Myristice are true & a Monimiacew rc - ‘Tropical regions of both hemispheres. Represented by 250 species, —_ belong to 30 genera. _.. Kibara Endlicher. Eastern India, Malay Archipelago. With a about 14 species. ___—_—C&K.. formicarum Beccari. New Guinea. The branches are hollow and ipehs swollen at the internodes just beneath the insertion of the leaves; ie live inside together with coccids (Beccari, 1877, ‘Malesia,’. I, pt. pp. 189-192). _Anthobembix Perkins. New Guinea. Contains two species, one ot which i is a myrmecophyte. © _ -—sOA«.: hospitans (Beccari) (=Kibara hospitans Beccari). Branches ¥ Pibahaped below the nodes; these swellings hollow, pierced with _ apertures and inhabited by ants (Iridomyrmex scrutator Smith) together ___with coccids (Myzolecanium kibare Targioni) (Beccari, 1877, loc. cit.). Lauraces z _ Tropical and subtropical regions of both pemmiepheres. Includes -_-- 1100 species, belonging to 48 genera. _ Pleurothyrium Nees. Brazil, Peru. There are five species, of which the following three have swollen, fistulose branches and probably are myrmecophytes (Mez, 1888 and 1889; K. Schumann, 1888). P. cuneifolium Nees. Peru, Brazil. Pceppig has mentioned the occurrence of ants on this plant: “in ramulis revera fistulosis degunt formicarum agmina pessime pungentia.”’ Slits, 1 to 2 mm. wide, serve as entrances to the cavities (Mez, 1889, p. 471). P. Peppigii Nees. Peru. P. chrysophyllum Nees. Peru. Ocotea Aublet. Tropics of both hemispheres. About 200 species, some of which have pouches or bull#, more or less pronounced, placed in the axils of the side-veins and projecting towards the upper side of the leaf; such species are O. phillyrwoides (Nees) of Brazil, O. Mandonii Mez of Bolivia, 0. Bernouilliana Mez of Guatemala, and O. bullaia E. Meyer of the coastal region of Cape Colony and Natal. In the last-named 510 Bulletin American Museum of Natural History [Vol. XLV species the pouches are large pits with ciliolate orifices on the under side in the axils of the lowest one or two pairs of nerves, the pits corre- sponding to large hollow tubercles on the upper side. Whether these pouches are merely acarodomatia or occasionally settled by ants is not known. Nepenthacez . Oriental Region, the Seychelles, and Madagascar. Only one genus, Nepenthes Linnzeus, with some 60 species, one of which has been recorded as myrmecophytic, but the case needs further investigation. N. bicalearata Hooker fil. Borneo. The petiole of the pitcher- shaped leaves is curled up and, in the curled part, swollen and hollow. According to Shelford (1916), there is no evidence that this cavity is inhabited by ants; while Beccari (1884) saw an opening leading inside and apparently found ants in the swelling. Rosacesz Cosmopolitan. Includes 1700 species, belonging to 102 genera. Hirtella Linneus. Tropical America, with forty species; one species occurs in Madagascar. Myrmecophytism seems to be exceptional in this genus, as is also the case in Cola and Randia. H. physophora Martius. The cordate leaves have at the base of the blade a pair of compresso-globose sacs inhabited by ants (Spruce, 1908). Leguminosx# Cosmopolitan, with 12,000 species and 530 genera. This and the Composit are the largest families of plants. Acacia Willdenow. Tropical and subtropical regions of both hemispheres. There are over 600 species. The so-called bull’s-horn acacias of Mexico, Central America, and Cuba are apparently true myrmecophytes; their stipular thorns are much enlarged and flattened or inflated; they are usually hollowed out by ants, which pierce an entrance below the tip of the thorn, more rarely near its base, and establish their nests inside; furthermore, the young leaves bear at the tips of their pinne, minute, bright yellow food-bodies (Beltian bodies)! which are eagerly collected by the ants and carried inside the thorns. These plants all grow in dry or semi-desert regions under conditions very different from those of other myrmecophytes. 1Meneghini and Savi (1844), Fr. Darwin (1877), and A. F. W. Schimper (1888), who have the inner structure and development of these Beltian bodies, all agree that they are homolcgues of glandular serrations which frequently occur on the margins of young leaves. Such glands mucus or resin and, as a rule, disappear at an early stage; while in the ant acacias they increase con- siderably, are filled with proteins and fats and, when not removed by the ants, finally drop off. _Whedler, Ants of the Belgian Congo 511 ee a ~ One of of the Mexican species was figured and described by Francisco 2 idienos 3 in 1651 (p. 86, Cap. LIII) as Arbor cornigera or the Huitz- - mamaxalli (“forked-thorn”) of the Aztecs. In accordance with the __ ideas of his time, Hernandez believed that the thorns themselves gener- __ ated the ants: “generantur preterea intra corniculas formic quedam ___ tenues fulvzeque et nigricantes.”” Linneus’ Acacia cornigera, however, is __ an altogether different plant and was described from a cultivated speci- __- men growing in the garden of George Clifford, between Haarlem and ____ Leyden, Holland; its origin is unknown. In fact, until quite recently, such confusion existed in the classification of bull’s-horn acacias that it is almost impossible to recognize the species on which ecological observa- ____ tions have been published by Belt (1874), Beccari (1884), Wheeler (1913), _ Wasmann (1915a), and others. s ‘HH. Schenck (1913, 1914) and W. E. Safford (1910, 1914, 1915) have _ shown thatthe bull’s-horn acacias contain a number of more or less re- _ lated forms which are probably only partly known; twenty-seven species have thus far been described. It must be expected that these numerous allied forms, which often differ markedly in size and shape of their thorns, will be found to harbor a corresponding variety of guest ants. _ Owing to the uncertainty of identification of the plants studied by vari- ous authors, the following list of ants will merely give a general hint as to the species which may be expected in these plants. 2 It may be of interest to note that some bull’s-horn acacias have been ___ eultivated in hothouses in Europe (Commelin; Linnzus; Beccari) and in certain botanical gardens of the tropics (in Java, Raciborski; in Ceylon, Ridley; in Gaboon and Cameroon, H. Schenck; also in Cuba, according to Wheeler). The thorns are then swollen and hollow, as on the _ wild-growing plants, but are not attacked by ants.’ Raciborski (1900) _ remarks that the food-bodies of such acacias are not collected by the ants in Java and that this is true also for the Miillerian bodies of the Cecropiz which he saw cultivated at Buitenzorg. Belt (1874), in Nicaragua, found in the thorns of his “‘ Acacia corni- gera”’ specimens of Pseudomyrma gracilis (Fabricius), = P. bicolor Smith, and more rarely of a Crematogaster. Emery (1890 and 1891) has given a long list of ants found by Alfaro in the thorns of unidentified Costa Rican acacias; only three of these, however, Pseudomyrma baits Emery, 'Dr. W. E. Sofiord ba: « we reg lorms me that a number of bull’s-horn acacias are now being cultivated eee oS oeingen. D.C. In each case the swollen thorns have maintained their char- in spite of the absence of ants. Prof. Wheeler saw two Central American bull’s-horn ore wowise in the Botanical Garden of Port of Spain, Trinidad. All their thorns were inhabited by ates po Be which even enveloped some of the thorns with carton. The ants were numerous 512 Bulletin American Museum of Natural History ([Vol. XLV P. spinicola Emery, and P. nigrocincta Emery, he considers obligatory acacia ants': “these species occur only on‘acacias, while other species of the same genus burrow their nests in wood; all three pierce the thorns close to the tip, when they are still young and soft, as Belt describes it; never was more than one of these three species found on a single tree and in each case the ant inhabited all the thorns on the living branches of the acacia.’”’ When the branches die, these Pseudomyrme leave the thorns, which are then occupied by many other ants: Pseudomyrma gracilis variety mexicana Roger, P. subtilissima Emery, P. nigropilosa Emery, P. kiinckeli Emery, Crematogaster brevispinosa Mayr, Cryptocerus minutus (Fabricius), Camponotus rectangularis Emery, and others; some of these species may occasionally invade young thorns of living branches, but, as they often occur elsewhere, they must be designated as facultative guests of the plant. Wheeler (1913) found Pseudomyrma spinicola Emery on “Acacia sphzxrocephala”’ in Panama, and P. belti Emery with its subspecies ful- vescens Emery on “A. cornigera”’ and “A. Hindsii” in Guatemala. Dr. P. P. Calvert, moreover, sent him P. belti and P. nigrocincta taken from acacia thorns in Costa Rica. Wheeler agrees with Emery that these four forms are, so far as known, the only obligatory acacia ants of Central América; among the facultative acacia ants he mentions Camponotus planatus Roger, Pseudomyrma gracilis (Fabricius), and Solenopsis species, taken by him in Guatemala; also Pseudomyrma nigropilosa Emery found by Calvert in Costa Rica. Wasmann (1915a) described Pseudomyrma wasmanni Whadiee'= P. canescens Wasmann, nec Smith, from the swollen thorns of “ Acacia sphzrocephala”’ collected at Tampico, Mexico. South American ant acacias are thus far known from Paraguay only. J. Bohls collected there in woody, expanded thorns of an unidentified acacia eleven species of ants: Pseudomyrma acanthobia Emery and variety fuscata Emery, Cryptocerus pilosus Emery, C. bohlsi Emery, C. peltatus Emery, C. quadratus Mayr, C. pallens Klug, C. pusillus Klug, C. grandinosus F. Smith, Crematogaster brevispinosa Mayr, and Myrme- lachista nodifera variety flavicornis Emery. In his report of this collec- tion Emery (1896a) remarks: “I have found most of the thorns (sent by Bohls) which still contained ants, inhabited by Pseudomyrma, which had its narrow galleries burrowed in the wood. The large Cryptocerus had completely hollowed otit the thorns occupied by them. The openings asi” terms “ obligatory” and “facultative” as applied to acacia ants were proposed by Wheeler Wheeler, Ants of the Belgian Congo 513 of the Pseudomyrma nests were placed not far from the tip, those of the other species pierced at various levels, often also several on one thorn.” The only other observations on these interesting plants were made by =. Fiebrig (1909), who studied Acacia cavenia Hooker and Arnott in the __ Chaco of northern Paraguay; the thorns of this species are very large, 90 mm. long and 8 mm. wide, and usually inhabited by Pseudomyrma . __ fiebrigi Forel; normal thorns are filled with pith; in those occupied by ants that substance is more or less removed and an opening is found be- a low the tip. Frequently, however, the pith is destroyed by a caterpillar oo which pupates inside, the moth escaping through a hole near the point Pe. of the thorn. Fiebrig believes that the ants appropriate these excavated a thorns, using apertures made by the moth. According to Chodat and ___ Carisso (1920), the swelling of the thorns of A. cavenia is due to the sting ___ of an insect, the gall thus produced being eventually settled by ants, after its maker has left it. I cannot agree with this explanation. ____ In a foregoing chapter (p. 372) I have discussed the so-called ant -_ acacias of East and South Africa and have given my reason for not re- garding them as true myrmecophytes. In their case, the swellings of the thorns are typical insect galls, probably produced by a lepidopterous larva. When the gall maker has left, the empty shelters may be invaded ___ by various ants, even before they are completely dry, thus simulating | mayemhccodomatin. _ Selerolobium Vogel. Tropical South America. Containing twelve species. _ Only one of the species, S. odoratissimum Spruce, of Brazil (Rio Negro), is said to be myrmecophilous; its leaves have a large sac, fur- rowed along the upper face and extending upward from the knee of the petiole to the base of the second pair of leaflets (Spruce, 1908). It is _ possible that this pouch is merely an insect gall which, when empty, be- comes settled by ants. ) Humboldtia Vahl (= Baischia Vahl). Ceylon and British India. Represented by four species, one of which is myrmecophilous. _ H. laurifolia Vahl. India. The swollen internodes are occupied by ants (Bower, 1886 and 1887; Schimper, 1903; Morteo, 1904; Ridley, 1910). Figured by Taubert, 1894, in Engler and Prantl, ‘Die Natiirl. Pflanzenfam.,’ III, pt. 3, p. 143, fig. 80, and by A. F. W. Schimper, 1903, ‘Plant Geography,’ p. 147, fig. 83; this figure is also copied by Escherich (19066) and Wheeler (1910b). Escherich (19lla, pp. 46-47) re-examined H. laurifolia in the Botanical Garden at Peradeniya, Ceylon. He found that only compara- 514 Bulletin American Museum of Natural History (Vol. XLV tively few of the swollen internodes (at most 20 per cent in the Garden, as contrasted with 50 per cent in the wild state, according to Green) contained a number of species of ants that are also found nesting in other locations (T'echnomyrmex, Tapinoma, Monomorium, Crematogaster, ete.). Since the ants are not in the least aggressive and, furthermore, often keep coccids inside the domatia, he concludes that they are decidedly noxious to the plant, the more so since they frequently attract woodpeckers which damage the branches in order to feed on them and their brood. Schotia Jacquin (= Theodora Medikus). Tropical Africa. There are twelve species, one of them possibly myrmecophytic. S. africana (Baillon) (=S. humboldtioides Oliver). Cameroon, Spanish Guinea, Gaboon. The young branches often have swollen and hollow internodes settled by ants. There is still a possibility that these enlargements are mere insect galls, which are invaded by ants after being left by their makers (see above, p. 409). Tachigalia Aublet (=Cubzxa Schreber; Tachia Persoon). South America. Includes six species, all of which have inflated petioles in- habited by ants. Pseudomyrma picta Stitz and Azteca brevicornis (Mayr) were found in Tachigalia by Ule in Brazil (Stitz, 1913a). T. caripes Spruce. Brazil. The trigonous petioles are mostly di- lated at the base into a fusiform sac tenanted by ants (Spruce, 1908). T. ptychophysca Spruce. Brazil. Like the preceding (Spruce, 1908). T. formicarum Harms. Eastern Peru. The petiole is swollen and inhabited by Pseudomyrma (Ule, 1908). Platymiscium Vogel. South America. Contains fifteen species. The stem is hollowed and inhabited by ants, and even sometimes dilated at the nodes (Spruce, 1908). Meliacesz Tropical and subtropical regions of the globe. Has 42 genera, with about 700 species. . Chisocheton Blume. Indomalayan Region. About thirty species. C. pachyrhachis Harms. New Guinea. A tree with the nodes of the branches and the base of the petiole swollen and hollow; several aper- tures leading into the cavity (K. Schumann and K. Lauterbach, 1901, p. 382). Aphanamixis Blume. Indomalayan Region. Includes eleven species. A. myrmecophila (Warburg) (=Amoora myrmecophila Warburg). New Guinea. The branches are often swollen and excavated, even the younger upper portions, the growing extremity narrowing very abruptly; 1922) . Wheeler, Ants of the Belgian Congo 515 several apertures lead inside the irregular cavities, which have smooth, brown walls; the swellings are inhabited by ants (Warburg, 1894, pp. gd Euphorbiacezw ae A large, cosmopolitan family, with 4500 species, belonging to about s Endospermum Bentham. Indomalayan Region to New Guinea. ____ Ineludes twelve species, two of which are to all appearances true myrme- _ E. moluccanum (Teysmann and Binnendijk). Amboina, Moluccas, _ Celebes. There is a question whether this species is myrmecophytic. According to Beccari, this is the plant figured by Rumphius (1741, I, pp. 257-259, Pl. uxxxv) as “ Arbor Regis’”’.' In the latter’s description, however, a confusion may have been made between several plants; so that it is by no means sure that the following remarks concerning the myrmecodomatia of his Arbor Regis apply to E. moluccanum: _ Truncus, omnesque crassi rami nullo constant corde, sed excavati sunt, ejusque loco referti sunt plurimis magnis et nigricantibus formicis, que in una alterave parte truncum perforant, et fenestras quasi formant, perambulantes illum usque ad ramorum extremum tanquam murum concavum, ita ut hee arbor solo ex cortice suum hauriat nutrimentum, tenuiores vero rami medullam gerunt, qualem Sambucus habet. Si quidam amputetur ramus, formic he magna vi ac celeritate excurrunt, mox circumstantes invadentes homines ac mordentes tanto impetu, ut periculosum valde sit huic accedere arbori, immo totum circa hance solum mordentibus hisce animalibus, repletur, que adpropinquantium etiam pedes infestant. Observavi autem Indos non ita horum morsus presentire per duram ipsorum cutim, ac nos, unde et intrepide ad illam accedunt arborem. The relations of EZ. moluccanum to ants have apparently not been studied in the field since Rumphius’ time. | E. formicarum Beecari. New Guinea, Bismarck Archipelago. In New Guinea, according to Beccari (1884), the branches are normally swollen and hollow toward their extremity; he found them inhabited by Camponotus angulatus Smith, which had apparently pierced the entrances to the cavities. Dahl (1901) describes this plant in the Bismarck Archi- pelago as having normal branches, filled with pith which is partly ex- cavated by ants, Camponotus (Colobopsis) quadriceps (Smith). Macaranga DuPétit-Thouars. Tropical and subtropical regions of the Old World. About 170 species, a number of which are myrmecophy- tic. ‘Merrill (1917) admits the correctness of Beccari's reduction of Rumphius’ “ Arbor Regis"’ to B. moluccanum. c | : bal we 516 Bulletin American Museum of Natural History (Vol. XLV Donisthorpe (1917) described Dolichoderus (Hypoclinea) crawleyi from Singapore, “associated with species of Lecanium (coccids) in hol- low stems of Macaranga.’’ Wheeler (1919, p. 77) also mentions Crema- togaster (Decacrema) decamera (Forel) “from Macaranga with slightly trifid leaves” at Kuching, Borneo. M. hypoleuca (Reichenbach fil. and Zollinger). Malay ‘Puiltleette’ Sumatra, Borneo. Inhabited by forms of Crematogaster re borneensis (Ern. André) (Viehmeyer, 1916). M. caladifolia Beccari. Borneo (Beccari, 1884). M. formicarum Pax and O. Hoffmann. Borneo. A low tree with thick, hollow branches which are pierced with an entrance and inhabited by ants (Pax, 1914). In these three species ants live within the hollow, slightly swollen stem and branches, and also underneath the lanceolate, erect, persistent bud-bracts in the axils of the leaves; food-bodies, white and globular, are scattered on the back of the young leaves between the raised veins. The food globules are most plentiful in plants not settled by ants, and have been seen carried about between the mandibles of these insects (Beccari, 1884; Ridley, 1910; Pax, 1914; Shelford, 1916). M. triloba (Reinwardt). Malay Peninsula and Archipelago. M. Griffithiana Mueller. Malay Peninsula. M. Hulletii King. Malay Peninsula. In these three species, the stems are also hollow and settled by ants; furthermore, the bud-bracts are reflexed into a ring-like pouch which almost completely surrounds the stem. The concave under side of the bracts bears abundant pear-shaped or globular, white food-bodies, which are much sought for by the ants and are conveyed to the nest in the hol- low stem, where the larve are fed on them; the ants not only hide be- neath the bracts but occasionally take their larve there. M. Hosei King possibly has similar myrmecodomatia. The ant of M. triloba is a Crematogaster near C. daisyi (Forel) (W. Smith, 1903; Ridley, 1910). In an unidentified species of Macaranga of Sarawak, the bracts are very large, lanceolate, acuminate, deflexed, coriaceous, not appressed to the stem, but concave, thus providing a nidus or feeding ground for ants (Ridley, 1910). M. saccifera Pax. M. Schweinfurthii Pax (=M. rosea Pax). The above two species are from Tropical Africa and have pouch- like stipules, which in M. saccifera are sometimes inhabited by ants of the genus Crematogaster (see above, p. 412). nn iil i Sl a ae ae meee hs : hd ' hl Wheeler, Ants of the Belgian Congo 517 ree __ Mabea Aublet. South America. Contains thirty species, some of Sterculiaces Tropical regions of both hemispheres. Represented by 820 species * - Cola Schott and Endlicher. Tropical Africa. With forty-five 5 species. The following three closely allied forms have at the base of the _ leaf-blade a pair of pouches which are often inhabited by small species of 2 a (see above, p. 417). _____—C. Dewevrei de Wildeman and Durand. ___ C. Laurentii De Wildeman. _ C. marsupium K. Schumann. p _ Scaphopetalum Masters. Tropical Africa. Includes eight species, = two of which have an elongate pouch at the base of the leaf-blade often -_ eceupied by ants of the genus Engramma (see above, p. 422). ____ §&. Dewevrei De Wildeman and Durand. Belgian Congo. __§. Thonneri De Wildeman and Durand. Belgian Congo, Cameroon. 4 Flacourtiaces Tropical regions of both hemispheres. With 650 species and 84 genera. Barteria J.D. Hooker. Tropical Africa. Includes four species, all ___ of which probably have hollow or swollen internodes, normally inhabited by Pachysima xthiops (F. Smith) or P. latifrons (Emery); accidentally __ by other ants (see above p. 432). B. Dewevrei De Wildeman and Durand. Belgian Congo. _ B. fistulosa Masters. Fernando Po, Cameroon, Belgian Congo. B. nigritana J. D. Hooker. Southern Nigeria, Cameroon, Spanish Guinea, Gaboon, (Belgian Congo?). __B. Stuhlmannii Engler and Gilg. German East Africa. Gertrudia K. Schumann. New Guinea. With one species, G. amplifolia K. Schumann. It is a tree or shrub with branches “strongly swollen at the apex below the leaf-bud, hollow and with an aperture leading into the cavity (probably a myrmecodomatium) ” (K. Schumann and K. Lauterbach, 1901, p. 455, Pl. xv). Perhaps this swelling is only an insect gall. Melastomatacesx Tropical and subtropical parts of both hemispheres; very abundant in America, where a few forms reach the Nearctic Region. Represented by 2800 species and 170 genera. With the exception of Pachycentria, 518 Bulletin American Museum of Natural History [Vol. XLV which is a doubtful myrmecophyte, all the myrmecophytic members of this family are restricted to the Neotropical Region. Tococa Martius. South America. Includes forty species which, with one or two exceptions, have ant-pouches on the leaves. Either all the leaves or only one of each pair have a hollow sac or pair of sacs at the base of the blade, or in the upper part of the petiole; these pouches are usually inhabited by ants (species of Azteca’). T.. disolenia Spruce. Brazil (Spruce, 1908). T. bullifera Spruce. Brazil (Spruce, 1908). T. macrophysca Spruce. Brazil (Spruce, 1908). T. formicaria Martius. Brazil (Spix and Martius, 1831). T. guianensis Aublet. Guiana. Aublet (1775) describes the two pouches which in this species are placed along the upper part of the petiole, each with an opening beneath the base of the leaf-blade; ants are usually found in them and from the description it would seem that they also inhabit the stem of the plant. Microphysca Naudin. Northern Brazil and Peru. Contains two species, M. quadrialata Naudin and M. rotundifolia (Spruce), with pouches on the leaves. Myrmidone Martius. South America. There are two species, both with sacs on the leaves shaped much as in certain forms of Tococa. M. macrosperma Martius. Brazil (Spruce, 1908). M. rotundifolia Spruce. Brazil (Spruce, 1908). Maieta Aublet (including Calophysca de Candolle). South America. Includes eight species, probably all with ascidia serving as abodes for ants. M. guianensis Aublet (=M. hypophysca Martius). Guiana, Brazil. The branches are fistulose and swollen at the nodes; the leaves also bear pouches (Spruce, 1908). To judge from his figure, this is the unidentified melastomataceous plant alluded to by Belt (1874, pp. 223-224) in the following passage: Tn each leaf, at the base of the laminew, the petiole or stalk is furnished with a couple of pouches, divided from each other by the midrib, as shown in the figure. Into each of these pouches there is an entrance from the lower side of the leaf. I noticed them first in Northern Brazil, in the province of Maranham; and afterwards at Paré. Every pouch was occupied by a nest of small black ants; and if the leaf was shaken ever so little, they would rush out and scour all over it in search of the aggressor. I must have tested some hundreds of leaves, and never shook one without the ants coming out, excepting one sickly-looking plant at Par4. In many of the 1 Azteca traili Emery was found in the ascidia of a melast t plant by Schulz at Pard. 1923) Wheeler, Ants of the Belgian Congo 519 the caus and young ants, and in some I saw a few dark-colored coccidee or cera _ _M. tococoidea (de Candolle). Brazil, Peru, Guatemala. A large bifid sac at the base of the petiole (Spruce, 1908). Pachycentria Blume. Malay Archipelago. Includes twelve species. __ These are woody epiphytes, some of which have tuberous swellings on the roots, filled with a spongy tissue. Ridley did not find any ants inside these enlargements and doubts whether the plants are true myrmeco- _ phytes. It is probable that the swellings are merely tubers. P. macrorhiza Beccari. Borneo. Tuberous and galleried roots in- habited by ants (Shelford, 1916). P. microstyla Beccari. Borneo. Like the preceding (Shelford, 1916). Medinilla Gaudichaud. India, Malay Archipelago, Oceania, Mada- gascar, tropical Africa. Contains over 100 species. M. loheri Merrill. Luzon, Philippine Islands. Only one of the leaves in each pair is normal; the other is modified into a crop-shaped ascidium opening on the upper side with a slit. According to Loher’s observations, this pouch is sometimes occupied by ants, the species of which is not stated (Solereder, 1920). M. disparifolia C. B. Robinson. Luzon, Philippine Islands. The leaves have a similar structure as in the foregoing, and are perhaps also used by ants. Loganiacez Tropical regions of both hemispheres. Represented by 400 species and 35 genera. : Fagrea Thunberg. Oriental Region. Contains twenty-five species. In the three forms enumerated below, the base of the petiole bears auri- culate appendages, which are curved downward and more or less pressed against the stem. The cavities thus formed are occupied by ants, which cover the opening with a papery substance and keep their brood inside (Burck, 1891). F. borneensis Scheffer. Borneo. F.. imperialis Miquel. Sumatra. _ F. auriculata Jack. Oriental Region. Gentianacee Cosmopolitan. Represented by 71 genera, with 900 species. Tachia Aublet (=Myrmecia Schreber). South America. There are four species. Bushes or small trees. The stem and the long, slender 520 Bulletin American Museum of Natural History [Vol. XLV branches are hollow. In the original description of 7’. guianensis Aublet, of Guiana, there is a note as follows: “Le trone et les branches qui sont creux, servent de retraite aux fourmis; e’est pour cette raison que cet arbrisseau est nommé ‘Tachi’ par les Galibis, ce qui en leur langue signifie, suivant leur rapport, ‘nid de fourmis’” (Huth, 1887; Spruce, 1908). Apocynacesz Cosmopolitan, though chiefly in tropical regions. Represented by 165 genera containing 1300 species. Epitaberna K.Schumann. One species, Z. myrmecia K. Schumann, in Cameroon: upper part of the internodes swollen, spindle-shaped, with a cavity inhabited by ants (see above p. 442). Asclepiadacez Cosmopolitan; chiefly in tropical and subtropical regions, and abundant in Africa. Represented by 267 genera, with 2200 species. Dischidia R. Brown (including Conchophyllum Blume). Oriental Region. Includes fifty species. They are all twining epiphytes; a few are associated with ants. D. Rafflesiana Wallich. Malay Region. D. timorensis Decne. Malay Region. In these two species a certain number of leaves are converted into cone- or pitcher-shaped pouches with an opening at the base through which roots project into the cavity; this pouch also contains soil and sometimes ants, which make regular nests there, with brood (Treub, 1883a; Beccari, 1884; Groom, 1893; Ridley, 1910). The seeds are scattered by ants (see above, p. 357). Beccari found D. Rafflesiana in Java inhabited by Dolichoderus bituberculatus Mayr and Crematogaster brevis Emery. D. complex Griffith. Malanch (Pearson, 1902). D. pectenoides Pearson. Philippines (Pearson, 1902). D. Shelfordii Pearson. Borneo (Pearson, 1903; Shelford, 1916). . In the above three species a certain number of leaves are double pitchers; a small pitcher is found inside each large pitcher; the inner surface of the former is thickly beset with glandular hairs; the larger, outer pitcher is filled with soil and numbers of rootlets, which spring from the petiole or stem and grow through the orifice; in the outer one are found also numbers of ants, Crematogaster difformis F. Smith. ‘Microscopic examination of the inner surface of the outer pitcher re- vealed the presence of a dense waft of superficial mycelium which was Wheeler, Ants of the Belgian Congo 521 easily removed on the point of a needle. The growth of this mycelium _ appeared to be radial, starting from the center of a curious rosette-like __ structure, formed by shorter hyphzx of a peculiar character. These bore a profuse crop of minute abstricted gmme. At the center of each rosette the tissue of the pitcher-wall appeared to have been punc- ___ tured” (Pearson, 1902, p. 387). _---_—-—- The following three species are doubtful myrmecophytes: D. Mer- ___— guiensis Beccari, of Tenasserim; D. clavata Wallich, of India; and D. __ digitiformis Beccari, of Celebes. ae Borraginacere Cosmopolitan. About 100 genera, with some 1600 species. _ Cordia Linnzus. Tropical regions of both hemispheres. Contains _ 250 species. A few of the South American forms are apparently true myrmecophytes. Ule collected Cryptocerus cordix Stitz from an uni- dentified Cordia in Brazil (Stitz, 19132) and Azteca longiceps Emery subspecies cordincola Forel was taken from the swellings of a Bolivian species (Forel, 19202). Chodat and Carisso (1920) regard the caulinary swellings of the species of Cordia examined by them in Paraguay, as mere insect galls, subsequently occupied by ants. It can hardly be _ doubted, however, that they are true myrmecodomatia. . ____-@, Gerascanthos Jacquin. Central and South America. Beccari (1884) noticed on herbarium specimens from Mexico, below the terminal verticil of branches, an obovate, hollow swelling of the stem with a lateral aperture; there were coccids, but no ants inside. Spruce (1908) found these swellings inhabited by ants; and Emery (1890) records Pseudomyrma belti subspecies fulvescens Emery from this plant in Guate- mala; while Azteca pittieri Forel variety emarginatisquamis Forel occurred in specimens from Costa Rica (Forel, 1920a). ’ @€. nodosa Lamarck. Brazil. Beccari (1884) and Spruce (1908) mention that the stems are swollen and hollow beneath the nodes and settled by ants. Schimper (1888) has studied the myrmecodomatia of this species near Pernambuco: below the false verticil of leaves, side twigs, and inflorescences which terminates the main branches, one frequently finds an elongate, pouch-like swelling which opens above by a small natural aperture placed between the leaves and branches of the false verticil. These pouches are often inhabited by small ants and in such cases their inner wall is covered with a dark brown, earthy crust evidently produced by the ants.' See also Rettig (1904). Adzteca stanleyuli Forel and A. olitrix Forel were taken from swellings of C. nodosa collected near Par&, Brazil (Forel, 1920a). ‘It may be supposed that this dirty layer contains a mycelium as in the case of Myrmecodia. 522 Bulletin American Museum of Natural History (Vol. XLV C. longituba Chodat and Vischer. Chodat found in the swellings of this species in Paraguay nests of Pseudomyrma chodati Forel (Forel, 1920a). C. miranda de Candolle and C. hispidissima de Candolle possess, according to Beccari (1884), similar myrmecodomatia; they form, together with C. nodosa, a special section of the genus (Physoclada A. de Candolle). Verbenaces Cosmopolitan, but mostly in tropical and subtropical climes. Represented by 900 species, belonging to 80 genera. Clerodendron Linnzus. Tropical regions of the Old World. About 200 species. C. myrmecophilum Ridley. Malay Peninsula, Sumatra, Borneo. C. breviflos Ridley. Malay Peninsula. C. fistulosum Beccari. Borneo. These three species have normally hollow branches, which are often inhabited by ants (Beccari, 1884; Ridley, 1910; Shelford, 1916). According to Beccari and Shelford, the ant of C. fistulosum is Camponotus (Colobopsis) clerodendri Emery; it gnaws entrances to the hollow stem ‘always directly below the insertion of the leaves, either on one or on both sides of each node; on plants free from ants, these spots are marked by a little circular patch of a texture and structure different from that of the surrounding parts. Beccari also describes and figures the internodes as markedly swollen, and more so towards their upper extremity. An unidentified species of Clerodendron in the Belgian Congo also shelters ants inside its hollow branches (see above, p. 443). Vitex Linneus. Tropical regions of both hemispheres. With 120 species. Two myrmecophytic species have been mentioned as occurring in Africa, and probably some others also shelter ants inside their stem. V. Staudtii Guerke. Togo, Cameroon, Spanish Guinea, Belgian Congo. Creeper with hollow stems and branches, which are inhabited by Viticicola tessmanni (Stitz) (see above, p. 447). V. yaundensis Guerke. Cameroon. Rubiacez One of the largest families of plants: over 5000 species, classed under some 400 genera, have been described. They are cosmopolitan, though the majority are found between the tropics. About sixty-five species belonging to eleven genera present myrmecodomatia, this family thus containing by far the largest number of myrmecophytes. Wheeler, Ants of the Belgian Congo 523 ‘Myrmecodia Jack.! Oriental Region, from Cochinchina and the 4 Malay Peninsula to New Guinea, northern Queensland, the Solomon a and Fiji Islands. There are eighteen species. All are epiphytic, low _ shrubs, with rhizomes swollen into basal pseudobulbs or tubers, occupied __ by anastomosing cavities which communicate with the exterior by means of numerous pores and are often inhabited by ants; the apertures seem to be formed naturally, without the intervention of the ants, at least in _ Beceari originally (1884) held that the galleries of the swollen rhizomes were the work of ants; that it was impossible for plants to reach maturity without the intervention of these insects; that the tun- _nelling by them caused the tuber to grow enormously, while its weight was not proportionally increased, the galleries thus enlarging the ab- sorbent surface of the rhizomes.? Later he altered his views somewhat, as can be seen in the following quotation from his ‘Wanderings in the - Great Forests of Borneo’ (1904, p. 405): At first I thought that the ants by the irritation they produced on young budding plants of Myrmecodia, favored the swelling of the base of the stem, and were the direct cause of such an hypertrophy. Further investigations and researches and the _ observations of Dr. Treub have, however, convinced me that from the very beginning * these swellings appear independently of any action of the ants, and that when the latter are absent the tubers develop much in the same manner. I do not, however, think it equally certain that ants have no part in the formation of the internal gal- leries. My observations tend to prove that in some cases, in non-Bornean species of Myrmecodia (M. alata and bullosa), ants take an active part in the formation of the galleries and especially in that of the apertures which lead to them. But be this as it may, the hospitating Rubiacez live on a footing of reciprocal utility or mutualism with their inhabitants, which act as a formidable army of defence, for no animal dares to meddle with a plant guarded by a host of biting ants, ready to assault the im- prudent invader in myriads. H. O. Forbes (1880 and 1885, pp. 79-82) and Treub (1883, 1888) raised young Myrmecodiz from seed and found that the tuber is a normal production of the plant and that the galleried inner structure arises in the absence of ants. Treub’s investigations are of such importance that they should be considered more in detail. He saw that soon after germination and before the first leaves are formed, the axis below the tuber te theca (1750, VI, roams PL. ty) first discovered the remarkable East Indian Rubiacew with ant- takers, 2 diningelahed wo Linde: Nudes Jormicarem niger, (Ulgdnopiaium embeineee Berar icarum? 3 ‘ia Rumphii i). that not wellin ane eee plant were ceed by the ants! in tisse has given a aioe e coer tum i ton Canes the tem ame $y blants the ants contin rritate and excavate & swelling, it assumes a may become than a He also believed that th lants can- 6 Gobel form thrive Aa they the oo ants. —_ r “Nees me | ” 524 Bulletin American Museum of Natural History (Vol. XLV cotyledons begins to enlarge and it is from this part of the plant that the whole tuber is produced. When the swelling is quite young the entire mass of cells, including the central bundle, is continuous; but when older, some of the central cells have dried up and thus form the first cavity whose inner walls are covered with a layer of suberose cells; later other galleries are formed, which at an early stage communicate with one another. Treub also apparently admits that the entrances to the cavities are produced by the Myrmecodia itself without any outside help. In his opinion, the tuber and inner labyrinth are normal ecological peculiarities of the plant, the latter being used for aérating purposes. The walls of the galleries are in some parts smooth and uniform, in others studded with little prominences, which Treub thinks are not, as origin- ally supposed, glands secreting some fluid attractive to ants or absorb- ing organs for nutritive substances, but lenticels or rudimentary breath- ing organs. The ants he regards as mere opportunists who have taken advantage of the secure shelter afforded by the excavated tubers, but are of no visible utility to the plant. G. Karsten (1895) also disclaims the supposed symbiotic relations between the Myrmecodiz and ants. He believes, however, that the cavities have not only a respiratory function, but that their inner walls can also absorb transpiration water condensed inside the tubers during the cooler nights and at the same time assimilate certain dissolved nutritive elements introduced by the ants or found in the excrement of these insects. Rettig (1904) agrees with Karsten and Treub in explaining the peculiarities of the Myrmecodiz and allied genera on the ground of the physiological needs of the plant. He notes that these epiphytes are light-loving, thriving in nature on branches which are much exposed to intense sunshine or even on rocks; the galleries of the tubers are filled with air and act as aérating tubes, which isolate the inner tissues and prevent the plant from drying out. This author does not discard Treub’s idea that the pimples on the inner walls may be for respiration; he even observes that there is undoubtedly a current of air through the apertures, _ since fresh air enters during the cooler nights and partly escapes during the day. He believes, however, that in many cases rain-water enters the cavities through the openings and is then absorbed by the tuber; he has shown experimentally that such absorption can actually take place. Our knowledge of the Myrmecodiz has been materially increased by Miehe’s (1911b) researches. According to his findings, the inner walls of the cavities of Myrmecodia tuberosa are, as a rule, clean; those in Whesler, Ants of the Belgian Congo 525 certain portions a1 are ed of a brownish-yellow color, never covered with fungi, and the pupz of the ants are always kept in such galleries only; others are blackish, strewn with paler papillx, the dark color being _ due to a covering of fungus. This growth occurs only in tubers occupied ; by ants and, when opened, such cavities exhale a fresh mushroom odor. i It is evident that the tips of the hyphe are cut off by the ants and in _ some places whole sods of these filaments are trimmed evenly, yet Miehe believes that the insects do not feed on the fungus, but merely cut the ____ hyphve down because their growth would interfere with the ants’ move- _ ments in the galleries. He thinks, however, that the mycelium grows ___ on the excrement voided by the ants on the papillose portions of the walls only. The papille are evidently not rudimentary roots or root- _ buds, but Miche calls them haustoria or suckers, since he learned from experiment that parts covered with them readily absorb water, while the ___ smooth portions do not. In wild and in cultivated specimens he often _ found rain-water accumulated in some of the cavities. He notes that the _ Iridomyrmex of Myrmecodia is seldom seen outside the galleries, unless the plant be disturbed, on which occasions the ants rush out at once. Their food was not ascertained, nor whether they come out at night. They seemed to him provided with very feeble weapons. Concerning Iridomyrmex myrmecodiz in the Solomon Islands, Wm. M. Mann (1909, p. 362) gives the following account: 3 This is one of the most abundant ants in the Solomons. It nests sometimes beneath bark or in crevices on standing trees, but usually in bulbs of an epiphyte, Myrmecodia species (?M.Guppyanum), which grows on the branches of several species of trees and is especially common on a lowland-inhabiting species of Barringtonia. It has been shown that Myrmecodia can thrive without the presence of ants, but I am sure that few of this species do, for among the many that I cut open, none were without them. Even very young bulbs, less than an inch in diameter, contained in- cipient colonies. Inamore recent paper W. M. Mann (1921, p. 406) mentions the com- mon occurrence of various species of Myrmecodia and Hydnophytum in the _ Fiji Islands. Their bulbs are often inhabited by colonies of ants, Irido- myrmex sororis Mann, I. nagasau Mann, and its subspecies alticola Mann being the more common forms. Pecilomyrma senirewer Mann subspecies myrmecodiz Mann and certain Camponotus and Pheidole also occasionally use such bulbs as nesting sites, though Mann remarks that many bulbs “‘contained no ants at all, but myriopods, spiders, scorpions, or geckos and their eggs.” Wheeler (1919, p. 111) records Camponotus quadrisectus (Smith) “from the distorted pseudobulb of a Myrmecodia” in Borneo and 526 Bulletin American Museum of Natural History (Vol. XLV Crematogaster difformis F. Smith subspecies sewardi (Forel) was also described from a Bornean Myrmecodia. M. armata de Candolle. Java. As Rettig remarks (1904, p. 12, footnote), this is evidently the plant so carefully investigated in Java by Treub, and originally called by him (1883) ‘‘ Myrmecodia echinata Gaud- ichaud.”’ Later (1888), Treub agreed with Beccari that his former identification was incorrect but claimed, apparently with reason, that his plant was not M. tuberosa Jack. It is the species used by Rettig (1904) for some of his experiments and the one studied by H. Miehe (1911) under the name “ M. tuberosa Beccari.’’? Miehe found most of his specimens inhabited by Iridomyrmex myrmecodie (Emery); in one locality, however, exclusively by Camponotus maculatus subspecies pallidus (Smith). Beceari (1884) also mentions the occurrence of Irido- myrmex myrmecodiz in the tubers of Javanese “M. tuberosa” (=M. echinata de Candolle). M. tuberosa Jack. Sumatra, Borneo, and probably elsewhere in the Malay Archipelago. Beccari (1884) found in Bornean specimens Cre- matogaster difformis Smith and Shelford (1916), alsoin Borneo, C. difformis and Iridomyrmex myrmecodiz (Emery). Shelford mentions that both ants are by no means restricted to the tubers of epiphytic Rubiacew, for they frequently nest in hollowed-out branches of various dead or living shrubs or trees. M. bullosa Beccari. New Guinea, Amboina (G. Karsten, 1895). Inhabited by Iridomyrmex cordata (Smith) in New Guinea (Beccari, 1884). M. Menadensis Beccari. Celebes. 8S. H. Koorders' gives the fol- lowing interesting remark concerning this plant: “ Especially common in the Minahasa in the lower plain to 1000 m. above sea-level in young forests, preferably in abandoned coffee-orchards. One sees there on most of the half-dead dadap trees (Erythrina) a number of these strange epi- phytes. It is remarkable that as a rule I have found, on the same trees, one or more specimens of the following other curious myrmecophilous epiphytes with tuberous stem divided into chambers, viz., Hydnophytum formicarum Jack, H. Selebicum Beccari, Polypodium sarcopus DeVr. and Teysm. and Polypodium carnosum Christ, and of the most peculiar Conchophyllum maximum Karsten.” Thus there seem to be regular “associations”’ of myrmecophytic epiphytes, in the sense plant ecolo- gists use this term. = 11898, ‘ Verslag eener botanische dienstreis door de Minahasa.’ (Batavia), p. 497. Wheeler, Ants of the Belgian Congo 527 )M-Rumphii Beccari. Amboina:~ Tubers inhabited by Pheidole Gidvetals (Fabricius) (Beccari, 1884). Merrill (1917, p. 489) posi- . tively identifies with this species Rumphius.’ “ Nidus germinans formi- = corem ruber.” _M. alata Beccari. New Guinea. One of the tubers contained Irido- % i myrmex scrutator Smith, Pheidole megacephala variety and Crematogaster species (Beccari, 1884). ~~ -M. Antonii Beccari. Professor Wheeler has contributed the fol- = lowing note with regard to this Australian species: “While I was at _ Kuranda, in northern Queensland, during the winter of 1914-1915, Mr. _ F. P. Dodd collected for me in the vicinity of the village a number of specimens of Myrmecodia Antonii, all of which were inhabited by colonies of Iridomyrmex myrmecodizx variety stewarti Forel, originally described from Torres Straits. The colonies were not populous and, as the ant is small and timid, I fail to see how it can protect the plant. This q ant sometimes nests about the roots and leaves of other epiphytes. At Cairns, near Kuranda, I found a colony nésting under the leaves ot a __ Dischidia that were applied to the branch of a tree. In northern Queens- land both Myrmecodia and Hydnophytum are called the ‘ant-house’ by the colo by ? _ M. Goramensis Beccari. Moluccas. Tubers settled by Iridomyrmex cordata Smith (Beccari, 1884). M. erinacea Beccari. New Guinea. Crematogaster species was found in the tubers (Beccari, 1884). M. Albertisii Beccari. New Guinea. Tubers inhabited by Irido- myrmex cordata Smith (Beccari, 1884). M. Dahlii K. Schumann. Bismarck Archipelago. Dahl (1901) found the galleries of the tubers inhabited by Iridomyrmex myrmecodiz subspecies decipiens Emery and a subspecies of Camponotus maculatus (Fabricius); both ants were also found nesting in other locations. M. pentasperma K. Schumann (erroneously quoted as M. penta- gona by Forel). Bismarck Archipelago. The tubers were inhabited by Tridomyrmezx cordata Smith and I. myrmecodiz (Emery) (Dahl, 1901). Myrmedoma Beccari. New Guinea. Contains two species: M. arfakiana Beccari and M. Naumanni Warburg. Myrmephytum Beccari. Malay Archipelago. With two species, M. Selebicum Beccari, Celebes, and M. Beccarii Elmer, Sibuyan (Philip- pines). Hydnophytum Jack. Oriental Region. Includes thirty-five species, with swollen excavated rhizomes as in Myrmecodia. Baie A ie! 528 Bulletin American Museum of Natural History _ [Vol. XLV H. montanum Blume (=H. formicarum Beccari). Malay Archi- pelago, northern Queensland. Miehe (1911b) found the tubers in- habited by Iridomyrmex myrmecodiz (Emery) in Java; the walls of the galleries are in places covered with a fungus-growth similar to that of the Myrmecodizx. Beccari (1884) mentions having found in the tubers Irido- myrmex myrmecodixz (in Java) and Crematogaster difformis (in Borneo). This species was also studied by Treub (1883) and others. H. petiolatum Beccari. New Guinea. Tubers inhabited by Irido- myrmex cordata Smith (Beccari, 1884). H. amboinense Beccari. Amboina. Merrill (1917, p. 488) positively identifies with this Rumphius’ “ Nidus germinans formicarum niger.” Squamellaria Beccari. Fiji Islands. With two species. This is related to the foregoing four genera and may possibly be myrmecophytic in a similar way; it is not known, however, whether it has tubers. Nauclea Linneus. Tropical Asia, Malay Archipelago, islands of the Pacific. Contains forty species. N. lanceolata Blume. Java. The branches present swellings in- habited by ants (K. Schumann, 1891), p. 57, fig. 22B). N. formicaria Elmer. Mindanao, Philippines. “Nearly all the Salas of the tree were teretely swollen, 3 to 7 cm. long and 1 em. thick. These eylindrical portions were punctured and inhabited by small black ants” (Elmer, 1911, p. 990). I have examined a specimen collected by Elmer (cotype) in the herbarium of the Arnold Arboretum of Harvard Univer- sity: the cylindric swelling is situated above the middle of one of the internodes and begins and ends abruptly, being very regular and slightly flattened on the two sides corresponding to the lower leaf-pair. On the same branch the two internodes above the swelling are perfectly nor- mal. Professor Wheeler recognized in the remains of ants found in- side the domatia a species of Crematogaster of the subgenus Decacrema. N. strigosa Korthals. Borneo and Luzon, Philippines. G. D. Havi- land (1887, p. 53, Pl. I), who has examined a number of herbarium specimens, writes that in most of them “some of the branchlets have hollow swellings which have been inhabited by ants.” He adds: “I suspect that this plant is the Sarcophelus macrocephalus of K. Schumann, of which I have not, however, been able to find any description.” S. macrocephalus was briefly characterized by K. Schumann in 1890; its branches present swellings inhabited by ants; it was found on Samar Island near Luzon, Philippines. Nauclea strigosa has been made the type of a distinct genus, Myrme- conauclea, by Merrill who says (1920, p. 376) that “‘a certain percentage Wheeler, Ants of the Belgian Congo 529 a ____ of the branchlets always present hollow swellings, perforated on one side, . which are inhabited by colonies’ of small ants.”’ — _N.« celebica G. D. Haviland. Celebes. “The branchlets present g icons hollow swellings which have been inhabited by ants” (Havi- s land, 1897, p. 54). ____N. cyrtopoda Miquel. Borneo, Sumatra, Java. The branches often Bs have hollow swellings occupied by ants (Haviland, 1897, p. 57). ___——s«Unearia Schreber (=Ourouparia Aublet). Tropical regions, mostly * in the Old World. Contains thirty-four species. «UU . africana G. Don. Tropical Africa, Madagascar, Comoros. The . “ myrmecodomatia have been described above (p. 458). __——-—« Sarcocephalus Afzelius. Tropical regions of the Old World. There | are thirteen species. An unidentified species of Sarcocephalus from the Belgian Congo is myrmecophilous (see above, p. 460). Duroia Linnzus fil. (=Schachtia Karsten; Amajoua Poeppig and Endlicher). Tropical South America. Includes ten species. D. hirsuta (Poeppig and Endlicher). _D. petiolaris J. D. Hooker. These two species have branches with spindle-shaped, hollow swel- lings which are inhabited by ants, species of Azteca, Myrmelachista schu- manni Emery, and Allomerus septemarticulatus Mayr (K. Schumann, 1888; Emery, 1891). D. saccifera (Martius) (= Amajoua saccifera Martius). Brazil. Two contiguous pouches, at the base of the leaf-blade, are often settled by ants (K. Schumann, 1888 and 1891b, p. 12, fig. 5B; Spruce, 1908). _ D. dioica (Karsten) (=Schachtia dioica Karsten). Colombia. The original description says: ‘‘ramulis . . . ad apicem internodii inferioris elongati innovationum tumidis.’"' According to Huth (1887), Karsten did not find ants in these swellings. Remijia de Candolle. South America. With fourteen species, only one of which is myrmecophytic. ’ _ R. physophora Bentham has two pouches, inhabited by ants, at the base of the leaf-blade (K. Schumann, 1890; Spruce, 1908). Randia Houston. Tropics of both hemispheres, especially in the Old World. Contains 150 species. R. Lujz B. De Wildeman. R. myrmecophyla £. De Wildeman. ‘Karsten, 1850, Linnwa, XXX, p. 157. i “Ee ow LT ’ Ree = ve 530 Bulletin American Museum of Natural History (Vol. XLV Both from the Belgian Congo; internodes swollen into spindle- shaped myrmecodomatia (see above, p. 465). R. physophylla K. Schumann. Cameroon, Belgian Congo. Rudi- mentary pouches with a nectary, attracting ants, at the base of the leaf-blade (see above, p. 467). Plectronia Linneus. Tropical and subtropical parts of the Old World. Includes 200 species. P. glabriflora (Hiern) of Tropical West Africa, P. Leureaall E. De Wildeman of the Belgian Congo, and some other species of Tropical Africa have swellings of the stems in which ants often nest (see above, p. 471). Cuviera de Candolle. Tropical Africa. There are fourteen species, a number of which have swellings of the internodes inhabited by ants of the genus Crematogaster (see above, p. 488). Such myrmeco- domatia are known with certainty for the following species: C. longiflora Hiern. Cameroon. C.latior Wernham. Belgian Congo. C. Ledermannii Krause. Cameroon. C. angolensis Hiern. Angola, Belgian Congo. C. physinodes K. Schumann. Gaboon. Psychotria Linneus. A very large genus with over 400 described species and distributed throughout the tropics of the Old and New World. P. myrmecophila Lauterbach and Schumann. New Guinea. A bush with pouch-like stipules; the margins are reflexed and the stipule itself much inflated; the cavity thus formed is divided into two by a median projecting carina; apertures are pierced through the wall and also through the inner partition. Remains of ants, together with coccids, have been found in these stipular pouches (K. Schumann and K. Lauterbach, 1901, p. 579, Pl. xxm). ; 4. BIBuiocRAPHY oF THE RELATIONS BETWEEN PLANTS AND ANTS Anonymous. 1821. ‘Die Ameise als Forstfrevlerin auf der That betroffen.’ Nie- mann’s Viterlandische Waldberichte, II, pp. 196-199. 1875. ‘Fungi from interior of a white ant-hill.’ Grevillea, ITI, No. 28, pp. 165-166. 1882a. ‘Fungi of ants’ nests. Gardeners’ Chronicle, N.8., XVII, p. 401. 1882b. ‘Riickschritte in der Blumentiichtigkeit durch Verlust der Fligel und durch Zersplitterung der Nahrungserwerbs-Thatigkeit auf ver- schiedenartige Bezugsquellen. (Ameisen).’ Ent. Nachr., VIII, pp. 233-237. Wheeler, Ants of the Belgian Congo 531 1888a. ‘Ameisen als Beschiitzer von Waldbiitumen.’ (sterreich. Forst- _~——-geitg., VI, p. 231. 1888). ‘Ants and plants.’ Gardeners’ Chronicle, (3) III, p. 279. __-1888e. ‘Leaf-stripping ants in Arizona.’ Insect Life, I, pp. 191-192. 1889. ‘Ants destroying young maples in Nebraska.’ Insect Life, I, p. 346, 1899. ‘Tropische Ameisen als Pilzziichter.’. Natur, XLVIII, pp. 135-137. 1901. ‘Mittel gegen Ameisen.’ Zeitschr. f. Pflanzenkrankh., XI, pp. 54-55. -_-—-:1907. (On fungi formed in ants’ nests). Gardeners’ Chronicle, (2) ) XVIII, p. 401. 1909. (Dolichoderus bituberculatus Mayr, the cacao-ant of Java, and its _ relations to Dactylopius crotonis Green.) Verslag omtrent den staat van het algemeen proefstation te Salatiga over het jaar 1908, pp. 109- 201. 1911. ‘Means of transporting fungi by the Ambrosia-beetles.’ Ent. News, Rene’ XXII, p. 470. «1912. ‘Pilzziichtende Ameisen.’ Fauna Exotica, , pp. 66-67. * - 1914. ‘Die Krankheitsiibertragung durch Ameisen.’ Entom. Zeitschr., sf Frankfurt a. M., XXVIII, p. 86. - 1918. ‘Ein Merkblatt fir die Ameisenbekampfung.’ Mitt. Gartenbauges. _ Steiermark, XLIV, pp. 66-70. 1920a. ‘La fourmi d’Argentine en France.’ Bull. Soc. Vulgar. Zool. Agric., Bordeaux, XTX, pp. 91-93. — —«44920. ‘La formica argentina.’ Cire. 1 Lab. Ent. Agrar. Portici, pp. 1-6. q aecunee, J. 1912. ‘Ueber Ameisenpflanzen.’ Jahresber. Preuss. Bot. Ver., eee (1911), pp. 53-55. ot 1913. ‘Ueber Ameisenpflanzen.’ Schrift. Phys-Okon. Ges. Kénigsberg, = 3 LIU, pp. 319-321. - ahdos, A. 1908. ‘Les fourmis sont-elles nuisibles?’ Cosmos, Paris, N.S., LVILI, a pp. 651-653. a 1913. ‘Les fourmis A miel des déserts mexicains.’ La Nature, LX XIX, pp. | a 161-163. Apa, A. W. 1914. ‘Some observations on the domestic ant: Monomorium = species.’ Bull. Soc. Ent. Egypte, IV, pp. 56-62. _ Apterz, G. 1886. ‘Myrmecologiska studier. II. Svenska myror och deras lef- oS nadférhAllanden.’ Bihang Svenska Vet. Ak. Handl., XI, No. 18, pp. a 1-329, Pls. vu. = _ 1896. ‘Myrmecologiska notiser.’ Ent. Tidskr., XVII, pp. 129-141. ; 1913. ‘Myrornas liv.’ Stockholm. 243 pp. 1914. ‘Formica fusca-picea Nylander, en torfmossarnas myra.’ Arkiv. f. Zool., VIII, No. 26, 5 pp. Arrxen, E. H. 1890. ‘Red ants’ nests.’ Journ. Bombay Nat. Hist. Soc., V, p. 422. _ Atexanper, D. A. 1895. ‘The sauba-ant in Brazil.’ Ent. News, VI, pp. 19-20. _ Avren, H. von. 1913. ‘Eine neue Ambrosiagalle an Chxrophyllum temulum 1..’ ee, 17. Jahresber. Ver. f. Naturw. Braunschweig, (1909-1912), pp. 57-62. 4 Racuany, P. 1902. ‘Uebertrigt die Ameise den Hausschwamm?’ Zoolog. Garten, sl ’ XLII, p. 202. > ANDERSSON, J. 1901. a skadedjur i sccttegeeuna int. Tidskr., XXII, 532 Bulletin American Museum of Natural History (Vol. XLV Anpré, Ep. 1882. ‘Relations des fourmis avec les pucerons et les gall-insectes.’ Bull. d’Insect. Agric., VII, p. 37. Anpr&, Ern. 1885. ‘Les fourmis.’ Paris. 345 pp. 1899. ‘Les fourmis champignonnistes.’ Bull. Soc. Grayloise 4’ Emul., II, pp. 271-280. ANNIBALE, R. 1907a. ‘Sopra due Bignoniacee mirmecofile africane.’ Bull. Orto Bot. Palermo, VI, pp. 83-85. 1907b. ‘Contributo allo studio delle Bignoniacee mirmecofile e acarofile.’ Boll. Soc. Nat. Napoli, XXI, pp. 60-70. AnTOINE, F, 1882. ‘Myrmecodia echinata Gaud. Eine Ameisenpflanze von den Molukken.’ Csterreich. Bot. Zeitschr., XXXII, pp. 347-353, Pl. Armit, W.E. 1878. ‘Agricultural ants.’ Nature, XTX, p. 643. Arnotp, G. 1913. {The food of ants.’ Proc. Rhodesia Scientif. Assoc., XII, pp. 11-24. 1915-1920. ‘A monograph of the Formicide of South Africa.’ Ann. South African Mus., XIV, pt. 1, 1915, pp. 1-158, Pl. 1; pt. 2, 1916, pp. 159-270, Pls. n-1v; pt. 3, 1917, pp. 271-402; pt. 4, 1920, pp. 403- 578, Pls. v-1x. Ascuerson, P. 1878. (Galls on Salix conifera Wang. and Acacia fistula Schweinf.) Verh. Bot. Ver. Brandenburg, XX, pp. 44-45. Assmutn, J. 1907. ‘Einige Notizen itiber Prenolepis longicornis Latreille.’ Zeitschr. Wiss. Insektenbiol., III, pp. 301-309, 328-334, and 357-368. Avs.et, J. B. 1775. ‘Histoire des plantes de la Guiane frangaise.’ Paris. Vols. I- IV; xxxii+976+52+160 pp., 392 Pls. Autmann, G. 1912. ‘Die Schidlinge der Baumwolle.’ Fauna d. Deutsch. Kolon., Reihe 5, Heft 4, pp. 1-166. AuLMANN, G. anp La Baume, W. 1911. ‘Die Schiidlinge des Kaffees.’ Fauna d. Deutsch. Kolon., Reihe 5, Heft 2, pp. 1-98. 1912. ‘Die Schidlinge des Kakaos.’ Fauna d. Deutsch. Kolon., Reihe 5, Heft 3, pp. 1-86. Axmann, A. 1895. ‘Vorbeugungsmittel gegen die Beschidigungen durch Lasius flavus Latreille.’ Zentralbl. f. d. Ges. Forstwesen, XXI, pp. 249-252. Baccarint, P. 1893. ‘Sopra un curioso cecidio della Capparis spinosa.” VII, pp. 405-414, PI. vir. 1909. ‘Sui micozoocecidii od “ Ambrosiagallen.’’’ Bull. Soc. Bot. Italiana, pp. 137-145. Bacu, M. 1859. ‘Ueber gewisse Veriinderungen und Auswiichse an verschiedenen Pflanzentheilen, welche durch den Einfluss der Insekten bewirkt werden. Die Ameisen.’ Natur u. Offenbarung, V, pp. 363-372, 385- 394. ; Barer, G. A. 1903. ‘Note sur un Membracide myrmécophile de la République Argentine.’ Bull. Soc. Ent. France, pp. 306-308. Bartey, I. W. 1920. ‘Some relations between ants and fungi.’ Ecology, I, pp. 174-189, Pls. v—vn. Baker, FE. G., Moore, 8., anp Renpie, A. B. 1905. ‘The botany of the Anglo- German Uganda boundary commission.’ Journ. Linn. Soc. London, Bot., XXXVII, pp. 116-227, Pls. av. Ps __Wheeler, Ants of the Belgian Congo 533 = x, E.D. 1915. ‘Adaptations to arid conditions in Cereopide and Membracide.’ Ann. Ent. Soe. America, VIII, pp. 365-368. , H. A. 1920. ‘An observation on ants.’ Agric. News, Barbados, XIX, pp. 10-11. , C.S. 1912. ‘A manual of Philippine silk culture.’ Manila. 53 pp., 20 Pls. :, E.R. 1916. ‘The Argentine ant. Distribution and control in the United -‘* States.’ Bull. U.S. Dept. Agric. No. 377, 23 pp. 1920. ‘The Argentine ant as a household pest.’ Farmers’ Bull. 1101 U. 8. a. Dept. Agric., pp. 1-11. Rike, M. A. 1914. ‘Cockroaches and ants as the carriers of the vibrios of Asiatic i ____ ¢holera.’ Philippine Journ. Sc., B, TX, pp. 1-4. weer, G. 1905. ‘Il micozoocecidio dei Verbascum.’ Nuovo Giorn. Bot. Etalinno, N.S., XII, pp. 709-722. Sentaion, Il; pp. 204-206. i, H. W. 1863. ‘The naturalist on the River Amazon.’ London. Vol. I, att 351 pp.; vol. II, 423 pp. (2d ed., 1892). Bavpouin, M. 1898. ‘L’emploi des fourmis en médecine opératoire.’ Rev. Scient. it ae Bourbonnais, XI, pp. 252-253. ” ms, W. J. 1898. ‘Seed dispersal.’ Boston. 87 pp. — J. 1910a. ‘Les champignons dits Ambrosia.’ Ann. Sci. Nat., Botani- que, (9) XI, pp. 31-73, 5 Pls. 1910b. ‘L’Ambrosia du Tomicus dispar.’ C. R. Ac. Sci. Paris, CL, pp. 1071- 1 eal 1074. Fe ve O. 1874. ‘Descrizione di una nuova spece del genere Myrmecodia della ee, _ famiglia delle Rubiacee.’ Nuovo Giorn. Bot. Italiano, VI, pp. 195-197, ee Pi. vt. X va ae 1884-1886. ‘Piante ospitatrici ossia piante formicarie della Malesia e della a Papuasia.’ Malesia, II, 340 pp., 65 Pls. _—s«43885. ‘Plantes A fourmis de |l’Archipel Indo-malais et de la Nouvelle- Eee Guinée (review by E. Levier).’ Arch. Italiennes Biol., VI, 41 pp. 1902. ‘Nelle foreste di Borneo.’ Florence. xvi+667 pp. aE 1904. ‘Wanderings in the great forests of Borneo.’ London. xxiv +424 pp. acme, T. 1873. ‘Beitrag zur Naturgeschichte des Bostrychus lineatus und des s _ Bostrychus domesticus.’ Tharandter Jahrb., XXXIIT, pp. 17-44. : BELLEvore, A. 1888. ‘Observations sur le Monomorium pharaonis Latreille at (domestica Schenk).’ Ann. Soc. Ent. France, (6) VIIT, Bull., pp 1890. ‘Observations on Monomorium pharaonis Latreille.’ Insect Life, I, pp. 230-233. a Ram Tu. 1874. ‘The naturalist in Nicaragua.’ London. xvi+403 pp. (2d ed., ; 1888). _ Beqvarrt, J. 1913. ‘Notes biologiques sur quelques fourmis et termites du Congo belge.’ Rev. Zool. Afr., II, pp. 396-431. 1920. ‘A new host of Laboulbenia formicarum Thaxter, with remarks on the : fungous parasites of ants.’ Bull. Brooklyn Ent. Soc., XV, pp. 71-79. ‘Bercevin, E pe. 1910. ‘Sur une larve de Pediopsis nana H. 8. nourricidre de fourmis.’ Bull. Soc. Hist. Nat. Afr. Nord, I, pp. 95-97. 534 Bulletin American Museum of Natural History [Vol. XLV Bercrotn, E. 1903. ‘Neue myrmecophile Hemipteren.’ Wien. Ent. Zeitg., XXTI, pp. 253-256. Berxetey, M. J. 1847. ‘Decades of fungi. Dee. XV-XIX. Ceylon fungi.’ Hooker's London Journ. of Bot., VI, pp. 479-514, Pls. xvir—xx. Berke.ey, M.J. anp Broome, C. E. 1871. ‘The fungi of Ceylon.’ Journ. Linn. Soc. London, Bot., XI, No. 56, pp. 494-567. 1873. ‘Enumeration of the fungi of Ceylon.’ Journ. Linn. Soc. London, Bot., XIV, pp. 29-140, Pls. m—x. Bervesk, A. 1898. ‘Rapporti fra la vite ed i saccaromiceti.’ Riv. Patol. Veg., V, pp. 211-237, 263-282, 295-342, 354-360. Bettany, G.T. 1879. “The galleries of the cutting ant of Texas.’ Nature, XX, p. 583. 1880. ‘The parasol ant of Texas.’ Nature, XXI, pp. 17-18. Brncuam, C. TI’. 1906. ‘A plague of ants in the Observatory District, Cape Town, South Africa.’ Trans. Ent. Soc. London, Proc., pp. xxiii-xxvi. Brré, L. 1897. ‘Biologische Mittheilungen aus Neu-Guinea.’ Berlin. Ent. Zeitschr., XLII, pp. 129-138. Bisxopr?, F. C. anv Jones, C. R. 1907. ‘The cotton bollworm: a summary of its life-history and habits, with some results of investigations in 1905 and 1906.’ Farmers’ Bull. 290 U. 8. Dept. Agr., pp. 1-32. Biscnory, H. 1912. ‘Pilzkultur aus an Ameisen gewachsenen Pilzen.’ Berlin. Ent. Zeitschr., LVL, Sitz. Ber. f. 1911, p. (2). Buake, J. 1873. ‘On the structure of the honey-bag in the honey-making ant Myrmecocystus mexicanus.’ Proc. California Ac. Sci., II, p. 98. Buancuarp, R. 1900. ‘Un enfant attaqué par des fourmis.’ Archives de Parasit- ologie, LIT, pp. 645-646. BLocuMann, F. 1884. ‘Ueber eine Metamorphose der Kerne in den Ovarialeiern und iiber den Beginn der Blastodermbildung bei den Ameisen.’ Verh. Naturh. Med. Ver. Heidelberg, N.8., III, pp. 243-247. 1888. ‘Ueher das regelmiissige Vorkommen von bakterienihnlichen Ge- bilden in den Geweben und Eiern verschiedener Insekten.’ Zeitschr. Biol., XXIV, pp. 1-15, PI. 1. Bocnart, 8. 1663. ‘Hierozoicon, sive bipartitum opus de animalibus sacre serip- ture.’ London. 2 vols. Bénner, W. 1914a. ‘Fréhéstende Myrer i Danmark. A seed-reaping ant, Tetra- morium c#spitum, in Denmark.’ Ent. Medd., Kjébenhavn, X, pp. 77-82. 1914b. ‘Formica fusca picea, eine Moorameise. Mit Schlussbemerkung von E. Wasmann.’ Biol. Centralbl., XXXIV, pp. 59-76 (Wasmann, - pp. 76-80). « 1915. ‘Die Ueberwinterung von Formica picea und andere biologische Beo- bachtungen.’ Biol. Zentralbl., XX XV, pp. 65-77, Pl. 1. Boutvar, I. 1901. ‘Un nuevo Ortéptero mirmec6filo, Attaphila bergi.’ Com. Mus. Nac. Buenos Aires, I, pp. 331-336. Bouie, G. 1907. ‘Die Bekimpfung der Ameisen- und Heuschreckenplage in Siid- amerika.’ Tropenpflanzer, XI, p. 392. Bonnter, G. 1879. ‘Les nectaires. Etude critique, anatomique et physiologique.’ Ann. Sc. Nat., Botanique, ‘6) VIII, pp. 5-212, Pls. 1-vut. Bos, H. 1888. ‘Mieren en bladluizen.’ Tijdschr. vy. Ent., XX-XI, pp. 235-244. __ Wheeler, Ants of the Belgian Congo 535 ‘Rovere, E. L. 1909a. ‘Sur les phénoménes qui caractérisent le déménagement ‘ chez la fourmi moissonneuse.’ C. R. Ac. Sci. Paris, CXLIX, pp. 649- F 654. 1909). ‘Les phénoménes qui caractérisent le déménagement chez la fourmi $4 moissonneuse.’ Cosmos, Paris, N. 8., LXJ, pp. 606-608. — 191la. ‘L’agriculture chez les fourmis.’ Rev. Progrés, No. 1. 1911b. ‘Sur les fourmis moissonneuses (Messor barbara) des environs de _ Royan.’ ler Congr. Intern. Entom. Bruxelles (1910), II, Mém., pp. 237-248. 1918. i vie psychique des Insectes.’ Paris (Bibl. Philos. Scientif.). Bown, F oO. 1886. ‘On Humboldtia laurifolia as a myrmecophilous plant.’ Nature, A XXXIV, No. 883, p. 538. 1887. ‘On Humboldtia laurifolia as a myrmecophilous plant.’ Proc. Philos. 5 | Soc. Glasgow, XVIII, pp. 320-326, Pl. x. ‘Brawc, H. E. 1913. ‘Morphology and biology of the Membracidie of Kansas.’ a Kansas Univ. Bull., VIII, No. 3, pp. 1-84. ee BRANES, G. 1893. ‘Die Blattliuse und der Honigthau.’ Zeitschr. Ges. Natur- >. wiss., LX VI, pp. 98-103. == ‘Buannrcover, V. 1902. ‘Pucerons et fourmis.’ La Nature, XXX, 2, p. 202. 4 a. J.G. 1896. ‘Decomposition of rocks in Brazil.’ Bull. Geol. Soc. America, “Des a VI, pp. 255-314, Pls. x—xtv. — ———s«*1900. ‘Ants as geologic agents in the tropics.’ Journ. Geology, Chicago, en VIII, pp. 151-153. 1910. ‘Geologie work of ants in tropical America.’ Bull. Geol. Soc. America, XXI, (1909), pp. 449-496, Pl. xxxv. 1912. ‘Geologie work of ants in tropical America.’ Ann. Rept. Smiths. ene Inst. for 1911, pp. 303-333, PI. 1. z ‘Seisteckure, A. G. H. 1919. ‘Ants attacking bees.’ Journ. Bombay Nat. Hist. oe Soc., XXVI, 3, p. 878. a Brenz, C. 1886. ‘Notes on the @eodomas, or leaf-cutting ants of Trinidad.’ Ameri- a can Naturalist, XX, 2, pp. 123-131. Britten, J. 1875. ‘Ant supporting plants.’ Popular Science Rev., XIV, p. 29. i eee, C. 1916. ‘Contribucién al estudio de las hormigas de la provincia de San = Luis.’ Rev. Mus. La Plata, XXIII, pt. 2, pp. 291-357, Pls. 1—x11. 1917. ‘Costumbres y nidos de hormigas.’ An. Soc. Cientif. Argentina, LXXXIJ, pp. 302-316; LXXXIV, pp. 154-168; 3 Pls. 1919. ‘Nidos y costumbres de hormigas.’ Physis, Buenos Aires, IV, pp. 539-541. __- Bron, R. 1913. ‘Zur Biologie von Formica rufa und Camponotus herculeanus i. sp.’ 4 Zeitschr. Wiss. Insektenbiol., IX, pp. 15-19. Bryan, G.H. 1879a. ‘Harvesting ants.’ Science Gossip, p. 238. c 1879b. ‘Harvesting ants.’ Nature, LX, p. 174. Bocuner, P. 1912a. ‘Ueber intrazellulare Symbionten bei zuckersaugenden Insekten und ihre Vererbung.’ Sitzb. Ges. Morphol. Physiol. Minchen XXVII, q (1911), pp. 89-96. a 1912b. ‘Studien an intracellularen Symbionten. 1. Die intracellularen Symbionten der Hemipteren.’ Archiv f. Protistenkunde, XXVI, pp. 1-116, Pls. x11. ) Bulletin American Museum of Natural History (Vol. XLV 1913. ‘Neue Erfahrungen tiber intrazelluliire Symbionten bei Insekten.’ Naturw. Wochensckr., XXVIII, pp. 401-406 and 420-425. 1918a. ‘Vergleichende Eistudien. 1. Die akzessorischen Kerne des Hymenoptereneies.’ Arch. Mikroskop. Anat., XCI, Abt. 2, pp. 1-202, Pls. t-x. 1918). ‘Studien an intracellularen Symbionten. 2. Die Symbionten von Aleurodes, ihre Uebertragung in das Ei und ihr Verhalten bei der Embry- onalentwicklung.’ Archiv f. Protistenkunde, XXXIX, pp. 34-61, Pls. tv—-v. 1919. ‘Zur Kenntnis der Symbiose niederer pflanzlicher Organismen und Pedikuliden.’ Biol. Centra!bl., XXX XTX, pp. 535-540. 1920. ‘Neue Beobachtungen an intrazellularen Symbionten.’ Sitzb. Ges. Morphol. Physiol. Miinchen, XX XT, (1919), pp. 54-65. 192la. ‘Studien an intracellularen Symbionten. 3. Die Symbiose der Anobiinen mit Hefepilzen.’ Archiv f. Protistenkunde, XLII, pp. 319-336. 1921b. ‘Tier und Pflanze in intracellulare Symbiose.’ Berlin. Buck xy, 8. B. 1860. ‘The cutting ant of Texas.’ Proc. Acad. Nat. Sci. Philadelphia, XII, pp. 233-236; Ann. Mag. Nat. Hist., (3) VI, pp. 386-389. 186la. ‘Myrmica (Atta) molefaciens, “stinging-ant,” or “mound-making ant,’”’ of Texas.’ Proc. Acad. Nat. Sci. Philadelphia, XII, (1860), pp. 445-447. 1861b. ‘Note on ants in Texas.’ Proc. Aéad. Nat. Sci. Philadelphia, XIII, pp. 9-10. Bucxton, G. B. 1876-1883. ‘Monograph of the British Aphidw.’ London. Vol. I, 1876, iii+193 pp., Pls. A-C and 1~-xxxvu1r; vol. II, 1879, 176 pp., Pls. XXXIX-Lxxxvt; vol. III, 1881, ii+142 pp., Pls. uxxxvi—cxtv; vol. IV, 1883, ix+228 pp., Pls. D-J and cxv—cxxxtv. 1901-1903. ‘A monograph of the Membracidx.’ London. xii+296 pp., 60 Pls. (1901, pp. 1-92, Pls. 1-xvii1; 1902, pp. 93-180, Pls. x1x—xxXxvIir; 1903, pp. 181-296, Pls. xxxrx-ix; remarks by Poulton, pp. 273-285). Btscen, M. 1891. ‘Der Honigtau. Biologische Studien an Pflanzen und Pflanzen- liusen.’ Jenaische Zeitschr. Naturwiss., XXV, pp. 339-428, Pls. xv-xv1; Biol. Centralbl., XI, pp. 193-200. Buenton, E. 1909a. ‘La fourmi rouge de Cevylan (@cophylla smaragdina).’ Arch. Sci. Phys. Nat. Soc. Helvétique Sci. Nat., pp. 105-107. 1909b. ‘Fourmi fileuse de Ceylan (cophylla smaragdina).’ Arch. Sei. Phys. Nat. Genéve, XXVIII, pp. 297-298, 511-513, 1 PI. 1910. ‘Observations relatives A l'industrie des termites.’ Ann. Soc. Ent. France, LX XIX, pp. 129-144. 1913. ‘Les mocurs des termites champignonnistes.’ Bibliothéque Univ. Lausanne, pp. 552-583, PI. 1914a. ‘La biologie des termites de Ceylan.’ Bull. Mus. Hist. Nat. Paris, pp. 170-204, 8 Pls. 1914b. ‘Les movurs des termites champignonnistes.’ Bull. Soc. Nat. Acclim. France, LXI, pp. 532-535. Burcx, W. 1884a. ‘Contributions to the fern flora of Borneo.’ Ann. Jard. Bot. Buitenzorg, IV, pp. 88-100, Pl. vir. _ __Wheeler, Ants of the Belgian Congo 7 537 1884). ‘Sur !’organisation florale de quelques Rubiacées.’ Ann. Jard. Bot. Be Buitenzorg, IV, pp. 12-87, Pls. 1v—v1. --—-—s«1891. ‘Beitrige zur Kenntnis der myrmecophilen Pflanzen und der _* _ ‘Bedeutung der extranuptialen Nectarien.’ Ann. Jard. Bot. Buiten- ; zorg, X, pp. 75-144, Pls. vu—x1. _Boscarios, L. anv Huser, J. 1900. ‘Eine neue Theorie der Ameisenpflanzen.’ im ___ Beih. z. Bot. Centralbl., IX, 2, pp. 85-88. Bosse, W. 1901. ‘Die Ausscheidung von Gummi arabicum an ostafrikanischen S __ Akazien.’ Naturw. Wochenschr., N. F., I, pp. 100-101. ; “ Borm-Rexre, H. v. 1907. ’Paychohidlogieche und biologische Beobachtungen _-———s aan, Ameisen, Bienen und Wespen.’ Naturw. Wochenschr., XXII, pp. Rs Se wv . _ 465-478. Ae _ Cane, P. P. 1911. ‘Studies on Costa Rican Odonata. II. The habits of the rank plant-dwelling larva of Mecistogaster modestus.’ Ent. News, XXII, pp. 7 402-410. | Cucem A. 1907. ‘Le alleanze tra le piante e le formiche.’ Emporium, ~s Bergamo, XXV, pp. 392-400. -—- 1909. ‘Le alleanze degli animale e delle piante.’ Turin. 167 pp. | Canon, T. 1872. ‘Illustrazione di una Rubiacea del genere Myrmecodia.’ Nuovo Giorn. Bot. Italiano, IV, pp. 170-176, PI. 1. Castuxs, J. 1790. ‘Observations on the sugar ants.’ Philos. Trans. Roy. Soc., LXXX, pp. 346-358. Cavara, F. 1899. ‘Di una nuova laboulbeniacea, Rickia Wasmannii nov. gen. e nov. spec.’ Malpighia, XIII, pp. 173-188, Pl. vt. Cavans, F. 1910. ‘Ants and seed-dispersal.’ Knowledge, VII, p. 319. ‘ 1911. ‘Ambrosia fungi.’ Knowledge, VIII, p. 148. 1912. ‘Ants and plants.’ Knowledge, IX. p. 150. _ Chirbor, C. 1914. ‘Etude des Laboulbéniacées européennes.’ Archives de Para- sitologie, XVI, pp. 373-403, PI. vr. ih oa C. anp Picarv, F. 1909. ‘Contribution a la biologie et &la systématique des Laboulbéniacées de la flore frangaise.’ Bull. Scientif. France et Belgique, XLII, pp. 247-263, Pls. m—v. Cesany, V. 1855. ‘Notice A servir d’appendix aux mémoires de Messieurs Leveillé, Tulasne et autres, sur la véritable nature des sclerotiums.’ Botan. Zeitung, XIII, pp. 73-80. 1861. ‘Appunti per una futura Crittogamologia Insubrica.’ Commentario della Soc. Crittogamol. Italiana, I, pt. 2, pp. 47-72, Pls. m-—tv. Craters, A. J. 1918. ‘Notes on some minor cutaneous affections seen in the Anglo- ian Sudan.’ Journ. Trop. Med. Hyg., London, XXI, pp. 197-200. 1919. ‘(dema of the eyelids caused by ants.’ Journ. Trop. Med. Hyg., London, XXII, p. 117, 1 Pl. Cuamrton, H.G. 1913. ‘Insects caught by an orchid.’ Ent. Monthly Mag., XLIX, pp. 254-255. Cuarman, T. A. 1904. ‘Notes on Xyleborus dispar Fabricius.’ Trans. Ent. Soc. London, pp. 100-102. _ CHarraty, N. 1886. ‘La fourmi Sauva.’ Revue Scientif., (3) XX XVIII, pp. 371- a 538 ’ Bulletin American Museum of Natural History {Vol. XLV Cuevatrer, A. 1908. ‘Le Cacaoyer dans |’Ouest Africain.’ Les Végétaux Utiles de l’Afrique Tropicale Frangaise, IV, 245 pp. Cntovenba, E. 1911. ‘Di due forme probubilmente mirmecofile dell’ Abissinia settentrionale.’ Ann. di Bot., [X, pp. 327-329. Curep, T. F. 1921. ‘A list of the fungi of the Malay Peninsula.’ The Gardens’ Bull. Straits Settlements, II, pp. 311-318. Cnropat, R. anp Viscuer, W. 1920, ‘La végétation du Paraguay. Résultats scientifiques d’une mission botanique suisse au Paraguay. IX. Urticiflores.’ Bull. Soc. Bot. Genéve, (2) XI, (1919), pp. 226-258. Cxopat, R. anp Carisso, L. 1920. ‘Une nouvelle théorie de la myrmécophilie.’ C. R. Séances Soc. Phys. Hist. Nat. Genéve, XX XVII, No. 1, pp. 9-12. Cuoparp, L. 1921. ‘Une fourmi exotique acclimatée dans le Midi de la France.’ Rev. Hist. Nat. Appliquée, lére partie, II, pp. 140-144. Crapton, J. F. 1865. ‘Ants and Cocci.’ Science Gossip, p. 234. CockereELL, T. D. A. 1893. ‘The use of ants to Coccide and Aphide.’ Nature, XLVUH, p 608. 1897. ‘Coccide associated with ants.’ Science Gossip, N.8., IIL, pp. 239- 241. 1902. ‘A new plant louse from Southern Califoraia (with description of the attending ant by Prof. W. M. Wheeler).’ Bull. South. California Ac. Sci., I, pp. 40-41. 1903. ‘Five new Coccidew from Mexico.’ Entomologist, XXXVI, pp. 45- 48. 1919. ‘Capture of ents by gummy exsudations.’ Ent. News, XXX, p. 115. CockeEreELL, T. D. A. anp Kine, G. B. 1897. ‘New Coccide found associated with ants.’ Canadian Ent., X XIX, pp. 90-93. CockERELL, W. P. 1903. ‘Some aphids associated with ants.’ Psyche, X, pp. 216— 218. Comstock, J. H. 1887. ‘Relations of ants and aphids.’ American Naturalist, XXT, p. 382. Conner, 8. D. 1919. ‘Aphids and ants on fruit trees.’ Proc. Indiana Ac. Sei., (1918), pp. 245-246. Coox, O. F. 1892.* ‘Do termites cultivate fungi? ’ Botan. Gazette, XVII, p. 282. 1904a. ‘An enemy of the cotton boll weevil.’ Science, N.8., XIX, pp. 862- 864. 1904b. ‘An enemy of the cotton boll weevil. U.S. Dept. Agric., Report No. 78, pp. 1-7. 1904c. ‘Pupation of the kelep ant.’ Science, N.S., XX, pp. 310-312. 1904d. ‘Professor William Morton Wheeler on the kelep.’ Science, N. 8, XX, pp. 611-612. 1904e. ‘Evolution of weevil-resistance in cotton.’ Science, N.S., XX, pp. 666-670. 1904f. ‘Report on the habits of the kelep, or Guatemalan cotton-boll- weevil ant.’ Bull. 49 U.S. Dept. Agric. Bur. Ent., pp. 1-15. 1905a. ‘Progress in the study of thekelep.’ Science, N. 8., XXI, pp. 552- 554. 1905b. ‘The social organization and breeding habits of the cotton-protect- ing kelep of Guatemala.’ U.S. Dept. Agric. Bur. Ent. Techn. Ser. No. 10, pp. 1-55. | Wheeler, Ants of the Belgian Congo 539 dl te 1906. ‘Please excuse the kelep.’ Science, N. 8.. XXIII, pp. 187-189. Pgletae, M. C. 1889. ‘New Australian fungi.’ Grevillea, XVIII, No. 85, Pp. 1-8. _———s«1891. ‘Exotic fungi.’ Grevillea, XX, No. 93, pp. 15-16. 2 «1892. ‘Vegetable wasps and plant worms. A popular history of entomogen- es ous fungi, or fungi parasitic upon insects.’ London. v+364 pp., 4 Pls. Cooter, R. “a 1903. ‘Ants as fruit pests.’ Bull. 51 Montana Agric. Exper. Sta., Ei Conner, ~ "1916. ‘Actes d’habitude et actes de nouveauté chez la fourmi Messor Eas. - barbarus.’ Bull. Soc. Hist. Nat. Afr. Nord, III, pp. 45-48 and p. 90. Commas, 1909. ‘Der Gartenbau der Ameisen.’ Sitzb. Sachs. Ges. Bot., N. F., ae sds XTI-X TIT (1907-09), pp. 42-66, 2 Pls. a ay Coma Tams, pA. 1916. ‘Consideragdes sobre a campanha contra a formiga Sativa [Atla sexdens (L.) Fabr.}.’ Arch. Mus. Nac. Rio de Janeiro, XIX, pp. Bef 179-192, 1 PI. 3 _ Cosransm, J. 1894. ‘Les fourmis champignonnistes.’ Le Naturaliste, XI, pp. et 207-208. ; Covren, W. 1863. ‘Remarks on tent-building ants.’ Proc. Ent. Soc. Philadelphia, I, pp. 373-374. oe Benois, H. 1898. ‘Les plantes myrmécophiles.’ La Nature, XXVII, pp. 70-74. _——s«44907. ‘La culture des plantes phanérogames par les fourmis.’ Le Natura- “3 1910. ‘Les termites champignonnistes.’ La Nature, XX XVIII, pp. 273- a te 274. «4911. ‘Nouveaux insectes champignonnistes.’ La Nature, XXXIX, 2e Sem., No. 1936, p. 193. Crate, C.F. 1898. ‘The transmission of disease by certain ticks, bedbugs, ants, ete.’ New York Med. Journ., LXVIII, pp. 593-599. Crawiey, W.C. 1916. ‘The alate Sinatles of the ant-aphis Forda.’ Ent. Record, XXVIII, pp. 272-273, Pl. vt. 1917. ‘Ants and aphids in West Somerset.’ Proc. Somerset Nat. Hist. Soc., LXI, pp. 148-163. Craw ey, W. C. anp Donistnorre, H. 1913. ‘The founding of colonies by queen ants.’ Trans. 2nd. Congr. Entom. Oxford (1912), II, pp. 11-77. Darert, F. W. 1896. Relatorio annual do Instituto Agronomico do Estado de S. Paulo (Brazil) em Campines. 8S. Paulo, VII and VIII, 451 pp. Dant, F. 1901. ‘Das Leben der Ameisen im Bismarck Archipel, nach eigenen Beobachtungen vergleichend dargestellt.’ Mitt. Zool. Mus. Berlin, IT, 1, No. 3, pp. 1-70, Pl. 1. Data Torre, K. W. v. 1919. ‘Honig und Wachs ohne Honigbienen.’ Tiroler Bienenzeitung, Innsbruck, p. 3. Damanti, P. 1885. ‘Rapporti tra i nettarii estranuziali della Silene fasciata Lk. e le formiche.’ Giorn. Soc. Accl. Agrar. Sicilia, XXV, 5, p. 101. Danxuer, M. 1907. ‘Spinnende Insekten.’ Natur u. Offenbarung, LITT, pp. 291- 300. Danuta, 8. T. 1913. ‘The part played by flies and other insects in the spread of infectious diseases in the tropics, with special reference to ants and the transmission of T'r. hippicum by Musca domestica.’ ‘Trans. 15th Inter- nat. Congr. Hyg. Demog. Sect. 5, Washington, D.C., IV, pp. 182-185. 540 Bulletin American Museum of Natural History (Vol. XLV Darwin, Fr. 1876. ‘On the glandular bodies on Acacia sphxrocephala and Cecropia peltata serving as food for ants. With an appendix on the nectar- glands of the common brake-fern, Pleris aquilina.’ Journ. Linn. Soe. Lendon, Bot., XV, pp. 398-409, PI. vr. Davis, Wm. T. 1914. ‘The fungus-growing ant on Long Island, New York.’ Journ. New York Ent. Soc., XXIT, pp. 64-65. Dean, G. A. 1905. ‘The mound-building prairie ant (Pogonomyrmex occidentalis).’ Trans. Kansas Ac. Sci., XIX, pp. 164-170, Pl. . De Gasparis, A. 1895. ‘Sopra alcune piante a funzione mirmecofoba.’ Boll. Soc. Natural. Napoli, [X, pp. 25-28. 1899. ‘Contributo alla biologia delle felci.’ Atti Accad. Sci. Napoli, (2) IX, No. 9, pp. 1-13, Pls. 1-1. De Groot, J.J.M. i899. ‘Vruchtboomen kunstmatig onder bescherming van mieren buiten onzen Archipel.’ Teysmannia, IX, pp. 535-536. Deacoux, —. 1848. ‘Notice sur les mceurs et les habitudes de quelques espéces de Formiciens des climats chauds.’ Revue Zoolog., XI, pp. 138-150. Dexrino, F. 1872. ‘Sui rapporti delle formiche colle tettigometre e sulla geneal- ogia degli Afidie dei Coecidi.’ Bull. Soc. Ent. Italiana, IV, pp. 343-351. 1874-1875. ‘Ulteriori osservazioni e considerazioni sulla dicogamia nel regno vegetale.’ Atti Soc. Italiana Se. Nat., Milano, XVI, pts. 3-4, 1874, pp. 151-349; XVII, pte. 3-4, 1875, pp. 266-407. 1875a. ‘Altre osservazioni sui rapporti tra cicadelle e formiche.’ Bull. Soe. Ent. Italiana, VII, pp. 61-65. 1875b. ‘On the fondness of ants for certain Homoptera.’ Ent. Monthly Mag., XII, p. 10. 1875c. ‘Rapporti tra Insetti e tra nettarii estranuziali in aleune piante.’ Bull. Soc. Ent. Italiana, VII, pp. 69-90. 1879. *Nettarii estrafiorali.’ Ann. Sci. Ital., Milano, XV, p. 76. 1881. ‘Nuove osservazioni sovra piante entomofile.’ Ann. Sci. Ital., Milano, XVI, pp. 27-39. 1886-1889. ‘Funzione mirmecofila nel regno vegetale; prodromo d’una monografia delle piante formicarie.’ Mem. Accad. Sc. Bologna, (4) VII, 1886, pp. 215-323; VIII, 1887, pp. 601-659; X, 1889, pp. 115-147. 1887-1888. ‘Weitere Bemerkungen tiber myrmecophile Pflanzen.’ Monatl. Mitth. Gesammtgeb. Naturwiss., V, pp. 25-26. 1889. ‘Osservazioni e note botaniche.’ Malpighia, ILI, pp. 337-357, Pl. xm (Galle quercine mirmecofile, pp. 349-352. Acacie africane a spine mirmecodiate, pp. 352-353). 1898. ‘Nuove specie mirmecofile fornite di nettarii estranuziali.’ Rendic. Accad. Sci. Napoli, XXXVII, pp. 330-334. De Srerani Perez, T. 1905. ‘Contributo all’entomofauna dei cecidii.’ Marcellia, IV, pp. 36-40. De Witpeman, FE. 1897. ‘Une plante myrmécophile nouvelle (Scaphopelalum Thonneri De Wild. et Th. Durand).’ Bull. Herbier Boissier, V, No. 6 (Juin), pp. 521-525. 1904a. ‘Notices sur des plantes utiles ou intéressantes de la flore du Congo.” Brussels. I, fase. 2, pp. 223-396, Pls. x1u1-xxXviIt. ena. Wheeler, Ants of the Belgian Congo 541 ~~ 1904b. ‘Sur le Randia Lujzx De Wild. n. sp., plante myrmécophile et acaro- phile nouvelle de !a famille des Rubiacées.’ C. R. Ac. Sci. Paris, CXXXVIII, pp. 913-914. 1905-1907. ‘Mission Emile Laurent (1903-1904). Enumération des ‘ plantes récoltées par Emile Laurent, pendant sa derniére Mission au re Congo.’ Brussels. Fasc. 1, 1905, pp. 1-112, Pls. -xxxvuimr; fase. 2, 1905, pp. 113-192, Pls. xxxrx—x1v1; fasc. 3, 1906, pp. 193-354, Pls. -xivir-cvi; fase. 4, 1907, pp. ix-cxx, 355-450, Pls. cvm—cxuim; fase. _ 5, 1907, pp. i-viii, exxi-cexxv, 451-617. 19192. ‘La myrmécophilie dans le genre Uncaria (Rubiacées), en Afrique., C. R. Soc. Biol. Paris, LX XXII, pp. 1076-1078. -1919b. ‘Sur le Macaranga saccifera Pax, Euphorbiacée myrmécophile de l’ Afrique tropicale.’ C. R. Ac. Sci. Paris, CLXIX, pp. 394-396. 1920. ‘Clercdendrcn a tiges fistuleuses.’ C. R. Soc. Biol. Paris, LX XXIII, pp. 582-584. 1921. ‘Sur les théories dela myrmécophilie.’ C. R. Ac. Sci. Paris, CLX XII, pp. 124-126. De Wipeman, E. anv Duranp, T.: 1901. ‘Reliquie Dewevreanz.’ Brussels. 291 pp. £ Tenkuance, E.L. 1913. (Root lice kept by Lasius latipes on the roots of Thalesia uniflora.) Journ. New York Ent. Soc., XXI, p. 314. Drepenicus, : 1906. ‘Die Ameisen als Baukiinstler.’ Aus der Natur, II, pp. 80- Disraxr, Ww. % "1900-1911. ‘Insecta Transvaaliensia. A contribution to the knowl- edge of the entomology of South Africa.’ London. Pt. 1, 1900, pp. 1-24, Pls. 1-11; pt. 2, 1902, pp. 25-48; pt. 3, 1903, pp. 49-80, Pls. m-tv; pt. 4, 1903, pp. 81-96, Pls. v-vi1; pt. 5, 1904, pp. 97-128, Pls. 1x—xu1; pt. 6, 1904, pp. 129-258, Pls. xu1—xv; pt. 7, 1906, pp. 159-180, Pls. xvi-xvui; pt. 8, 1907, pp. 181-204, Pls. xvi—xrx; pt. 9, 1908, pp. 205-228, Pls. xx-xx1; pt. 10, 1910, pp. 229-252, Pls. xx1—xxim1; pts. 11-12, 1911, pp. 253-299, Pls. xx1v—xxvn. Dosps, C. M. 1916. ‘A curious fungus growth.’ Journ. East Afric. Uganda Nat. Hist. Soc., V, No. 9, pp. 59-60. Docrors van Leeuwen-Reynvaan, W. anv J. 1911. ‘Over de verspreiding der zaden van enkele Dischidia-soorten door middel van een miersoort: Tridomyrmex myrmecodie Emery.’ Versl. Vergad. Wis- Natuurk. : Afd. Ak. Wet. Amsterdam, XX, pt. 1, pp. 131-136. 1913. ‘Beitrige zur Kenntnis der Lebensweise einiger Dischidia-Arten.’ Ann. Jardin Bot. Buitenzorg, XX VII, pp. 65-91, Pls. xu-xv1. Dopp, F. P. 1902. ‘Notes on the Queensland green tree ants (Ee»phylla smaragdina Fab.?).’ Victorian Natural., XVIII, pp. 136-140, 140-142. Déorartes, W. 1910. ‘Ueber eine neue Galle an Caucalis daucoides.’ Botan. Zeitung, LXVIII, Abt. 2, pp. 313-316. Dor.ein, F. 1905a. ‘Beobachtungen an den Weberameisen (@cophylla smaragdina).’ Biol. Centralbl., XXV, pp. 497-507. 19056. ‘Die Pilzkulturen der Termiten.’ Verh. Deutsch. Zool. Ges., XV, pp. 140-149. 19062. ‘Ostasienfahrt. Erlebnisse und Beobachtungen eines Natur- forschers in China, Japan und Ceylon.’ Berlin. xiii+511 pp. 542 Bulletin American Museum of Natural History (Vol. XLV 1906). ‘Termite truffles.’ Spolia Zeylanica, III, pp. 203-209. 1920. ‘Mazedonische Ameisen. Beobachtungen tiber ihre Lebensweise.’ | Jena. iii+74 pp. 1921. ‘Mazedonien. Erlebnisse und Beobachtungen eines Naturforschers im Gefolge des deutschen Heeres.’ Jena. viii+592 pp. | Dominiqus, J. 1900. ‘Fourmis jexdiniices,’ Bull. Soc. Sci. Nat. Ouest, Nantes, 4 X, pp. 163-168. } DonistruorrPe, H. St. J. K. 1907. ‘Myrmecophilous notes for 1906 (with descrip- tions of two new Coccide by R. Newstead).’ Ent. Record, XTX, pp. 4-7, Pl. 1. 1912a. ‘Ants as honey pots.’ Marvels of the Universe, XXIII, pp. 940— 1912b. ‘Dairying ants.’ Marvels of the Universe, XX, pp. 804-806. 1912c. ‘The agricultural ant of Texas.’ Marvels of the Universe, XXVII, pp. 1099-1103. 1912d. ‘Myrmecophilous notes for 1911.’ Ent. Record, XXIV, pp. 4-10, 3440. 1913. ‘Myrmecophilous notes for 1912.’ Ent. Record, XXV, pp. 61-68, 89-97. 1914a. ‘Myrmecophilous notes for 1913.’ Ent. Record, XXVI, pp. 37-45. 1914b. ‘Nest of Lasius fuliginosus.’ Trans. Ent. Soc. London, Proc., pp. xviii—xix: 1914c. ‘Ant larve as sewing machines.’. Trans. Ent. Soc. London, Proc., p. ii. 1915. ‘British ants, their life-history and classification.’ Plymouth. xvi+379 pp., 18 Pls. 1917. ‘Dolichoderus (Hypcclinea) crawleyi n. sp., a species of ant new to science; with a few notes on the genus.’ Ent. Record, XXTX, pp. 201- 202. Doveras, J. W. 1865. ‘On the occurrence of Systellonotus triguttatus in company with Formica fusca.’ Ent. Monthly Mag., II, pp. 30-31. 1886. ‘Ants and Coccide.’ Ent. Monthly Mag., XXIII, p. 6. 1891. ‘Notes on some British and exotic Coccide (No. 21).’ Ent. Monthly Mag., X XVII, pp. 244-247. ‘6 Dovuior, H. 1889. ‘Les fourmis moissonreuses.’ Le Naturaliste, XI, p. 268. Drake, C. J. 1921. ‘A new ambrosia beetle from the Adirondacks; notes on the work of Xyloterinus politus Say.’ Ohio Journ. Science, XXI, pp. 201— 205. . Duparp, F. 1878. ‘Note relative aux mceurs des fourmis.’ Ann. Soc. Ent. France, (5) VIII, Bull., p. civ. Ducetuier, L. 1912. ‘Les fourmis moissonneuses en Algérie.’ Revue des Colons de |’ Afrique du Nord, Alger, I, pp. 223-227. Durour, L. 1856. ‘Note sur la Fcrmica barbara.’ Ann. Soc. Ent. France, (3) IV, pp. 341-343. Durr, G. R. 1912. ‘Life-histories of Indian insects. IV. Hymenoptera.’ Mem. Dept. Agric. India, Ent. Ser., IV, No. 4, pp. 183-267, Pls. x1—xtv. Dyes, T. A. 1916. ‘The seed-mass and dispersal of Hellebcrus fotidus Linn.’ Journ. Linn. Soe. London, Bot., XLITII, pp. 433-455, Pl. xxxvt. ‘a 1922} Wheeler, Ants of the Belgian Congo 543 --Epwanps, H. 1873a. ‘Notes on the honey-making ant of Texas and New Mexico.’ American Naturalist, VII, pp. 722-726. 18736. ‘Notes on the honey-making ant of Texas and New Mexico, Myr- mecocystus mexicanus of Westwood.’ Proc. California Ac. Sci., V, pp. 72-75. Ecorns, H. 1899. ‘Zur Lebensweise des Xyleborus cryptographus Ratz.’ Illustr. Zeitschr. Ent., IV, pp. 291-292. Eensoan, E.M. 1915. ‘Ants.’ Rept. Div. Ent. Biennial Period Ending Dee. 31, 1914, Hawaii Bd. Agric. For., pp. 139-140. rome, — 1892. (Fruit attacked by ants.) Zeitsehr. f. Obst-u. Gartenb., Dresden, No. 9, p. 99. Siasore, Jesste 8S. B. 1915. ‘Fungiin the nests of ants.’ Trans. Brit. Mycol. Soc., V, pt. 1, (1914), pp. 138-142, Pl. 1. Exus, J. B. anp Evernart, B. M. 1886-1887. ‘Synopsis of the North American Hypocreacee, with descriptions of the species.’ Journ. Mycol., II, 1886, pp. 28-31, 49-51, 61-69, 71, 73-80, 97-99, 109-111, 119, 121-125, 133-137; III, 1887, pp. 1-6, 113-116, 119. ae Tiaene, A. D.E. 1911. ‘New and noteworthy Rubiacee.’ Leaflets of Philippine Botany, Manila, III, pp. 971-1046. 1913. ‘Rubiacee from Mount Urdaneta.’ Leaflets of Philippine Botany, Manila, V, pp. 1855-1905. - Exvsier, E. 1907. ‘Das extraflorale Nectarium und die Papillen der Blattunter- ; seite bei Diospyros di:calor Willd.’ Anz. Ak. Wiss. Wien, XLIV, p. 419; Sitzb. Ak. Wiss. Wien, math. naturw. KI., CXVI, Abt. 1, pp. 1563- 1590, Pls. 1-11. Emery, C. 1888. ‘Catalogo delle formiche esistenti nelle collezioni del Museo civico di Genova. III. (Supplemento). Formiche raccolti dal sig. Elio Modigliani in Sumatra e nell’isola Nias.’ Ann. Mus. Civ. Genova, XXV, pp. 528-534, Pl. rx. 1889. ‘Alleanze difensive tra piante e formiche.’ Nuova Antologia Sci., (3) XIX, pp. 578-591. 1890. ‘Studii sulle formiche della fauna neotropica. I. Formiche di Costa Rica. II. Sopra alcune specie del genere Pseudomyrma. III. Nuove forme americane dei generi Strumizenys e Epitritus. IV. Gly- plemyrmes uncinatus Mayr. VY. Note sopra alcune specie di Crypto- cerus.’ Bull. Soc. Ent. Italiana, XXII, pp. 38-80, Pls. v-rx. 1891. ‘Zur Biologie der Ameisen. I. Die in Akaziendornen lebenden Ameisen von Costa Rica (pp. 163-168). II. Lirmetopum microce- phalum Panz., eine europiiische Raubameise. III. Ueber den Hoch- zeitsflug der Aueleen. IV. Die Ernte der Ameisen in Siideuropa (pp. 176-177). V. Ueber Beziehungen der Insekten zu den Ameisen.’ Biol. Centralbl., XI, pp. 165-180. 1892a. ‘Aelteres tiber Ameisen in Dornen afrikanischer Akazien.’ Zool. Anzeiger, XV, p. 237. 1892). ‘Estudio contributivo a la biologia de las hormigas.’ An. Mus. Nac. Costa Rica, (1888-1889), pp. 65-67. 1893. ‘Studio monografico sul genere Azteca Forel.’ Mem. Accad. Sc. Bologna, (5) III, pp. 319-352, 2 Pls. ae S44 Bulletin American Museum of Natural History (Vol. XLV 1894a. ‘Studi sulle formiche della fauna neotropica. VI-XVI.’ Bull. Soc. Ent. Italiana, XXVI, pp. 137-241, Pls. 1-1v. 1894). ‘Estudios sobre las hormigas de Costa Rica. I. Hormigas recogidas en 1889 por Anastasio Alfaro.’ An. Mus. Nac. Costa Rica, (1889), pp. 45-64, 2 Pls. 1896a. ‘Formiciden gesammelt in Paraguay von Dr. J. Bohls.’ Zool. Jahrb. Abt. Syst., [X, pp. 625-638. 18¢6b. ‘Aleune forme nuove del genere Azteca For. e note biologiche.’ Boll. Mus. Zool. Anat. Comp. Torino, XI, No. 230, pp. 1-7. 1897. ‘Formiche raccolte da Don Eugenio dei Principi Ruspoli, durante l’ultimo suo viaggio nelle regioni dei Somali e dei Galla.’ Ann. Mus. Civ. Genova, XX XVIII, pp. 595-605. 1899. ‘Végétarianisme chez les fourmis.’ Arch. Sci. Phys. Nat. Genéve, (4) VIII, pp. 488-490. Lave 1902. ‘Intorno alle larve di aleune formiche.’ Mem. Accad. Sc. Bologna, (5) VIII, pp. 3-10, Pls. r-1. : 1910. ‘Le formiche e gli alberi in Italia.’ Pro Montibus (Alpe organo ufficiale), Bologna, Nos. 19-20, (9 pp.). 1912a. ‘Les plantes A fourmis.’ Scientia, XII, xxiv-4, Suppl., pp. 41-56. ‘Le piante formicarie.’ Ibid., XII, pp. 48-62. 1912b. ‘Alcune esperienze sulle formiche granivore.’ Rend. Accad. Se. Bologna, N.8., XVI, pp. 107-117, 1 Pl. 1915. ‘La vita delle formiche.’ Turin. 254 pp. Ra 9 A. 1892. ‘Passifloracee africanw.’ Engler’s Bot. Jahrb., XIV, pp. 374— 393. Ensuin, E. 1911. ‘Gargara geniste F. und Formica cinerea Mayr.’ Zeitschr. Wiss. Insektenbiol., VII, pp. 19-21. Errincton pE LA Crorx, Mrs. 1900. ‘Observations sur le Termes carbonarius Haviland.’ Bull. Mus. Hist. Nat. Paris, VI, pp. 22-23. Escuenricu, K. 1900. ‘Ueber das regelmiissige Vorkommen von Sprosspilzen in dem Darmepithel eines Kifers.’ Biol. Centralbl., XXII, pp. 349-357. 1903. ‘Ueber die Biologie der Ameisen. Uebersicht tiber die neueren Arbeiten.’ Zool. Zentralbl., X, pp. 209-250. 1906a. ‘Ueber die Biologie der. Ameisen. 2. Teil. Die neueren Arbei'en (1905-1906).’ Zool. Zentralbl., XIII, pp. 405-440. 1906b. ‘Die Ameise. Schilderung ‘Sheer Lebenweise.’ Braunschweig. xx-+232 pp. (2d ed., Braunschweig, 1917, xvi+348 pp.). 19082. ‘Eine Ferienreise nach Erythrea.’ Aus der Natur, IV, pp. 115-124. 1908b. ‘Kleinere biologische Beobachtungen aus Erythrea.’ Verh. Ges. Deutsch. Nat. Aerzte Vers. LX XIX, 2, 1, pp. 247-248. 1908c. ‘Nochmals: Die Theorie der Ameisen und Pflanzen, ein Irrtum der Biologie.’ Beil. z. Allgem. Zeitung, Miinchen. 1908d. ‘Eine Ferienreise nach Erythrea.’ Leipzig. 44 pp. 1909a. ‘Die pilzziichtenden Termiten.’ Biol. Centralbl., X XTX, pp. 16-27, Pi. 1909b. ‘Ameisen und Pflanzen. Eine kritische Skizze mit besonderer Beriicksichtigung der forstlichen Seite:’ Tharandter Forstl. Jahrb., LX, pp. 66-96. Wheeler, Ants of the Belgian Congo 545 a 1909¢. ‘Die Termiten oder weissen Ameisen. Eine biologische Studie.’ Leipzig. xii+198 pp., 1 Pl. 1909-1910. ‘Die myrmekologische Literatur von Januar 1906 bis Juni ber ob 1909.’ Zeitschr. Wiss. Insektenbiol., V, 1909, pp. 285-289, 320-325, 364-368, 405-415; VI, 1910, pp. 25-36, 73-80. 19lla. ‘Zwei Beitrige zum Kapitel: “‘Ameisen und Pflanzen.’’’ Biol. Centralbl., XX XI, pp. 44-51. 1911b. “Termitenleben auf Ceylon.’ Jena. 247 pp. f Bram 8. 1876. ‘Plant fertilization.’ Nature, XIII, p. 427. Panza, F. C.v. 1909. ‘Bekaimpfung von Kakaowanzen durch Ameisen.’ Tropen- pflanzer, XIII, pp. 41-42. 1911. ‘Een en ander over symbiose in planten- en dierenrijk.’ Teysman- nia, XXIII, pp. 444460. F. 4. D. 1909. ‘Is the association of ants with trees a true symbiosis?’ Nature, LXXXI, No. 2070, p. 23. Faren, D.G. anv Cook, O. F. 1898. ‘Fungus gardening as practised by the termites in West Africa and Java.’ Science, VIII, pp. 659-660. Fangtmanson, C. O., Lamporn, W., and Perkins, R. C. 1914. ‘The growth of fungi on the shelters built over Coccide by Cremastcgaster-ants.’ Trans. Ent. Soc. London, Proc., pp. xlii-1. Fawcett, H. 8. 1913. ‘Fungus gardens cultivated by ants.’ Monthly Bull. State Comm. Horticulture, Sacramento, Cal., II, pp. £39-540. Fawcertr, W. 1886. ‘An entomogenous fungus.’ Ann. Sinn: Nat. Hist., (5) XVIII, pp. 316-318. Feope, F. 1897. ‘Ueber die Verbreitung von Samen und Pflanzen durch Thiere.’ Jahresb. Schles. Ges. Vaterl. Cult. Breslau, LX XIV, (1896), Abth. 2c, pp. 45-53. : 1900. ‘Ueber Symbiose zwischen Pflanzen und Thieren.’ Jahresb. Schles. Ges. Vaterl. Cult. Breslau, LX XVII, (1899), Abth. 2c, pp. 2-15. Ferprwanpsen, G. anv Wixce, 0. 1908. ‘Fungi from the Danish West Indies collected 1905-1906 by G. Raunkier. I. Phycomycetex, Ustilaginee, Uredinee, Discomycevee, Pyrenomycetex et Fungi imperfecti.’ Botanisk Tids- skr., XXIX, pp. 1-25, Pls. 1-11. : Fiusata, K. 1907. ‘Eine Wespen zérstérende Ameise aus Paraguay. Eciton vagans Olivier.’ Zeitschr. Wiss. Insektenbiol., III, pp. 83-87, 154-156. 1909. ‘Cecropia peltata und ihr Verhiltnis zu Azteca Alfari, zu Atta serdens und anderen Insekten, mit einer Notiz iiber Ameisendornen bei Acacia cavenia.’ Biol. Centralbl., XXIX, pp. 1-16, 33-55, 56-77, 5 Pls. Fiercner, T.B. 1907. ‘Leaf-nesting ants.’ Spolia Zeylanica, V, p. 64. Fut, W. P. 1914. ‘On the capture of living insects by the corn-field ant (Lastus niger americanus). Journ. Econ. Ent., VIII, pp. 476-478. Fiorent, P. 1920. ‘La fourmi d’Argentine.’ Journ. Agric. Pratique, Paris, XXXIV, No. 35, p. 172. FonTanet,—1918. ‘Ants andaphids.’ 10th Rep. Quebec Soc. Protect. Plants from Insects and Fungous Diseases, (1917-1918), pp. 52-57. Forses, H.O. 1880. ‘Notes from Java.’ Nature, XXII, p. 148. 1885. ‘A naturalist’s wanderings in the Eastern Archipelago.’ New York. xix +536 pp. 546 Bulletin American Museum of Natural History [Vol. XLV Forses, 8. A. 1887. ‘Relations of ants and aphids.’ American Naturalist, XXT, pp. 579-580. 1891. ‘A summary history of the cornroot Aphis (Aphis maidiradicis).’ Insect Life, III, pp. 233-238; 17th Rept. State Entom. Illinois, pp. 64-70. 1894. ‘A monograph of insects injurious to Indian corn,’ 18th Rept. State Entom. Illinois, pp. 1-171, 15 Pls. 1905. ‘The more important insect injuries to Indian corn.’ 23rd Rept. Nox. Benef. Ins. Illinois, pp. 1-273, 8 Pls. 1906. ‘The corn-root Aphis and its attendant ant.’ Bull. 60 U. 8. Dept. Agric. Div. Ent., pp. 29-39. 1908a. ‘Experiments with repellents against the corn-root Aphis, 1905 and 1906.’ Univ. Illinois Agric. Exper. Stat., Bull. 130, pp. 1-28. 1908). ‘Habits and behavior of the cornfield ant, Lasius niger americanus.’ Univ. Illinois Agric. Exper. Stat., Bull. 131, pp. 1-45, 1 Pl. 1909.2 ‘The general entomological ecology of the Indian corn-plant.’ American Naturalist, XLIII, pp. 283-301. 1909b. ‘Habits and behavior of the corn-field ant, Lasius niger americanus.’ 25th Rept. State Entom. Illinois, pp. 27-40, Pl. 1. 1915. ‘Recent Illinois work on the corn-root-A phis and the control of its injuries.’ 28th Rept. State Entom. Illinois, pp. 1-62. Foret, A. 1874. ‘Les fourmis de la Suisse.’ Nouv. Mém. Soc. Helvét. Sci. Nat Zurich, XXVI, 447 pp., 2 Pls. 1885. ‘Etudes myrmécologiques en 1884, avec une description des organes sensoriels des antennes.’ Bull. Soc. Vaudoise Sc. Nat., (2) XX, pp. 316-380, Pl. x1. 1892a. ‘Die Akazien-Crematogaster von Prof. Keller aus dem Somaliland.’ Zool. Anzeiger, XV, pp. 140-143. 1892b. ‘Die Nester der Ameisen.’ Neujahrsbl. Naturf. Ges. Zurich, XCV, (1893), pp. 1-36, PI. 1892c. ‘Liste der aus dem Somaliland von Hrn. Prof. Dr. Keller aus der Expedition des Prinzen Ruspoli im August und September 1891 zuriick- gebrachten Ameisen.’ Mitth. Schweiz. Ent. Ges., VIII, pp. 349-354. 1894. “Abessinische und andere afrikanische Asneseees.? Mitth. Schweis. Ent. Ges., [X, 2, pp. 64-100. 1895. ‘Une nouvelle fourmi melligére.’ Ann. Soc. Ent. Belgique, XX XIX, pp. 429-430. 1896a. ‘Die Fauna und die Lebensweise der heeiliek im kolumbischen Urwald und in den Antillen.’ Verh. Schweiz. Naturf. Ges., 79. Jahresv., pp. 148-150. 18966. ‘Zur Fauna und Lebensweise der Ameisen im columbischen Urwald.’ Mitth. Schweiz. Ent. Ges., IX, pp. 401-411. 1896c. ‘Ants’ nests.’ Ann. Rept. Smiths. Inst. for 1894, pp. 479-505, Pls. LV-LVI. 1896d. ‘Quelques particularités de l’habitat des fourmis de l’Amérique tropicale.’ Ann. Soc. Ent. Belgique, XL, pp. 167-171. j 1897. ‘Communication verbale sur les moeurs des fourmis de l’Amérique méridionale.’ Ann. Soc. Ent. Belgique, XLI, pp. 329-332. Wheeler, Ants of the Belgian Congo 547 ee 1898. ‘La parabiose chez les fourmis.’ Bull. Soc. Vaudoise Sc. Nat., (4) XXXIV, pp. 380-384. 1899. (Letter from North Carolina on the habits of North American ants.) Ann. Soc. Ent. Belgique, XLII, pp. 438-447. 1900. ‘Ebauche sur les meeurs des fourmis de l’Amérique du Nord.’ Rivista Se. Biol., II, pp. 1-13. 190la. ‘Variétés myrmécologiques.’ Ann. Soc. Ent. Belgique, XLV, pp. 334-382. ae 1901. ‘Sketch of the habits of North American ants.’ Psyche, [X, pp. q 231-239, 243-245. Fee 1902. ‘Fourmis d’Algérie récoltées par M. le Dr. K. Escherich. Un 4 Camponotus habitant des tiges renflées.’ Ann. Soc. Ent. Belgique, a XLVI, pp. 462-463. 1903a. ‘Die Sitten und Nester einiger Ameisen der Sahara bei Tugurt und Biskra beobachtet von Dr. A. Diehl.’ Mitth. Schweiz. Ent. Ges., X, } pp. 453-459. - : 1903b. ‘Faune myrmécologique des noyers dans le canton de Vaud.’ Bull. Soc. Vaudoise Sc. Nat., (4) XX XIX, pp. 83-94. 1903c. ‘Mélanges entomologiques, biologiques et autres.’ Ann. Soc. Ent. Belgique, XLVII, pp. 249-268. 1904a. ‘Miscellanea myrmécologiques.’ Rev. Suisse Zool., XII, pp. 1-52. 1904b. ‘In und mit Pflanzen lebende Ameisen aus dem Amazonas-Gebiet und aus Peru gesammelt von Herrn E. Ule.’ Zool. Jahrb. Abt. Syst., XX, pp. 677-707. 1905. ‘Einige biologische Beobachtungen des Herrn Prof. Dr. Geeldi an brazilianischen Ameisen.’ Biol. Centralbl., XXV, pp. 170-181. 1909a. ‘Ameisen aus Java and Krakatau beobachtet und gesammelt von Edward Jacobson. 1. Theil.’ Notes Leyden Mus., XXXI, pp. 221- 253 (ethological part by E. Jacobson, pp. 233-251; remarks by A. Forel, ? pp. 252-253). 19095. ‘Fourmis du Musée de Bruxelles. Fourmis de Benguela récoltées par M. Creighton Wellman et fourmis du Congo récoltées par M. Luja, Kohl et Laurent.’ Ann. Soc. Ent. Belgique, LIII, pp. 51-73. 1910. ‘Ameisen aus der Kolonie Erythriia, gesammelt von Prof. Dr. K. Escherich (nebst einigen in West-Abessinien von Herrn A. Gilg gesammelten Ameisen).’ Zool. Jahrb. Abt. Syst., X XIX, pp. 243-274. 1911. ‘Ameisen aus Java beobachtet ‘und gesammelt von Herrn Edward Jacobson. 2. Theil.’ Notes Leyden Mus., XXXIII, pp. 193-198. 1913. ‘Wissenschaftliche Ergebnisse einer Forschungsreise nach Ostindien. II. Ameisen aus Sumatra, Java, Malacca und Ceylon gesammelt von v. Buttel-Reepen in 1911-1912.’ Zool. Jahrb. Abt. Syst., XXXVI, pp. 1-148. 1916. ‘Fourmis du Congo et d’autres provenances.’ Revue Suisse Zool., XXIV, pp. 397-460. 1920a. ‘Fourmis trouvées dans des galles de Cordia et d’Agcnandra.’ Bull. Soc. Bot. Genéve, (2) XII, pp. 201-298. 1920). ‘Les fourmis de la Suisse.’ 2d Ed. La Chaux-de-Fonds. xvi+333 pp. 548 Bulletin American Museum of Natural History (Vol. XLV Foster, E. 1908. ‘The introduction of Iridomyrmex humilis (Mayr) into New Orleans.’ Journ. Econ. Ent., I, pp. 289-293. Fow er, W.W. 1893. ‘Coccids in ants’ nests.’ Ent. Monthly Mag., X XIX, p. 17. Fresentvs, J. B. 1850-1863. ‘Beitriige zur Mykologie.’ Frankfurt a. M. Pt. 1, 1850, pp. 1-38; pt. 2, 1852, pp. 39-80; pt. 3, 1863, pp. 81-111; Pls. I-x1ml. Frrepericus, K. 1920. ‘Weberameisen und Pflanzenschutz.’ Tropenpflanzer, XXIII, pp. 142-150. Frocearr, W. W. 18€6. ‘Honey ants.’ Horn Exp. Central Australia, pt. 2, pp. 385-392, Pl. xxvit. 1902. ‘Insects of the wattle trees (Acacia).’ Agric. Gaz. New South Wales, XIII, pp. 701-720, Pls. 1-1. 1905. ‘Domestic insects: ants.’ Agric. Gaz. New South Wales, XVI, pp. 861-866, PI. 1906a. ‘Domestic insects: ants. With catalogue of Australasian species.’ Miscell. Publ. Agric. Dept. New South Wales, No. 889, 35 pp., 1 Pl. 1906. ‘Rabbits and ants.’ Agric. Gaz. New South Wales, XVII, pp. 1113-1119. 1919. ‘Ants and flowers.’ Australian Naturalist, IV, p. 88. Fromuoiz, C. 1885. ‘Die digyptische Hausameise, Mcncmorium pharaonis L., in Berlin.’ Ent. Nachr., XII, p. 122. Fucus, G. 1907. ‘Ueber die Fortpflanzungsverhaltnisse der rindenbriitenden Borkerkifer, verbunden mit einer geschichtlichen und kritischen Dar- stellung der bisherigen Literatur.’ Munich. v+83 pp., 10 Pls. Fuiier, C. 1915. ‘Observations on some South African termites.’ Ann. Natal Mus., III, pt. 2, pp. 329-504, Pls. xxv—xxxv. Funxuouser, W.D. 1915. ‘Life-history of Vanduzea arquata Say (Membracide).’ Psyche, XXII, pp. 183-198, Pl. xvu- 1917. ‘Biology of the Membracide of the Cayuga Lake Basin.’ Cornell Univ., Agric. Exp. Stat., Mem. XI, pp. 177-445, Pls. xxim—xty. F.W. 1884. ‘Die Ameisen als Waldschidlinge.’ (sterreich. Forstzeitg., II, p. 244. Gapeceau, E. 1907a. ‘Les fourmis mycophages.’ La Nature, XXXVI, 1, pp. 49-51. 1907b. ‘Les plantes myrmécophiles.’ La Nature, XXXV, 2, pp. 295-298. Gavow, H. 1908. ‘Through Southern Mexico.’ London. xvi+527 pp. Gatiarpo, A. 1912. ‘Observaciones sobra una hormiga invasora. (Iridomyrmex humilis Mayr).’ Bol. Soc. Physis, Buenos Aires, I, pp. 133-138. 1915. ‘Observaciones sobre algunas hormigas de la Repdblica Argentina.’ An. Mus. Nac. Hist. Nat. Buenos Aires, XX VII, pp. 1-35. 1916a. . ‘Notes systématiques et éthologiques sur les fourmis Attines de la République Argentine.’ An. Mus. Nac. Hist. Nat. Buenos Aires, XXVIII, pp. 317-344. 1916). ‘Notas acerca de la hormiga Trachymyrmex pruindsus Emery.’ An. Mus. Nac. Hist. Nat. Buenos Aires, XXVIII, pp. 241-252, Pls. vi— Ix. 1916c. ‘Las hormigas de la Reptiblica Argentina. Subfamilia Dolicoderi- nas.’ An. Mus. Nac. Hist. Nat. Buenos Aires, XXVIII, pp. 1-130. Gams, H. 1921. ‘Zur Ameisengeographie von Mitteleuropa.’ Naturwiss. Wochen- schr., XXXVI, pp. 414-416. _ Wheeler, Ants of the Belgian Congo 549 Garman, H. 1895. ‘The bean root-louse (Tychea phasecli, Passerini).’ 7th Ann. ae Rept. Kentucky Agric. Exper. Stat., (1894), pp. xxii—xxvii. - Garrerr,A.O. 1910. ‘Honey ants in Utah.’ Science, N. S., XXXII, pp. 342-343. a. Gara, 8. 1893. ‘Ants as plant-guardians.’ The Great World’s Farm, London, pp. Ts 306-342 Gexé, G. 1844. ‘Memorio per servire alla storia naturale di aleuni Imenotteri.’ Mem. Soc. Italiana Res. Modena, XXIII, (1842), pp. 30-62. . Garey, T.G. 1873. ‘The milch-cows of the ants.’ Canadian Ent., V, pp. 207-208. Gemaeex, F. 1912. ‘Untersuchungen iiber den Bau und die Echeneweles der Lymexyloniden, speziell des Hylecetus dermestoides L.’ Zeitschr. Wiss. + Zool., CI, pp. 683-735, Pls. xxx—xxx1. ‘ Gensreceen, A. 1871. ‘Beitrag zur Insektenfauna von Zanzibar. IV. Hymen- optera.’ Arch. f. Naturg., XX XVII, 1, pp. 349-357. Gran, A. 1893. (Coceids living with ants.) Ann. Soc. Ent. France, LXII, Bull., . pp. excix—ce. 1894a. ‘Sur une cochenille souterraine des vignes du Chili.’ C. R. Soe. Biol. Paris, (10) I, pp. 126-128. 1894b. ‘Sur les transformations du Margarcdes vitium Gd.’ C. R. Soe. Biol. Paris, (10) I, pp. 412-414. 1894c. ‘Troisiéme note sur le genre Margarodes.’ C. R. Soc. Biol. Paris, (10) I, pp. 710-713. 1897. ‘Sur la distribution géographique des cochenilles du genre Marga- rodes et sur deux espéces nouvelles de ce genre.’ C. R. Soc. Biol. Paris, (10) IV, pp. 683-685. 1901. ‘Sur deux champignons parasites des cécidies.’ Bull. Soc. Ent. France, pp. 46-48. Ginsox, R. J. H. 1894. ‘The mushroom beds of the South American ants.’ Proc. Liverpool Lit. Soc., XLVIII, pp. 99-105. Sei E. 1908. ‘Flacourtiaceer dtwann Engler’s Bot. Jahrb., XL, 4, pp. 444-518. Giitet, J. anp PAque, E. 1910. ‘Plantes principales de la région de Kisantu.’ Brussels. 120 pp. Guerre, C. P. 1898. ‘An insect-catching plant.’ Ent. News, IX, pp. 169-170. 1918. ‘Some grass-root aphids.’ Ent. News, X XIX, pp. 281-284. Gapet, K. 1888. ‘Ueber epiphytische Farne und Muscineen.’ Ann. Jardin Bot. * * Buitenzorg, VII, pp. 1-73, Pls. 1-1x. 1889-1893. ‘Pflanzenbiologische Schilderungen.’ Jena. I, 1889, iv-+239 pp.; Il, 1893, iv+386 pp. Gewor, E. A. 1904. ‘Formigas cuyabanas.” Journal do Commercio, Rio de Janeiro, 27 de Abril, p. 2; and 7 de Maio. 1905a. ‘Beobachtungen iiber die erste Anlage einer neuen Kolonie von Alta cephalotes.’ C. R. 6° Congr. Intern. Zool., Berne, pp. 457-458. 1905b. ‘Myrmecologische Mitteilung das Wachsen des Pilzgartens von Alta cephalotes betreffend.’ C. R. 6° Congr. Intern. Zool., Berne, pp. Gérne, R. 1895. (On ambrosia beetles.) Bericht d. Lehranst. f, Obst-, Wein- und Gartenbau, Geisenheim, (1894-95), p. 25. Goopcuitp, J.G. 1903. ‘Ants in relation to flowers.’ Trans. Edinburgh Field Soc., V, pp. 10-23. 550 Bulletin American Museum of Natural History ‘ (Vol. XLV Goor, T. vAN peR. 1915. ‘Over de biologie der gramangmier (Plagiclepis longipes Jerd.). Meded. Proefstat. Midden-Java, Salatiga, XIX, pp. 1-60. 1916. ‘Verdere ondersoekingen omtrent de economische beteekenis der gramang-mier.’ Meded. Proefstat. Midden-Java, Salatiga, XXII, pp. 1-120, 6 Pls. 1917. ‘De zwarte cacao-mier (Dolichoderus bituberculatus Mayr).’ Meded. Proefstat. Midden-Java, Salatiga, XXV, pp. 1-142, 2 Pls. Govupte, J.C. 1898. ‘Ants and aphides.’ Victorian Natural., XTV, p. 148. GovuneE.LLz, E. 1896. ‘Transport de terres effectué par des fourmis au Brésil.’ Bull. Soc. Ent. France, pp. 332-333. Govureav, O. 1867. (Aphids living on roots with Lasius flavus.) Ann. Soc. Ent. France, (4) VII, Bull., pp. Ixxxviii-—Ixxxix. , Gowpey, C. G. 1917. ‘A list of Uganda Coccide, their food-plants and natural enemies.’ Bull. Ent. Research, VIII, pt. 2, pp. 187-189. Grapuam, M. 1918. ‘The house ants of Jamaica as carriers of pathogenic microor- ganisms.’ Jamaica Publ. Health Bull., (1917), pp. 29-34. Granam, S.A. 1918. ‘The carpenter ant as a destroyer of sound wood.’ 17th Rept. State Entom. Minnesota, pp. 32-40. Graumnitz,C.v. 1913. ‘Die Blattschneider-Ameisen Siidamerikas.’ Intern. Ent. , Zeitschr., VII, pp. 233, 240-242. G. (Gray), G. R. 1858. ‘Notices of insects that are known to form the bases of fungoid parasites.’ Hampstead. 4to, 22 pp., 5 Pls. (privately printed). Green, E.E. 1896. ‘On the habits of the Indian ant, @cophylla smaragdina Latr.’ Trans. Ent. Soc. London, Proc., pp. ix-x; Zdologist, (3) XX, pp. 110. 1900a. ‘Note on Dorylus orientalis West.’ Indian Museum Notes, V, p. 39. 1900b. ‘Note on the web-spinning habits of the “red ant’’ Gcophila smarag- dina.’ Journ. Bombay Nat. Hist. Soc., XIII, p. 181. 1900c. ‘Note on the attractive properties of certain larval Hemiptera.’ Ent. Monthly Mag., XX XVII, p. 185. 1903. ‘Pupe of the “red ant”? (Ecophylla smaragdina).’ Spolia Zeylanica, I, pp. 73-74. 1904. ‘On some Javanese Coccide. With descriptions of new species.’ Ent. Monthly Mag., XL, pp. 204-210. Groom, P. 1893. ‘On Dischidia Rafflesiana (Wall.).’ Ann. of Botany, Vi, pp. 223-242, Pl. x. Guentuer, K. 1913. ‘Die lebenden Bewohner der Kannen der insektenfressenden Pflanze Nepenthes destillatoria auf Ceylon.’ Zeitschr. Wiss. Insekten- biol., [X, pp. 90-95, 122-130, 156-160, 198-207, 259-270. Guérin, P. 1906. ‘Sur les domaties des feuilles de Diptérocarpées.’ Bull. Boe. Bot. France, LIII, pp. 186-192. Guerxke, M. 1894. ‘Verbenacew africane. I.’ Engler’s Bot. Jahrb., XVIII, pp. 165-183. 1903. ‘Verbenacee africane. III.’ Engler’s Bot. Jahrb., XX XIII, 2, pp. 292-300. 2 Guitpine, L. 1829. ‘An account of Margarcdes, a new genus of insects found in the neighborhood of ants’ nests.’ Trans. Linn. Soc. nese XVI, pt. 1, pp. 115-119, Pl. xn. Wheeler, Ants of the Belgian Congo 551 a - Haasis, FW. 1917. a ‘Dying of young pines in circles about anthills.’ Journ. Forestry, XV, pp. 763-771. Hazertanpt, G. 1893. ‘Eine botanische Tropenreise. Indo-Malayische Vegeta- tionsbilder und Reiseskizzen.’ Leipzig. viii+300 pp. os Hacrpons, M. 1907. ‘Pilzziichtende Borkenkifer.’ Naturwiss. Wochenschr., XXII, pp. 289-293. 1910. ‘Coleoptera. Fam. Ipide,’ Genera Insectorum, fase. 111, pp. 1-178, . ate 14 Pls. = ‘Hauum, E. 1887. ‘Die symbiose zwischen Ameisen und Pflanzen.’ Humboldt, ees VI, pp. 453-456. Hamnz0x, J. 1890. ‘The inkghitents of a hickory nut hull.’ Ent. News, I, pp. 49-50. liceee. A. 1910. ‘Grundziige zur Biologie der Laubblitter.’ Beih. z. Bot. Centralbl., XXV, Abt. 1, pp. 137-182. Haxa, K. 1914. ‘On fungi parasitic on insects found in Gifu Profecture,’ Botan. Magaz., Tokyo, XXVIII pp. 339-351. Harpwicxe, T. 1828. ‘Observations on the loves of the ants and the aphids.’ Zool. Journ., IV, pp. 113-115. Hanes, J. A. 1907. ‘The fungi of termite nests.’ American Naturalist, XLI, pp. . 536-539. Harr, J. H. 1895a. ‘Natural history notes. No. 14. The hunter ant of Trinidad.’ Bull. Trinidad Bot. Gard., II, pp. 82-83. 1895b. ‘Diacrium (Epidendrum) biccrnutum Hook. Diacrium (Epidendrum) indivisum Bradf. Mss.’ Bull. Trinidad Bot. Gard., II, pp. 79-80. 1897. ‘Life history of the parasol ant.’ Bull. Trinidad Bot. Gard., IT, pp. 166-178. Hartia, T. 1844. (On ambrosia beetles.) Allgem. Forst- und Jagdzeitung, XIII, pp. 73-74. 1872a. ‘Der Fichtensplintkafer Bostrychus (Xyloterus) lineatus.’ Allgem. Forst- und Jagdzeitung, XLVIII, pp. 181-183. 1872b. ‘Der Buchensplintkifer Bostrychus (Xyloterus) domesticus.’ Allgem. Forst- und Jagdzeitung, XLVIII, pp. 183-184. Havitanp, G. D. 1897. ‘A revision of the tribe Naucleew (Nat. Ord. Rubiacez).’ Journ. Linn. Soc. London, Bot., X XXIII, pp. 1-94, Pls. 1-1v. 1898. ‘Observations on termites, or white ants.’ Journ. Linn. Soc. London, Zool., XX VI, pp. 358-442, Pls. xxu—xxv. (Reprinted in 1902, Ann. Rept. Smithson. Inst. for 1901, pp. 667-678, 4 Pls.) Hawtey, R.C. anv Recorp, S.J. 1916. ‘Do ants kill trees about their colonies?’ American Forestry, XXII, pp. 685-686. Hayes, W. P. 1920. ‘Sclenopsis molesta (Say): a biological study.’ Agricult. Exper. Stat. Kansas State Agric. Coll., Techn. Bull. 7, pp. 1-55. Heapuer, T. J. anp Dean, G. A. 1908. ‘The mound-building prairie ant.’ Bull. 154 Kansas State Agric. Coll., pp. 165-180. Hepecock, G. G. 1906. ‘Studies upon some chromogenic fungi which discolor wood.’ 17th Ann. Rept. Missouri Bot. Gard., pp. 59-114, Pls. m—x1. Heer, O. 1852. ‘Ueber die Haus-Ameise Madeiras.’ An die Ziircherische Jugend, auf das Jahr 1852, von der Naturf. Ges., LIV, pp. 1-24, PI. 1. 1856. ‘On the house ant of Madeira’ (translated by R. T. Lowe). Ann. Mag. Nat. Hist., (2) XVII, pp. 209-224, 322-333, Pl. r. 552 Bulletin American Museum of Natural History [Vol. XLV Herm, F. 1894. ‘DégAts occasionnés sur les tubercules de pomme de terre par les fourmis.’ Ann. Soc. Ent. France, LXIII, pp. 29-32. 1895. ‘Plantes et fourmis. Relations biologiques.’ Assoc. Frang. Avance. Sci., C. R. 24° Sess., Paris, 1°°° partie, pp. 31-75. 1897. ‘Ants as guests of plants.’ Popular Science Mo., L, pp. 827-831. 1898. ‘The biologic relations between plants and ants.’ Ann. Rept. Smiths. Inst. for 1896, pp. 411-455, Pls. xvu—xxu. 1908. ‘Myrmécophilie probable des Pavia Boerh.’ Notes Bot. Pure et Appliquée, 2 pp. Hems.ey, W.B. 1883. ‘Social life of ants and plants.’ Gardeners’ Chronicle, N.58., XX, pp. 71-72. Henninos, P. 1902a. ‘Einige neue Ccrdiceps-Arten aus Surinam.’ Hedwigia, XLI, pp. 167-169. 1902b. ‘Fungi blumenavienses II a cl. Alfr. Méller lecti.’ Hedwigia, XLI, pp. 1-33. 1904a. ‘Fungi amazonici II a cl. Ernesto Ule collecti.’ Hedwigia, XLIII, pt. 4, pp. 242-273, Pl. rv. 1904b. ‘Ueber Cordiceps-Arten.’ Nerthus, Altona, VI, pp. 1-4, Pl. 1. Henninos, P. anp Nyman, E. 1899. ‘Vorldufige Mittheilungen tiber einige neue Agaricineen aus javanischen Termitenbauten.’ Naturwiss. Wochen- schr., XIV, p. 28. . Henstow, G. 1907. ‘On some remarkable adaptations of plants to insects.’ Journ. Roy. Hortic. Soc., XXII, pp. 97-103. Henscuet, G. 1876. ‘Schonet die Waldameise!’ Centralbl. Ges. Forstwesen, II, p. 160. Hernanvez, F. 1651. ‘Rerum medicarum Nove Hispanize thesaurus seu planta- rum, animalium et mineralium Mexicanorum historia.’ Rome. xiv +950+22+90+6 pp. Herrera, A. L. 1902. ‘Hormiga destructora del picudo.’ Bol. Com. Parasitologfa Agricola, Mexico, I, pp. 404—406. 1905. ‘Modo de usar el aparato exterminador de hormigas.’ Com. Para- sitologia Agricola, Mexico, Cire. 28, pp. 1-3. Herscuxo, A. 1907. * ‘Der Ameisenbesuch bei Centaurea montana L.’ Wien. Ent. Zeitg., XX VI, pp. 329-332. 1908. ‘Ueber den Insektenbesuch bei einigen Vicia-Arten mit extrafloralen Nektarien.’ Wien. Ent. Zeitg., XX VII, pp. 299-305. 1916. ‘Ueber den Insektenbesuch bei Vicia Faba L.’ Wien. Ent. Zeitg., XXXYV, pp. 123-125. 1919. ‘Ameisen und Psylliden.’ Wien. Ent. Zeitg., XX XVII, pp. 212-213. Hewitt, C.G. 1909. ‘Tent-building habits of ants.’ Ottawa Naturalist, XXIII, pp. 168-170. HitpEBRAND, F. 1892. ‘Einige Beobachtungen an Keimlingen und Stecklingen.’ Botan. Zeitung, L, pp. 1-11, 17-24, 33-42, PI. 1. 1898. ‘Die Gattung Cyclamen, L., eine systematische und biologische Monographie.’ Jena. 190 pp., 6 Pls. Hinps, W. E. 1907a. ‘Some factors in the natural control of the Mexican cotton boll weevil.’ Bull. 74 U. 8. Dept. Agric. Bur. Ent., pp. 1-79. 1907b. ‘An ant enemy of the cotton boll weevil.’ Bull. 63 U.S. Dept. Agric. Bur. Ent., pp. 45-48. a) | Wheeler, Ants of the Belgian Congo 553 x Mier, B. R. W. G. 1920. ‘A naturalist in Himalaya.’ London. xii+300 pp. a F. vy. 1908. ‘Fragmente zur Mykologie, V.’ Sitzb. Ak. Wiss. Wien, math. naturw. Kl., CXVII, Abt. 1, pp. 985-1032, Pls. r-1v (169. Ueber Termitenpilze, pp. 985-999). 1909. ‘Fragmente zur Mykologie. VI.’ Sitzb. Ak. Wiss. Wien, math. naturw. Kl., CX VIII, Abt. 1, pp. 275-452, Pl. Hocscumn, J. 1898. ‘Ueber Ameisenpflanzen.’ Jahresb. Schles. Ver. Vaterl. Cult. Breslau, LX XV, (1897), Abth. 2c, pp. 11-16. Bi. Houaren, N. 1904. ‘Ameisen (Formica exsecta Ny.) als Hiigelbildner in Simpfen.’ Zool. Jahrb. Abt. Syst., XX, pp. 353-370. Bg Sores, C. 1898. ‘Mycologische Untersuchungen aus den Tropen.’ Berlin. 8+122 pp., 12 Pls. 1899. ‘Pilzbauende Termiten.’ Botan. Untersuch. Festschr. f. Schwen- _dener, pp. 411-420. 1901. ‘Fungus cultures in the tropics.’ Ann. Roy. Botan. Gard. Peradeniya, I, pp. 27-37, Pl. 1. 1912. ‘In der Tropenwelt.’ Leipzig. vi+210 pp. ae J.D. 1871. ‘On Barteria, a new genus of the Passifloree from the Niger River.’ Journ. Linn. Soc. London, Bot., V, pp. 14-15, Pl. 1. Hore, F.W. 1837. ‘Inquiries into the ground for the opinion that ants lay up stores of food for the winter.’ Trans. Ent. Soc. London, II, Proc., p. xxxvii. 1840. ‘On some doubts respecting the economy of ants.’ Tenn. Ent. Soc. | London, II, pp. 211-213. ees A.D. 1898. ‘On the history and habits of the “wood engraver” ambrosia beetle—Xyleborus zylographus (Say), Xyleborus saxeseni (Ratz.)—with brief descriptions of different stages.’ Canadian Ent., XXX, pp. 21-29, Pls. 1-11. 1899. ‘Report on investigations to determine the cause of unhealthy condi- tions of the spruce and pine from 1880-1893.’ Bull. 56 West Virginia Agric. Exper. Stat., pp. 197-461. 1904a. ‘Catalogue of exhibits of insect enemies of forest and forest products at the Louisiana purchase exposition.’ Washington; pp. 15-16. 19045. ‘Insect injuries to hardwood forest trees.’ Yearbook U. 8. Dept. Agric., (1903), pp. 313-328, Pl. xxxrx. 1905. ‘Insect injuries to forest products.’ Yearbook U. 8. Dept. Agric., (1904), pp. 381-398. Horne, C. 1872. (Seed-collecting ants in India.) Science Gossip, No. 89, p. 109. Hoven, W. 1888. ‘Ants transporting the scale-bug.’ Science, XII, p. 204. Howarp, L.O. 1895. ‘Margarcdes in the United States.’ Insect Life, VIII, p. 359. 1901. ‘On the habits of Entilia sinuata.’ Bull. 30 U. 8. Dept. Agric. Div. Ent., N.8., pp. 75-77. Husparv, H.G. 1897a. ‘Ambrosta-beetles.’ Yearbook U.S. Dept. of Agric., (1896), pp. 421-430. 18976. ‘The ambrosia-beetles of the United States.’ Bull. 7 U. 8. Dept. Agric. Div. Ent., N. 8., pp. 9-30. Huser, J. 1905. ‘Ueber die Kolonienertadung bei Atta serdens.’ Biol. Centralbl., XXV, pp. 606-619, 624-635. 1907. ‘The founding of colonies by Atta serdens.’ Ann. Rept. Smiths. Inst. for 1906, pp. 355-367, Pls. 1-v. 554 Bulletin American Museum of Natural History [Vol. XLV 1908. ‘A origem das colonias de Satiba (Atta serdens).’ Bol. Museu Geeldi Para, V, (1907-08), pp. 223-241. Huser, P. 1810. ‘Recherches sur les mceurs des fourmis indigénes.’ Paris and Geneva. xv+328 pp., 2 Pls. 1820. ‘The natural history of ants.’ Translated by J. R. Johnson. London. xlv+398 pp., 1 PIL. Honcerrorp, H. B. anp Witu1aMs, F. X. 1912. ‘Biological notes on some Kansas Hymenoptera.’ Ent. News, XXIII, pp. 241-260, Pls. xrv—xvr. Hunter, W. D. 1907. ‘Some recent studies of the Mexican cotton boll weevil.’ Yearbook U.8. Dept. Agric., (1906), pp. 313-324, Pl. xv1. 1912. ‘Two destructive Texas ants.’ U. 8. Dept. Agric. Bur. Ent. Cire. No. 148, pp. 1-7. Hours, E. 1886. ‘Ameisen als Pflanzenschutz. Verzeichniss der bisher bekannten myrmekophilen Pflanzen.’ Frankfurt a. Oder. 15 pp., 3 Pls. 1887. ‘Myrmekophile und myrmekophobe Pflanzen.’ Berlin. 27 pp., 2 Pls. 1888. ‘Neue myrmekophile Pflanzen.’ Monatl. Mitth. Gesammtgeb. Naturwiss., V, pp. 219-220. c InertnG, H. vy. 1882. ‘Ueber Schichtenbildung durch Ameisen (Atta cephalotes).’ Neues Jahrb. Mineral., pt. 1, pp. 156-157. 1891. ‘Die Weehecllicsichunaéan zwischen Pflanzen und Ameisen in den Tropen.’ Ausland, pp. 474-477. 1894. ‘Die Ameisen von Rio Grande do Sul.’ Berlin. Ent. Zeitschr., XXXIKX, pp. 321-446, 1 Pl. 1898. ‘Die Anlage neuer Colonien und Pilzgirten bei Atta sexdens.’ Zool. Anzeiger, X XI, pp. 238-245. ; 1905. ‘A formiga cuyabana.’ Revista Agricola 8. Paulo, XI, pp. 511-522. 1907. ‘Die Cecropien und ihre Schutzameisen.’ Engler’s Bot. Jahrb., XXXIX, pp. 666-714, Pls. vi-x (Review in Zeitschr. Wiss. Insekten- biol., VI, pp. 34-36). 1917. ‘As formigas cuyabanas empregadas como meio de destrucefio das formigas cortadeiras.’ Physis, Buenos Aires, III, pp. 352-360. ILtuincworts, J. F. 1913. ‘Little brown ant doing good work in Hawaii.’ Hawaiian Forester and Agric., X, pp. 370-371. 1916. ‘Coconut leaf-roller [Omioides blackburni (Butl.)] destroyed by ants.’ Proc. Hawaiian Ent. Soc., III, p. 142. 1917. ‘Economic aspects of our predaceous ant (Pheidole megacephala.)’ Proe. Hawaiian Ent. Soc., ITI, pp. 349-368. Inv, J. R. 1907. ‘La plaga de las hormigas y los procedimientos para su destruc- cién.’ Com. Parasitologia Agricola, Mexico, Circ. 68, pp. 1-11, 6 Pls. IstvAnrFI, G. 1894. ‘Gombatenyészté hangfak.’ Termes. Kozl. Magyar Tars., pp. 378-387 (habits of fungus-growing and honey ants). Jack, W. 1825. ‘Account of the Lansium and some other genera of Malayan plants.’ Trans. Linn. Soc. London, XIV, pt. 1, pp. 114-130, PL. rv. Jacosson, E, 1907. ‘Notes on web-spinning ants.’ Victorian Natural., XXTV, pp. 36-38. Wheeler, Ants of the Belgian Congo 555 _ 1908.-‘Zur Verfertigung der Gespinnstnester von Polyrhachis bicolor Sm. ada auf Java (mitgeteilt von E. Wasmann).’ Notes Leyden Mus., ; XXX, pp. 63-66, Pl. vt. Saoaer N. J. 1763. ‘Selectarum stirpium americanarum historia.’ Vienna. é vii+284 pp., 183 Pls. Jae, C 1895. ‘Etudes sur les fourmis. 8° note. Sur l’organe de nettoyage tibio- . tarsien de Myrmica rubra L., race levinodis Nyl.’ Ann. Soc. Ent. _-—s«é France, XLII, (1894), pp. 691-704. 1896. Les fourmis.’ Bull. Soc. Zool. France, XXI, pp. 60-93. i ___- 1899. ‘Sur les corpuscules de nettoyage des fourmis.’ Bull. Soc. Zool. : i France, XXIV, pp. 177-178. ‘ Taxxowsny, R. 1894. ‘Ein neuer Forstschidling.’ Zentralbl. f. d. Ges. Forst- wesen, XX, pp. 431-434. -& Janse, H. 1910. ‘Nepenthes-Tiere. II. Biologische Notizen.’ Ann. Jardin fa ?- Bot. Buitenzorg, 3 Suppl., pp. 941-946. Br bascaead ‘T.C. 1851. ‘A catalogue of the species of ants found in Southern India.’ _ Madras Journ. Litt. Se., XVII, pp. 103-127; Ann. Mag. Nat. Be Hist., (2) XIII, 1854, pp. 45-56, 100-110. Sonnet, E.L. 1912. ‘The western harvesting ants.’ Guide to Nature, V, No.7, Bec. pp. 210-214. a Jones, G. A. 1912. ‘The structure and pollination of the cacao flower.’ West ‘'? Indian Bull., XII, 3, pp. 347-350. “De Sows, T.H. 1915. ‘Sunsets affecting vegetable crops in Porto Rico.’ Bull. U.S. Dept. Agric., No. 192, pp. 1-11. _ 1917. ‘Occurrence of a fungus-growing ant in Louisiana.’ Journ. Econ. Ent., X, p. 561. doom, J. F. anv Nirscue, H. 1885-1895. ‘Lehrbuch der mitteleuropiischen gy Forstinesktenkunde.’ Vienna. I, 1885-1893, pp. 1-736; II, 1893-1895, pp. 737-1421. JuMELLe, H. anv Perrier ve LA Batute, H. 1907a. ‘Les termites champignon- nistes A Madagascar.’ C. R. Ac. Sci. Paris, CXLIV, pp. 1449-1451. 1907b. ‘Les champignons des termitiéres de Madagascar.’ C. R. Ac. Sci. Paris, CXLV, pp. 274-276. 1910. ‘Termites champignonnistes et champignons des termitiéres A Madagascar.’ Rev. Gén. Bot., Paris, XXII, pp. 30-64. > = ee Kavrensacn, J.H. 1843. ‘Monographie der Familien der Pflanzenliuse (Phytoph- thires). I. Die Blatt- und Erdliuse.’ Aachen. xliii+223 pp., Pl. 1. KanneNnperc,—1900. (Eigenartige Ameisenwohnungen.) Mitth. Deutsch. Schutz- geb., XIII, 1, p. 11. Karawatew, V. 1906. ‘Systematisch-biologisches tber drei Ameisen aus Buiten- zorg.’ Zeitschr. Wiss. Insektenbiol., II, pp. 369-376. 1914. ‘Eine neue Weberameise, Polyrhachis armata Le Guillou.’ Biol. ; Centralbl., XXXIV, pp. 440-444. Karsten, G. 1895. ‘Morphologische und biologische Untersuchungen tiber einige Epiphytenformen der Molukken.’ Ann. Jardin Bot. Buitenzorg, XII, pp. 117-195, Pls. x11—xrx. 556 Bulletin American Museum of Natural History (Vol. XLV Karsten, H. 1886. ‘Ameisenpflanzen.’ Flora, LXIX, pp. 304-306. Kavurrmann, H. 1912. ‘Der Adlerfarn in seiner Jugend eine Ameisenpflanze.’ Uns. Welt, Godesberg, IV, pp. 194-199. Keeste, F. 1910. ‘Plant animals, a study in symbiosis.’ Cambridge. viii+163 pp., 1 Pl. ‘ 1917. ‘Plant animals: a study in symbiosis.’ Proc. Roy. Inst. Great Britain, XXI, p. 160. Kewier, C. 1892a. ‘Neue Beobachtungen tiber Symbiose zwischen Ameisen und Akazien.’ Zool. Anzeiger, XV, pp. 137-140. ; 1892b. ‘Ueber die biologischen Verhiltnisse der ostafrikanischen Steppen.” Verh. Schweiz. Naturf. Ges., 75. Jahresvers., pp. 47-48. 1906. ‘Eine Sammlung von Gallen aus dem Mittelmeergebiet.’ Verh. Schweiz. Naturf. Ges., 89. Jahresvers., pp. 76-77. KeRNER VON MariLaun, A. 1876. ‘Die Schutzmittel der Bliten gegen unberufene Giste.’ Festschr. Zool. Bot. Ges. Wien, pp. 189-261. 1878. ‘Flowers and their unbidden guests’ (translated by W. ‘Ogle). London. xvi+164 pp., 3 Pls. 1887-1891. ‘Pflanzenleben.’ Leipzig. I, 1887, x+734 pp., 20 Pls.; II, ¢ 1891, viii+8€6 pp., 20 pls. 1894-1895. ‘The natural history of plants: their forms, growth, reproduc- tion, and distribution.’ Translated by F. W. Oliver. London. Vol. I, 1894, 777 pp.; vol. II, 1895, xi+983 pp. Kersuaw, J.G.C. 1913. ‘Anatomical notes on a membracid.’ Ann. Soc. Ent. Belgique, LVII, pp. 191-201. Kevucuenivus, P. E. 1914a. ‘Enkele beschouwingen over de schildluizen van de koffie.” Meded. Besoekisch Proefstation (Djember, Java), No. 8, pp. 9-19. 1914b. ‘De beteekenis van twee bekende mieren, in verband met het groeneluizenvraagstuk van de koffie.’ Teysmannia, XXV, pp. 711-716. 1915. ‘Het vraagstuk van de gramang-mier (Plagislepis longipes); tevens een kritiek.’ Teysmannia, XXVI, pp. 382-395. 1916. ‘Een repliek.’ Teysmannia, XXVII, pp. 27-32. KILLERMANN, 8S. 1908. ‘Getreidesammelnde Ameisen.’ Naturw. Wochenschr., VU, p. 378. KINDERMANN, V. 1905. ‘Lamium album L., eine myrmekophile Pflanze.’ Lotos, LILI, pp. 339-341. 1907a. ‘Ueber die Verbreitungsmittel der Gelegenheitsepiphyten unserer heimischen Flora.’ Lotos, LV, pp. 25-31. 1907b. ‘Ueber Gelegenheitsepiphyten.’ Aus der Natur, III, pp. 875-377. 1908. ‘Ueber myrmekochore Pflanzen.’ Lotos, LVI, pp. 104-105. Kine, G. B. 1897a. ‘Aphids and coccids associating with ants. An account of their habits with bibliographical notes.’ Ent. News, VIII, pp. 125-129. 1897b. ‘A new ant-nest coccid.’ Psyche, VIII, pp. 150-151. 1898. ‘China asters infested by a coccid.’ Psyche, VIII, p. 312. 1902. ‘Two new ants’ nest coccids from Texas.’ Canadian Ent., XXXIV, pp. 285-286. Kina, G. B. anp Cocxrerett, T. D. A. 1897. ‘New Coccide found associated with ants.’ Canadian Ent., X XTX, pp. 90-93. ie cs ra 1923) Wheeler, Ants of the Belgian Congo 557 Bo, ‘Kine, , G. B. anp-Tinsiey, J.D. 1898. ‘A new ant-nest coccid.’ Psyche, VIII, an pp. 297-298. os Kino, H. 1911. ‘The black garden ant (Aphznogaster barbara L.).’ 4th Rept. Well- ike come Research Lab. Khartoum, B, pp. 142-143. “i Kxavre, F. 1906. ‘Die Ameisen.’ Aus Natur u. Geisteswelt, XCIV, 156 pp. 1908. ‘Die Symbiose der Ambrosiakiifer mit Pilzen.’ Zentralbl. f. d. Ges. Forstwesen, XXXIV, pp. 498-501. a Kenissr, L. 1909. ‘Beitriige zur Trophobiose bei den Ameisen.’ Zeitschr. Wiss. Insektenbiol., V, pp. 76-82. 1 ‘Kxorn, P. 1899. ‘Termiten und ihre Pilzgirten.’ Ilustr. Wochenschr. Ent., IV, pp. 257-259. es i a L. 1887. ‘Die Ameisen im Dienste des Gartenbaues.’ Gartenflora, XXXVI, pp. 370-375. . Koen, €.L. 1854-1857. ‘Die Pflanzenlause, Aphiden’ Narnberg. viii +835 pp., Kascu, K.A. 1908. ‘Die Theorie der Ameisenpflanzen,— ein Irrtum der Biologie.’ Beil. z. Allgem. Zeitung, Miinchen, No. 8, pp. 59-61 (Review in Zeitschr. : Wiss. Insektenbiol., VI, 1910, p. 36). . Kom, J.G. 1779. ‘Naturgeschichte der sogenannten weissen Ameise.’ Beschiift. Ges. Naturf. Fr. Berlin, IV, pp. 1-28, PI. 1. . Kanvicke, M . 1918. ‘Ueber die extrafloralen Nectarien auf den Laubblittern einiger Hibisceen.’ Flora, CXI-CXII, (Festschrift E. Stahl), pp. 526—- 540, Pl. vu, pom H. 1906. ‘Zur Biologie der spinnenden Ameisen.’ Natur u. Offenbarung, LU, pp. 166-169. 1909. ‘Die Ameisenpflanzen des tropischen Afrika mit besonderer Beriick- sichtigung ihrer biologischen Verhiltnisse.’ Natur u. Offenbarung, LV, pp. 89-111 and 148-175. . aad H. J. 1884. ‘Beitrag zur Biologie der Aphididen.’ Berlin. Ent. Zeitschr., XXVIII, pp. 343-345. 1890. ‘Die getreidesammelnden und die ackerbautreibenden Ameisen.’ Naturw. Wochenschr., V, pp. 193-195. Koninospercer, J. C. 1908. ‘Tweede overzicht der schadelijke en nuttige insecten van Java.’ Meded. Dept. Landbouw, Batavia, No. 6, pp. 1- 114. Kornuavser, 8. I. 1919. ‘The sexual characteristics of the membracid Thelia bimaculata (Fabr.). I. External changes induced by Aphelopus thelix Gahan.’ Journ. Morphology, XXXII, pp. 531-636. Krereuin, K. 1913. ‘Die Beziehungen der Tiere und Pflanzen zueinander.’ 2d. Ed., Leipzig. Vol. 1, 107 pp.; vol. IT, 99 pp. Kravse, K. 1912. ‘Rubiacew africane. III.’ Engler’s Bot. Jahrb., XLVIII, 3+, pp. 403-433. Kravsse, A. H. Molla. ‘Ueber Aphznogaster sardoa Mayr.’ Arch. f. Naturg., LXXVII, Bd. 1, Suppl. Heft 1, pp. 39-41. 1911b. ‘Ueber Mezxr strucckr Ltr. und einige andere Ameisen auf Sardinien.’ Bull. Soc. Ent. Italiana, XLI, (1909), pp. 14-18. 1913a. ‘Campcnctus herculeanus vagus Scop. als Korkschidling.’ Arch. f. Naturg., LX XIX, A, Heft 6, pp. 34-35, 2 Pls. 558 Bulletin American Museum of Natural History [Vol. XLV 1913b. ‘Ueber die Beschidigung der Korkeiche durch Cremastogaster scutel- laris Oliv.’ Arch. f. Naturg., LXXIX, A, Heft 1, pp. 56-58, 1 Pl. 1916. ‘Die mechanische Einwirkung von Fcrmica fusca cinerea (Mayr) For. auf Sandboden.’ Naturw. Wochenschr., N. F., XV, pp. 371-373. 1919. ‘Uebersicht tiber die Ameisen Sardiniens nebst einigen biolog- ischen Beobachtungen.’ Zeitschr. Wiss. Insektenbiol., XV, pp. 96-103. Kronre.p, M. 1890. ‘Ueber die kiinstliche Besiedelung einer Pflanze mit Ameisen.’ Tag. Deutsch. Nat. Vers., LXII, (1889), pp. 262. Kuster, E. 1903. ‘Pathologische Pflanzenanatomie.’ Jena. vii+312 pp. 1911. ‘Die Gallen der Pflanzen. Ein Lehrbuch fiir Botaniker und Entomo- logen.’ Leipzig. x+437 pp. Kuntze, O. 1877. ‘Die Schutzmittel der Pflanzen gegen Thiere und Wetterun- gunst.’ Beilage zur Botan. Zeitung, XX XV, pp. 1-91. La Baume, W. 1907. ‘Wie entstent die Kolonie der Blattschneiderameise?’ Aus der Natur, III, pp. 158-159. 1912. ‘Myrmecophile Zikaden.’ Prometheus, XXIII, pp. 21-22. Lasse, M. 1921. ‘Les fourmis et |’élevage des faisandeaux en Tunisie.’ Bull. Soc. Nat. Acclimat. France, LX VIII, pp. 20-21. Lacxowitz, W. 1888. ‘Ameisen in Dienste des Gartenbaues.’ Humboldt, VII, p. 157. Laceruem, G. 1900. ‘Ueber Lasius fuliginosus (Latr.) und seine Pilzzucht.’ Ent. Tidskr., XXI, pp. 17-29. Lavoy, L. 1906. ‘Parasitisme et mutualisme dans la nature.’ Paris. 284 pp. 1909. ‘Les jardins de termites.’ Le Naturaliste, XXX, pp. 57-58. 1910a. ‘Les fourmis moissonneuses.’ Le Naturaliste, XXXII, pp. 107-108. 19106. ‘La myrmécophilie des Cecropia.’ Le Naturaliste, XXXII, p. 155. Lamporn, W.E. 1914. ‘On the relationship between certain West African insects, especially ants, Lycenide# and Homoptera.’ Trans. Ent. Soc. London, (1913), pt. 3, pp. 436-498. LaMeereE, A. 1902. ‘Note sur les mceurs des fourmis du Sahara.’ Ann. Soc. Ent. Belgique, XLVI, pp. 160-169. Lanog, D. pe. 1910. ‘Le réle des fourmis dans la lutte contre la punaise du cacaoyer A Java.’ Journ. Agric. Tropicale, X, p. 284. Le Cerr, F. 1914. ‘Chenilles et cocons recueillis dans les galles de Formicides.’ Voy. Alluaud et Jeannel Afr. Orient., Lépidoptéres, I, pp. 1-32, 2 Pls. 1914. ‘Sur une chenille de Lycénide élevée dans les galles d’ Acacia par des fournisdu genre Cremastogaster.’ C. R. Ac. Sci. Paris, CLVIII, pp. 1127-1129. Levesur, A. 1901. ‘Ueber Pilzeziichtende Ameisen.’ Nerthus, Altona, III, pp. 411-414, 422-425. Leermans, 8. 1906. ‘Een mierenherberg.’ De Levende Natuur, XI, pp. 130-131. Lerpy, J. 1877a. ‘Remarks on ants.’ Proc. Acad. Nat. Sci. Philadelphia, XXIX, pp. 304-305. 1877b. ‘Remarks on the yellow ant.’ Proc. Acad. Nat. Sci. Philadelphia, XXIX, p. 145. 1882. ‘The yellow ant with its flocks of Aphis and Coccus.’ Proc. Acad. Nat. Sci. Philadelphia, XXXIV, pp. 148-149. SS es Be ee ren lee 559 4884. Ant infected with a fungus.’ Proc. Acad. Nat. Sci. Philadelphia, oO XXXVI, p. 9. ‘Leney, A. 1893. ‘Le chéne-lidge, sa culture et son exploitation? Paris and Nancy. Lxoxan, P. 1911. ‘The honey-ants of Point Loma, California.’ Trans. San Diego Soc. Nat. Hist., I, No. 3, pp. 85-97. 1912. ‘The emmet harvesters of Point Loma.’ The Thedeophical Path, II, No. 3, pp. 215-216. - 1914a. ‘The living honey-jars of Lomaland.’ Raja-Yoga Messenger, July, pp. 19-20. -1914b. ‘A marvel of motherhood: a record of observations on the founding _of a colony of honey ants (Myrmecocystus mexicana).’ The Theosophical . ‘Les relations des fourmis avec les Hémiptéres Homoptéres de la famille des Fulgorides. Domestication des Tettigometra.’ C. R. Soc. «Biol. Paris, LVIII, pp. 1005-1007; Bull. Soc. Ent. France, pp. 161-164. * “Tawis, 1. 1912. ‘A few notes on Sclenc psis geminata.’ Proc. Hawaiian Ent. Soc., ee _ I, No. 14, pp. 175-178. o-? Lacurexsre, W.A.J. 1870. (Relations of Tettigometra to ants.) © Petites Nouv. ee ee Entom., No. 19, p. 74. ia 2 ‘ 1877. ‘Quelques feuillets de mon journal.’ Mitth. Schweiz. Ent. Ges., V, a aid No. 5, pp. 297-303. 1880. (Relations between root-aphids and Lasius.) Ann. Soc. Ent. France, (5) X, Bull. pp. ciii-—ev. 1885. ‘Les Pucerons. Monographie des Aphidiens.’ I. Montpellier. 188 Lincecum, G. 1862. ‘Notice on the habits of the “agricultural ants” of Texas (“stinging ant”’ or “mound-making ant’’), Myrmica (Atta) molefaciens ‘ Buckley.’ Communicated by Charles Darwin. Journ. Proc. Linn. Soe. London, ., VI, pp. 29-31. 1866. ‘On the agriceliaral ant of Texas (Myrmica mclefaciens).’ Proc. Acad. Nat. Sci. Philadelphia, XVIII, pp. 323-331. 1867. ‘The cutting ant of Texas. (codoma terana Buckley.’ Proc. Acad. Nat. Sci. Philadelphia, XTX, pp. 24-31. 1874. ‘The agricultural ant.’ American Naturalist, VIII, pp. 513-518. Lanpincer, L. 1906. ‘Verbreitung der Corydalis sclida durch Ameisen.’ Mitt. Bayer. Bot. Ges., XX XIX, pp. 518-519. Livspaver, L. 1897. ‘Ueber Ameisenpflanzen.’ Verh. Zool. Bot. Ges. Wien, XLVII, pp. 45-46; Bot. Centralbl., LXX, p. 12. 1913. ‘Pflanzen und Ameisen.’ Carinthia, II, Jahrg. 103, pp. 221-223. Luave, Pasto pe LA. 1832. ‘Sobre las busileras u hormigas de miel.’ Registro Trimestre 6 Colleccion de Memorias de Historia, Literatura, Ciencias y Artes, Mexico, I, No. 4, Oct., pp. 455-463. 1961, ‘Sur les busileras ou founnis A miel.’ Rev. Mag. Zool., (2) XIII, pp. 457-462. Lion, C. G. 1919a. ‘Mycological notes.’ Cincinnati. No. 59, pp. 845-860. 1919). ‘Mycological notes.’ Cincinnati. No. 61, pp. 877-903, Pls. CXXIV-CXXXIX. 1920. ‘Mycological notes.’ Cincinnati. No. 62, pp. 904-044, Pls. cxt- cLIiv. B 560 Bulletin American Museum of Natural History (Vol. XLV Lock, R.C. 1904. ‘Ecological notes on Turnera ulmifolia L. var. elegans Urban.’ Ann. Roy. Botan. Gard. Peradeniya, I, 1, pp. 107-119. Lércren, A. 1905. ‘As formigas cuyabanas.’ Bol. da Agricultura, 8. Paulo, (6) V, p. 218. Law, O. 1874. ‘The honey ants.’ American Naturalist, VIL, pp. 365-366. Lomparpi, D. 1912. ‘Alcune osservazioni morfologiche e biologiche intorno alla Forda formicaria Heyden.’ Atti Acc. Lincei Roma, Rend., (5) XXI Sem. 1, pp. 809-814. Lussock, J. 1880. ‘Observations on ants, bees, and wasps; with a description of a new species of honey-ant. Part VII. Ants.’ Journ. Linn. Soc. London, Zool., XV, pp. 107-187, Pl. vi. 1883. ‘Observations on ants, bees, and wasps. Part X. With description of a new species of honey-ant.’ Journ. Linn. Soc. London, Zool., XVII, pp. 41-52, Pl. m. 1888. ‘Observations on ants, bees, and wasps. Part XI.’ Journ. Linn. Soc. London, Zool., XX, pp. 118-136, Pl. Lucas, H. 1855. ‘Note sur le Myrmececystus mexicanus.’ Ann. Soc. Ent. France, (3) ILL, pp. 54-55. 1858. ‘Sur un Hyménoptére qui cause des dégits dans un magasin de chocolat.’ Ann. Soc. Ent. France, (3) VI, Bull., p. Ixxxi. 1860. ‘Observations sur les busileras ou fourmis & mniel du Mexique (Myrme- cocystus melligerus).’ Rev. Mag. Zool., (2) XII, pp. 271-280. 1873. ‘Note relative aux Alta barbara et structor, Hyménoptéres de la tribu des Formicides.’ Ann. Soc. Ent. France, (5) III, Bull., pp. clxxx— clxxxi. Lupwic, F. 188%a. ‘Weitere Untersuchungen itiber Ameisenpflanzen.’ Biol. Centralbl., VIL, pp. 577-580. 1888), (Hchimeer; A. F. W. Die Wechselbeziehungen zwischen Pflanzen und Ameisen im tropischen Amerika (review).’ Biol. Centralbl., VII, pp. 321-330. 1889. ‘Extranuptiale Saftmale bei Ameisenpflanzen.’ Humboldt, VIII, pp. 294-297. 1898. ‘Biologische Beobachtungen an Helleborus fetidus.’ C&sterreich. Bot. Zeitschr., XLVIII, pp. 281-284. 1899. ‘Die Ameisen im Dienste der Pflanzenverbreitung.’ Illustr. Wochen- schr. Ent., IV, pp. 38-41. 1902. ‘Insekten- und pflanzenbiologische Beitrige. Zur Karpobiologie von Leucojum vernum L.’ Allgem. Zeitschr. Ent., VIII, p. 451. 1907. ‘Weiteres zur Biologie des Helleborus fetidus.’ Zeitschr. Wiss. Insektenbiol., III, pp. 45-50. 1908. ‘Neuere Beobachtungen iiber Pilze ziichtende Insekten.’ Prome- theus, XIX, pp. 373-374. 1909. ‘Potentilla fragariastrum, eine neue Ameisenpflanze mit postfloraler Blattvergrésserung.’ Aus der Natur, V, pp. 254-256. Liverwapt, H. 1909. ‘Beobachtungen tiber die Lebensweise von Camponotus rufipes F.’ Zeitschr. Wiss. Insektenbiol., V, pp. 226-229, 269-271, 305-312. eta Wheeler, Ants of the Belgian Congo 561 . a 1915. ‘Insekten- und sonstiges Tierleben an brasilianischen Bromeliaceen.’ _-—sé«Feittsschr. Wiisss.. Insektenbiol., IX, pp. 78-84. s 1918. ‘Notas myrmecologicas.’ Rev. Mus. Paulista, X, pp. 31-64, Pl. ; Lom, M. 1911. ‘Ueber myrmekophile Pflanzen und die siz bewohnenden Ameisen.’ ze ___ Schrift. Phys.-Okon. Ges. Kénigsberg, LII, pp. 232-233. “Lown, A. W. 1831. ‘Lettre sur les habitudes de quelques fourmis du Brésil, ad- 2 et dressée 4 M. Audouin.’ Ann. Sc. Nat., XXIII, pp. 113-138. _Texoarnd A. N. 1887. ‘Pflanzenbiologische Studien. II. Die mye der Pflanzen an Thiere.’ Nova Acta Soc. Sci. Upsaliensis, (3) XIIT No. 10, 88 pp., 4 Pls. = MeCoxsoc, J. W. anp Hayes, W. P. 1916. ‘A preliminary report on the life 4 a economy of Solenopsis molesta Say.’ Journ. Econ. Ent., [X, pp. 23-38. | Motoor, H.C. 1877. ‘The agricultural ants of Texas.’ Proc. Acad. Nat. Sel. ___— Philadelphia, XXTX, pp. 299-304. 1879a. ‘The natural history of the agricultural ant of Texas. A monograph Re ae of the habits, architecture and structure of Pogonomyrmex barbatus.’ Gale es Philadelphia. 310 pp., 24 Pls. 18796. ‘Cutting or parasol ant. Atta fervens Say.’ Proc. Acad. Nat. Sci. 1879c. ‘On the architecture and habits of the cutting ant of Texas (Atta 9% fervens).’ Ann. Mag. Nat. Hist., (5) II, pp. 442-449. «879d. ‘On Myrmecocystus mexicanus Wesm.’ Proc. Acad. Nat. Sci. Philadelphia, pp. 197-198. 1880. ‘Note on a new northern cutting ant, Atta septentrionalis.’ Proc. Acad. Nat. Sci. Philadelphia, pp. 359-360. 1881. ‘The honey ants of the Garden of the Gods.’ Proc. Acad. Nat. Sci. Philadelphia, XX XIII, pp. 17-77, Pls. 1—x. 1882a. ‘The honey ants of the Garden of the Gods and the Occident ants of the American plains.’ Philadelphia. 188 pp., 13 Pls. 1882b. ‘Ants as beneficial insecticides.’ Proc. Acad. Nat. Sci. Philadelphia, pp. 263-271. 1906. ‘Honey ants of the Garden of the Gods.’ " Blasper’ s Monthly, pp. 126- 133. 1909. ‘Ant communities and how they are governed.’ NewYork. 321 pp. _ Macerra, G. 1903. Insectos dafiosos al aleornoque en Estramadura y Castilla la vieja. Madrid. 46 pp. (Review in Allgem. Forst- und Jagdzeitung, 1904, p. 397). Macer, L. 1874. ‘Sopra un nido singolare della Formica fuliginosa Latr.’ Atti Soc. Italiana Se. Nat., Milano, XVII, pp. 64-98, Pls. 11—v1. Maontn, A. 1905. ‘Les plantes myrmécophiles d’aprés les travaux de Ule.’ Mém. Soc. Emul. Doubs, (7) X, p. xiii. Macowan, —. 1882. ‘Utilization of ants as grub-destroyers in China.’ North- China Herald, April 4. Mancer, —. 1919. ‘Ein Ueberfall der Rossameise auf ein Bienvolk.’ Bayrische Bienenzeitung, Munich, p. 219 (1920, Arch. f. Bienenkunde, II, 3—4, pp. 73-74). 562 Bulletin American Museum of Natural History [Vol. XLV Mann, W.M. 1915. ‘Some myrmecophilous insects from Hayti.’ Psyche, XXTI, pp. 161-166. 1919. ‘The ants of the British Solomon Islands.’ Bull, Mus. Comp. Zoél. Cambridge, Mass., LXIII, No. 7, pp. 271-391, Pls. 1-11. 1921. ‘The ants of the Fiji Islands.’ Bull. Mus. Comp. Zoél. Cambridge, Mass., LXIV, No. 5, pp. 401-499. Marceravius bE Ligerstap, G. 1648. ‘Historia rerum naturalium Brasili# libri octo.’ Amsterdam. viii+300 pp. Marcuat, P. 1909a. ‘Sur les cochenilles de l’ Afrique Occidentale.’ C. R. Soc. Biol. Paris, LXVI, pp. 586-588. 1909). ‘Contribution A l'étude des Coccides de l'Afrique Occidentale.’ Mém. Soc. Zool. France, XXII, pp. 165-182, 2 Pls. Marcuat, P. anp Poutters, R. 1920. ‘La fourmi d’Argentine.’ La Vie Agric. et Rurale, Paris, XVI, p. 351; Bull. Agric. Algérie, XXVI, pp. 215-217. Mar.artt, C. L. 1896. ‘Cockroaches and house ants.’ Bull. 4 U. 8. Dept. Agric. Bur. Ent., N.8., pp. 84-99. Marsa, H.O. 1910. ‘Notes on a Colorado ant (Fcrmica cinereorufibarbis Forel) .’ Bull, 64 U.S. Dept. Agric. Bur. Ent., N.8., pt. 9, pp. 73-78. Martins, M. N. 1907. ‘Une fourmi terrible envahissant l'Europe.’ Broteria, Ser. Zool., VI, pp. 101-102. Massarr, J. 1904. ‘La guerre et les alliances entre animaux et végétaux.’ Brussels. 28 pp. 1906. ‘Les fourmis-jardiniéres.’ Bull. Soc. Bot. Belgique, XLIII, 2, pp. 100-103. Maske.i, W. W. 1891. ‘Further coccid notes; with descriptions of new species and remarks on coccids from New Zealand, Australia, and elsewhere.’ Trans. New Zealand Inst., XXIV, pp. 1-64. Massezr, G. 1891. ‘Anew Cordyceps.’ Ann. of Botany, V, No. 20, pp. 510-511. 1893.. ‘Fungus-gardens of some South American ants.’ Grevillea, XXII, No. 102, pp. 48-49. 1895. ‘A revision of the genus Cordyceps.’ Ann. of Botany, TX, pp. 1-4, Pls. 1-11. 1898-1899. ‘Révision du genre Cordyceps.’ Revue Mycologique, XX, 1898, pp. 49-57, 85-94; XXI, 1899, pp 1-16; Pls. cLxxvill, CLXxrx, and CLXXxill. 1899. ‘Fungiexotici. II.’ Bull. Misc. Inform. Bot. Gard. Kew, Nos. 153— 154, pp. 164-185. Massias, O. 1901. ‘Schomburgkia tibicini: Batem.’ Die Gartenwelt, VI, pp. 9-10. Marrrro.o, O. 1896. ‘Sopra alecune larve micofaghe.’ Bull. Soc. Bot. Italiana, pp. 180-183. ‘ Marrte!, G. E. 1909. ‘Acacie africane a spine mirmecodiate.’ Boll. Orto Bot. Palermo, VIII, pp. 131-133. Marrer, G. E. anp Tropes, C. 1908. ‘Graminacee provviste di nettarii estranuziali, Nota prevent.’ Boll. Orto Bot. Palermo, VII, pp. 113-117. 1909. ‘Ricerche e studt sul genere Eragrostis in rapporto ai nettarii estra- nuziali.’ Contrib. Biol. Veget. Palermo, IV, 2, pp. 205-286, 1 Pl. Maxwe.t-Lerroy, H. anp Howrerr, F. M. 1909. ‘Indian insect life. A manual of the insects of the plains (tropical India.)’ Caleutta. xii+-786 pp., 84 Pls. (ants, pp. 224-233). Wheeler, Ants of the Belgian Congo 563 Mayr, -G. 1862. er tetials Studien.’ Verh. Zool. Bot. Ges. Wien, XII, i- pp. 649-776, Pl. xx. 1907. ‘Formicide.’ In Sjéstedt, Exped. Kilimandjaro, Meru, etc., II, Abt. 8, 2, pp. 7-23, 1 Pl. Marner, F. 1860. ‘Bidrag til de danske Myrers Naturhistorie.’ Danshia Videusk: Selsk. Skrift., naturv. math. Afd., (5) V, pp. 275-340. 1892. ‘Trek af Insektlivet i i Venesuela.’ Ent. Medd. Kjébenhavn, III, Ne pp. 125-166. as Maxon G. anv Savi, P. 1844. ‘Sulle appendici apiculari proprie alle foglioline dell’ Acacia cornigera.’ Giorn. Bot. Italiano, I, pp. 106-112. | Mancer, R. G. 1910. ‘Relaciones entre las plantas y los insectos.’ Asoc. Espaii. F ' Se. Nat. Zaragoza, IV, Sec. 3, pp. 365-375. | Menctax,L. 1911. ‘Sur le réle des Insectes comme agents de propagation de | ergot xf des Graminées.’ C. R. Soc. Biol. Paris, LXX, pp. 300-302. | Mem, E. D. 1917. ‘An interpretation of Rumphius’s Herbarium Amboinense.’ "ee Philippine Bur. of Sci. Publ. No. 9, 596 pp., 2 Maps. 1920. ‘Myrmeconauclea, a new genus of rubiaceous plants from Palawan Ae cae! and Borneo.’ Philippine Journ. Sc., XVII, No. 4, pp. 375-376. 2 Mr, C. 1888a. ‘Morphologische Studien iiber die Familie der Lauraceen.’ Verh. m5 Bot. Ver. Brandenburg, XXX, pp. 1-31. 1888b. ‘Myrmecophilie der Lauraceen-Gattung Pleurothyrium.’ Verh. Bot. Ver. Brandenburg, X XIX, (1887), pp. xxiii-xxiv. 1889. ‘Lauracew americane monographice descripte.’ Jahrb. Bot. Gart. ho Berlin, V, pp. 1-5£6, Pls. au. 1890. ‘Morphologische und anatomische Studien iiber die Gruppe der Cordiex.’ Engler’s Bot. Jahrb., XII, pp. 526-588, Pls. rv—v. “Musne, H. 191la. ‘Ueber die javanische Myrmeccdia und die Beziehung zu ihren 3 Ameisen.’ Biol. Centralbl., XX XI, pp. 733-738. Ss 1911). ‘Javanische Studien. I. Untersuchungen iiber die javanische o- Myrmecodia.’ Abhandl. Math. Phys. Kl. Siichs. Ges. Wiss. Leipzig, XXXIHI, No. 4, pp. 312-361. 1912. ‘Ameisenpflanzen.’ Handwérterbuch der Naturwissenschaften, I, pp. 255-265. ere, t. 1910. ‘Sull’impollinazione di Rohdea japonica Roth per mezzo delle formiche.’ Ann. di Bot., VIII, pp. 241-242. -cneesaada A.W. 1838. ‘Qbeérvations sur le canal médullaire et les diaphragmes du trone de Cecropia palmata L.’ Bull. Sci. Phys. Nat. Néerlande, pp. 29-31, 168-172. Miséperc, E. 1906. ‘Ueber Systellonotus triguttatus L. und sein Verhiltnis zu Lasius niger.’ Zeitschr. Wiss. Insektenbiol., II, pp. 107-109. Méuier, A. 1893. ‘Die Pilzgirten einiger sidamerikanischer Ameisen.’ Botan. Mitt. aus den Tropen, Jena, Heft 6, vi+-127 pp., 7 Pls. 1901. ‘Phycomyceten und Ascomyceten. Untersuchungen aus Brasilien.’ Botan. Mitt. aus den Tropen, Jena, Heft 9, xii+319 pp., 11 Pls. Mawes, — 1889. ‘Symbiose zwischen Pflanzen und Ameisen im tropischen Amerika.’ : Humboldt, VII, pp. 456-459. ___- Moaeriwer, J.T. 1872. (Habits of certain species of ants belonging to the genus =. Aphenogaster.) Trans. Ent. Soc. London, (1871), Proc., pp. xlvii- xlviii. 564 Bulletin American Museum of Natural History (Vol. XLV 1873. ‘Harvesting ants and trapdoor spiders. Notes and observations on their habits and dwellings.’ London. xi+1£6 pp., 12 Pls. Mott, F. 1912. ‘Die ZerstSrung des Bauholzes durch Tiere und der Schutz dagegen.’ Naturw. Zeitsehr. f. Forst- u. Landwirtsch., X, pp. 487-497, 518-532. Mo .uiarp, M. 1902. ‘ Basieperium gallarum n. g., n. sp.’ Bull. Soc. Mycol. France, XVIII, pp. 167-170, Pl. rx. Moore, H. W. B. 1913. ‘The planters insect friends.’ Timehri, Journ. Agric. Brit. Guiana, (3) III, pp. 33-42. Morpwixo, A. 1907. ‘Die Ameisen und Blattliuse in ihren gegenseitigen Bezieh- ungen und das Zusammenleben von Lebewesen iiberhaupt.’ Biol. Centralbl., XXVII, pp. 212-224, 233-252. Moreno, A. 1900. ‘Observaciones acerca de las costumbres de las hormigas.’ Mem. Rev. Soc. Cient. Antonio Alzate, XTV, pp. 60-62. Morrnt, F. 1886. ‘Contributo all’anatomia ed alla fisiologia dei nettariiestranuziali.’ Mem. Accad. Sci. Bologna, (4) VII, pp. 325-391, Pls. 1-v1. Morris, G. K. 1880a. ‘Harvesting ants in New Jersey.’ Amer. Entom., III, pp. 228-229. 1880b. ‘A new leaf-cutting ant in New Jersey.’ Amer. Entom., III, pp. 264-265. 1880c. ‘A new harvesting ant.’ American Naturalist, XTV, pp. 669-670. 1881. ‘A new leaf-cutting ant.’ American Naturalist, XV, pp. 100-102. Morstartr, H. 1910. ‘Mittel gegen die Treiberameisen (“Siafu’’).’ Pflanzer, VI, p. 105. 1914. ‘Die Schidlinge der Baumwolle in Deutsch-Ostafrika,’ Pflanzer, X, Beiheft, p. 2. 1920. ‘Ueber einige Ergebnisse der Termitenforschung.’ Biol. Zentralbl., XLIX, pp. 415-427. Morteo, E. 1904. ‘Sopra due piante formicarie, Humboldtia laurifolia L. e Tri- plaris americana Vahl.’ Malpighia, XVIII, pp. 504-511, Pls. x—x1. Morton, F. 1912. ‘Die Bedeutung der Ameisen fiir die Verbreitung der Pflanzen- samen.’ Mitt. Naturw. Ver. Univ. Wien, pp. 77-85, 89-100 and 101— 112. Mose.ey, H.N. 1879. ‘Notes by a naturalist on the “Challenger”.’ London. xvi+ 620 pp. : : Mittier, Frirz. 1874. ‘The habits of various insects.’ Nature, X, pp. 102-103. 1876a. ‘On Brazil kitchen middens, habits of ants, etc.’ Nature, XII, pp. 304-305. 1876b. ‘Ueber das Haarkissen am Blattstiel der Imbauba (Cecropia), das Gemiisebeet der Imbauba-Ameise.’ Jenaische Zeitschr. Naturwiss., X, pp. 281-286. 1880-1881. ‘Die Imbauba und ihre Beschiitzer.’ Kosmos, VIII, pp. 109- 116. 1883a. (Article on Atta.) Blumenauer Zeitung. 1883b. (On Cecropia.) Blumenauer Zeitung, January. 1888. ‘Zur Frage der myrmecophilen Pflanzen.’ Monatl. Mitth. Gesammt- geb. Naturwiss., V, pp. 113-114. Miier-Tuurcav, H. 1892. ‘Die Ameisen an den Obstbiumen.’ Zeitschr. f. Pflanzenkrankh., II, pp. 134-135. a _ Wheeler, Ants of the Belgian Congo 565 OE a a— Narauer, G. von. 1913. ‘Die Ameisen der Grunewaldmoore.’ Naturw. Wochensehr., ai XXVIII, pp. 491-492. 1915. ‘Ueber Konvergenzen im Leben der Ameisen und Termiten.’ ___—_—— Heitschr. Wiss. Insektenbiol., XI, pp. 161-165. ‘4s, L. 1912. ‘Biologische Beobachtungen.’ Ent. Mitt. Deutsch. Ent. Mus. F. W. 1908a. ‘Die Pilzkulturen der Nutzholzborkenkifer.’ Centralbl. _-—~—~—*é«&@Badkkttler.. Pasiitenk., Abt. 2, XX, pp. 279-282. _-—«*-1908b. ‘Die pilzziichtenden Bostrychiden.’ Naturw. Zeitschr. f. Forst- u. Bm tS Landwirtsch., Stuttgart, VI, pp. 274-280. --—-*:1908e. ‘Ueber Ambrosiakifer.’ Aus der Natur, IV, pp. 321-330. --1908d. ‘Ambrosiapilze.’ Ber. Deutsch. Bot. Ges., XX VIa, pp. 735-754, Pl. xm. --—s«;4909a. ‘Ueber Ambrosiagallen.’ Natur und Kultur, VI, 6 pp. -—-1909b. ‘Ambrosiapilze. II. Die Ambrosia der Holzbohrkifer.’ Ber. ll Deutsch. Bot. Ges., X XVII, pp. 372-389. - 1910a. ‘Neue Beobachtungen an kérnersammelnden Ameisen.’ _ Biol. “ARSE Centralbl., XXX, No. 4, pp. 138-150. ‘ -—-4910b. ‘Ambrosiapilze. III. Weitere Beobachtungen an Ambrosiagallen.’ : Ber. Deutsch. Bot. Ges., XXVIII, p. 455-480, Pl. xrv. 1910c. ‘Tiere als Pflanzenziichter.’ Kosmos, Stuttgart, VII, pp. 298-301. ___-:1910d. ‘Neues iiber die kérnersammelnden Ameisen der Mittelmeerlinder.’ MSE, Aus der Natur, V, pp. 737-744. _—-:4911a. ‘Ueber pilzziichtende Gallmiicken (Ambrosiagallen).’ Aus der Natur, VII, pp. 97-106. 1911b. ‘Ambrosiapilze. IV. Tropische Ambrosiapilze.’ Ber. Deutsch. Bot. Ges., X XIX, pp. 50-58. 191le. ‘Zur Uebertragung des Ambrosiapilzes von Xyleborus dispar F.’ Naturw. Zeitschr. f. Forst- u. Landwirtsch., 1X, pp. 223-225. 1913. ‘Biologie der Pflanzen auf experimenteller Grundlage (Bionomie).’ Stuttgart. xxix+775 pp. (Myrmecophilie, pp. 516-533; Myrme- kochorie, pp. 716-720; Tiere als Pilzziichter, pp. 490-516). 1914. ‘Zur Frage der systematischen Stellung der sog. Ambrosiapilze.’ pe bares Centralbl. Bakter. Parasitenk., Abt. 2, XLII, pp. 45-49. _ Nenreune,H. 1884. (On Alta terana.) Zoolog. Garten, XXV, p. 265. _ Newer, W. 1907. “The New Orleans Ant.’ Cire. 13 Louisiana State Crop Pest Com., p. 9. 19082. ‘Notes on the habits of the Argentine or “New Orleans” ant, Iridomyrmex humilis Mayr.’ Journ. Econ. Ent., I, pp. 21-34 and 152. 1908). ‘The Argentine or “ New Orleans” ant.’ 2nd. Bien. Rep. Louisiana State Crop Pest Com. (1906-07), pp. 7-9. 1908e. ‘Two interesting inquilines occurring in the nests of the Argentine ant.’ Journ. Econ. Ent., I, 1908, pp. 262-265. 1909a. ‘The life history of the Argentine ant, Jridomyrmex humilis Mayr.’ Journ. Econ. Ent., I, pp. 174-192. 1909). ‘Measures suggested against the Argentine ant as a household pest.’ Journ. Econ. Ent., I, pp. 324-332, Pl. vii. 1914. ‘A natural enemy of the Argentine ant.’ Journ. Econ. Ent., VII, p. 147. 566 Bulletin American Museum of Natural History (Vol. XLV Newe.t, W. anp Barner, B.S. 1913. ‘The Argentine ant.’ Bull. 122 U.S. Dept. Agric. Bur. Ent., pp. 1-98, Pls. 1—x11. NEWELL, W. anp Rosenrevp, A. H. 1909. ‘Some common insects injurious to truck crops.’ Cire. 27 Louisiana Crop Pest Com., pp. 93-131. NewsTeap, R. 1892-1893. ‘On new or little known Coccida, chiefly Rnglish. (No. 2 and No. 3).’ Ent. Monthly Mag., XXVIII, 1892, pp. 141-148; XXIX, 1893, pp. 77-79. 1893a. ‘Paracletus cimiciformis in ants’ nests in North Wales.’ Ent. Monthly Mag., XXIX, p. 115. 1893b. ‘A new coccid in an ants’ nest.’ Ent. Monthly Mag., XXIX, p. 138. 1893c. ‘Observations on Coccide.’ Ent. Monthly Mag., XXIX, pp. 205- 210. 1900. ‘Observations on Coccide. No. 18.’ Ent. Monthly Mag., XXXVI, pp. 247-251. 1901-1902. ‘Monograph of the Coccide of the British Isles.’ London. Vol. I, 1901, xii+220 pp., Pls. A-E and 1-xxxiv; vol. I, 1903, viii+ 270 pp., Pls. F and xxxv—Lxxv. 1910a. ‘On two new species of African Coccide.’ Journ. Econ. Biol., London, V, pp. 18-22. 1910b. ‘Some insect pests affecting cultivated plants in the West Indies.’ Journ. Roy. Hortic. Soc. London, XXXVI, Not. 53-63. 1914. ‘Homoptera (Psyllide and Coccidz) collected in the Lagos district by W. A. Lamborn.’ Trans. Ent, Soc. London, (1913), pp. 520-524, Pl. xxix. 1920. ‘Observations on scale-insects (Coccidw). VI.’ Bull. Ent. Research, X, pt. 2, pp. 175-207, Pl. xvi. Nickes, L.J. 1911. ‘Field work in the control of the Argentine ant.’ Journ. Econ. Ent., IV, pp. 353-358. Nieuwenuvuis von Uxxiit1-Giitpenpranpt, Mrs. M. 1906. ‘De schadelijke gevolgen der suikerafscheiding bij eenige myrmecophiele planten.’ Versl. Vergad. Wis.- Naturk. Afd. Ak. Wet. Amsterdam, XV, pp. 69- 75. 1907. ‘Extraflorale Zuckerausscheidungen und Ameisenschutz.’ Ann. Jardin Bot. Buitenzorg, XXI, pp. 195-328, Pls. xx—xxrx. Niezasitowskl, E. L. 1911. (Production of earth-mounds by ants in meadows.) Kosmos, Lemberg, XX XV, pp. 159-163. Nitison, A. 1890. ‘Vaxter och myror.’ ’ Arsbok Svenska Vet. Ak. Norton, E. 1868. ‘Notes on Mexican ants.’ American Naturalist, IT, pp. 57-72, Pl. u. NyLanvER, W. 1869. ‘Observationes circa Pezizas Fenniz.’ Notis. Siallsk. pro Fauna et Flora Fenn. Férhandl., X, pp. 1-97, Pls. 1-11; published as a separate in 1868. Outers, H. 1920. ‘Om Udnyddelser af Myrer.’ Haven, Copenhagen, pp. 65-68. Ona, R. pe. 1916, ‘Municipal control of the Argentine ant.’ Journ. Econ. Ent., LX, pp. 468-472. Ono, K. 1907. ‘Studies on some extranuptial nectaries.’ Journ. Coll. Sci. Imp. Univ. Tokyo, XXIII, 3, 28 pp., 3 Pls. Wheeler, Ants of the Belgian Congo 567 ~ Osren-Sacken; “R-V. 1862. ‘Entomologische Notizen. VII. Stallfiitternde GES Ameisen.’ Stettin. Ent. Zeitg., XXIII, pp. 127-128. 1882. “Ants and aphides.’ Psyche, III, p. 343. : Pe 1905. ‘Fourmis bitiseuses, fileuses et agronomes.’ Rev. Scientif., (5) IV. ’ 310-312. a » Pacensrecne, H. A. 186la. ‘Ueber Honig-producirende Ameisen.’ Froriep’s 2 . Notiz., I, 13, pp. 194-197. a — 1861b. ‘Ueber Myrmecocystus mericanus.’ Heidelberg Jahrb. Litt., III, g _————spp. 172-175; Verh. Nat. Med. Ver. Heidelberg, (1859-60), pp. 78-82. - Passanor, L. 1904. ‘Die Kalahari.’ Berlin. xvi+-822 pp., 3 Pls.; Atlas, 10 Pls., eS PaTourLiarn, N. 1892. ‘Une Clavariée entomogéne (Hirsutella entomophila).’ Fede Revue Mycologique, XIV , p. 67. - arto, W. H. 1879. ‘A gall-inhabiting ant.’ American Naturalist, XIII, pp. 126- 5. 127. Pax, F. 1895. ‘Euphorbiacez africane.’ Engler’s Bot. Jahrb., XIX, pp. 76-126. all 1912. ‘Euphorbiacew-Geloniee.’ Das Pflanzenreich. Regni vegetabilis . conspectus, IV, 147, 4, (pt. 52), pp. 1-41. 1914. ‘Euphorbiacew-Acalyphex-Mercurialine.’ Das Pflanzenreich. Regni vegetabilis conspectus, IV, 147, 7, (pt. 63), 473 pp. 1915. ‘Beobachtungen iiber das Auftreten der “argentinischen Ameise,”’ Tridomyrmex humilis Mayr, in Schlesien.’ Ill. Schles. Monatschr. : Obst-, Gemiise- u. Gartenbau, IV, p. 33. Pearson, H. H. W. 1902. ‘On some species of Dischidia with double pitchers.’ Journ. Linn. Soc. London, Botany, XX XV, No. 245, pp. 375-390, Pl. rx. 1903. ‘The double pitchers of Di.chidia Shelfordii, sp. nov.’ Ann. of Botany, XVII, pp. 617-618. Pecxout, T. anp Pecxout, G. 1888-1889. ‘Historia das plantas medicinaes e uteis do Brazil.’ Rio de Janeiro. 1370 pp. (Cecropia, pp. 858-869, fase. 5, 1893). Pienix. J. 1917. ‘Notes sur les Aphides radicicoles.’ Ann. Service des Epiphyties, Paris, IV (1915), pp. 277-285. Penzic,O. 1886. ‘O. Beccari’s neuere Arbeiten tiber die myrmekophilen Pflanzen de& malayischen und papuanischen Archipels.’ Engler’s Bot. Jahrb., VII, Litteraturber., pp. 51-72. 1893. ‘Ueber die Perldriisen des Weinstockes und anderer Pflanzen.’ Atti Congr. Bot. Internaz. Genova, (1892), pp. 237-245, Pl. xv. 1894. ‘Sopra una nuova pianta formicaria d’ Africa (Stereospermum dentatum Rich.).’ Malpighia, VIII, pp. 466-471, Pls. 1x—x. 1904. ‘Noterelle biologiche. I. Sopra una galla di Toddalia aculeata Pers. Il. Un caso di simbiosi fra formiche e Cicadelle.’ Atti Soc. Ligust. Se. Nat. Genova, XV, pp. 62-72, Pls. 1-1. Penzic, O. ann Cutaprera, C. 1903. ‘Contributo alla conoscenza delle piante acarofile.’ Malpighia, XVII, pp. 429-487, Pls. xvi—xvutt. Penzic, O. anv Saccanpo, P. A. 1897. ‘Diagnoses fungorum novorum in insula Java collectorum. Series secunda.’ Malpighia, XI, pp. 496-530. PeraGa.wo, A. 1882. ‘L’olivier, son histoire, sa culture, ses ennemis, ses maladies et ses amis.—Le frelon (Vespa crabro) et son nid.’ Nice. 180 pp., 1 Pl. EO ——— —— 568 Bulletin American Museum of Natural History [Vol. XLV Pérez, J. 1906. (Tetramcrium cxespitum injurious to live plants; Aphwn«gaster barbarus stealing seeds of vegetables; T'apinuma erraticum attending Tettig: metra obliqua.) Actes Soc. Linn. Bordeaux, LXI, Proc. Verb., PP. XXxXii—XxXv. ; Perris, E. 1873. ‘Notes sur quelques Hémiptéres myrmécophiles.’ Pet. Nouv. Entom., No. 84, pp. 336-337. Percn, T. 1906. ‘The fungi of certain termite nests (Termes redemanni Wasm. and T. cbscuriceps Wasm.).’ Ann. Roy. Botan. Gard. Peradeniya, III, pt. 2, pp. 185-270, Pls. v—xx1. 1907. ‘Insects and fungi.’ Science Progress, II, pp. 229-238. 1913a. ‘White ants and fungi.’ Ann. Roy. Botan. Gard. Peradeniya, V, pt. 6, pp. 389-393. 1913b. ‘The black termite of Ceylon (Eutermes monoceros Koen.),’ Ann. Roy. Botan. Gard. Peradenyia, V, pt. 6, pp. 395-420, Pls. vi—x1v. 1913c. ‘Termite fungi: A résumé.’ Ann. Roy. Botan. Gard. Peradeniya, V, pt. 5, pp. 303-341. Perersen, E. 1891. ‘Schutzringe gegen Ameisen.’ Prakt. Ratgeb. f. Obst- u. Gartenbau, p. 335 (Zeitschr. f. Pflanzenkrankh., II, p. 122). Prcapo, C. 1913. ‘Les Broméliacées épiphytes, considérées comme milieu biolog- ique.’ Bull. Scientif. France et Belgique, XLVII, pp. 215-360, Pls. VI-XXxIVv. Picarp, F, 1913. ‘Contribution A 1’étude des Laboulbéniacées d’Europe et du Nord de l'Afrique.’ Bull. Soc. Mycol. France, XXIX, pp. 503-571, Pls. XXIX—XXXII. Pierce, W.D. 1912. ‘The insect enemies of the cotton boll weevil.’ Bull. 100 U.S. Dept. Agric. Bur. Ent., pp. 1-99. Préron, H. 1907. ‘Sur la fondation de nouvelles colonies d’Aphzxnogaster (Messor) barbara nigra.’ Bull. Soc. Ent. France, pp. 280-282. Pintner, T. 1906. ‘Aus dem Leben der Ameisen.’ Vortr. Ver. Verbreit. Naturw. Kenntn. Wien, XLVI, pp. 103-146. Pinto, 8S. 1905. ‘Méfaits de la fourmi Quissondé.’ Archives de Parasitologie, X, p. 121. Praut, M. 1912. ‘Ueber die Symbiose bei‘Ameisenpflanzen. Neuere Arbeiten zur Kritik der Theorie Schimpers.’ Monatschr. Nat. Unterr., V, pp. 33-37. Purr, C.C. 1907. ‘Webera sessilis and ants.’ Bryologist, Brooklyn, N. Y., X, pp. 54-55. Porenog, E. A. 1904. ‘Pogonomyrmex occidentalis.’ Canadian Ent., XXXVI, p. 360. Porenoz, P. 1921. ‘Biological control of destructive insects.’ Science, N. S., LIV, pp. 113-114. Portier, P. 1918. ‘Les symbiontes.’ Paris. xx+315 pp., 1 Pl. Povtsen, V. 1875. ‘Om nogle Trikomer og Nektarier.’ Vidensk. Medd. Naturh. — Foren. Kjébenhavn, (3) VII, pp. 242-383, Pls. v—vt. Pricer, J. L. 1908. ‘The life history of the carpenter ant.’ Biol. Bull., XIV, pp. 177-218. Racrsorskt, M. 1898. ‘Biologische Mitteilungen aus Java.’ Flora, LXXXYV, pp. 325-361. ——— Wheeler, Ants of the Belgian Congo 569 1900. ‘Ueber myrmekophile Pflanzen.’ Flora, LXX XVII, pp. 38-45. 1902. ‘Rosliny i mréwki (Plants and ants).’ Kosmos, Lemberg, XXVII, . pp. 11-18. s Rima, E. 1883. ‘Untersuchungen iiber die Spermogonien der Rostpilze.’ Denk- eee schr. Ak. Wiss. Wien, XLVI, 2, pp. 1-51. ~~‘: 1889. ‘Ueber extraflorale Nectarien.’ Verh. Zool. Bot. Ges. Wien, XX XIX, oie Sitzb., pp. 14-21. ed es 1891. ‘Ueber myrmecophile Eichengallen.’ Verh. Zool. Bot. Ges. Wien, hee XLI, Sitzb., pp. 88-92. a . J. C. 1837-1844. ‘Die Forstinsekten.’ Berlin. Vol. I, 1837, 10+ eet 202 pp., 22 Pls.; vol. I, 1840, 252 pp., 17 Pls.; vol. III, 1844, 314 pp., 16 Pls. 1866-1868. ‘Die Waldverderbniss.’ Berlin. Vol. I, 1866, x+298 pp., 35 Pls.; vol. II, 1868, xvi+464 pp., 21 Pls. 1871. ‘Ueber die kiinstliche Vermehrung der Hiigelameise.’ Zeitschr. : f. Forst- u. Jagdwesen, III, pp. 433-434. Ray, J. (Rasus). 1686-1704. ‘Historia plantarum.’ London. 3 Vols. (Myrmeco- - phytes in Vol. III, p. 1373). Rea, G. 1914. ‘New and rare British fungi.’ Trans. Brit. Mycol. Soc. (1913), IV, pt. 2, pp. 307-317. Rex, L. 1897. ‘‘Biologische Beobachtungen an brasilianischen Ameisen.’ Illustr. Wochenschr. Ent., I, pp. 600-603, 612-616. Rescue, L. 1844. (Habits of Cfcodoma.) Ann. Soc. Ent. France, (2) II, Bull., p. Ixvii. alsin, H. 1906. ‘Weben der Ameisen.’ Natur und Haus, XV, pp. 248-249. Rertic, E. 1904. ‘Ameisenpflanzen-Pflanzenameisen.’ Beih. z. Bot. Centralbl., XVII, pp. 89-122. Ricuter, A. 1898. ‘Ueber die Blattstructur der Gattung Cecropia, insbesondere einiger bisher unbekannter Imbauba-Baume des tropischen Amerika.’ Bibliotheca Botanica, VIII, Heft 43, pp. 1-25, 8 Pls. Sete, M. E., Mrs. 1893. ‘Notes on Entilia sinuata.’ Insect Life, V, No. 4, pp. 243- 245. RIcHARDSON, vali 1886. ‘Ants and sunflowers.’ American Naturalist, XXI, pp. 297. Rick, J. 1903. ‘Zur Pilzkunde Vorarlbergs. V.’ Césterreich. Bot. Zeitschr., LIII, pp. 159-164. Ripitey, H. N. 1890. ‘On the habits of the caringa (Fcrmica gracilipes Gray).’ Journ. Straits Branch Roy. As. Soc., No. 22, Dec., pp. 345-347 (correc- tion, ibid., No. 23, 1891, p. 147). 1894. (Some remarks on Fcrmica smaragdina.) Trans. Ent. Soc. London, . Proc., p. xxxiii. 1908. ‘A list of the ferns of the Malay Peninsula.’ Journ. Straits Branch Roy. As. Soe., No. 50, pp. 1-59. 1910. ‘Symbiosis of ants and plants.’ Ann. of Botany, XXIV, pp. 457- 483, Pls. xxxv—-XxXxvI. Ruy, C. V. 1880. ‘Honey-producing oak-gall.’ Amer. Entom., (III) N. 8., I, p. 298. 1881. ‘Ants injurious in Arizona.’ American Naturalist, XV, pp. 573-574. 570 Bulletin American Museum of Natural History {Vol. XLV 1882. ‘The utilization of ants in horticulture.’ Nature, XXVI, p. 126. 1889. ‘Some insect pests of the household. Bed bugs and red ants.’ Insect Life, II, pp. 104-108. Rosert, E. 1845. ‘Observations sur les rapports des fourmis avee les pucerons.’ Ann. Sci. Nat., Zool., (3) III, pp. 99-102. Rosertson, C. 1897. ‘Seed-crests and myrmecophilous dissemination in certain plants.’ Botan. Gazette, XXIII, pp. 288-289. Roccuetti, B. 1905. ‘Richerche sugli acarodomazi.’ Contrib. Biol. Veget. Paler- mo, IV, pp. 1-37, 2 Pls. Ropway, J. 1894. ‘In the Guiana forest. Studies of nature in relation to the struggle for life.’ London. xxiii+242 pp., 16 Pls. (2d. Ed., 1911, 326 pp., 24 Pls.) Rorerxe, W. 1915. ‘Ons standpunt in zake het gramang en luizenvraagstuk voor de koffiecultuur, tevens een verweerschrift en een kritiek op de beschouwingen van den Heer Keuchenius over hetzelfde onderwerp.’ Teysmannia, X XVI, pp. 636-671. 1918. ‘Zur Myrmekophilie von Gerydus boisduvali Moore.’ Tijdschr. Vv. Ent., LXI, pp. 1-16, Pls. 1-1. Rouwer, 8. A. ‘The aphid genus Forda.’ Psyche, XV, pp. 67-68. Romanes, G. J. 1893. ‘The use of ants to aphids and Coccide.’ Nature, XLVIII, p. 54. Roquete Pinto, E. 1915. ‘Dinoponera grandis.’ Rio de Janeiro. (Review in Rev. Mus. Paulista, XI, 1919, pp. 808-810). Ross, H. 1909. ‘Pflanzen und Ameisen im tropischen Mexiko.’ Naturw. Wothen- schr., N. F., VIII, pp. 822-830. 1914. ‘Ueber veipiinbe Tiergallen.’ Ber. Deutsch. Bot. Ges., XXXII, pp. 574-597. Roruney, G. A. J. 1889. ‘Notes on Indian ants.’ Trans. Ent. Soc. London, pt. 3, pp. 347-374. Rovpat, J. 1905. ‘Notizen zur Biologie von Tettigometra atra Hagenb.’ Wien. Ent. Zeitg., XXIV, pp. 167-168. Roveer, A. 1867. (Tettigometra leta in nests of Tapinoma erraticum.) Ann. Soc. Ent. France, (4) VII, Bull., pp. bxxxiii-Ixxxiv. 1870. (Relations between Tetligometra obliqua and Formica congerens.) Ann. Soc. Ent. France, (4) X, Bull., pp. Ixxvi-Ixxvii. Ruvow, F. 1896. ‘Einige Ameisenwohnungen. ; Illustr. Wochenschr. Ent., I, pp. 473-475. 1905. ‘Ameisen als Gartner.’ Insektenbérse, XXII, pp. 199-200. 1906a. ‘Die Wohnungen der Ameisen.’ Ent. Jahrb., XV, pp. 148-171. 1906b. ‘Einige merkwiirdige Gallenbildungen.’ Ent. Jahrb., XVI, pp. 73- 105. Rumer, G. E. (Rumputivs). 1741-1750. ‘Herbarium Amboinense, plurimas complec- tens arbores, frutices, herbas, plantas terrestres et aquaticas, que in Amboina et adjacentibus reperiuntur insulis.’ Amsterdam. 6 vols. — Ruzsxy, M. 1913. ‘Myrmekologische Notizen. 1. Ueber Ameisen der Gattung Lasius Fbr. 2. Ueber das Nisten des braunen Lasius in Apfeln.’ Arch. f. Naturg., LX. XIX, Abt. A, Heft 9, pp. 58-63. — Wheeler, Ants of the Belgian Congo 571 Saccanvo, P- A. 1883. ‘Sylloge Pyrenomycetum.’ Vol. II. Padua. 813+Ixix+ eg s . 77 pp. 1886. ‘Sylloge Fungorum.’ Vol. IV. Padua. 810 pp. 1891: ‘Sylloge Fungorum.’ Vol. 1X. Padua. 1141 pp. 1895. ‘Sylloge Fungorum.’ Vol. XI. Padua. 753 pp. — ——s«41905. ‘Sylloge Fungorum.’ Vol. XVII. Padua. cvii+991 pp. / Saccanno, P. anp Trotrer, A. 1913. ‘Sylloge Fungorum.’ Vol. XXII. Padua. 1612 pp. - Sarrorn, W.E. 1910. “Bull-horn acacias in botanical literature, with a description of two new species.’ Science, N. 8., XX XI, pp. 676-677. 1914. ‘Acacia cornigera and its allies.’ Journ. Washington Ac. Sci., IV, pp. _ 1915. ‘New or imperfectly known species of bull-horn acacias.’ Journ. , _-———s Washington Ac. Sci., V, pp. 355-360. a SanLnEne, J. 1876. ‘Om forekomsten af Formica gagates hos en réd stackmyrart.’ Meddel. Soc. Fauna Flora Fenn., I, pp. 134-136. Sad, K. 1902. ‘Niitzlichkeit der Ameisen.’ Zeitschr. f. Pflanzenkrankh., XII, pp. 4 279-281. - 1910. ‘Die argentinische Ameise, eine neue sechsfiissige Grossmacht.’ : Prometheus, X XI, pp. 634-638. ; | Saussony, E.J. 1909. ‘The extrafloral nectaries of the genus Polygonum.’ Ann. of Botany, XXIII, pp. 229-242, Pl. xvi. ‘SamParo pe Azevepo, A. G. 1894. ‘Sativa ou Manhtadra. Monographia.’ Sad Paulo. SANDERSON, E.D. 1904. ‘The kelep and the cotton plant.’ Science, N.S., XX, pp. 887-888. Sanrscmt, F. 1914. ‘Formicide.’ Voy. Alluaud et Jeannel Afr. Orient., II, pp. 41- 148, Pls. 1-111. Savepans, E. 1896. ‘Honey bees destroyed by wood ants (Formica rufa).’ Ent. Monthly Mag., XXXII, p. 161. Saunpers, W. 1875. ‘The Mexican honey ant (Myrmecocystus mericanus).’ Cana- dian Ent., VII, pp. 12-14. _ Savace, T.S. 1847. ‘On the habits of the drivers, or visiting ants of West Africa.’ Trans. Ent. Soc. London, V, pp. 1-15. 1849a. ‘The driver ants of Western Africa.’ Proc. Acad. Nat. Sci. Phila- delphia, IV, pp. 195-202. : 1849). ‘Observations on the species of Termitide of West Africa, described by Smeathman as 7’. bellicosus, and by Linnwus as 7’. fatalis.’ Proc. | Acad. Nat. Sci. Philadelphia, IV, pp. 211-221. ‘Savirie-Kent, W. 1891. ‘Observations on the weaving habits of the North Queens- * land green ant, Formica virescen. Fabr.’ Proc. Roy. Soc. Queensland, VIII, 2, p. xii (title only). 1897. ‘The naturalist in Australia.’ Epadbin: 302 pp., 59 Pls. Scuencx, H. 1900. ‘Ueber die Wechselbeziehungen zwischen Pflanzen und Ameisen im tropischen Wald.’ Ber, Senckenberg. Naturf. Ges. Frankfurt a/M., pp. civ-evi. 1913. ‘Acacia myrmecophilse nove.’ Fedde’s Repertorium, XII, p. 370. 1914. ‘Die myrmekophilen Acacia-Arten.’ Engler’s Bot. Jahrb., L, Suppl., pp. 449-487. eg — Nae 572 Bulletin American Museum of Natural History (Vol. XLV _ Scnenkuine-Prévor. 1896a. ‘Ameisen als Pilz-Ziichter und -Esser.’ Tlustr. Wochenschr. Ent., VI, pp. 89-93. 18965. ‘Die Héckerameisen und ihre Pilzgiirten.’ Insektenbérse, XIII, p. 264. 1896c. ‘Die Pilzgiirten der Haarameisen.’ Insektenbérse, XIII, pp. 153- 154. 1897. ‘Rozites gongylophora, die Kulturpflanze der Blattschneider-Ameisen.’ Illustr. Wochenschr. Ent., I, pp. 56-60. Scnitter-Trerz,— 1902. ‘Ameisen, Blattliuse und Honigthau an Kulturpflanzen.’ Schweiz. Bauer, LVI, p. 70. Scuimprer, A. F. W. 1888. ‘Die Wechselbeziehungen zwischen Pflanzen und Ameisen im tropischen Amerika.’ Botan. Mitt. aus den Tropen, Jena, Heft 1, pp. 1-95, 3 Pls. 1889. ‘Zur Frage der Myrmekophilie von Myrmecodia und Hydnophytum.’ Botan. Zeitung, XX XI, pp. 507-511. 1898. ‘Pflanzengeographie auf physiologischer Grundlage.’ Jena. xviii+ 876 pp. (2d. Ed., 1908; English translation, Oxford, 1903, xxx+ 839 pp.). Scuinz, H. 1894. ‘Ueber myrmekodome Akazien.’ Ber. Schweiz. Bot. Ges., IV, pp. xx-xxii. Scurps, M. 1920. ‘Ueber Wanderameisen.’ Naturw. Wochenschr., XXXV, pp. 618-619. ScuiecuTENDAL, D. von. 1846. ‘Review of “Klotzsch, Herbar. Mycol., Centuria XI cura L. Rabenhorst.”’ Botan. Zeitung, IV, pp. 876-880. 1858. ‘Review of ‘Klotzsch, Herbar. Mycol. Ed. Nov., Centuria VIII cura L. Rabenhorst.”’ Botan. Zeitung, XVI, pp. 302-304. ScHLECHTENDAL, D. von. AND Cuamisso, A. pe. 1830. ‘Plantarum Mexicanarum a cl. vir. Schiede et Deppe collectarum recensio brevis.’ Linnea, V, pp. 554-625. Scumacx, A. 1909. ‘Lebensweise der siidbrasilianischen Ameise Azteca muelleri.’ Berlin. Ent. Zeitschr., LIV, Sitz. Ber. f. 1908, p. (4). ScumipBercer, J. 1837. ‘Ueber Bostrychus dispar.’ Kollar’s Naturgeschichte der schidlichen Insekten. Vienna. Scumipt, H. 1914. ‘Einige Notizen itiber das Zusammenleben von Gallinsekten und Pilzen an einheimischen Pflanzen.’ Fihlings Landw. Zeit., LXIII, 4, pp. 143-146. ScuNEIDER-ORELLI, O. 191la. ‘Ueber die Symbiose eines einheimischen pilzziich- tenden Borkenkiifers (Xyleborus dispar F.) mit seinem Niahrpilze.’ Verh. Schweiz. Naturf. Ges., 94. Jahresv., Bd. 1, pp. 279-280. 1911b. ‘Die Uebertragung und Keimung des Ambrosia-Pilzes von Xyleborus (Anisandrus) dispar F.’ Naturw. Zeitschr. f. Forst- u. Landwirtsch., IX, pp. 186-192, 223-225. 1912. ‘Der ungleiche Borkenkifer (X ylebc rus dispar F.) an Obstbiumen und sein Nihrpilz.’ Landwirtsch. Jahrb. d. Schweiz, pp. 326-334. 1913. ‘Untersuchungen tiber den pilzziichtenden Obstbaumborkenkifer Xylebcrus (Anisandrus) dispar und seinen Nihrpilz.’ Centralbl. Bakter. Parasitenk., XX XVIII, Abt. 2, pp. 25-110, 3 Pls. 1920. ‘Beitriige zur Biologie des pilzziichtenden Kifers, Hylecatus dermes- toides.’ Mitt. Schweiz. Ent. Ges., XIII, 2, pp. 64-67. , Wheeler, Ants of the Belgian Congo 573 ae —_— emkase’: 1067-1868. ‘Reioon in British-Guiana in den Jahren 1840- 1844.’ Leipzig. Vol. I, 1847, 468 pp.; vol. Il, 1848,530pp.; vol. III, ee 1848, pp. 531-1260. . Rosert. 1838. ‘On the ant tree of Guiana (Triplaris americana).’ ay H. 1902. ‘Les Aphides radicicoles de Belgique et les fourmis.’ Ann. Soe. Ent. Belgique, XLVI, pp. 136-142. H. von. 1903. ‘The “bluing” and the “red rot” of the western yellow _ pine, with special reference to the Black Hills Forest Reserve.’ U.S. Dept. Agric. Bur. Plant Ind., Bull. No. 36, 40 pp., 14 Pls. Ser C. 1899. ‘Formica rufa 8 als Bliitenbestiuber.’ Illustr. Zeitschr. Pe, Ent., IV, p. 283. ‘Somonaa, J. 1894. ‘Pilze. II.’ in Cohn, ‘Kryptogamen-Flora von Schlesien,’ PaaS 9 III, pt. 2. _ Scnumann, K. 1888. Sith ies ‘Aaisiotanfonoen.’ Pringsheim’s Jahrb. Wiss. ; 1889. ‘Ueber Asuiieentinnme.’ Virchow-Holtzendorff’s Gemeinverst. _ Vortriige, 38 pp., 1 Pl. 1890. ‘Einige weitere Ameisenpflanzen.’ Verh. Bot. Ver. Brandenburg, - XXXI, 2, pp. 113-123. _ 1891a. ‘Ueber afrikanische Ameisenpflanzen.’ Ber. Deutsch. Bot. Ges., EX, pp. 54-72. . 1891b. ‘Rubiacee.’ Engler and Prantl, ‘Die Natiirlichen Pflanzen- oe familien,’ IV, pt. 4, pp. 1-156. _-«:1902. ‘Ameisenpflanzen.’ Gartenflora, LI, pp. 95-101. 1903. ‘Apocynacez Africanz.' Engler’s Bot. Jahrb., XX XIII, 2, pp. 316- 321. Scuoman, K. anp Lavrersacn, K. 1901. ‘Die Flora der Deutschen Schutz- = gebiete in der Siidsee.’ Berlin. xvi+613 pp., 22 Pls. = Somwans, E. A. 1905. ‘Note on Dorymyrmex pyramicus Roger.’ Proc. Ent. Soc. Washington, VII, p. 4. 1916. ‘Ants protecting Acacia trees in Central America.’ Proc. Ent. Soc. 4 Washington, XVIII, p. 211. = Scuwanze, W. 1902. ‘Beitriige zur Kenntniss der Symbiose im Tierreich.’ Ham- : burg, Progr. Johanneum, 40 pp. (review in Zool. Zentralbl., 1902, pp. 331-334). ScHWEINFURTH, G. 1867. ‘Aufzihlung der Acacien-Arten des Nilgebiets.’ Lin- nwa, XXXV, pp. 309-376, Pls. rv—xxim. ‘Scuwenvt, E. 1907. ‘Zur Kenntnis der extrafloralen Nectarien.’ Beih. z. Bot. Centralbl., XXII, Abt. 1, pp. 245-286, 2 Pls. Soorr, D. H. wp Sancewr, Era. 1893. ‘On the pitchers of Dischidia Raffesiana a. (Wall.).’ Ann. of Botany, VII, pp. 243-269, Pls. x1-xn. Senn, H. 1910. ‘Die Knollen von Polypodium Brunei Werckle.’ Verh. Naturf. = Ges. Basel, XXI, pp. 115-125. eaee, R. 1901. ‘Den Skandinaviska vegetationens spridningsbiologi.’ Upsala. iv+459 pp. 1906a. ‘Ueber postflorale Nektarien. Fin Beitrag zur Kenntnis der myr- mekotrophen Anpassungen im Diente der Verbreitungsbiologie.’ Botaniska Studier tillignade F. R. Kjellman, Upsala, pp. 275-287. 574 Bulletin American Museum of Natural History (Vol. XLV 19065. ‘Entwurf einer Monographie der europiischen Myrmecochoren,’ Kungl. Svenska Vetensk. Ak. Handl., XLI, No. 7, pp. 1-410, 11 Pls. Serre, P. 1909. ‘La lutte contre les fourmis 4 Cuba.’ Bull. Mus. Hist. Nat. Paris, XV, pp. 188-192. 1913. ‘La peste des fourmis-manioc A Bahia.’ Bull. Mus. Hist. Nat. Paris, XIX, pp. 438-441. Seurat, L.G. 1901. (Crematcgaster scutellaris nesting in the bark of cork-oaks in Tunis) Revue des Cultures Coloniales, No. 86, p. 197. Seuricu, P. 1891. ‘Ueber die Wechselbeziehungen zwischen Pflanzen und Ameisen.’ 11. Ber. Naturw. Ges. Chemnitz, Mitt. Sitz., (1890), pp. xxxviii—xliv. Suetrorp, R. W.C. 1916. ‘A naturalist in Borneo.’ London. xxvii+331 pp. Sirvesrri, F. 1902. ‘Ergebnisse biologischer Studien an siidamerikanischen Termi- ten.’ Allgem. Zeitschr. Ent., VII, pp. 173-178, 257-260, 289-292, 326-334. Sséstept, Y. 1896. ‘Termes Lilljebcrgi, eine neue, wahrscheinlich pilzanbauende Tagtermite aus Cameroon,’ Festskrift fr Lilljeborg, Upsala, pp. 269— 280, Pl. xrv. 1900. ‘Monographie der Termiten Afrikas.’ Kongl. Svenska Vet. Ak. Handl., XXXIV, No. 4, pp. 1-236, Pls. 11x. 1903. ‘Termiterna och deras biologi.’ Arsbok Svenska Vet. Ak., pp. 89- 101. 1908. ‘Akaziengallen und Ameisen auf den ostafrikanischen Steppen.’ In Sjéstedt, ‘Exped. Kilimandjaro, Meru, etc.,’ II, Abt. 8, pt. 4, pp. 97-118, Pls. vi—virr. 1909. ‘Symbios mellan acacior och myror p& de ostafrikanska stipperna.’ Fauna och Flora, Upsala, IV, pp. 34-44. SmeaTHMAN, H. 1781. ‘Some account of the termites, which are found in Africa and other hot climates.’ Philos. Trans. Roy. Soc. London, LXXI, pp. 139-192, 4 Pls. Smiru, E. F. 1896a. ‘Ambrosia.’ American Naturalist, XXX, pp. 318-319. 18966. ‘White ants as cultivators of fungi.’ American Naturalist, XXX, pp. 319-321. Smirn, F. 1876. ‘Descriptions of new species of Cryptoceride, belonging to the genera Cryptocerus, Meranoplus, and Cataulacus.’ Trans. Ent. Soc. London, pp. 603-612, Pl. x1. Smirn, M.R. 1916a. ‘Observations on ants in South Carolina.’ Ent. News, XX VII, pp. 279-280. 1916). ‘Notes on South Carolina ants.’ Ent. News, XX VII, p. 468. Smirxa, WINIFRED. 1903. ‘Macaranga triloba; a new myrmecophilous plant.’ New Phytologist, II, pp. 79-82, Pls. v—vr. Samira, W.W. 1892a. ‘On the origin of ants’ nests.’ Ent. Monthly Mag., XXVIII, pp. 60-65. 1892b. ‘Coccids in ants’ nests.’ Ent. Monthly Mag., XXVIII, p. 307. SotereDER, H. 1920. Ueber eine heterophylle philippinische Ameisenpflanze aus der Familie der Melastomatacee, nebst Bemerkungen itiber das Auftreten von Amylodextrin-K6éruern in den sog. Perldriisen.’ Naturw. Wochenschr., XXXYV, pp. 689-691. _ Wheeler, Ants of the Belgian Congo 575 Seman, A. T. 1920. ‘On certain entomogenous fungi.’ Mycologia, New York, XII, pp. 62-76, Pls. m1-—v. Sreosnn, C. 1881-1882. ‘Fungi argentini, additis nonnullis brasiliensibus Poids montevideensibusque. Pugillus quartus.’ An. Soc. Cientif. Argentina, ; XII, pp. 13-30, 63-82, 97-117, 174-189, 208-227, 241-258, Pl.1; XIII, et. 1882, pp. 11-35, 60-64. . 1889. ‘Fungi Puiggariani. Pugillus I.’ Bol. Ac. Nac. Ciencias Cérdoba, ss XI, pt. 4, pp. 381-622. 1899. ‘Fungi argentini novi v. critici.’ An. Mus. Nac. Hist. Nat. Buenos ter. Aires, VI, pp. 81-365. «1910. ‘Mycetes argentinenses. Series V.’ An. Mus. Nac. Hist. Nat. _-—«é#Bueenos Aires, (3) XIII, pp. 329-467. 1912. ‘Mycetes argentinenses. Series VI.’ An. Mus. Nac. Hist. Nat. oteed t- Buenos Aires, XXIII, pp. 1-146. --—s«4$915. += ‘Primo contributo alla conoscenza delle Laboulbeniali Italiane.’ _ —sKRedia, X, pp. 21-75, Pls. 1-1x (separate in 1914). _ 1917. ‘Revisién de las Laboulbeniales argentinas.’ An. Mus. Nac. Hist. a * Nat. Buenos Aires, XXIX, pp. 445-688. ‘oe Sensex, HL 1894. ‘Origin of classes among the “parasol” ants.’ Nature, LI, pp. ey . 125-126. Serx, J. vow. AND Martius, C. von. 1831. (On the habits of ants; among others ee ae those found inside leaf-pouches of Melastomatacer.) ‘Reise in ___ Brasilien in den Jahren 1817 bis 1820.’ III, pp. 1283-1284. Srnuce, Rica. 1908. ‘Notes of a botanist on the Amazon and Andes.’ Edited by . A. R. Wallace. London. Vol. I, lii+518 pp.; vol. Il, xii+542 pp. al Som, 2. 1909. ‘Beitrag zur schweizerischen “Epiphytenflora”.’ Mitt. Naturf. Ges. Bern, (1908), pp. 17-90, 1 PI. 1917. ‘Beitrag zur Kenntnis stengelbewohnender Ameisen in der Schweiz.” Rev. Suisse Zool., XXV, pp. 95-109. 1919. ‘Meine Reise auf alte Brombeerstengel.’ Natur u. Technik, I, pp. 219-222. Sremnvorru,— 1897. ‘Ameisenpflanzen.’ 44-47. Jahresb. Nat. Ges. Hannover, ‘ (1893-1897), p. 32. §Srer,E. 1913. ‘Messmates. A book of strange companionships in nature.’ London. xii+220 pp Srernan, J. 1912. Tes wiwehés Hausgenossen aus dem Insektenreich.’ Naturw. “7 Techn. Volksbiicherei, No. 29, 48 pp. Siz, H. 1910. ‘Westafrikanische Ameisen. I.’ Mitt. Zool. Mus. Berlin, V, pp. 125-151. 1911. ‘Formicid#.’ Wiss. Ergebn. Deutsch. Zentr. Afr. Exp. (1907-08), III, pp. 375-392. 1913a. ‘Ameisen aus Brasilien, gesammelt von Ule.’ Deutsch. Ent. Zeitschr., pp. 207-212. 19136. ‘Ameisen und Pflanzen.’ Die Naturwissenschaften, I, pp. 1281- 1288. 1917. ‘Die Beziehungen der Ameisen zum Menschen und ihre wirtschaft- aq liche Bedeutung.’ Zeitschr. f. Angew. Ent., IV, 1, pp. 71-128. _ Srone,J.H. 1881. (Lychnis viscaria, a trap for ants.) Nature, XXV, pp. 151-152 576 Bulletin American Museum of Natural History (Vol. XLV Stores, M. C. anv Hewirr, C.G. 1909. ‘On the tent-building habits of the ant, Lasius niger L., in Japan.’ Mem. Manchester Lit. Phil. Soe., LIII, No. 20, pp. 1-6, 1 Pl. Srroumeyer, H. 1906. ‘Neue Untersuchungen tiber Biologie, Schiidlichkeit und Vorkommen des Eichenkernkiifers Platypus cylindrus var.? cylin- driformis Reitter.’ Naturw. Zeitschr. f. Forst- u. Landwirtsch., IV, pp. 329-341, 409-420, 506-511, 2 Pls. 1911. ‘Die biologische Bedeutung sekundiirer Geschlechtscharaktere am Kopfe weiblicher Platypodiden.’ Entom. Blitter, VII, pp. 103-107, 2 Pls. SrupHavter, R. A. anp Ruaeies, A.G. 1915. ‘Insects as carriers of the chestnut blight fungus.’ Pennsylvania Dept. For. Bull., XII, pp. 1-33. Srumprr, E. 1905. ‘Ein Fall von Symbiose zwischen Ameisen und Cicaden.’ Aus der Natur, I, pp. 443-444. Surrace, H. A. 1913. ‘Pests of domestic animals, households and buildings, bush fruits and lawn plants.’ Zool. Bull. Div. Zool. Pennsylvania Dept. Agric., II, pp. 1-30. Swarm, J. 1864. (Ants caring for aphids inside shelters on Vernonia noveboracensis near Philadelphia.) C.R. Ac. Sci. Paris, LIX, p. 230. Swarn, A. F. 1920. ‘Argentine ant control from an economic standpoint.’ Monthly Bull. California State Dept. Agric. Sacramento, [X, No. 8, pp. 333-338. Swinate, W. T. 1896. ‘Fungus gardens in the nest of an ant (Atta tardigrada Buckley) near Washington, D. C.’ Proce. American Ass. Ady. Sci., 44th Meet., pp. 185-186. Swynnerton, C.F. M. 1916. ‘Experiments on some carnivorous insects especially the driver ant Dcrylus; and with butterflies eggs as prey.’ Trans. Ent. Soc. London, (1915), pp. 317-350. Syxes, M. L. 1899. ‘Termites and ants of West Africa.’ Trans. Manchester Microsc. Soc., pp. 85-91. Syxes, W.H. 1835. ‘Descriptions of new species of Indian ants.’ Trans. Ent. Soc. London, I, pt. 2, pp. 99-107, Pl. xm. SamApy, Z. 1911. ‘Hangya-névénybardétsag (Ameisen-Pflanzen-Freundschaft).’ Termes. K6zl. Magyar Tars., XLIII, pp. 107—108. Tatrew, W. 1898. ‘Bemerkungen iiber einige russische mirmecophile Pflanzen.’ Arbeiten Naturf. Ges. Charkow, 1898, Suppl., pp. xli-xliv (Beih. z. Bot. Centralbl., VIII, p. 93-94). Tanner, J. E. 1892. ‘The parasol or leaf cutting ant. (#£ccdema cephalctes.’ Journ. Trinidad Field Nat. Club, I, pp. 68-69 and 123-127. Terractano, A. 1898. ‘I nettarii estranuziali nelle “Bombacee.”’ Contrib. Biol. Veget. Palermo, II, 2, pp. 137-191, Pls. rx—xvut. Tuaxter, R. 1888. ‘The Entomophthorew of the United States.’ Mem. Boston Soc. Nat. Hist., IV, pt. 6, pp. 133-201, Pls. xrv—xx1. 1891. ‘On certain new or peculiar North American Hyphomycetes. vt Botan. Gazette, XVI, pp. 201-205, Pls. xrx—xx. 1902. ‘Preliminary diagnoses of new species of Laboulbeniacesw.’ Proc. American Ac. Arts Sci. Boston, XX XVIII, pp. 7-57. 1908. ‘Contribution toward a monograph of the Laboulbeniacew. Part II.’ Mem. American Ac. Arts Sci. Boston, XIII, pt. 6, pp. 217-469, Pls. XXVIII-LXXI. Wheeler, Ants of the Belgian Congo 577 4914. ‘On certain peculiar fungus-parasites of living insects.’ Botan. —_ Gazette, LVIII, pp. 235-253, Pls. xvi—xrx. Tanonato, F.V. 1913. ‘First list of Aphides found with Myrmica.’ Ent. Record, XXV, pp. 48-51. 1914. ‘Two new myrmecophilous Aphides from Algeria.’ Entomologist, XLVI, pp. 28-31. 1915. ‘New myrmecophilous Aphides.’ Ent. Record, X XVII, pp. 52-55. 1916a. ‘Notes on Aphidie found in ants’ nests.’ Entomologist, XLIX, p. 49. 19165. ‘A new myrmecophilous aphid from Africa.’ Ent. Record, XXVIII, . pp. 37-38. 7" Tanase, C. 1908. ‘Ueber den Nestbau einiger Ameisen.’ Aus der Natur, I, pp. 664-668. Turse1.0N—Drer, W.T. 1902. ‘Morphological notes. VII. Evolution of pitchers __ in Dischidia Rafflesiana.’ Ann. of Botany, XVI, pp. 365-369. TaoMmann, H. .1908. ‘Schmetterlinge und Ameisen. Ueber das Zusammenleben der Raupen von Psecadia pusiella Rém. und P. decemguttella Hb. mi Formiciden.’ Jahresb. Nat. Ges. Graubindens, pp. 21-31. THwarres, G. 1854. ‘On ants destructive to cocci.’ Trans. Ent. Soc. London, (2) Ill, Proc., p. 10. Tauren, 0. 1914. ‘Woran leiden die meisten Kakaopflanzungen?’ Tropenpflanzer, ; XVIII, p. 403. ‘Fixstey, J.D. 1898. ‘An ants’-nest coccid from New Mexico.’ Canadian Ent., XXX, pp. 47-48. Treat, J. D. anv Kina, G. B. 1899. ‘The tenth ants’-nest species of coccid from Massachusetts.’ Ent. News, X, p. 57. Trrvs, E. 8. G. 1905. ‘Report on the “New Orleans” ant (Iridcemyrmex humilis Mayr).’ Bull. 52 U.S. Dept. Agric. Bur. Ent., pp. 79-84. TorKka, V. 1905. ‘Tettigometra obliqua Panz.’ Zeitschr. Wiss. Insektenbiol., I, pp. OE —— , 451-455. Townsenp, B. R. 1870. ‘The red ant of Texas.’ Amer. Entom. and Botan., II, TricArpu, I. 1904. ‘Termiten aus den Sudan.’ Jagerskiéld Exped., I, pp. 1-47, ; 3 Pls. 1914. ‘Notes biologiques.’ Med. Géteborgs Mus. Zool. Afd., III, pp. 43-47. Treat, M. 1878. ‘The harvesting ants of Florida.’ Lippincott’s Mag. Pop. Lit. Sci., XXII, pp. 555-562. 1880. ‘Notes on harvesting ants in New Jersey.’ Amer. Entom., III, pp. 225-226. Tretease, W. 1879. ‘Nectar, its nature, occurrence and uses.’ In Comstock, J. H. ‘Report on Cotton Insects.’ Ithaca, pp. 317-343, 1 Pl. 1881. ‘The foliar nectar glands of Populus.’ Botan. Gazette, Chicago, VI, pp. 384-390. 1882. ‘Unusual care of ants for Aphides.’ Psyche, III, pp. 310-311; (additional note), ibid., IV, 1883, pp. 13-14. 1889. ‘Myrmecophilism.’ Psyche, V, pp. 171-180. Trevs, M. 1883a. ‘Sur les urnes du Dischidia Rafflesiana (Wall.).’ Ann. Jardin Bot. Buitenzorg, III, pp. 13-36. oy i, a ee 578 Bulletin American Museum of Natural History (Vol. XLV 1883b. ‘Sur le Myrmecodia echinata.’ Ann. Jardin Bot. Buitenzorg, IIT, pp. 129-157. 1888. ‘Nouvelles recherches sur le Myrmeccdia de Java [M yrenecedis tuberosa Beccari (non Jack )}.’’ Ann. Jardin Bot. Buitenzorg, VI, pp. 191-212, Pls. xxmI—xxv. Tren, R. 1886. ‘Notes on insects apparently of the genus Margarodes, Lansd.- Guilding, stated to occur abundantly in the nests of white ants, and also of true ants, in certain western districts of the Cape Colony,’ Trans. Ent. Soc. London, pp. 461-463. 1895. (Note on a South African honey-ant.) Trans. Ent. Soc. London, Proc., p. xxiii. Trotrer, A. 1900a. ‘I micromiceti delle galle.’ Atti Ist. Veneto Sci., LEX, pp. 715-736. 1900b. ‘Ricerche intorno agli entomocecidi della flora italiana.’ Nuovo Giorn. Bot. Italiano, N.8., VII, pp. 187-206, PI. rx. 1904. ‘Osservazioni sugli acarodomazii.’ Bull. Soc. Bot. Italiana, XIII, pp. 82-86. 1905. ‘Nuove ricerche sui micromiceti delle galle e sulla natura dei loro rapporti ecologici.’ Ann. Mycologici, III, pp. 521-547 (review in Marcellia, V, 1906, pp. xvii—xix). Tryon, H. 1888. ‘Notes on the Queensland ants.’ Proc. Roy. Soc. Queensland, II, pp. 146-162. 1900. ‘Harvesting ants.’ Queensland Agric. Journ., pp. 71-79. Tucker, E, X. 1911. ‘Random notes on entomological field work.’ Canadian Ent., XLIII, pp. 22-32. Tuasne, L. AND C.. 1861-1865. ‘Selecta fungorum carpologia.’ Paris. 3 vols. Turner, C.H. 1909. ‘The mound of Pogonomyrmex badius Latr. and its relation to the breeding habits of the species.’ Biol. Bull., X VII, pp. 161-169. Uxsricu, E. 1907. ‘Ueber europiische Myrmekochoren.’ Verh. Bot. Ver. Branden- burg, XL, pp. 214-241. 1919. ‘Deutsche Myrmekochoren. Beobachtungen tiber die Verbreitung heimischer Pflanzen durch Ameisen.’ Leipzig and Berlin. 60 pp. Ure, E. 1897. ‘Symbiose zwischen Asclepias curassavica und einem Schmetterling nebst Beitrag zu derjenigen zwischen Ameisen und Cecropia.’ Ber. Deutsch. Bot. Ges., XV, pp. 385-387. 1900. ‘Verschiedenes tiber den Einfluss der Thiere auf das Pflanzenleben.” Ber. Deutsch. Bot. Ges., XVIII, pp. 122-130. 1901. ‘Ameisengiirten im Amazonasgebiet.’ Engler’s Bot. Jahrb., XXX, : Beibl. No. 68, pp. 45-52, Pl. xx11. 1905a. ‘Blumengiirten der Ameisen am Amazonenstrome.’ Karsten u. Schenck, Vegetationsbilder, 3. Reihe, Heft 1, 6 Pls., 14 pp. of text. 1905b. ‘Wechselbeziehungen zwischen Aimelens und Pflanzen.’ Flora, XCIV, pp. 491-497. 1905c. ‘Biologische Eigentiimlichkeiten der Friichte in der Hylea.’ Engler’s Bot. Jahrb., XX VI, Beibl. No. 81, pp. 91-98. 1905d. ‘Ueber Blumengiirten der Ameisen am Amazonenstrom.’ Verh. Ges. Deutsch. Nat. A2rtze, Vers. LX XVI, 2, 1, pp. 245-249. { . : ee 579 ; ‘Bigentiimliche von Pflanzen durchwachsene Ameisennester am _ Amazonenstrom.’ Naturw. Wochenschr., XXI, pp. 145-150, 1 PI. ‘Ameisenpflanzen des Amazonasgebietes.’ Karsten u. Schenck, -Vegetationsbilder, 4. Reihe, Heft 1, 6 Pls., 14 pp. of text. [-1908. ‘Die Pflanzenformationen des Amazonas-Gebietes. Pflanzen- _ geographische Ergebnisse meiner in den Jahren 1900-1903 in Bra- ~ silien und Peru unternommenen Reisen.’ Engler’s Bot. Jahrb., XL, 2, 1907, pp. 114-172; 3-4, 1908, pp. 398-443; Pls. m—vii and rx—x1. 1915. ‘Die Vegetation des Amazonasgebietes.’ Verh. Bot. Ver. Branden- burg, LVII, 1, pp. 56-75. . 1893. ‘Notes on some insect pests of Trinidad, British West Indies.’ Insect Life, VI, pp. 196-198. -1895a. ‘Notes on some fungus-growing ants in Trinidad.’ Journ. Trinidad _ Field Nat. Club, II, 7, pp. 175-182. . ‘Notes on the fungus growing and eating habit of Sericomyrmex __ @pacus Mayr.’ Trans. Ent. Soc. London, pp. 77-78. Laéene, C.F. 1885. ‘Brazil and Java. Report on coffee-culture in _ America, Asia and Africa.’ London. xii+637 pp. R. 1915. ‘Eenige interessante gevallen van symbiose tusschen plant en ‘mier.’ Entom. Berichten, IV, pp. 174-176. FELT, E. 1909. ‘Plantenstoffen en mieren.’ Pharmac. Weekbi., XLVI, ppp. 1002-1009. R.A. 1910. ‘Contributions to a knowledge of the corn root-aphis.’ Bull. r _ 85 U.S. Dept. Agric. Bur. Ent., pt. 6, pp. 97-118. H. 1909. ‘Bilder ans dem Ameisenleben.’ Leipzig. viii+159 pp. 116. ‘Ameicen von Singapore. Beobachtet und gesammelt von H. Over- beck.’ Arch. f. Naturg., LX XXI, Abt. A, Heft 8, (1915), pp. 108-168. KENS,G. 1920. ‘Ameisenpflanzen.’ Deutsches Kvlonial-Lexikon, I, p. 41. man, A.G. 1895. ‘Mangg‘s kunstmatig onder bescherming van mieren te ____‘Tj'intjing.’ ~Teysmannia, VI, pp. 673-676. SI ern, J. 1905a. ‘Die ostafrikanische Treiberameise (Siafu).’ Der Pflanzer, No. 19, pp. 289-302. 19056. ‘Die Baumwollkifer-Ameise aus Guatemala.’ Pflanzer, I, p. 362. _ 1906a. ‘Insektenwanderungen in Usambara.’ Insektenbérse, XXIII, : pp. 68, 70-71, 75, 77-78. 1906). ‘Verhinderung des Fruchtansatzes bei Cobea scandens dureh Amei- sen.’ Zeitschr. Wiss. Insektenbiol., II, pp. 204-206. ‘ “1878. (Special relations between ants and the vegetation.) ‘Tropical Nature and Other Essays,’ London, pp. 89-90. q Wass, B.D. anv Ritxy, C. V. 1869. ‘Ants and aphids.’ Amer. Entom., I, p. 110. a o- H.T. 1891. ‘On the habits of certain “harvesting” ants.’ Scient. Mem. 3 ot Med. Off. Army of India, VI, 6 pp., 1 PI. 580 Bulletin American Museum of Natural History {Vol. XLV Wanacn, B. 1911. ‘@cophylla smaragdina F.’ Berlin. Ent. Zeitschr., LVI, Sitz. Ber. f. 1910, p. (32). . Warsura, 0. 1892. ‘Ueber Ameisenpflanzen.’ Biol. Centralbl., XII, pp. 129-142. 1894. ‘Plante Hellwigiane. Beitrag zur Flora von Kaiser Wilhelms- Land.’ Engler’s Bot. Jahrb., XVIII, pp. 184-212. 1897. ‘Monographie der Myristicaceen.’ Nova Acta Ac. Nat. Curios. Halle, LX VIII, 680 pp., 25 Pls. Warmina, E. 1894. ‘Om et Par af Myrer beboede Trwer.’ Vidensk. Meddel. Naturh. Foren. Kjébenhavn, (5) V, (1893), pp. 173-187. Wasmann, E. 1884. ‘Die Honigameise des Géttergartens.’ Stimmen aus Maria- Laach, XXVII, pp. 275-285. 1887. ‘Die Getreidesammelnden Ameisen in alter und neuer Zeit.’ Stim- men aus Maria-Laach, XXXIII, pp. 560-374. (Tijdschr. v. Ent., XXXII, 1889, pp. 104-108). 1889. ‘Wechselbeziehungen zwischen Pflanzen und Ameisen im tropischen Amerika.’ Natur u. Offenbarung, XXXV, pp. 487-489. 1892. ‘Ein neuer Paussus vom Somaliland.’ Mitth. Schweiz. Ent. Ges., VIII, pp. 355-357. 1899. ‘Mit Schimmelpilzen behaftete Ameisenkolonien.’ Natur u. Offen- barung, XLV, p. 505. 1905a. ‘Versuche mit einem brasilianischen Ameisennest in Holland.’ Tijdschr. v. Ent., XLVIII, pp. 209-213, Pl. vir. 1905b. ‘Beobachtungen tiber Polyrhachis dives auf Java, die ihre Larven zum spinnen der Nester beniitzt.’ Notes Leyden Mus., XXV, pp. 133- 140. 1908. ‘Beschreibung des Gespinnstnestes von Polyrhachis laboriosa Sm.’ Notes Leyden Mus., XXX, pp. 66-67, Pl. vr. 1914. ‘Ameisenplagen im Gefolge der Kultur.’ Stimmen aus Maria-Laach, LXXXVII, pp. 589-598. 1915a. ‘Eine neue Pseudomyrma aus der Ochsenhorndornakazie in Mexiko, mit Bemerkungen tiber Ameisen in Akaziendornen und ihre Giste. Ein kritischer Beitrag zur Pflanzen-Myrmekophilie.’ Tijdschr. v. Ent., LVILI, pp. 296-325, Pls. x1v—xvit. 1915). ‘Nachtrag zu “Eine neue Pseudcmyrma aus der Ochsenhorndorna- kazie in Mexiko.”’’ Tijdschr. v. Ent., LVIII, Suppl., pp. 125-131. 1917. ‘Die Ausbreitung der argentinischen Ameise in der Kapkolonie und ihr Einfluss auf die einheimische Ameisenfauna.’ Ent. Mitt. Deutsch. Ent. Mus. Berlin, VI, pp. 184-186. Wesster, F.M. 1887-1888. ‘The relation of ants to the corn-aphis.’ Rept. Comm. Agric. Washington, pp. 148-149. 1890. ‘Some hitherto unrecorded enemies of raspberries and blackberries.’ Insect Life, II, Nos. 7-8, pp. 257-258. 1899. ‘On the relations of'a species of ant, Lasius americanus, to the peach root louse, Aphis prunicola.’ Canadian Ent., XX XI, pp. 15-16. 1907. ‘The corn-leaf aphis and corn-root aphis.’ U. 8. Dept. Agric. Bur. Ent. Cire. No. 86, 13 pp. _ Wheeler, Ants of the Belgian Congo 581 _- Weppert, H. A. 1849. ‘Additions a la flore de I’ Amérique du Sud.’ Ann. Se. Nat., x Botanique, (3) XIII, pp. 40-113, 249-268. 1a = C.M. 1891. ‘Sixth contribution to a knowledge of the life history of certain Bs little known Aphidide.’ Bull. Illinois State Lab. Nat. Hist., II, ese pp. 207-214. ‘3 Wass, F.E. 1908. ‘The dispersal of fruits and seeds by ants.’ New Phytologist, VII, pp. 23-28. a 1909. ‘The dispersal of the seeds of the gorse and the broom by ants.’ New ; . Phytologist, VIII, pp. 81-89. & eee: H. 1913. ‘Vorratskammern der Tiere.’ Kosmos, Stuttgart, X, pp. 252- 254. WESMAEL, C. 1838. ‘Sur une nouvelle espéce de fourmi de Mexique.’ Bull. Ac. Se. E Belgique, V, pp. 766-771, Pl. xrx. Wernerttt,C.M. 1852. ‘Chemical investigation of the Mexican honey ant.’ Proc. Acad. Nat. Sci. Philadelphia, VI, pp. 111-112. Werrsrais, R. v. 1888. ‘Ueber die Compositen der dsterreichisch-ungarischen Flora mit zuckerausscheidenden Hiillschuppen.’ Sitzb. Ak. Wiss. * . Wien, math. naturw. K1., XCVII, Abt. 1, pp. 570-589. 7 1889. ‘Pflanzen und Ameisen.’ Vortr. Ver. Verbreit. Naturw. Kenntn. Wien, XXIX, pp. 307-327. WHEELES, W.M. 1900. ‘A new myrmecophile from the mushroom gardens of the ‘Texan leaf-cutting ant.’ American Naturalist, XXXIV, pp. 851-862. 1901a. ‘Compound and mixed nests of American ants.’ American Nat- uralist, XX XV, pp. 431-448, 513-539, 701-724, 791-818. 1901b. ‘Notes on Mexican ants collected by W. M. Wheeler.’ Ann. Soc. Ent. Belgique, XLV, pp. 199-205. 1902a. ‘A new agricultural ant from Texas. With remarks on the known North American species.’ American Naturalist, XXXVI, pp. 85-100. 1902). ‘New agricultural ants from Texas.’ Psyche, IX, pp. 387-393. 1904a. ‘The American ants of the subgenus Colobopsis.’ Bull. American Mus. Nat. Hist., XX, pp. 139-158. - 1904b. ‘Some further comments on the Guatemalan boll-weevil ant.’ Science, N.8., XX, pp. 766-768. 1904c. ‘On the pupation of ants and the feasibility of establishing the Guatemalan kelep or cotton-weevil ant in the United States.’ Science, _ N.S., XX, pp. 437-440. 1905a. ‘Dr. O. F. Cook’s “Social organization and breeding habits of the cotton-protecting kelep of Guatemala’’.’ Science, N. 8., XXI, pp. 706-710. 1905b. ‘The ants of the Bahamas, with a list of the known West Indian species.’ Bull. American Mus. Nat. Hist., XXI, pp. 79-135, PI. vi. 1905c. ‘An annotated list of the ants of New Jersey.’ Bull. American Mus. Nat. Hist., XXI, pp. 371-403. 1906a. ‘The kelep excused.’ Science, N.S8., XXIII, pp. 348-350. 1906). ‘An ethological study of certain maladjustments in the relations of ants to plants.’ Bull. American Mus. Nat. Hist., XXII, pp. 403-418, Pils. uxit—_xvut. 1906c. “The habits of the tent-building ant (Cremastcgaster lineclata Say).’ Bull. American Mus, Nat. Hist., XXII, pp. 1-18, Pls. 1-v1. 582 Bulletin American Museum of Natural History (Vol. XLV 1907. ‘The fungus-growing ants of North America.’ Bull. American Mus. Nat. Hist., XXIII, pp. 669-807, Pls. xurx—nn1, 1908a. ‘The ants of Porto Rico and the Virgin Islands.’ Bull. American Mus. Nat. Hist., XXIV, pp. 117-158, Pls. x1—xu. 1908). ‘Honey ants, with a revision of the American Myrmecocy.ti.’ Bull. American Mus. Nat. Hi:t., XXIV, pp. 345-397. 1910a. ‘Colonies of ants (Lasius neonizer Emery) infested with Labculbenia formicarum Thaxter.’ Psyche, XVII, pp. 83-86. 1910b. ‘Ants, their structure, development, and behavior.’ New York. xxv-+663 pp. (Ch. XVI. ‘The harvesting ants,’ pp. 267-293; Ch. XVII. ‘The relations of ants to vascular plants,’ pp. 294-317; Ch. XVII. ‘The fungus-growing ants,’ pp. 318-338; Ch. XIX. The relations of ants to plant lice,’ pp. 339-360; Ch. XX. ‘Honey ants,’ pp. 361-377). 1910c. ‘The North American forms of Camponctus fallax Nylander.’ Journ. New York Ent. Soc., XVIII, pp. 216-232. 19lla. ‘Descriptions of some new fungus-growing ants from Texas, with Mr. C. G. Hartman’s observations on their habits.’ Journ. New York Ent. Soc., XIX, pp. 245-255, Pl. vu. 1911b. ‘Two fungus-growing ants from Arizona.’ Psyche, XVIII, pp. 93— 101. 1912a. ‘Notes on a mistletoe ant.’ Journ. New York Ent. Soc., XX, pp. 130-133. 1912b. ‘Additions to our knowledge mi the ants of the genus Myrmecocystus Wesmael.’ Psyche, XIX, pp. 172-181. 1913. ‘Observations on the Central American acacia ants.’ Trans. 2nd Congr. Entom. Oxford, (1912), II, pp. 109-139. 1914. ‘The ants of the Baltic amber.’ Schrift. Phys.-Okon. Ges. Kénigs- berg, LV, pp. 1-142. 1915. ‘On the presence or absence of cocoons among ants, the nest-spin- ning habits of the larve and the significance of the black cocoons among certain Australian species.’ Ann. Ent. Soc. America, VIII, pp. 323-342. 1916. ‘The Australian ants of the genus Aphznogaster Mayr.’ Trans. Roy. Soc. South Australia, XL, pp. 213-223, Pls. xx1—xxm. 1919. ‘The ants of Borneo.’ Bull. Mus. Comp. Zoél. Cambridge, Mass LXIII, pp. 43-147. 1921. ‘A new case of parabiosis and the “ant gardens”’ of British Guiana.’ Ecology, I, pp. 89-103. Wuee ter, W. M. anv Battey, 1. W. 1920. ‘The feeding habits of pseudomyrmine and other ants.’ Trans. Amer. Phil. Soc. Philadelphia, N.8., XXII, pt. 4, pp. 235-279, Pls. 1-v. Wuire, F.B. 1872. (Seed-storingants.) Trans. Ent. Soc. London, Proc., pp. v-vi. Wierenca,O.M. 1916. ‘Waarschuwing tegen de Anggrang (roode mier).’ Meded. Proefstation Malang, Soerabaya, No. 13, pp. 15-16. Winker, H. 1905. ‘Einige tierische Schidlinge an Kakaofriichten’ (with zoological notes by L. Reh). Zeitschr. f. Pflanzenkrankh., XV, pp. 129-137. 1906. ‘Beitrige zur Morphologie und Biologie tropischer Bliten und Friichte.’ Engler’s Bot. Jahrb., XX XVIII, pp. 233-271. OM _____ Wheeler, Ants of the Belgian Congo 583 -_-:1908. ‘Der Laubwechsel in unseren Breiten und in den Tropen.’ Aus der Natur, III, pp. 654-661. 1913. ‘Die Pflanzenwelt der Tropen.’ ‘Das Leben der Pflanzen,’ VI, -—s—=is*é«éS tutttgart, pp. 245-523 (Ameisen und Pflanzen, pp. 383-398). Woexten, H. anp Zimmer,-C. 1912. ‘Eine akademische Studienfahrt nach Ost- SE oe: afrika.’ Breslau. 120 pp. Wertacm, E. 1882a. ‘Zur Anatomie der Aphiden.’ Zool. Anzeiger, V, pp. 239-241. _-~—«1882b. ‘Zur Anatomie der Aphiden.’ Arb. Zool. Inst. Wien, IV, pp. 397— as 441 (abstract in Journ. Microsc. Soc., III, 1883, pp. 49-51). waa Wostem, R. S. anp Borpen, A. D. 1919. ‘A comprehensive demonstration of ee. 4) Argentine ant control.’ The California Citrograph, IV, No. 6, p. 147. ‘aa Wooowowr, C.W. 1908. ‘The Argentine ant in California.’ Cire. 38 California ar en Agric. Exper. Stat., p. 11. me : 1910. ‘The control of the Argentine ant.’ Bull. 207 California Agric. Exper. Bs Woonworrs, E. 1920. ‘Damage to nursery plants by the fire ants (Solenopsis eae es. geminata subsp. manicsa Wheeler).’ Monthly Bull. California State oe ae Dept. Agric. Sacramento, IX, No. 3, pp. 87-88. Waovanmos, RC 1892. ‘Our ants.’ Journ. Bombay Nat. Hist. Soc., VII, pp. 13-60, 175-203, Pls. A—D. i R. H. 1902. ‘Two Malayan “myrmecophilous” ferns, Pclypcdium (Lecan- ¢pleris) carn sum (Blume), and Pclypcdium sinuesum Wall.’ Ann. of Botany, XVI, pp. 185-231, Pls. x-xm. — Zamtenmann, — 1896. ‘Untersuchungen iiber die Pilzgirten stidamerikanischer % Ameisen.’ Ber. Naturw. Ges. Chemnitz, XIII, pp. xix—xx. ‘ZIMMERMANN, A. 1908. ‘Ueber Ambrosiakifer und ihre Beziehungen zur Gummi- bildung bei Acacia decurrens.’ Centralbl. Bakter. Parasitenk., X, Abt. 2, pp. 716-724. I * oo Fa + yt . . , 4 ‘ a a « ‘ ee t * 2 M « ‘ . y : ip 4 zy, ' i’ ates 3 } eae a 3 ”* ¥ ; . 1 t Ad Fig. 1. Nest of harvesting ant (Aor specie) in Beurerin A. M. N, H. Vou. XLV, Prare XXVI 1914. Extremity of a branch Many of it. ‘under side part of photograph) Fig. 2. Cola Laurentii De the leaves ITAXX ®491d ‘ATX “TOA HN IVY Steering the branch; ants. Medje, October, 1910. Bourerm A. M.N. H. Vout. XLV, Pirate XXVIII October, 1910. Characteristic are the were from this specimen. Bouizierix A. M. N. H. Vor. XLV, Pratre XXIX a : V.—THE ANATOMY OF CERTAIN PLANTS FROM THE BELGIAN CONGO, WITH SPECIAL REFERENCE TO MYRMECOPHYTISM By Irvine W. Bartey INTRODUCTION ____ In examining a dried specimen of the myrmecophytic Vitex Staudtii _ Guerke collected by Lang and Chapin at Medje in the Belgian Congo, I was impressed by the close correlation between the distribution of the lateral cavities (supposed excavations) and the phyllotaxy of the plant. Through the courtesy of my colleague, Professor Wheeler, to whom the material had been sent by the collectors, I secured the opportunity of sectioning nodes and internodes of portions of the specimen and of study- ing their anatomical structure. So many features of unusual interest were encountered that it seemed advisable to study in detail the anatomy of other myrmecophytes from tropical Africa. All of the available material of ant-inhabited species of Sarcocephalus, Barteria, Plectronia, and Cuviera secured by Dr. Bequaert and Messrs. Lang and Chapin was very kindly turned over for my use. The myrmecodomatia of these genera proved to be fully as interesting as those of Vitex. Indeed, in so far as the anatomy of the host plants is concerned, the African myrme- cophytes are even more remarkable than the much discussed Acaciz, Cecropiz, and Myrmecodiz and are specially significant in any general consideration of myrmecophily. Minute ANATOMY OF AFRICAN MYRMECOPHYTES Vitex Staudtii Guerke' The taxonomy and general gross morphology of this verbenaceous liana are described by Dr. Bequaert on pages 447-452. The plant exhibits a typical decussate phyllotaxy. In other words, pairs of opposite leaves alternate with each other at right angles, resulting in four vertical rows or orthostichies of leaves. As is frequently the case in plants having this typeof phyllotaxy, the stems are four-sided and quadrangular, each side corresponding to one of the orthostichies of leaves. At each node there are two circular apertures which are located at approximately the same level on opposite sides of the stem and midway between the leaf-scars or points of attachment of the leaves (Fig. 88e). These orifices, which serve as exit holes for the ants, shift their position from one pair of opposite sides of the stem to the alternating pair at each succeeding internode. 'The following avcount of the anatomy and histology of this is based w the study of material (No. 743) collected by Lang and Chapin at Medje, July, 1914." om 585 586 Bulletin American Museum of Natural History (Vol. XLV When the stem is split open it is found to be hollow. The central cavity is continuous from internode to internode, but is considerably constricted just above each node by a thicker peripheral layer of medullary tissue (Fig. 88, A-A). In addition to the nodal apertures which communicate with tlie exterior, there are numerous lateral internodal cavities which per- forate the xylem and end blindly just under the bast. They are arranged symmetrically, one over the other, in longitudinal rows subtending the nodal apertures (Fig. 881). In other words, there is an obvious and close correlation between the distribution of the lateral cavities and the phyllo- taxy of the plant. In specimens which are inhabited by Viticicola tessmanni (Stitz) Wheeler, the conditions described above are characteristic of all portions of the stems and branches having fully elongated internodes. Only during the earlier stages in the differentiation of the nodes and internodes —in relatively close proximity to the growing-points or apical meristems —are they filled internally with medullary tissue and devoid of lateral cavities. The longitudinal and lateral cavities are also absent in very young plants that are not inhabited by ants. Such facts as these suggest that the central cavities and their bisymmetrically arranged lateral rami- fications are excavated by the ants, which leave the supernodal constric- tions or projecting rings of medullary tissue to separate the broods and prevent them from falling downward into the lower internodes. In his field studies of Vitex Staudtii, Dr. Bequaert was unable to find any external evidences of preformed structures like the “pros- tomata” of Cecropia, described by Fritz Miiller (1880-81), Schimper (1888) and others, which might account for the curiously symmetrical arrangement of the nodal apertures and lateral internodal cavities; nor was he able to find any clue to the functional significance of the latter. They are not used as receptacles for eggs, larve, or pupx, nor do they contain coccids or other organisms. The only explanation that suggested itself was that they might be constructed for purposes of aération. This suggestion is negatived, however, by the structure of those portions of the “bark”? which cover the outer ends of the supposed lateral excavations. © There are no lenticels or patches of aérenchyma suitably located, through which air might readily penetrate into the interior. On the contrary, the overlying tissues are compact and devoid of conspicuous intercellular spaces, and, in old stems, there are thick disks of impervious scleren- chyma situated just opposite the ends of the lateral cavities (Pl. XXXII fig. 4). Wheeler, Ants of the Belgian Congo 587 “ Score Fore Cavities A; bits stout, dry stems and branches of Viter Staudtii, the lateral ; Bet appes: superficially to be galleries or pits excavated in the woody portion of the “central cylinder,” or stele. As is shown in Pl. XXXII, fig. 4, they perforate the xylem and end blindly on a general level with the cambium. However, a more detailed and critical study of the histology of the tissues figured in this photomicrograph indicates con- clusively that the cavities are not mere tunnels in the normal xylem, : - since there is no fringe of chewed and broken prosenchyma such as occurs in galleries excavated by wood-boring insects. The outer two- _ thirds of the cavity are jacketed by the remains of a layer of very thin- _ walled, unlignified cells (TPa). In the xylem this layer rests upon _ heavily lignified, thick-walled parenchyma (LXmPa) which, in turn, __ merges more or less abruptly into prosenchyma (XmPr) (Pl. XXXII, eae fig. 3). That these layers were not formed subsequent to the excavation __ of a cavity in the prosenchyma is shown by the fact that there is no unconformity—torn or broken libriform fibers—between the paren- chyma and surrounding prosenchyma. What, then, is the mode of formation of the lateral cavities? Are they preformed structures or galleries excavated from a pre-existing core of delicate tissue? The section, illustrated in Pl. XX XIII, fig. 1, which - was made from a freshly cut stem preserved in alcohol is of considerable interest in this connection. The soft tissues are in situ and have not contracted and collapsed as in preparations made from dried specimens. The lateral cavity does not extend to the general level of the surface of the woody cylinder, but its convex end subtends a plug of soft tissue (CT) which projects into the xylem. The inner portion of this intruding mass of unlignified tissue is jacketed laterally by an extension of the layer of thin-walled, unlignified cells (TPa) previously mentioned, which re- semble in size, shape, and orientation those of the adjoining layer of heavily lignified xylem-parenchyma (LXmPa). Externally, the central core is constricted more or less by the phloem and cortex, into which it gradually merges. Since its outer margin is somewhat convex and its inner surface is concave, it forms a roughly dome-shaped layer of con- siderable thickness. Dark-colored, elongated, conducting cells, ramify- ing from the phloem (Pm), proliferate through it, and, as is shown in Pl. XXXIII, fig. 2, the thin-walled isodiametric cells which form its ground mass are arranged more or less symmetrically in radiating rows. This is particularly true of its inner portion (NL), where the cells are very much smaller and of more uniform size and shape (Pl. XX XIII, ee ee ae 588 Bulletin American Museum of Natural History (Vol. XLV fig. 4). In examining the photomicrographs, it should be noted that the concave inner margin of this intruding core of delicate, unlignified tissue is irregularly serrate (Pl. XX XIII, figs. 2 and 4) and shows unmistakable evidences of having been gnawed by the ants. Small chunks of tissue have been nipped out of the surface, leaving a fringe of torn and injured cells. This suggests that the lateral cavities are not natural gaps in the woody portion of the central cylinder but are galleries excavated by the ants in peculiar cores of delicate, unlignified tissue, that are symmet- rically distributed in certain radii of the stem. The question presents itself, accordingly, as to whether these paren- chymatous areas of the xylem are normal features in the anatomy of Vitex Staudtii or traumatic structures produced by the ants. They have the appearance of abnormalities and resemble certain complex zoécecidia or so-called prosoplasmatic galls, with their histological differentiation into ‘nutritive,’ ‘‘mechanical,’”’ and “conducting’’ tissues. The similarity becomes very striking indeed when these peculiar structures are studied microchemically. The cells—septate libriform fibers and parenchyma—of the normal xylem are packed with starch (Pl. XXXII, figs. 1 and 3) and are separated from the core of thin-walled tissue by the layer of thick-walled, heavily lignified parenchyma corresponding to a mechanical layer, which is devoid of starch, as is the thin-walled un- lignified parenchyma which adjoins it. The dome-shaped, central core of soft tissue is abundantly supplied with proteins and fats, which reach their highest concentration in the cloudy protoplasts of the small, regularly arranged cells of the inner zone (Pl. XXXII, fig. 5). In stout stems the cap or disk of sclerenchyma, which is formed by the periderm and overlies the soft core of unlignified tissue (Pl. XXXII, fig. 4), corresponds to a second mechanical layer. It is evident, accord- ingly, that in the tissues which surround the lateral cavities there are the equivalents of a “starch layer,” a “protein-fat,” or “nutritive” layer, two ‘‘mechanical” layers, and a ramifying system of conducting cells. _ As in certain insect galls,' the starch and protein-fat containing tissues are separated by a layer of heavily lignified cells, and the second, mechanical or “ protective” layer is situated near the exterior. However, in view of the fact that ants are not known to produce gall-heteroplasias, particularly of this highly: differentiated and structurally complex type, more critical evidence is required before these structures can safely be considered to be of traumatic origin and due solely to the activity of is. ants. q pare cvnipid and other prosoplasmatic zoocecidia described by Houard (1903), Kister (1903, 1911)-c Coens (1912), and others. ibecier, ate of Gia Belgian Conge 589 ae ae | have al already nied that the lateral cavities do not contain coccids oz cxithe larve of gall-making insects; this is true of young as of old, stout stems and branches. Furthermore, the tissues in question do not appear to contain bacteria or other micro-organisms. In very young, tender = stems, just subsequent to the formation of the longitudinal central — eavity, the flat sides of the stele, or so-called fibrovascular cylinder, are iz tous and devoid of gaps or oval patches of delicate tissue, sz and the peripheral layer of medullary tissue is homogeneous and entire (PL. XXXII, fig. 2). At a little later stage in the differentiation of the * internodes, the lateral cavities make their appearance. Macroscopic __ and microscopic studies of the tissues in the interior of such internodes __ reveal very clearly the mode of origin of the oval lateral pits. Patches of _ the medullary tissue are ripped and torn away and the prosenchyma is perforated, revealing the cambium. The exposed portion of the latter ___ divides actively, producing callus, which projects toward the interior of the stem (Pl. XXXIV, fig. 2). With further increase in the size of ____ the stem, the cavities, which are somewhat irregular at first with jagged __ margins, are enlarged by the removal of additional elements of the medulla and prosenchymatous xylem and are smoothed till they finally assume their characteristic symmetrical, oval outlines (Pl. XXXII, fig. 3). These facts indicate very clearly not only that the lateral cavi- ties are excavated by the ants, but that the peculiar tissues that sur- round them are abnormalities, comparable to zodcecidia or heteroples- matic galls. Although the heteroplasias are relatively simple at their inception, they soon become complex and highly differentiated from the histological point of view. Thus, at first, there is merely a simple callus, which projects into a perforation in the prosenchymatous xylem. This is accompanied by more or less hypertrophy of the cells of the overlying cortex and a slight retardation in the development of the subepidermal periderm (Pl. XXXIV, fig. 2 and Pl. XXX, fig. 2). As growth proceeds, _ other structural abnormalities make their appearance. Owing to trau- matic stimuli, the peripheral layer of meristematic cells of the cambium, adjoining the callus, ceases to form prosenchymatous xylem and lays _ down thick-walled, heavily lignified parenchyma next to the prosen- _ ehyma and thin-walled, unlignified tissue next to the callus. As the mass of the callus increases, extensions of the phloem proliferate through it, accelerating a flow of nutritive substances to its innermost portions. When the overlying periderm is differentiated, it remains for a time un- modified or only slightly modified in structure (Pl. X XXIII, fig. 1), 590 Bulletin American Museum of Natural History [Vol. XLV but later forms a disk of very dense, heavily lignified tissue, or scleren- chymatous cap, opposite the central core of delicate cells (Pl. XXXII, fig. 4). As the woody cylinder increases in diameter, the ants continu- ally, 7. e., at relatively frequent intervals, excavate the inner margin of the intruding mass of callus, for if they do not do so the lateral cavities become occluded by wound-wood. Such occluded galleries are of com- mon occurrence, particularly in long internodes having very numerous heteroplasias and small broods. The next question to be considered then is, why are these zoécecidia so symmetrically arranged in obvious correlation with the decussate phyllotaxy of the plant? Pl. XXX, fig. 1 illustrates a transverse section of a young internode, cut just ABOVE a node (at the level A—A of Fig. 88). The four sides of the quadrilateral stem are histologically similar, with one notable exception. The vessels or principal water-conducting pas- sageways are largely concentrated in one pair of opposite sides of the stele. A similar condition is shown in Pl. XXX, fig. 2, a transverse section cut just BELOW the node (at the level B—B, Fig. 88), but the vessels in this case are aggregated in the alternating pair of sides. In other words, the principal water-conducting passageways in each inter- node are largely confined to those sides of the stele which are to pass out to the leaves at the next (higher) node. Therefore, their orientation changes at each succeeding internode in accordance with the decussate phyllotaxy of the plant. As shown in Pl. XXX, fig. 2, the lateral cavities are excavated in the sides of the stele which are poorly supplied with vessels. Furthermore, they are located in those portions of the xylem which are devoid even of a narrow fringe of small primary trachee (PT). Why should the ants select these radii of the stem for the construe- tion of the lateral pits? Of course, the breaking of the conduits would certainly interfere with the normal flow of water to the leaves, and, inas- much as in vines and lianas the area of water-conducting tissue is rela- tively small in proportion to the area of transpiring leaf-surface, this might affect the normal physiological processes of the plant and even lead to the drying up of the leaves. It seems probable, however, that the ants avoid the vessels because when the conduits are ruptured there is an excessive flow of water. That the cambia on the four sides of the stem are equally capable of producing the hyperplasias, is indicated by the fact that when the ants make an excavation in the wrong surface, as occasionally happens, it results in the formation of a heteroplasia which resembles those that occur so abundantly in the alternating pair of sur- faces. Wheeler, Ants of the Belgian Congo 591 Exit Apertures bile “— stout stems the inner portions of the exit holes resemble struc- tar thw of the lateral pits which subtend them. The prosenchymat- ; Raticatty below the general level of the woody cylinder (Pl. XXXIV, 2 fig. 1). In this depression rests a-peculiar ring of sclerenchyma (PI. XXXIV, figs. 1 and 3) which extends to the outer surface of the stem and q _ ‘amuaally projects considerably beyond it, so that the nodal apertures have be externally an embossed or crater-like profile. These rings of extremely dense, tough tissue make their appearance in young stems and un- ____ doubtedly tend to prevent the cambium and cortex from forming lateral __ eallus which would seal the exit hole unless periodically removed by the __ants.! Similar structures may be formed in the internodal portions of the plant whenever, as occasionally occurs, one of the lateral galleries is ___ extended beyond the cambium, through the cortex and epidermis, to the __ exterior of the stem. Under these circumstances the usual sclerenchym- __ atous disk (Pl. XXXII, fig. 4) is replaced by a projecting ring of ‘selerenchyma (Pl. XXXTV, fig. 1). I have already called attention to the bisymmetrical arrangement of these nodal apertures and their obvious relation to the decussate phy]- lotaxy of Vitex Staudtii. At each node there are two exit holes excavated on opposite sides of the stem and at approximately the same level (Fig. 88). It is interesting to note in this connection that in the verbenaceous myrmecophyte, Clerodendron fistulosum, described and figured by ____ Bececari (1884-86), the nodal apertures’ are located just below the points of attachment of the leaves, in the alternating pair of surfaces of the _ stem. This striking contrast in the location of the exit holes in the two myrmecophytes is due apparently to differences in their nodal and inter- nodal anatomy. In Vitex Staudtii two entire sides of the stem pass out into the pair of opposite leaves at the node. These sides—‘‘leaf traces” ___ which are considerably narrower than the alternate pair (Pl. XXX, _ fig. 2) become more and more abundantly supplied with vessels in the vicinity of the node. With the passing out of the leaf-traces, leaving two large gaps in the stele, there is an abrupt transition in the remaining sides of the stele from xylem that is nearly devoid of vessels to woody In a number of other vies that T have studied the exits or entrances soon, become veehimnper (588) by wound-wood unless hegt ope open by the ants. apertures arise spéntancoualy and con- (1888) questions Becrar's statement that t 592 Bulletin American Museum of Natural History (Vol. XLV tissue that is crowded with water-conducting passageways. Therefore, the most favorable situations for the location of the nodal apertures are to be found midway between the points of attachment of the leaves and just below the level where these structural transitions occur. In the stems of certain species of Clerodendron, on the other hand, the elements of the xylem are differentiated and distributed in such a manner that the “‘prostomata”’ subtend the bases of the leaves. The gaps left in the woody cylinder by the exit of leaf-traces are filled with very delicate, soft tissue. A priori, it would seem as if these gaps would be the most favorable places for the excavation of the exits. It should be noted, however, that if they were so utilized there would inevitably result a destruction of the axillary buds and the formation of lateral branches be prevented. In Vitex Staudtii, although the ants occasionally remove a portion of the tissue, I have never seen a node in which they had cut their way through to the exterior. Yet, as soon as the buds form branches, the cavities in the latter are found to communi- cate freely with those of the main axes. From the point of view of insect behavior, it would be extremely interesting to discover whether there are structures or substances in the buds which prevent their destruction by the ants. Origin of the Central Cavities Many of the earlier investigators assumed that the central chambers of various myrmecodomatia are excavated by ants. Thus, Rumphius (1750) and others considered that the anastomosing galleries in the ‘“pseudobulbs” of Myrmecodia and Hydnophytum are constructed by ants. There is a considerable element of danger, however, in making hasty generalizations in regard to the origin of these structures. Forbes (1880), Treub (1883), and others have shown that the domatia of Myr- mecodia and Hydnophytum occur normally in plants from which ants are entirely excluded. Furthermore, it is well known that plants, both herbaceous and arborescent, having hollow internodes are widely distrib- uted in both temperate and tropical regions. It has been stated that the stems of very young seedling plants of Vitex Staudtii, which are not inhabited by Viticicola, are filled with medullary tissue, whereas those of older plants, which are occupied by the ants, are hollow except in the vicinity of the growing points. This might be considered to indicate that the domatia are excavated by Viti- cicola. It must be admitted, however, that evidence of this character is purely circumstantial and not necessarily conclusive. Hollow internodes _ Wheeler, Ants of the Belgian Congo 593 - may be pres present in bie vigorous adult shoots when they are entirely _ absent in smaller stems, such as are frequently formed during the earlier q oma in the ontogeny of the plant or under unfavorable growth condi- tions. Furthermore, in examining herbarium material of other represen- Ry _tatives of this genus, I find that, although certain species possess solid ss stems, others normally have well-developed central cavities in their core of medullary tissue (Pl. XXXI, fig. 1). The structure of the delicate, ss immature internodes of ant-inhabited specimens of Vitex Staudtii is of 4 ~ interest in this connection: Pl. XXXII, fig. 1 illustrates a transverse ___ section of such an internode cut a relatively short distance behind the ____ terminal growing-point. The medulla is not homogeneous but consists ____ of an oval central core of very large, thin-walled cells and a peripheral layer of denser medullary tissue, which is richly protoplasmic. As the stem increases in diameter the oval core of delicate tissue shrivels up and is trimmed away by the ants leaving the oval cavity shown in Pl. XXXII, fig. 2. From the point of view of this histological evidence, it seems probable that in Viter Staudtii there is an inherent tendency towards the formation of hollow stems and branches. Whether the ants accelerate the formation of the central cavities, as has been shown by Fiebrig (1909) to be the case in Cécropia, can only be determined by careful field observations. Cuviera A number of species of the African, rubiaceous genus Cuviera are myrmecophytic and characterized by having elongated, spindle- shaped swellings on the branches (Figs. 99 and 100) which are inhabited ___ by ants.' Two lots of these myrmecodomatia, one collected by Dr. Bequaert at Stanleyville, March 8, 1915 and the other (No. 1031) by Messrs. Lang and Chapin at Kunga near Malela, July, 1915, were sent to me for anatomical study. The former, occupied by Crematogaster africana subspecies laurenti variety zeta (Forel), were obtained from Cuviera angolensis Hiern; the latter, inhabited by Crematogaster impres- siceps variety frontalis Santschi, from an unidentified species of Cuviera. The myrmecodomatia of the two species are similar and differ only in certain minor morphological and histological details. Externally, those of Cuviera angolensis are shorter, slimmer, and of a deep olive green color, whereas the others are longer, stouter, and of a reddish green color The color differences are due largely to differences in the internal a) 4See Schumann (1891). 3In making these comparisons I am dealing with material preserved in alcohol 594 Bulletin American Museum of Natural History [Vol. XLV distribution of a translucent, amorphous, amber-colored substance com- monly referred to as “‘tanniniferous”’ or “resiniferous,”” which occurs normally in many of the Rubiacezw. As is shown in Pl]. XXXV, fig. 1, there is a considerable concentration of this amber-colored substance in the subepidermal and other cortical cells of the unidentified species of Cuviera, which gives to the branches their reddish tinge. In C. angolensis, on the other hand, the cortex and bast are relatively free from it (Pl. XXXV, figs. 2 and 3; Pl. XXXVI). The structure and shape of the core of medullary tissue is not the same in the two species. In the former (Pl. XXXYV, fig. 1), the cells which contain the amber-colored substance are aggregated in the center of the more deeply lobed medulla, whereas in the latter (Pl. XXXV, fig. 2) they are diffused, with a peripheral row scattered along the inner margin of the stele. . Mode of Origin of the Myrmecodomatia As has been shown by Dr. Bequaert (Part IV, p. 490), the myrmeco- domatia are not abnormalities produced by ants or gall-forming organ- isms. They are preformed, localized, hollow hypertrophies of the branches that become occupied by ants, or occasionally by the larve of certain beetles, after they have become fully differentiated. This is _ shown very clearly by the structure of swollen branches which have no entrance holes and are entirely devoid internally of insects or other organisms during the various stages of their “ontogeny.” Pl. XXXVI, fig. 1 illustrates a transverse section of the swollen, lower portion of a young, tender internode. The cortex, stele, and medulla are already clearly outlined, but their finer histological details are’ still in process of differentiation. The pith is not homogeneous, as is the case in the un- swollen, upper portion of the internode, but consists of a large central core of succulent pulp and a peripheral layer of denser tissue. Further- more, it is evident that the differentiation of the so-called fibrovascular cylinder is not proceeding uniformly, since two opposite sides of the stele are considerably thicker than the intervening pair. As the lower part of the internode becomes more and more hypertrophied, the juicy delicate tissue in its interior gradually collapses and dries up; except for a fringe of thin-walled cells, filled with the amber-colored, hyaline substance, which jacket the inner margin of the peripheral layer of thick-walled medullary tissue (Pl. XXXVI, fig. 2). Wheeler, Ants of the Belgian Congo 595 Entrance Holes | i; Although the circular apertures, through which the ants gain access ae an interior of the swollen internodes, are not so regularly arranged as ee in Vitex Staudtii, yet they are more or less symmetrically oriented in relation to the somewhat distorted—by torsion—decussate phyllotaxy Uae of the plant, and shift their position regularly at each succeeding inter- 4 ; node. This i is determined apparently by the structure of the walls of the Bere r tia. As is shown in P]. XXXV, fig. 3 and Pl. XXXVI, x fig. 1, the four sides of the swollen portion of the branch are not similar BS. - either i in size or in thickness. The ants almost invariably make their __ primary excavations in the thin side of the stem which has the widest y surface. Therefore, since the position of this surface is closely related ; ins to the decussate phyllotaxy, the orientation of the entrance holes shows a ae similar correlation. ___ The structure of the tissues surrounding the entrance holes indicates , ‘that the ants make the perforations subsequent to, or during the later stages of, the formation of the central cavities, and before the cambium an _ has Sochied numerous ee elements in the thinner sides of the swollen internodes. Lateral Excavations and Heteroplasias As shown in Figs. 99b and 100a-b, the myrmecodomatia of Cuviera are characterized internally by a number of elongated or oval pits which are located in the thinner sides of the branch. These depressions, which are arranged in one or two longitudinal rows, or more irregularly scat- tered, are commonly occupied by coccids of different ages and sizes.* In certain myrmecodomatia each pit contains a coccid which more or less completely fills it, whereas in others only two or three of the pits are so occupied, or the coccids may be entirely absent. When studied microscopically, these pits are found to contain growths of thin-walled callus (Pl. XXXV, fig. 3; Pl. XX XVII, fig. 2) which recall the dome-shaped callus formations that occur in the lateral galleries of Vitex Staudtii. As in the latter, the cells are arranged in F _ radiating rows but, instead of containing cloudy, opaque protoplasts, _ are filled with the golden yellow, or reddish brown, hyaline substance which occurs normally in certain cells of the cortex, phloem, xylem, and medulla. In addition, it may be noted that the akan margin 'The coccids in Cuviera angolensis have been identified by Newstead as Peeudococeus citri variety congoensia Newstead. fa BL oe a! . ta ie y 596 Bulletin American Museum of Natural History (Vol. XLV of the xylem is not jacketed by two clearly defined layers of parenchyma, and there is no elaborate, ramifying system of conducting cells (PI. XXXVII, fig. 2). The structure of these heteroplasias, as well as their entire absence in swellings which have not been occupied by insects, indicates that they are abnormalities produced by traumatic stimuli. The question pre- sents itself, accordingly, are they due to the activities of the ants or the coccids? That they serve as food-reservoirs for the coccids is shown by the fact that the sucking mouth-parts of the latter are embedded in the callus. The torn and chewed inner margin of the heteroplasias (Pl. XXXVII, fig. 1) suggests, however, that these layers are also fed upon by the ants. Furthermore, the pits or depressions originate as excavations in the thinner walls of the domatia and not by the collapse of soft tissue or the retardation of the growth of tissues underlying the coccids. Since the cells of the peripheral layer of medullary tissue and the first-formed elements of the prosenchymatous xylem are relatively tough and thick-walled at the time when the excavations are started, they must have been made by insects with strong mandibles and not by the delicate sucking mouth-parts of the coccids. In this connection, it should be emphasized in passing that all of the myrmecodomatia examined contained the lateral pits and heteroplasias regardless of the presence or absence of the coccids. As in Vitex Staudtii, the ants evidently cut through to the cambium and induce the formation of a nutritive callus before the stele becomes considerably thickened. “Fungus Gardens” As has been stated previously, the internal cavities of hypertrophied branches which have no entrance apertures are entirely devoid of para- sitic or saprophytic organisms. The walls of the myrmecodomatia, on the contrary, form the substratum fora more or less luxuriant growth of fungi. In recently occupied cavities, which are jacketed by a layer of cells containing the amber-colored substance, scattered aérial hyphe project into the interior from all sides (Pl. XX XVII, fig. 3). In the case of the domatia of the unidentified species of Cuviera, which contain numerous ants, the hyphx and amber-colored substratum have been more or less cleanly gnawed away except at the two ends of the spindle- shaped cavity (Pl. XX XVIII, fig. 1) where there are dense growths of delicate, white hyphz, resembling ‘“ambrosia.’”” As shown in Pl. XXXVIII, fig. 4, there is a considerable layer of detritus between this crop of aérial hyphx and the basic substratum of medullary tissue. _ Wheeler, Ants of the Belgian Congo 597 _ - When Sablyaed microscopically, the detritus is found to consist largely _ of the more or less disintegrated castings from the infrabuccal chambers of the ants and to contain numerous nematodes (Pl. XX XVIII, fig. 3). The hyphe radiate through this layer in all directions and penetrate into the underlying medullary cells. In the myrmecodomatia of C. angolensis, which contain relatively few ants, there is a similar terminal concentra- tion of hyphz, but it is less conspicuous, owing to the fact that the lateral a: —_ of the cavities have been less thoroughly scarified. Plectronia species A ¥ ay - Among the specimens turned over to me for anatomical study were a ‘Sas myrmecodomatia, preserved in alcohol, which were collected by Dr. Bequaert near Stanleyville, March 6, 1915. They were inhabited by Engramma kohli Forel and were collected from an unnamed species of the __ rubiaceous genus Plectronia, referred to as species A in Part IV (p. a4), The myrmecodomatia of this plant resemble those of Cuviera in 4 ‘that they are preformed, localized, hollow swellings of the branches which __ are oceupied by ants (Fig. 96). Sections of the normal and abnormal _ portions of the branches are illustrated by photomicrographs on PI. XXXIX. Asin the case of Cuviera angolensis (Pl. XXXV, figs. 2 and 3), _ the pith in the former regions is homogeneous, whereas in the letter portions it is heterogeneous, consisting of a succulent central pulp, which collapses, and a peripheral layer of denser medullary tissue. Two opposite sides of the swollen branch are considerably thinner than the alternate pair, and the differentiation of the cambium and fibrovascular tissues is retarded in them. The ants pierce one of there thin walls and searify its inner surface, as well as that of the opposite side. From these injured surfaces and from the margins of the irregularly shaped apertures heteroplasias originate, which superficially resemble thore that occur in Cecropia under somewhat similar circumstances (Pl. XL, fig. 1). As shown in PI. XL, fig. 2, the heteroplasias consist of two distinct tissues— a central core of thick-walled, heavily pitted parenchyma, that is packed * with starch, as is the parenchyma of the normal and abnormal xylem; and an outer layer of thin-walled, isodiametric cells, which are filled with an amber-colored, hyaline substance such as occurs in the hetero- plasias of Cuviera. The torn and chewed inner margin of this leyer in- _ dicates that it is eaten by the ants. Furthermore, in a number of the _ myrmecodomatia small coccids were found attached to this callus-tissue by their proboscides. 598 Bulletin American Museum of Natural History [Vol. XLV In addition to the myrmecophytes that have been described in the preceding pages, and of which there was an abundance of well-preserved material suitable for anatomical study, there are four others of which un- fortunately only fragmentary material is available. However, in view of the fact that more abundant and better preserved specimens cannot be obtained in the immediate future, it seems desirable to call attention to certain features of these plants, upon which some light is thrown by the evidence at hand. Plectronia Laurentii De Wildeman The myrmecodomatia of this interesting plant have been described and figured by De Wildeman (1905-07) after E. Laurent’s notes and by Kohl (1909). Additional notes upon their gross morphology are given by Dr. Bequaert (p. 471; Fig. 95a and b). My own observations are based upon the study of two dried myrmecodomatia, collected by Dr. Bequaert near Stanleyville, March 6, 1915, which contain coecids and Crematogaster africana subspecies laurenti variety zeta (Forel). The myrmecodomatia are preformed, more or less pronounced hol- low swellings of the stems and branches, of the two general types figured on Pl. XLI. The internal cavities originate apparently by the falling away of the central succulent portion of the heterogeneous medulla. As in the case of the previously described myrmecophytic plants, two opposite sides of the domatia differ in thickness and in histological differ- entiation from the alternating pair. Entrance apertures and longitudi- nally arranged lateral excavations, which frequently coalesce to form grooves or trenches, are located in these walls and change their position at each succeeding internode in correlation with the decussate phyllo- taxy of the plant. The pits contain heteroplasias which resemble those of Vitex Staudtii in that their nutritive layers do not contain an amber- colored substance. Many of the coccids are attached to this tissue by their sucking mouth-parts. The zoédcecidia resemble those of Vitex Stavdtii in the frequent occurrence of an outer cap of sclerenchyma, - which is formed by a periderm opposite the delicate core of callus (Pl. XLI, fig. 2). The less quadrangular of the two myrmecodomatia examined by me contains “fungus gardens” at the two ends of its spindle- shaped inner cavity, which closely resemble those that occur in Cuviera. However, in these domatia the substratum of thick-walled medullary tissue, upon which the gardens rest, does not contain an amber-colored substance. The parenchyma of the woody cylinder and the cells of the medulla, except those in the vicinity of the central cavity, are densely packed with starch. Wheeler, Ants of the Belgian Congo 599 — taxonomy and eet gross morphology of these flacourtiace- a Eermmccophytes have been discussed in detail by Schumann sing Seiliich are frequently more or less fapcinted in appearance , B. Dewevrei by hollow stems and branches of normal dimen- S ~The sinterial that I have studied consists of several myrmeco- _ domatia of B. fistulosa (No. 933), collected by Messrs. Lang and Chapin __ near Stanleyville, larvee of Pachysima xthiops (F. Smith) (No. 747), ‘removed from the myrmecodomatia of B. fistulosa at Medje, and a sec- g - tion of : a hollow twig (No. 175) of B. Dewevrei, secured by Dr. Bequaert near Leopoldville. The myrmecodomatia of B. fistulosa contained _ Crematogaster buchneri subspecies biimpressa (Mayr) and Pseudococcus aie itré variety congoénsis Newstead; the hollow twig of B. Dewevrei, ae (Atopogyne) africana variety schumanni- (Mayr), and Ae Lecanium (Saissetia) barteriz Newstead. Pl. XLII, fig. 1 illustrates a transverse section of the normal, un- ae swollen, basal pcrtion of a deciduous branch of B. fistulosa. The pith is | homogeneous and consists of compact, relatively thin-walled parenchyma. The fibrovascular cylinder or stele is well developed and of normal structure. In the hypertrophied portion of the branch or myrmecodo- matium, on the contrary, the medulla is heterogeneous and the stele is feebly developed and broken up into separate strands or bundles (PI. XLII, fig. 2). A core of succulent pulp has evidently collapsed except for a layer of thin-walled cells which jacket the cavity and are filled with an amber-colored substance such as occurs in Cuviera (Pl. XXXVI, fig. 2). The amber-colored substance also occurs in cells which are scattered _ through the peripheral layer of denser medullary tissue, and in the sub- epidermal cells of the cortex. £. Laurent (1903-04; see De Wildeman, 1905-07) and Kohl (1909) found lateral pits or depressions in the walls of the myrmecodomatia of Barteria fistulosa, many of which were occu- pied by coccids. These structures are not present in the myrmecodo- matia that I have sectioned and the coccids are attached to various portions of the walls of the domatia by their proboscides. It should be noted, however, in this connection, that my material consists entirely of young, succulent branches in which processes of secondary growth are still in their incipient stages. The depressions are probably excavated in older or more woody myrmecodomatia. 600 Bulletin American Museum of Natural History _—«([(Vol. XLV A transverse section of the branch of B. Dewevret (Pl. XLII, fig. 1) resembles a similar section of the normal, unswollen portion of a decidu- ous branch of B, fistulosa (Pl. XLII, fig. 1). It differs from it, however, in two important features. The pith is heterogeneous and the internal cavity, which serves as the domatium of the ants, is jacketed by a layer of thick-walled, medullary parenchyma containing an amber- colored, hyaline substance. Furthermore, one side of the myrmecodo- matium is much thinner than the others and is nearly devoid of vessels. The entrance apertures are commonly located in this wall, as are rows of depressions or oval pits (Fig. 87d). The latter and the heteroplasias which partially fill them (Pl. XLIV, fig. 2) recall those that occur in Cuviera (Pl. XX XVII, fig. 2). The large, thin-walled cells of the callus are filled with an amber-colored, hyaline substance, which occurs normally in many of the parenchymatous cells of the medulla and stele. The coccids evidently feed on the substances in these callus-formations, for their sucking mouth-parts are embedded in them. That the lateral pits are excavations made by the ants and not depressions made by the collapse of delicate cells underlying the coccids is indicated by the struc- ture of the tissues upon the margins of the pits. Sarcocephalus species Among the myrmecophytes described by Dr. Bequaert is an inter- esting species of the rubiaceous genus Sarcocephalus (p. 460; Fig. 92). Unfortunately, the only available structural material of this plant con- sists of a dried section of a single myrmecodomatium (No. 161) which was collected at Masongo between Walikale and Lubutu. It was in- habited by Crematogaster (Atopogyne) africana subspecies winkleri variety fickendeyi (Forel) and coccids, which are identified by Newstead as Pseudococcus crassipes Newstead. As shown in Pl. XLIII, fig. 2, this myrmecodomatium differs from the others that have been described in having four thin sides which alter- nate with four thick sides. Longitudinal rows of pits or grooves, which are more or less completed filled by callus-growths, are situated in these thinner walls. Numerous coccids are attached to these heteroplasias, so that when one looks through the hollow twig there are four rows of these insects hanging from the four corners of the quadrangular cavity. The heteroplasias resemble those of Vitex Staudtii and Plectronia Laurentii, since their cells do not contain a golden or reddish brown, hyaline sub- stance; nor is the central cavity jacketed by a layer of cells which are filled with this material. The pith is heterogeneous and apparently once Wheeler; Ants of the Belgian Congo. 601 a Di aiaed a central core of succulent pulp, which has aollapaed and has ; he by the ants. 3 " vA __ Saurenr Fearures or Arrican Myrmecornyies Although these African myrmecophytes belong to three distinct Pe irs (Parietales, Tubiflore, and Rubiales) and to different growth Es forms (trees, shrubs, and lianas) they are fundamentally similar from the _ anatomical and histological points of view. In all, there is apparently an inherent tendency towards the formation of a heterogeneous pith, the central succulent portion of which collapses and dries up leaving an _ internal chamber or cavity. They are all characterized by similar _ peculiarities in the differentiation of their fibrovascular cylinders, _ which are more or less closely correlated with phyllotaxy. Certain sides or radii of the stele tend to be thinner, to contain fewer vessels, and to. _ differentiate later than others. In Cuviera, Plectronia, Sarcocephalus, and Barteria fistulosa, the internal cavities and revobatities i in the differ- entiation of the medullary and fibrovascular tissues tend to be localized. in certain shoots, or certain portions of the stems and branches, and are _ eoncomitants of more or less pronounced hypertrophies of these organs. In Vitex Staudtii and Barteria Dewevrei, on the other hand, in which these conditions are more generalized, there are no external indications of swellings or abnormal enlargements of the stem and branches. All of these myrmecophytes differ from previously described extra- _ African myrmecophytic plants in the occurrence, within their myrmeco- _ domatia, of excavations that contain peculiar callus-heteroplasias.' _ These traumatic structures, which are situated, like the entrance and _ exit apertures, in the thinner or evascularized sides of the myrmecodo- _ matia, are arranged more or less symmetrically in relation to the phyllo- _ taxy of the plants, and are formed by the young cambium and cortex _ when these tissues are exposed by the removal of the underlying cells of _ the medulla and xylem. Another unique feature of this group of myrme- | cophytes is the occurrence in the Cuviere and Plectronia Laurentii of _ “fungus gardens.” Furthermore, with the exception of Vitex Staudtii, all of the African myrmecophytes are characterized by containing more or less numerous coccids. 1Other merrrccohy tes, upon further and more gation! study, may be found to contain similar e - * of Cecropia, dearribed by Frits Maller (1880-81) and H. 4 ae (1907). spoven not located in pit-like ee resemble to a certain extent the hetero- ‘ of tified species of Plectronia. 602 Bulletin American Museum of Natural History (Vol. XLV The myrmecodomatia of the Cuviere, Barteriw, and certain of the Plectroniz resemble each other in being jacketed internally by layers of cells which are filled with an amber-colored substance. The large, thin- walled cells of the heteroplasias in these myrmecophytes are also char- acterized by being filled with this substance. The heteroplasias in Vitex Staudtii, Plectronia Laurentii, and the species of Sarcocephalus, on the contrary, are devoid of this hyaline substance. Foop or TWIG-INHABITING ANTS As is shown in the appended bibliography, a considerable number of investigators have devoted more or less attention to the study of myrme- cophytism but, in spite of the significance attached to the supposed symbiotic relation between plants and ants, there is comparatively little reliable information concerning the feeding habits of the latter and the extent to which they are dependent upon the former for food. It is true that one finds in the literature many conclusions in regard to the general feeding habits of the ants, but these are frequently mere assump- tions and are not supported by conclusive evidence. Even in the case of the much discussed “Beltian and Miillerian food-bodies” and extra- floral nectaries, although it has been definitely established that the former are collected and the latter frequently visited by the ants, there are no critical and detailed field observations or carefully planned __ experimental investigations which reveal the exact nutritional signifi- cance of these structures and the réle that they play in the feeding of larvee, queens, and various castes of workers. Many writers assume that, because a tissue which contains carbo- hydrates or other possible nutrient substances, is excavated or gnawed by ants, it serves as an important reservoir of food for these insects. Thus, a number of investigators have concluded that the ants feed upon the fresh, succulent, or more or less collapsed and dried, medullary tissue of various myrmecophytes. For example, Belt (1874) states that in the “bull’s-horn” Acaciz, “the thorns, when first developed, are soft and filled with a sweetish pulp substance; so that the ant, when it makes an entrance into them, finds its new house full of food.” Similarly, Fiebrig (1909) and H. von Ihering (1907) assume that in Cecropia the juicy pith of young internodes forms an important item of food in the diet of the Aztece. It should be noted, however, that in the absence of reliable collateral evidence, the mere fact that the medullary tissue is excavated or gnawed by the ants does not indicate necessarily that it is actually eaten by them, since it may be removed solely for the purpose of clean- ) ____ Wheeler, Ants of the Belgian Congo 603 2 Dee Ge tallarging the domatia. Thus, Fiebrig (1909) records having seen 3 Axtece, in young internodes of Cecropia, busily engaged in excavating ‘ _ the medulla and casting the fragments out of the domatia through the a ‘There is a sdiather uncertainty as to whether the more or less ‘“‘ pure 4 cultures” of fungi, that grow upon the walls of ant nests, actually are _ fungus gardens, or are mats of weeds which are periodically trimmed _ away by the ants. Emery (1899), Lagerheim (1900), Ferdinandsen and _ Winge (1908), Ferquharson (1914), J.S. B. Elliott (1915), Donisthorpe _ (1915), and others assume that various Formicide, other than the g _ Attini, are fungivorous. These investigators base their conclusions upon one or more of the following lines of evidence: (1) the association of a particular fungus with a particular species of ant; (2) the occurrence of _ the fungus in “pure cultures”; (3) the cropping of aérial and other _ hyph by the ants; and (4) analogies with the remarkable fungus- _ growing and fungus-feeding habits of the attine ants. It must be ad- _ mitted, however, that evidence of this character is not at all conclusive. - Thus, the cropping of mycelia does not prove that a fungus is eaten by an insect. Miehe (1911) found localized, }uxuriant growths of certain _ fungi in the pseudobulbs of Myrmecodia and Hydnophytum, but con- cluded that the hyphx were cropped by the ants merely to prevent them _ from occluding or obstructing the galleries. The occurrence, in ant nests, of more or less pure cultures of fungi does not indicate necessarily _ that the mycelium is actually cultivated and eaten by the insects, since, as suggested by Perkins (1914), the mats of hyphe may be purely ad- ventitious. Furthermore, there is much uncertainty in reasoning from analogy with the highly specialized, phytophagous Attini that cosmopoli- _ tan, more or less omnivorous representatives of the Formicide are fungus-farmers. F A number of investigators assume that the coccids; which occur in so many myrmecophytes, are introduced into the myrmecodomatia by the ants and tended by them as “milch cows.’’ For example, Belt _ (1874) states that in Cecropia the Aztece ‘do not obtain their food _ directly from the tree, but keep brown scale insects (Coccidse)—which _- suck the juices from the tree and secrete a honey-like fluid that exudes _ from a pore on the back and is lapped up by the ants.” Ule (1906) _ notes that most myrmecophytes contain coccids and assumes that they are brought into the domatia by the ants. FE. Laurent (1903-04; see _ De Wildeman, 1905-07) and Kohl (1909) reach similar conclusions in __ regard to the coccids in Barteria fistulosa, and the latter infers that “im 604 Bulletin American Museum of Natural History [Vol. XLV Inneren ihrer Wohnungen geben sich die Ameisen fleissig mit der Schild- liusezucht ab, in deren Exkrementen ihre hauptsichlichste Nahrung besteht.”” Fiebrig (1909), on the contrary, believes that the coecids in Cecropia “in keinem direkten Verhiltnis zu diesen Ameisen stehen,”’ and Dr. Bequaert (Part IV, p. 436) is of the opinion that the coecids find their way into the myrmecodomatia of their own accord, just as they do into other cavities (normal and abnormal) that are not inhabited by ants. It is difficult to observe the normal activities of ants in myrmeco- phytes, even when living plants and insects are available. Therefore, they must be studied in artificial nests or by indirect methods. Valuable clues in regard to the feeding habits of other animals have been obtained by analyses of feces or the contents of alimentary tracts. Unfortunately, ants (imagines) take only fluids or semifluids into their crops and stomachs. They do not masticate their food with their mandibles or maxille. The act of feeding consists in lapping or rasping nutrient sub- stances—which previously may have been dismembered or cut into fragments of suitable sizes by means of the mandibles—with the rough- ened surface of the protrusible tongue. Meinert (1860) discovered that ants, like the social wasps,’ have a curious ‘mouth sac’’ which he considered to be a crop or “‘social stomach.’’ Although Meinert’s conclusions were severely criticised by Adlerz (1886), the latter was unable to determine the true function of this infrabuccal chamber. In fact, Janet (1895, 1899) is the only student of ants who has studied the function of the infrabuccal sac with any degree of care. He demonstrated, by feeding experiments and by dusting ants with various powders, that the sac acts as a receptacle for detritus— which the imagines remove by means of their toilet organs from their _ own bodies, from their progeny, and from their companions—and food- residues. The latter substances may be of two distinct types: (1) coarse fragments of animal or plant tissues, which adhere to. the surfaces of the ant during the process of cutting up or dismembering the food and subsequently are swept into the sac; and (2) finely divided solids rasped off by the tongue and segregated from the fluid or semi-fluid substances that pass into the crop. The material that accumulates in the infrabuccal cavity in the ai of a peculiarly moulded pellet, ‘‘corpuscule enroulé”’ of Janet, is finally _ cast out as a useless residuum or fed to a larva (Pseudomyrming). 1Previously investigated by Brants (1841). _ Wheeler, Ants of the Belgian Congo 605 q ‘Although it + it was Sie eisai of these infrabuccal pellets that H. von x Thering (1898) and Huber (1905) were able to throw so much light upon the origin of fungus gardens during the founding of new colonies of attine 3 _ ants, I have not succeeded in finding a single investigator who has analyzed an extensive series of them in searching for clues concerning the 4 - feeding habits of ants. _____ The strongly hypocephalic larve of Viticicola and Pachysima are : fed with pellets voided from the infrabuccal pockets of the workers. The pellets are inserted into a curious depression or pouch (trophothylax) __ which is located on the ventral surface of the larva close to its head. _ Owing to their relatively large size and dark color, these pellets are quite = conspicuous and easily obtained from larve preserved in alcohol. . , | have been able to secure a large number of them and to *- compare Bice with pellets dissected from the heads of imagines. I : — have’ also dissected numerous workers of Crematogaster africana sub- i species laurenti variety zeta, C. impressiceps variety frontalis, C. buchneri a subspecies biimpressa, C. (Atopogyne) africana variety schumanni, and 3 _ other ants which inhabit myrmecophytes, and have analyzed the con- 3 — ws ged infrabuccal pockets. Food of Viticicola tessmanni et Cae his field studies of Vitex Staudtii, Dr. Bequaert was unable to _ secure any clue in regard to the chief sources of food of the ants, Viticicola, __ which are “obligatory”’ inhabitants of this myrmecophyte. The plants grow in wet, swampy regions and are not provided with food-bodies or _ extrafloral nectaries. Unless disturbed, the ants are not found running _ over the vegetation. They appear to remain, at least during the daytime, _ ‘in the myrmecodomatia and, as shown by my colleague, Dr. Wheeler, _ (p. 108), seem to be structurally modified for their life in the dark, tube- like domatia of Viter Staudtii. Their eyes are somewhat rudimentary for the group to which they belong and their body color is light yellowish _ brown, an unusual color in species of T'etraponera, to which V. tessmanni was originally referred. a I have shown at the beginning of this paper that certain of the _ tissues in the myrmecodomatia of Vitex Staudtii are abundantly supplied _ with nutrient substances. The cells of the outer portion of the peripheral layer of medullary tissue in young twigs are filled with starch, and the _ elements (parenchyma and libriform fibers) of the xylem are densely _ packed with grains of this carbohydrate. The ‘nutritive layer’’ of the _ gall-like heteroplasias, on. the other hand, is rich in nitrogenous sub- eee 606 Bulletin American Museum of Natural History (Vol. XLV stances and fats. The ants gnaw this protein-fat layer, but, although they clean and smooth the interior of the domatia by removing more or less of the pith, they do not appear to cut through to the starch-contain- ing cells, except at an early stage in the formation of the heteroplasias. That the latter structures provide the principal food of Viticicola tessmanni is suggested by the following facts. 1. Field observations have failed to reveal an external food supply. 2. The pellets do not contain food from an outside source. ; 3. The myrmecodomatia do not contain coccids. 4. The starch-containing tissues are not excavated extensively by the ants. 5. The heteroplasias are traumatic structures produced by the ants. 6. They resemble gall-heteroplasias and have a protein-fat layer which is gnawed by the ants. 7. The pellets in the larval trophothylaces are composed of frag- ments of this tissue (Pl. XLV, fig. 1) and bits of ant larve or triturated eggs (Pl. XLV, fig. 3). | Pellets of Pachysima The pellets in the larval trophothylaces of Pachysima xthiops and P. latifrons are composed of substances obtained both from within and without the myrmecodomatia. Almost every pellet contains either a whole coccid larva (Pl. XLV, fig. 2), or one or more chunks of an adult coccid (Pl. XLIV, fig. 3). In addition, they have a relatively large ad- mixture of fragments of medullary tissue, containing the amber-colored substance (Pl. XLV, fig. 6) ; aerial hyphe from the walls of the domatia (Pl. XLV, fig. 5); mites; nematodes; unicellular hairs of Barteria fistulosa; dirt; and numerous spores of many different types (Pl. XLIV, fig. 1; Pl. XLV, fig. 4). Occasionally, they contain bits of extraneous plant tissues, pollen, fragments of malaxated insects, ete. Pellets dissected from the heads of imagines contain a similar assort- ment of substances, but usually in somewhat different proportions. In other words, they have a smaller admixture of animal tissue. This is due, in all probability, to the fact that the worker nurses add fragments of coccids to their pellets before feeding them to the larve. Although the larve undoubtedly feed upon the fragments of coccids, it is difficult to determine how much of the miscellaneous material — is actually eaten by them. The larval stomach contains a structureless — mush, so that analyses of the contents of this organ are of little signifi- _ cance in this connection. There is a similar difficulty in distinguishing _ Wheeler, Ants of the Belgian Congo 607 between detritus and food residues in the infrabuccal chambers of the 2 _ imagines. The fragments of malaxated insects are food residues, but the _ aérial hyphe and fragments of medullary tissue may be vegetable débris, which adhered to the surfaces of the imagines during the process of x Meppins and enlarging the domatia. Pellets of Peodiglianelses The infrabuccal sacs of the Crematogasters which inhabit the Cuvierz one Plectronia Laurentii also contain substances both from within and without the domatia. Although many of the pellets have an admixture of malaxated insects, they do not contain fragments of the coecids which inhabit the myrmecophytes. Not infrequently, the pellets are composed entirely of bits of the ambrosia-like mycelia, or fragments of parenchyma ___ which are packed with this fungus. Most of the pellets, as in the case of _ Pachysima, contain malaxated medullary tissue, and dirt, pollen, hairs, spores, and other extraneous substances. ‘The infrabuccal sacs of the Crematogasters which inhabit the _ Barteriz and Sarcocephalus are more or less completely filled with the same general assortment of substances which occur in the infrabuccal pellets of Crematogaster africana subspecies laurenti variety zeta and C. impressiceps variety frontalis; except that the bits of ambrosia-like mycelia are replaced by fragments of chromogenic, aérial hyphe which grow within the ant-inhabited cavities of these myrmecophytes. Nutritive Significance of the Callus-heteroplasias The problem of determining the exact nutritional significance of the callus-heteroplasias in the Cuvierx, Plectroniez, Barteriz, and Sarco- cephalus is complicated by the presence of coccids which feed upon these tissues. In other words, the question arises as to whether these struc- tures are induced by the ants primarily for their own consumption or for feeding the coccids. Histological evidence indicates that the nutritive layer is gnawed and rasped by the ants, and fragments of this tissue occur in the pellets of the Crematogasters, but these facts in themselves do not afford a solution of the difficulty. Nor does the absence of coccids in certain of the myrmecodomatia which contain callus-heteroplasias prove that the chief function of these traumatic tissues is not the provi- sion of food-reservoirs for Coccide. 608 Bulletin American Museum of Natural History (Vol. XLV ‘. Relations Between Ants and Coccids The well known aphid- and coccid-tending habits of the Cremato- gastrini suggest that the coccids are introduced into the myrmeco- phytes by the Crematogasters and carefully tended by them as “milch cows.” That the Crematogasiers actually devote considerable attention to the coccids is indicated by the following observation of Kohl (1909): Nicht selten fand ich Cocciden-Kolonien auch ausserhalb der Hohlungen, so auf Plectronia Laurentii, wo sie sich in einem Gehiuse befinden, das von den Ameisen aus fein zerkleinerten ‘pflanzlichen Bestandteilen hergestellt wird, nachdem letztere mittels einer von ihnen ausgeschiedenen klebrigen Flissigkeit verkittet wurden. Diese Schildliiusekammern standen vielfach mit den innern Nistriumen durch eine oder mehrere Ofinungen i in Verbindung. It seems probable, however, that the large colonies of coccids in Barteria fistulosa may be purely adventitious. The Pseudomyrmine are not known to tend Aphidide, Coccide, Membracide, Fulgoride, or Psyllide, and, as I have indicated on preceding pages, the Pachysima imagines actually carve up the coceids and feed them to their larve. - Nutritive Value of Medullary Tissue The ants scarify the walls of the myrmecodomatia of ‘the Cuvierez, Barterix, Plectronizx, and Sarcocephalus, excavating the remains of the succulent, inner portions of the pith and removing more or less of the thick-walled medullary tissue, Although fragments of these tissues occur abundantly in the infrabuccal pellets of the imagines, there is, unfortu- nately, no eonclusive evidence to indicate whether the ants actually feed upon them, or remove them merely for the purpose of cleaning and en-- larging the domatia; in other words, whether the fragments in the infra- ‘buccal sacs are food residues or detritus. Nutritive Value of Fungi All of the ant-inhabited plants (species of Acacia, Triplaris, Cecro- pia, Nauclea, Enterolobium, Myrmecodia, and Hydnophytum, as well as Cuviera, Vitex, Plectronia, Barteria, and Sarcocephalus) of which I have succeeded in obtaining suitable material have a more or less luxuriant growth of fungi upon the inner walls of their myrmecodomatia. These fungi are cropped by the ants and fragments of them are taken into the infrabuccal cavities of the imagines. However, as I have shown elsewhere (Bailey, 1920), there is no reliable evidence to indicate that these fungi are cultivated and fed upon by the ants, and are not purely adventitious and merely cropped in order to prevent them from obstructing the domatia and interfering with the broods. __ Wheeler, Ants of the Belgian Congo 609 Ants and Nematodes _ I have shown that the infrabuccal pellets of Pachysima and the se etistdations of detritus of Crematogaster impressiceps variety frontalis ____ contain numerous nematodes (PI. XX XVIII, fig. 3). Janet (1893) has ___ pointed out that certain nematodes pass their larval stages in the pharyn- ___ geal glands of Formica rufa Linnzus, Lasius flavus (Fabricius), ete., and ____ that sexed individuals occur in the detritus of the ant colonies. te the ase of Pachysima xthiops, the nematodes evidently work their way into the infrabuccal sac and subsequently are transferred to the larval tropho- 4 , thylaces or are cast out in voided pellets. Amber-colored Substance _. Thave called attention to the peculiar amber-colored substance in the normal and abnormal tissues of the Cuviere, Barteriz, and certain of the Plectroniz, which serves not only as an excellent ‘culture- medium’’ for fungi, but is fed upon by coccids and ants. When seen en masse, it is dark-colored and opaque, but in freshly-cut, thin, micro- scopic preparations is a translucent or hyaline, bright-golden-yellow, amorphous substance. It gradually darkens, if sections are left in alcohol or water, turning a reddish brown. The substance shrinks in drying and swells very considerably when remoistened with alcohol or water, but is insoluble in these liquids and also in ether, chloroform, benzol, carbon bisulphide, acetic, hydrochloric, and sulphuric acids, and cuprammonia. It dissolves readily, however, in dilute solutions of caustic soda, caustic potash, and nitric acid. It does not give a red color-reaction in Sudan III, aleanna, corallin, or hot or cold phloroglucin and hydrochloric acid, but takes on a reddish tinge in Hanstein’s aniline and in hydrochloric acid. It turns dark in iron salts and stains readily in aniline-blue. In iodine and cold Millon’s reagent, it takes on a dirty greenish color, but, when heated in the latter reagent, turns a dark brick-red. It retains its original color in hot concentrated nitric acid, but turns a darker yellow or orange brown on moistening with strong ammonia.’ I suspect that the substance is a complex mixture, containing pro- teins and carbohydrates but, in the absence of abundant material and an extensive chemical investigation, I have not been able to secure any reliable clue in regard to its composition. Of course, it should be kept in mind that the material which I have studied consists entirely of dried a and of myrmecodomatia 'When the substance has turned dark reddish brown, by standing in water or alcohol or by drying in air, the original bright-golden-yellow color is restored by treatment with nitric acid 610 Bulletin American Museum of Natural History [Vol. XLV preserved in 70%-90% alcohol. Therefore, the substance in question may have been considerably modified (by phenomena of coagulation, oxidation, ete.) and may have existed in a more fluid phase in living plants. If it has not undergone solidification, the coccids must secrete substances which act upon it in order to render it sufficiently fluid to pass through their sucking mouth-parts. Similar amber-colored substances occur in a number of other myr- mecophytes, particularly in rubiaceous plants. The cells of the peri- pheral layer of medullary tissue in the myrmecodomatia of certain Cecropiz are filled with a translucent amorphous material' whose micro- chemical reactions parallel those that have been outlined above. The exact role which these substances play in the nutrition of ants, and of the coccids and fungi which are frequently associated with them, deserves to be carefully investigated. THEORIES OF MYRMECOPHILY AND MyYRMECOPHYTISM Hypothesis of Richard Spruce During the last sixty years there has been considerable speculation concerning the significance of myrmecophytism. In his explorations of the Amazonian and Andean regions of South America (1849-1864) Richard Spruce encountered many plants having peculiar structural modifications, foliar sacs, hollow fistulose stems, etc., which were occu- pied by ants. Upon returning to England, Spruce prepared a paper? in which he endeavored to account for the origin of these remarkable myrmecophytes. This paper was read before the Linnwan Society on April 15, 1869, but unfortunately did not appear in print until 1908, owing to the fact that the Council of the Society would not authorize its publication unless Spruce made certain fundamental changes in the text. This Spruce refused to do. : The kernel of Spruce’s hypothesis is contained in the following paragraph: I have reason to believe that all of these apparently abnormal structures have been originated by ants, and are still sustained by them; so that if their agency were withdrawn, the sacs would immediately tend to disappear from the leaves, the dilated branches to become cylindrical, and the lengthened branches to contract; and al- though the inheritance of structures no longer needed might in many cases be main- tained for thousands of years without sensible declension, I suppose that in some it would rapidly subside and the leaf or branch revert to its original form. 1Fiebrig’s (1909) conclusion that the reddish brown color is produced by oxidations induced by formic acid is not likely to be substantiated. : 2“ Ant-agency in plant-structure; or the modifications in the structure of plants which have been caused by ants by whose long-continued agency they have acquired sufficient permanence to be em- ployed as botanical ters.”’ : ‘ __ Wheeler, Ants of the Belgian Congo 611 pa Efile Scena that the ants induced the formation of the abnormali- ties i in order to provide themselves with domatia during periods of in- ae und: and drew largely upon his remarkable fund of information a _ concerning the flora of tropical America in support of this idea. He pointed out that the majority of the myrmecophytes occur in low _ Swampy regions or regions that are periodically inundated, and em- _ phasized the fact that species which are entirely submerged during periods of high water are normal, whereas species which are only partly covered tend to be myrmecophytic. Seeing, then, how the sacs on the leaves have originated, and what purpose they serve, it is plain that a species of Tococa, like T. planifolia, inhabiting the very a _ river's brink, and liable to be completely submerged for several months of ___ every year, could never serve as a permanent residence for ants, nor consequently ___ have any character impressed on it by their merely temporary sojourn; even if their ____ instinct did not teach them to avoid it altogether, as they actually seem to do; where- as the species of Tococa growing far enough inland to maintain their heads above _ water even at the height of flood are thereby fitted to be permanently inhabited, and __—s are consequently never destitute of saccate leaves, nor at any season of the year clear ‘a é “Siete. Nearly all tree-dwelling ants, although in the dry season they may descend to the ground and make their summer-houses there, retain the sacs and tubes above- mentioned as permanent habitations; and some kinds of ants appear never to reside _ elsewhere, at any time of year. There are some ants which apparently must always ____ live aloft; and the Tococa dwellers continue to inhabit Tococas where there is never any risk of flood, as in the case of T'. pterocalyxr, which grows on wooded ridges of the Andes. Their case is parallel to that of the lake-dwellers of the mouth of the Orinoco ‘and the inundated savannas of Guayaquil, whose descendants must needs elevate ___ their houses on stages six feet or more in height, although nowadays erected on rising ground far beyond the reach of floods or ocean-tides (Spruce). _ Spruce did not consider that there was any true symbiosis between the ants and the plants as is indicated by the following statement in a letter to Darwin: The ants cannot be said to be useful to the plants, any more than fleas and lice are to animals; and the plants have to accommodate to their parasites as best they may. Belt’s Theory of Symbiosis In 1874 Belt published the results of his observations upon leaf- cutting ants and certain myrmecophytic Acaciz, and formulated an ingenious theory of myrmecophily. He concluded that the ants which inhabit the “bull’s-horn”’ Acaciz. form a most efficient standing army for the plant, which prevents not only the mam- malia from browsing on the leaves, but delivers it from the attacks of a much more dangerous enemy—the leaf-cutting ants. For these services the ants are not only securely housed by the plant, but are provided with a bountiful supply of food; and 612 Bulletin American Museum of Natural History [Vol. XLV to secure their attendance at the right time and place this food is so arranged and distributed as to effect that object with wonderful perfection. The leaves are bi-pin- nate. At the base of each pair of leaflets, on the mid-rib, is a crater-formed gland, which when the leaves are young, secretes a honey-like liquid. Of this the ants are very fond; and they are constantly running about from one gland to another to sip up the honey as it is secreted. But this is not all; there is a still more wonderful provision of more solid food. At the end of each of the small divisions of the com- pound leaflet there is, when the leaf first unfolds, a little yellow fruit-like body united by a point at its base to the end of the pinnule. Examined through the microscope, this little appendage looks like a golden pear. When the leaf first unfolds, the little pears are not quite ripe and the ants are continually going from one to another, examining them. When an ant finds one sufficiently advanced, it bites the small point of attachment; then, bending down the fruit-like body, it breaks it off and bears it away in triumph to the nest. All the fruit-like bodies do not ripen at once, but suc- cessively, so that the ants are kept about the young leaf for some time after it un- folds. Thus the young leaf is always guarded by the ants; and no caterpillar or larger animal could attempt to injure them without being attacked by the little warriors. These facts lead Belt to the conclusion that the function of the honey-secreting glands of plants is to attract insects which protect the flower-buds and leaves from the attacks of phytophagous insects and herbivorous mammals; and, by analogy, that the sugary secretions of various plant lice, scale insects, and leaf-hoppers have a similar fune- tion in attracting ants for the protection of these insects. Delpino’s Hypothesis A similar explanation of the function of extrafloral nectaries was put forward by Delpino in a paper read in 1873 (published in 1874-75). What then is the function of the extranuptial nectaries, which are found on the caulinary leaves, on the bracts, and on the calyx? Though I reserve for another paper the publication of my studies of such and other extra-dichogamie relations between plants and insects, I do not hesitate to announce now that the chief function of these nectaries is to place the ants, wasps, and Polistes in the position of sentries and guards, to prevent the tender parts of the plant from being destroyed by larvae.’ He elaborated this hypothesis in subsequent papers and reached the following conclusion (1889) in regard to myrmecophilous plants: 1 Qual @ dunque la funzione dei nettarii estranuziali, sia che si trovino sulle foglie cauline, sulle brattee o sul calice? Quantumque noi ci riserbiamo in altro lavoro di publicare i nostri studii sovra siffatte ed altre relazioni estradicogamiche tra le piante e gl'insetti, non esitiamo fin d’ora ad enunciare che riffatti nettarii hanno per funzione principale di costituire nelle formiche, nelle vespe, nei Polistes alieensngte vigili sentinelle e guardiani per impedire che le parti tenere delle piante siano divorate dai rucni. Kerner von Marilaun (1876) has also advanced the view that the function of extrafloral nectaries is to prevent ants from plundering the nectar from flowers and thus interfering with the normal proces- ses of insect-fertilization. ; Wheeler, Ants of the Belgian Congo 613 _ ae _ In the case of myrmecophilous species, the plant works for the ant in two ways, ne "either in supplying a sugared food, or in furnishing suitable lodgings, and the ant 3 ‘works for the plant in defending it against its enemies.’ Beccari’s Hypothesis * 4 E tnccari (1884-86), like Richard Spruce, endeavored to account for as origin of the peculiar structural modifications of various myrmeco- ___ phytes upon the basis of the Inheritance of Acquired Characters. The ___ following extracts from his beautifully illustrated work upon the myrme- Rok sophytes of Malasia are significant in this connection: ss Among insects, e. g. the ants, endowed with burrowing habits and also attracted wie TEE itiies qubotances and fleshy tissues, may have sought, by piercing the bark and _ the ligneous parts, to enter certain branches or certain twigs. Having taken away the ‘removable tissues, they may have succeeded in obtaining a commodious cavity in which to nest. During this performance the ants may have stimulated, although es unconsciously, the walls of the inhabited cavity, which in consequence of the stimulus, _ may have grown and become ulteriorly modified. So that finally, in the course of time, they may have produced twigs that were normally provided with sheltering eavities. The argument, if valid for the twigs, may also be applied to the spines, to the pouches on leaves or roots, or to any other organ. If the ants did not always use a definite spot in order to gain access into the cavity, or if the entrance was made on a place where no sensitive protoplasma was present, the aperture would not be- come hereditary (Acacia cornigera, Endospermum); if, on the contrary, the ants con- tinually sélected some particular spot | for perforation, and especially if the latter was of Acacia cornigera seem to me to have perhaps had this origin. . . . I am therefore of opinion that, because of the long time during which the stimulus of the ants acted upon the bulbose hypocotyle of Myrmecodia and related genera, a period of hereditary production (perhaps more pronounced in certain species than in others) has begun even of the galleries of the tuber, which are the essential part of the organ with which the plant acts as a host. In this way it may happen that this organ at once assumes the growth of a tuber, which, under certain circumstances, may remain in life and also grow independent of ants.‘ ™ + dal un alent Sastharinn 0 tormendole coemcaeh a piven ly Ae ae ig Belem la piante ited - per cs. fra gli insetti, dotate di abitudini perforatric, attratte forse da sostanse o da tessuto aver cercato, perforando la scorza ¢ la C. | introdurel <7 Eg tly rem di certi rami o ou fest, oes vi ‘0 il tessuto facil Sarees 2, provarees wae eavith comoda ella quale fa uale fare il nido. Sees ceregeh oom emnng ——— ayvranno stimolato, anche sense volerlo, le pareti della cavita abitata, la quale in causa d li stimoli Seoretl accrescersi © modificarsi a seconda del pp sapere. gor Bare oo andar del tempo a dare per eredita a dei fusti, che naturalmente prod tatrici. Il regionamente ti piesiol chia tes heen a alle gon oda ue altro organo. Se le f hon si saranno sempre servi poh rh ar ceeaen eal ° pol suesters pe sard stata praticata in luogo dove non esiste protoplasma pr l'apertura non taria (Acacia qvaraee, Endospermum); se perd le formiche approfitteranno costante- mente di qualche speciale per la orazione e scpratete se questo sarA un luogo dove esistono accumulamente f game he we succherine 0 mucillaginose, si es urre delle areole, dove il offrirA meno resistenza che altrove rerromse), a che lunga d “~~ eapetatanagate pervio A ( peitorum).. I _ toad * dell’ Acacia cornigera, mi he potre bbero avere avuto questa act me be ere ag) in causa del lungo tempo "erent il quale lo atimolo delle formiche ry hati dohe Myrmecodia o generi affini, sia incominciato (forse in talune specie pi ena in altre) un di produsione ereditaria anche per le gallerie del tubero, che costituiscono la parte easenziale gor ospitatore, per cui pud accadere che una volta avviata la vegetazione dei tuberi, questi *, certe circostanze possano mantenersi in vita ed anche accrescersi 614 Bulletin American Museum of Natural History (Vol. XLV Other Hypotheses Fritz Miller (1880-81) and A. F. W. Schimper (1888) brought together considerable evidence, particularly in the case of Cecropia, to show that myrmecophytism is a true symbiosis, that the ants actually protect their hosts and that the food-bodies, extraflorai nectaries and many of the peculiar structural modifications of myrmecophytes are adaptations for the purpose of attracting ants. It should be noted in passing, however, that “‘Schimper’s theory”’ of myrmecophily is not essentially different from the earlier hypothesis of Belt. In 1900 Buscalioni and Huber published a short paper, “ Eine neue Theorie der Ameisenpflanzen,’”’ in which they noted the abundance of myrmecophytic plants in swampy or inundated regions, and suggested that myrmecophytism originated under environmental conditions of this character, a view which, as I have shown, was championed by Spruce. It is evident, accordingly, that there are in reality but two distinct theories of myrmecophytism; the Belt-Delpino hypothesis, or theory of symbiosis (myrmecophily), and the Lamarckian theory of Richard Spruce. The “theories” of Fritz Miller, Beccari, Schimper, Buscalioni and Huber, and others resemble in their fundamental features one or the other of these hypotheses. Critics of the Theory of Myrmecophily The Belt-Delpino hypothesis and the adherents of myrmecophily ‘have been severely criticised in recent years by a number of different investigators. Moller (1893), Ule (1900), Rettig (1904), H. von Thering (1907), Madame Nieuwenhuis von Uxkiill-Giildenbrandt (1907), Fiebrig (1909), Wheeler (1913), and others have assailed the principal bastions of this theory and have succeeded in demolishing many of the principal arguments advanced in its support. Thus, it has been demon- strated that the ants are not sufficiently effective guardians of the plants to account for the origin of the various structural modifications of myrmecophytes through the action of Natural Selection. This is true even in the classical cases of the myrmecophytie Cecropiz and Acaciz. Furthermore, many plants which are not inhabited or frequently visited by ants are provided with pseudo-domatia, prostomata, food-bodies, extrafloral nectaries, ete. Present Status of the Problem of Myrmecophytism Although these investigators have succeeded in overthrowing the ingenious theory of Belt-Delpino and in showing that myrmecophytism is, in all probability, not a phenomenon of true symbiosis, but rather one 1922] Wheeler, Ants of the Belgian Congo 615 of parasitism, they have failed to provide a satisfactory working hypo- thesis to account for the origin and functional significance of the various inherited structural peculiarities of myrmecophytes. For example, Madame Nieuwenhuis von Uxkiill-Giildenbrandt (1907) states, at the end of her comprehensive monograph upon extrafloral nectaries, that biologists have no more idea in regard to the true meaning, origin, and functional significance of these glandular structures than they did in the time of Linnzus. We are equally in the dark concerning the so-called Beltian and Miillerian food-bodies, saccate leaves, fistulose or swollen, hollow stems and branches, prostomata, etc. It is true that the work of Treub (1883, 1888), Rettig (1904), Miehe (1911), and others indicates that the pseudobulbs, galleries, and papille of the Myrmecodiz have a physiological origin and function, but their results are not entirely con- clusive. _In view of the taboo that has been placed upon the Inheritance of Acquired Characters (owing to the seeming impossibility of proving, or _ disproving, the validity of this phenomenon) the simple Lamarckian ex- planations of Richard Spruce and Beccari are not likely to find many ardent supporters. Furthermore, the frequent occurrence of the same peculiar structural modifications, in plants which are not inhabited or visited by ants, is as serious a stumbling block in the way of the Spruce- Beccari hypothesis as it is in that of Belt-Delpino. The present status of the problem of myrmecophytism may be summarized, therefore, as follows. Certain plants tend—for reasons which are at present obscure—to form extrafloral nectaries, food-bodies, prostomata, saccate leaves, fistulose branches, and other pseudo- domatia, etc. In many cases, but by no means in all, these structural modifications of plants are taken advantage of by ants in their search for food and domatia. The myrmecophytic relationship which results is purely a case of parasitism in which all of the advantage lies with the ants. My own observations upon myrmecophytism among African plants lead me to believe that the relationship is solely one of parasitism; but one in which the behavior of the parasitic insects is particularly signi- ficant. Although there appears to be no valid reason for supposing that the ants have been, or are now, concerned in the origin and development of the pseudo-domatia or peculiar modifications of the central cylinder that are associated with phyllotaxy, histological and other’ evidence indicates conclusively that the callus-heteroplasias, as well as the re- markable exit apertures of Vitex Staudiii, are traumatisms induced by the ants. 616 Bulletin American Museum of Natural History [Vol. XLV Origin of the Pseudo-gall-forming Habit It is to be emphasized that the pit-like excavations, containing nutritive callus-heteroplasias, occur in plants of different growth forms (belonging to three distinct orders, Parietales, Tubifloree, and Rubiales) and are produced, not only by the “obligatory’’ Viticicola and Pachysi- mz, but also by the ubiquitous, ‘facultative’? Crematogasters. In the case of the highly modified and specialized V2ticicola tessmanni, the gali- like structures of Vitex Staudtii are very complex histologically and their production appears to be under more delicate control, as is evidenced by their characteristic form and structure and very symmetrical distribu- tion. In the flacourtiaceous and rubiaceous myrmecophytes, on the other hand, not only do the individual excavations vary greatly in size and shape, but their distribution is more or less erratic and there are numerous evidences of ‘trial and error” in their production. Whereas in Vitex Staudtii the ants almost invariably cut their excavations to just the right depth (the level of the cambium and inner cortex) to produce ingrowths of delicate, undifferentiated, nutritious parenchyma and gnaw away the inner margin of this layer at a rate which yields the most favor- able results, the Crematogasters, Engrammata, and “obligatory”? Pachy- sime frequently cut their way to the outer cortex, epidermis, or exterior and induce the formation of heteroplasias which, owing to the differen- tiation of more or less sclerenchyma, wound-wood, wound-cork, etc., are less nutritious and more difficult to control during the subsequent proc- ess of feeding. The question naturally suggests itself, how did these pseudo-gall- forming habits originate among ants? There are certain general tenden- cies in the growth of plants and in the activities of ants which appear to throw some light upon this problem. In woody plants, the cortex and cambium (and its derivative tissues) are under a certain equilibrium as regards mechanical strains and stresses, osmotic forces, distribution of moisture and food-substances, etc. When this equilibrium is disturbed by mechanical injuries, the polarity, rate of division, physiological activity and differentiation of the cells of the cambium and cortical parenchyma are more or less profoundly modified, depending upon the type and severity of the injury, the kind of plant, and its stage of development, etc. Abrasions and perforations, which extend through the cortex, phloem, and cambium to the xylem or medulla, tend to alter the polarity of the cambial cells in the immediate vicinity of the injury and to cause them to divide more actively. This change in polarity and acceleration in growth, which commonly extends to the Wheelér, Ants of the Belgian Congo . 617 —— parenchyma of the overlying tissues, produces lateral out-growths of delicate, thin-walled, unlignified, more or less isodiametric cells from the sides of the wound. These callus-formations gradually bridge over the gap in the side of the stem and !ead ultimately to a regeneration of the missing portions of the cambium, cortex, xylem, and phloem. However, the differentiation of the cells of the callus does not result at once in the production of norma! tissues, but of so-called wound-wood, wound-cork, etc. These wound-tissues vary greatly in form, structure, and arrange- ment, depending upon a number of different factors. Furthermore, there tends to be a considerable concentration of food-substances (protein, fats, sugars, etc.) in callus-heteroplasias and frequently of gums, muci- lages, resins, tannins, etc. in the tissues which are differentiated from them. _ The entrance and exit apertures, made by ants in the sides of myr- mecodomatia, usually become occluded by wound-tissues unless they are kept open by the ants. For example, F. Miiller (1880-81) found that in Cecropia the perforations made by Azteca queens in entering the primordial, internodal chambers frequently heal over and have to be reopened upon the exit of the young colony. Similar occluded apertures are of common occurrence in abandoned myrmecodomatia and domatia whose inhabitants have died or been killed by other insects. In clearing the entrance and exit holes, the ants are forced to gnaw upon the callus- heteroplasias. Therefore, since these tissues are usually well supplied with nutrient substances, as is indicated not only by microchemical analyses, but also by the fact that they are frequently fed upon by coccids, it might be expected a priori that the ants would discover the nutritive value of the callus and in many cases endeavor to increase its volume. The simplest and most direct method of producing additional callus-heteroplasias in the African myrmecophytes, and in other plants having similar myrmecodomatia, is for the ants to cut through to the cambium from the inside of their domatia. The moist, dark environ- ment in the interior of the domatia favors a luxuriant growth of callus and tends to retard the differentiation of its cells, whereas the general environment upon the exterior of the myrmecodomatis appears in most cases to have an opposite effect. This is indicated by the structure, form, and development of callus-outgrowths from perforations in the walls of fistulose stems and branches. The heteroplasias tend to grow more rapidly in the direction of the moist central chambers than they do laterally or towards the exterior, so that they frequently project more or less into the domatia; and, as shown in Pl. XL, fig. 1, the differentia-— 618 Bulletin American Museum of Natural History [Vol. XLV tion of the cells proceeds more rapidly in the external than in the inner portions of the heteroplasias.' In the myrmecophytic Cuviere, Plectroniz, and Barteriz, the callus heteroplasias which grow out from the margins of the entrance apertures resemble those that develop in the pit-like excavations. When the cells of the latter are filled with an amber-colored substance, the elements of the former contain this material which is fed upon by coccids and also apparently by the ants. In view of these facts, it is not surprising that the pseudo-gall- forming habit should have originated among several genera of African twig-inhabiting ants, the facultative Crematogasters as well as the obligatory Pachysime and Viticicola. The remarkable fact is that the pseudo-cecidia should occur in a number of African myrmecopbytes, belonging to three distinct orders, and not have been recorded in similar myrmecophytes from other tropical or subtropical regions. Of course, this may be due largely to structural and other differences in the myrme- codomatia, or to differences in the general growth-conditions of the vegetation, but I am inclined to believe that the phenomenon’ occurs in extra-African myrmecophytes and has been overlooked. For example, I find that the myrmecodomatia of Nauclea formicaria Elmer and of an unidentified Philippine myrmecophyte contain bisymmetrically arranged lateral excavations and callus-heteroplasias which resemble those of the African Cuviere. These myrmecodomatia were inhabited by species of Decacrema and Camponotus respectively. Furthermore, as has been pointed out earlier in this paper, the ‘‘stomatomes”’ of Cecropia, which contain colonies of Aztece, resemble the heteroplasias of Plectronia species A, and, although they are not situated in pit-like ex- cavations, may prove to be homologous structures. Sclerenchymatous Rings and Caps of Vitex Staudtii Just as certain abnormal environmental conditions, caused by mechanical injuries, lead to the formation of delicate, undifferentiated callus, others produce the transformation of thin-walled parenchyma intosclerenchyma. Thus, one commonly finds more or less sclerenchyma in those portions of the bark which overlie the excavations in the walls of the myrmecodomatia of the Cuviere and Plectroniz. Similarly, the outer surfaces of the exit apertures are often more or less completely jacketed by sclerenchyma. When the inwardly projecting callus has 1In Cecropia these ingrowths of callus, which are unusually large and conspicu may continue to grow oad oven prolilerate after the entrance apertures are por hace closed by wounded Wheeler, Ants of the Belgian Congo 619 di j : been completely gnawed away by the ants, as frequently happens, this > dense layer of tissue must serve as a more or less effective barrier to the growth of additional occluding tissue from the sides of the apertures. ¢ = a have emphasized the fact that in Vitex Staudtii the lateral ex- q ee and callus-heteroplasias are remarkabiy similar in form and structure and very symmetrically distributed. There is an equal uni- _ formity in the production of sclerenchyma, as is evidenced by the peculiar _ selerenchymatous caps and rings which are such characteristic features g in the anatomy of this myrmecophyte. In the absence of necessary experimental investigations, it is not possible to determine, however, _ whether Vitex Staudtii possesses a more pronounced tendency towards 3 - the formation of sclerenchyma than other myrmecophytes or whether = the production of these peculiar structures is due solely to a more delicate control of ordinary traumatic phenomena by the highly 4 - specialized Viticicola tessmanni. are | conclusion, it is to be emphasized that these tropical “bio- F eeessints,”” in which representatives of the higher plants, fungi, ants, coecids, and nematodes are intimately associated, deserve to be carefully and critically studied in the field. They should form the basis for some | emery interesting physiological and ecological investigations. » BrsplioGRAPHY Aviary, G. 1886. ‘Myrmecologiska studier. II. Svenska myror och deras lef- nadférhAllanden.’ Bihang Svenska Vet. Ak. Handl., X1, No. 18, pp. 3 1-329, Pls. vu. _ Bamey, I. W. 1920. ‘Some relations between ants and fungi.’ Ecology, I, pp. 174-189, Pls. v—vu. gemma O. 1884-86. ‘Piante ospitatrici ossia piante formicarie della Malesia e : della Papuasia.’ Malesia, II, 340 pp., 65 Pls. - Bm, Tu. 1874. ‘The naturalist in Nicaragua.’ London. xvi+403 pp. | Bras 1841. ‘Bijdrage tot de kennis der monddeelen van eenige vliesvleugelige gekorvenen.’ Van der Hoeven, Tijdschr. Nat. Geschied., VIII, pp. ; Buscarion1, L. anp Huser, J. 1900. ‘Eine neue Theorie der Ameisenpflanzen.’ Beih. z. Bot. Centralbl., [X, 2, pp. 85-88. _ Cosens, E. 1912. ‘A contribution to the morphology and biology of insect galls.’ } Trans. Canadian Inst., LX, 3, Pp. 297-387, Pls. 1—xu1. : Dewrio, F. 1874-75. ‘Ulteriori osservazioni e considerazioni sulla dicogamia nel q regno vegetale.’ Atti Soc. Italiana Sci. Nat. Milano, XVI, 3-4, 1874, pp. 151-349; XVII, 3-4, 1875, pp. 266-407. 1886-89. ‘Funzione mirmecofila nel regno vegetale; prodromo d’una monografia delle piante formicarie.. Mem. Ac. Sci. Bologna, (4) VII, 1886, pp. 215-323; VIII, 1887, pp. 601-659; X, 1889, pp. 115-147. 620 Bulletin American Museum of Natural History _ [Vol. XLV De Witpeman, &. 1905-07. ‘Mission Emile Laurent (1903-1904). Bnumération des plantes récoltées par Emile Laurent, pendant sa derniére Mission au Congo.’ Brussels; fasc. 1, 1905, pp. 1-112, Pls. r-xxxvi; fase. 2, 1905, pp. 113-192, Pls. xxxrx—xtv1; fase. 3, 1906, pp. 193-354, Pls. xivi-cv1; fase. 4, 1907, pp. ix-exx, 355-450, Pls. evii-cxiu; fasc. 5, 1907, pp. i-viii, exxi-exxv, 451-617. Donistuorre, H.S. J. K. 1915. ‘British ants, their life history and classification.’ Plymouth. xvi+379 pp., 18 Pls. E..iott, Jesse 8, B. 1915. ‘Fungi in the nests of ants.’ Trans. Brit. Mycol. Soc., V, 1, (1914), pp. 138-142, Pl. u. Emery, C. 1899. ‘Végétarianisme chez les fourmis.’ Arch. Sci. Phys. Nat. Genéve, (4) VIII, pp. 488-490. Farquuarson,}C. O. 1914. ‘The growth of fungi on the shelters built over Coccide by Cremas‘ogaster-ants.’ Trans. Ent. Soc. London, Proe., pp. xli-l. FERDINANDSEN, C. anv Winer, 0. 1908. ‘Fungi from the Danish West Indies col- lected 1905-1906 by C. Raunkier. I. Phycomycetew, Ustilagines, Uredinesw, Discomycetee, Pyrenomyceter et Fungi imperfecti.’ Botanisk Tidsskr., XIX, pp. 1-25, Pls. 1-n. Fresric, K. 1909. Cecropia poliale und ihr Verhiiltnis zu Azteca Alfari, zu Alta sexdens und anderen Insekten, mit einer Notiz tiber Ameisendornen bei Acacia cavenia.’ Biol. Centralbl., XXIX, pp. 1-16, 33-55, 65-77, 5 Pls. Forses, H. O. 1880. ‘Notes from Java.’ Nature, XXII, p. 148 (Myrmecodia and Hydnophytum). Hovarp, C. 1903. ‘Recherches sur la nutrition des tissus dans les galles de tiges.’ C. R. Ac. Sci. Paris, CXX XVI, pp. 1489-1491. Huser, J. 1905. ‘Die Koloniengriindung bei Atta sexdens.’ Biol. Centralbl., ys XXV, pp. 606-619, 624-635. InerinG, H. von. 1898. ‘Die Anlage neuer Colonien und Pilzgirten bei Atta sexdens.’ Zool. Anzeiger, XXI, pp. 238-245. 1907. ‘Die Cecropien und ihre Schutzameisen,’ Engler’s Bot. Jahrb., XXXIX, pp. 666-714, Pls. vi—x. Janet, C. 1893. ‘Sur les Nématodes des glandes pharyngiennes des fourmis (Pelodera, sp.).’ C.R. Ac. Sci. Paris, CX VII, pp. 700-701. 1895. ‘Etudes sur les fourmis. 8° note. Sur l’organe de nettoyage tibio- tarsien de Myrmica rubra L., race levinodis Nyl.’ Ann. Soc. Ent. France, LXIII (1894), pp. 691-704. 1899. ‘Sur les corpuscules de nettoyage des fourmis.’ Bull. Soc. Zool. France, XXIV, pp. 177-178. KERNER VON MariLaun, A. 1876. ‘Die Schutzmittel der Bliiten gegen unberufene Giste.’ Festschr. Zool. Bot. Ges. Wien, pp. 189-261. Kout, H. 1909. ‘Die Ameisenpflanzen des tropischen Afrika mit besonderer Beriicksichtigung ihrer biologischén Verhiltnisse.’ Natur u. Offen- barung, Miinster, LV, pp. 89-111, 148-175. Kuster, E. 1903. ‘Pathologische Pflanzenanatomie.’ Jena. 312 pp. 1911. ‘Die Gallen der Pflanzen.’ Leipzig. x+-437 pp. LacerHem, G. 1900. ‘Ueber Lasius fuliginosus (Latr.) und seine Pilzzucht.’ Ent. Tidskr., X XI, pp. 17-29. Wheeler, Ants of the Belgian Congo 621 _ Mervert, F. 1860. ‘Bidrag til de danske Myrers Naturhistorie.’ Danske Vidensk. Es Selsk. Skrift., naturv. math. Afd., (5) V, pp. 275-340. ; — H. 1911. ‘Javanische Studien. II. Untersuchungen iiber die javanische ; Myrmecodia.’ Abhandl. Math. Phys. Kl. Sichs. Ges. Wiss., XXXII, ei No. 4, pp. 312-361. Mourn, A. 1893. ‘Die Pilzgirten einiger siidamerikanischer Ameisen.’ Botanische Mitteilungen aus den Tropen, Jena, VI, vi+127 pp., 7 Pls. __ Murer, Frrrz. 1880-81. ‘Die Imbauba und ihre Beschiitzer.’ Kosmos, VIII, pp. 109-116. us Nrevwensuts von Uxxtttr-Gtpensranpt, Mrs. M. 1907. ‘Extraflorale Zuckerausscheidungen und Ameisenschutz.’ Ann. Jardin Bot. Buiten-* 2 zorg, X XI, pp. 195-328, Pls. xx—xxrx. Perkins, R. C. L. 1914. (Fungus growths on the shelters of Crematogaster). Trans. Ent. Soc. London, Proc., pp. xlii-l. Berna, E. 1904. ‘Ameisenpflanzen-Pflanzenameisen.’ Beih. z. Bot. Centralbl., XVII, pp. 89-122. am G. E. 1741-1755. ‘Herbarium ‘Aucinian, plurimas complectens arbores, frutices, herbas, plantas terrestres et aquaticas, que in Amboina et adjacentibus reperiuntur insulis.’ Amsterdam. 6 vols. Scnticom, A. F. W. i888. ‘Die Wechselbeziehungen zwischen Pflanzen und Ameisen im tronischen Amerika.’ Botanische Mitteilungen aus den q Tropen, Jena, I, pp. 1-95, Pls. 1-11. Scuumann, K. 1891. ‘Ueber afrikanische Ameisenpflanzen.’ Ber. Deutsch. Bot. Ges., [X, pp. 54-72. Spruce, Ricuarp. 1908. ‘Notes of a botanist on the Amazon and Andes.’ Edited by A. R. Wallace. London. I, lii+518 pp.; Il, xii+542 pp. Trevs, M. 1883. ‘Sur le Myrmecodia echinata.’ Ann. Jardin Bot. Buitenzorg, III, pp. 129-157. 1888. ‘Nouvelles recherches sur le Myrmecodia ‘de Java [M yrmecodia tuberosa Beccari (non Jack)].’ Ann. Jardin Bot. Buitenzorg, VII, pp. 191-212, Pls. xxm—xxv. Une, E. 1900. ‘Verschiedenes iiber den Einfluss der Thiere auf das Pflanzenleben.’ Ber. Deutsch. Bot. Ges., XVIII, pp. 122-130. 1906. ‘Ameisenpflanzen.’ Engler’s Bot. Jahrb., XX XVII, 3, pp. 335-352, Pls. v1, vu. Wueeter, W. M. 1913. ‘Observations on the Central American Acacia ants.’ Trans. 2nd Congr. Entom. Oxford (1912), II, pp. 109-139. Wuee er, W. M. anv Bamey, I. W. 1920. ‘The feeding habits of pseudomyrmine and other ants.’ Trans. Amer. Phil. Soc. Philadelphia, N. 8., XXII, 4, pp. 235-279, Pls. 1~v. . ™~ ‘Vv Pirate XXX Vitex Staudtii Guerke Fig. 1. Transverse section of a young stem, cut just above a node (at the level A-A shown in Text Figure 88, p. 449): CC, central cavity; Ph, peripheral layer of pith; PT, primary trachee; Xm, secondary xylem containing vessels; VXm, secondary xylem devoid of vessels; Cm, cambium; Bk, bark. X 19. Fig. 2. Transverse section of a young stem, cut just below the node (at the level B-B in Text Figure 88, p. 449): CC, central cavity; LC, lateral cavity; C7’, shriveled callus; Ph, remains of peripheral layer of medullary tissue; PT, primary tracheew; Xm, secondary xylem containing vessels; VXm, vesselless secondary xylem; Cm, cambium; Bk, bark. X 19. The arrows indicate the sides of the stele which pass out into the leaves at the next (higher) node. Bouieri~ A. M.N.H Vor. XLV, Prare XXX Ph2 Xm Be Pirate XXXI Fig. 1. Vitex littoralis Decne. Transverse section of a young stem, showing a central cavity in a species which is not inhabited by ants: CC, central cavity; Pht, thin-walled pith; Ph2, thick-walled pith; Xm, xylem; Cm, cambium; Bk, bark. X 12. Fig. 2. Viter Staudtii Guerke. Transverse section of a decorticated stem 18 mm. in diameter, showing six growth layers: CC, central cavity; Ph, remains of pith; Xm1, growth layers containing few vessels; Xm2, growth layers containing numerous large vessels; X mPa, radii of stem devoid of vessels, in which the forma- — tion of xylem parenchyma has been greatly accentuated by traumatic stimuli. 7. XXXI H. N. Beiietrx A. M . UG eee eae ee ee ‘ * Pirate XXXII Vitex Staudtii Guerke Fig. 1. Transverse section of a very young, tender stem, illustrating heterogene- _ ous medulla and early stage in the differentiation of the ‘‘fibrovascular cylinder” — A or stele: Phi, large-celled, thin-walled medullary tissue; Ph2, peripheral layer of medullary tissue composed of small cells; P7', primary trachew; Xm, xylem; Cm, ; ecambium; Bk, bark. The arrows indicate the sides of the stele which pass out ins the leaves at the next (higher) node. X 19. Fig. 2. Transverse section of normal stem cut a short distance below fi illustrated in Fig. 1: CC, central cavity; Ph2, peripheral layer of medullary ti PT, primary tracheew; Xm, xylem; Cm, cambium; Bk, bark. The arrows indicate. , the sides of the stele which pass out into the leaves at the next (higher) node. x a Fig. 3. Tangential longitudinal section of a stout stem, showing a cross : of lateral cavity: LC, lateral cavity; T’Pa, shriveled remains of thin-walled, u nified parenchyma; LXmPa, thick-walled, lignified xylem parenchyma; 4 prosenchymatous portion of xylem. X 33. Fig. 4. Radial longitudinal section of stout dry stem, illustrating lateral es and outer cap of sclerenchyma: LC, lateral cavity; 7’Pa, shriveled remains of walled, unlignified parenchyma; LXmPa, thick-walled, lignified xylem parench XmPr, prosenchymatous portion of xylem; Cm, cambium; Pm, phloewa periderm; CS, sclerenchymatous disk or cap. X 26. Fig. 5. Section of “nutritive” layer stained with Sudan III to differentiate x 90. H. N. Bouverim A. M. eee Pirate XXXII Vitex Staudtii Guerke Fig. 1. Radial longitudinal section of stout stem preserved in alcohol, : convex end of eter cavity and tissues which surround it: pert lateral cavity; ( lignified xylem dhe dieoo XmPr, prosenchymatous pa packed with s' Cm, cambium; Pm, phloem; Cz, cortex; Pd, periderm. Stained with chlorios zine. X 35. Fig. 2. More highly magnified view of the tissues shown in Fig. 1. 10h cavity; NL, “nutritive,” inner layer of callus; C7’, outer, larger-celled por' callus; 7’Pa, thin-walled, unlignified parenchyma; LXmPa, thick-walled, xylem parenchyma; XmPr, prosenchymatous xylem; Cm, cambium; Pm, p Stained with hwmatoxylin-safranin. X 60. : Fig. 3. Radial longitudinal section of xylem, showing septate, libriform f packed with starch. Section stained with chloriodide of zinc. X 170. Fig. 4. Section of “nutritive” layer, illustrating ground mass ol wiealich walled cells and dark-colored strands of conducting tissue. 200. : XXXII PLATE XLV, Bouretin A. M. N. H. Vitex Staudtii Guerke Fig. 1. Radial longitudinal section of stout stem with exit cavity (2) s by a ring of sclerenchyma (S): Xm, xylem; Cm, cambium. X 43. Fig. 2. Sector of transverse section cut just above the section illustrated i XXXII, Fig. 2, showing early stage in the formation of lateral cavity and nm layer: CC, central cavity; LC, lateral cavity; Ph, ee tissue; Xm, xylem; callus; Cm, cambium; Bk, bark. X 38. Fig. 3. Tangential longitudinal section of stout stem with exit cavity E ring of sclerenchyma; XmPa, parenchymatous portion of xylem; XmPr, pr chymatous portion of xylem. X 38. f Bouierm A. M.N. H. Vou. XLV, Pratre XXXIV Cr a: Cm ma \\\ Pe Pirate XXXV Fig. 1. Cuviera species? (collected at Kunga). Transverse section of normal, unswollen portion of internode: Ph, pith; LT, leaf trace bundles which pass out at the next (higher) node; Xm, cylinder of xylem; Pm, phloem; Cm, cambium; Cz, cortex. X 14. Fig. 2. Cuviera angolensis Hiern (from the Tshopo River). Transverse section of normal, unswollen portion of internode: Ph, pith; LT, leaf trace bundles which pass out at the next (higher) node; Xm, cylinder of xylem; Cm, cambium; Pm, phloem; Cz, cortex. X 14. Fig. 3. Cuviera angolensis Hiern. Transverse section of myrmecodomatium: Ph1, remains of thin-walled pith; Ph2, thick-walled pith; LT, leaf trace bundles which pass out at the next (higher) node; Xm, cylinder of xylem; Pm, phloem; Cm, cambium; NL, nutritive layer; Cz, cortex; CC, central cavity. X 11. Buiieti~n A. M. N. H. Vor. XLV, Prare XXXV ee oe ee Pirate XXXVI Cuviera angolensis Hiern Fig. 1. Transverse section of the swollen portion of a very young stem, showing pulpy pith which later collapses and dries up: Ph1, thin-walled pith; Ph2, thick- walled pith; LT, leaf trace bundles which pass out at the next (higher) node; Cm, cambium; Cz, cortex. X 14. a a Fig. 2. Transverse section of.the base of a swelling on a stout stem, illustrating ; one phase in the formation of a cavity without the intervention of the ants: CC, central cavity; Phi, remains of thin-walled pith; Ph2, thick-walled pith; Xm, xylem cylinder; L7', leaf trace bundles which pass out at the next (higher) node; Gp, gaps made by the exit of leaf trace bundles; Cm, cambium; Pm, phloem; Cz, cortex. m2: f Vou. XLV, Prats XXXVI H, M.N. Bouierin A. Fig. 1. Cuviera species? (collected at Kunga).— ing chewed inner portion. X 210. — Fig. 2. Cuviera angolensis Hiern. Sector of a transverse section of matium, showing nutritive layer: NL, nutritive layer; C7, callus; CC, ¢ Ph2, thick-walled pith; Xm, xylem; Cm, cambium, Pm, phloem; Cz, ¢ Fig. 3. Cuviera angolensis Hiern. Bection of thnbt edge of eontiak thick-walled cells of pith, thin-walled cells of pith containing mbe stance, and aérial hyphe of fungus. X 210. = Botzrerin A. M.N. H. Vou XLV, Prate XXXVII Botzrerm A. M.N.H VoL. XLV, Prate XXXVIII Cm ce Bk Pirate XXXIX Plectrcnia species A (from the Tshopo River) Fig. 1. Transverse section of young, normal internode: Ph, pith; Xm, xylem — cylinder; Cm, cambium; Bk, bark. X 9. Fig. 2. Transverse section of swollen portion of young stem, showing central cavity formed by the drying up of the thin-walled cells of the pith: CC, central — 19 cavity; Phi, remains of thin-walled portion of pith; Ph2, thick-walled portion of pith; Xm, xylem; Cm, cambium; Bk, bark. xX 9. Butcerm A. M. N. H. Vou. XLV, Prare XXXIX SA Maer y Piate XL : Fig. 1. Cecropia species. Sector of a transverse section of a young stem, show- ing normal structure (right) and callus formation or ‘‘stomatome”’ (left). Thelatter is differentiated into two distinct portions: a darker outer layer and a light-colored _ hyperplasia which projects into the cavity of the stem; these two layers are separated _ by a meristematic layer which is continuous with the cambium: Ph, pith; NXm, normal xylem; NBk, normal bark; AbXm, abnormal xylem; AbPm, abnormal phloem; C71, dark outer layer of callus formation; C72, light-colored hyperplasia projecting into the cavity of the stem; Cm, normal cambium; ACm, meristematic iY layer of callus formation. X 10. a Fig. 2. Plectronia species A (from the Tshopo River). Sector of a transverse section of a myrmecodomatium, showing hyperplasia projecting into the cavity of the twig: CC, central cavity; NL, nutritive layer; SL, starch containing parenchyma; Ph, pith; NXm, normal xylem; AbXm, abnormal xylem; NBk, normal bark; CT1, external callus; Cm, cambium. The layers SL and NL together represent the tissue —__ designated as CT in Fig.1. X 26. i Buiietin A. M. NH Voit. XLV, Pratre XL oy ii ‘Lp Puate XLI Plectronia Laurentii De Wildeman Fig. 1. Transverse section of quadrangular, hypertrophied branch: CC, central cavity; NL, nutritive layer or callus; Ph, pith; FPh, layer of medullary tissue which consists of flattened, thick-walled cells; T’Ph, remains of thin-walled pith tissue; VXm, vesselless xylem; Xm, xylem containing numerous vessels; Cm, cambium; Bk, bark. xX 6. ; Fig. 2. Transverse section of less swollen myrmecodomatium: CC, central cavity; LC, lateral cavity; CS, cap of sclerenchyma; NL, nutritive layer; Phi, layer of thin-walled medullary tissue; Ph2, layer of medullary tissue having thick- walled, flattened cells; Xm, xylem containing numerous vessels; VXm, vesselless xylem; Cm, cambium; Bk, bark. X 7. Boureti A. M. N. H Vou. XLV, Prats XLI ™ “ — - 7 PLate XLII Barteria fistulosa Masters Fig. 1. Transverse section of normal stem: Ph, pith; Xm, cylinder of xylem; LT, leaf trace bundles; Cm, eambium; Bk, bark. 11. 2a Fig. 2. Transverse section of sw sillan stem, showing central cavity formed by the ; collapse and drying up of the thin-walled cells of the pith: CC, central cavity; Phi, remains of thin-walled cells of pith; Ph2, thick-walled cells of pith; Xm, xylem; se leaf trace bundles; Cm, cambium; Bk, bark. xX 10. ; Buurerm A.M. N. H. Vou. XLV, Pratre XLII Puate XLII Fig. 1. Barteria Dewevrei De Wildeman and Durand. Transverse section of __ stout stem, showing heterogeneous pith, central cavity, and thin side of myrmeco- domatium: CC, central cavity; APh2, layer of thick-walled, flattened pith cells that __ are filled with amber-colored, hyaline substance; Ph2, peripheral layer of medullary tissue; LT, leaf trace bundles; VXm, vesselless xylem; Xm, xylem containing numerous vessels; Cm, cambium; Bk, bark. X 10. Fig. 2. Sarc cephalus species. Transverse section of dried myrmecodomatium, showing central cavity, heterogeneous medulla, and four nutritive layers: CC, central — cavity; Ph, peripheral layer of medullary tissue; Cd, sections of coecids; NZ, oS nutritive layer or callus; Xm, xylem; Cm, cambium; Bk, bark. X 8. aan Boureti~n A. M. N. H. Vou. XLV, Pirate XLII qesecetereesen, Fig. 1. postin 6.4 pale es es < (F. Smith), showing aetna be Wika ree Fig. 2. Barteria Dewevrei De Wildeman and Durand. thick-walled medullary tissue; C7, callus containing oto stance; VXm, vesselless xylem; Cm, cambium; Pm, phloem; Ca, pellet of Pechystna wthiops (F. Smith). x 100. - Pirate XLIV XLV, Vou. Buiietix A. M.N.H. LS Se Se eS Fig. 1. Fragments of nuzitive laee of heteropasia ts larval tessmanni (Stitz). XX 90. a Fig. 2. Coccidl larva (Sticlococous formicariue Newstead) from. Pachysima xthiops (F. Smith). X 58. Fig. 3. Fragen of ant ara fom tarp of il numerous spores of different kinds. X 330. _ ne Fig. 5. Portion of pellet of Pachysima zthiops (F. Smith) aérial hyphx, spores, and other detritus. X 330. x Fig. 6. Portion of pellet of Pachysima zxthiops (F. Smith, sh medullary tissue containing amber-colored substance. xX 96 Vou. XLV, Pratre XLV 3% VI.—NOTES ON A COLLECTION OF WEST AFRICAN MYRMECOPHILES Tiga By Wa. M. Mann __ Prof. Wheeler has kindly given me for study an interesting collec- tion of West African myrmecophiles, most of them collected by the Rev. G. Schwab from nests of several species of Dorylus subgenus Anomma, some larve of Microdon taken by Messrs. Lang and Chapin, and two _ Pausside, not found with their host ant, collected by Dr. J. Bequaert _ and Messrs. Lang and Chapin. Rey. G. Schwab had before sent quantities of material to Father E. _ Wasmann, who has recently written much on the guests of the doryline ants, increasing their number from fourteen species in 1900' to an ex- tensive favna, rich in highly specialized genera and species. Most of __ the species before me have been described by him. There is in the col- lection, however, an additional species of the interesting genus _ Dorylophila and a new variety of Ocyplanus kohli which I venture _____ Four specimens (one adult and three larv) of an aradid bug and the eurious Microdon pup hereafter described were taken with Pheidole ___ megacephala (Fabricius). The other species in the following list are guests a COLEOPTERA Pausside Pleuropterus lujw (Wasmann) Text Figure 101 dohrni Wasmann, 1907, Deutsch. Ent. Zeitschr., p. 152, Pl. 1, fig. 3 (2) (nec fig. 4; nec Ritsema). dohrni subspecies lujz WaAsMANN, 1907, Deutsch. Ent. Zeitschr., p. 152; ee 1910, Ann. Soc. Ent. Belgique, LIV, pp. 394 and 396. ____ Pleuropterus lujz Wasmann, 1918, Tijdschr. v. Ent., LXI, p. 81. Belgian Congo: Medje (Lang and Chapin). ‘The one specimen is without host ant. Originally described from - Kondué, Kasai, also without indication of the host. Paussus wthiops Westwood Text Figure 102 Paussus xthiops Westwoop, 1845, Arcana Ent., IT, p. 186, Pl. xcmt, fig. 6. Bian- ; cuarp, in Cuvier, Régne Animal, 3d Ed., Ins., Pl. uxt, fig. 8 (before 1845, but without descripticn). asmann, WS. linengiiste dem neotropischen und dem Athiopischen Fee Es Mearb. ADL Syst, XIV, pp. 215-280, Pla, Xitt-xiv. if wil, a VV P FOU AIS ) Fig. 101 foriFig. @@251 oviana} Fig, 101. Pleuropterus luja (Wasmann); a, insect from above; Rpardahapoesc only Fig. 102. Paussus xthiops Westwood: insect from above. yor notes! Belgian Congo: Between Beni and Kasindi (J. Bequaert); 90 “At Lisasa, a village in the Savannah of the Semliki Valley, about midway between Beni and Kasindi, a great many specimens of this Paussus were attracted by lights in the evening (August 12, 1914): When taken between the fingers these beetles would ‘explode’ in the same manner as bombardier-beetles (Brachinus, Pheropsophus, ‘ete.): They emit at the same time»a-volatile,substance with a strong odor of bromine which stains the skin browny’ (J. Bequaert). Staphylinide Sympolemon anommatis Wasmann . ‘4 Sympolemon anommatis WasMANN, 1900, Zool. Jahrb. Abt. Syst, XIV, 0. 258: 1917, Zeitsehr. Wiss. Zool., CX VII, p. 311, Pl. 1x, fig. 30. TS TT EY be | Cameroon: Akono-Linga (Schwab). i Host: Doryius (Anomma) nigricans sjestedli Emery. Three specimens. which agree closely with a cotype received 3 Father Wasmann. The species has been found with various tern in the Belgian Congo senetesies St. Gabriel) and ‘Cameroon (Gran Batanga; Yukaduma), Mimanomma spectrum Wasmann Mimanomma spectrum WaAsMANn, 1912, Zool. Anzeiger, XXXIX, p. Lo Mt Ls; 1917, Zeitschr. Wiss. Zool., CX VII, p. 302, Pl. rx, fig. 27, bericht Cameroon: -Akono-Linga (Schwab). weceinsseendel Host: Dorylus (Anomma) nigricans sjestedti Emety. oh TRY) 0 Wheeler, ‘Ants of the Belgian Congo 625 1) Prof. Wheeler sends me the following notes in regard to this extra- . dorylophile: ‘In conversation with Mr. Geo. Schwab I learned __ that, although he investigated as many as 1000 to 1200 marching armies of) Durylus and Ancmma during his sojourn of many years in the - Cameroon, he succeeded in finding Mimanomma only on two oceasions. ‘Phe first lot, comprising the types, was sent to Father Wasmann in two vials which led him to cite them erroneously as from two armies (Zool. Anzeiger, XX XIX, 1912, p. 473). The second lot, which Mr. Schwab ‘sent to me, was taken with the same host (Anomma nigricans subspecies _ &jeestedti) about 60 miles farther inland and 30 miles north of Akono- _ Linga, August 19, 1916. The beetles walk in the Anomma files but more aris than the ants. Mr. Schwab says he has never seen the ants & touching or paying the slightest attention to the Mimanomma. Be. The same i is true of the other staphylinids which are often very numerous ; ions or bring up the rear after the ants have passed. He pre ei the dorylophiles are most abundant in August and may be very scarce in the processions during the rainy season. He captured beetles as voluntarily and persistently returned to the ant- trail after they had been removed from it. “Wasmann, in dealing with the ecitophiles of the Neotropical and the dorylophiles of the Ethiopian Region, has elaborated hypotheses of mimicry, hypertely, etc., to account for the ant-like appearance of some of these insects. Mimanomma he regards as a case of hypertely—one in which the insect has become an example of greatly and uselessly exag- gerated mimicry of its host (‘iiber das Ziel hinausschiessende Mimicry’). As it is rather important that such speculations, which are easily excogi- tated in laboratories and museums, should not be left in undisputed possession of the field cf theoretical biology, I advance another hypo- thesis which seems to me worthy of consideration: It is well known that bivouacking dorylines, and especially the species of Anomma, form great masses, like swarming bees, with their long legs, antenne snd bodies d and enveloping the brood, booty, and guests. Long, slender insects like Mimanomma and even those of Wasmann’s ‘Trutztypus,’ which have the very opposite shape, being short and broadly rounded anteriorly, with rapidly tapering posterior end, would be beautifully adapted for foreing their way through and moving about in the forest of , Antenne and bodies of the bivouacking ants, much as both very , long, insinuating and small, rotund, pushing people seem to be ee. adapted for shouldering their way through a crowd than people of average stature. Hence, the peculiarities of form referred by Was- 626 Bulletin American Museum of Natural History [Vol. XLV mann to mimicry, hypertely, etc. may be really direct and useful adaptations to the very peculiar nest environment created by the densely agglomerated bodies of their hosts. I have seen such conditions in ecitophile-containing artificial nests of our North American Eezton (Acamatus) schmitti imery and opacithorar Emery, and have no doubt: that future observers will be able to make similar observations on Anomma and its guests. Of course, M. spectrum is really ‘phasmoid,’ rather than ‘ant-like’.’’ Dorylomimus brevicornis Wasmann Dorylomimus brevicornis WASMANN, 1917, Zeitschr. Wiss. Zool., CX VII, p. 298. Cameroon: Batanga (Schwab). Host: Dorylus (Anomma) nigricans burmeisteri variety rubellus (Savage). Originally taken from the columns of the same ant at St. Gabriel near Stanleyville.. A single specimen before me agrees closely with the description of the type. It is very distinct from a cotype of D. kohli Wasmann in having the head shorter and broader and the antenne shorter. Dorylophila rotundicollis Wasmann Dorylophila rotundicollis WasMann, 1904, Zool. Jarhb. Abt. Syst., Suppl. ate p. 633, Pl. xxx1, fig. 7. - Cameroon: Akono-Linga (Schwab). Host: Dorylus (Anomma) nigricans sjestedti Emery. Several specimens in the collection agree closely with Wasmann’s description and figure of this species, which was described from speci- mens taken with Dorylus wilverthi Emery in the Congo. Dorylophila schwabi, new species Length 2 mm. Dark reddish brown, antennz yellowish brown; very feebly shining; head, thorax, and elytra finely granulose-punctate and with a dense covering of short hairs; abdomen with fine, silky, semirecumbant hairs which are longest on the margins and apex. Head broader than long, wider behind than in front, sides in back of eyes feebly convex and rounding into the feebly convex posterior border. Eyes a little more than half as long as sides of head, the surface a little convex. Antennz stout, first joint as long as the second and third together, second and third joints elongate- cylindrical, the third shorter than the second, fourth joint slightly longer than broad, remaining joints transverse, becoming strongly so apically, terminal joint a little longer than the two preceding. Pronotum broader than long, with a strong semi- circular impression at the posterior portion and the posterior two-thirds of sides; 1922] Wheeler, Ants of the Belgian Congo 627 - middle of posterior b Siedee slightly produced and rounded; surface in front of semi- ‘ow circular impression convex, with a broad, shallow impression behind middle. Elytra at base a little broader than prothorax, broader behind than in front, sides and ____ posterior border nearly straight, sides elevated into blunt margins, surface flat _____ behind, elevated and feebly convex in front of middle. Abdomen narrow, about as ____ Jong as remainder of body, at base a little narrower r than the elytra, first five seg- - ments margined at sides. Cameroon: Efulen to Elat (Schwab). Host: Dorylus (Anomma) nigricans burmeisteri variety rubellus (Savage). This is the second species in the genus and differs from D. rotundi- collis Wasmann in its smaller size, more delicate punctatiion, in the broader and thicker antennal joints, and in not having the posterior corners of the elytra angulately projecting. Znictonia (Anommatonia) anommatophila Wasmann _ Binictonia anommatophila W asMann, 1904, Zool. Jahrb. Abt. Syst., Suppl. VII, p. 636. Zinictonia (Anommatonia) anommatophila Wasmann, 1915, Ent. Mitt. Deutsch. _ Ent. Mus. Berlin, IV, p. 31, Pl. 1, figs. 2, 2a—b. Cameroon: Akono-Linga; Mful Aja (Schwab). Host: Dorylus (Anomma) nigricans sjestedti Emery. Znictonia (Anommatochara) rubella Wasmann Ainictonia (Anommatochara) rubella WasMANN, 1915, Ent. Mitt. Deutsch. Ent. Mus. Berlin, IV, p. 33, Pl. 1, figs. 4, 5, and 5a. Cameroon: Akono-Linga (Schwab). Host: Dorylus (Anomma) nigricans sjestedti Emery. Ocyplanus kohli Wasmann! variety niger, new variety Differing from the typical form (from nest of Dorylus wilverthi Emery) in color, being black, with the appendages brown and the apical portions of femora dark brown to black. The difference is constant in a series of thirty specimens before me, which apparently belong to a distinct variety. Cameroon: Mful Aja (Schwab). Host: Dorylus (Anomma) nigricans sjestedti Emery. Demera kohli Wasmann Demera kohli Wasmann, 1904, Zool. Jahrb. Abt. Syst., Suppl. VII, p. 630. Cameroon: Metit (Schwab). Host: Dorylus (Anomma) kohli variety congolensis Santschi. Several specimens, one of which has been compared with the type, are in the collection. Mg kohli Wasmann, 1916, Ent. Mitt. Deutsch. Ent. Mus. Berlin, V, pp. 137 and 139, 1, 5. 628 Bulletin American Museum of Natural History (Vol. XLV Pygostenus bicolor Wasmann 1 to elbbint Pygostenus bicolor WASMANN, 1904, Zool. Jahrb. Abt. Syst., Suppl. VII, p. ea. Cameroon: Batanga (Schwab). { ‘oitet20q Host: Dorylus (Anomma) nigricans burmeistert variety’ rubellus (Savage). ‘ YT 28 ito yf One specimen. STI SPOR re : rocraata’ > Pygostenus luje Wasmann rol Pygostenus lujz WAsMANN, 1904, Zool. Jahrb. Abt. Syst., Suppl. Vaz D. 648. id Cameroon: Batanga (Schwab). af = eW estos Host: Dory!us (Anomma) nigricans burmeisteri variety. rubellus (Savage). “IT109 Four specimens. |. z romA) alcotolatt Pygostenus alutaceus Wasmann re Pygostenus alutaceus WAsMANN, 1904, Zool, Jahrb. Abt. Syst, sini VII, p. 649. Cameroon: Batanga (Schwab). ootsita’) Host: Dorylus (Anomma) wioripant aE velletan rubellus (Savage). T The single specimen in the eclleasiies runs in Masnaial s be and answers to the short description of this Species, which was first taken with D. wilverthi Emery in Congo. otemadD if OL sdaoH Phyllodinarda xenocephala Wasmann Phyllodinarda xenocephala Wisuany, 1917, Zeitschr. Bech Zool., Pica! p. 330, Pl. rx, figs. 35 and 36. firth Cameroon: Akono-Linga (Setiwab). foo mi (yaad Host: Dorylus (Anomma) nigricans sjestedti Emery‘) 1o 2oitoq Originally found with the’same ant in Cameroon (a pees Lolodorf). av socijeib Darra Ge: Syrphide Bye: Microdon species ~ - ‘Text Figure 103 Hi vided exionatl Larva. Length 6 to 7.5 mm. Dark brown, opaque (except stigmal plates), granulose-punctate.. Fann asad oval, convex above, concave beneath. Dorsum with a strong median longitudinal ridge extending from the posterior spiracle to anterior end and a series of n similar transverse ridges which are interrupted at middle; these ridges thickly coy with coarse, conical spines, some of which appear to be composed of elongate flattened hairs; surface between ridges reticulate, the reticule made up of rows of clusters 0 1922] .. Wheeler, Ants of the Belgian Congo 629 3 to 5 erystalline-like particles. Lateral margins with an interrupted, moderately coarse longitudinal ridge beneath which is a series of four fine parallel ridges and a membraneous margin. Posterior spiracle elongate, tubercular, dull grayish in color, stigmal plates shining, amber-colored, each divided into four stubby finger-like projections, two above and two below, above with two very large pores. Congo: Zambi (Lang and Chapin). Host: Pheidole megacephala (Fabricius). Several specimens. ssa These pup are remarkable on account of the pronounced ridges on the upper surface and the structure of their bristles. The latter vary, those at the sides of the ridges being elongate, whitish flat hairs arranged in groups of 2 to 6; the others thick, conical, brown structures, seeming- ingly composed of masses of hairs coalesced. Most of the conical spines are subequal in size but among them are a few much larger than the others. All have at the tips whitish particles which are somewhat glistening and may possibly be exudations. BPTSsSt Ea a \ SACs }a4 Fig. 103. Microdon species: larva living with Pheidole megacephala (Fabricius) at Zambi; from HOMOPTERA Coccidee The following scale insects were found in the domatia of various ant-plants collected by Lang, Chapin, and Bequaert in the Belgian Congo. They have been identified by Prof. R. Newstead, of the Liver- pool School of Tropical Medicine. The three forms first enumerated are apparently still undescribed. Pseudococcus citri (Risso) variety congotnsis Newstead Taken from myrmecodomatia of Barteria fistulosa inhabited by Pachysima zxthiops (F. Smith) at Medje (Lang and Chapin). Also from domatia of Cuviera angolensis inhabited by Crematogaster africana sub- species laurenti variety zeta (Forel) near Stanleyville (Lang and Chapin). . 630 * Bulletin American Museum of Natural History ([Vol. XLV Pseudococcus crassipes Newstead Taken from myrmecodomatia of Sarcocephalus species inhabited by Crematogaster africana subspecies winkleri variety fickendeyi (Forel) at Masongo, between Walikale and Lubutu (J. Bequaert). Lecanium (Saissetia) barteria Newstead Taken from hollow stems of Barteria Dewevrei inhabited by Crema- togaster africana variety schumanni (Mayr) at Leopoldville (J. Bequaert). Stictococcus formicarius Newstead Stictococcus formicarius NewsTEaD, 1910, Journ. Econ. Biol., V, p. 19, fig. Larve of this species were recognized in the pellets taken from the trophothylax of larve of Pachysima xthiops (F. Smith) living in Barteria fistulosa at Medje (Lang and Chapin). This scale insect was described from specimens found in the hollow stems of Barteria fistulosa and Cuviera angolensis. VIL—KEYS TO THE GENERA AND SUBGENERA OF ANTS By Wn. M. WHEELER Key TO THE SUBFAMILIES! 8, 9 nS ok: Saneeal orifice round, terminal, surrounded by a fringe of hairs; sting transformed into a sustentacular apparatus for the orifice of the poison vesicle, which has a peculiar structure called by Forel ‘‘ pulviniferous vesicle”’ (vessie 2 coussinet). Abdominal Ke pedicel consisting of a single segment; no constriction between the second and third segments. Male genitalia not retractile. Nymphs rarely naked, most frequently enclosed in a cocoon. FORMICIN#®. Cloacal orifice in the shape of aslit....................0..06. 2. 2. Sting rudimentary (except Aneuretus); abdominal pedicel con- sisting of a single segment; no constriction between the second and third segments of the abdomen; the poison glands are often vestigial and there are anal glands which secrete an aromatic product of characteristic odor (T’apinoma-odor). Nymphs without a cocoon........... DOLICHODERIN &. _ Sting developed, though sometimes very small, but capable never- theless of being exserted from the abdomen. The first two segments of the abdomen usually modified, either forming together a two-jointed pedicel, or the first alone (petiole) forming the pedicel, the second (postpetiole) being merely constricted posteriorly and articulating with a spheroidal surface of the third segment, which is usually transversely striated (stridulatory organ); rarely the second segment is not appreciably modified. . “se at 3. Pedicel of two segments, the pedis ana the Siatioeicls- lparely (in Melissotarsus, e. g.) the postpetiole is attached to the follow- ing segment over its whole extent. Frontal carine usually separated from each other (except in the Melissotarsini and certain Attini). In the male the copulatory organs are almost always exserted (being entirely retractile in certain genera of the Solenopsidini only); cerci nearly always present (except Anergates). Nymphs naked.................... 4. ‘Wheeler, Wm. M. 1920. ‘The subfamilies of Formicidw, and other taxonomic notes.’ Psyche, XXVIII, pp. 46-55. 631 632 Bulletin American Museum of Natural History (Vol. XLV Pedicel consisting of a single segment, more rarely of two, but in this case the frontal caring are very close to each other and do not cover the insertions of the antenne (Doryline) or the mandibles are linear and denticulate (Myrmecia).......... 5. 4. Clypeus not prolonged back between the frontal carine (in some species of Pseudomyrma there is an apparent prolongation which, however, is the equivalent of the frontal area and is often separated from the clypeus), its posterior margin rounded. Median spurs of middle and hind tibiw pectinate. Ocelli almost always developed in the worker. Antenna 12- jointed in worker, female, and male. Fore wings with two closed cubital cells, rarely with one. Larve hypocephalic and with a trophothylax; the thoracic and first abdominal segments furnished with peculiar exudatory papille (exuda- toria) which form a cluster around the mouth. PSEUDOMYRMIN#&, Clypeus almost always prolonged between the frontal carinew; if not, the spurs of the middle and hind tibiz are simple or absent, or the antennz are 11-jointed in worker and female, 12-jointed in the male and the fore wings have one closed cubital cell. Larve orthocephalic, without exudatoria around the mouth. MYRMICIN 2. 5. Frontal carine very close to each other, almost vertical, not at all covering the antennal insertions; abdominal pedicel of one or two segments. In the male the genitalia are completely re- tractile (except in Leptanilla) and the subgenital lamina is usually (if not always) furcate; cerci absent. Nymphs usually naked (eyes and ocelli absent in the 8 of all African genera). DORYLIN#&. Frontal carine separated or close together; in the latter case they are dilated anteriorly to form an oblique or horizontal lamina, covering in part. the insertion of the antenne; abdominal pedicel of a single segment (except Myrmecia). Copulatory organs of the male incompletely retractile; subgenital lamina never furcate (except in Paraponera); cerci nearly always present. Nymphs usually enveloped in a cocoon (eyes pres- ent in the 8 of most African genera). CERAPACHYIN-® and PONERIN #&. Wheeler, Ants of the Belgian Congo 633 DORYLINZ Leach Key to the Tribes ot besa and Soldier: pygidium tridentate, with a median impres- sion; maxillary and labial palpi 2-jointed; cheeks without a longitudinal carina; first abdominal segment not separated from the second by a constriction. Female: cloaca open, leaving the sting uncovered; hypopygium forked and extend- ing considerably beyond the pygidium; thorax unsegmented. “Male: pterostigma of fore wing very elongate and narrow; radial cell open and elongate, one closed cubital cell; genital INN MUEEMIMEND Ss 6. ccs ee cece ww ne Doryuint Forel. ~ Worker and Soldier: pygidium simple; maxillary palpi 2- or 3- jointed, labial palpi 2-jointed; cheeks longitudinally carinate ; a constriction usually separates the second and third abdominal segments, making the petiole 2-jointed. Female: cloaca covered by the pygidium; hypopygium not considerably ex- tended. Male: pterostigma of fore wing broad or narrow; radial cell elongate, one or two closed cubital cells; genital armature retractile.............. i .Ecrroninti Forel. Worker: pygidium siunle: eaesitlagy ‘and. labial palpi 1-jointed; cheeks not carinate; petiole 2-jointed. Female: cloaca open, leaving the sting uncovered; hypopygium lobed and extending beyond the pygidium; thorax with a suture behind the anterior pair of legs, which is effaced on the dorsum. Male: fore wings without pterostigma or nervures. , Genital armature extended, not retractile............... LEPTANILLINI Emery. 1. Dorylini Forel Dorylus Fabricius. (Ethiopian, North Africa, the Mediterranean coast of Asia Minor, Indomalayan, Papuan). 8,24 ‘a. Antennw 12-jointed in the soldier and in the large- and medium-sized worker. (Indomalayan)......... Subgenus Dichthadia Gerstecker. Antenne 9-jointed. (Ethiopian, North Africa, Indomalayan, Papuan). Subgenus Alaopone Emery. eee ee ee Cnn enn he ee b. b. Pygidium with a semicircular impression, the margins of which are sharp. RM ket ccneeediensccctetmecsnvnsess » ane The impressed area of the pygidium without distinct margins......... d. c. Antennw short and thick; all the joints of the funiculus, except the last, much wider than long. (Ethiopian)........... Dorylus, sensu stricto. 634 Bulletin American Museum of Natural History (Vol. XLV Antenne elongate; at least some of the joints of the funiculus longer than wide. (Ethiopian)............. Subgenus Anomma Shuckard. d. Subapical tooth of mandibles simple; antenne 11-jointed; worker major 13 mm. long. (Same distribution as the genus). Subgenus Typhlopone Westwood. Subapical tooth of mandibles double, or truncate; worker major 8 mm. long. (Eitistoniom) <> soi +. ss wcxuuee Subgenus Rhogmus Shuckard. The female of Typhlopone Westwood is unknown. a. Antenne 12-jointed................ Subgenus Dichthadia Gerstecker. Antenne 11-jointed... 2... cio cee ccccsnseveecnnss ss suntan b. b. Hypopygium having the shape of a cleft plate which is narrowed behind. Subgenus Dorylus, sensu stricto. c Anomma Shuckard. “ Rhogmus Shuckard. Hypopygium wide, forming two lobes which are divergent behind. Subgenus Alaopone Emery. of a. Mandibles wide at the base and prolonged into a point, with the inner margin deeply excised............ Subgenus Dichthadia Gerstecker. Mandibles shaped differently. ...... .\-rs 6. 0000s nee tase oe b. b. Petiole wider than long, its posterior face concavely excavated......... c. Petiole nearly square, or round. .. 20%... 5.0/0. < He< + 0s ois Sule ie d. c. Mandibles less than 4 times as long as wide. 4 Subgenus Dorylus, sensu stricto. Mandibles more than 4 times as long as wide. Subgenus Anomma Shuckard. d, Mandibles about 3 times longer than wide. Subgenus Typhlopone Westwood. Mandibles much shorter. 2.0... ios oe cae oe so os lw alee Se e€. e. Wings with a second recurrent nervure. . .Subgenus Rhogmus Shuckard. Wings without a second recurrent nervure. .Subgenus Alaopone Emery. 2. Ecitonini Forel The female of Cheliomyrmex and the worker and female of nicto- giton are unknown. 8, 4 1, Antenne 10-jointed. No vestiges of eyes. (Ethiopian, North Africa, Indomalayan, Papuan, northeast Australia). Znictus Shuckard. Antenne 12-jointed . oii cece ideee's pies b's cabin eu his ae 2. ain ~~ Wheeler, Ants of the Belgian Congo 635 Pedicel composed of one segment, the ggg not sharply | pctore with a median tooth. (Neotropical).! ity Cheliomyrmex Mayr. fi " f Saiaeke dateith a. median tooth. (Neotropical, except Gis Ano ; “te tilles and Chile; central and southern United States). 3 Eciton Latreille. n ae ss otns ca ah dae the ecsipitel angle. Eyes distinct. Soldier with hook-shaped mandibles (E£. rapar has ae NE ss beac ekew aes Subgenus Eciton, sensu stricto. , SWIG jolnt of the fusiodkis moro than half the length of the second. ___ Head without spines or at most with a simple, straight spine at the occipital angle. Mandibles of the soldier not hook-shaped. Subgenus Labidus Jurine. Any ? faibediee Creo es 5 ne .....nictus Shuckard. e Antenne 5 sss y's. ce vein en ke Eciton Latreille. a a Subgenus Acamatus Emery. E Claws with a distinct median tooth Ceo Te Al OMS Fae Cae eke + bok aes b. 7 Subgenus Eciton, sensu etriele. _ Epinotum and petiole above without projections . Subgenus Labidus Jurine. os ___ Pterostigma of fore wing broad; radial cell open.............. 3. Be a Mandibles very long, slender, and falcate, with a peculiar cluster of short, erect hairs at the base on the inner side; subgenital _--—-: plate with four apical teeth; hind femora distinctly flattened. ae Cheliomyrmex Mayr. _Mandibles shorter and of a different shape, or if of the same shape then without the peculiar cluster of hairs; subgenital plate with pawn ate teeth; hind femora not or only feebly flattened. Eciton Latreille. a. Legs short, the hind femur not reaching the hind margin of the second seg- ment of the gaster; head narrow; thorax hump-backed, much raised a eae be Subgenus Acamatus Emery. 4‘ have recently placed Cheliomyrmez in an independent tribe, the Cheliomyrmicini. 636 Bulletin American Museum of Natural History [Vol. XLV Legs long, the hind femur reaching to or beyond the hind margin of the second segment of the gaster; head large, the thorax moderately inflated. ...... Subgenera Eciton, sensu stricto and Labidus Jurine. 3. Two closed cubital cells. Thorax long and narrow; scutellum not prominent. Legs short and thick; tibie with a long spur (Congo) a5 pines: i cult deeeeaaaes Z£nictogiton Emery. One closed cubital cell. Thorax with the mesonotum much raised above the pronotum; scutellum prominent. Legs usually slender; tibiz with a rudimentary spur. .4nictus Shuckard. 3. Leptanillini Emery Leptanilla Emery. (Corsica, Sardinia, Barbary, Singapore, and Borneo). CERAPACHYIN2 Forel and PONERIN& Lepeletier Key to the Tribes! Pa 8,° 1. Claws pectinate. Mandibles articulated near the anterior angles of the head. Constriction behind the postpetiole feebly marked. LeproGenyinI Forel. Claws simple or toothed; in certain Simopone pectinate, but in these the postpetiole is separated by a strong constriction behind. ..:...500:56 0: «06 OU eS hig Se Epi lace 5 igs an 2. 2. Mandibles articulated to the middle of the anterior margin of the head, when closed placed parallel to each other in front of the clypeus; when opened they lie in a straight line parallel to the anterior margin of the head. Postpetiole not separated by a constriction behind................ ODONTOMACHINI Mayr. Mandibles articulated to the anterior angles of the head........ 3. 3. Postpetiole narrower than the following segment, forming with the petiole a two-jointed pedicel. Mandibles linear, very long. Antenne 12-jointed. Claws toothed. Metanotum developed dorsally, between the mesoscutellum and the epinotum. Myrmecint Emery. Not having all these characters...............00.+cereceeets 4. 4. Head flattened, much as in Dorylus: the face with two deep and broad antennal fosse below, in which the *antenne are in- serted close together, just above the short and obtuse clypeus. Frontal carine very approximate. Eyes absent. Mandibles For the convenience of identification of specimens the tribes of the Cerapachyine and Ponerine have been united in one key. sil Ct a ee eee ee ee ee) ee ee ee — _ Wheeler, Ants of the Belgian Congo 637 narrow, with three apical teeth. Antenne 11-jointed, their funiculus much thickened. Postpetiole very feebly con- stricted behind. Female unknown. . DoryLozELINI, new tribe. Head of the usual Ponerine shape. . Not with all these characters..5. Face on each side with a scrobe which extends to the hind margin of the head and is recurved behind the eye, so that it can take the scape and funiculus of the antenna. Mandibles triangular, robust. Antennz 12-jointed. Petiole with a ventral spine near its base; postpetiole separated by a constriction behind. IEEE os EE) SE PARAPONERINI Emery. Facial scrobes usually absent; when feebly marked (as in Acantho- ponera and Prodiscothyrea) they never extend behind the eyes; but the frontal carine often take up the base of the scape. In Paranomopone a deep facial scrobe on each side in front of the eye accommodates scape and funiculus................ 6. At least one ocellus in the worker. Body of worker and female elongate, cylindrical; pygidium impressed, armed at the sides with several stumpy spines (female as far as known winged). Antenne 12-jointed........... CYLINDROMYRMICINI Emery. Usually no ocellus in the worker....................---e008- 7; Petiole depressed, articulated over its whole width with the post- petiole. Antenne 12-jointed.....................00-.45: 8. Articulation between the petiole and postpetiole narrow; if broader (as in Prionopelta) the hind tibie# have one or no spur and the pygidium is not bordered by a row of small spines ........ 9. Two spurs on the hind tibia. Mandibles narrow. Thorax with dis- tinct sutures in the worker; the metanotum not developed dorsally. Pygidium not bordered by spines. AMBLYOPONINI Forel. Tibiz with a single, broad, pectinate spur. Mandibles in the work- er subtriangular. Thorax without distinct sutures. Pygidium impressed, with a lateral row of spines in the worker. ACANTHOSTICHINI Emery. Insertion of the antenne nearer the sides than the middle line of the head. Mandibles narrow, arcuate, with spiniform teeth. Antenne 12-jointed. Petiole with a high scale above; con- striction behind the postpetiole indistinct. Claws simple. THAUMATOMYRMICINI Emery. 638 Bulletin American Museum of Natural History (Vol. XLV Not having all these characters. ...............05-0c0ceeee> 10. 10. . Insertion of the antennz exposed...............-.0+4+-00008 At. Insertion of the antenne at least partly covered by the frontal carine. In Ophthalmopone almost exposed.............. 12. 11. Gaster strongly reflexed ventrally, or if not (as in Probolomyrmex and Escherichia) the frontal carine are fused together and with the elypeus. Mandibles subtriangular. Antenne 9-, 10- or 12-jointed. Tibie# with one or no spur. Claws simple. PROCERATINI Emery. Gaster straight. Frontal carine distinct from each other. Anten- nal fossa margined by lateral carine of the cheeks. CERAPACHYINI Forel. 12. Frontal carine remote, more or less parallel, or feebly diverging behind, without lateral lobe (except in the Neotropical genus Alaris) 6c siiawitee eee eee EctTaTomMINI Emery. Frontal carine with a lateral lobe..................see+e00- 13 13. Insertion of the antennz approximated; the frontal carine usually converging behind the lobe. . ... ............. ss +++ asd geen 14. Insertions of the antenne remote. Clypeus flat. The body entirely covered with a very fine pruinose pubescence. Mandibles subtriangular. Antenne 12-jointed. Middle and hind tibiz with two spurs. Claws toothed. PLATYTHYREINI Emery. 14. Middle and hind tibiz without spurs. Legs very long. Claws very large, simple. Clypeus short, its anterior margin arcuate, with little teeth. Mandibles elongate, narrow at the base, broad- ened toward the middle, with small, unequal teeth along their inner margin. -Eyes small. Antenne 12-jointed, filiform. Female ergatoid............... ONYCHOMYRMICINI Ashmead. Middle and hind tibize with one or two spurs. Antenne 12-jointed. Ponerint Forel. CERAPACHYIN Forel 1. Cerapachyini Forel In a recent paper on the Australian members of this tribe! I have followed Ern. André in restricting Sphinctomyrmex Mayr to the genotype S. stali Mayr, from South America; that species is only known by the female and the genus is therefore not included in the following key. This 1Wheeler, Wm. M. 1918. ‘The Australian ants of the ponerine tribe Cerapachyini.’ Proc. Ameri- can Acad. Arts Sciences, LIII, pp. 215-265. Wheeler, Ants of the Belgian Congo 639 _ female has the segments of the gaster separated by constrictions; but the E eyes are well developed; the thorax has distinct sclerites and was prob- _ ably winged; the antenne are 12-jointed and the pygidium is emarginate. _ Santschi refers certain African male ants to Sphinctomyrmex, but it is __ very improbable that they correctly belong there; and the same remark % applies to the male he describes as Lioponera. pss g SBE A -Gaster elongate, cylindrical, the segments separated from each ee other by pronounced constrictions. Female as far as known ergatomorphic or dichthadiiform, wingless and without dis- tinct sutures on the dorsal face of thorax. (Indomalayan, Papuan, Australian).................. Eusphinctus Emery. a. Antennz 11-jointed in worker and female. Subgenus Eusphinctus, sensu stricto. Antenne 12-jointed in worker and female........................... b. b. Worker with well-developed eyes and emarginate pygidium. Large, Og EES ee ee Subgenus Zasphinctus Wheeler. Worker eyeless or with very minute eyes. Pygidium entire. Smaller, 0 ferruginous or yellow species... .Subgenus Nothosphinctus Wheeler . Segments of the gaster not thus separated...................... 2. 2. Last antennal joint much thicker and larger than the preceding joint, forming a one-jointed club. Petiole not marginate on the sides. (Syria, Ethiopian, Malagasy, Indomalayan, Papuan, Australian, Neotropical, Texas). Cerapachys I. Smith. a. Antennw 12-jointed............... Subgenus Cerapachys, sensu stricto. Antenn# ll-jointed..................... Subgenus Parasyscia Emery. Mees MODMIREGGL g, ee cee ee Subgenus Oocerwa Roger. Ee ee Subgenus Syscia Roger. Last antennal joint not enlarged, though longer than the preceding joint, and not forming a distinct club.................... 3. 8. Funiculus of antenna terminating in a 4-jointed club. (Indo- malayan and Australian; Ethiopian and North African species EOI Es So ara Lioponera Mayr. Funiculus not terminating in a 4-jointed club. Petiole marginate on sides. (Ethiopian, Malagasy, Indomalayan, Papuan, and NE) OE ER a ae a Phyracaces Emery. 640 Bulletin American Museum of Natural History [Vol. XLV 2. Acanthostichini Emery Acanthostichus Mayr. (Neotropical and Texas). a. Female wingless, dichthadiiform; eyes small, flattened; ocelli replaced by three depressions. Male rather stout, with short and thickened antenn; thorax without Mayrian furrows. Worker known, with the characters given in the key to the tribes. Subgenus Acanthostichus, sensu stricto. Female winged and slender, with lengthened, cylindric gaster; eyes and ocelli well developed. Male rather slender, with slender antenne; thorax with well-developed Mayrian furrows. Worker unknown. Subgenus Ctenopyga Ashmead. PONERINZ Lepeletier 1. Cylindromyrmicini Emery 8 . 1. Antenne 12-jointed. Middle and hind tibie with two spurs. Claws simple. Female winged, similar to the worker. (Neo- Combe) fo crs ei cnen os ee Cylindromyrmex Mayr. Antenne 11-jointed. Tibize with a single, pectinate spur. Claws toothed or pectinate. Female unknown. (Ethiopian, Mala- MOY) os: os 0s: + ving tyne SR ota gta Simopone Forel. 2. Myrmeciini Emery Myrmecia Fabricius. (Australia, Tasmania; one species described from New Caledonia doubtfully belongs here). a. Worker: mandibles short and broad. Scape not extending beyond three quarters of the length of the head. Female and male unknown. Subgenus Promyrmecia Emery. Mandibles long and narrow. Scape almost reaching to or even sti beyond the occipital margin of the head. ...............0000005: b. Worker and female: mandibles with a long, recurved apical tooth, ee unequal teeth along the inner margin. Subgenus Myrmecia, sensu stricto. Worker and female: mandibles linear, always straight and serrate. Male WREAOWN 5565 oc. ohsa eee Subgenus Pristomyrmecia Emery. - 3. Amblyoponini Forel 3 The female, where known, is winged. 1. Mandibles blunt at the apex, with two teeth-rows on their inner margin. Clypeus denticulate along the anterior margin. — Frontal carine remote. Eyes present, but very small. — Sculpture coarse. (Malagasy, Ethiopian, Indomalayan). — Mystrium a Mandibles pointed |, «0. icc isaeeee secs nee +2. 2 ‘ 1922) Wheeler, Ants of the Belgian Congo 641 _ 2. Anterior margin of the clypeus arcuate and as a rule denticulate. a Antenne slender, not club-shaped. Eyes usually small. Teeth on the inner margin of the mandibles partly in two rows. Integument partly dull. (Mediterranean, Nearctic, Neo- tropical, Indomalayan, Papuan, New Zealand). Stigmatomma Roger. Be q _--~—s a._-—s«* Frontal lobes approximate. (Papuan). eas Subgenus Fulakora Mann. (Type: S. (Fulakora) celata Mann). Frontal lobes widely separated . . .. Subgenus Stigmatomma, sensu stricto. i _ Anterior margin of the clypeus without teeth, often straight or emarginate. Middle teeth of the mandibles in one row. Inte- meement smooth and shining.........................-.. 3. 3 Funiculus club-shaped, short and thick, slightly flattened. Eyes small. (Indomalayan, Papuan, Australian). Myopopone Roger. Dexienlos slender, filiform, hardly thickened toward the apex. . .4. 4. Frontal carine remote. Inner margin of the mandibles with a few teeth. Eyes very small. (Australian, Papuan, New aes Sire Sa ub oo e'g oom b 98 Amblyopone Erichson. Frontal carine apparently more approximate. Mandibles denticu- late along the entire inner margin. Clypeus produced into a rectangular lobe. Eyes absent or vestigial. Female unknown. SS ES Aer ere Xymmer (Santschi). Pt The male of Amblyopone is imperfectly described; that of Xymmer is unknown. 1. Frontal carine datinet. No cerci. Middle tibie with two spurs. Mystrium Roger. Frontal carine vestigial. Cerci developed................... 2. 2. Integument dull. Middle tibie with one spur. Stigmatomma Roger. Integument smooth and shining. Middle tibie with two spurs (generic identity still doubtful)........... Myopopone Roger. 4. Paraponerini Emery Paraponera F’. Smith. (Neotropical). 5. Platythyreini Emery Platythyrea Roger. (Tropicopolitan). 642 Bulletin American Museum of Natural History (Vol. XLV 6. Ectatommini Emery 8 (2 when known) Worker: face on each side with a deep scrobe in front of the eye; this scrobe incompletely divided by a longitudinal ridge into two compartments, one for the accommodation of the scape and one for the funiculus. Antenngw ending in an indistinct, 3-jointed club. Claws simple. Female ergatoid, with one ocellus. (Australian)............ Paranomopone Wheeler. Face without deep scrobes to accommodate the whole of the BNEONNB. ooo noicces.c 5 aus + one onne ae > 405 67 oe 2. Antenne ending in a 3- or 4-jointed club. Claws simple. Female winged, with eyes and ocelli..................2eeceeees: 3. Antenne not ending in a distinct club. Claws as a rule toothed or Wifi. oo osec cole cs Sn bee so cele e053 ob 0 0 nicoys (slain 5. Articulation of petiole and postpetiole not remarkably narrower than the postpetiole (as in the Amblyoponini). Middle and hind tibie with one spur. (Neotropical, Indomalayan, Pancaays 2 ivi 65 Rae a ea Prionopelta Mayr. Articulation of petiole and postpetiole much narrowed, as usual in this tribe. 6..i..0.06 0) A A a. PE 4. Petiole distinctly narrowed into a peduncle at the base. (Neo. tropical). ke A eee Typhlomyrmex Mayr. Petiole not pedunculate at the base. (Borneo, Papuan, Neo- tropical) ..5%c25 sade. eee Rhopalopone Emery. Basal segment of the gaster strongly curved or vaulted price so that its hind part is directed downward or even anteriorly..6 Basal segment of the gaster of normal shape.................. 8. Thorax with pronounced promesonotal and mesoépinotal sutures. Eyes normally developed. Basal segment of the gaster moderately vaulted. Petiole with a basal spine ventrally; postpetiole also with a ventral, flattened tooth, directed an- teriorly. Female and male unknown. (Papuan). Wheeleripone Mann. (Type: Wheeleripone albiclava Mann). Thorax of the worker without traces of sutures dorsally.......... 7. Eyes of the worker small. Basal segment of the gaster very strongly curved. Female winged, with two closed cubital cells, or er- gatoid. (Neotropical).................65.. Alfaria Emery. — Eyes of the worker larger. Basal segment of the gaster more feebly curved. Female winged, with one closed cubital cell. (Indo- — malayan, Papuan). 052550 di ten ete Stictoponera Mayr. Wheeler, Ants of the Belgian Congo 643 aa er Antennal fosse extending backward above the eyes. Epinotum with teeth or spines. Promesonotal suture very distinct in the worker. Posterior coxe unarmed. (Neotropical, Australian, _ New Zealand; including Heteroponera Mayr). Acanthoponera Mayr. Pixtennel fosse short or indistinct, as usual.................. 9. j - 9) Promesonotal suture very distinct in the worker; often mobile, or . at least interrupting the striation...................... 10. Promesonotal suture entirely obsolete, or impressed but not inter- | MNS. 3 Ue eee ohare tacos as 12. 0. Posterior coxe armed with a spine. Female winged, with one -__ elosed cubital cell. (Neotropical).....Holeoponera Mayr. asi Posterior cox unarmed. Female ergatoid or winged with two I cata eg kee i oe oe 11. a Worker: small; first joint of the funiculus very little shorter or even longer than the second; the latter as a rule less than twice as long as thick. Spurs of the middle and hind tibie sinuate and broadly pectinate. Female winged. (Indomalayan, [eenpeen, Australian)................ Chalcoponera Emery. - Worker: larger; first joint of the funiculus distinctly shorter than the second; the latter slender, at least twice as long as thick. Spurs of the middle and hind tibie straight or feebly sinuate, shortly pectinate. Female unknown, probably highly ergatoid. (Australian, Papuan)...... Rhytidoponera Mayr. Worker: mandibles long and narrow, obliquely truncated at the tip, denticulate along the inner margin. Female unknown. di oo, cp viediic os «vices Sages ces Emeryella Forel. Worker: mandibles triangular or linear, but not toothed along the inner (basal) margin. Female when known, winged, with two closed cubital cells. (Neotropical, Texas). Ectatomma I. Smith. a. Clypeus on each side with a tuberculate swelling covering the insertions of the antenne; mandibles triangular. Posterior coxe# unarmed. Subgenus Ectatomma, sensu stricto. Clypeus not swollen above the insertions of the antennw. Posterior coxe ee Ue OE NE sc vn ws Atel kc esa Sin Ok Rais d kp gasberss b. b. Epinotum with a pair of long spines... .Subgenus Poneracantha Emery. Epinotum unarmed or at most with short teeth...................... C. ¢. Mandibles triangular; the apical margin denticulate and separated from the inner or basal margin by an angle. Subgenus Parectatomma Emery. O44 Bulletin American Museum of Natural History (Vol. XLV Mandibles narrow and more or less linear; the inner or basal margin curves gradually into the apical margin which is not denticulate. Subgenus Gnamptogenys Roger. og The male of Paranomopone and Rhopalopone is unknown; that of Alfaria is doubtful. 1. Antennal scape much longer than the two following joints together. One closed cubital cell. ....:. 0 :....02s% 5 > »sle 0) oe ale ee 2. Antennal scape not longer than the third joint or if longer, wings with two closed cubital cells. ......:5.....2.--+eceevcrees’ 3. 2. Articulation of the petiole and postpetiole not remarkably narrower than the postpetiole................. Prionopelta Mayr. Articulation of petiole and postpetiole as usual, much narrowed. Typhlomyrmex Mayr. 3. - Two closed cubital cells: ............ 5.02. +<+ 55 ee 4. One closed cubital cell. ..... 0.0). .2...5...0005++s) se 5. 4. Scape as long as or longer than the second joint of the funiculus. Mayrian furrows on the mesonotum feeble or indistinct. Rhytidoponera Mayr. Scape shorter than the second joint of the funiculus. Mayrian furrows pronounced.....................Chaleoponera Emery. Ectatomma fF’. Smith. Emeryella Forel. ?Alfaria Emery. 5. Sculpture of fine, close striae. Petiole short, thickened behind into & W0GO. os a ee eee te tea Holcoponera Mayr Sculpture of coarse foveole, sometimes confluent. Petiole elongate not swollen into a node............... Stictoponera Mayr 7. Thaumatomyrmicini Emery Thaumatomyrmex Mayr. (Neotropical). 8. Proceratiini Emery 8 (2 when known) 1. Clypeus separated by a distinct suture from the front, cheeks and frontal carine. Antenne 12-jointed.................-.. 2.4 Clypeus fused with the cheeks and frontal carine, the whole form- — ing a plate projecting out over the mandibles; the antennz © are inserted close to the anterior margin of this structure. ...4. _ _ Wheeler, Ants of the Belgian Congo 645 . 2 cliieice 1 margin of the clypeus projecting in the middle. Thorax a" without dorsal sutures. Petiole more or less nodiform. Female winged. (Ethiopian, Mediterranean, Japan, Nearctic, Neo- = RI Srl), Niel. Giese lia ee. Seats Sysphincta Roger. ____ Anterior margin of the clypeus not projecting in front.......... 3. _ 3. Dorsal sutures of the thorax faint or absent. Petiole scale-like. Antenne not ending in a club. Female winged. (Nearctic, ee Indomalayan, Papuan)................ Proceratium Roger. _ ‘Thorax with distinct promesonotal and mesoépinotal sutures. Petiole decidedly transverse, less squamiform, the anterior surface being flattened. Funiculus ending in a distinct, 3- argrs jointed club. Female unknown. (Haiti). Bure. Spaniopone Wheeler and Mann. aa 53 “ Antenne RES a EA eg ge a a 5. __ Antenne 9- or 10-jointed. Basal segment of the gaster vaulted, : the remaining segments forming an anteriorly directed cone. i ‘Thorax without dorsal sutures......................... 7. 5. Basal segment of the gaster vaulted, the remaining segments form- ____ ing an anteriorly directed cone. Eyes present. Dorsal sutures of the thorax faint or absent. Female unknown. (South va SE a Pseudosphincta Arnold. aah Segments of the gaster straight, directed posteriorly........... 6. _ Eyes present, very small. Thorax with obsolete promesonotal and distinct mesoépinotal sutures. (Ethiopian). Escherichia Forel. Eyes absent. Thorax without dorsal sutures. Female unknown. EEN Probolomyrmex Mayr. Antenne 9-jointed. Clypeus forming a semicircular disc. Frontal caring small and short. Face without lateral depressions for the antennal scape. Female winged. (Nearctic, Ethiopian, New Zealand, Indonesian, Java, Papuan). Discothyrea Roger. Antennz 10-jointed. Clypeus forming a very short, transverse plate. Frontal carine large, the face deeply and broadly ex- cavated at their sides, forming scrobes for the accommodation of the antennal scape. Female winged. (Australian, Indo- On ig ga des ces cc cas 6 Prodiscothyrea Wheeler. Known only for two genera. 1. Frontal carine not fused with each other. Wings with one closed ene, sce ac prceckacceeds Sysphincta Roger. Bulletin American Museum of Natural History (Vol. XLV Frontal carine fused with each other. Wings with one closed cubital cell......................!...,--Discothyrea Roger. 9. Dorylozelini, new tribe Dorylozelus Forel. (Australian). co unknown. 10. Ponerini Forel 8 Middle and hind tibiw with two spurs.....................--. 2. Middle and hind tibie with a single, well-developed spur, page . always pectinate; the lateral spur rudimentary or absent. . The two spurs of the middle tibie simple, small. Median aa of the hind tibiz pectinate, the lateral one simple. Mandibles elongate subtriangular, curved downward. Eyes absent. Antenne thickened. Petiole with a ventral tooth. Female winged and with eyes and ocelli. (Neotropical, Indomalayan, — PDR) 630, ee es Centromyrmex Mayr. Median spur of both middle and hind tibie well developed, pee- tinate. Eyes usually present (in Pseudoponera very small or abwehit) 6... ee oa ee ae 3. Mandibles narrow, converging near the base where they are pro- vided with a strong tooth beneath; in front of this tooth they are projecting into a beak. Eyes very large, placed near the base of the mandibles. Antenne filiform. Claws bifid. Female winged. (Indomalayan)...... Harpegnathos Jerdon. — Mandibles of normal shape... ............ 2... 4 ae ee 4. 2 Anterior margin of the clypeus arcuate, with numerous denticula- _ tions. Antenne filiform. Pronotum with two spines on its _ anterior margin. Claws simple. Female winged. (Indo- malayai, Papudit). (sso teers c ss Odontoponera Mayr. Anterior margin of the clypeus unarmed or with two teeth. ..... 5. & Node of the petiole compressed above and forming a sharp edge, with a slight notch behind followed by a terminal blunt tooth. Antenne filiform. Anterior margin of the clypeus emarginate, — on each side with an obtuse tooth. Claws simple. Female © —e Te! unknown. (South Africa)............ Streblognathus Mayr. | Petiole shaped differently: ........ 5 0i.0545.4. bcs 0 oe a 6. Clypeus with a median, raised portion, produced in front. Female © WIIBOG, 55. win a «win, ona nse of vs p vip lhe tu Sept gee sac tee (fe Clypeus without a raised, projecting, median area.............. 9. ‘ 7. Decale: wings with three closed cubital, two discoidal and two sub- a _ Wheeler, Ants of the Belgian Congo 647 eae median cells, one of the latter very small. Raised portion of the clypeus excavated in the middle and bordered by lateral ridges. Postpetiole without ventral tooth at the base. Middle tibie and metatarsi furnished above with rows of spines. Worker unknown. (Ethiopian)........... Glyphopone Forel. _ Female: wings with two closed cubital, two discoidal, and one sub- 10. eS ga «ag dc vip a dsiin Wa vindeve me © 8 Be < 8. Female: median area of the clypeus moderately raised, convex, slightly produced in front, hardly carinate on the sides, almost - flat or very shallowly concave in the middle. Middle tibie and metatarsi furnished with rows of spines. Claws simple. Worker unknown. (Ethiopian)........ Leptopone Arnold. Worker and female: median area of the clypeus deeply excavated in the middle, shining, with heavy strie, bordered laterally by strong ridges. Antenne filiform. Middle tibiae without rows _of heavy spines. Postpetiole with a ventral, blunt, compressed tooth near its junction with the petiole. Claws with a small tooth near the middle. (Ethiopian)....Paltothyreus Mayr. Anterior margin of the clypeus with a tooth on each side of a median emargination. Claws with a median tooth. Female unknown. ME ag CPCS. eee sk ccc cea cs 9% Dinoponera Roger. Anterior margin of the clypeus not or bluntly bidentate; in the latter case the claws simple.........................-.. 10. Mesepisternum with an oval cavity leading to the first stigma which is covered by a small pronotal lobe. Petiole with a pair of spines directed backwards. Claws simple. Female unknown, probably ergatoid. (Indomalayan, Papuan, Australian). Diacamma Mayr. Mesepisternum of the usual shape.......................... 11. Claws with a tooth near their base. Cheeks carinate. Scape of the antenne compressed. Scale of the petiole thick, more or less nodiform. Female wingless, ergatoid. (Ethiopian). Megaponera Mayr. NG pete sc ths ila So cdwilct's ev oo ees 12. Eyes large, placed behind the middle of the head length. Female unknown. (Ethiopian)............... Ophthalmopone Forel base othe midi of the claws” Emery and B. Rectionds (Feral) ove deserined 09 having o email tooth nese the the claws. Yet they can not well be placed in Megaponera. as ip porate that, ped, for which the Ei Forel (t ohn gre eens, Sher wih po sepeeeie goe panes, & w mee ‘agensia Forel (type: Megale- 648 13. 14, 15. 16. 17. 18. Bulletin American Museum of Natural History [Vol. XLV Eyes placed in the middle or before the middle of the sides of the Cheeks carinate. Eyes placed about or slightly before the middle of the sides of the head. Female winged. (Neotropical). Neoponera Emery. a. Mandibles almost sublinear. Node of the petiole club-shaped, being gradually raised behind............ Subgenus Eumecopone Forel. Mandibles subtriangular. Node of the petiole not club-shaped. Subgenus Neoponera, sensu stricto. Cheeks not carinate. Eyes placed before the middle of the sides of the header a i is aol ge 14. Mesoépinotal suture obsolete in the worker. Usually large-sized species. Female winged. .......57....... 0. <. oe 15. Mesoépinotal suture more or less distinct.................... 18. Pronotum more or less a enareeate on the sides. (Neotropical to ROBOT... ocd ck ch bite ee hie Pachycondyla IF’. Smith. Upper part of the head separated from the sides and from the occi- put by a blunt ridge. Mesepisternum of the worker divided from the sternum by a distinct suture. (Ethiopian, Indo- malayan, Japan, Papuan, Australian). .Ectomomyrmex Mayr. Upper part of the head not separated from the sides and from the occiput by a ridge. Mesepisternum fused with the sternum in the worker (except in one Malagasy species).............. 17. Petiole surmounted by a flattened scale which curves back over the postpetiole and terminates in a comb of five teeth. Gaster without constriction behind the postpetiole. Epinotum with two stout spines. Mandibles elongate. (West Africa). Phrynoponera Wheeler. Petiole with a thick node, rarely somewhat compressed and dentate above or behind. Gaster with pronounced constriction between the postpetiole and succeeding segment. Epinotum usually unarmed. Mandibles subtriangular. (Ethiopian, Malagasy, Indomalayan, Papuan, Australian). Bothroponera Mayr. Mandibles subtriangular, with a distinct edge between the apical or masticating and the basal, inner margin. Female winged. (Tropicopolitan, Mediterranean, Japan, New Zealand). Euponera Forel. ~~ Wheeler, Ants of the Belgian Congo 649 a. Mandibles elongate, with an extensive masticating margin, which is armed with numerous teeth. First joint of the funiculus as a rule shorter _ than, or as long as, the following, seldom longer (including Xiphopelta Mandibles comparatively short, with small or a few teeth. First joint of the i funiculus noticeably longer than the following.................. ; ‘- Length 9to10mm. Mandibles with 8 teeth. (Malagasy). Subgenus Euponera, sensu stricto. ESS cali: dy io tua a's £C akg dn Ge hack ke Fdeeelaaestp c. c. Mesonotum convexly swollen, surrounded by a deeply impressed suture. Metatarsus of the middle legs without stiff hairs on its dorsal face. Subgenus Brachyponera Emery. Mesonotum depressed; surrounding suture not deeply impressed. Middle legs short, their metatarsi furnished with stiff hairs or spines on their ge a aa eR Subgenus Trachymesopus Emery. Mandibles long and narrow; their masticating, apical margin passes through a curve into their basal, inner margin. Middle legs short, their metatarsi with stiff hairs above. Pseudoponera Emery. a. Apex and masticating margin of the mandibles strongly dentate. Eyes very small. Female winged. (Indomalayan). Subgenus Pseudoponera, sensu stricto. Apex of the mandibles dentate, their masticating margin with feeble traces of teeth. Eyes obsolete or absent. Female unknown. (Ethiopian). Subgenus Promyopias Santschi. Mandibles subtriangular, very long, ending in a very elongate apical tooth, the.apical margin also with three long teeth. Clypeus unarmed, without projecting lobe. Eyes absent. Mesoépin- otal suture obsolete. Integument moderately punctate. Abdomen with feeble pubescence. Female unknown. (Indo- EN See gia a'o os od he's .s lads Hie 4 Emeryopone Forel. meee maving all these characters............................ 20. Integument very finely and densely punctate. At least the abdo- men with abundant pubescence. No lateral spur on hind and Rei Sig ae oie Ne wa UNG Gab c aciee's'o.< oe © « 21. Integument smooth, or sparsely punctate, or coarsely sculptured. Abdomen without pubescence. ........ 2.0.0... ...00005. 23. Clypeus with a sharp point in the middle of its anterior margin. Mandibles elongate, narrow, with 3 to 5 strong teeth. Eyes vestigial. Female unknown. (Neotropical). Belonopelta Mayr. re conics cco xts's ck Weer ss cate 22. 650 22. 24. 25. 26. 27. Bulletin American Museum of Natural History [Vol. XLV Antenne ending in a 4-jointed club. Eyes vestigial or absent. Female winged, with eyes. (Ethiopian, Indomalayan, Pain En Es Pe a eo eee eae Cryptopone Emery. Antenne without distinct club or the club is 5-jointed. Eyes small, sometimes obsolete or absent. Female winged; sometimes also an ergatoid female present. (Cosmopolitan). Ponera Latreille. The two lobes of the frontal carine fused into a plate which is slightly notched in front above the clypeus. Clypeus much produced into a broad plate, truncate in front and with sharp lateral angles. Mandibles subtriangular, their apical margin strongly dentate. Eyes small. Female unknown. (Ethio- pnt) PRONE et part, <<) ebay ge, Asphinctopone Santschi. Clypeus not projecting, or with a narrow median lobe (Trapezio- pelta), but in that case the mandibles are linear. .......... . 24. Clypeus short; the frontal carine contiguous and forming a plate which is raised above the clypeus. Mandibles slightly curved, linear, broadened and spear-shaped in their apical third, ending in a blunt apex; hollowed out into a rim along their inner margin. Eyes vestigial. Female unknown. (Ethiopian). Cacopone Santschi. Clypeus, frontal carine, and mandibles shaped differently. .... . 25. Clypeus with a projecting median lobe. Eyes present or vestigial. -Mandibles linear. Lateral spur of the middle and hind tibie small, but present. Claws simple. Female winged. (Malayan, Papuan).................. Trapeziopelta Mayr. Clypeus without projecting median lobe..................... 26. Mandibles falciform, flattened, broadened towards the third of their length, ending in a sharp point. Eyes present. Female winged. (Ethiopian).......... Psalidomyrmex Ern. André. Mandibles not falciform, nor flattened...................... 27. Mandibles linear, arcuate, pointed, with one tooth or two spaced teeth along their basal, inner margin. Eyes vestigial. Female winged, with eyes. (Indomalayan, Papuan, Ethiopian). Myopias Roger. Mandibles blunt at the apex, linear, with a few irregular teeth along their inner margin. Eyes small. Female winged (or in some species ergatoid?). (Ethiopian). .Plectroctena F. Smith. _ Wheeler, Ants of the Belgian Congo 651 oi _____‘The male of the following genera is unknown: Glyphopone Forel, _Leptopone Santschi, Ectomomyrmex Mayr, Phrynoponera Wheeler, _ Pseudoponera Mayr, Emeryopone Forel, Belonopelta Mayr, Cryptopone Emery, Trapeziopelta Mayr, Myopias Roger, Asphinctopone Santschi, _ and Cacopone Santschi. Bs oe Middle and hind tibie with two spurs, those on the middle tibie +. WT MONE oS ce eccrt peor cept eee wedee ks 2. ______ Middle and hind tibiew with a single spur.................... 12. __ 2. The two spurs of the middle tibie very small, simple. Median spur ; of the hind tibie large, pectinate. Mandibles very short, without teeth. Antenne short. Pygidium without spine. Mesonotum with Mayrian furrows... .Centromyrmex Mayr. _____ Median spur of both middle and hind tibie large and pectinate. . .3. _ 3. Postpetiole pyriform, almost forming a second node of the pedicel. Mandibles rather vestigial. Antenne long. Pygidium with a long terminal spine. Mesonotum with two deep, con- Ny wa os ace $14 405 Harpegnathos Jerdon. Postpetiole as usual, with a feeble or indistinct constriction be- EE UR OE nos a 0 5 op cl bag keer oe