BOTANICAL MUSEUM LEAFLETS HARVARD UNIVERSITY BOTANICAL MUSEUM LEAFLETS HARVARD UNIVERSITY VOLUME XXVII BOTANICAL MUSEUM CAMBRIDGE, MASSACHUSETTS 1979 DATES OF PUBLICATION - VOLUME 27 PN Se hadadoaee vamena tae cdaersyaxaasiod uae einueneciaseies March 6, 1980 PU ee ccs ia ens arses dass ens aes ee en css March 28, 1980 POs Broa gessiacnuveenaciuaueneinoeunssapatswacicsnmcdes May 25, 1980 NOS. 7-9 ..cccccccccsccccecccceccecccceceecsnceecuceeseevevacs June 16, 1980 PVs A piseauccaaeeaana vents nesucrcssvceeveurisadaecouens October 28, 1980 CORRECTION Page 177: Fifth line from bottom should read: aa. Petals never mucronate; plants rhizomatous. .... 14. S. brevilabris alliance TABLE OF CONTENTS NuMBERS | -2 (January-February 1979) Salpiglossis, Leptoglossis and Reyesia (Solanaceae). A Synoptical Survey by ARMANDO T. HUNZIKER AND ROSA SUBILS .........+0000. l The Identity of Amazonian and Trujillo Coca BY TIMOTHY PLOW MAN sieccciaverssciexsadeicis tunes tener sesencieae’ 45 NuMBERS 3-4 (March-April 1979) Monopteryx angustifolia and Erisma Japura: Their Use by Indigenous Peoples in the Northwestern Amazon by DARNA L. DUFOUR AND JAMES L. ZARUCCHI ......-.++++ 69 Studies in the Genus Micrandra II by RICHARD EVANS SCHULTES......:cccssosceccsscesscsnsceveoenes 93 NuMBERS 5-6 (May-June 1979) Ancient Gold Pectorals from Colombia: Mushroom Effigies? by RICHARD EVANS SCHULTES AND ALEC BRIGHT.........- 113 Discovery of an Ancient Guayusa Plantation in Colombia by RICHARD EVANS SCHULTES.......scsccsescesssecscecrecencess 143 The Flowers of Ilex Guayusa by MELVIN SHEMLUCK .......cccccscccncecccescceeescseeseensenees 155 Chemical Test for Silica Determinations as an Archaeological Feature of Ethnobotany by ELIZABETH A. COUGHLIN........scseeeceecescesceeceesenseaes 161] NuMBERS 7-9 (September- November 1979) Systematics of the Genus Stelis Sw. DY LESLIE Ay GARKY sisieicrsers cami mene 167 NUMBER 10 (December 31, 1979) The Ethnobotany of the Flathead Indians of Western Montana Dy J RPPREY A. FART viccinsricvesonssndatdicinisieersaeeaioeeee’ 261 INDEX OF ILLUSTRATIONS PLATE Apatostelis BDO ics th inca pds aaay ahaa araranivany ey eas wrelaned Mieneaneteseniaeetts 88 PUTO VIS iis 0h code binned civ saave cans eeceserneaenuseensederesseereces 88 RAC CALA 2 povcccreciuseduscnbansewsannnsdseensterenesiaeeereeanser ani ens 88 PAIRS SECS cacccecaciecbednenvicckaudenn tanndva ese seanenaeteseeseseeenes 91 ONE ois. d wicca pviwnnscunssens dhcesk beanies sanesemeapeamareeiatas 89 BOS Scares oink Sa Se sence ceay ce sksansdisa Gaerne 91 RSs ee CON i sian eric ederssscavadewninshssceremecipeienietatovanate 90 PIOSSUNAG hi cigsnvarcaueixeersahter tenes yp ceepcaxclveanyenreieeahoueas 89, 92 Be Meg ceusnvescceeseeseuceneohinsicrar ine nines erat eeatenseuseer ens 91 DOA ciccxstetuewsisn ean sosanersrtaveieseuanass seyraneteanataeinaees 9] TE OES sataesae in edcenns ina gateseedncerreveesaenemsaneer 89,92 RCE ccc scsi iitonceeae caer eiennhenaavsertedeavaldasanenes 93 PARLE sacs capscdsiacdotccsactraedincerseeerpanenanesntaneneeduccosce 94 AOU Os focctagnce ciasnn er ceenacesvpeevesera ewer eee eokerne es 97 PRN aie vic d ate vinnie eecwneeadeieexcersondoesuaa rue eeienaesedes 96 PVT IIE fsck caaunaeencneneans bencqedeeenineetusenstieeesessa arise on 89 CLIC Eb nae ee eee enn Teena eer ere cre here rr rere 9] GOS Beta eck cienesteoccdtertiadliiciedwewertisa Peenes ee ieaeneeea eee 91 DONS OIC As icnascicssncéesasiule nen term nce eee aaatanears 94 PCT SUAS aad cue cenenanasasssdesdevestes4tensiamenine Masanaieenundens 95 PUDONG cece vincnnnsesscessdedutiencsesderiebewiscnctseniscanaceesedawienrrs 95 Qihisls gi litt |; Seen enn te Tree TT orem mer rrr ore rr nr T er reer ret 95 BB el cscs warn Cenonaterveenseiesdsacsedasnansueoeseasees esse 96 SA ee secrveciecercire ii ewricarsastiineiosenntsetneir teense 96 POM 6 cacrsaaigvnwe css eiad eves eas ad ae issaseniedcceace 96 Colombian gold “mushroom” pectorals showing more or less Tealistic AMpPHiDIAN TOMS v1.cesrcssaxcaservies 26-29 highly stylized amphibian TOrMssi.sishsnnerseceesisattesesens 30,31 SOUP AVIAN CMU ES, 502 ca cacniyeesaarsicsasivnsnevareenestyanienads 32,33 Stipe: Of MUSHTOOM-HKE COMES ncuisascosicsecs cis crepeees 34,35 three (instead of two) dome-shaped objects................006 36 Colombian Quimbaya Culture, gold seated man.............66. a] Bie specks cs hitdaesecaphcrwesewe deneciebansaseinmeaecnranesac 16 WAG KOTE OT SCCUS co iiscieecc che ceri rset idee eens kee shawserns es 22 freshly Narvested leaves and: IF Wits) cc scscsenvesnisasabstangeeewens 20 PUMICE OL wai e esraneversaresseneeterscees se rionel opie aries 17 Erythroxylon Coca var. Ipadu ef) Vase ort 0g | 2) go, ener ge ee ee 9 PAIGE DIAG AN, TOW Cl ise ys ececectresaarene sas acento eae 10 habit of plants in cultivation .............cccccceeeeeceseeeceneeees 1] Erytnroxyion. Hardin, HOW pe Oli sacacccdcascsesevaxvreeexacivacs 15 Erythroxylon novogranatense var truxillense BA At TV WN os 2c cy. cr vnctonptnavenyeat caiaeo tive eesen seeeeees 12 MOWETING OLARCH OF jos o55 is eect isidenirensinacsenderein cess 13 plantation of Trujillo coca showing use of pacay (Inga Feuillei) as a shade tree................c.cece eee 14 Ilex Guayusa DUA CT A TW irises easa ie ie inn tn naaueaunepe nes deeeieucneceees 42 old trees in Alto Afan, Pueblo Viejo, FUNIUMAVO OLOM DIS 5. icsseccsivicinenarapisasdesuseesaaancaays 40,41 Leptoglossis BUD GV Ay ci spexdsepeveiaucoceseucnsedcopsecasasniesacncaeaiaweuasexsannaes 5 geographical distribution of specieS............ccccceeeeceeeeeeees : eRe C: lo Een eee on ne en ee ee ere 4 ME OU Ais cas sad eecesndeue vere auneiss io oatn tacos eins denis eevaneissoas 6 Micrandra RAV op pap enes a eeoiet brat ecleaup ea pendec neon ea sswatsauaieanahandadeereennves 25 siphonoides, flowering branch Of............cccccccceeeeeeeeeees 23 SpEiccana, DULLVESSEd TOOTS Of pediescnaisiciesoncerccssbeaxsnedeed 24 Monopteryx angustifolia ..............ccccccceceeeeeeeveeeeeueeeseens 18 PU CCS ON SOO TUE ia sce eee tees lordsseneeeeesnewens 22 ee acs eeu weneia vite vgn oe ser eeerecenvplaweinisess 19 BOO A TOUS OE pee ssstcgiauencisseen epaiaeae ane thsuaeacieescrcsens 21 Pottery vessel in form of human head. Viru and Chicamac Valley, Peru, Early Intermediate Period, Mochica TV siciccsccscsxsasvesssusancesex 38,39 Reyesia, geographical distribution (os Wy acres (cle 0} Sere ere tee ree re ee ne RoE ee a eee a Te eee 7 SY CIA Pat VORA yyy vecrncevesciverdelueenetceesedsuniaiectdrieteenornes 8 Salpiglossis, geographical distribution ON BCG Ol ashe paicn sees satan aeaucu ces yerescaersuceasnteneatess tie l a DIPIOSS IS BUA lctrani cers cs acacia gsaaieereesoniasiaeciases 2 Stelis OC Uae cottey pawn ce cavers sag atasruivnndacsans xg gn txenaneev ere 47 CTO a hie cacanancnenna ya ceceaseekitoiaueustsevaraneviandendnesiareteded 56 ATMS exiotaeeseniaiavsdaasask buibeventsdvn channel btepeeiaysatesssaueaeusa 58 Stelis (continued) BNE asta caaauiicuny ea cecnceest ers eneckowees sansstassanauiiaugunneeiaess 81 Re se can ktrs cide dhe ee enone tanta nenens aaa eaieseaeoe eer eneeaia 50 SOARS Ea chase Whi eoreakewuehsseanensnaarasunuenincns a ederses 74 BE ova ive nnd nena vceeitdavasuesaskladauwainseie+saveesaneesseer 65 DRS AG Foch udexsicccnexisunenieseydcnseeeieacnenesasasauniegs besten cones 44 RR st cpchg vtinaaau wes cuntinans sakaen rider miniwiateneenep eres 69 DINO Pes) os cape anp nanan ieeeelsdecan neers gacasenaesyeereuenecasyaescas 78 OIA one ctcadeseieteciancaiansiacrannieusecies i hlacnseenenentekapers 67 OIG caieconwaresasuxaresiaka i xiscuseseireauvocstesexsaeuaiiepssedes 68 PT ITA COA 5 ac cach ccevasadeeanrneeereimiaseebaKesdeasesakaewsaseieuess 62 i MEN cae cane caviar ahaha ye ade ae a Sete a eae ese aes 48 BRE ruses ve sc leds places sncven sess iuen ei auawinuas ons vceeeagcokbesess 85 Ba ces ates eure ind men eb anor ss enna ne SERPS 47 Dt eo dsc haan aces veaeneu sinscuneusd ous beneneuneehaesanersaen 70 PB ONS 5545 sa sacacdnretetvtecieadsayeeneesesns a gesinureyceedereeserens 77 BU os ls.a sb eran big dessa evn da hada verse eon sy ers tanensanwnaceteacess 50 POON d tess titative paidacdeteauasasacneeeeserecseonienerees at Bh EIR a ices sera wee es deagss bene ceencneincsoreunneeseerecseees 78 Be ncn g esa acute aero erecta cree essa iadadaienvorteaxee ss 59 BROS OW OS As 05444 2dincadexnenedicaatsverconsegesaetenminasedeneouests 56 reel [ef cle) 1 2 ee a ae ee ae ee Onn Te Pee eee me ere os en ene 72 aro LE re ee errr re Tore er eter nr rem ret c rrr 44 CATE INCE, aise crac ecacpalek ghtnrhawenss Lecgsereatestuamenzeminays 76 el PIS Di tadcuiensvancaicas iebeeneneseaeieeb i vec dsaemyseseeean nee 56 NE ox weeny t'ccceeyviadinsaanaekaeuinaeuianssavederkiexaweesians 48 COTO IISIS os hcccnnsaeeeseededa ese seeeswiadsevesncxcacecesediesseanisas 79 es coca hci a casnnecuncnwovcecndctenstaesnseeasacusanieiesPises 69 CAE ADO VOSS goscsuscasssrsacsasdan er vanearer encerieetenestsenness 48 CTS 1G ode wcahends ka dudes sundnvuwlaiacacshesceceesictraesaseneas 57 CRO TGAinsdutrexnierscwssuarscosavicrasistsdeee eiejrcenesiees 65 NN civ nsaawae cas sete een vaaeee ea sonn nies oaeese teases 69 ered eG 405. | ee ee ae oe nee ee rene nen ee eee ee renee 49 NOE coca tiecc canine ci assidanedaencapensanag pecadioutaetewnnenies 45 VR Ea gcse nde od earlobe sans Fonduendnvestaresesececennaness 8] COM ccc ss retusa enka eens asawed ie eeneabaesnsenaesedaaus 51 COORG Fen ican se eee es 66 Ce Sis io pv ia wie rnas put ewkeened ie kalaenasers Consensastewtsaas 49 aU EG iccaeaerssabvesscetseqruedessaccstsseuncespeesassaneeresens 73 CURED dsakcsneaes oqeenzcinsinanteceerressaraantinantseseaneeennats 87 Stelis (continued) UG oe ce econ gna cone ardesoenes ean aasecnsauecosin gen eeees 86 OUMOMIVALAL 56a coca scene se sude cacao rs oeeacecrereste tans pueeed ene neen ue 55 AY 62 [1 6: | Bea een en Oe ne en eT Aer er 78 “cyl cel (a) eLe ro) FR eee eee ag Pe ere ee ME eee eS ee 7 AGN ICY As cons score iaaticn ss vai usheccduaiagn saieysancen inne tieees 82 C1 ofc} 601 1.1 6 a eee eee eae eee Ee ee Cen Ae eee ener nt renee V2 TBS hs 2a esse eee choses a pean aera eeanvaseneieumnenoeys 44 ETS cvncay sce svanustayacveansnasadeaieeuereesssusassecatsiaunieeiats 67 SCO ashore scteceasen se aciaessaces fecsenaaaendssassesmeesamnraienests 64 POUPNS Or VU) citeuddianeacsweduenrnes ceuanereinveanreecaanriceas 46 CUP tC POU Sic sisieisiesisesiaivisenneieneieeeiaaaneiniwarianssees 82 1 AUSSIE 1s ta-centeasnua tuvanencnsnsenniete anenkuser It may be worth pointing out that we apply the term “staminode”™ to a structure derived froma partially or almost totally atrophied stamen, usually consisting of a filament and a tiny distal sterile vestige of an anther: in most cases, this vestige appears merely as an apical swelling of the filament, but in extreme cases it disappears, and the only structure left is the filament itself. the level of adnation of the filaments to the corolla tube. Lepto- glossis linifolia is unique in this respect, because its very short filaments are adnate to the base of the enlarged distal sector of the corolla tube (PI. V. M: Pl. VI, K,O, T); on the contrary, the other SIX species display long filaments adnate up to halfway of the corolla tube or at lower levels (PI. IV, K; Pl. V, A. K). The pollen grains are separated in all the species of the genus. The disk is usually conspicuous, large, fleshy and with a bright reddish- orange colour. Gametic chromosome number: n = 10 (Subils, 1979: 19) Zygotic chromosome number: n = 20 (Diers, 1961: 451) Diagnostic features at the species level.—L. lomana and L. acutiloba are the only species in the genus with two kinds of leaves: the basal ones (with a long petiole and a broad blade) forming a rosette at ground level, and the cauline ones (sessile and with a linear blade). The other five species, on the contrary, exhibit only cauline leaves, either with long petioles and broad blades (PI. IV, C, D), or sessile with narrow blades (PI. VI, B, L). The pubescence of stems, leaves and calyces is another useful character for diagnostic purposes: 1. Darcyvana, L. Ferrevraei and L. /omana show predominantly branched hairs (Pl. IV, B, H, J); L. albiflora, on the other hand, displays only simple hairs (PI. V, D). In L. schwenckioides both simple and glandular hairs are noted (PI. V, L,O). When capitate hairs are present, the head and the foot may be either unicellular or pluricellular (Pl. V, O, P: Pl. VI, A). Another noteworthy feature, the slightly arcuate hairs, is known only in L. acutiloba (Pl. V, E, F). The calyx is, as a rule, actinomorphic and S5-cleft (Pl. IV, G, I; Pl. V, N), the only exception being L. a/biflora, with a zygomorphic and 7-10-cleft calyx (Pl. V, B). KEY TO THE SPECIES |. Filaments inserted at the middle or low part of the narrow portion of the corolla tube. Calyx 5-cleft, actinomorphic, or 7-10-cleft and 7zygomorphic. 2. Calyx 5-cleft, actinomorphic; plants with or without a rosette of leaves. 3. Herbs with stems lignified at lower part. Leaves cauline: the rosette leaves wanting. 4. Hairs simple and capitate. Plants (0.1) 0.3-1 min height. Flowers in lax cymes aggregated on lateral branches, forming false racemes. LANGA SOS Ueucu. daa klay iene s-caeeae eee |. L. schwenckioides 4’. Hairs predominantly branched (a few simple, and/or glandular ones as well). Plants 0.05-0.2 m in height. 5. Branched hairs most frequently of dendroid type, with more or less stiff cell walls. Sessile or subsessile flowers arranged in contracted cymes. Petiolate leaves with elliptic blades. Vegeta- tive internodes comparatively long (1.2-5.5 cm long) Werth a ee MG el he assy Shae 4 Waa bee AON 8 4 We 2. L. Ferreyraei 5’. Branched hairs prevailingly of the bifurcate or trifurcate type, with soft walls. Flowers pedicellate arranged in lax cymes. Sessile leaves linear or linear-spathulate. Internodes compara- tively short (0.5-3 cm long). ............... 3. L. Dareyana 3’. Short-lived tiny herbs lacking lignification, 0.05-0.3 min height, with two types of leaves: the basal ones forming a rosette at ground level (with a long petiole and a broad blade), and the cauline ones (sessile and with a linear blade). 4. Calyx with a dense pubescence of branched hairs (mainly bifur- pated). mies DUNCLOUS cine 4e dees oe eaN wen ews 4. L. lomana 4. Calyx with a few simple and capitate hairs. Stems glabrous Pree Eas Wiaipes ae haa. a0\S 0, 4/18) a cat & Gieyal ee’ eect ie-ins 9 Gud @ a a a hian'e 5. L. acutiloba 2.’ Calyx 7-10 cleft, zygomorphic. Annuals 0.1-0.3 m. alt., lacking a rosette Oe AVES: 65 ins dhe Ree eee ean oe we eee be ee 6. L. albiflora 1.’ Filaments inserted near the base of the enlarged portion of the corolla tube. Calyx 5-cleft, actinomorphic. Geophyte with gemmiferous roots, (11) 14-35 (40) cm in height. only with cauline leaves. .............. 7. L. linifolia 1. Leptoglossis schwenckioides Benth. (PL. 3c Peo no Bentham, Bot. Voy. Sulphur: 143. 1845. Peru: Huaman- tango. HOLOTYPE: Huamantango, Barclay s.n. (K). TOPO- TYPE: Quebrada de Guamantango, G. W. Barclay 2316, 10 Jun 1938. “Shrub 3 ft. high; fl. whitish” (BM).— Bentham, in De Candolle Prodr. 10: 196. 1846.—Miers in Ann. Mag. Nat. Hist. Ser. 2, 5 (25): 35. 1850.— Miers, Ill. South Am. PI. 2: 64, pl. 53. 1849-1857.—D7’Arcy in Ann. Missouri Bot. Gard. 65: 710, f. 3A. 1978. Salpiglossis schwenkioides (Benth.) Wettst. in Engl. u. Prantl, Natitirl. Pflanzenfam. 4 (3 b): 36. 1891. Lack of field observations make it imposible to state whether or not the plants are annual or perennial. Specimens with a lignified base may reach | m in height: on the other hand, small floriferous herbarium specimens (v. gr. Sagastegui et al. 7528) may measure scarcely 10 cm. Zygotic chromosome number: 2n = 20 (Diers 1961: 451); this chromosome count is based on Diers 1/08; we have seen a dupli- cate of this collection at Stockholm. ADDITIONAL MATERIAL. Peru. Dept. Ancash: Prov. Recuay, Marca (Chuccho), elev. 1700 m, Gomez 473, 14 Apr. 1965. “En arenal” (USM). —Dept. La Libertad: Prov. Trujillo, Cerro Chiputur al N de Trujillo, 1000 m.s.m., 4. Lopez M. 0739, 9 Sept 1951. “Loma arcillosa, fl. amarillentas” (USM; US).— Dept. Lima: Zapan, 1800 m.s.m., Vilcapoma 97, 20 Aug 1972. “Hab. pedregoso, borde de carretera, flores amarillas” (USM).—Cerca de ruinas de Cajamarquilla, 400 m.s.m., Ferrevra 8363, 24 Jun 1952. “En cauce seco; flores amarillas” (USM).—Prov. Canta: cerca de Canta, 2200-2500 m.s.m., Ferrevra 7255, 8 May 1950. “Hab. pedregoso, falda de cerro” (US). Canta Valley. road to Huaral, 20 Km NE of Trapiche, 1000 m.s.m., Hutchison 1032, 5 Aug 1957 (M; US; G: S: USM; NY; GH; K; U).— Entre Puente Trapiche y Cerro Huachoc, 800-900 m.s.m., Ferrevra 14826, 5 Dec 1962. “*Hierba en cauce seco; fl. amarillentas’ (USM). Prov. Huarochiri: hillsides above Santa Eulalia, 1200 m alt., Goodspeed 33098, 10 Apr 1942. “Bush | m high (G; GH; US).--Santa Eulalia, Weberbauer, [year] 1924. “Planta sufrutescente, fl. amarillo-parduzco” (USM). Bei S. Bartolome, 6Ostl. von Lima, ca. 1800 malt., Diers 1/08, 26 May 1959. “Bliite weiss-gelb; krduterarme Bezirk der Wiistenpflanzen” (S).—San Bartolome, (Lima-Oroya Railroad), 1500-1600 m alt... Weberbauer 5290, Apr 1910 (GH). Prov. Chancay: entre Trapiche y Huaral, camino antiguo por el cerro Huachoc, 800-900 m.s.m., Ferrevra 11767, 20 Mar 1956(USM).— Prov. Cajatambo: Ambar, 2000 m alt., Stork 11455, 16 Apr 1939. “Herb to 0.3 m; fl. pale yellow; a root perennial, in sandy gravelly hillsides very dry except in rainy season” (GH). — Dept. ?: Cum- ing 1010 (K).— Cuesta of Purruchuca, Cuming 1071 (K).— Peru, Mathews 1011 (BM). Dept. Cajamarca: Prov. Contumaza : Portachuelo (Ascope San Benito), 850 m.s.m., Sagastegui et al. 7528, 6 May 1971 (MQ). Distribution. Endemic to sandy and stony slopes, usually between 1000 and 2200 m high, in the Peruvian departments of Lima, Ancash, La Libertad and Cajamarca. 2. Leptoglossis Ferreyraei A.T. Hunz. et Subils sp. nov. (Pl. 3,3 PL 4.) Herbae vestitae ca. 10-20 cm altae. Internodia vegetativa | ,2- 5.5 cm longa. Pili diversi: plerumque ramificati raro glandulosi (capita unicellularia vel pluricellularia; pedes semper unicellu- lares). Folia petiolata; laminae ellipticae 1,5-3,5 cm longae, 0,3-1 cm latae. Inflorescentiae cymosae, densae, internodiis brevissi- mis. Flores sessiles vel subsessiles. Calyx 3,5-4 mm longus, lobulis 5 aequalibus. Corota 9-15 mm longa. Stamina 2; staminodia 3. 16 centrale breviore quam lateralia. Fructus quam tubi calycini aequantes vel leviter longiores. Semina matura non adsunt. Piliferous herbs about 10-20 cm alt., with the basal parts ligni- fied. Vegetative internodes 1.2-5.5 cm long. Hairs mostly branched with rigid and rugose cell walls, a few glanduliferous in calyces and stems (the head unicellular or pluricellular, but the stalk always unicellular). Leaves petiolate exclusively cauline: blades elliptic 1.5-3.5 cm long, 0.3-1 cm wide. Cymes contracted with very short internodes. Flowers sessile or subsessile. Calyx 3.5-4 mm long, with 5 more or less equal lobes. Corolla 9-15 mm long. Stamens 2; staminodes 3, the central one shorter than the 2 laterals. Disc red. Capsule as long as the calyx tube or slightly longer. Mature seeds not seen. Peru. Dept. Arequipa: Prov. Caraveli: Lomas de Jahuay, entre Nazca y Chala, 300/400 m.s.m., Ferrevra 14020, 2 Dec 1959. “Fl. amar.-verd.” Holotyvypus (USM). Obs.— Together with 1. Darcyvana, this species differs from L. fomana and L. acutiloba in the basally lignified stems, and by not having leaves disposed in rosettes. As indicated in the key, L. Ferrevraei is distinguished from L. Darcvana mainly by the char- actertistics of the pubescence, the petiolate leaves with elliptic blades, the contracted cymes with sessile or subsessile flowers. and the longer vegetative internodes. In honor of Dr. Ramon Ferreyra of Lima, who collected the holotype specimen. His name is latinized as Ferreyraeus. ADDITIONAL MATERIAL. PERU. Dept. Arequipa: Prov. Caraveli: Lomas de Pongo, Velarde Nunez 534, 2/3 August 1947 (US).—Lomas de Jahuay, entre Nazca y Chala, 400/450 m.s.m., Ferrevra 11490, 11 Oct 1955. “Fl. parduzco-purptrea; hierba. Hab. arenos.” (USM). Distribution. Endemic to the coastal xerophytic vegetation of southern Peru (Dept. Arequipa). 3. Leptoglossis Darcyana Hunz. et Subils (Pl. 3, 6.) Hunziker & Subils in Lorentzia 3: 15. f. 1. 1979. Pert. 17 HOLOTYPE: Dept. Tacna: Prov. Tacna: Lomas de Sama, 530 m alt., Zegarra 271, 6 Oct 1972 (US).—PARATYPES: Dept. Arequipa: Prov. Caraveli, Lomas de Atico, 50-100 m alt., Ferreyvyra 13923 (USM).—Dpto. Tacna: Unos 50 Km al N de Tacna, 500/600 m.s.m., Ferrevra 11660 A, 1 Dec 1953(USM). A rare endemic species of the peculiar “lomas” type of vegeta- tion, found at the desertic Pacific coast in southern Peru, from Arequipa to Tacna; it is closely related to L. schwencktoides, and even more to L. /omana and L. acutiloba. These last two species are sympatric with L. Darcvana, differing in their herbaceous ephemeral stems, with two types of leaves: basal petiolate leaves (at congested internodes), and sessile cauline ones at the remain- ing upper nodes. L. schwenckioides, on the other hand, grows at higher elevations in central and northwestern Peru, showing taller stems, and its pubescence lacks any sort of branched hairs. 4. Leptoglossis lomana (Diels) A. T. Hunz. (Pl. 3, 4; Pl. 5, H, 1.) Hunziker in Kurtziana 10: 46. 1977 (based on Leptofeddea lomana Diels).—-D’Arcy in Ann. Missouri Bot. Gard. 65: 709, f. 3 B. 1978. Leptofeddea lomana Diels in Repert. Spec. Nov. Regn. Veg. 16: 193. 1919. ISOTYPE: Peru, Mollendo, 20-100 m.s.m., Weber- hauer 1486, 2 Oct 1902 (K; also photograph Field Mus. Series nr. 3064 at CORD et GH, of the holotype from B). Tiny ephemerophyte unique to the “lomas”, a xerophytic type of vegetation, insouthern Peru(Dept. Arequipa), near the Pacific coast (20 to 500 m alt.). ADDITIONAL MATERIAL. Peru. Dept. Arequipa: Prov. Camana: Lomas de Camana, 100 m.s.m., Ferreyra 11697, 2 Dec 1965. “Flor amarillenta; hab. arenoso” (USM). Entre Camana y Arequipa (Km 161-162), 400-500 m.s.m., Ferrevra 2574, 10 Nov 1946. “Hab. arenoso; fl. amarillas” (USM).—-Prov. Caraveli: Lomas de Capac, Km 651, cerca de Chala, Sco/nik 1020, 27 Aug 1948. “Flores amarillas” (CORD). Lomas de Capac, Km 648, 200 m.s.m., Hutchison 1298, 14 Sept 1957. “Not over 10 cm tall incl. inflorescences; fl. dirty cream” (M; NY; G; K). Prov. Islay: south of Mollendo, Mexia 4176 & 7776, 17 Nov 1935. “Herb frequent, scattered; 18 fl. cream colored” (US; GH; K).— Arriba de Mejia, al sur de Mollendo, 180-200 m.s.m., Ferrevra 6416, 12 Nov 1949. “Hab. arenoso; hierba de fl. amarillas” (USM).—Mollendo, A. W. Hill 343, 1 Jan 1903 (K).— 5. Leptoglossis acutiloba (Johnst.) A. T. Hunz, et Subils (Pi 3. s-Pl. 3,C, EF) Hunziker et Subils in Lorentzia 3:17. 1979. Basionym: Sa/lpiglossis acutiloba Johnston in Contr. Gray Herb. 85: 179.1929. Saliglossis linearis Johnston in Contr. Gray Herb. 81: 96. 1928, not Hooker 1831. Type collection: Peru: Dept. Arequipa: Prov. Arequipa, Tiabaya, 2100-2200 m.s.m., Pennell 13063, 8 Apr 1925. “Annual herb; white sand dunes; corolla purplish (Holotype: GH; Isotype: S). Annual plants native to mountain ranges at altitudes slightly exceeding 2000 m, in southern Peru (Arequipa and Moquegua). ADDITIONAL MATERIAL. Peru. Dept. Arequipa: Prov. Arequipa: Tiabaya, ca. 2150 m.s.m., Pennell 13081, 6 Apr 1925. “Corolla tube near apex constricted, then abruptly inflated and decurved; tube pale orange yellow, streaked with or becoming wholly mulberry purple; lobes all spreading internally pale orange yellow, reticulate- lined with mulberry purple” (USM; US; S; GH; NY; K).— Tingo, 2100-2300 m.s.m., Pennell 13119, 8 Apr 1925. “Annual herb, corolla yellow and purple (detailed descr. with 13081) (NY: S: US: K; USM).—Jura (Bafios), 2500-2600 m alt., Vargas 7984, 29 Mar 1949(BM; US). — Dept. Moquegua: Prov. Moquegua: Hills SE of Moquegua, 1500-1600 m alt., Weberhbauer 7456, 22) 24-I11-1925. “Upper part of corolla greenish, brown veined, tube brown (S: US; BM; G: K). 6. Leptoglossis albiflora (Johnst.) A. T. Hunz. et Subils comb, nov. (Pl. 3, 2; Pl. 5, A, B, D, G.) Basionym: Salpiglossis albiflora Johnston in Contr. Gray Herb. 85: 178. 1929. TYPE COLLECTION: Peru: Dept. Moquegua: NW of Moquegua, Mt. Estuquina, 1600-1700 m.s.m., Weber- hauer 7424 a, 22 Mar 1925. “Corolla white with yellow center” (Holotype: F:; isotypes: K; NY; US; BM; G). Like L. /omana and L. acutiloba, sharing a small size (10 to 30 cm in height) and an annual growth, L. albiflora is readily identif- 19 iable by the absence of basal leaves, and, especially, by its unique 7-10-cleft zygomorphic calyx. The geographical distribution is comparatively wide: it has been collected (always at intermediate elevations: 1500 to 2000 m), in the Peruvian departments of Lima, Ayacucho and even as far south as Moquegua®, ADDITIONAL MATERIAL. PERU. Dept. Ayacucho: Entre Nazca y Puquio, 1500-1600 m alt.. Ferrevra 5465, 19 Mar 1969. “Hab. pedregoso, fl. blancas” (US; USM).—Dept. Lima: Prov. Lima, Cerros al N de Chosica, 1800-1900 m alt., Weberbauer, Apr 1923. “FI. blancas, veget. rala; xerofita con hierbas anuales, arbustos y Cactaceas” (USM).—Matucana, 8000 ft. alt., Machride & Featherstone 375, 12 Apr/ 3 May 1922 (NY).—Santa Eulalia, K. Maisch, 22 Aug 1923 (USM).— Mountains near Chosica (Lima-Oroya railroad), 1600-1700 m alt., Weberbauer $325, Apr 1910 (GH; US).— San Bartolomé (Lima-Oroya railroad), 1500-1600 m alt.. Weber- hauer 5297, Apr 1910 (GH; US).—Proyv. Cajatambo, Ambar, 2010 m.s.m.. Stork 11438, 16 Apr 1939. “Annual herb 0.2-0.3 m; calyx white with green or purple ridges: corolla white with reddish (GH).— Unknown Dept.: Huaytara, 6-8000 ft.. Pearce, May 1867. “Annual” (K). 7. Leptoglossis linifolia (Miers) Griseb. (Pl. 3, 7; Pl. 6.) Grisebach in Abh. KGnigl. Ges. Wiss. Géttingen, Phys. Cl. (1) 24: 241. 1879. (Based on Nierembergia linifolia Miers). Transfer incorrectly attributed to Bentham & Hooker by Jack- son (Index Kewensis 2: 63. 1895), and to Wettstein by Fries (Kungl. Sv. Vetenskapsak Handl. 46, 5: 38. 1911).—D’Arey in Ann. Missouri Bot. Gard. 65: 709. 1978.— Hunziker & Subils in Lorentzia 3: 13. 1979, Nrerembergia linifolia Miers in London Journ. Bot. 5: 174. 1846. “In Prov. Argentinis, v. v.”. LECTOTYPE: Pampas B.[uenos] Ayres, Canada de Lucas, Miers 835 (K). Nierembergia linifolia a macrophylla Dunal, in De Candolle Prodr. 13 (1): 587. 1852. “Ad Mendoza. . .”. [leg. Gillies]. Nierembergia linifolia B parviflora Dunal, in De Candolle Prodr. /. ¢. 1852. “In provincia Cordovae”, [leg. Gillies]. Crelostigma tenue Philippi in Anal. Univ. Chile 36: 197. 1870. "We have been unable to locate specimens of Weberhauer 3178; according to Macbride (1962: 143) it belongs to L. albiflora; should this identification be correct, the range of JL. albiflora would be extended northwards to the Dept. La Libertad. 20 Argentina: Mendoza, Wenceslao Diaz [year] 1868-1869 (Ho- lotype: SGO nr. 042803. Isotypes: W, and photogr. Field Mus. nr. 33004; G, and photogr. Field Mus. nr. 26798). Schwenckia tenuis (Phil.) Griseb. in Abh. K6nigl. Ges. Wiss. Got- tingen, Phys. Cl. (1) 19: 214. 1874. (Pl. Lorentz.: 166). Salpiglossis linifolia (Miers) Wettstein in Engl. & Prantl. Nat. Pflanzenfam. 4 (3b): 36. 1891.—Millan in Bol. Min. Agric. Nac. (Buenos Aires) 30 (1): 21, f. 7, t, u. 1931. Salpiglossis tenuis (Phil.) Wettst., /. c. 1891.—O. Kuntze, Rev. Gen. Pl. 3 (3): 224. 1898. Leptoglossis tenuis (Phil.) Jackson, Index Kewensis | (3): 63. 1894. Leptoglossis schwenckioides var. linifolia (Miers) Monachino in Lilloa 5: 435. 1940. Leptoglossis schwenckioides var. tenuis (Phil.) Monachino, /. c. 1940. Leptoglossis schwenckioides var. xerophytica Monachino, thid. 436. 1940. Argentina: San Luis: Alto Pencoso, Bruch & Car- ette, Feb 1914 (Holotype: NY). Endemic perennial species of the dry and low areas of central and western Argentina (usually ca. 300 to 600 or 700 m; in the West sometimes up to 1200 m); towards the East it is found up to 62° long. W, and in the North it has been collected at the “chaco” oriental region of Tucuman province. Owing to its gemmiferous roots, it grows frequently at disturbed places bordering the highways. Gametic chromosome number: n = 10 (Subils 1979: 19; mate- rials from La Rioja and San Luis). ADDITIONAL MATERIAL. ARGENTINA. WITHOUT COMPLETE DATA: W. 34° Pampas, O. Kuntze, 16 Jan 1892 (CORD, US: NY).— Pampas Reise, O. Kuntze, Jan 1892 (NY). Prov. Catamarca: Dept. Paclin: San Antonio, leg. O. Tarter, 22 Feb 1886 (CORD: Herb. Kurtz 4288). — Prov. Cordoba: Dept. Cruz del Eye: Arroyo de La Higuera, cerca de La Higuera, A. 7. Hunziker 9871, 18 Feb 1952. “Corola magenta (mas o menos violacea)” (CORD).— Pichana, Stuckert 14516, 20 Nov 1904 (G; CORD).— Dept. Ischilin: Quilino, Lorentz et Hieronymus 516, 8 Nov 1872 (CORD: G; US).— Quilino, cerca de la captacion de agua, Hosseus 683, 9 21 May 1942 (CORD). Alrededores de Villa Quilino, 4. T. Hunziker 9986, 26 Oct 1952. “Poco frecuente. Cara superior del limbo corolino granate lilacea; al envejecer la flor se aclara mucho, volviéndose apenas lilacea” (CORD). — Dept. Marcos Juarez: Barrancas del Rio Carcarafia: casi 30 KM al sur de Marcos Juarez, al cruzar el rio por la ruta que va a Inriville, A. 7. Hunziker 18960, 30 Nov 1966 (CORD). Dept. Minas: Cacapiche, Hieronymus 833, 19/20 Mar 1877(CORD).~— Entre La Higuera y Rumi Huasi, 4. 7. Hunziker 9155, 15 Apr 1951. “Rarisima. Limbo corolino violeta”(CORD). Dept. Rio Primero: Entre La Paloma y Las Saladas, Kurtz 4540, 6 Mar 1887(CORD)..— Inmediaciones de Rio Primero, a orillas del Rio Primero, 4. 7. Hunziker 10653, 4 Feb 1955. “Frecuente. Corola rosada” (CORD). Estancia San Teodoro, Sruckert 22136, Jan 1911 & 23038, Jan 1915 (CORD). Dept. Rio Seco: Margenes del Rio Dulce, al norte de Mar Chiquita, Savago 1721, Dec 1953 (CORD). — Dept. Rio Segundo: Villa del Rosario, leg. &. Piomti (CORD: Herb. Kurtz 14824). Dept. San Alberto: Altautina, Stuckert 10321 & 10322, 5 Dec 1901 (CORD; G). Dept. San Justo: Jeanmaire, inmediaciones de la estacién, Ariza & Astegiano 2488, 6 Dec 1970 (CORD). Ruta 19, Jeanmaire, Subils et al, 710, 4 Jan 1964 (CORD). — Miramar, Di Fulvio & Articd 144, 15 Nov 1969(CORD).— Orilla de Mar Chiquita, entre Miramar y la desembocadura del Plujunta, A. 7. Hunziker 13279 & 13295, 21 Jan 1957 “Poco frecuente, en manchones, en la vegetacion vecina a la orilla. Corola “roseus” (Chromat. Saccardo Nr. 17)” (CORD). Cerca de Jeanmaire, yendo desde La Francia, A. 7. Hunziker 11359, 26 Nov 1955. “Manchones algo frecuentes en cafiada salitrosa. Corola discolor: limbo por dentro purpureo-lilacino y por fuera blanquecino. A veces todo rosaceo y, mas raramente aun, todo blanco” (CORD.)—Camino entre J. Cortez y Altos de Chipion, Savago 1720, Dec 1953 (CORD). Mar Chiquita, leg. Bodenbender (CORD: Herb. Kurtz 9872). Sacanta, Stuckert 5993, Dec 1899 & 7078, Apr 1899(CORD). Dept. Totoral: Totoral (Villa General Mitre), leg. J. 4. Domin- guez, [year] 1901 (CORD: Herb. Aurtz 1/964). Prov. La Rioja: Dept. Gral. Belgrano: cerca de Olta, Roig & Mendez 8672, 12 Apr 1975. “A orillas del camino, fl. azul-violado” (CORD).— Dept. Gral. Ocampo: Ambil, en la ruta 79, entre Tello y Santa Rosa de Catuna, ca. 700 m.s.m.,.4. 7. Hunziker etal. 13870, 18 Feb 1959. “Manchon denso. Corola rojiza. Perenne con 6rganos subter- raneos” (CORD.— Ruta 79, al cruzar el Rio Saladillo, entre Ambil y Santa Rita de Catuna, A. 7. Hunziker etal. 14096, 4 Mar 1959. “Manchones abundantes a orilla del rio” (CORD). Dept. J. F. Quiroga: Malanzan, leg. Bodenhender, Dec 1895) Jan 1896 (CORD: Herb. Kurtz 8940). Dept. Gob. Gordillo: Sierra de Malanzan: La Aguadita, Chamical, Roig 8608, 10 Apr 1975 (CORD). La Aguadita (Ruta prov. 29), Biurrun & Lasso 170, 9 Apr 1976. “Flores azul- violaceas” (CORD). — Dept. Independencia: Cerro Morado prope Baldecito, leg. Bodenbender, Feb 1896 (CORD: Herb. Aurtz 90/8). Dept. R. Vera Penaloza: Salinas de Mascasin, Ruta 20, Km 1042. cerca de limite con San Juan, 510 m.s.m., A. T. Hunziker, Subils & Bernardello 23185, 9-1-1979. “Limbo blanco, tubo oscurecido por lineas purpureas; flores levemente perfumadas. Manchones con varios individuos; poco frecuente” (CORD).—Prov. La Pampa: Dpto. Chalileo: Santa Isabel, Troiani & Steibel 5554, 15 Feb 1978. "En el puente sobre el Rio Salado” (CORD).— Santa Isabel, Cano 3032, 7 Jan 1964 (US).— Prov. Mendoza: Unknown Dept.: Leg. Gillies s. n., sub nom. Nierem- hergia linifolia n. sp. Miers (K: Herb. Hook.). Leg. Gillies s. n., sub nom. Nicotiana linoides n.sp. Gill. (K: Herb. Hook).).—- Dept. La Paz: Desaguadero, 22 inmediaciones del Paso de Las Tropas. Ruiz Leal 8840, 16/19 Feb 1944, “FI, rosada hasta blanca!” (RL; LIL). — Entre Paso de las Tropas y Bajada del Gato, Ruiz Leal 8879, 29-11-1944 (RL: LIL). Dept. Las Heras: El Challao, Ruiz Leal 960, 29 Jan 1933 (RL: LIL). Dept. Lavalle: (Lagunas del Rosario). Las Tuni- tas, Ruiz Leal 14519, 9 Jan 1952 (RL). Puesto El Tapon, Roig 6892, 11 Mar 1970 (Herb. F. Roig). Dept. Maipt: Fray Luis Beltran, aprox. | km al norte del cruce de Ruta 20 por calle Las Margaritas, Ambrosetti & del Vitto s.n., 14 May 1977 (Herb. Ruiz Leal 28990).— Dept. Malargiie: Cordillera de Mendoza, Cordon de Santa Elena, leg. José Figueroa (CORD: Herb. Kurtz 15739). Dept. San Rafael: San Rafael, Puente Nuevo, a orillas del Rio Diamante, Suhils & Di Fulvio 43, 28 Jan 1959. “Flores lilacinas, hasta blancas” (CORD).— Rio Diamante, entre Fortin Nuevo y Monte Coman: bei Monte Tucuman. Kurtz 7056 & 7056 a, 14/16 Jan 1892. “FI. coerul., in nassem Flussand: hier un da” (CORD).—Punta del Agua (Nevado), Ruiz Leal 17469, 29 Dec 1955. “Fre- cuente en cerros aridos” (RL). Dept. Tupungato: Tupungato, en las huayque- rias del Carrizal, leg. £. Mendez (Herb. Roig 9540), 11 May 1975 (CORD). Prov. San Juan: Dept. Calingasta: Entre Villa Corral y Villa Nueva, Castellanos s.n., 27 Jan 1950 (US). Dept. Caucete: Ruta 20, Km 1043, cerca del limite con La Rioja, A. T. Hunziker, Subils & Bernardello 23356, “fl. blanca”: 23357 & 23360, “fl. amarilla”; 23358, “fl. cremosa™; 23359, “fl. lilacina™: 13 Jan 1979 (CORD).— Dept. Rivadavia: Marquesado, a orillas del Rio San Juan y cerros vecinos al dique, Ruiz Leal 16371, 29 Oct 1954(RL). Dept. Santa Lucia: Alto de Sierra, Hicken 85, 15 Dec 1907. “Suelo pedregoso” (SI). Dept. Valle Fertil: San Agustin del Valle Fertil, en las inmediac. del dique, 4. 7. Hunziker 16673, 12 Dec 1963, “Rara, corola lila*(CORD).— Sierra de Valle Fértil, Toma de Los Alvarez, a orillas del rio,aca. 8 Kmal SW de Valle Fertil, A. 7. Hunziker 16735, 12 Dec 1963. “Rara. Corola lilacina” (CORD).—Dept. Zonda: Pachaco, a orillas del Rio San Juan (Ruta 20), 4. 7. Hunziker 12996, 9 Nov 1956. “Limbo y lobulos al principio blancos, luego amarillentos; tubo por fuera con numerosas rayas long. oscuras” (CORD).--Unknown Dept.: Orillas del Rio San Juan, Hicken 78, | Nov 1907. “Poco difundida” (SI).—En las cercanias de la Estancia Experimental, Hosseus & Cerceau 2074, 3 Feb 1921. “Flor blanca” (CORD). Prov. San Luis: Dept. Ayacucho: Entre la Villa de Lujan y San Francisco, Galander s.n., 15 Mar 1882 (G; CORD).— Dept. Gral. Pedernera: Entre Cra- mer y Villa Reynolds, 485 m.s.m., Anderson 1500, 8 Jan 1969. “En terreno salitroso” (CORD). Dept. Junin: Santa Rosa, 650 m.s.m., Varela 472, 7 Feb 1944. “Flores lilas* (NY: LIL). Dept. La Capital: Rio Desaguadero, margen este, proximidades de la confluencia con el Rio Jarilla, 450 m.s.m.. 4. T. Hunziker, Subils & Bernardello 23417, “flor blanca”; 23479, “fl. amarillento- lilacina™, 23423, “fl, cremosa™;, 23424, “fl. lilacina”, 27 Jan 1979 (CORD). Salinas de Bebedero, alrededores del Establecimiento CIBA, 390 m.s.m.. 4. T. Hunziker, Subils & Bernardello 23449, “fl. lilacina”, 23454, “fl. purpurea”: 23465, “fl. blanca”, 27 Jan 1979 (CORD). Prov. Santiago del Estero: Dept. Aguirre: Pinto, leg. O. Kuntze, Oct 1892 (US: NY).- Dept. Atamisqui: Sobre ruta 9, atravesando Salinas Grandes, entre el Rio Saladillo e Isla Verde. ca. 200 m.s.m., A. 7. Hunziker & Subils 21285, 5 Jul 1971 (CORD). Dept. Banda: Dique Los Quiroga, Mever 12804, 19 Nov 1947 L; BR). Dept. Capital: Santi- ago del Estero, Mever 6887, 29 Oct 1944. “Fl. lila” (LIL). —Dept. Loreto: Ruta Nacional 9, entre Loreto y San Vicente, A. T. Hunziker, Subils & Bernardello 23160, 23 Dec 1978. “Flores purptireas” (CORD). Dept. Quebrachos: Camino 23 a Sumampa Viejo, Balegno 89, 25 Jan 1944. “FI. lila” (LIL). Sumampa, Ragonese 6328, 27 Oct 1946. “FL. rosadas™ (LIL). Dept. Rio Hondo: Rio Hondo. Cabrera 15588, | Mar 1963. “Flor lila’(CORD). — Las Termas, Bartlett 20469, 11 Jun 1943 (US: GH). Dept. Rivadavia: Unos 10 Km al oeste de La Isleta (al WSW de Ceres), 4. 7. Hunziker 10463, 11 Nov 1954. “Corola rosada: en hondonada salitrosa: frecuente” (CORD). Dept. Robles: Turena, Ma/ldo- nado 254, 6 Dec 1939 (LIL). Unknown Dept.: Rio Saladillo, ademas cerca del pueblo de Santiago, Lorentz 5, Princ. Dec 1871 (CORD). — Prov. Tucuman: Dept. Burruyacu: Sierra de Medina, 1200 m.s.m.. Venturi 2721, 1-1-1924. “Flores azules, en los prados” (LIL; US).--Dept. Cruz Alta: La Florida, Monetti 1218, (U; US). Dept. Graneros: La Madrid, leg. O. Tarter, 26 Jan 1886. “Im Schatten an sandigen Stellen, selten” (CORD: Herb. Aurtz 4235). Dept. Leales: Chaflar Pozo, Venturi 483, Oct 1919. “Flor violeta oscuro. A orillas de bosques ralos” (SI; US; GH). Chanar Pozo, 400 m.s.m., Fenturt 5439, 25 Oct 1927. “Flor morada. En terreno salado” (CORD: GH: US). EXCLUDED NAMES Leptoglossis Coulteri A. Gray in Proc. Am. Acad. 12: 165. 1877 = Hunzikera Coulteri (A. Gray) D’Arcy, Ann. Missouri Bot. Gard. 65: 705. 1978. Leptoglossis texana (Torrey) A. Gray in Proc. Am. Acad. 12: 164. 1877. = Hunzikera texana (Torrey) D’Arcy, Phytologia 34: 283. 1976. Leptoglossis viscosa (Torrey) Millan in Darwiniana 5: 489. 1941. = Hunzikera texana (Torrey) D’Arcy in Ann. Missouri Bot. Gard. 65: 706. 1978. ° x 7 Reyesia Gay Gay, Hist. Fis. Pol. Chile, Bot. 4: 418, lam. 52. 1849.— Bureau in Bull. Soc. Bot. France 10: 39-45. 1863. The author discusses the peculiar morphology of the reproductive struc- tures of Revesia. Bentham & Hooker, Gen. PI. 2 (2): 908. 1876. Hunziker in Kurtziana 10: 46. 1977.—D/’Arcy in Ann. Missouri Bot. Gard. 65: 710. 1978. Preroglossis Miers in Ann. Mag. Nat. Hist. Ser. 2, 5 (25): 32. January 1850. Type species: P. /axa Miers = Revesia chilensis Gay. It should be noted that the authority of Revesia has frequently been wrongly attributed to Clos. Starting with Bentham & Hooker (2.2908. 1876), Wettstein (1891:36) and the Index Kewensis (1: 700. 1895), this mistake was repeated by Reiche (1910: 398), Weder- mann (1928: 472), D’Arcy.(1978: 710, 712) and Hunziker (1979: 77, 82); the latter reference, however, has already been corrected (Hunziker, May 1977). 24 Type species: Revesia chilensis Gay. Four species in northern Chile (one of which also extending into adjoining Argentinian territory), usually at rather high alti- tudes in the Andes (2500-3500 m), except R. chilensis, which occurs at lower elevations (100-800 m) near the Pacific Coast. Although the plants are apparently similar in habit to those of Leptoglossis, Revesia is a more advanced genus, because in addi- tion to a number of differentiating characters mentioned in the key, it shows a more developed zygomorphic corolla and an androecium composed of four didynamous elements (these are the two posterior stamens with shorter filaments, and the two lateral ones with longer filaments; the anterior stamen is missing). Moreover, these lateral stamens may be either functional. as is the case In R. chilensis, or they may become sterile staminodes in more advanced species (for example: R. parviflora). Another peculiarity of R. chilensis is shown by the anthers of the shorter posterior pair of stamens: the thecae are unequal and divergent, whereas the ones of the longer pair are more parallel and equal- sized. Diagnostic features at the species level.—The multicellular glandular hairs of the calyx may be either oblong (R. chilensis and R. cactorum), or globose (R. parviflora and R. juniperioides). The stamens are usually glabrous, except in R. parviflora which has piliferous filaments; this latter character is correlated with the aggregation of pollen grains, which are united in tetrads in R. parviflora, and free in the other three species. The size of the corolla is another useful character. KEY TO THE SPECIES 1. Capitate hairs of the calyx with an oblong head. Filaments glabrous; pollen grains free. 2. Corolla 6-8 mm long.................cccceeeee 1. R. chilensis 2’. Corolla (10) 11-12 (15) mm long .............. 2. R. cactorum I’. Capitate hairs of the calyx with a globose head. Corolla 8-10 mm long. 2. Filaments piliferous; pollen grains in tetrads. Stems comparatively delicate, with long internodes ................. 3. R. parviflora 2’. Filaments glabrous; pollen grains free. Stems robust. rigid, intri- cately-branched, with short internodes......... 4. R. juniperoides 25 1. Reyesia chilensis Gay CPL. Ty 2) Gay, Hist. Fis. Pol. Chile, Bot. 4: 418, lam. 52. 1849. TYPE: Chile, Prov. Coquimbo, in siccis pr. Copiapo, Gay s. n. [ann.] 1838 (Holotype: P; isotype: K).— Hunziker, Kurtziana 10: 46. 1977.—D’Arcy in Ann. Missouri Bot. Gard. 65: 712, f. 1 D. 1978. Pteroglossis laxa Miers in Ann. Mag. Nat. Hist. Ser. 2,5 (25): 33. January 1850. HOLOTYPE: Chile, Coquimbo, Bridges 1389 in herb. Hooker (K); the collection number is not /839, as was wrongly transcribed by Miers (1. c.). The calyx has glandular hairs with oblong heads; furthermore, the plate subsequently published by Miers (Ill. South Amer. PI. 2: 32, tab. 52. 1849- 1857) shows that the filaments are glabrous. It follows, then, that this name isasynonym of R. chilensis, and not of R. parvi- flora as was believed by D’Arcy (1978: 713). Salpiglossis chilensis (Gay) Wettstein in Engl. u. Prantl, Natur. Pflanzenfam. 4 (3 b): 36, f. 16 H-J. [891]. Salpiglossis brachysiphon Johnston in Contr. Gray Herb. 85: 161. 1929. HOLOTYPE: Chile: Prov. Antofagasta, steep hill- side ca. 6 Km N of port [Tocopilla] and about opposite Caleta Duendes, Johnston 3625, 18 Oct 1925. “A brittle glandular plant; flowers dilute blue (GH). Revesia laxa (Miers) D'Arcy, in Ann. Missouri Bot. Gard. 65: 713. 1978, Apparently a perennial species (or biennial?); each plant has several stems without a basal rosette of leaves, and the roots attain 4 mm in diameter. The lower leaves have comparatively large blades (almost 6 cm long and 2.2 cm broad) pinnately cleft or almost pinnatisect, and conspicuous petioles (1.2-4 cm long); but at the upper nodes the petioles are shorter and even disappear, and the blades become smaller until the uppermost ones look like filiform, linear scales 2-10 mm long. 26 ADDITIONAL MATERIAL. CuiLe. Prov. Antofagasta: Dept. Taltal: Taltal, ca. 100 m alt., Werdermann 821, Oct 1925 (CORD; G; GH: US; U:; K; BM).— Hills SE of Taltal, Johnston 5084, 25 Nov 1925. “Rock crevices in dry quebrada” (US; GH). — Western end of Llano Colorado. ca. lat. 25° 29’, ca. long. 70° 32’, Johnston 5650, 13 Dec 1925 (S; GH).—Ca. 5KmSE of Taltal, 200 malt., Morrison 17095, 14 Jan 1939. “Annual herb 0,2 m, fl. white, tiny on filiform stems” (K; GH; G; S).—Ca. 10 Km E of Taltal. Worth er al. 15823, 13 Oct 1938. “Perennial herb to I dm alt.; fl. pale bluish: stems slender; leaves pinnatifid; infrequent” (GH). —Cerros de Toco- pilla, Barros 6543, 22 Sept 1940 (US).—Tocopilla, Barros 5575, 9 Jan 1941 (US).—Tocopilla, Jaffuel 1003, 27 Oct 1930 (GH).—Prov. Atacama: Dept. Vallenar: Vallenar, Alto del Carmen, 800 m alt., Werdermann 156, Nov 1923 (BM: G: GH; K; U; US).—Dept. Huasco, ca. 3 Km SW of Huasco, ca 35 malt., Worth et al. 16243, 27 Oct 1938. “Perennial herb 0,1 m alt.; fl. bluish, white outside; calyx glandular-pubescent; stem slender” (GH; K).—Huasco, C. Roberts, Febr. 1931. “Sea level”. (K).— Prov. Coquimbo: Dept. La Serena: Tres Cruces (Est. Junta de Chingoles, 29° 22’ lat.), Mufioz 325, 15 Sept 1935 (GH).- Laguna de Elqui, Barros 5619, Jan 1945 (US).—Atacama, R. A. Philippi (CORD). 2. Reyesia cactorum (Johnston) D’Arcy (Pl. 7, 4.) D’Arcy in Ann. Missouri Bot. Gard. 65: 712. 1978. Salpiglossis cactorum Johnston in Contr. Gray Herb. 85: 114. 1929. HOLOTYPE: Chile: Dept. Taltal, rocky hillside near Aguada del Cardon, Johnston 5258, 30 Nov 1925 (GH). A very rare species, sympatric with, and closely related to R. chilensis. It is known froma single collection, and it differs from R. chilensts in having longer corollas. More material is needed in order to determine the status of this taxonomic entity. 3. Reyesia parviflora (Phil.) A. T. Hunz. (Pl. 7, 32 PL3.) Hunziker in Kurtziana 10: 46. 1977 Salpiglossis parviflora R. A. Philippi, Viaje al Desierto de Atacama pag. 219. 1860. HOLOTYPE: Chile: Desertum Ata- cama vallis, Sandon, Febr 1854 (SGO 055827; also a photo- graph at GH). ISOTYPE: W. Apparently an erect annual plant (12) 16 to 40 (60) cm tall, often with only one main stem, which is more or less branched; zi usually it has a basal rosette of broad leaves with long petioles (petioles 10-20 mm long, by exception only 5 mm, or even 44 mm). As a sharp contrast, the linear, cauline leaves are inconspic- uous, sessile, squamiform; their length vary between | and 3mm, although in some cases they may reach 12 mminlengthand 8 mm in breadth. ADDITIONAL MATERIAL. ARGENTINA. Prov. Mendoza: Dept. San Carlos: Puesto Rio Colorado. Herb. Kurtz (CORD).— Prov. San Juan: Dept. Calingasta: En las colinas entre Villa Nueva y El Leoncito, Castellanos, 28 Jan 1950 (LIL 15728).—Precordillera entre Barreales, Tontal y Retamito: Cuesta de La Cortadera, Kurtz 9772 (leg. Bodenbender), 22 Febr 1897. “In coll. glareos.” (CORD: S).—-Camino a Paso del Espinacito: Pampa Negra, Castellanos, 10 Jan 1953 (LIL 35324: US). Cordillera del Espinacito: inter Ciénega Redonda et Quebrada Colorada. Kurtz 9564 (leg. Bodenbender), 5 Febr 1897 (CORD: S). Dept. Iglesia: Rio de la Tagua, below its confluence with Rio de la Sal (ca. lat. 29° 2’ 5” S, long. 69° 28’ 42” W). ca. 2900 malt., Johnston 6146, 12/13 Jan 1926. “Erect plant growing in gravel; corolla yellow” (K; GH; S).— Vicinity of Bafios de San Crispin (ca. lat. 29" II’ S, long. 69° 44’ W), ca. 3300 m alt., Johnston 6115, 10/12 Jan 1926. “Erect; corolla yellow” (GH; S; LIL).— Rio de la Sal, 3060 malt... /. H. Hunziker et al. 4874, 20 Mar 1951. “Corola amarilla. Ladera de grava” (CORD). CHILE. Guanaqueros, R. A. Philippi, Jan 1885 (GH).—Prov. Atacama: Geisse, 1889-1890 (GH). Dept. Vallenar, vicinity of Laguna Grande, ca. 3500 malt., Johnston 5957, 5/6 Jan 1926. “Erect plant; corolla yellow” (GH). Dept. Chanaral, vic. of Potrerillos, ca. 2800 m alt., Johnston 4743, 26 Oct 1925. “Corolla yellow” (GH; K).— Quebrada de Dofia Inés Chica. Gigoux, Jan 1886 (GH).—Des. Atacama, R. A. Philippi(BM; W).— Prov. Coquimbo: Coquimbo, R. A. Philippi (CORD; K; Fotogr. ex B, Field Mus. 3066: GH: CORD). Dept. Elqui: Banos del Toro, Reed (K).— Bafios del Toro, ca. 3500 m alt., Werder- mann 195, Dec 1923 (BM; GH; U: G).—Bafios del Toro, ca. 3300 m alt.. Morrison 17266, 5 Feb 1939. “Annual herb 0,4-0.5 m: fl. yellow” (G; GH: K).—Ca. 95 Km along the road from Rivadavia to La Laguna Dam, 2700) 3000 m alt.. Morrison et al 17089, 6 Jan 1939. “Rocky slopes along the roadside; annual herb 0,35 m; fl. yellow; very rare” (GH).— Ca. 98 Km from Rivadavia. ca. 2900 m alt.. Worth et al. 16399, 5 Nov 1938. “Annual herb 0,12 malt.: fls. lined with brown inside” (GH).—-Fundo Rio Seco, 14 Km east of Nueva Elqui, near the river, 3200 m alt., Wagenknecht 18119, 16 Dec 1940. (S). Distribution: Always at high elevations in the Andes (2900 to 3500 m). This is the only species in the genus that occurs also in Argentinian territory (Prov. Mendoza and San Juan): in Chile it is confined to the northern provinces of Atacama and Coquimbo. Obs.— This species was wrongly referred by D’Arcy (1978: 713) to R. laxa, undoubtedly due to his lack of access to the holotype 28 of Preroglossis laxa. As was already mentioned, P. /axa is a synonym of R. chilensis, a view already offered by Bentham & Hooker (1876: 909). 4. Reyesia juniperoides (Werdem.) D’Arcy cs a a | D’Arcy in Ann. Missouri Bot. Gard. 65: 712. 1978. Salpiglossis juniperoides Werdermann in Notizbl. Bot. Gart. Mus. Berlin-Dahlem 10 (95): 474. 1928. ISOTYPES: Chile: Prov. Tarapaca: Dept. Tarapaca, Parca, Cordillera de Qui- pisca, ca. 2500 m alt., Werdermann 1054, Mar 1926. A perennial, erect species, intricately branched, attaining 0.8 m in height (fide Werdermann |. c.); the stems are also aphyllous with their terminal branches spiniform. ADDITIONAL MATERIAL. CHILE. Prov. Tarapaca: Dept. Tarapaca: Iquique, Caritaya (Alto de Camina), Barros 6549, 16 Jan 1941 (US).—Quebrada a 43 Km de Azapa, camino a Chapiquifia, Ricardi et Marticorena 25509, 24 Sept 1958 (CORD).— Maminha, A. Pfister 9467, 12 Jan 1950 (CORD). SUMMARY AND CONCLUSIONS In this paper a taxonomic revision of Salpiglossis R. et P. (2 spp.), Leptoglossis Benth. (7 spp.), and Revesia (4 spp.) (three solanaceous genera of the tribe Sa/piglossideae) is presented. The long standing disagreement of the systematic position of the tribe is discussed and a key is offered to distinguish these three and related genera, as well as keys for the species within genera. Morphological and phytogeographic observations are included at both the generic and specific levels, supplemented by illustra- tions and maps. The more important results are as follows: 1. Salpiglossis cannot be united with Bouchetia Dunal in DC (the authors will discuss this problem in more details in a forth- coming paper). The genus consists of two species: S. sinuata R. et P. (Chile and Argentina), and S. spinescens Clos (Chile). 2. Leptoglossis is a disjunct genus of xerophytes with six Peruvian species [L. schwenckioides Benth., L. Ferrevrae Hunz. et Subils sp. nov., L. Darcvana Hunz. et Subils, L. /omana (Diels) 29 A. T. Hunz., L. acutiloba (Johnst.) A. T. Hunz. et Subils. and J. albiflora (Johnst.) A. T. Hunz. et Subils comb. nov.], and one Argentinian [/. /inifolia (Miers) Griseb.]. All the species have separate pollen grains. 3. Revesia has four species also of xerophytic habitats: R. parviflora (Phil.) A. T. Hunz. in Argentina and Chile, and the three others in northern Chile [R. chilensis Gay, R. cactorum (Johnst.) D’Arcy, and R. juniperoides (Werd.) D’Arcy]. Revesia parviflora is reinstated as the correct name for the first species, because the type collection of Preroglossis laxa Miers belongs to R. chilensis. The only species with pollen grains in tetrads is R. parviflora. BIBLIOGRAPHY Baillon, H. E. 1888. Histoire des plantes 9: 281-359, t. 343-461. Paris. Bartling, F. T. 1830. Ordines Naturales Plantarum eorumque Characteres et Affinitates adjecta Generum Enumeratione. 1-1V, 1-498. Gottingae. Bentham, G. 1835. Revision of Scrophulariaceae: trib. Salpiglossideae. In Bot. Reg. 21: Sub t. 1770. London. . 1845. The botany of the vovage of H.M.S. Sulphur. 1-194, 1-60 [ pag. 143: Leptoglossis gen. nov.] London. . 1846. Scrophulariaceae. Subord. et Trib. 1. Salpiglossideae. In De Candolle (Ed.), Prodr. Syst. Nat. 10: 190-203. Bentham, G. et J. D. Hooker. 1876. Genera Plantarum 2 (2): 882-980, 1244- 1245. Brongniart, A. 1843. Enumeration des genres de plantes cultivés au Muséum @'Histoire Naturelle de Paris... . 1-136 (2e ed. 1850, 1-237). Paris. Clos, D. 1849. Escrofularineas. In Gay, Hist. Fis. Pol. Chile. Bot. 5: 100- 188. Paris. Dale, FE. FE. 1933. Multiple allelomorphs in Salpiglossis. Amer. Nat. 67: 71. . 1937. A series of multiple alleles especially affecting the corolla in Salpi- glossis. Amer. Journ. Bot. 24 (10): 651-656, f. 1-14. Dale, E. E. & O. L. Rees-Leonard. 1939. Plastid variegation and concurrent anthocyanin variegation in Sa/piglossis. Genetics 24: 356-367. D'Arcy, W. G. 1978. A preliminary synopsis of Salpiglossis and other Cestreae (Solanaceae). Ann. Missouri Bot. Gard. 65: 698-724. f. 1-5. Diers, L. 1961. Der Anteil an Polyploiden in den Vegetationsgtrteln der West- kordillere Perus. Zeitschr. Bot. 49 (5): 437-488. Don, G. 1837. A general system of gardening and botany 4 [Solanaceae: 397 -488, f. 40-44]. London. Eichler, A. W. 1969. Einige Bemerkungen tiber den Bau der Cruciferen bliithe und das Dédoublement. In Flora 52: 102-109 [Solanaceae: 104-108], Tab. I. Eichler, A. W. 1875. Bliithendiagramme 1: 199-206. f. 117-118. W. Engelmann, Leipzig. 30 Endlicher, S. 1839. Genera Plantarum secundum Ordines Naturales disposita. I-LX, 1-1483. Part 9 [Solanaceae et Scrophularinae: 662-696]. Vindo- bonae. Fedde, F. 1896. Beitrage zur vergleichenden Anatomie der Solanaceae. 1-48. f. 1-6. Arbeit aus dem Bot. Gart. Univ. Breslau. Feuillée, L. 1714. Journal des observations physiques, mathematiques et bo- taniques. . . . 2: 504-767, pl. 1-50. Paris. Gay. C. 1849. Bignoniaceas. In Gay, Hist. Fis. Pol. Chile, Bot. 4: 408-420, lam. 52 [pag. 418: Reyesia]. Paris. Gray, A. 1877. Characters of some little-known or new genera of plants. In Proc. Amer. Acad. Arts & Sc. 12: 159-165 [Leptoglossis: 164-165]. Hassall, A. H. 1841-1842. Observations on the structure of the pollen granule. . . etc. In Ann. & Mag. Nat. Hist. 8 (49): 93-108 (Oct 1841); 9 (Suppl.): 544- 573, p. 13-18 (Sept 1842). Hunziker, A. T. 1977. Estudios sobre Solanceae. VIII. Novedades varias sobre tribus, generos, secciones y especies de Sud America. In Kurtziana 10: 7-50, f. 1-12 (Mayo 1977). . 1979. South American Solanaceae: a synoptic survey. In Hawkes er al. The Biology and Taxonomy of the Solanaceae. Linn. Soc. Symposium Ser. 7: 49-85, f. I-11. London. Hooker, W. J. 1827. Salpiglossis straminea. Exot. Fl. 3, t. 299. 1827. . 1834. Salpiglossis straminea var. picta. In Curtis’ Bot. Mag. 71: 3365. Hemsley, W. B. 1882. In Biol. Centr. Am. (ed. F. D. Godman & O. Salvin) Botany 2: 404-438. London. Johnston, I. M. 1928. Some undescribed American Spermatophytes. In Conrr. Grav Herb. 81: 85-98. . 1929. Papers on the Flora of Northern Chile. In Contr. Grav Herb. 85: 1-171. . 1929. Some undescribed species from Peru. In Contr. Grav Herb. 85: 172-180. Lindley, J. 1846. The Vegetable Kingdom. I-LX VIII, 1-908, f. 1-526. [Solan- aceae: 618-622, f. 417]. London. MacBride, J. F. 1962. Solanaceae. Flora of Peru. In Field Mus. Nat. Hist. 13. 5-B, 1: 1-267. Chicago. Miers, J. 1849. Observations upon several genera hitherto placed in Solanaceae, and upon others intermediate between that family and the Scrophular- iaceae. In Ann. Mag. Nat. Hist. 2nd. Ser. 3 (15): 161-183. Millan, A. R. 1931. Solanaceas argentinas. Clave para la determinacién de los generos. In Bol. Ministerio Agr. Nacidn 30 (1): 3-21, f. 1-7. Buenos Aires. . 1941. Revision de las especies del genero Nierembergia (Solanaceae) In Darwiniana 5:487-547. Munoz Pizzaro, C. 1966. Flores silvestres de Chile. 1-245, p. 1-51 (Laminas de E. Sierra Rafols). Ediciones de la Universidad de Chile. Remy, J. 1849. Solaneas. In Gay, Hist. Fis. Pol. Chile, Bot. 5: 38-100. Paris. San- tlago. Robyns, W. 1931. L’organisation florale des Solanacées zygomorphes. In Mémi. Acad. Roy. Belgique Cl. Sc. 11 (8): 1-84, pl. I-VI. Bruxelles. Ruiz, H. et J. Pavon. 1794. Flora peruvianae et Chilensis prodromus. 1-X XII, 1-154, tab. 1-37. Reprint 1965. J. Cramer. Lehre. Schlepegrell, G. von. 1892. Beitrége zur vergleichenden Anatomie der Tubi- 3] floren. In Bot. Centralbl. 49 (7): 193-200; 49 (8): 225-231; 49 (9): 257-263; 49 (10-11): 289-299; 49 (12): 353-360; 49 (13): 385-393, Taf. 1-4; 50 (1): 1-7: 50 (2): 33-39. Subils, R. 1979. Cromosomas gaméticos de Leptoglossis linifolia (Solanaceae) In Lorentzia 3: 19-20, f. I. Sweet, R. 1827. Salpiglossis straminea. In Brit. Fl. Gard. 3: pl. 231. _ 1828. Salpiglossis picta. In Brit. Fl. Gard. 3: pl. 258. _ «1830. Hortus Britannicus. 2nd. ed. London. _ 1833. Salpiglossis barclayana. \n. Brit. Fl. Gard. Ser. 2, 2, pl. 112. Torrey, J. 1859. General Botany of the Boundary. In Emory, Reporton the U.S. and Mexican Boundary Survey 2 (1) [Solanaceae: 151-156]. Washington. Vesque, J. 1885. Caractéres des principales familles gamopéetales tirés de l’ana- tomie de la feuille. Ann. Sc. Nat. 7e. Sér. 1: 183-360, pl. 9-15. Vilmorin R. de et M. Simonet. 1928. Recherches sur les nombres des chromo- somes chez les Solanées. In Zeitsch. indukt. Abstamm. u. Vererbungs- lehre Suppl. 2: 1520-1536. Werdermann, E. 1928. Ubersicht iiber die in Chile vorkommenden Arten der Gattung Salpiglossis Ruiz et Pav. (Solanaceae). In Notizhl. Bot. Gart. u. Mus. Berlin-Dahlem 10 (95): 472-475. Wettstein. R. v. 1891. Solanaceae. In Engler u. Prantl, Nattirl. Pflanzenfam. 4 (3b): 4-38, f. 12-16. Wettstein. R. v. 1892. Ueber die Systematik der Solanaceae. In Verhandl. d. kais.-kénigl. zool.-bot. Gessell. in Wien 42: 29-33. Wodehouse, R. P. 1929. The origin of symmetry patterns of pollen grains. In Bull. Torrey Bot. Club 56: 342-350, pl 16. Wydler, H. 1866. Morphologische Mitteilungen. In Flora 49: 513-525, f. 1-8. Museo BoTANICO (UNIV. NACIONAL CORDOBA) CASILLA DE Correo 495 5000 CORDOBA, ARGENTINA 32 PLATE 1 1. S. spinescens 2. S. sinuata = \ ESCALA 100 200 300 400 © 6©$00 600 MILLAS . a . a 400 600 800 KILOMETROS h gat Po ; \ 3 ; \ : | | ‘vane (. i \ Oy SK j Plate 1. Geographical distribution of the species of Sa/lpiglossis. 33 Plate 2. Salpiglossis sinuata. (A, B, C, from Joseph 808; others from Bauer Nov 1855). A: dehiscing fruit, X 2.6; B: seed (lateral view), X 26.4; C: seed cross section, X 26.4; D: anther (ventral view), 8.8; E: glandular hair of the calyx, X 132: F: lateral internal view of a theca to show the dorsal insertion of the filament (dorsal side to the left), X 8.8; G: upper part of a floriferous branch, X 0.88; H: pollen tetrad, X 320; I: flower bud, * 1.8; J: longitundinal section of seed, x 26.4; K: portion of calyx showing venation, X x 4.4; L: basal portion of a branch, < 0.88; M: anther (dorsal view), X 8.8: N: ovary cross-section, * 13.2; O: expanded corolla (internal view) showing four didynamous stamens and the anterior staminode, X 1.8; P: gynoecium, 2.6. 34 PLATE 2 35 PLATE 3 Plate ND WRF WHE 40}-— . albiflora . ferreyrae + acutiloba L L L L. lomana L L. darcyana L .» linifolia + schwenckioides —_ OS eee ESCALA 100 200 300 400 500 600 MILLAS ° 200 400 600 800 MILOMETROS 3. Geographical distribution of the species of Leproglossis. 36 PLATE 4 Plate 4. Leptoglossis Ferreyraei(Ferreryra 14020). A: ovary cross section, X 18; B. H. J: branched hairs of the calyx, X 440; C: basal leaf, X 5.3; D: habit, x 0.88; E: gynoecium, X 13.2; F: anther (ventral view), X 18: G: flower, 3.5; I: flower bud, X 7; K: expanded corolla (internal view) showing two posterior stamens and three staminodes, X 7. 37 Plate 5. Leptoglossis albiflora (Weberbauer 7424 a). A: expanded corolla (internal view) displaying three staminodes and the two posterior stamens ata very low level of adnation, X 14; B: zygomorphic calyx with eight dissimilar lobes (note its peculiar pattern of venation), X 4.4; D: hair of the calyx, X 220; G: stigma situated at the peculiar flat wings of the style (see also figs. C, H. Jand Q), 18. —L. acutiloba (Pennell 13081). C: stigma, X 18: E, F: curved hairs of the calyx, X 220.——L. lomana (Scolnik 1020). H: stigma X 18:1: branched hair of the calyx, X 220.—L. schwenckioides (Hutchison 1032). J: stigma, X 18: K: expanded corolla (internal view) of a flower with two pairs of fertile stamens (each pair with unequal anthers) and the anterior staminode, X 7; L, O: calycinal hairs, X 220; N: calyx, X 4.4—L. linifolia (Di Fulvio 144). M: expanded corolla of a flower bud, X 7; P: hair of the calyx, X 220; Q: stigma, X 18, 38 PLATE 5 MI ay oy Oe svonnyaye pny yey eennuyan ayy Rye ar 39 PLATE 6 Plate 6. Leptoglossis linifolia (fruit and seed: A. T. Hunziker 11359; others from DiFulvio 144). A: glandular hair of the coroHa (outer surface), * 125; B: flowering branch, X 1; C: flower (top view), * 2; D: seed (hilar view), X 25; E: stamen (ventral face), X 10; F, G, N, S: embryos, X 25; H-J: seeds cross sections (the small triangle at the base indicating the position of the hilum, 25; K, O: the two halves of the upper part of the corolla, one with three staminodes, the other with the two posterior stamens, X 4; L: habit (note the sprouting roots), X 0.25; M: ovary and disk, X 7.5; P: cross section of ovary, X 15; Q: gynoecium X 2.5; R: fruit, X 2.5; T: upper part of the flower (vertical section) showing the arrange- ment of the stamens and staminodes in relation to the stigma, X 4; U: longitudi- nal part of the calyx showing venation, X 3.75. 40 PLATE 7 iP a.* Fer ° os — “eeeet? solaee ie ! i: z ' “oa oa) 1. R. juniperoides ‘Ce 2. R. chilensis 3. R. parviflora 4. R. cactorum 40}— “+ 500 600 MILLAS @00 KILOMETROS | | Plate 7. Geographical distribution of the species of Revesia. 4] Plate 8. Revesia parviflora (K, l: LIL 35324; the others from J. H. Hunziker 4874). A: habit, X 0.88; B: pollen tetrad, ¥ 352; C: calyx hair, X 132; D: flower bud. X 7: E. H: anther (ventral and basal views), X 18: F: fruit, X 3.9: G: gynoecium, X 13.2; I: flower, x 10.5: J. L. M: lateral view, cross section and longitudinal section of seed, X 26.4; K: expanded corolla (internal view), * 10.5. 42 PLATE 8 43 BOTANICAL MUSEUM LEAFLETS VOL. 27, No. 1-2 JANUARY-FEBRUARY 1979, THE IDENTITY OF AMAZONIAN AND TRUJILLO COCA TIMOTHY PLOWMAN! In the course of preparing a taxonomic treatment of the culti- vated species of Erythroxylum?, it has become necessary to make some nomenclatural additions and changes. The cultivated coca plants of South America are generally considered to belong to two species of Erythroxylum: E. Coca Lam., which includes so-called Bolivian or Huanuco coca, and E. novogranatense (Morris) Hieron., which is Colombian coca (Plowman, 1979a). Both of these species have one distinct culti- vated variety. Since they were first recognized, these varieties have been the source of considerable taxonomic confusion and misinterpretation in both the botanical and pharmaceutical literature. The coca which is typical of the Amazon basin, known as Amazonian coca, is a variety of Erythroxylum Coca. Trujillo coca, which is grown primarily on the north coast of Peru, isa variety of E. novogranatense. Both of these varieties will be discussed in greater detail in separate papers now in preparation. My intention here is to identify these cultivated taxa with correct scientific names, descriptions and with the designation of types. In view of the current broad interest in coca and its derivatives, it is essential to stabilize the names of the plants without further delay. AMAZONIAN COCA Erythroxylum Coca is the most important commercial species of coca, furnishing by far the greatest portion of the world’s ' Assistant Curator, Botany Department, Field Museum of Natural History, Chicago, Illinois 60605, and Research Associate, Botanical Museum of Harvard University, Cambridge, Massachusetts 02138. 2Botanists have not been consistent in the spelling of the generic name of coca. For a discussion of the correct orthography of Erythroxylum, see Plowman (1976). 45 supply of coca leaves and cocaine. This species is native to the montana region of the eastern Andes, extending from Ecuador south to Bolivia. It is cultivated mostly between 500 and 1500 m. elevation in an area characterized by a favorable tropical envir- onment with high rainfall, moderate temperatures, and well drained, mineral-rich soils. A distinct variety of Erythroxylum Coca is also found in South America but has been largely ignored by both botanists and anthropologists. This is the coca of the Amazon valley, which continues to be cultivated today on a small scale by a number of Indian tribes in the western Amazon of Brazil, Colombia and Peru. The history, cultivation, method of use and alkaloid chemistry of Amazonian coca was discussed recently in a paper read before the 43rd International Congress of America- nists (Plowman, 1979b). Amazonian coca is readily distinguished from typical Ery- throxylum Coca (E. Coca var. Coca) of the Andean foothills. The Amazonian variety grows typically as a tall, spindly shrub with long, weak branches and relatively large, elliptical leaves. The leave are usually blunt or rounded at the apex in contrast to montana coca which usually has more or less pointed leaves. The parallel longitudinal lines found on the leaf undersides and usually considered characteristic of E. Coca are often faint or even lacking in the Amazonian variety. Furthermore, the flow- ers of Amazonian coca have a shorter, thicker pedicel and a markedly denticulate staminal tube. Erythroxylum Coca is a self-incompatible, distylous species with both long-styled (pin) and short-styled (thrum) morphs, which occur in approximately equal numbers (Plowman, unpub- lished data; Ganders, 1979). However, most populations of Amazonian coca contain only the short-styled morphs, with the exception of one or two collections from the easternmost Amazon in Brazil. The occurrence of only short-styled morphs in the Amazonian variety results from the fact that the plants are propagated vegetatively by cuttings, in contrast to Andean coca which is grown mainly from seeds. In Amazonia, entire planta- tions may be based on a single short-styled clone. Since only one stylar form is present in Amazonian coca, there is, under field conditions, no fertilization in this obligately 46 outcrossing species. As a result, Amazonian coca rarely if ever produces viable seed. Even though fruits may be formed, at times in abundance, the seeds are devoid of embryos and/or endosperm and will not germinate. Amazonian coca appears to cross readily with long-styled morphs of its Andean counterpart. Experimental reciprocal crosses in the greenhouse produced normal offspring. These plants have not yet reached flowering size, so their degree of fertility cannot be reported at this time. Amazonian coca is apparently known only in cultivation and is incapable of competing with the dense secondary vegetation which grows up after cultivated plots are abandoned. In this feature, it further differs from Andean E. Coca, which fre- quently escapes and forms a component of the forest understory in forested areas around coca plantations (Plowman, 1979a). Amazonian coca also differs in the manner of preparation of the leaves for chewing. The leaves are always toasted to dryness and pulverized in a mortar and pestle. The resulting powder is then mixed with the ashes of leaves of Cecropia or Pourouma species (Moraceae). Martius made the first collections of Amazonian coca during his sojourn in the Brazilian Amazon in 1819-1820. He described in detail the unique preparation and use of the plant (Spix & Martius, 1831) and later published a description and illustration of Amazonian coca under the name Erythroxylum Coca Lam., since he considered it to be the same as the Peruvian plant described by earlier authors (Martius, 1843). There has been only one other attempt to name Amazonian coca scientifically; and this was the result of an unfortunate error in identification. In the 1870’s, a variety of coca was intro- duced from Europe into cultivation on the island of Java. In 1890, this plant, which became known in the trade as “Java coca”, was described as a new variety by the Dutch botanist Burck working at the Botanical Gardens at Bogor (Buitenzorg). Burck named the plant Erythroxylum Coca var. Spruceanum commemorating Richard Spruce, the famous English botanist who made extensive plant collections in the Amazon region in the mid-19th century. In 1889, Morris a botanist at the Royal Botanic Gardens at 47 Kew, published an article? in which he described another new variety of coca, Erythroxylum Coca var. novogranatense, which is the basionym of EF. novogranatense. In his article, Morris also discussed the relationships of other cultivated coca plants and remarked that specimens of Java coca “corresponded exactly” to herbarium specimens of coca at Kew collected by Richard Spruce on the Rio Negro in 18514. On the basis of this state- ment, Burck injudiciously chose the name “Spruceanum” for Java coca. The resemblance of Java coca to Spruce’s original collection, which I have studied at Kew, is entirely superficial. Morris’s identification was based merely on the variable characters of leaf shape and the lack of the characteristic parallel lines on the underside of the leaves. Spruce’s collection at Kew is labeled “No. 73, ‘Ipadu’, sitio on the Rio Janauri (affluent of the Rio Negro)”, and clearly represents Amazonian coca as delimited here. This herbarium specimen also serves as a voucher for the first of two samples of coca powder which Spruce sent to the Kew Museum (Spruce, 1853a). The holotype of Erythroxylum Coca var. Spruceanum Burck is a specimen preserved at the Herbarium Bogoriense in Bogor, Indonesia. Isotypes are found in the Institute for Systematic Botany in Utrecht, and at the Rijksherbarium in Leiden. All of these specimens belong to Erythroxylum novogranatense, and the name E. Coca var. Spruceanum is placed in synonymy with it. This identification agrees with the treatment of Payens who revised Erythroxylum for the Flora Malesiana in 1958. Even though Burck indirectly cited the Spruce collection as a para- type, this specimen must be excluded from the type material of E. Coca var. Spruceanum, since it belongs to a different species and does not conform to the type. I would here like to name Amazonian coca as Erythroxylum coca var. ipadu, after the widely used Brazilian name of the plant, and offer the following description: ’Morris’ original article of 1889 was unsigned and was erroneously attributed by Burck and others to Dr. Thiselton Dyer, then Director at Kew. 4Although Morris reported the date of Spruce’s coca collections as 1854, Spruce began his studies and collections of coca as early as 1851, as reported two years later ina letter to Sir William Hooker (Spruce, 1853b). 48 Erythroxylum Coca var. Ipadu Plowman var. nov. Ab Erythroxylo Coca typico ramis fortiter erectis virgatis, ramentis paucis vel nullis, foliis apice rotundatis, lineis abaxiali- bus obscuris, pedicellis brevioribus, stamineo urceolo valde 10- denticulato, stigma promienti, ovato-oblonga differt. TYPE: PERU: Dept. Loreto: Prov. Maynas. Rio Ampiyacu, Puca Urquillo and vicinity, approx. lat. 3°05’, long. 71°55’. Cultivated shrub 2.5 m tall in open coca plantation in jungle clearing, interplanted with yuca and fruit trees. Branches slender, erect, then arching, leaves mostly near the tips, bright green. Flowers cream. N.v. jibina (Witoto). 5 Apr 1977. T. Plowman, R. E. Schultes & O. Tovar 6663 (holotype, ECON; isotypes, F 1824462, GH, K, MO, S, U, US, USM). Slender shrub to 3 m. tall. Single trunk to 2 cm. in diameter. Bark light greyish brown with fine longitudinal cracks. Branches strongly erect, spindly, sometimes arching over with age. Branch- lets straight, not knobby at the nodes, light yellowish green becoming reddish brown, with lenticular or elongated lenticels. Ramenta inconspicuous, sparse or sometimes absent, if present found only at the base of new shoots. Leaves present mostly at the branch tips, the blade elliptic to broadly elliptic or elliptic- ovate, apically obtuse to rounded, basally obtuse to acute, some- times briefly attenuate, 35-115 mm long, 20-42 mm wide, firmly membranaceous, dark to medium green above, pale glaucous green beneath, dull to somewhat shiny above, dull beneath, mid- rib yellowish green, with an acute adaxial ridge, lateral nerves often prominulous above, abaxial longitudinal lines obscure or wanting, central panel often faint or concolorous. Petiole 2—4 mm. long. Flowers in axils of past season’s growth, usually on naked stems but sometimes axillary in subterminal leaf axils, 1-5(6) per node, scattered or sometimes crowded, creamy white to yellowish throughout, odorless. Pedicel short, thickened dis- tally, 2-4 mm long, to 1.5 mm in diameter. Lamina of petals spreading, sometimes suberect and not fully opening, incurved at apex. Margin of staminal cup markedly 10-denticulate, the ‘The term “ramenta” was first used by Martius (1843) for the persistent, often congested. scale-like structures which frequently cover the twigs of Erythroxylum species. Mor- phologically, ramenta are simply stipules which are produced without accompanying leaves. 49 teeth short, triangular. Short-styled flowers predominant, style short 1.5-2 mm. long, stigma large, ovate-oblong, | mm long, 0.5 mm broad, bright yellowish green. Drupe 7-8 mm long, 4-6 mm in diameter, embryo and endosperm usually aborted. Chromosome number n = 12. ETYMOLOGY: This variety is named for the Brazilian vernacular name of Amazonian coca. COMMON NAMES: idapu, yvpadu (Amazonian Brazil), coca (Amazonian Peru), jibina (Witoto, Peru), (pi (Bora, Peru), patu (Cubeo, Colombia), botd (Maku, Brazil). DISTRIBUTION: Cultivated in the Amazon basin of Peru, Co- lumbia and Brazil, along the Amazon River and its major tributaries. SPECIMENS STUDIED: BRAZIL: AMAZONAS: Basin of Rio Ica, near San Antonio, “padu” or “coca”, Dec 1935, Krukoff 7645 (NY); Rio Purus basin, Rio Uneiuxi, Maku Indian village, 300 km above mouth, “bot6” (Maku), 23 Oct 1971, Prance et al. 15572 (ECON, INPA, K, NY, U, US); Rio Jurua, Marary, Sep 1900, Ule 5039 (K, L); Rio Solimées, Tefe, Lago de Tejé, Mujeira, “padu”, 13 Jul 1973, Lieras et al. P16681 (GH); Tefe, “padu”, “ipadu”, 16 Jul 1972, PLK & Urbana 12231 (INPA); Tefe, Fazenda Experimental, 20 Feb 1973, PLK & Marilene 12553 (INPA); “prope Ega et Sao Paulo d’Olivencga”, 1819-1820, Martius s.n. (M); “in sylvis Japurensibus”, “ypadu”, 1819, Martius S.N. (M); Ilha Parauary, 20 Oct 1874, Traill 77 (K); Rio Negro, Providentia, “ipadu”, Jul 1888, Prinzessin Therese von Bayern s.n. (M); Rio Negro, Cucui, 7.5.1973, Silva et al. 1310 (F); Rio Janauari, “paida”, Spruce 73 (K): Manaus, igarape do Buiao, “ipadu”, 5-7-1958, Coélho s.n. (F). PARA: Belem, “ipadu”, Sep-Oct 1961, Pires 51919 (NY, US). [Belem, Sao Jozé to the Arsenal, 1828-1830], Burchell 9588 (K), Belém, Botanical Garden, Museu Goeldi, 15 May 1908, Baker 70 (A, BO, GH, LE, MO, NY, US); Rio Guama, Apr 1899, E. Poisson s.n. (P); Santarem, May 1929, Dahlgren & Sella 36 (F). COLOMBIA: PUTUMAYO: Rio Putumayo, entre Puerto Asis y Puerto Legu- isamo, 300-400 m, 14-15 Oct 1954, Garcia-Barriga et al. 18713 (COL). AMAZONAS: Leticia, camino a Tarapaca, km. 17, 18 Jul 1965, Lozano & Ospina 488 (COL); Tikuna village, km. 12 on trail from Leticia to La Pedrera, 14 Apr 1975, Cabrera 3367(COL). Rio Igaraparana, La Chorrera, 180 m, 6 Jun 1942, Schultes 3899(ECON); Ro Igaraparana, “kudu jibina” (Witoto), 1 Mar 1974, Idrobo 6859 (COL, ECON); Rio Igaraparana, 15 km. abajo de La Chor- 50 rera, “jibina”, 12 Feb 1974, Idrobo 6775 (COL). VAUPES: Rio Piraparana, Raudal, 4 Sep 1952, Schultes & Cabrera 17169 (A, COL, ECON, F): Rio Piraparana, San Miguel, environs of Catholic mission, 23 Oct 1976, “canea patu”, “hokquhe”, 23 Oct 1976, Davis 119 (ECON, F), “naa k kwahe” (Bara- sana), Davis 120n (ECON, F); Rio Apoporis, Soratama, Raudal Jirijirimo, below mouth of Kananari, 900 m, 21 Jan 1952, Schultes & Cabrera 14979 (ECON): Rio Apoporis, Jino-Goje, entre los Rios Piraparana y Popeyaka, alt. 250 m, 3-11 Sep 1952, Garcia- Barriga 14453 (ECON); Rio Kanari, “pa-too” (Kabuyayi), 6-8-1951, Schultes & Cabrera 13415 (F); Rio Kananari, Cerro Isibukuri, 250-750 m, 13 Jun 1951, Schultes & Cabrera 12419 (AAH, F, GH). Rio Kubiyu, 2 hrs. upriver from confluence with Rio Vaupés, “hok+ patu” (Cubeo), 9 Apr 1975, Davis 10 (ECON, GH, K, MO, P, US), “karika patu” (Cubeo), Davis 11 (ECON, F, MO, US) “patu”, Davis 12 (F), “wehki patu”, Davis 13 (ECON, F, MO, NY, S, U); same locality, 12 Apr 1975, “karika patu”, Davis 18 (ECON, F, US); “hoké patu”, Davis 19 (ECON, F, K, NY, MO, US), “wehki patu”, Davis 20 (ECON, US); lower Rio Kubiyu, | Apr 1975, Zarucchi et al. 1145 (ECON, F). Bocas del Caruru, Casa Alvarez, 240 m, 25 Sep 1939, Cuatrecasas 7012 (F). PERU: LORETO: Prov. Maynas: Rio Napo, Negro Urco, | mile downriver, “jibe” (Witoto), 17 Aug 1966, Martin 1318 (ECON); Mishuyacu, near Iquitos, Apr 1930, Klug 1117 (F, NY, US); Iquitos, alt. 120 m, “coca”, 10 Oct 1929, Li. Williams 3551 (F); Morona Vieja, mouth of Quebrada Versalles, 11 km. N.O. de Iquitos, 17 Aug 1966, Torres 167 (ECON, F); road from Iquitos to Rio Nanay, 12 Jun 1966, Martin 1001 (ECON); Rio Nanay, road to Picuruyacu, near house of Jose Pina, 4 Aug 1966, Martin 1209 (ECON). Rio Ampiyacu, 24 Sep 1972, Croat 20864 (DUKE, F, K); Pebas, Rio Amazonas, 18 Apr 1977, Plowman et al. 6922 (ECON, F, K), Plowman et al. 6923 (ECON, F, MO), Plowman et al. 6924 (ECON, F, US); Brillo Nuevo, Rio Yaguasyacu, affluent of Rio Ampiyacu, “ipi” (Bora), 12 Apr 1977, Plowman et al. 6748 (ECON, F, K, NCU, USM), Plowman et al. 6750 (ECON, F, USM), “huangana coca”, Plowman et al. 6802 (ECON, F, GH, K, USM), “mojarra coca”, “ts+-paa” (Bora), | May 1977, Plowman et al. 7136 (ECON, F, K, NCU, USM), “pelejo coca”, “daa-llimti” (Bora), Plowman et al. 7137 (ECON, USM); Florida, Rio Putumayo, mouth of Rio Zubineta, alt. 200 m, Mar-Apr 1931, Klug 2002 (A, F, GH, MO, NY). TRUJILLO COCA Trujillo® coca is a variety of coca cultivated on the desert coast of Peru and in the adjacent arid valley of the Rio Maranon. This plant has been grown there for at least 3000 years and is ecologi- cally well adapted to its desert habitat, which both ancient and modern farmers watered with complex irrigation systems (Plow- 6The older spelling of the Spanish name Trujillo is “Truxillo” and is often encountered in the earlier literature. “Trujillo” is the correct form. 51 man, 1979a). Archeological and ethnohistorical evidence dem- onstrates that Trujillo coca was once cultivated in most if not all the coastal river valleys of Peru (Rostworowski, 1973). In the pharmaceutical trade, Trujillo coca has also been referred to as “small-leaved” or “Peruvian” coca to distinguish it from “Bolivian” or “Huanuco” coca (referring to the species Erythroxylum Coca). Even in pre-Conquest Peru, Trujillo coca was recognized as a distinct variety and was called “tupa” coca, meaning “royal” or “noble” coca, to distinguish it from “mamox” coca, which was the name applied to the “large leaved” coca grown on the eastern slopes of the Andes, i.e. E. Coca (Rostwo- rowski, 1973; Plowman, 1979a). Trujillo coca is still cultivated on a small scale, especially in the region around Trujillo on the north coast of Peru. It consti- tutes less than 5% of the total coca production in Peru but is the principal variety used in the beverage industry owing to its high content of essential oils and flavors. The identity of Trujillo coca became the object of much debate and confusion when the leaf first appeared in world phar- maceutical markets in the 1880’s. The controversy centered on the botanical identification of the sundry commercial varieties of coca and involved a number of prominent botanists and phar- macists of the day, including E. R. Squibb, E. M. Holmes, D. Morris, W. Burck and H. H. Rusby. The history of this contro- versy is long and complicated and resembles the early attempts to identify the botanical sources of quinine and curare. Much of the difficulty in identifying commercial coca leaves stemmed from the lack of experience of early workers with the plants in the field. Instead, they were attempting to describe and name samples of dried leaves encountered in commerce or iso- lated living plants of unknown provenience, cultivated in conser- vatories and experimental gardens. My purpose here is not to elaborate all the intriguing details in the history of the identification of coca. This will be treated in depth in a separate paper which is in preparation. However, it is necessary to describe the events which have obfuscated the iden- tification of Trujillo coca, in order to clarify its present taxo- nomic position and correct name. As early as 1889, Trujillo coca was associated with Erythroxy- 52 lum novogranatense (as E. Coca var. novogranatense) by Morris at Kew, who stated that his new variety approached “very nearly (although not so coriaceous) as what are known in commerce as Truxillo leaves”. In 1900, H. H. Rusby, then Professor at the New York College of Pharmacy, described Trujillo coca as a new species, Erythroxylum truxillense. He asserted that this coca differed from both E. Coca of Lamarck and from E. Coca var. novogranatense of Morris (which Rusby erroneously wrote as “neo-granatense”). Neither Rusby nor his English-speaking contemporaries were aware that Hieronymus, working in Berlin, had made the correct combination Erythroxylum novograna- tense in 1895 in identifying the Colombian coca collections of F. C. Lehmann. In his rambling and confused paper of 1900, Rusby discussed several different kinds of coca, both wild and cultivated, which he had encountered both in his extensive travels in South Amer- ica and in the pharmaceutical trade. Rusby departed from acceptable taxonomic procedures, even then in common prac- tice, of describing plants clearly and concisely and of citing spe- cific collections as types or otherwise authentic specimens. He described his E. truxillense casually and in prose discourse. His brief description of the “Trujillo” or “small green leaf which we get directly from Peru” was given as follows: “It is mostly from 3 xX 1-1/4 to 4 X 1-1/2 cm. It is obovate, with narrowed base, mostly acute or acutish at the apex and minutely apiculate. In commercial leaves the lateral lines are commonly faint or even wanting”. He also provided an illustration of this leaf in his Fig. 14. Rusby did not cite a type specimen of commercial Trujillo coca. However, his intention was to describe the Trujillo leaf found in the New York pharmaceutical trade, and a specimen of this material should be designated as a lectotype. No such speci- mens are preserved today at the New York Botanical Garden where the best set of Rusby’s herbarium is deposited. However, an appropriate specimen has been discovered among Rusby’s vast materia medica collection now housed at the Harvard Botanical Museum. This entire collection was transferred to Harvard in 1973 “on indefinite loan from the New York Botani- cal Garden.” 53 In 1979, Susan Marie Rossi began at the Botanical Museum the task of curating and re-cataloguing Rusby’s collections of medicinal plant products, contained largely in glass jars. She recovered, among other coca specimens, a jar labeled “No. 2684, Erythroxylon truxillense Rusby, Truxillo coca, small coca, native of Peru and cultivated. The commercial drug presented by E. Merck & Co., New York City.” This specimen corresponds to a listing in Rusby’s catalogue (1921) of plant products included in the now defunct Economic Museum of the New York Botanical Garden. The provenience of this specimen is confirmed by the occasional presence in the sample of leaves of pacay, Inga Feuillei D. C. This leguminous tree, a native of Peru, is commonly planted as a shade tree in plantations of Trujillo coca and its leaves often appear as a contaminant of commercial Trujillo leaf. This sample, identified by Rusby, is the most appropriate of Rusby’s collections to serve as the lectotype of his species EF. truxillense. In addition to his cursory description of Erythroxylum truxil- lense, Rusby created further confusion by describing a living plant which he found in flower at the conservatory of the New York Botanical Garden in August, 1900. He knew nothing of the origin of this plant but likened it to his E. truxillense, but not without some reservations. He provided a brief description of it and stated that this plant was “unquestionably the same thing which Dr. Burck speaks of as Java coca and for which he pro- poses the name EF. Coca Spruceanum.” Rusby made an herbarium specimen of this living plant and deposited it at the New York Botanical Garden herbarium labeled E. truxillense Rusby. The following year (1901), he pro- vided further descriptive details of the plant and published a line drawing of it (Rusby 1901, fig. 2). From both the specimen and the drawing, it is certain that this plant represents typical Colombian coca, E. novogranatense var. novogranatense. It is likely that the living plant which Rusby confused with Trujillo coca’ was derived from progeny of the “Kew Plant” originally named by Morris, in spite of Rusby’s contention that his plant did not match Morris’s £. Coca var. novogranatense. We now know that FE. novogranatense is at least partially self- compatible (Plowman, unpublished data; Ganders, 1979) and 54 spontaneously produces prodigious amounts of seed. Seeds of the original “Kew Plant” were sent out to conservatories and botanical gardens throughout the world, and it is likely that some of them found their way to the New York Botanical Garden. In short, the specimen which Rusby collected in cultivation must be excluded as a type of Erythroxylum truxillense. It was misconstrued by Rusby as being identical with the Trujillo leaves of commerce. In spite of his extensive field experience with coca in South America dating from 1885, Rusby never visited Trujillo and never saw or collected Trujillo coca in the field. Since this plant is grown only in a limited area in South -America and is not cultivated in other countries, it is under- standable that Rusby failed to recognize the subtle differences between the dried leaves of commerce and the living specimen of E. novogranatense in the greenhouse. Most botanists since the time of Rusby have agreed that the cultivated coca plants belong to two closely related species of E. Coca and E. novogranatense (Schulz, 1907; Payens, 1958; Towle, 1961; Gentner, 1972; Machado, 1972; Plowman, 1979a). Until recently Trujillo coca was generally included within E. novogra- natense. But no botanists had ever studied Trujillo coca in the field or had examined herbarium specimens collected in Peru. The only exception to this were studies of archeological coca leaves from coastal Peru, which certainly represent Trujillo coca (Harms, 1922: Griffiths, 1930; Towle, 1961; Plowman, 1979a). The first documented herbarium specimens of Trujillo coca which I have found in major herbaria are those collected by Augusto Weberbauer in 1914 in the Province of Pataz in Peru, along the upper Rio Maranon. These were originally identified as Erythroxylum Coca. No further specimens of Trujillo coca were made until the 1960’s when interested botanists began to sample the local varieties of coca. From these recent collections, it is finally possible to assess intelligently the taxonomic status of Trujillo coca. From herbarium studies, transplant experiments, chemical analyses and anatomical studies, it may be affirmed that Trujillo coca is in fact best placed in the species FE. novogranatense. However, it differs from the typical form of this species suffi- a ciently to justify treating it as a distinct variety within E. novogranatense. Disjunct populations of Trujillo coca have recently been dis- covered in northwestern Ecuador and adjacent Colombia, where it is occasionally grown as a medicinal dooryard plant. These populations are of great interest because of their geographical isolation from plants in coastal Peru and because of their prox- imity to areas where typical Colombian coca is grown. Further- more, they are growing in wet, montane habitat which 1s ecologically very different from the arid Peruvian coast. Contin- uing efforts by police authorities to annihilate coca in Ecuador now threaten these last remnant populations with extinction before they can be fully studied and before their possible role in the evolutionary history of coca can be properly assessed. In 1972, Machado, in his treatment of the Peruvian species of Erythroxylum, attempted to reduce Rusby’s E. truxillense to a variety of E. novogranatense, making the new combination E. novogranatense var. truxillense. Lamentably, this combination was not validly published, a fact which I overlooked in previous publications (Holmstedt et al., 1977; Plowman et al., 1978; Plowman, 1979a). Article 33.2 of the International Code of Botanical Nomenclature (Stafleu et al., 1978) plainly states that new combinations made on or after Jan. 1, 1953, must be accompanied by clear indication of the basionym (in this case, E. truxillense Rusby) and a full and direct reference given to its author and the original publication with page or plate reference and date. Machado (1972) neglected to include either the basio- nym or the original reference to Rusby’s publication. I would therefore like to validate the combination Erythroxy- lum novogranatense var. truxillense here. | also append a new description of the variety and designation of the lectotype. It should also be noted that Machado’s new species Erythroxylum Hardinii, published in 1969, is placed in synonymy with EF. novo- granatense var. truxillense on the basis of its morphology, leaf venation and anatomy. Erythroxylum novogranatense var. truxillense (Rusby) Plow- man, comb. nov. 56 Erythroxylum truxillense Rusby, Druggists Circular & Chemi- al Gazette 44: 220. t. 14. 1900. 45: 49. 1901. LECTOTYPE: “Truxillo coca, small coca. Native of Peru and cultivated.” The commercial drug presented by E. Merck & Co., New York City. Acces- sion No. 2684, Economic Museum of the New York Botanical Garden, no date. Specimen of dried leaves in a glass jar. (lectotype, ECON; iso- lectotypes consisting of small samples of this collection deposited at F, NY, USM). Erythroxylum Hardinii E. Machado, Anales Ci. (Lima) 7(1-2): 14. 1969. Type: PERU: Dept. San Martin, “Quebrada cerca de Crisnejas y el pueblo de Uchiza. Aparentamente en el lugar y sitios alendafios se cultivo hace mucho tiempo.” 15 Feb 1965, E. Machado 1256 (holotype, US 2803917: photograph of holotype, F, neg. 55486; isotypes, MOL, NCSC), Shrub to 3 m. tall, usually with multiple trunks reaching 4 cm. in diameter. Bark greyish brown with transverse and longitudi- nal cracks. Branches relatively dense, erect and spreading, straight or bending with age, light reddish to greyish brown, becoming longitudinally fissured. Branchlets straight, slender, not markedly zigzag in extension growth, nodes scarcely knobby, usually as thick as the internodes, stems mostly smooth but sometimes minutely scaly-ramentaceous, light green becoming reddish or greyish brown, lenticels punctate or rarely lenticular, usually not breaking the surface. Internodes alternatively ex- tended 6-20 mm. long, or shortened during ramenta production, I-3 mm. long. Ramenta often present but short, occurring mainly at the base of new shoots or on slow-growing short Shoots, inconspicuous. Stipules scarcely diverging from axis, membranaceous, pale green, turning light brown with age and disintegrating, keels in young stipules entire or rarely minutely fimbriate towards apex. Leaves usually persisting on the branches, weakly distichous, blade plane, narrowly elliptic to oblong-lanceolate, sometimes elliptic, apically acute to obtuse or rounded, basally acute to attenuate, 20-65 mm. long, 10-25 mm. wide, membranaceous, medium to light green above, pale green to glaucous green beneath, midrib flat adaxially or with only a slight medial ridge, pale green, often drying whitish beneath, 57 lateral nerves and veinlets usually obscure above, rarely promin- ulous, abaxial longitudinal lines and central panel usually inconspicuous or obscure. Petiole 1-5 mm. long. Flowers in axils of previous season’s growth, often near the tips of the branchlets or sometimes among the persisting leaves, usually 1-3(10) per node, scattered, if congested then only briefly, witha strong, foetid odor resembling raw pumpkin. Pedicels long, 3-11 mm. long, median length 9 mm. Petals fugaceous or sometimes persisting, appearing rotate, 3.5-5 mm. long, 2-3 mm. wide, the lamina subcymbiform or the midrib depressed above, yellowish green to cream, ligule creamy white. Staminal cup half the length to equaling the calyx. Drupe ovoid to ellipsoid, some- times fusiform, obtuse to acute at apex, 10-13 mm. long, 4-7 mm. in diameter. Whole plant, especially the leaves, suffused with odor of wintergreen (methyl salicylate). Chromosome number 2n = 24. ETYMOLOGY: The variety is named for the city of Trujillo in northern Peru, near the area of primary cultivation and from which the leaves are exported. COMMON NAMES: coca de Trujillo, tupa (Peru): coca (Ecuador). DISTRIBUTION: Cultivated in northern Peru on the western slopes of the Andes and in the valley of the Alto Rio Maranon. Disjunct populations known from northwestern Ecuador and adjacent Colombia. SPECIMENS STUDIED: COLOMBIA: NARINO: Rio San Juan, across river from Maldonado, 2000 m., “coca”, 26 Jan 1977, Boeke 854 (F); Tumaco, 9 May 1926, O.F. Cook 98 (US). ECUADOR: CARCHI: Maldonado, 1450-1650 m., “coca”, 2 June 1978, Madi- son etal. 4920 (F, SEL); Environs of Chical, 12 km. below Maldonado, along Rio San Juan, 1200 m., “coca”, 25 May 1978, Madison et al. 4447 (F, NCU, SEL). PERU: AMAZONAS: Prov. Chachapoyas: Rio Utcubamba, km. 6 from La Caclid, 1600 m., 25 Feb 1976, Plowman 5563 (ECON, F, USM), Plowman 5564 (ECON, F, K, USM), Plowman 5565 (ECON, F, NCU, RB, USM). Yumbay, above Balsas near Hornopampa, alt. 1100 m, 27 Feb 1976, Plowman 5583 (ECON, F, K, NCU, US, USM), Plowman 5587 (ECON, F, USM), Plowman 5588 (ECON, F, NCU, USM), Plowman 5589 (ECON, F, USM), Plowman 5590 (ECON, F, K, USM). Prov. Contumaza: Dist. Simbron, Fundo Farrat, 8 Jan 1966, Machado 2960 58 (NCSC), Machado 2961 (MOL), Machado 2962 (MOL), Machado 2963 (NCSC). DEPT. HUANUCO: PROV. LEONCIO PRADO: Pumahuasi, Fundo “Villa Gloria”, Experimental planting of Ing. Rodolfo Collantes, 830 m., “coca de Trujillo”, 6 Apr 1976, Plowman 5828 (ECON, F, K, USM), 19 Jul 1967 Machado 1051 (NCSC), Machado 1201 (NCSC), Machado 1214 (NCSC), Machado 1239 (NCSC); 20 Jul 1967, Machado 2573 (NCSC). LA LIBERTAD: Prov. Boliar: Laderas del Rio Marafion, Mar 1965, Machado 2989 (MOL). Prov. Otuzco: Hacienda Collambay, between Simbal and La Cuesta, 800 m., 2 Sep 1973, Lopez & Sagastegui 7999 (GH): 850 m., 3 Mar 1976, Plowman 5603 (ECON, F, UPS, USM), Plowman 5604 (ECON, F, USM), Plowman 5605 (ECON, F, USM), Plowman 5606 (B, ECON, F, USM, VT), Plowman 5607 (ECON, F, USM), Plowman 5608 (ECON, F, USM), Plowman 5609 (ECON, USM), Plowman 5610 (ECON, F, USM), Plowman 5611 (ECON, F, USM), Plowman 5612 (ECON, F, K, NCU, USM), Plowman 5613 (ECON, F, K, USM), Plowman 5614 (ECON, F, U), Plowman 5616 (ECON, F). PROV. PATAZ: Rio Maranon, abajo de Pataz, 1400 m, 18 Aug 1914, Weberbauer 7079 (MOL); Cerca del Rio Maranon, 18 Feb 1965, Machado 2987 (NCSC); Cerca de Tayabamba, orillas del Rio Maranon, 2 Mar 1965, Machado 2934 (MOL), Machado 2935 (MOL); Cerca de Huancaspata, orillas del Rio Maranon, 15 Jan 1966, Machado 2929 (MOL); Prov. Trujillo: Laderas de la Vertiente Occidental, Mar 1965, Machado 1248 (NCSC); Hacienda del Sr. Lopez, 1965, Machado 1018 (NCSC); Simbal, El Sacra- mento, 580 m, 3 Mar 1976, Plowman 5600 (ECON, F, USM), Plowman 5601 (ECON), F, USM), Plowman 5602 (ECON, USM), 17 Apr 1967, Ferreyra s.n. (ECON, USM), 16 Apr 1967, Ferreyra s.n. (USM); Simbal, La Banda, 500-600 m, “tupa”, 17 Apr 1967, Ferreyra s.n. (ECON, USM); Simbal, San Lorenzo, 500-600 m, “tupa”, 16 Apr 1967, Ferreyra s.n. (ECON, USM); Sim- bal, Las Animas, “tupa”, “Coca de Trujillo”, 17 Apr 1967, Ferreyra s.n. (ECON, USM); Simbal, Fundo de Rosario Gutierrez, “tupa”, 17 Apr 1967, Ferreyra s.n. (ECON, F, USM); Simbal, 4 Mar 1976, Plowman 5618 (ECON, F, USM); 3 km. from Simbal, “tupa”, 16 Apr 1967, Ferreyra s.n. (ECON, USM). LIMA: Lima, Universidad Nacional Agraria de La Molina, Jardin Botanico, 6 Apr 1975, Plowman 5209 (BH, ECON, F), 6 Feb 1976, Plowman 5388 (BH, COL, ECON). ACKNOWLEDGMENTS Part of the research reported in this paper was conducted at the Botanical Museum of Harvard University under a con- tract with the U. S. Department of Agriculture (No. 12-14— 1001-230). Post-doctoral fellowship support was also received from the Katharine A. Atkins Fund at Harvard University. Field work on Amazonian coca was carried out as part of Phase VII of the Alpha Helix Amazon Expedition 1976-1977 and was funded by the National Science Foundation Grant No. 76-80874. I am particularly grateful to Professor R. E. Schultes of 39 Harvard University for his continuing encouragement and sup- port of coca research and for providing access to the H. H. Rusby collections. | would like to thank Susan Marie Rossi for her long hours spent in searching for Rusby’s invaluable coca collections. Drs. Ramon Ferreyra and Oscar Tovar of the Museo Nacional de Historia Natural in Lima graciously gave valuable advice and assistance in field work. I also wish to thank the curators of the herbaria (cited according to the stan- dard abbreviations of Holmgren and Keuken, Index Herbario- rum, 1974) who kindly supplied herbarium specimens on loan for this study. REFERENCES Burck, W., 1890. Opmerkingen over de onder den Naam van Erythroxylon Coca in nederlansch Indie gecultiveerde gewassen. Teysmannia I: 385- 398, 449-464. t. 1-3. Ganders, F. R., 1979. Heterostyly in Erythroxylum Coca (Erythroxylaceae). J. Linn. Soc. Bot. 78: 11-20. Gentner, W. A., 1972. The Genus Erythroxy/um in Colombia. Cespedesia 1: 481-554. Griffiths, C. O., 1930. Examination of coca leaves found in a pre-Incan grave. J. Pharm. Pharmcol. 3: 52-58. Harms, H. von, 1922. Ubersicht der bisher in altperuanischen Grabern gefundenen Pflanzenreste. Festschrift Eduard Seler. Stuttgart. p. 180. Hieronymus, G., 1895. Plantae Lehmannianae in Guatemala, Costarica, Colombia et Ecuador collectae. Bot. Jahrb. Syst. 20, Beibl. 49: 35. Holmstedt, B., E. Jaatmaa, K. Leander & T. Plowman, 1977. Determination of cocaine in South American species of Erythroxylum using mass fragmen- tography. Phytochemistry 16: 1753-175S. Machado, E., 1969. Erythroxylaceae Novae II. Anales Ci. (Lima) 7(1~—2): 14-17. Machado, E., 1972. El genero Erythroxylon enel Peru. Raymondiana 5: 5-101. Martius, C. F. P. von, 1843. Beitrage zur Kenntnis der Gattung Erythro- xylon. Abh. Akad. Miinchen 3: 367-369. t. 6. Morris, D., 1889. Coca. Bull. Misc. Inform. (Kew) 25: 1-13. Payens, J. P. D. W., 1958. Erythroxylaceae. Flora Malesiana 5: 543-552. Plowman, T., 1976. Orthography of Erythroxylum (Erythroxylaceae). Taxon 25: 141-144. Plowman, T., L. Riidenberg & C. W. Greene, 1978. Chromosome numbers in neotropical Erythroxylum (Erythroxylaceae). Bot. Mus. Leafl. Harvard Univ. 26: 203-209. Plowman, T., 1979a. Botanical Perspectives on Coca. J. Psyched. Drugs 11: 103-117. Plowman, T., 1979b. Amazonian Coca. Paper presented before the 43rd International Congress of Americanists, Aug. 15, 1979, Vancouver, Canada. 60 Rostworowski de Diez Canseco, M., 1973. Plantaciones prehispanicas de coca en la vertiente del Pacifico. Revista Mus. Nac. (Lima) 39: 193- 224. Rusby, H. H., 1900. The botanical origin of coca leaves. Druggists Cir- cilar & Chemical Gazette, November: 220-223. Rusby, H. H., 1901. More concerning Truxillo coca leaves. Druggists Cir- cular & Chemical Gazette, March: 47-49. Rusby, H. H., 1921. Guide to the Economic Museum of the New York Botanical Garden. Bull. New York Bot. Gard. 11 (41): 1-138. Schulz, O. E., 1907. Erythroxylaceae. In A. Engler, ed. Das Pflanzenreich 4 (134): 1-164. Spix, J. B. von & C. F. P. von Martius, 1831. Reise in Brailien. Munchen. 3: 1169, 1180. Spruce, R., 1853a. Journal of a Voyage up the Amazon and Rio Negro. J. Bot. Kew Gard. Misc. 5: 212. Spruce, R., 1953b. Botanical objects communicated to the Kew Museum from the Amazon River, in 1851 and 1852. J. Bot. Kew Gard. Misc. 5: 240. Stafleu, F. A., et al., 1978. International Code of Botanical Nomenclature. IAPT Regnum Vegetabile 97. Utrecht. Towle, M. A., 1961. The Ethnobotany of Pre-Colombian Peru. Wenner- Gren Foundation. New York. 61 PLATE 9 FI Plate 9. Erythroxylum Coca var. Ipadu. Original specimen of Richard Spruce (no. 73), collected on the Rio Janauari, affluent of the Rio Negro, Brazil. From the Royal Botanic Gardens, Kew. 62 PLATE 10 Plate 10.) Erythroxylum Coca var. Ipadu. 1, Habit of plant. 2, flower with one petal removed. 3, petal showing ligule, adaxial view. 4, androecium, staminal tube opened up showing denticular margin. 5, gynoecium, showing large ovoid-oblong stigma. Drawing by L. T. Bates. 63 PLATE 11 Plate 11.) Eryvthroxylum Coca var. Ipadu. Habit of plants in cultivation, Brillo Nuevo, Rio Yaguasyacu, affluent of Rio Ampiyacu, Prov. Maynas, Dept. Loreto, Peru. Photograph by R. E. Schultes. 64 PLATE: 12 ERYTHROXYLUM novogranatense var. truxillense (Rusby) Plowman Plate 12. Erythroxyvlum novogranatense var. truxillense. 1, habit of plant, with flowers and fruits. 2, stipule, abaxial view. 3, flower, lateral view. 4, flower with petals and one sepal removed. Drawing by L. T. Bates. 65 PLATE 13 Plate 13.) Erythroxylum novogranatense var. truxillense. Flowering branch of Trujillo coca at Collambay, Dept. La Libertad, Peru (Plowman 5606). Photograph by T. Plowman. 66 PLATE 14 Plate 14. Erythroxylum novogranatense var. truxillense. Trujillo coca showing the use of pacay (Inga Feuillei) as a shade tree. Simbal, Dept. La Libertad, Peru (Plowman 5600). Photograph by T. Plow- man. Plantation of 67 PLATE 15 SIFEE Jo, Inales fad A Molina PLANTAS DED PERT Holotype Plate 15.) Ervthroxylum novogranatense var. truxillense. Holotype of Ery- throxylum Hardinii E. Machado from Crisnejas, near Uchiza, Dept. San Martin, Peru. Preserved at U. S. National Herbarium, Field Museum type neg. 55486. 68 BOTANICAL MUSEUM LEAFLETS HARVARD UNIVERSITY CAMBRIDGE, MASSACHUSETTS, MARCH-APRIL 1979 VOL. 27. No. 3-4 MONOPTERYX ANGUSTIFOLIA and ERISMA JAPURA: Their Use by Indigenous Peoples in the Northwestern Amazon! DARNA L. DUFOUR* AND JAMES L. ZARUCCHI** The collection of wild vegetable foods is an integral part of the subsistence pattern of indigenous peoples in the tropical forest of South America. The role of these foods in the indigenous diet ranges from trail snacks and emergency foods to important sources of nutrients. Outstanding examples are palm fruits, palm hearts and Brazil nuts. This paper is focused on two wild vegetable foods collected in the northwestern Amazon, the seeds of Monopteryx angustifolia Spruce ex Bentham and Erisma Japura Spruce ex Warming. Ethnobotanical information on these species is meager, but they are seasonally important food resources in the northwestern Amazon. The seeds of both are collected in large quantities, can be stored for long periods of time, and are available during the rainy season when animal protein is not particularly abundant. | A preliminary draft of this paper, entitled “Vegetable Protein in the Diet of Indians in the Northwestern Amazon” by Darna L. Dufour, was presented at the sixth annual meeting of the Canadian Association for Physical Anthropology, Niagara-on-the- Lake, Ontario, November 1978. *Department of Anthropology, State University of New York, Binghamton, New York 13901. **Botanical Museum and Department of Biology, Harvard University. the Botanical Museum, Harvard University, Cambridge, Massachusetts 02138. Subscription: $25.00 a year, net, postpaid. Orders should be directed to Secretary of Publications at the above address. Second-Class Postage Paid at Boston, Massachusetts. COLOMBIA | BRAZIL \ mm PIO CUDUYARI RIO VAUPES 4 +. % \ id 4 | «a { % \ 5 Kos 4 ACARICUARA = r St a q ~ Y args mn , MONFORT gt RIO PAPURI 4 COLOMBIA | i | | COLOMBIA Y RIO TIQUIE 2} 4 a ' VAUPES TERRITORY COLOMBIA | aha oe J ose | TERRITORIAL CAPITAL e INDIGENOUS VILLAGE e CATHOLIC MISSION POST AG RIVER OR STREAM ~~ KX Fy sg mMOUEL ‘ NATIONAL BOUNDARY eS Oy ee > | RESOURCE EXPLOITATION AREA OF YAPU PO PIR A - PARANA r $ : Prepared by Pew 8 Farmers Oat 1078 o30'w | Figure |. Map of the eastern Vaupes showing location of Yapu fieldsite. 70 The observations presented here are based on field work by the senior author with Tatuyo Indians living in the headwaters of the Papuri River at the village of Yapu, Vaupés, Colombia (figure 1).? The Tatuyo? are root crop horticulturists relying on bitter manioc, Manihot esculenta, as a caloric staple. Dietary animal protein is obtained from fish, and to a lesser extent from game. Insects and other invertebrates are minor sources of protein. The availability of animal protein shows a distinct pattern of sea- sonal variation. In general, fish and game are more abundant in the dry season and less so at the height of the rainy season. A wide variety of fruits, nuts and other vegetable products are collected and make important contributions to the diet. In addi- tion to Erisma and Monopteryx, these include Caryocar spp., Caryodendron orinocense, Euterpe spp., Hevea spp., Inga spp., Jessenia Bataua, Mauritia flexuosa, Micrandra spp. and Pora- queiba sericea. The upper Papuri area is transitional humid to very humid tropical forest with patches of caatinga and savanna vegetation. The terrain is predominantly rolling plain with occasional hilly uplands breaking the monotony of the skyline. Mean annual temperature is relatively high at about 26°C. and rainfall abun- dant at 340 cm./yr*. Seasonal differences in temperature and rainfall are not well marked. There is, however, a dry season of slightly less rainfall and slightly higher mean daily temperature from November to February. The rainy season begins in March and reaches a maximum in July. Both Monopteryx angustifolia and Erisma Japura are com- mon in virgin forest of the upper Papuri but not evenly dispersed ? The field work in the upper Papuri region was done by the senior author between 1976 and 1978. The research was carried out in collaboration with the Instituto de Ciencias Naturales, Universidad Nacional de Colombia, Bogota, and supported in part by an SSRC Dissertation Fellowship and NSF Grant BNS 75-20169. ow Tatuyo is the Spanish name for one of the linguistically distinct, exogamous groups of Tukanoan Indians in the northwest Amazon. In this paper, we refer to the inhabitants of the village of Yapu as Tatuyo. This is somewhat of an oversimplification, as there are other language groups represented in the village as well. However, the core of the village is Tatuyo, and the site of the village is within their traditional territory. + General ecological and meteorological data from Instituto Geografico “Augustin Codazzi” (1979:104). Bogota, Colombia. 71 within it. Monopteryx is encountered most frequently on the banks of rivers, in areas not generally subject to inundation, and along small forest streams. Erisma is generally found away from the rivers on terra firma. ERISMA, in the Vochysiaceae, is a genus of neotropical trees containing about sixteen species (Stafleu, 1954). The species are distributed from the Guianas to the Amazon basin. The genus 1s characterized by the fruits (see plates 16 and 17) which usually contain large wing-shaped calyx lobes. It was first described by Rudge in 1805 based on a collection by Martin from French Guiana. The description of Erisma Japura Spruce ex Warming 1s based upon a large forest tree collected by Richard Spruce at Panuré [Ipanore] on the Rio Uaupes, Brazil. Stafleu (1954) notes that the original material, Spruce 26/3, consists of two separate collections: November 1852 (flowering) and February 1853 (fruiting). The specific epithet is derived from the local name for the tree. Spruce notes that “the kernels are pleasant eating”, and that a “Japura-butter” was made by the natives. Other species of Erisma, notably FE. calcaratum and E. uncina- tum, have been reported as sources of “Jaboty butter” or “Ja- boty tallow” (Eckey 1954:562). MONOPTERYX is a genus in the Leguminosae, tribe Sopho- reae, which was described in 1862 by Spruce in Bentham’s treat- ment of the family in Martius, Flora Brasiliensis. Spruce considered the genus to contain two species, M. angustifolia and M. Uaucu, both found in the northwest Amazon basin and upper Orinoco River area. Two flowering collections made by Spruce in the 1850’s were the basis for the genus. Pittier in 1915 described M. Jahnii and provided a revised, although inaccu- rate, description of the genus. Monopteryx Jahnii is a synonym for Fissicalyx Fendleri Bentham (V. E. Rudd, pers. comm.). Both M. angustifolia and M. Uaucu are large forest trees with characteristic buttresses. Spruce (1908:20) referred to these but- tresses by the native word “sapopemas” [sapo, a root; pema, flat]. He gave the following account of M. angustifolia (1908: 335): “The trunk of this tree is 4 feet thick and 80 feet high. It has dz racemes of rose-coloured papilionaceous flowers. It grew on the rocky banks of the cataracts”. Recently a third species in the genus, M. inpae, has been described by Dr. William Rodrigues from the region of Manaus, Amazonas, Brazil. (Rodrigues, 1975). METHODS OF DATA COLLECTION The data were gathered as part of an energy flow study of the village of Yapu. All foods brought into the village were routinely weighed on a spring scale, capacity 30 kg. The edible portion was determined as percent edible by weight, using a scale accu- Pate to 1B: Collections of all important food plants were made in the general area of the village. Voucher specimens are deposited in the herbarium of the Instituto de Ciencias Naturales, Bogota, Colombia (COL); a duplicate set is deposited in the Economic Herbarium of Oakes Ames, Botanical Museum of Harvard Uni- versity, Cambridge, Massachusetts (ECON). Foods of unknown nutrient composition were prepared in the field for later biochemical analysis in Bogota. Indigenous methods of preservation, such as drying, smoking, and ferment- ing, were used when possible. In other cases, benzoic acid was used as a preservative at a concentration of 3% by weight. Dietary surveys were conducted on randomly chosen subjects. The method used was a 24-hour weighed dietary survey. The observer accompanied each subject for a 24-hour period and weighed all food portions on a 500 g. capacity dietary scale accurate to | g. Food eaten at a site where the dietary scale was not available was weighed on a 16 oz. spring scale or carefully estimated from a table of mean weights of food portions. OBSERVATIONS ON THE UPPER PAPURI Erisma Japura is known in Tatuyo as “bati” and in Lingua Geral as “yapura”5. Its pale yellow flowers appear in the dry 5 The name “japura” by Spruce (1908:399) is probably a variant of the pronunciation or spelling of “yapura™ as the glide /y/ in Latin America alternates with the fricative / zz (Pers. comm. Arthur P. Sorensen, Jr.). i3 season, and the fruits are ready for collection toward the middle of the rainy season’. The fruits are winged and roughly oblong in Shape. The average weight per fruit is 14 g., most of which is the woody pericarp. The edible portion of the fruit is nut-like, reminiscent of a large cashew in color and texture and, when fresh, accounts for about 24% of the total fruit weight (see plate 20). The fruits are gathered from the forest floor complete with woody pericarp. The Tatuyo are selective in their gathering and avoid malformed, worm-infested and sprouted fruits. In a sam- ple gathered by the author, it was found that about 20% were deemed inedible by Tatuyo standards. The trees are dispersed in the forest surrounding the village, but there are well-known areas which have a somewhat higher frequency of trees than others. Monopteryx angustifolia is referred to by the Tatuyo as “jimio” and by neighboring Tukano speakers as “simio”. The tree flowers toward the end of the dry season in January and February and, like Erisma Japura, fruits toward the middle of the rainy season in May and June’. The fruit is a long flattened pod, approximately 18-20 cm. long with an average weight of over 50 g. (plate 21). The pod is cracked open by the heat of the sun, allowing the seed to fall free to the ground. In the forest during the heat of the day, one can hear the characteristic crack of the pod, followed by the sound of the pod and seed falling through the leaves overhead. The pulse is a large flattened disk measuring approximately 4.5 < 3.5 * I cm., with an average weight of about 7 g. The seed coat is usually slipped off as the seed is collected. Sprouted seeds, with sprouts of up to 3 or 4 cm., are acceptable and frequently collected toward the end of the harvest season. As is also true of Erisma, the trees are not uniformly distrib- ° Flowers were collected in January 1977. Fruits were collected from a different tree in late April 1977, In 1978, the Tatuyo said that the trees were not going to flower, and the first flowers located appeared in mid March. ’ Fruits were collected in April 1977. A late-fruiting tree was spotted the first of August. As was the case with Erisma, the Tatuyo said that the trees were not going to flower in 1978. However, a flowering tree was located on March 21, 1978. 74 uted in the forest, and the location of prime collecting sites is well known by the Tatuyo. SEED COLLECTION In 1977, the harvest of Erisma and Monopteryx overlapped during an eight-week period toward the middle of the rainy season. Women collected small amounts of seeds as they trav- eled to and from their cultivated plots; and, during the height of the harvest, both men and women devoted entire working days to seed collection. Figure 2 shows the cumulative collection rates for both seeds from April 12 through June 12, 1977. The total amount of Erisma collected was 1311 kg. Considering that this seed has an edible portion of about 24%, this represents some 314 kg. of edible food. A total of 723 kg. of Monopteryx seed was collected during the same period. This is considered to be 100% edible as collected. During a period of intensified collec- tion in the fifth week, some 310 kg. of Erisma fruits and 180 kg. of Monopteryx seeds were gathered by the men of the village as part of a traditional “Yurupari” ritual. SEED UTILIZATION AND STORAGE Erisma. During the harvest season, small quantities of Erisma seeds are eaten raw or toasted. They have a pleasant although occasionally bitter taste. Eaten raw, they leave a thin film of wax-like fat on the roof of the mouth. Collected in quantity, Erisma seeds are cooked and prepared for storage in the form of a “butter”, referred to below as “bati- butter”. The method of preparation is as follows: the fruits are boiled in water until the inner seed is cooked and the woody pericarp softened. The outer shell, or pericarp, is then easily removed, taking with it the thin skin covering the seed. The ® “Yurupari” is used here as a general term for a type of dance festival, communal ritual common in the northwest Amazon involving the use of sacred trumpets and the exchange of forest fruits. Wallace (1889:241), who traveled the Vaupes River in the 1850’s, was the first to describe these rites. A recent analysis of this ritual among the neighboring Barasana is provided by Hugh-Jones (1974). 75 z ° 7 e o a Ww WwW 3 @ Be a = WwW = = = q = q rea) = o @ O ' ' \ * 0 A \ z ‘ re) = ~ \ OZ _ O 5 \ WO “4 Ww wreEaeac a es &—~——$—<_g95SP— a ~ 2 yee WwW s a. ae © s s eo * oO 2 ~N =O ‘a uw P ara e) ou \ YM > ¥yv < *. iw = * = Ld l l l _l J oO O ie) ie) oO oO oO fe) oO oO oO oO oO fe) vt N 12) @ wo +t N ‘SOM ‘NOILOS1NIOO Iiva 8 OIWIT 3AILYINWND Figure 2. Cumulative collection rates for “jimio” (Monopteryx angustifolia ) and “bati” (Erisma Japura) during the 1977 harvest at Yapu. 76 peeled seeds are washed thoroughly in a basket. Once washed, the seeds are reboiled until very soft (about 45 minutes) and then, while still hot, either pressed through a basket sieve or crushed to a butter-like consistency with a mortar and pestle. The crushing of the seeds is done quickly so that they are still warm when buried in a leaf-lined pit in the ground. Storage pits used for bati-butter are dug neatly into the house floors in areas free from ants and other insects. The pits are about 30 cm. in diameter and of sufficient depth to allow the stored food to be covered with 6 to 8 cm. of soil. A typical pit is approximately 60 cm. deep. Once dug, the pit is carefully lined with the juvenile leaves of a tree tentatively identified as Clathro- tropis macrocarpa (Leguminosae). This leaf, known in Tatuyo as “miapu”, is commonly used to package food. The leaves are placed against the sides of the pit with the lower surface of the leaf toward the pit wall, petiole up, and the pointed apex bent to cover the pit floor. Additional leaves are bent in the pit to cover the pit floor thoroughly. When complete, the leaf-lining 1s 3-, or 4-leaves deep at all points and prepared in such a way that the inner leaves are lower in the pit than the outer ones. The bati- butter is placed in the pit and compacted. The leaves are then wrapped over it in a systematic manner, beginning with the inner leaves and finishing with the outermost ones. When complete, the entire leaf packet is covered with soil. An estimated 85% of the Erisma collected in the period of observation was prepared as bati-butter and stored in leaf-lined pits in house floors. According to the Tatuyo, this butter can be stored for up to a year if proper care is exercised in changing the leaf-lining periodically and in securing insect-free storage pits. Under such anaerobic conditions, the butter ferments, develop- ing a sharp taste and an odor not unlike a strong cheese. This odor is no doubt what Spruce referred to in his note on the Kew specimen: “. . . People who can get over its vile smell (which 1s never lost) find it exceedingly savoury.”? The Tatuyo do indeed relish the strong flavor it attains after prolonged storage. Bati- butter, either fresh or fermented, is prepared by cooking it with water into a thick dip or adding it to “puné”, a fish and manioc 9 From Stafleu, 1954:474. 77 porridge!’. As stored supplies of bati-butter decrease after the harvest, they become increasingly precious as food. It is then that women dig up portions of their supply for special occasions and as gifts. One further use of bati-butter should be mentioned. The Tatuyo occasionally prepare a paste of bati-butter, charcoal, and the larvae of an unidentified species of wasp, which they use as a fish poison. The paste is fashioned into pellets which the fisherman throws into the stream as both bait and poison for “boteka” (Leporinus alternus''), a much sought-after river fish. In the latter part of the rainy season, in late August and September, the Tatuyo collect Lepidoptera larvae and pupae known as “batiya” (family: Noctuidae!2) in great abundance froé Erisma trees further downstream on the Rio Papuri and in the area of Acaricuara!}. The larvae are collected as they des- cend from the canopy to pupate in the forest floor. According to informants, these trees are identical in flower and fruit to the collections made at Yapu and are assumed to be Erisma Japura. Monopteryx. During their harvest, Monopteryx seeds are occasionally eaten peeled and roasted, although they have an exceedingly bitter taste. The vast majority of the harvest, how- ever, 1S prepared for consumption as follows: the flexible seed coats are removed if this was not already done when the seeds were collected. Each seed is then split neatly in half and the remaining paper-thin integument peeled off with the aid of a knife or fingernails. This peeling is tedious and is most often done by groups of women in the early evening. The split pulses are then softened by boiling in water for a considerable period of '° Porridges of fish and manioc starch are very common dishes in the northwest Amazon. They are referred to in Spanish as “munica”. The general term in Tatuyo is “pune”. '! Determined by German Galvis V., Unidad Ictiologia, Seccion Zoologia, Instituto de Ciencias Naturales, Universidad Nacional de Colombia, Bogota. A voucher specimen is deposited at the same institution. '2 Determined by D. M. Weisman of the Insect Identification and Beneficial Insect Introduction Institute of the United States Department of Agriculture. '\ The traditional spelling is Wacaricuara, but current usage is now Acaricuara. 78 time, up to three or more hours. As the pulses boil, a small quantity of a light green oily substance is liberated, forming a scum on the sides of the pot. The cooked pulses are edible but still very bitter. To remove the bitter taste, they are soaked in the river for two days. This soaking is done by emptying the cooked pulses into a large, loosely woven basket and placing the basket in the river so that the water circulates through it. Prepared pulses are eaten plain or as “jimio janique”, a drink made by adding mashed pulses to the boiled liquid removed from freshly grated manioc (Manihot esculenta). Prepared pulses not destined for immediate consumption are mashed and buried in leaf-lined pits in much the same way as described above for bati-butter. Generally, however, the leaves of Phenakospermum guyanense (Musaceae) are used to form the leaf packet, and the pulses are not heated prior to burial. If proper care is exercised, Monopteryx can be stored in this manner for about six months. Under such anaerobic storage conditions, Monopteryx ferments, developing a sharp, pleasant taste. In the fermented state, it is occasionally eaten plain but most often prepared as “jimio janique”. An estimated 50 to 75% of the total amount of Monopteryx seed collected during the 1977 harvest was stored for periods of time in the ground. COMPOSITION OF SEEDS AND SEED PRODUCTS In order to determine the composition of Monopteryx and Erisma seeds, samples of freshly cooked seed were prepared as for eating and preserved with benzoic acid. Duplicate samples of each were stored in leaf-lined pits as described above, until just prior to shipment, at which time they were removed to sterile glass jars and sealed with paraffin wax. The fermentation time of the pit-stored samples was four to six weeks. The results of food composition analyses of “jimio” and “bati” are shown in Table 1. The analyses were done on partially dry material by the Instituto Colombiano Agropecuario in Bogota. As indicated in Table 1, Erisma seed is high in fat (70.5g./ 100g.). and food energy (768 kcal./100g.). Fermentation in- creases the fat content (to 80.1g./ 100g.) and the energy value (to 79 O8 TABLE 1 COMPOSITION OF BATI-BUTTER AND JIMIO, 100 GRAMS EDIBLE PORTION FOOD CARBOHYDRATE MOISTURE” ENERGY PROTEIN FAT Total Fiber ASH FOOD % kcal. (MJ) g. g. g. g. g. Bati-butter (Erisma Japura) Fresh cooked (64) 2.3 768 (3.2) 9.4 70.5 13.3 1.8 4.5 Cooked, fermented (54) 2.5 850 (3.6) 9.9 80.6 5.8 1.3 1.2 Jimio (Monopteryx angustifolia) Fresh cooked (78) 5.7 554 (2.3) 12.6 23.2 53.0 3.4 5.5 Cooked, fermented (76) 3.1 $79 (2.4) 17.5 26.6 $1.7 2.9 1.1 Nuts and oil seeds” 37 596 (2.5) 17.0 54.3 21.7 3.7 3.3 Pulses” 11.3 354 (1.5) 25.4 5.0 55.0 5.5 3.3 a. Value in parentheses is the percent moisture of food as prepared for eating. Analysis was done on partially dry material. b. Mean value for 8 nuts and oil seeds. See note 14. c. Mean value of 6 cultivated pulses. See note [5. 850kcal./100g.). In comparison with average values for eight cultivated nuts and oil seeds!4, bati-butter is considerably higher in fat and calories and lower in protein. On a dry weight basis, the values for bati-butter are similar to those for pecans (Carya illinoensis) (Wu Lueng and Flores, 1961:68~71). Monopteryx is moderately high in fat (23.2g/ 100g) and pro- tein (12.6g./100g.). In a fermented state, it is slightly higher in protein (17.5g./100g.) and food energy (579 kcal./100g.). In comparison with average values for six cultivated pulses, Monopteryx is relatively high in fat and low in protein!>. The protein values reported here are for crude protein only. Although it is a question of much interest, the biological value of Monopteryx protein is not currently known. The oil composi- tion of both Monopteryx and Erisma seeds is also of interest. The oil of Monopteryx angustifolia is not well known, but it has all the properties of a good edible oil (Mors and Rizzini 1966:27). Little is known of the oil of Erisma Japura, but two closely related species, FE. calcaratum and E. uncinatum, have been shown to have oils of good quality (Pesce 1941:91-94). CONSUMPTION PATTERNS The consumption of Monopteryx angustifolia and Erisma Japura was measured during the 1977 May-June harvest period in the village of Yapu. The 24-hour weighed dietary survey method was used to record food intakes on a sample of eighteen adults. During the survey period, Monopteryx seeds were con- sumed freshly prepared as the main part of a meal and as a beverage. Erisma seeds were also eaten freshly cooked as the principal dish in a meal and, in two instances, as a dip made of fermented bati-butter. In terms of the total food intake, the two seeds accounted for 14 Values shown are the average values for the following eight nuts and oil seeds: Anacardium occidentale, Bertholletia excelsa, Corylus spp., Helianthus annuus, Pinus cembroides, Prunus Amygdalus, Sesamum indicum, and Terminalia Catappa (Wu Lueng and Flores, 1961:68 71). iS. The values shown are the average values for the following six cultivated pulses: Cicer arietinum, Glycine Max, Phaseolus vulgaris, Pisum sativum, Lens spp., and Vicia Faba (Wu Lueng and Flores, 1961:66 68). om about 9% of the mean daily caloric intake and about 10% of the protein intake. Of the two seeds, Monopteryx was consumed more frequently and in larger quantities. It accounted for about 70% of the total caloric and over 80% of the protein contribution of the two seeds. There is a general tendency among the Tatuyo to use vegeta- ble products gathered from the wild in order to supplement meals in which there is little or no animal protein. This pattern was clear in the consumption of fresh Monopteryx seed. Only six out of the eighteen adults surveyed during the dietary study consumed fresh cooked Monopteryx seeds as part of a meal. In all of those meals, animal protein was not only absent but unavailable in the household. The exception was one meal con- taining a small amount of roasted insects contributing less than | g. of animal protein. The mean per capita intake of these individuals, five women and one man, was about 300 g. of fresh cooked seed, giving them an average intake of about 370 kcal. (1.5 MJ) and 9 g. crude vegetable protein. The highest consumption of Monopteryx in this group of six was one woman whose intake over a 24-hour period was 548 g. of fresh seed, and an additional 57 g. as “jimio janique”, making the daily total 605 g. of fresh seed. The seed provided her with about 750 kcal. (3.1 MJ) and 17 g. vegetable protein. She explic- itly stated that she was eating “jimio” because there was no fish or meat available. The use of Monopteryx seed as a beverage (jimio janique) tends to be reserved for days of low animal protein availability. This is especially true of stored supplies of fermented “jimio” after the harvest. This investigator’s carefully stored supply, set aside early in the harvest specifically for biochemical analysis, proved to be no exception. One dreary rainy morning when we had breakfasted without fish or meat, the family with whom we were living dug up the sample of fermented “jimio” and prepared it as “jimio janique” so that we would not be hungry. Like Monopteryx, fresh Erisma seed is also consumed at meals in which there is little or no animal protein available. The dietary survey data, however, indicate a much lower consump- tion of these seeds. The mean per capita intake of the five out of the eighteen adults in the dietary survey who consumed freshly 82 cooked Erisma seed was 44 g. This provided them with approxi- mately 125 kcal. (0.5 MJ) and 1.5 g. crude vegetable protein. Since both Erisma seed and bati-butter are very high in fat, they are more important as sources of calories than of vegetable protein. The ability of the Tatuyo to store both Monopteryx and Erisma in a fermented state extends their availability. The har- vest period lasts into the mid-rainy season when neither hunting nor fishing is very productive. Stored supplies of both seeds are important food resources at the height of the rainy season in July, when animal protein of almost any kind is difficult to obtain. Stored supplies of “jimio” in the village were almost all exhausted by late August, two months after the harvest. One household held a supply for two additional months. Bati-butter is eaten in smaller amounts than “jimio” and stored supplies were stretched to at least six months in many households. Much of the bati-butter, however, was eaten during the height of the rainy season, even though it was explicitly stated by the Tatuyo that it was being saved to eat with fish porridge in October when fish are more abundant. Our purpose here has been to describe and illustrate the use of two little-known food plants in the diet of indigenous peoples in the northwestern Amazon. By doing so, we suggest that closer attention be paid to collected vegetable foods. Although the caloric staple of these people is provided by cultivation, and animal protein is obtained from wild fauna, gathered vegetable foods are important supplementary nutritional resources. ACKNOWLEDGMENTS We are indebted to Dr. Polidoro Pinto E., Director of the Instituto de Ciencias Naturales, Universidad Nacional de Colombia, Bogota, and to the specialists at this institute for their generous hospitality and technical assistance. We would also like to thank Drs. V. E. Rudd, R. S. Cowan, and J. J. Wurdack for their assistance in verifying plant determinations; Dr. Ger- ardo Pérez Gomez, Departamento de Quimica, Universidad Nacional de Colombia, for his help with the biochemical anal- 83 yses; Dr. Lloyd Knutson, Chairman of the Insect Identification and Beneficial Insect Introduction Institute, United States Department of Agriculture, Beltsville, Maryland, for his cooper- ation in the identification of insect specimens; and to Dr. Arthur P. Sorensen, Jr. for reading a draft of the manuscript. In particular, the senior author would like to express her gratitude to Dr. Gonzalo Correal U., Secci6n de Antropologia, Universidad Nacional de Colombia, for his support and supervi- sion during the field project, and to her husband, Paul N. Pat- more, who provided immeasurable help in the field study. REFERENCES Eckey, E. W. 1954. VEGETABLE FATS AND OILS. Reinhold Publishing Corp., New York. Hugh-Jones, S. 1974. MALF INITIATION AND COSMOLOGY AMONG THE BARASANA INDIANS OF THE VAUPES AREA OF COLOMBIA. Ph.D. Dis- sertation, University of Cambridge, Cambridge, England. Instituto Geografico “Agustin Codazzi”. 1970. ATLAS BASICO DE COLOMBIA, Banco de la Republica, Bogota, Colombia. Mors, W. B. and C. T. Rizzini. 1966. USEFUL PLANTS OF BRAZIL. Holden- Day Inc., San Francisco. Pesce, C. 1941. OLEAGINOSAS DA AMAZONIA. Revista da Veterinaria, Belém, Para, Brazil. Pittier, H. 1915. On the characters and relationships of the genus Mono- pteryx Spruce. Bull. Torrey Bot. Club 42: 623-627. Rodrigues, W. A. 1975. Contribuigao para o estudo do género Monopteryx Spr. ex Benth. (Leguminosae) da Amaz6nia. Acta Amazonica 5(2): 153-155. Spruce, R. in G. Bentham. 1862. in Martius, FLORA BRASILIENSIS /5(1): 307 t. 122. Spruce, R. in E. Warming. 1875. in Martius, FLORA BRASILIENSIS 13(2): 109 t. 21 fig. 1. Spruce, R. 1908. NOTES OF A BOTANIST ON THE AMAZON AND ANDES, 2 Vols. A. R. Wallace, editor, Macmillan and Co., Ltd., London. Stafleu, F. A. 1954. A Monograph of the Vochysiaceae IV. Erisma. Acta Bot. Neerl. 3(4):459- 480. Wallace, A. R. 1889. A NARRATIVE OF TRAVELS ON THE AMAZON AND RIO NEGRO. Second edition. Ward, Lock and Co., London. Wu Lueng, W. and M. Flores. 1961. TABLA DE COMPOSICION DE ALIMENTOS PARA USO EN AMERICA LATINA. Instituto de Nutricién de Centro America y Panama and Interdepartmental Committee on Nutrition for National Defense, U.S. Government Printing Office, Washington, be oe 84 PLATE 16 ERISMA L Japura Plate 16. Illustration of Erisma Japura Spruce ex Warming from Martius, FLORA BRASILIENSIS. PLATE 17 Plate 17. Fruit of Erisma Japura Spruce ex Warming. s6 PLATE 18 A CSSA TEL ad) Ranta MONOPTERYX angustifolia. Plate 18. Illustration of Monopteryx angustifolia Spruce ex Bentham from Martius, FLORA BRASILIENSIS. 87 PLATE 19 Plate 19. Buttresses of Monopteryx angustifolia Spruce ex Bentham from Spruce’s NOTES OF A BOTANIST ON THE AMAZON AND ANDES, page 335. 88 PLATE 20 Plate 20. Freshly harvested leaves and fruits of Erisma Japura. One fruit toward the center of the group is split open to show the edible kernel. Photo- graph by Paul N. Patmore. 89 PLATE 21 Plate 21. Seeds and fruits of Monopteryx angustifolia Spruce ex Bentham. One pod has been opened to demonstrate the relative size and position of the seed. Photograph by Paul N. Patmore. 90 Pia 22 ; fe 4 ~ * sen tn, a ¢ * . ' : § ’ * : Q = Baskets of Erisma Japura and Monopteryx angustifolia gathered in 1. Patmore. Plate 22. preparation for the Yurupari ceremony. Photograph by Paul N 9] BOTANICAL MUSEUM LEAFLETS VOL. 27, No. 3-4 MARCH-APRIL 1979 STUDIES IN THE GENUS MICRANDRA II* Miscellaneous Taxonomic and Economic Notes RICHARD EVANS SCHULTES During the course of monographic studies on the euphorbi- aceous genus Hevea, source of most natural rubber of com- merce, related genera—Micrandra (including Cunuria), Vau- pesia and Joannesia—have been investigated in the field and herbarium. Sundry notes of interest from the viewpoints of tax- onomy, floristics or economic value of these genera have accu- mulated. The following notes concerning various species of Micrandra are offered towards an eventual monographic treat- ment of the genus. Little is known about the commercial value of the latex of Micrandra. \t has therefore seemed appropriate to publish a short survey of the possible utilitarian importance of the genus based upon studies of the literature, field work and significant notes on herbarium specimens. During my taxonomic studies, numerous herbarium collec- tions have turned up which, for one reason or another, merit special citing. Amongst these specimens are those of the late Paul H. Allen who worked in the Colombian Vaupés on rubber in the early 1940’s during the shortage of this commodity due to the war emergency. Allen’s collections are notable because of his extremely detailed field notes. This study has been assisted by a grant from the Cabot Foun- dation of Harvard University. UTILIZATION OF MICRANDRA AS A SOURCE OF RUBBER As a source of commercial rubber, Micrandra is of very secondary, if any real, importance. That rubber is produced from these trees at the present time is doubtful, but an apprecia- ble amount from several species may have found its way into *The first contribution in this series appeared in Bor. Mus Leafl., Harvard Univ. 15(1952)201-222. 93 commerce in the past, especially during periods of feverish need to increase production. To judge from the literature, one might assume that Micran- dra had enjoyed a relatively major role in the world rubber supply. Most literature reports of the exploitation of Micrandra for rubber are either without foundation or else are erroneous repetitions of one or more of the earlier botanical explorers who noted this exploitation as an isolated observation not connected with a critical economic study of the industry as a whole. In the earlier literature, it seems that there are but two reports on Micrandra by field botanists with first hand experience and whose statements should, for this reason, carry greater weight than most other literature references. Richard Spruce, who col- lected the type material of the genus Micrandra (including Cunuria) in the Rio Negro basin of Amazonian Brazil, failed to mention the utilization of Micrandra latex in commercial rubber production, an industry which, during his sojourn in the region (1851-1854), was beginning its great development on the basis of seringa or Hevea. In his posthumously published notes, he wrote (Spruce, R.: [ed. A. R. Wallace] “Notes of a botanist on the Amazon and Andes” | (1908) 508): “On the Vaupés, I met with two trees (2427, 2479 hb.) of a genus apparently not far removed from Siphonia [Hevea], which yield pure rubber and are also called by the Indians xeringui .. .” E. Ule, who carried out extensive surveys on rubber plants in the Amazon regions in the early years of the present century, stated (Ule, E.: “Veranlassung und Verlauf von Ules Expedition nach den Kautschukgebieten des Amazonenstromes” in Tropenpflanz. Beih. 6 (1905) 1) that, although Micrandra produces a good rubber, the trees are sel- dom exploited because the latex cannot be mixed with that of Hevea and because it is too troublesome for the rubber tappers to cut it to the exclusion of Hevea. In another report on his rubber studies, Ule (Ule, E.: “Die Kautschukpflanzen der Ama- zonas Expedition und ihre Bedeutung fiir die Pflanzengeogra- phie” in Engler Bot. Jahrb. 35 (1905) 670) reported merely that Micrandra siphonioides is found to be frequent in the rubber forests of the Rio Negro and elsewhere. From these two references, apparently, has stemmed a flood of reports, occasionally highly misleading, in both popular and 94 technical literature. Some include Micrandra in lists of rubber- yielding species. While this is strictly correct, the unfortunate impression is often given that Micrandra should be counted amongst the commercial sources or rubber (Clouth, F.: “Rubber, gutta-percha and balata” (1903) 30; Walle, P.: “Au pays de l’or noir, le caoutchouc de Bresil,” ed. 2 (1912) 124; Carneiro, A. J. de Souza: “Rubber in Brazil” (1913) 8; Ferguson, Jr. ed. 3: “All about rubber and gutta-percha” (1899) iii, clxxxiv). This last source states that it grows along the most “steamy valleys” of the Amazon and is indiscriminately cut by the natives “to furnish Para rubber”! Others definitely assert that Micrandra-rubber was actively exploited and entered into commerce either alone or as an adulterant of Hevea-rubber (Morris, D.: “Plantes pro- duisant le caoutchouc du commerce” in Bull. Soc. Etudes Colon. no. 5 (1899) 178; Ehrhardt, K.: “Die geographisch Verbreitung der fiir die Industrie wichtigen Kautschuk- und Guttaperchap- flanzen” (1903) 26; Seeligmann, T., G. Lamy-Torrilhon & H. Falconnet: “Indian rubber and gutta-percha,” ed. 2 (1910) 15; Ramondt, A. S.: “Caoutchouc, guttah-pertja en balata (1907) 6). Still others quite correctly assert that rubber from Micrandra is said to enter the trade, passing as a grade of Para rubber (Hevea), but that no precise information on this point can be cited to substantiate the report (Warburg, O.: “Les plantes a caoutchouc et leur culture” (1902) 21, 48; Jumelle H.: “Les plantes a caoutchouc et a gutta” (1903) 60; Reintgen, P.: “Die Geographie der Kautschukpflanzen” (1905) 25; Schidro- witz, P.: “Rubber” (1911) 33; ed. 2 (1916) 33; Brown, H.: “Rubber, its sources, cultivation and preparation” (1914) 33). Occasionally, a report will credit Micrandra with producing in the Hevea-rubber districts of the Amazon the most highly esteemed grade of rubber (Pontio, M.: “Analysis du caoutchouc et de la gutta-percha” (no date) 8). During my twelve-year study of lacticiferous plants in the Amazon Valley—especially in northwestern Brazil and south- eastern Colombia—I saw no evidence of exploitation of Micran- dra trees for their rubber. This was true even in areas where Micrandra was extremely abundant and easily accessible. Furth- ermore, I saw no evidence, such as scars on the bark, that such exploitation had been carried on in the past, although former 95 workers may well have felled the trees for ringing and would, therefore, have left no traces. The period of my field work (1941-1953) coincided with a renaissance of the wild rubber industry with high prices due to the world shortage of the commodity during and immediately after the war. I paid very special attention to Micrandra, since | was collecting material for a monographic study of Hevea and its relatives, such as Micrandra. Had any tapping of Micrandra been under way or had it formerly been carried out in the northwest Amazon, I believe that some indication of it could not have escaped my ken. The late Dr. Adolfo Ducke, much of whose half century of botanical exploration in the Amazon of Brazil was dedicated to the study of latex-yielding trees, failed to mention on any of his field labels or in his writings the use of Micrandra as a commercial source of rubber. It would seem, therefore, that we are wholly justified in assuming that Micran- dra is not now, and probably has not in the past been, exploited in either the Colombian or the Brazilian sectors of the Amazon basin as a major source of rubber. Fortunately, however, we now have definite and reliable information concerning the former use of Micrandra as a rubber tree. It appears that Micrandra was tapped in Venezuela, in localities where the trees abound and where rubber-yielding spe- cies of Hevea are lacking. Two reports from field technicians to Mr. Oliver E. Nelson, Special Representative of the Rubber Development Corporation in Caracas, give valuable informa- tion on Micrandra in Venezuela. Copies of these reports are preserved in the National Archives in Washington, D. C. and in the Botanical Museum of Harvard University. On May 4, 1943, Mr. Harry J. Fuller, writing on his explora- tory trip to the Paragua-Caura area of southern Venezuela, stated: Caracas to Ciudad Bolivar by air, thence to La Paragua on the Paragua River by truck, thence by boat up the Paragua River. First Micrandra at the Auraima rapids. Rio Oris enters the Paragua at Auraima and was ascended for some distance. Thick stands of Micrandra. Continued up Paragua. Numerous small cafos enter the Paragua. Micrandra much more numerous on these canos than on banks of Paragua itself. Two days above the Oris reached Rio Torono. About 300 yards wide at confluence. A half day up the 96 Torona is an abundance of Micrandra. Established camp and started experiments. Expt. 1. 2 men with machetes tapped 10 trees in 3!4 hours, on one side up to 7 ft. 475 grams of rubber after 3 weeks air drying. Men walked a total of 350 yds. Fuller thinks that tapping to 20 or 30 ft. with spurs for climbing would make it possible for one man to produce | to 14 kilos per day. Expt. II. Two Indians tapped 8 trees in 3 hours to a height of 514 ft. 350 kg. rubber after 3 weeks air drying. Expt. II]. 75-foot tree felled. Diam. 19 in. at 3 ft. Two men made transverse cuts on trunk with machetes at 3-inch intervals, extending about 2/3 circumference. Cuts extended to main branches, about 45 ft. from ground. Three hours required. Less than | kg. rubber (after air drying) collected next day. Continued up Paragua. Major rivers with large concentrations of Micrandra are: Rio Carapo and its tributaries. Rio Carum and its tributaries. Rio Mari which is only | km. from the Rio Capapaipa, an important tributary of the Caura. The best way of reaching the upper Caura is by way of the Paragua and its tributary, the Mari, from which, by | km. portage, Caura basin may be reached. Micrandras grow close to rivers and canos, rarely at distances greater than 300-400 yds. Number of Micrandras increases as one ascends the Paragua. Not scattered but commonly grow in groups of 5 to 25 or 30 trees, possibly the offspring of a single seed, developing from root sprouts. Seldom found isolated. Individual trees in a group 10 to 50 yds. apart. Fuller did not get into the upper Caura but had reliable informa- tion that Micrandra was the only rubber tree there and is present in considerable quantity. On July 8 of the same year, Mr. William O. Hansen, reporting on his studies of Micrandra in the Territorio Federal del Amazo- nas in Venezuela, had the following data to offer: Rubber known locally by several names: 1. “Caucho de Gauca” because the Gauca or Guacamaya (macaw parrot) eats the fruit. 2. “Caucho de Invierno” (wet season rubber), because it yields best in the wet months. 3. “Arara Seringa” which seems to be the Portuguese or Yeral Indian equivalent for Caucho de Guaca. Reported used commonly in Brazil. 4. “Seringa Irari.” Used from Manaos north to the Colombian and Venezuelan borders wherever the Yeral Indian dialect is spoken. The word “Irari” means unreal or imitation and is used to distin- guish Micrandra from Hevea, the real rubber. Micrandra is said to exist in most parts of the territory. In groups of 20 to 50. Mostly in low land flooded yearly. Reported along the Orinoco, Casiquiare, Guainia and Rio Negro. The greatest concen- 97 trations are reported in the areas around Maroa, San Carlos, the Desecho de San Miguel (Casiquiare) and the Cano de Guami (Rio Negro near Santa Rosa de Amanadona). Some of the men who worked rubber 1900-1914, said they ex- ploited Micrandra. Trees felled and ring-tapped every 3 ft. Latex collected as tree scrap. After 2 to 3 weeks, when the leaves were wilted, trees ring-tapped again at 2 to 3 inch intervals and the latex collected in gourdes of leaves. A tree 3 ft. in diam. would give about | gal. latex. Fairly stable. Coagulate in 5 to 10 hours. Various tapping tests tried mostly with low yields. In a few cases fair yields were obtained. One tree 24” DBH gave | 1/4 quarts latex at one tapping. This tree was just beginning to bud and other trees that gave good yields were just starting to grow. The latex coagulates rapidly on the addition of water. A native said that her father collected Micrandra rubber and always found a large lump of rubber at the base of the trees. Used a sharp stick to probe for the rubber. The lumps weighed 5 or 6 kg. each. Average yield about 150 gms. air dry scrap of lump per tree per day. An expert tapper with climbing equipment can tap about 10 trees per day. It is interesting to note that the first of these reports does not mention exploitation but that the second maintains that Micran- dra was exploited from 1900 to 1914. The trees were felled, ringed and the latex allowed to coagulate as scrap on the trees. This was undoubtedly the so-called Caura rubber formerly exported from the Orinoco basin of Venezuela. Neither of these reports indicates that Micrandra was in production in the 1940’s, and I have been unable to find any data which would support the belief that these trees were ever used to any extent during the recent war-shortage of rubber. Unfortunately, I have been unable to find voucher herbarium specimens upon which these foregoing reports might have been based. In the Venezuelan region concerned two closely related species are known to occur: Micrandra minor and M. siphoni- oldes. Allen, who carried out surveys for the Rubber Reserve Corpo- ration in Colombia at the same time, has recorded on voucher specimens exceedingly interesting notes concerning the type of rubber and its possible exploitation. All of his observations on Micrandra were made amongst the Tukano Indians on the Rio Papuri, an affluent of the Rio Vaupes, which forms part of the boundary between Brazil and Colombia. Of Micrandra minor, 98 he reported (A//len 306/): “Latex pale yellow, seldom flowing freely, usually coagulating in the cuts ina manner reminiscent of Castilla. This scrap can be removed after an interval of about three days, usually being wound into balls. Tensile strength excellent, considerably better than weak fine Hevea. Some resin content.” On a separate label (A//en 3067), Allen noted: “Coagu- lated latex has much higher tensile strength than the best Acre Hevea. Yield very variable, but averages less than a quarter pound from felled trees. Cannot be tapped daily as Hevea.” He similarly states of M. siphonioides: “Latex pale yellow, seldom flowing freely, usually coagulating in the cuts in the manner reminiscent of some species of Castilloa [sic], being removed after an interval of about three days as scrap, which is wound into balls. Tensile strength excellent, being considerably better than Acre Fine Hevea.” Allen offered notes also on the latex of other species of Micrandra. Of M. Spruceana, he wrote: “Bark very thin... with scanty latex which coagulates with difficulty, producing a non- elastic gum.” Micrandra Sprucei, he annotated, had: “Latex yel- lowish, scant, coagulating to a non-elastic gum.” For M. Rossiana, he reported: “Latex very scanty, coagulating to a gummy non-elastic mass.” Although Allen’s notes suggest the possible exploitation along the Rio Papuri of M. minor and M. siphonioides, there is every indication that the other three spe- cies which he collected could not be worked and gave too little rubber or rubber of no commercial value. The latex of all species of Micrandra is white, or cream- coloured, except for some individuals of M. Rossiana, where it may often be yellowish. Ule’s statement (Ule, E. in Tropenflanz. Beih. 6 (1905) 1) that Micrandra milk may be distinguished from that of Hevea by its orange colour must be an error, for of the thousands of trees which I examined only a few had even a slightly yellowish latex. It is usually thick and slow-flowing. In taste, it is somewhat sweet, quite unlike the burning and bitter taste of most Hevea latex. It coagulates slowly on the trees, remaining tacky and resinous for weeks. The addition of river- water causes Micrandra latex to coagulate; whether or not this is due to some mineral or organic constituent present in this impure water is not known. Most rubber tappers whom I ques- 99 tioned maintain that Micrandra latex cannot be mixed with that of the rubber-yielding species of Hevea, the acid-coagulation of which it prevents. This unexpected effect is brought about like- wise if the latex of Hevea nitida Mart. ex Muell-Arg. be added to that of species, such as H. guianensis Aubl., which furnish good rubber (Schultes, R. E.: “The genus Hevea in Colombia” in Bot. Mus. Leafl., Harvard Univ. 12 (1945) 11; Seibert, R. J.: “A study of Hevea (with its economic aspects) in the Republic of Peru” in Ann. Mo. Bot. Gard. 34 (1947) 268). In quantity (and perhaps also in quality of rubber), latex varies from tree to tree, but we do not yet have precise data on this phenomenon, as we do for some species of Hevea. Micran- dra minor, growing along or very near water courses, yields much more latex than does M. siphonioides, an inhabitant of higher, well drained sandy savannahs situated well back from streams. The latex of the former species is thinner and freer flowing than that of the latter. For this reason, we might justifia- bly assume that, if Micrandra has ever been commercially ex- ploited, it was M. minor which was cut usually in preference to M. siphonioides, especially so since it is a much more abundant tree and much more accessible by canoe along the river banks. Little indeed is known about the rubber from Micrandra. According to Spruce, Micrandra yields “pure rubber.” Ule, like- wise, stated that Micrandra rubber is of good quality. Allen’s visual evaluation of the rubber of M. minor and M. siphonioides likewise suggests that it is of good quality. There is very little chemical information available on this rubber. Mr. A. V. McMullan of the United States Department of Agriculture reported (letter to R. E. Schultes, June 16, 1949) the following data after studying an air-dried specimen of rubber from Micrandra minor (Schultes et Murca Pires 9075a): “Resins (acetone extract) 3.78%; rubber hydrocarbon (benzene extract) 87.08%; insolubles 9.14%. Appeared to be somewhat softer and weaker than brasiliensis. This sample very difficult to enter solu- tion which may indicate a high polymer rubber. Merits more study.” Rubber from the leaves and bark of two trees of M. Lopezii were examined—Schultes et Lopez 9638: (bark) Resins 5.40%, rubber hydrocarbons 2.06%, “excellent rubber, clear, strong and elastic”; (leaves) resins 13.93%, rubber hydrocarbons 100 0.18%, “typical leaf rubber”. Schultes et Lopez 9663: (bark) res- ins 4.24%, rubber hydrocarbons 0.83%, “sticky and weak, not near the quality of above sample”; (leaves) resins 13.65%, rubber hydrocarbons, 0.21%, “typical leaf rubber”. For M. Spruceana (Schultes et Lopez 9641), the following results were recorded for bark rubber: Resins 3.92%, rubber hydrocarbons 1.31%; “poor, soft, sticky.” Micrandra Sprucei (Schultes et Lopez 9640) gave the following data: (bark) Resins 5.63%, rubber hydrocarbons 5.51%, “poor, soft and sticky”; (leaves) resins 16.84%, rubber hydrocarbons 0.16%, “typical leaf rubber.” An analysis of specimens sent in from Venezuela (probably by one of the expeditions the reports of which are quoted above) had a rubber content of 85.74 (Polhamus, L. G.: “Rubber con- tent of miscellaneous plants” U.S.D.A. Prod. Research Dept., No. 10 (1957) 22). The species from which the sample was taken is stated to have been Micrandra siphonioides, but this identifi- cation is open to some doubt. If the specimen were collected by either Fuller or Hansen who prepared the reports, its prov- enience was more probably Micrandra minor: the trees grew “in low land flooded annually” and “close to rivers... rarely at distances greater than 300-400 yards”: this is precisely the habi- tat of M. minor, not of M. siphonioides. A somewhat more complete chemical study appeared in 1956 (Wisniewski, A.: “Borrachas amazOnicas pouco conhecidas” in Bol. Tecn. Inst. Agron. Norte 31 (1956) 301) in which, nonethe- less, the author confessed to a lack of knowledge of Micrandra rubber in general. Wisniewski’s samples were air-dried. He had an average of five samples, and he compared Micrandra rubber to a piece of Acre Fina (the highest grade of rubber from Hevea brasiliensis) in the following summary: CR kg./cm.? AM Resin Micrandra 266 760 4.86 Acre Fina 210 805 2.53 TAXONOMIC AND FLORISTIC NOTES Micrandra elata (Didr.) Mueller-Argoviensis in Linnaea 34 (1865)) 142. Micrandra_ bracteosa Mueller-Argoviensis in Martius FI. Bras. 11 (1873) 290. Micrandra Glaziovii Pax in Engler Pflanzenr. iv, 147 1910 Euphorbiaceae-Jatropheae) 20. Micrandra brownsbergensis Lanjouw, Euphorbiaceae of Suri- nam (1931) 34, t. 7, 8. Micrandra santanderensis Croizat in Journ. Arn. Arb. 24 (1943) 169. BRAZIL: Sao Paulo, Mogy das Cruzes, in forest. 1876-77. A. Glaziou 1847a (M. Glaziovii TYPUS). Sao Paulo, Araracuara. Sep- tember 18, 1888. Loefgren 920. Bahia. Martius s.n. (M. bracteosae TYPUS). Rio Parana. July 1834. Riedel 23 (M. elatae TYPUS). COLOMBIA: Departamento de Santander, Barranca Bermeja, between Sogomoso and Carare Rivers. Alt. 100 m. October 9, 1936. O. Haught 2011. ~— Between Carare and Magdalena River, Puerto Berrio, Dorado Creek, 5 km. south of Raizubo. April 30, 1937. Haught 2189 (M. santanderensis TYPUS). FRENCH GUIANA: Montagne de Kaw. Alt. 250-270 m. “Tree 7 m. tall. Fls. pale green. Fruit green. Occasional in forest on bauxite.” December 14, 1954. R. S. Cowan 38815. PERU: Departamento del Loreto, Provincia de Maynas, Iquitos, Rio Nanay, Picuruyacu. Alt. c. 150-180 m. “Sandy upland, partially disturbed forest. 20 m. Fr. green. Sap milky.” September 24, 1975. S. McDaniel et M. Rimachi Y. 20204. —— Iquitos, Rio Nanay, Carreterado Picuruyacu. Alt. c. 150-180 m. “Upland. 10 m. Sap white. Fr. black. Shiringarana.” March 16, 1976. S. McDaniel, M. Rimachi Y. et J. Folsom 20533. SURINAM: Surinam River, Brownsberg Forest Reserve. Tree No. 1146. HBW No. 6687, October 31, 1924, FLORIS TYPUS; HBW No. 4267, February 24, 1919, FRUCTUS TYPUS. It is clear that these four binomials are referable to the widely distributed Micrandra elata. Although the species is most abundant along the coastal regions of Brazil and the Guianas, it has appeared in the Magdalena Valley of Colombia (from which locality is was described as a distinct species) and has recently been found far inland in the Estado do Para in Brazil and as far west as Iquitos, Peru. This representation of Micrandra in the Departamento de Santander in Colombia is the only record of 102 the genus west of the Andes; it is undoubtedly a coastal intrusion that proceeded inland up the Magdalena Valley. Micrandra bracteosa, described from material collected in Bahia, Brazil, was presumed to differ from M. e/ata by lacking glands at the base of the leaves or of having them only weakly developed and by having a panicle shorter than the leaves. Both are unreliable characters, and it seems advisable to treat M. bracteosa as a synonym of the widespread and variable M. elata. The distinguishing characters between Micrandra elata and the Surinam material described as M. brownsbergensis are of a minor and inconstant nature. Lanjouw did not distinguish M. brownsbergensis from M. elata when he described the concept. He did distinguish it from the very distinct M. siphonioides. Thanks to the very active Dutch collectors in Surinam, there is a wealth of herbarium material from the Brownsberg Forest Reserve. Careful examination of this ample material fails to provide one with any major character of distinction. The leaves are occasionally larger than is expected in M. e/ata and there are often fewer secondary nerves, but these and other fine differ- ences are too trivial to separate M. brownsbergensis even as a variety. Perhaps the most interesting collections of Micrandra are those made in 1936 and 1937 by O. Haught in Colombia and described by Croizat as M. santanderensis. When these collec- tions are viewed alone, they do look different; their distinction fades, however, when one has the whole series of collections from Colombia to southern Brazil along the coastal lowlands. Croizat stated that this species was distinct from all other known species because of its “differently colored foliage, but there are no floral differences. The conspicuously axillary tufts of hairs on the leaves of this new species are not found on M. elata Muell.- Arg. or... M. siphonioides Benth., to judge from the photo- graphs of the type specimens . . .” Haught 2189, the type of M. santanderensis, is, indeed, a good match for the type of M. brownsbergensis, although its leaves are somewhat larger; in Haught 2011, however, the leaves are very similar in size, shape and colouration to those of the type of M. brownsbergensis. We should point out the inexactness in several points of the 103 drawing of Micrandra elata provided by Pax (loc. cit., t. 5). The illustration which is published with this text has been prepared after careful study of ample material assembled from European, American and South American herbaria and represents, we believe, a more accurate and complete record of the species. The collections McDaniel et Rimachi Y. 20204 and McDa- niel, Rimachi Y. et Folsom 20533 are noteworthy as being the first from Peru. They represent also the material collected far within the Amazon—some 2000 miles upstream on the Ama- zon—of this species which is usually found nearer the coastal areas. Cowan 38815 represents the first collection of Micrandra elata from French Guiana. Micrandra Lopezii R. E. Schultes var. microcarpa R. E. Schultes var. nov. Arbor usque ad 25 pedes alta, a Micrandra Lopezii capsula semineque multo minoribus (valvis siccitate ad 20 mm. longis) principaliter differt. BRAZIL: Estado do Amazonas, Rio Alary Jabaru, Igana. “Caatinga. Arvore, 7 ms. 35 cms. Latex branco, coagulante rapido, espresso, forte e abundante.” November 8, 1947. Ricardo de Lemos Froes 21364 (YTypus in Herb. Instituto Agronomico do Norte 16829). In addition to the significant difference in size of the fruit (28 mm. long in dried condition in Micrandra Lopezii, 20 mm. or less in this new variety), the leaves appear to be generally subcu- neate instead of conspicuously rounded or even strongly cordate at the base. Micrandra Rossiana R. FE. Schultes in Bot. Mus. Leafl., Harvard Univ. 15 (1952) 211. COLOMBIA: Comisaria del Vaupés, Rio Papuri, vicinity of Monfort Mission. Alt. c. 200 m. “Slender, infrequent trees averaging 25~30 m. in height and 50 cm. in diameter. Bark grey, thin and hard and difficult to tap. Latex very scanty, coagulating to a gummy, non-elastic mass. Small tri-spermate capsule, resembling that of the arara-siringa [Micrandra siphonioides] but considerably elongated. Not well known, one or two individuals hesitat- ingly advancing either maha-wakpuh or buhawakpuh [Micrandra minor] but it is doubtful if either would be generally recognized as applicable to this species.” August 28, 1943. P. H. Allen 3109. 104 This collection of Allen was the first made in Colombian territory. The astuteness of Allen and his native informants in recognizing this as a different species is extraordinary. Later collections have shown that is is not a rare species in the north- west Amazon of Brazil and Colombia. Micrandra Rossiana has been known from Brazil, Colombia and Venezuela. Intensive studies of Micrandra have indicated that the distribution of M. Rossiana is much wider and more abundant in the Colombian Vaupés—especially in the Rio Apa- poris and its affluents—than in any other area of the northwest Amazon. Micrandra siphonioides Bentham in Hooker, Bot. Journ. 6 (1854) 371. COLOMBIA: Comisaria del Vaupeés, Rio Papuri, Macu-Parana. Alt. c. 200 m. “Trees averaging 25 m. in manchas but with numerous isolated specimens measuring 35 m. when felled. Average diameter about 75 cm above the often prominently developed buttresses, which in large specimens often extend upward for 3-4 m. from the ground. Flowers greenish yellow, on axillary new growth. Bark reddish brown, of varying thickness, that on the trunk often 2-3 cm. while on the buttresses it seldom exceeds | cm. Latex pale yellow, seldom flowing freely, usually coagulating in the cuts in the manner reminiscent of some species of Castilloa [sic], being removed after an interval of about three days as scrap, which is wound into balls. Tensile strength excellent, being considerably better than Acre Fine Hevea. Known locally as arara siringa or buhawakpuh (Tucano).” August 11, 1943. P. H. Allen 3050.—Same locality. “Tall trees, averaging 25 m. tall in manchas on margins of low sandy caatinga, but with numerous isolated specimens on clay ridges reaching 35 m. Average diameter about 75 cm. above the often prominently developed buttresses, which in old specimens often extend upward for 3-4 m. from the ground. Flowers greenish yellow, on axillary new growth. Bark reddish brown, of varying thickness, that on the trunk often 2-3 cm., while on the plank buttresses it seldom exceeds | cm. Latex pale yellow, seldom flowing freely, usually coagulating in the cuts in a manner reminiscent of Castilla. This scrap can be removed after an interval of about three days, usually being wound into balls. Tensile strength excellent, considerably better than weak fine Hevea. Some resin content. Known locally as arara-siringa (Geral) and bu-ha-wak-puh (Tucano). See pickled fruits, which resemble those of Hevea.” August 15, 1943. Allen 3061. Micrandra siphonioides is a very abundant species in the northwest Amazon. Seldom, however, have such detailed field notes been appended to voucher specimens. For this reason, it 105 seems advisable to quote the notes on Allen 3050 and 3061. The collection Allen 3061 has been identified erroneously as M. minor Benth. Micrandra Spruceana (Bai/llon) R. E. Schultes in Bot. Mus. Leafl., Harvard Univ. 15 (1952) 217. COLOMBIA: Comisaria del Vaupés, Rio Papuri, Santa Teresita Mission. Alt. c. 200 m. “Tall trees, averaging 35 m. in height and 80 cm. in diameter above the conspicuously developed, nearly unique stilt buttresses which are produced to a height of 3-4 m. as laterally compressed board-like flanges which act as ‘flying-buttresses’ often high enough to allow a man to stand beneath them. Leaves simple, with two basal disk-like glands. Inflorescen- ces of small green flowers from axillary new growth. Large tri-spermate seed capsule typical of Hevea. Bark very thin (6-8 mm.) with scanty latex which coagulates with difficulty, producing a non-elastic gum. Seeds collected for food. Known locally as wak-puh (Tucano).” August 15, 1943. P. H. Allen 3063. PERU: Departamento de Loreto, Provincia de Maynas, Rio Ampiyacu, Pebas and vicinity. April 10, 1977. 7. Plowman, R. E. Schultes et O. Tovar 6735, Puca Urquillo and vicinity. “Tree 65 m. tall with large buttresses, forming interwoven props at base. Growing in upland pri mary forest. Latex white, sparse. Fruit green. Seeds brilliant, smooth, red-brown.” April 27, 1977. Plowman, Schultes et Tovar 695]. Although a very abundant element of the flora of the north- west Amazon, little has been known about the latex of Micrandra Spruceana. Allen 3063 has an unusually detailed set of notes which have not hitherto been published. The two Peruvian collections appear to be the first from Peru. They were made during Phase VII of the Alpha Helix Amazon Expedition 1976-1977. Plowman, Schultes et Tovar 6951 is the voucher collection for Dragendorff spot test for alkaloids made in the Alpha Helix laboratory during the expedition: the species was found to be alkaloid-negative. Micrandra Sprucei (Mueill.-Arg.) R. E. Schultes in Bot. Mus. Leafl., Harvard Univ. 15 (1952) 218. COLOMBIA: Comisaria del Vaupés, Rio Papuri, vicinity of Piracuara Mission. Alt. c. 200 m. “Tall trees, slightly buttressed, averaging 30 m. in height and 60 cm. in diameter. Bark thin, averaging less than 1 cm. Latex yellowish, scant, coagulating to a non-elastic gum. Tri-spermate, Hevea- like capsule. Known locally as wak-so-ne-ne (Tucano).” August 18, 1943. P. H. Allen 3068. 106 Little has been known of this curious caatinga species of Micrandra. The collection Allen 3068 has been cited and its notes quoted in full because of their significance to our under- standing of the genus. VERNACULAR NAMES REPORTED FOR MICRANDRA arara-seringa Rio Vaupes, Brazil and Colombia M. minor; M. siphonioides arvore de mammora_ Rio Parana, Brazil M. elata bartabalie balli (Arawak) Dutch Guiana M. elata boo-ha-wa-puch (Tukano) Rio Vaupés, Colombia M. Rossiana bo-wapuch (Tukano) Rio Vaupes, Colombia M. Rossiana bu-ha-wak-puh (Tukano) Rio Papuri, Colombia M. Rossiana caucho Rio Caura, Venezuela M. minor caucho de guaco Territorio Federal, Venezuela M. minor and/ or M. siphonioides caucho de invierno Territorio Federal, Venezuela M. Minor and/or M. siphonioides caucho kunudi (Maquiretare) Rio Orinoco, Venezuela M. si- phonioides cauchorana’ Rio Solim6ées, Brazil M. minor; M. siphonioides caucho tomoro Rio Caura, Venezuela M. siphonioides cunuri Amazonas, Brazil and Colombia M. Spruceana cunuri da caatinga Amazonas, Brazil M. Sprucei efacone (Witoto?) Rio Igaraparana, Colombia M. minor huermega (Witoto) Rio Igaraparana, Colombia M. minor ka-ro-a (Yukuna) Rio Miritiparana, Colombia M. siphoni- oides koedi biose balli (Arawak) Dutch Guiana M. elata ko-no-ko (Mirana) Rio Caqueta, Colombia M. Spruceana ma-ha-puch (Tukano) Rio Vaupes, Colombia M. minor ma-ha‘puh (Tukano) Rio Vaupes, Colombia M. minor ma-ha-ree (Taiwano) Rio Kananari, Colombia M. minor ma-ha-wa-he (“macaw seed”) (Makuna) Rio Piraparana, Colombia M. minor ma-ha-wa-ho (“macaw seed”) (Barasana) Rio Piraparana, Colombia M. minor 107 ma-ha-w6-he (“macaw seed”) (Makuna) Rio Popeyaca, Colombia M. siphonioides ma-ha-wa-po-k6 (Barasana) Rio Piraparana, Colombia M. Rossiana ma-ha-wa-p6-k6 (Gwanano) Rio Vaupes, Colombia ™. Rossiana man-je-ka (man = macaw) (Kubeo) Rio Vaupés, Colombia M. minor ma-poo-a (Tanimuka) Rio Miritiparana, Colombia M. sipho- nioides moereidam (Karib) Dutch Guiana M. elata momofi (Waika) Territorio do Roraima, Brazil M. siphoni- oides 6-bai-me-ko (seringa de la sabola) (Mirana) Rio Miritiparana, Colombia M. siphonioides pai-re (Puinave) Rio Apaporis, Colombia H. minor pai-root (Puinave) Rio Apaporis, Colombia H. minor pen (Maku) Rio Piraparana, Colombia M. Spruceana py-root (Puinave) Rio Apaporis, Colombia M. minor seringa irari Rio Negro, Brazil M. minor and/or M. siphoni- oides seringarana Rio Solimoes, Brazil; Rio Amazonas, Peru M. minor, M. siphonioides; M. elata suru-wali-yek Bolivar, Venezuela M. glabra topoeloe alome (Karib) Dutch Guiana M. elata wa-ho (Makuna) Rio Piraparana, Colombia M. Spruceana wah-puch (Tukano) Rio Vaupes, Colombia M. Spruceana wakati-erwicheri (Bora?) Rio Igaraparana, Colombia M. minor wa-so-ne-ne (Tukano) Rio Vaupes, Colombia M. Sprucei was-so-roo-re (Gwanano) Rio Vaupes, Colombia M. Sprucei wer-meger (Witoto) Rio Igaraparana, Colombia M. minor; M siphonioides woo-sheen (Puinave) Rio Apaporis, Colombia M. minor ye-cha (Yukuna) Rio Miritiparana, Colombia M. Spruceana ye-ka (Kuripako) Rio Guainia, Colombia M. Spruceana 108 PLATE 23 Plate 23. Flowering branch of Micrandra siphonioides. Rio Apaporis, Comisaria del Vaupés, Colombia. Photograph: R. E. Schultes. 109 PLATE 24 7 3 x ’ Plate 24. Extensively buttressed roots are typical of Micrandra Spruceana. La Pedrera, Comisaria del Amazonas, Colombia. Photograph by R.E. Schultes. 110 PLATE,25 MICRANDRA elata _ (Didrichs) Muell-Arg. * 111 BOTANICAL MUSEUM LEAFLETS HARVARD UNIVERSITY CAMBRIDGE, MASSACHUSETTS, May-JuNE 1979 . VoL. 27, Nos. 5-6 ANCIENT GOLD PECTORALS FROM COLOMBIA: MUSHROOM EFFIGIES?! RICHARD EVANS SCHULTES AND ALEC BRIGHT One of the most fascinating and enigmatic archaeological objects in the Americas is a certain type of anthropomorphic gold pectoral found in southern Panama and, most especially, in Colombia. Called “Darien pectorals,” these ornaments are not confined to one region, although their greatest concentration seems to be in the Sinu country in northwestern Colombia, near the border with Panama’s Darien Province. They are found also in the Quimbaya region of Colombia, farther south. The dating of the Colombian gold objects and the styles of these objects are still rather indefinite, although archaeologists would generally place Sinu and Quimbaya goldwork in the late pre-Columbian centuries, most likely in the span of 1,000—1,500 A.D., but with the possibility of beginnings as early as 500 A.D. Interestingly, one such “Darien pectoral” has been found as far north as Chichen Itza in Yucatan, where it undoubtedly found its way as an item of long-distance trade, along with other lower Central American and Colombian gold artifacts. The Maya centre of Chichen Itza—and its famed cenote of sacrifice, 'This paper is published in English in the Botanical Museum Leaflets of Harvard University and in Spanish in the Boletin Museo del Oro (Banco de la Republica) Ano 3, Bogota, Colombia. Botanical Museum Leaflets (USPS 404-990). Published monthly except during July and August by the Botanical Museum, Harvard University, Cambridge, Massachusetts 02138. Subscription: $25.00 a year, net, postpaid. Orders should be directed to Secretary of Publications at the above address. Second-Class Postage Paid at Boston, Massachusetts. where such gold objects were thrown as offerings—was particu- larly active between 1,000 and 1,250 A.D. This, however, does not help in placing the manufacturing date of the object in ques- tion, as many of the other objects found in the Chichen Itza cenote were clearly heirloom pieces (Willey, pers. comm.). Although they vary slightly, these ornaments all follow a gen- eral plan. They are anthropomorphic, usually highly stylized. The most prominent feature is the pair of dome-shaped or rounded objects arranged side by side on the head. Lateral wing- like ornaments with spiral decorations made up of double spirals almost invariably frame the head of the pectoral. A flat face or mask, sometimes more or less natural but usually with complex filigree ornamentation, is discernible. Arms and hands hold two sticks or wands usually in an inverted V-shape. A frog or toad, sometimes very natural but usually extremely stylized, is almost always present immediately beneath the face: that is, on the chest. These pectorals have been divided into two general types: “Darien pectorals” and “Darien-related pectorals.” The former are the more typical, with most of the principal diagnostic fea- tures; the latter have very stylistic variations and fewer of the main diagnostic features, diverting in one or several ways from the basic morphological pattern of the Darien type. Although a few specimens are to be found in private collec- tions and in several museums, the greatest concentration of these gold pectorals is preserved in the Museo del Oro in Bogota, Colombia. Thanks to the director, Dr. Luis Duque Gomez, we have had the opportunity of examining in great detail the Muse- um’s collection of more than 150 specimens and of conferring with Mrs. Ana Maria Falchetti de Saenz, whose meticulous research is embodied in her thesis entitled The Goldwork of the Sinu Region, (Northern) Colombia (Falchetti-Saenz, 1976). Because of the two dome-like objects on the head, these pec- torals have popularly been called “telephone-bell gods.” This term originated apparently from the description given by Dr. Jose Perez de Barrada in 1954, when he mentioned “the semi- spherical buttons to which I have referred that remind one of the bells of old fashioned telephones or of a pair of mushrooms” (Perez, 1954). 114 ae Following the intense ethnomycological activities of Dr. R. Gordon Wasson (Wasson, 1962, 1968, 1972, 1979, 1980), his late wife Dr. Valentina P. Wasson (Wasson and Wasson, 1957), and the late Professor Roger Heim (Heim, 1967, 1978; Heim and Wasson, 1958) unveiling the ancient and the contemporary reli- gious use of hallucinogenic mushrooms in southern Mexico, Pérez de Barrada wrote: “It would not be strange to reconsider with great reserve this casual attribution. It should be noted that these semi-spherical buttons are not fixed directly to the head but are attached by means of filaments soldered to the back of the piece... We know nothing about the ritual use of mush- rooms amongst the Indians of Darien at the time of the discov- ery nor later, but if we remember, but keep it in mind, since no trace was found in the indigenous pharmacopoea of the Catios of the Golfo de Uraba, notwithstanding the excellent mono- graph by Father Severino Santa Teresa. On the other hand, the secrecy with which these Indians guard their knowledge of the properties of plants and their shamanistic ceremonies could have hidden a possible use of hallucinogenic mushrooms—a use which might be very ancient and which possibly existed in dis- tinct forms. The bridge between Guatemala and Darien is diffi- cult to establish but easy to suspect. Our suggestion that these buttons represent mushrooms is accepted by A. Emmerich” (Perez, 1954). This reference to Emmerich leads us directly to the second mention in the literature that these dome-like objects represent mushrooms. “It appears likely that the puzzling, hitherto uni- dentified hemi-spherical head dress ornaments in fact represent a pair of mushrooms, probably of hallucinogenic properties. It is significant that such mushroms are to this day traditionally counted, ceremonially used and consumed in pairs . . . the mush- room head dress ornaments were hammered out separately, riveted to short stems and then soldered to the body” (Emme- rich, 1965).? 2In a note, Emmerich states: “I am indebted to Mrs. Mary U. Light for her original insight in identifying these ornaments as mushrooms.” 115 A third consideration of the “telephone bell gods” as mush- room effigies was offered in 1974 by Professor Peter T. Furst. Referring to Emmerich, he reported: “Andre Emmerich developed the interesting theory that the pairs of telephone bell-like, semi-spherical, hollow-stemmed ornaments surmounting the head dress of a certain class of con- ventionalized anthropomorphic gold pectorals in the Darién style from Colombia (“telephone gods”) are in fact mushrooms. Emmerich demonstrated convincingly that over time these orna- ments gradually changed position as the effigies themselves became more and more stylized. On early, more realistic pieces, the mushroom form is unmistakable, the semi-spherical caps being separated from the head dress by stems or stipes attached to the top of the head. Subsequently, the stipes became shorter and the caps were slightly inclined forward. Eventually, the stipes, though still present beneath the cap, disappeared alto- gether from view and the two caps faced forward like a pair of female breasts. By this time the human characters had also been stylized to the point of abstraction” (Furst, 1974, 1976). Hf. Our own studies of the many gold pectorals in the Museo del Oro and our familiarity with the complexities of magico- religious, shamanic or ceremonial use of hallucinogenic plants, together with consideration of the natural range of psilocybine- containing genera of mushrooms in the New World, lead us to the belief that this identification of the dome-shaped head dress ornaments is indeed correct and that, further, they strongly sug- gest the religious use in prehispanic Colombia of intoxicating mushrooms. This interpretation of the gold pectorals has already twice been supported (Schultes and Hofmann, 1979, 1980). No other explanation of the possible significance of these domes has ever, so far as we know, been advanced. Significance they most certainly must have had. We are left, then, with the inescapable conclusion that they cannot represent anything else than mushroms. In a number of the pectorals, the domes are elevated on a stipe. Furthermore, a few of the domes have a mammiform tip 116 characteristic of some species of Psilocybe, and several have a design along the margin of the cap which could be interpreted as indicative of the scalloped edge of the cap of Panaeolus sphinc- trinus. We have, in addition, several tangential arguments which have not hitherto been offered and which, we believe, lend weight to this interpretation. As with the intoxications induced by many hallucinogens, levitation—the sensation of flying or soaring through the air and visiting distant places—is a commonly experienced psychic effect of psilocybine, the principal active constituent of species of Panaeolus, Psilocybe and Stropharia (Schultes and Hof- mann, 1979; Brown, 1972). In her long and involved chant during the Mazatec mushroom ceremony, for example, the famous shaman Maria Sabina repeatedly sings such descriptive phrases as “Whirling woman am I,” “Look, I feel as though I’m going to the sky,” “Woman like the big eagle am I,” “Woman of space am I” (Halifax, 1979). Dr. Albert Hofmann, the chemist who first isolated psilocybine and psilocine from the sacred mushrooms and who elucidated their structures and synthesized them, mentions levitation amongst other symptoms produced by small doses of psilocy- bine: “. . . bodily relaxation and detachment from the environ- ment .. . effects . . . associated with a pleasant feeling of extraordinary lightness, a bodily hovering” (Schultes and Hof- mann, 1979, 1980). Wasson, the first to provide a detailed de- scription of psilocybine-mushroom intoxication, reported spe- cifically, amongst many other effects, the experience of levita- tion: “. . . the bemushroomed person is poised in space, a disembodied eye, invisible, incorporeal, seeing but not seen... your body lies in darkness, heavy as lead, but your spirit seems to soar. . . and with the speed of thought to travel where it listeth, in time and space . . .” (Schultes, 1973). The Colombian gold pectorals almost invariably have two wings formed by spirals and arising at an angle lateral to the head dress ornaments. They vary somewhat in shape but always involve spiral filagree work. While occasionally abbreviated, they are usually elongated. We believe that these represent wings, the wings of a mythical bird, and are directly associated with levitation. Furthermore, we need not detail how frequently 117 and how generally, not only in the Americas but in the Old World as well, are birds associated with hallucinogens, but sev- eral examples will suffice. Amongst the Koryaks of Siberia, the culture hero, Big Raven, discovered the hallucinogenic mushroom Amanita muscaria from the spittle of the god Vihiyinin (Schultes and Hofmann, 1980). The mythical Thunder Bird carries the prayers of the peyote-eating Indians of the United States to heaven (LaBarre, 1938), and levitation is important in the Huichol peyote ritual in Mexico (Furst and Anguiano, 1977). In eastern Brazil, the Indi- ans who drank vinho de jurema (Mimosa hostilis) flew all night carried on the back of a huge bird that skirted thundering rapids and showed its passengers the abodes of the dead (Goncalves de Lima, 1946). In several “Darien-related pectorals,” the hands hold a bar on which are perched four birds. In one of the pectorals, the birds are movable. We suggest that, in these examples, too, the avian ornamentation indicates association with the flying sensation experienced in mushroom intoxication. But there is an even more compelling argument for the hallu- cinogenic connection of the pectorals: the frog or toad. Almost all of the pectorals are ornamented with these amphibian fig- ures. In some cases, the figure is realistic; in others, it is flat but easily discernible—with eyes, legs and a median dorsal band, indicating undoubtedly the coloured band on some of these animals. In most, however, it is highly stylized, their eyes and legs represented by circles of double spirals, the tail portion indicated by a flat triangular projection, and the snout repre- sented occasionally by a knob-like protuberance. This extraordinarily constant association of the frog or toad with the pectorals would seem to have deep significance. No other animals represent change and transition more sharply with their dramatic metamorphosis and fertility passing from the egg to a wholly water-living, gill-breathing creature resembling a fish to a terrestrial, four-legged amphibian. Furthermore, cer- tain frogs of the Dendrobatidae are frighteningly toxic—one species producing the most highly poisonous substance known (Daly and Myers, 1967; Daly and Witkop, 1971). For millenia and in widely separated parts of the globe, frogs and toads have 118 been associated with origin myths, mysticism, rain and fecun- dity, the moon, magic and, especially, with intoxication from hallucinogenic agents (Wasson, 1980). This significance of the toad-frog motif has been emphasized by Furst (Furst, 1974, 1976) who succinctly states: “. . . there is clearly much more than only the ‘obvious’ connection with rain to account for the impor- tance of the toad-frog motif in the indigenous symbolic system aa Purse, 1979). Hallucinogenic constituents found in plants employed for their psychoactive properties have been isolated from frogs and toads. The alkaloid bufotenine, known from the leguminous tree Anadenanthera peregrina from which a highly psychoactive snuff is prepared in South America (Holmstedt and Lindgren, 1967), is present in the skin glands of Bufo marinus and other amphibians (Schultes and Holmstedt, 1968). The related and more potent hallucinogen, 5-methoxy-N, N-dimethyltryptamine, one of the active components of the snuff prepared in South America from several species of Virola trees (Schultes and Holmstedt, 1968), has recently been found in the North Ameri- can desert toad, Bufo alvarius (Furst, 1974). Extremely toxic substances occur in the skin of some species of Phyllobates and Dendrobates, colourful amphibians of northwesternmost South America in the general region where the gold pectorals are found in greatest concentration. The venoms of some South American frogs and toads are employed in magical contexts, sometimes producing even ecstatic or hallucinogenic states. It was the Swedish anthropologist S. Henry Wassen who, many years ago, reviewed the literature and concluded “. . . that the ubiquitous frog/toad motif in South American mythology and art, includ- ing the great number of effigies of cast gold from prehispanic Colombia and Panama, was inseparable from the practical use of frog venom for blowgun dart poison (which in any event had a magical component) and from the widespread magico- religious beliefs and practices involving the toxins of different species of these amphibians” (Wassén, 1934). Frog poison— probably from a species of Phyllobates or Dendrobates—is used by hunters among the Amahuaca Indians of Amazonian Peru for inducing hallucinations: the poison Is rubbed on self-inflicted cuts, inducing a violent illness, characterized by vomiting, diar- 119 rhoea, convulsions, unconsciousness. Supernatural expertness in the hunt results from the hallucinations which are interpreted as communication with forest spirits (Furst, 1974). There can no longer be any doubt that the high place occupied in magico-religious spheres by frogs and toads must be attrib- uted in great part to the toxic properties of some species. Although the potently poisonous South American species can- not be termed hallucinogenic in the usual sense of that word, their toxins do act upon the central nervous system with effects so unreal as to induce the Indian to ascribe supernatural powers to the animal and actually visual and other hallucinations may indeed accompany the violent intoxications caused by agents that can in no wise be considered true hallucinogens (Furst, 1974, 1976). For, as has been correctly pointed out: “. . . the massive assault on the system brought on by _ bufotenine- containing Bufo venom is of a very different order than the shift from one state of consciousness to another triggered by bufo- tenine-containing snuff” (Furst, 1974). Perhaps it may not be wholly coincidental that toads were so frequently added as an ingredient of the hallucinogenic witches brews of medieval Europe. Whether as hallucinogen-inducing organisms, or as poisonous animals causing what might be termed pseudohallucination syn- dromes, these amphibians assumed —for this and other peculiari- ties—a significance in aboriginal mythology and magic, an xalted position amongst the peoples who created the Colombian gold work. In almost every gold-working culture of Colombia, there are numerous examples of the toad-frog motif. There are hundreds of specimens in the Museo del Oro in Bogota. These are espe- cially abundant in the Tairona area of the Sierra Nevada de Santa Marta. From the surviving aboriginal groups of Indians still living in this region, it is known that the frog is considered a mythological being at the center of the cosmos (Reichel- Dolmatoff, 1963). We believe that, especially when other characteristics of the gold pectorals (wings, birds, mushrooms) are taken into account, the constancy of the frog-toad motif on these artifacts must of necessity be interpreted as an indication that they played some 120 role in a magico-religious system based on the hallucinatory experience. That no reference to the hallucinogenic use of mush- rooms amongst the Indians of Colombia is found in the writings of the Spanish conquerors hardly argues against the possibility of such employment: the use and the cults connected with the mushrooms may well have died out between the dates of the latest gold pectorals and the 16th Century. Ly; It would, of course, be a futile exercise to presume that the gold pectorals in question—or any other artifact, for that matter—represent the use of hallucinogenic mushrooms, unless mushrooms possessing psychotomimetic constituents actually do occur in the region where the artifacts were made and used. Although the collection and study of mushrooms in Colombia is still in a very preliminary stage, psilocybine-containing species have been reported. Species of Psilocybe are known to be widely distributed in the world, and the field studies of Dr. Gaston Guzman in 1964 and 1971 have indicated that hallucinogenic species of Psilocybe occur in Colombia (Guzman and Varela, 1978). The localities are widely distributed throughout the nation and vary from the warm lowlands to paramos at high elevation. According to Guzman, the following species of Psilo- cybe have been registered from Colombia: P. antillarum (Fr.) Quel., P. argentina (Speg.) Sing., P. bullacea (Bull. ex Fr.) Kumm., P. columbia Guzman, P. Pintonii Guzman and P. sub- cubensis Guzman (which is occasionally considered to be a var- iant of Stropharia cubensis). A number of these mushrooms are presumably psilocybine-containing, since they are cyanescent and have a farinaceous odor and taste—indicative, according to Guzman’s experience, that they are hallucinogenic. Further- more, Panaeolus sphinctrinus (Fr.) Bresadola, known as one of the hallucinogenic mushrooms of Mexico, has been collected in Colombia (Guzman and Varela, 1978). Guzman writes in a letter: “I agree with you that the South American Indians used hallucinogenic species of Psilocybe. | reported 30 species from South America, but I think that there are more, but we need more explorations. Even | think that the 121 Indians from the Atlantic zone and not only those of the Andes used the hallucinogenic mushrooms.” The recent field work of Dr. Kenneth Dumont has resulted in the registration from Colombia of other species of Psilocybe, a number of which may likewise be psilocybine-containing (Dumont, pers. comm.). V. Our studies of the Darien and Darien-related gold pectorals of Colombia have strengthened our belief that mushrooms perhaps enjoyed a widespread magico-religious place in aboriginal cul- tures from Mexico, through Central America and in the Andes south to Peru. There are many pieces of evidence that may be interpreted as supporting such a belief. One manifestation of this cultural trait is provided by the numerous “mushroom stones” of Mexico, Guatemala and El Salvador, which have been extensively stu- died and documented (Borhegyi, 1957, 1961; Wasson, 1957: Rose, 1977). Four of these stone figures significantly have a frog-figure at their base (Wasson, 1980). Further south in Mex- ico and extending southeastward to El Salvador are various ceramic representations which have been interpreted as mush- room-related artifacts. From Colombia, there is one ceramic object from the area of greatest concentration of gold Darien pectorals which, although far less convincing than the gold objects, might be interpreted as representing a mushroom, since it has an undulating “cap” which occurs in some species of Psilo- cybe. There are, furthermore, numerous clay artifacts from Mexico with representations of mushrooms in contexts sugges- tive of their significance in magico-religious rituals (Borhegyi, 1963; Furst, 1974), We might also mention at this point that several unusual “mushroom-like” ceramic pieces have been excavated at Arara- cuara in the Colombian Amazon (Herrera, pers. comm.). Further to the south in the Andes, ceramic pots in the form of a human head with a mushroom-like protuberance from the forehead are frequent in the Mochica. This protuberance could not function as a handle: if the pot were filled with liquid, it 122 would be too heavy to be supported by such a brittle handle. To the best of our knowledge, no functional explanation of this type of protuberance has been offered. They are all in shape very suggestive of mushrooms. One of the pots figured actually has painted flecks on the cap which might lead one to suspect that it represents Amanita muscaria, even though the species is believed not to have existed this far south in pre-Columbian times. These and other pottery artifacts, considered in detail by Furst (Furst, 1974), tend to support the belief that mushrooms were impor- tant in pre-Columbian art in more than one locality in the Andes of South America. VI. It has only recently been discovered that the fly agaric Aman- ita muscaria, has deep-rooted religious significance and is still ceremonically used by the Ojibway Indians living on Lake Superior in the United States (Wasson, 1979) and that Indians of the Mackensie River area of British Columbia, Canada, sim- ilarly employ this hallucinogen (Halifax, pers. comm.). It is possibly significant that the Mackensie River area, being largely glacier-free in the Pleistocene, may have been one of the main routes taken by early man on his migrations from Siberia to the Americas. Perhaps it may not be superfluous, in closing, to point out that from Asia ancient mythological ideas stemming from the use of hallucinogenic mushrooms and their concomitant associ- ations are traceable in European witchcraft, in the cult of soma in the Indian subcontinent, in the use of fly agaric in Siberia and at least in two contemporary and unrelated and widely separated indigenous groups in North America. In view of the recognized widespread magico-religious use of mushrooms in the New World, we believe that the interpretation of the Colombian “telephone bell gods,” as mushroom-related artifacts, represents perhaps the most plausible explanation thus far advanced and that it may be of the utmost importance in our studies of the role of hallucinogens in aboriginal societies of the New World. PX) & %q@San Salyodor SX ANTILLES °97Crooked Isle So aix MEXICO eracru. © Chichen Itza 2 a a! © Pp BELICE GUATEMALA HONDURAS COSTA RICA PANAMA eLambayeque Cajamarca PERU BOLIVIA PREHISPANIC GOLD WORK IN THE AMERICAS — 2000 B.C. 1500 A.D. Compania Bajate ARGENTINA 124 ACKNOWLEDGMENTS Permission to publish plates 23 to 33 has graciously been given by Dr. Luis Duque-Gomez, Director, Museo del Oro, Bogota, Colombia. We gratefully acknowledge permission to publish plate 34 which has been granted by the Cleveland Museum of Art, Cleveland, Ohio (gift of Mr. and Mrs. R. Henry Norweb). We thank the director of the Peabody Museum, Har- vard University, Prof. Clifford Lamberg-Karlovsky, for permis- sion to reproduce plates 35 and 36 (Photograph by Hillel Burger). REFERENCES Borhegyi, S. F. De, “Mushroom stones of Middle America. A geographically and chronologically arranged distribution chart” in V. P. and R. G. Wasson, Mushrooms, Russia and History. Pantheon, New York 2 (1957) Appendix. Borhegyi, S. F. De, “Miniature mushroom stones from Guatemala” in Am. Antiqu. 26(1961) 498-504. Borhegyi, S. F. De, “Pre-Columbian pottery mushrooms from Meso- america” in Am. Antiqu. 28(1963) 328 338. Brown, F. C. Hallucinogenic Drugs. Charles C. Thomas, Springfield, III. (1972) 80. Daly, J. W. and C. W. Myers, “Toxicity of Panamanian poison frogs (Den- drobates): some biological and chemical aspects” in Science 156(1967) 970 1973. Daly, J. W. and B. Witkop, “Chemistry and pharmacology of frog venoms” in W. Bucheri, E. E. Buckley and V. Deulofeu (Eds.) Venomous Animals and their Venoms. Academic Press, New York (1971) 497-519. Dumont, K. Personal communication. Emmerich, A. Sweat of the Sun and Tears of the Moon. University of Washington Press, Seattle, Wash. (1965). Falchetti de Saenz, A. M. The Goldwork of the Sinu Region, (Northern) Colombia. University of London (Unpubl. Thesis) (1976). Falchetti de Saenz, A. M., “Colgantes ‘Darien’” in Bol. Mus. Oro 2(1979) 1- aa: Furst, P. T., “Hallucinogens in pre-Columbian art” in Spec. Publ. Mus. Texas Technical University 7(1974) 55-102. Furst. P. T... Hallucinogens and Culture. Chandler and Sharp Publishers, San Francisco (1976). Furst, P. T. and M. Anguiano, “‘To fly as birds’: myth and ritual as agents of enculturation among Huichol Indians of Mexico” in J. Wilbert (Ed.) Enculturation in Latin America: an Anthology. UCLA Latin American Center Publications, University of California, Los Angeles (1977) 95-181. Goncalves de Lima, O., “Observagoes sObre o vinho da jurema utilizado pelos indios Pancaru de Tacaratu (Pernambuco)” in Arqu. Inst. Pesquisas Agron. 4(1946) 45-80. 12 Guzman, G. and L. Varela, “Hongos de Colombia III” in Caldasia 12(1978) 309-338. Guzman-Huerta, G., Estudio Taxonémico vy Ecologico de los Hongos Neutropicos Mexicanos. Tésis Profesional, Instituto Politécnico Nacion- al, Ciencias Bioldgicas, Mexico (1959). Halifax, J... Shamanic Voices. E. P. Dutton, New York (1979). Halifax, J., Personal communication. Harner, M. J... “Common themes in South American Indian yajé experien- ces” in M. J. Harner (Ed.) Hallucinogens and Shamanism. Oxford University Press, New York (1973) 155-175, Heim, R., Nouvelles Investigations sur les Champignons Hallucinogenes. Museum National d’Histoire Naturelle, Paris (1967). Heim, R., Les Champignons Toxiques et Hallucinogénes (Ed. 2) Société Nouvelle des Editions Boubée, Paris (1978). Heim, R. and R. G. Wasson, Les Champignons Hallucinogenes du Mexique. Museum National d’Histoire Naturelle, Paris (1958). Herrera, W. Personal communication. Hofmann, A., LSD-Mein Sorgenkind. Klett-Cotta, Stuttgart (1979). Holmstedt, B. and J.-E. Lindgren, “Chemical constituents and pharma- cology of South American snuffs” in D. H. Efron (Ed.) Ethnopharma- cologic Search for Psychoactive Drugs, Public Health Service Publ. No. 1645, U. S. Government Printing Office, Washington, D. C. (1967) 339- 373. LaBarre, W., The Peyote Cult. Yale University Publications in Anthro- pology, New Haven, Conn. (1938). Perez de Barradas, J... Orfebreria Prehispanica de Colombia: Estilo Calima. Graficos Jura, Madrid (1954). Reichel-Dolmatoff, G., “Apuntes etnograficos sobre los indios del alto Rio Sinu” in Rev. Acad. Col. Cienc. Exact., Fis. Nat. 12, no. 45(1963) 29-40, Rose, R. M., Mushroom Stones of Mesoamerica. Unpubl. Ph.D. Thesis, Harvard University, Cambridge, Mass. (1977). Schultes, R.E., “An overview of hallucinogens in the Western Hemisphere” in P. T. Furst (Ed.) Flesh of the Gods. Praeger, New York (1972) 28. Schultes, R. E., Hallucinogenic Plants. Golden Press. New York (1976). Schultes, R. E. and A. Hofmann, The Botany and Chemistry of Hallucino- gens. Ed. 2 Charles C. Thomas, Springfield, Ill. (1979). Schultes, R. E. and A. Hofmann, Plants of the Gods: Origins of Hallucino- genic Use. McGraw Hill Co., New York (1980). Schultes, R. E. and B. Holmstedt, “De plantis toxicariis e Mundo Novo tropicale commentationes II. The vegetal ingredients of the myristica- ceous snuffs of the northwest Amazon” in Rhodora 70 (1968) 113-160. Wassén, S. H., “The frog-motive among South American Indians” in Anthro- pos 29(1934) 319-370. “Part Il. The frog in Indian mythology and imaginative world” Ibid., 613-658. Wasson, R.G., “The hallucinogenic mushrooms of Mexico and psilocybine: a bibliography” in Bot. Mus. Leafl. Harvard Univ. 20(1962) 25-73. Wasson, R. G., Soma-Divine Mushroom of Immortality. Harcourt, Brace and World, New York (1968). Wasson, R. G., “The divine mushroom of immortality” in P. T. Furst (Ed.) Flesh of the Gods. Praeger Publishers, New York (1972) 185-200. 126 Wasson, R.G., “Traditional use in North America of Amanita muscaria for divinatory purposes” in Journ. Psyched. Drugs 11(1979) 25-28. Wasson, R. G., The Wondrous Mushroom: Mycolatry in Mesoamerica. McGraw Hill, New York (1980). Wasson, V. P. and R. G. Wasson, Mushrooms, Russia and History. Pantheon, New York (1957). Willey, G. R., Personal communication. 127 PLATE 26 Plate 26-29 Colombian gold “mushroom” pectorals showing more or less realistic amphibian forms. Museo del Oro, Bogota. PLATE 27 129 PLATE 28 130 PLATE 29 ioe) PLATE 30 } § of i} . i i i ; Plate 30-31 Colombian gold “mushroom” pectorals showing highly stylized amphibian forms. Museo del Oro, Bogota. 132 PEALE.) 133 PLATE 32 Oa, : *?% 3 Plate 32-33 Colombian gold “mushroom” pectorals showing four avian figures. Museo del Oro, Bogota. 134 PLATE 33 PLATE 34 Plate 34-35. Views of the back of Colombian gold “mushroom” pectorals showing “stipes” of the mushroom-like domes. Museo del Oro, Bogota. PLATE 35 PLATE 36 Plate 36 Two of the rare Colombian gold “mushroom” pectorals showing three (instead of two) dome-shaped objects at the top. Museo del Oro, Bogota. 138 PLATE 37 Plate 37. Seated man, Colombian Quimbaya Culture, gold. Note the two mushroom-like objects on the head and the two bird (toucan?) beaks in place of legs. PLATE 38 Plate 38 Pottery vessel in form of human head. Viru and Chicamac Valley, Peru, Early Intermediate Period (AD 200-500), Mochica IV. Note the mush- room-like object on the left side of the forehead and the trance-like expression of the face. 140 PLATE 39 Plate 39 Pottery vessel in form of human head. Viru and Chicamac Valley, Peru. Early Intermediate Period (AD 200-500), Mochica IV. Note the mush- room-like object at centre of the forehead. 14] BOTANICAL MUSEUM LEAFLETS VoL. 27, No. 5-6 May-JUNE 1979. DISCOVERY OF AN ANCIENT GUAYUSA PLANTATION IN COLOMBIA RICHARD EVANS SCHULTES I. Ilex Guayusa Loes.* represents one of the caffeine yielding species of /lex which is used, and has been from early times, asa stimulant, emetic and medicine (Schultes, 1972). It is also one of the most poorly understood species of the genus. Loesener described //lex Guayusa from sterile material as a “species nova atque dubia” (Loesener, 1901). Flowering material was not available to him in 1901. Nor has any flowering mate- rial, to the best of my knowledge, been collected until very recently. Notwithstanding the uncertainty that Loesener himself expressed concerning the validity of his species, //lex Guayusa has always been accepted as a “good” species. Actually, there have been very few specimens collected of this locally important plant. It is now apparently native to eastern Ecuador and adjacent Peru and Colombia. Although the Eng- lish plant-explorer Richard Spruce was thoroughly familiar with guayusa, he seems never to have collected it during his assiduous collecting in Amazonian Peru and Ecuador 120 years ago (Schultes, 1978; Spruce, 1901). There is in the herbarium at Kew an unidentified collection of Ilex made during the last century in Guilaguiza and Zamora, Ecuador (E. C. Lehmann 5581). 1 have studied the specimen, which has a few remnants of dried fruits still adhering to the twigs, and I believe that, while it may possibly represent Ilex Guayusa, it can be so determined only with strong reservation. In 1939, Dr. Erik Asplund of the Riksmuseet in Stockholm made an excellent collection of /lex Guayusa in Tena, Provincia Napo-Pastaza, Ecuador (£. Asplund 9485). This collection is sterile. 143 Several significant collections, all from cultivated sources, were made by my former student, Dr. Homer V. Pinkley, in 1966 in eastern Ecuador (H. V. Pinkley 199, 454, 455, 456) at Dureno, Rio Aguarico; Puerto Napo; between Tena and Archid- ona, Guayura, respectively. All are sterile. Notes on these collec- tions, preserved in the Economic Herbarium of Oakes Ames in the Botanical Museum of Harvard University, state that the leaves are prepared as a tea drunk as a “health tonic” and amongst the Jivaro as an “emetic.” Il. lex Guayuisa has, until recently, not been known from Colombia. Several collections are now available, however, and indicate that this tree is recognized and used medicinally and does grow in the lowland (more or less 2100 feet altitude) areas of the Putumayo on the eastern slopes of the Andes near the Ecuadorian frontier. COLOMBIA: Comisaria del Putumayo, Valley of Sibundoy. “Obtained at a distance by a Sibundoy medicine man. Medicinal. Guayusa.” October 31, 1962. M. L. Bristol 352-A2. Comisaria del Putumayo, Pepino, near Mocoa. Alt. 680 m. “Cultivated.” May 7, 1972. R. E. Schultes 26359. Comisaria del Putumayo, Rio Mocoa, Alto Afan, near Mocoa. Alt. 700 m. May 8, 1972. Schultes 26360. The Bristo/ collection constitutes a few leaves purchased from a medicine man of the Kamsa Indian tribe in the highland Valley of Sibundoy (8500 feet altitude). lex Guayusa, of course, does not grow at this altitude, but the leaves are acquired by trade or purchase from lower, warmer regions. The two Schultes collec- tions demonstrate for the first time that the species does indeed form part of the Colombian flora. These collections are pre- served in the Economic Herbarium of Oakes Ames, Harvard Botanical Museum, and in the Herbario Nacional de Colombia in Bogota. III. It is suspected that guayusa was formerly employed over a far greater area than today (Schultes, 1972). We know that the Jesuits exploited /lex Guayusa as a medi- 144 cine over three centuries ago in the Rio Maranon of Peru. Padre Juan Lorenzo Lucero wrote in 1682 that the Jivaro Indians“... put together these evil herbs [Banisteriopsis, Brugmansia and other narcotics] with guafiusa and tobacco, also invented by the devil, and allow them to boil until the small remaining quantity of juice becomes the quintescence of evil, and the faith of those who drink it is rewarded by the devil with the fruit of maledic- tion, and always to the great misfortune of many . . .” (Jiménez, 1889). Later, in 1738, an Italian missionary wrote that the priests employed it as a stomach tonic. At the same time, in 1739, Padre Andrés de Zarate reported that one product of the Jesuit mis- sions was “guayusa.” The Jesuits exported guayusa from their missions and sold the leaves in Quito (five leaves for half a real) as a medicine (Figueroa, 1904). When the Jesuits were expelled from Ecuador in 1766, the business which they had established with guayusa as a cure of venereal disease fell apart. This did not affect, however, the use of Ilex Guayusa amongst the Jivaro and Kanelo of the Rios Napo and Pastaza; these natives continued to cultivate it. It was at this time, in 1857, that Spruce encountered guayusa extensively under cultivation amongst these same natives at Antombos, near Bafios, in Ecuador. Spruce’s report (Spruce, 1908) is a most detailed record of guayusa and deserves, there- fore, to be quoted in full. | am unable to explain why Spruce failed to make herbarium specimens of //ex Guayusa, unless his reason for neglecting this task was absence of flowers or fruits on the trees which he found. He wrote: “Instead of Cupana or Guarana [Paullinia Cupana HBK.], the Zaparos and Jibaros, who inhabit the eastern side of the Equatorial Andes, have Guayusa, a plant of very similar properties, but used by them in a totally different way. The Guayusa is a true Holly [//ex], allied to the maté or Paraguay tea (//ex paraguayensis), but with much larger leaves. | was unable to find it in flower or fruit, and cannot say if it be a described species. The tree is planted near villages, and small clumps of it in the forest on the ascent of the Cordillera indicate deserted Indian sites. The highest point at which I have seen it is at about 5000 feet above the sea, in the gorge of the Pastasa below Bafios, on an ancient site called Antombos, a little above a modern cane-farm of the same name. 145 There, in 1857, was a group of Guayusa trees, supposed to date from before the Conquest, that is, to be considerably over 300 years old. They were not unlike old Holly trees in England, except that the shining leaves were much larger, thinner and unarmed. “When I travelled overland through the forest of Canelos, and my coffee gave out, I made tea of guayusa leaves, and found it very palatable. The Jibaros make the infusion so strong that it becomes positively emetic. The guayusa-pot, carefully covered up, is kept simmering on the fire all night, and when the Indian wakes up in the morning he drinks enough guayusa to make him vomit, his notion being that if any food remain undigested on the stomach, that organ should be aided to free itself of the encumbrance. Mothers give a strong draught of it, and a feather to tickle the throat with, to male children of very tender age. | rather think its use is tabooed to females of all ages, like caapi on the Uaupes.” An hitherto apparently unpublished Spruce note on guayusa, preserved at Kew, in a letter to the “Agent of Ecuador Land Company (Mr. G. P. Pritchett) in a reply to enquiries about the feasibility of forming a colony of Europeans in Forests of Cane- los (written at Bafios, Dec. 1857)” gives additional information on guayusa and would seem to support the suspicion that the centre of distribution of the plant was the eastern slopes of the Ecuadorian Andes (Schultes, 1978). “I am not sure that the Guayusa, which the wild Inds plant near their houses might not successfully compete in the English market against the inferior sorts of Tea. This is the leaf of a sort of holly, perfectly difft from the Mate or Paraguay Tea, tho somewhat allied to it, and it has much the same aromatic flavour without the bitterness of Chi- nese tea. I have used it for weeks of thogt instead of tea, & I believe you have drunk [?] some.” Although the centre of use of /lex Guayusa in the 18th Cen- tury appears to have been the eastern Andean slopes of Ecuador and northern Peru, the plant was recorded from Colombia, somewhat to the north of this area. A missionary, Padre Juan Serra (Serra, 1956), who worked in the Putumayo-Caqueta region of Colombia from 1756 to 1767, wrote that guayusa was used by the head Franciscan priest, 146 Padre José Berrutieta, at Santa Rosa. In view of the extraordi- nary detail of Padre Serra’s report, it may be worth transcribing it here in full. “The day after the arrival of the President [head priest], I saw strings being hung out in the patio and hanging from them bundles of leaves. I went into the kitchen and asked their pur- pose. A woman answered: ‘Father, this is guayusa. The Presi- dent drinks it twice a day, and we have hung it in the sun to dry.’ I told her that I would like to try it, and she said that she would give me some in the afternoon. Later . . . I tasted it; but as it was already sweetened, I did not drink more, but told her: ‘I do not like it sweet, but unsweetened, in order to discover its true taste.’ Later, they brought me more, and I drank a whole cupful. It has the color of dark honey, and five leaves are enough to make a chocolate pot full of its juice. Its taste is like tea but finer and more pleasant. When I drank it, I began to sweat and expecto- rate so much that I was obliged to change my habit, and within half an hour coughed enough phlegm to fill a large cup. These effects seemed to me to be very good. I went to the President and asked him about guayusa. He said that the beverage was excel- lent for the treatment of venereal diseases, that it .. . cleansed the blood and improved the digestion and appetite, because, when taken in the morning, one does not feel hunger until the afternoon. It strengthened the body and removed all impurities through perspiration and phlegm. All these effects are true, and I have experienced them many times. Father Berrutieta told me also that guayusa taken with honey caused women to become fertile, and, if the honey was that of the bee called apate, the woman, if married, would become pregnant immediately. This fact is well known and proven in Quito and the highlands. The Jesuits brought the plant from their mission and sold it in Quito at five leaves for a half real. I asked him where it might be found, and he told me that in the village of La Concepcion, Fr. Jose Garvo had a big tree, but in Pueblo Viejo, the first town one reaches from here, about four days distant, there is a grove of more than one league in area, entirely of guayusa trees. I at once wrote the name of the village and the name of the tree, in order not to forget them, in order that I might provide myself with supplies for my journey and destination.” 147 When Padre Serra arrived at Pueblo Viejo in December, 1756, he “.. . asked the... Mayor about Guayusa. He said that there was a great deal, and that if I so wished he would have some brought, because it grew in the forest somewhat outside the town. I told him that I wanted to go there myself and see the guayusa trees. He said that I could not go, that the mountain growth of brush was dense, but I insisted, and he assigned to me three Indians, each with a machete... We took with us two saparos or baskets ... We arrived at the guayusa grove, which is on a plain. The guayusa tree is the most beautiful and luxuriant tree that I have ever seen. It grows to be rather large in girth, so much so that three men could not encircle it, and tall in propor- tion, with a heavy crown. The trunk is ash-color, like the trunk of the poplar, the leaf a gentle and delightful green. So much so that, seeing it, I considered the hardship of the journey well worth while. From the first tree I came to, I took some leaves and began to eat them to find out their taste. I found that it was very agreeable, somewhat similar to tea, but finer and more pleasant. Seeing that there were many seedlings in the field, while the Indians gathered leaves. ..1... cut six internodes of bamboo, and, with the machete, took out eighteen seedlings with roots, placing three in each internode with earth from the same place. I took them with me and, in each village of the Putumayo, I planted three guayusa trees, and they all grew, so that, at the end of three years, they were giving many leaves. In this way, all the priests were provided with guayusa for their own consumption.” When Padre Serra finished his stay in Colombia and went to Peru, he took half a hundredweight of guayusa leaves with him, as well as a supply to display in Bogota and Popayan. That guayusa was well known in the Colombian Putumayo in those years is attested also by the reports of several Franciscan missionaries who had a mission on the Rio Putumayo slightly downstream from its confluence with the Rio Sucumbios (Cuervo, 1894; Zawadzky, 1947). “Among the medicinal plants cultivated by our missionaries .. . for the relief of the poor Indians and themselves, the guayusa tree is outstanding. A de- scription of this tree is being sent, at his request, to Don Pedro de Valencia, treasurer of the Royal Mint at Popayan. Its leaves, 148 which are the most valued part of the plant, are eagerly sought in various parts of New Granada by those acquainted with its bene- ficial properties as a purgative and an aid to digestion.” About a century later, Padre Manuel Maria Albis (Albis, 1936) wrote about his trips to the Macaguaje Indians along the Rios Mecaya, Senseya and Caucaya, in the same Putumayo- Caqueta area of Colombia. Of guayusa, he reported that “it is hot and used in poisonings; the burned leaves, when mixed with barley and honey, are given to women suffering amenorrhoea; when boiled and mixed with yoco, a caffein-containing liana [Paullinia Yoco R. E. Schult. et Killip] the preparation is used to cure dysentery; the liquid is used for stomach aches.” Patino (Patino, 1968) insinuates that guayusa “grows both wild and cultivated.” Pinkley (pers. comm.) also believes that the species may grow in a truly wild state, although he has never encountered it outside of cultivation. Except for the vague state- ment by Padre Serra that “it grew in the forest,” I find no evidence in the literature to suggest its occurrence in an undoubt- edly wild state. All references indicate that guayusa, when not planted, grows as an escape or vestige of former plantings around abandoned human habitation sites. Patino further inti- mates that, since guayusa, according to early reports, grew so prolifically in the Colombian Putumayo-Caqueta region; that since Padre Serra’s experiments in guayusa propagation were so easily successful; and that since, in Pueblo Viejo, there was a grove “more than a league in size” — guayusa might be still found in the area. Although I once doubted that any vestiges of these ancient plantings still exist in the Putumayo, Patifio’s sus- picion has proven to be correct. IV. In May, 1972, together with Dr. Andrew T. Weil of the Botan- ical Museum of Harvard University and Mr. Enrique Hernan- dez of the Universidad de Narino, I was able to make two collections in southern Colombia. Both are sterile, but both are extremely interesting from the point of view of history and of economic botany. 149 The first (Schultes 26359) was from a small bushy shrub culti- vated in the dooryard of a curandera who lived in Pepino, a small town near Mocoa. She regularly pruned the bush, selling the leaves to Indian medicine men in the nearby highland village of Sibundoy and to the public herb market in the city of Pasto. They are valued medicinally as a “tonic.” The source of the collection Bristol 352-A2 was undoubtedly this bush, since our questioning always elicited the same answer: that this curandera was the only source of guayusa leaves for export in the whole Mocoa area. The curandera informed us that the original material for her bush came from “large and ancient trees” in the very old town of Pueblo Viejo near Mocoa. Pueblo Viejo is now a very tiny and poor hamlet completely off the main highway and, in 1972 at least, accessible only on foot. We decided to make the trip. Upon arrival, we began to enquire about guayusa. Few people knew the plant, but the school teacher found that several of the school children were aware of the location of a number of large trees. They agreed to guide us. A walk of an hour and a half over rude and aban- doned, muddy mule trails brought us to Alto Afan. There we were rewarded by finding tall trees of great age which, from their manner of growth, indicated definitely former cultivation. The trees were sterile, but the natives maintained that they did flower and that the flowers were “tiny and greenish white.” The eldest farmers living in Alto Afan—some of them appar- ently of great age——assured us that the trees were large and old when they were children. We have not the slightest doubt that these trees are vestiges of the guayusa plantations described by Padre Serra, who encoun- tered a grove in Pueblo Viejo “of more than one league in area” some 200 years ago. BIBLIOGRAPHY Albis, M. M. Memorias de un Viajero (1854). Rev. Popayan, 26, no. 163-165 (1936) 28-32. Cuervo, A. B. Coleccién de Documentos Inéditos sobre la Geografia y la Historia de Colombia 1V. Secc. 2a. Geografia-Viajes- Misiones-Limites. Casanare y el Caqueta durante la Colonia (1894). [Quoted in Patino, 1968]. 150 Figueroa, F. Relacidn de las misiones de la Compania de Jesus en el pais de los magnas. Coleccion de Libros y Documentos Referentes a la Historia de América (1904) 406 [Quoted in Patino, 1968]. Jiménez de la Espada, M. Noticias auténticas del famoso rio Maranon y mision apostolica de la Compania de Jesus de la provincia de Quito .. .” (1889) 13, 373, 626 [Quoted in Patinno, 1968]. Loesener, L.E.T. ‘“Monographia Aquifoliacearum” in Nov. Acta C. L. C. G. Nat. Cur. 78 (1901) 310. Patifio, V. M. ‘“Guayusa, a neglected stimulant from the eastern Andean foothills” in Econ. Bot. 22 (1968) 311-316. Schultes, R. E. “/lex Guayusa from 500 A.D. to the present” in Etnolog. Stud. 32 (1972) 115-138. Schultes, R. E. “Richard Spruce and the potential for European settlement of the Amazon: an unpublished letter” in Bot. Journ. Linn. Soc. 77 (1978) 131-139. Serra, Juan de Santa Gertrudis. Maravillas del Peru 1, pt. 1, 2 (1956) 28, 29; 2, pt. 3, 4. Biblioteca Presidencia la Republica, Lima. [Quoted in Patino]. Spruce, R. Notes of a Botanist on the Amazon and Andes [Ed. A. R. Wallace] Macmillan, London 2 (1908) 453-454. Zawadzky C., A. Viajes misioneros del R. P. Fr. Fernando de Jesus Lrrea, franascano, 1770-1773 (1947). [Quoted in Patino]. 151 PLATE 40 Plate 40. One of the large, old trees of /lex Guayusa in Alto Afan, Pueblo Viejo, Putumayo, Colombia. Photograph: R. E. Schultes. 152 PLATE 41 Plate 41. The present inhabitants of Pueblo Viejo know that the leaves of //ex Guayusa have presumed medicinal value, but they apparently do not use the plant. One of the trees in the ancient abandoned guayusa plantation in Alto Afan, Pueblo Viejo, Putumayo, Colombia. Photograph: R. E. Schultes. BOTANICAL MUSEUM LEAFLETS VoL. 27, No. 5-6 May-JUNE 1979 THE FLOWERS OF ILEX GUAYUSA MELVIN SHEMLUCK* In 1901, Theodor Loesener described a new species of holly from sterile material collected in 1898 by Warsczewic in eastern Peru. Loesener named the holly //ex Guayusa, because the Indi- ans of eastern Colombia, Ecuador, and Peru used its leaves to prepare a medicinal and beverage tea called guayusa (or huayusa). Although the plant is well known to local people, few bota- nists have collected it, and it is poorly represented in the world’s herbaria. The great English plant-explorer, Richard Spruce, did not collect guayusa, although he was well acquainted with it and even employed it as a substitute for tea when his supply of the latter ran out (Schultes: “Richard Spruce and the potential for European settlement of the Amazon: an unpublished letter” in Bot. Journ. Linn. Soc. 77 (1978) 131-139). It 1s so poorly repre- sented in herbaria that a major United States institution mounted leaves from a commercially prepared wreath as their sole representative of the species. Two reasons can be offered for this poor representation. One is that guayusa, a cultivated plant, has been poorly collected by researchers interested in more academic botanical problems con- nected with the wild flora. Another reason, the more conceiva- ble, is that /lex Guayusa could not be found in flower: botanists are hesitant, given the time and weight constraints of collecting expeditions, to collect non-flowering or non-fruiting material. Consequently, it has been those botanists interested in economi- cally important species who have collected most of the speci- mens of this plant. The absence of flowers and fruit has led some botanists to speculate that guayusa 1s an asexually reproduced cultigen which has lost its flowering ability through years of selection and vegetative propagation by man. Like many other cultivated *Technical Assistant, Botanical Museum of Harvard University. plants, it seems to grow only in association with human habita- tion, either presently or in the past (Schultes: “Discovery of an ancient guayusa plantation in Colombia” in Bot. Mus. Leafl. Harvard Univ. 27 (1979) 143). The view that this plant has been in association with man fora long period of time is also reinforced by speculation that //ex Guayusa may possibly represent only a cultivated variant of /. paraguariensis. It was, in fact, Loesener himself who, in describ- ing guayusa, first pointed out the possible relationships between L. Guayusa and I. paraguariensis: “Species ob flores adhuc igno- tos valde dubia. Sed si re vera de Ilice agitur, species /. paragua- riensis St. Hil. et 1. nitidae (Vahl) Maxim. sine dubio affinis. Folia maiora et acutius acuminata quam in affinibus.” /. para- guariensis, known as yerba maté or Paraguay tea, grows wild in the mountains of southern Brazil, Paraguay and Argentina; it is also cultivated in these countries. My interest in //ex Guayusa was sparked by a conversation with Prof. Richard Evans Schultes of the Botanical Museum of Harvard University in 1979. I was preparing an ethnobotanical expedition to Ecuador, and he suggested that a search for guayusa flowers be included in my plans. Prof. Schultes told me of the plant’s meager taxonomic representation, its interesting use as a caffeine-containing plant used as a snuff in ancient Bolivia (Schultes: “//lex Guayusa from 500 AD to the present” in Etnolog. Stud. 32 (1972) 115-138) and its importance as a medicinal and ritual tea amongst contemporary Indians in Ecua- dor (Villavicencio, M.: Geografid de la Republica del Ecuador (1858) 371-374; Patino, M.: Guayusa—a neglected stimulant from the eastern Andean foothills” in Econ. Bot. 22 (1968) 311-316). Soon after arriving in Quito in August 1979, I discovered that the question of guayusa’s ability to flower had been answered by a collection in the herbarium at the Universidad Central of that city. On October 4, 1975, an expedition led by Ing. Agr. Alberto Ortega, Professor of Botany at this university, discovered flow- ers on a tree growing near a market in Sacua, Ecuador. Thanks to Ing. Ortega, I have examined this material. It consists entirely of staminate flowers. This collection represents the first flowers 156 of Ilex Guayusa ever collected by botanists: the original material collected by Warsczewicz 77 years earlier was sterile—as are all subsequent collections. Just one week after examining the specimen at the Univer- sidad Central, I was also able to locate flowering material of guayusa. With Mr. Fred Ness of the University of New Mexico, I found a small tree in the tiny village of Rid Chico, ten kilome- ters south of Puyo, Ecuador. Rid Chico is inhabited by people known variously as the Canelos Quechua, Sacha Runa or Puyo Runo. The village stands at an altitude of approximately 1000 meters, similar to that of Ortega’s collection. Nevertheless, flow- ering was occuring two months earlier. The tree, growing next to the school master’s house, did not belong to any single local family but was used by all members of the immediate community. My guide and informant, Sr. Rafael Santi, told me that guayusa tea 1s used before and after drinking the hallucinogen called ayvahuasca (Banisteriopsis Caapi). Drinking the tea kills the bitter taste of the ayahuasca, and its use afterwards prevents hangovers. In addition, it gives a person strength to cope with the powerful hallucinogen. Guayusa is also used by local people as a coffee substitute and for stomach trouble, and it is reputedly an aphrodisiac. In Puyo, numerous grocery stores sell “leis” of folded and strung leaves for local consumption. The tree was 30 feet tall, with a diameter of about 10 inches at breast height. It had whitish bark on all stems older than three years. After waiting for three weeks, I was able to collect speci- mens with about one third of the flowers open. Sr. Santi told me that the tree produced seeds, but at the time I could verify only that abundant pollen was being released. I located two other guayusa trees growing several hundred meters outside of the same village, but they were not in flower. These trees were approximately 12 feet in height and had the appearance of large shrubs. I was told by several residents that these trees flowered many months later, in the early part of the year. I examined the flowers and found them all to be staminate. Numerous flowers and buds were cross-sectioned, and only two tiny vestigial locules could be found in the cushion-shaped 157 gynoecium. No hint of ovules or placental tissue exists. I can offer no reasonable explanation for my guide’s asserting that this tree produced seeds, especially since this man is a very astute observer and would not mistake other floral parts for seeds. It is of note, however, that one flower amongst the 60 to 80 examined did resemble the pistillate form of //ex paraguariensis. This one flower possessed a gynoecium slightly constricted into four car- pel regions, each bearing a stigma-like structure directly upon it, and it did produce pollen. With the discovery of flowers, questions about the natural history of /lex Guayusa can be explored. It is interesting that, upon cursory examination, flower morphology and leaf shape are very much like those of /. paraguariensis. In contrast, how- ever, guayusa attains heights of 75 feet, while /. paraguariensis is a shrub or small tree growing only to 24 feet. Additionally, guayusa leaves are usually much larger. Guayusa is apparently dioecious and may well be accommodated in the reproductive pattern of the genus. However, little more can be said with certainty, and fundamental questions concerning sexual repro- duction and lack of vernal synchronization of trees in the same area must be answered. In conclusion, the next logical step is to search for pistillate flowers of guayusa and to determine whether or not they pro- duce viable seeds. Furthermore, field data on all species of //lex in the New World tropics are needed in order to evaluate effec- tively the taxonomic position of this species. It is to be hoped that the historical difficulty in finding flowering material will not hinder future studies and that 77 years need not pass before the reproductive biology of this interesting economic plant is more fully understood. Ilex Guayusa Loesener emend. Shemluck Arbor magna, robusta, usque ad 75 ped. alta sed praecipue in cultura usualiter minor. Rami glabri vel subglabri, frequenter minute et longitudinaliter delicate striolati. Folia magna, glabra, supra vulgo nitidula viridiaque, subtus pallida, interstitis saepis- ime 1.5-3 cm. longis dissita; stipulis conspicuis inaequaliter subulato-deltoideis, cinereis, usque ad 1.8 mm. longis, basin Imm. latis; petiol glabri, plus minusve rugulosi, plerumque 158 1-1.2 cm. longi; lamina chartacea vel submembranacea, oblonga vel elliptico—oblonga (interdum subovato-lanceolata), margine leviter recurvata, crenato-serrata, base acuta vel subcuneato— obtusa, apice manifeste acuminata, vulgo 9-12 cm. (sed saepe usque ad 21 cm.) longa et 4-5 cm. (sed frequenter usque ad 7.5 cm.) lata, nervis lateralibus 8 ad 10, in angulis 45—60° dispositis, apicem versus curvis, marginem versus reticulatis, supra incon- spicuis et subtus prominentibus. Inflorescentiae imperfectae in axillis fasciculatae, cymosae, pedunculo plus minusve 5-9 mm. longo, sparsissime piloso. Flores staminati parvi, albo-virides gemmis 1.5-2 mm. in diametro; pedicelli minute strigosi, ple- rumque 5 mm. longi; bracteoli brunnei, granulosi, subulati, usque ad Imm. longi; calyx crateriformis, lobis 4, brunneis extus marginem versus albidis, dense granulosis, triangularibus, apice obtusulis, plerumque I-1.4 mm. longis, 1.2-2 mm. latis; petalis plerumque 4, basin conspicue connatis, toto 3.5-4 mm. longis, membranaceis viridulis sed marginem versus albis, extus pilosu- lis, intus in lineis minutissime papilosis, rotundato-obtongis, 3 mm. longis, base 2.5 mm. latis; stamina plerumque 4 (saepe 5), petalis base adnata, filamentis 1.5 mm. longis, antheris 1.5 mm. longis; ovario rudimentario pulvinato; 1-1.5 mm. in diametro, glabro. Flores pistillati ignoti. ECUADOR: Provincia de Pastaza, Rio Chico, affluent of Rio Pastaza. Village of Rio Chico and vicinity. alt. c. 1000 m. 31 August 1979. M. Shemluck 236.—Same tree 25 August 1979, 1979 M. Shemluck et F. Ness 221. 159 EXPLANATION OF PLATE 42. l 2. 3. 4 oF Flowering branch, about 1/2 natural size. Inflorescence (with incipient branch distally), about | 1/2 natural size. Single staminate flower, about 6 times natural size. Variant, uncommon form of pistil in pollen-bearing staminate flower, about 10 times natural size. Bud, about 5 times natural size. Drawn by E. W. Smith ILEX Guayusa Loesener 160 BOTANICAL MUSEUM LEAFLETS VoL. 27, No. 5-6. May-JUNE 1979 CHEMICAL TEST FOR SILICA DETERMINATIONS AS AN ARCHAEOLOGICAL FEATURE IN ETHNOBOTANY ELIZABETH A. COUGHLIN* Silica determinations of archaeological soils and ethnobotani- cal materials have consisted primarily of visual and microscopic identifications based on morphological measurements of dia- toms of sediment cores and of plant opal or phytoliths in the absence of other botanical evidence. Bacillariophyceae, or diatoms, are unicellular algae character- ized by a cell wall of silica, and classified on the basis of cell wall markings. Species are specific in their requirements and their relative abundance is useful in determining past environmental conditions. Phytoliths are opalline silica bodies found in the epidermal cells of certain plant groups. Unlike the silicification process in petrified wood, phytolith deposition produces distinct types associated with specific genera. Chemical analysis of archaeological soils and materials offers not only further substantiation for ethnobotanical identifica- tions, but also affords the possible mapping of a distinct archaeological zone consisting of silica deposition indicative of the past presence of standing or running waters, or of agricultur- al sites. A gravimetric determination of silica (as SiO.) can be meas- ured as loss on volatilization by means of hydrofluoric acid, after dehydration by use of both hydrochloric and perchloric acid. PROCEDURE 1. PREPARATION OF SAMPLES A. Samples should be stored and tested in non-borosilicate containers, such as those of polyethylene, plastics or rubber. *Ethnobotanical Laboratory, Botanical Museum of Harvard University. 161 B. Samples should be previously unprepared or unprocessed: that is, they should not have undergone any other type of analysis that involves acid digestion as sample preparation (such as in pollen analysis). C. Samples should be reduced to particulates either with a porcelian mortar and pestle, or a metallic grinder. D. At all times samples should be protected against atmos- pheric dust and any other contaminant. . PREPARATION OF REAGENTS A. All reagents and distilled water should be stored in non- borosilicate containers (see 1.,A) and protected from contam- ination (see 1.,D). B. Distilled water should be known to be low in silica. C. HYDROCHLORIC ACID (1:1): prepare HCL (1:1) by adding equal volumes of HCL and distilled water. D. HYDROCHLORIC ACID (1:50): prepare HCL (1:50) by diluting one volume of HCL with fifty volume of distilled water. E. SULFURIC ACID (1:1): prepare H,SO, (1:1) by adding equal volumes of H»SO, and distilled water. F. HYDROFLUORIC ACID (1:1): prepare HF (1:1) by adding equal volumes of HF and distilled water. G. PERCHLORIC ACID (7:3): prepare HCLO, (7:3) by diluting seven volumes of HCLO, with three volumes of dis- tilled water. (Perchloric acid is explosive in the presence of Organic matter; exercise due caution). . PREPARATION OF BLANKS A. Prepare a reagent blank of five milliliters of distilled water. B. Run test procedure as in 4, including additions of acids, evaporations and ignitions, on this reagent blank. 4. Test! A. DEHYDRATION 'This test eliminates interferences due to tannin, color and turbidity. 162 1. Place 5 gms. of pulverized materials to be tested in a platinum or acid-leached glazed porcelain evaporating dish. 2. To this sample, add 5 ml. of HCL (1:1) to form a slurry. Stir or mix to expose as much surface area of the sample as possible to the HCL. 3. Evaporate over a water bath or other suitable heating apparatus. 4. During this evaporation, add 5 ml. of HCL (1:1) at three appropriately spaced intervals. 5. Add 5 ml. HCLO, (7:3), and proceed to total evaporation. 6. After evaporation to dryness, place the dish in an oven (110°C) or other suitable apparatus to bake for forty minutes. 7. Remove evaporating dish and allow to cool until moder- ately warm. 8. Add to residue in dish, 5 ml. HCL (1:1), and 50 ml. distilled water to form a slurry. Stir with porcelain spatula or other suitable apparatus (no glass). 9. Pass this heated mixture through an ashless, quantitative grade filter paper by means of vacuum filtration. 10. Wash evaporating dish with small amounts of HCL (1:50), and pass this through the same filter. 11. Pass 50 ml. of hot distilled water in 10 ml. increments also through the residue on the filter. 12. Set aside this filter with residue. 13. Place the filtrate (see 9) and washings (see 10, 11) in the original evaporating dish, in portions if necessary, and evapo- rate to dryness as in 6. 14. Repeat steps 7 through 12. 15. Place the two filters with residue (see 12), in a tared covered platinum crucible and place the crucible in a drying oven (110°C) for forty minutes. 16. After drying, remove the covered crucible from the drying oven; place in a muffle furnace, and ignite to 1000°C for fifteen minutes. 17. Remove the crucible from the muffle furnace and cool in a dessicator; weigh and repeat step 16, until cooled crucible demonstrates a constant weight. 18. Record this constant weight. 163 B. VOLATILIZATION 1. Add small amount of distilled water to moisten residue in crucible. 2. Add 0.2 ml. of H,SO, (1:1). 3. Add 10 ml. HF (1:1). 4. Slowly evaporate to dryness. Do not use oven; avoid loss by spattering. 5. After evaporating to dryness, place in a muffle furnace, and ignite to 1000°C for fifteen minutes. 6. Remove the crucible from the muffle furnace and cool ina dessicator; weigh and repeat step 5 until cooled crucible dem- onstrates a constant weight. 7. Record this constant weight. 5. CALCULATIONS A. Subtract values of reagent blanks (see 3., A., B.) from weights after evaporation (see 4., A., 18), and after volatiliza- tion (see 4., B., 7). B. Subtract the corrected weight of the crucible after volatili- zation (4., B., 7) from the corresponding corrected weight after evaporation and before volatilization (4., A., 18). C. The difference X, in milligrams, is the loss on volatiliza- tion and represents silica. D. Mg. Si0,= X = loss on volatilization. 6. EVALUATION OF RESULTS Compared to lower levels in surrounding soil or material, significantly high levels of silica in association with the pres- ence of diatoms, indicates past environmental conditions of either standing or running water; and in the presence of cer- tain phytoliths indicates past grasslands rather than wood- lands, and more specifically, in accordance with the phytolith classification scheme suggested by D. M. Pearsall, supports the identification of the presence of maize in archaeological contents. 164 This procedure was developed in the Ethnobotanical Labora- tory of the Botanical Museum of Harvard University with the support and encouragement of Professor Richard Evans Schultes, Director of the Museum. I thank Professor Schultes and the Museum staff for their contribution to this project, and in particular, Professor Elso S. Barghoorn, Lillian Hanscom, Publications Secretary, and Wesley Wong, Librarian. SELECTED BIBLIOGRAPHY 1. Dimbleby, G. W., Plants and Archaeology, Humanities Press, Inc., New York, 1967. 2. Helbaek, H., ‘Palaeo-Ethnobotany’: in Science in Archaeology, (Eds. D. Brothwell and E. Higgs), Thames and Hudson, Bristol, 1963. 3. Leo, R. F. and Barghoorn, E. S., “Silicification of Wood”, in Bot. Mus. Leafl., Harvard Univ., Vol. 25, #1, December 7, 1976. Pearsall, D.M., “Phytolith Analyses of Archaeological Soils; Evidence for Maize Cultivation in Formative Ecuador”: in Science, Vol. 199, January 13, 1978. + 165 BOTANICAL MUSEUM LEAFLETS HARVARD UNIVERSITY CAMBRIDGE, MASSACHUSETTS, SEPT.-NOV. 1979 _ VOL. 27. No. 7-9 SYSTEMATICS OF THE GENUS STELIS SW. LESLIE A. GARAY In preparing the manuscript for the second volume of the “Orchids of Ecuador” which includes the genera of the Pleuro- thallidinae, | had the opportunity to look at the genus Sre/is critically again after some 23 years. Namely, in 1957, as a privi- leged recipient of a Guggenheim Fellowship for Canadians which allowed me to spend a year in the Orchid Herbarium of Oakes Ames at Harvard University, I had prepared a critical study of the genus Sre/is. Although the interest in this unpub- lished, book-length monograph has been fairly constant throughout the years by both professionals and amateurs alike, the chances for its publication grew slimmer for financial rea- sons as the years went by. The present synopsis 1s based essen- tially on tha® manuscript, but the position of each species has been re-evaluated. Consequently the internal structure of the genus is quite different from my classification published in 1956 (Canadian Journ. Bot. 34: 346-359). In that paper I also have covered in sufficient details the historical background of the genus including the infrageneric classifications. These aspects, therefore, are not repeated here. In this synopsis the systematics of the genus are completely revised and brought in line with the articles of the International Code of Botanical Nomenclature. The key to the sections has been expanded to include the 22 alliances into which the genus is now divided. In revising the genus it became apparent that about 32 species share in common a special character which cannot be accomo- Botanical Museum Leaflets (USPS 404-990). Published monthly except during July and August by the Botanical Museum, Harvard University, Cambridge, Massachusetts 02138. Subscription: $25.00 a year, net, postpaid. Orders should be directed to Secretary of Publications at the above address. Second-Class Postage Paid at Boston, Massachusetts. dated within the circumscription of the genus Ste/is. Here I refer to those species in which the flowers have a single stigma as opposed to the two stigmata in the genus Sre/is. For these spe- cies the generic name Apatostelis is proposed here. In recent years a lot of attention has been focused on the whole group of Pleurothallidinae, resulting in the publication of a number of new genera. As an introduction to the Pleurothalli- dinae a key to all genera within this subtribe is given here. Attention must be called to the fact that the following generic names are excluded for various reasons: Luere//la Braas differs from Masdevallia solely in the character of the petals which are calliferous externally in the former; Pabstiella Brieger differs from Pleurothallis only in the prominent columnfoot; Garayella Brieger is based on the same type as Chamelophyton Garay; Dubois- Reymondia Karst., recently resurrected by Brieger, had been reduced already in 1859 to a synonym of Pleurothallis by Lindley due to lack of differentiating characters; Restrepiopsis Luer is a recent segregate of Restrepiella Garay, yet both genera have the same circumscription; Octandrorchis Brieger is only an Octomeria with connate lateral sepals; Sreliopsis Brieger is not validly published—the obscure description in German and the photograph of the general habit of the plant leave little doubt that it is referable to Srelis Allenii L.O.Wms. I would like to call special attention to the genus Specklinia Lindl. which I recognize only as a subgenus of Pleurothallis. It certainly deserves a recognition at the generic level as it is evi- dent from the “Key to Genera” provided on subsequent pages. Such a step, however, will require some 800 nomenclatorial transfers. STUDY OF FLORAL DETAILS. The flowers of most Ste/is species are very small to minute and commonly are of a rather fleshy texture. The conventional method of preparing herbarium specimens through drying and pressing, frequently damages the intricate floral structures. Therefore, care must be taken in preparation for examination. There are two simple methods available: 1, boiling the dried flowers for about one minute, after which the flowers should be dissected submerged in water: 2, placing the dry flowers in con- 168 centrated ammonia for about 30 minutes, then washing in cold running water for 5-10 minutes, after which they should be dissected submerged in water. If drawings are to be prepared, they should be made from the submerged flowers and floral parts. If the flowers are exposed to air on open slides, they will desiccate very rapidly and become distorted. Most of the Ste/is drawings published by Schlechter show such distortions. The plates attached to this paper were prepared from flowers sub- merged in water. It should be mentioned that the advantage of using the ammo- nia method is that after the flowers have been washed thor- oughly in running water, they can be transferred to FAA or alcohol and can be stored indefinitely. Such material will not differ substantially from actual live material preserved in liquid in the field. TYPIFICATION OF THE GENUS. When Swartz established the genus Sre/is, he explicitly stated that its characters are those given for Humboldtia Ruiz & Pavon. At the same time he also noted that the generic name of Humboldtia was previously used by Vahl. Therefore, Ste/is is merely a new name for Humboldtia Ruiz & Pavon, and the lectotype for it must be chosen from among the eleven specific epithets provided by Ruiz and Pavon in their Systema Vegetabi- lium. There are five Humboldtia species which are definitely referable to Ste/is, and of these only Humboldtia purpurea exhibits the characters depicted by Ruiz and Pavon on plate 27 for Humboldtia; these drawings are part of the original proto- logue. The holotype of Humboldtia purpurea is in the herba- rium of Jardin Botanico, Madrid; duplicates are in the Reichen- bach Herbarium, Vienna, and in the Willdenow Herbarium, Berlin. The generic name Ste/is was conserved in 1905, although, as Stafleu has shown in Taxon 8:258,1959, the conservation was superfluous at that time. In 1929, Epidendrum ophioglossoides Jacq. was selected as the lectotype by Green in Prop. Brit. Bot. p. 100, and this typification was conserved in 1930. Unfortu- nately, Epidendrum ophioglossoides is based on Plumier’s poly- nomial of 1703, as well as on Plumier’s drawing of it published 169 subsequently by Burmann in 1759. A closer examination of Plumier’s drawing reveals at once that it is referable to the genus Pleurothallis R.Br. as it is understood today, and is conspecific with Pleurothallis floribunda (Lind1.)Lindl. Therefore, a retypi- fiction of Sve/is is necessary to prevent a mass transfer of some 1,000 specific epithets from Pleurothallis which is inevitable if the old typification is maintained. In using this reasoning, Garay and Sweet in the Journal of the Arnold Arboretum 53: 528, 1972 have selected Humboldtia pur- purea Ruiz & Pavon as a new lectotype. ACKNOWLEDGMENTS My special thanks and appreciation are recorded here to the John Simon Guggenheim Memorial Foundation for their Ca- nadian Fellowship awarded to me in 1957 for the purpose of a monographic revision of the genus Sre/is. This same foundation has also provided extra financial help toward the preparation of the illustrations published for the first time in this paper. The financial help in the form of a grant from the American Orchid Society’s Fund for Education and Research and the con- tribution of Mr. Fritz Hamer of El Salvador, both of which helped to defray in part the publication costs of this paper, are here gratefully acknowledged. KEY TO THE GENERA OF THE SUBTRIBE PLEUROTHALLIDINAE bs) POUND 2) oe pet ce deve deus dee pear nena keees Z fa; Poninia A OF 6-09 S: s